Principles and Practice of Endocrinology and Metabolism (December 2002): by Kenneth L. Becker (Editor), C.

Ronald Kahn (Editor), Robert W. Rebar (Editor) By Lippincott Williams & Wilkins Publishers

By OkDoKeY

Principles and Practice of Endocrinology and Metabolism

CONTENTS
Editors Contributing Authors Preface Preface to the First Edition Part I General Principles of Endocrinology Kenneth L. Becker, Editor Part II The Endocrine Brain and Pituitary Gland Gary L. Robertson, Editor Part III The Thyroid Gland Leonard Wartofsky, Editor Part IV Calcium and Bone Metabolism John P. Bilezikian, Editor Part V The Adrenal Glands D. Lynn Loriaux, Editor Part VI Sex Determination and Development Robert W. Rebar and William J. Bremner, Editors Part VII Endocrinology of the Female Robert W. Rebar, Editor Part VIII Endocrinology of the Male William J. Bremner, Editor Part IX Disorders of Fuel Metabolism C. Ronald Kahn, Editor Part X Diffuse Hormonal Secretion Eric S. Nyln, Editor Part XI Heritable Abnormalities of Endocrinology and Metabolism Kenneth L. Becker, Editor Part XII Immunologic Basis of Endocrine Disorders Leonard Wartofsky, Editor Part XIII Endocrine and Metabolic Dysfunction in the Growing Child and in the Aged Wellington Hung, Editor Part XIV Interrelationships Between Hormones and the Body Kenneth L. Becker, Editor Part XV Hormones and Cancer Kenneth L. Becker, Editor Part XVI Endocrinology of Critical Illness Eric S. Nyln, Editor Part XVII Endocrine and Metabolic Effects of Toxic Agents Kenneth L. Becker, Editor Part XVIII Endocrine Drugs and Values Kenneth L. Becker, Editor

PART I GENERAL PRINCIPLES OF ENDOCRINOLOGY

Kenneth L. Becker, Editor
Chapter 1 Endocrinology and the Endocrine Patient KENNETH L. BECKER, ERIC S. NYLN, and RICHARD H. SNIDER, JR. Chapter 2 Molecular Biology: Present and Future MEHBOOB A. HUSSAIN and JOEL F. HABENER Chapter 3 Biosynthesis and Secretion of Peptide Hormones WILLIAM W. CHIN Chapter 4 Hormonal Action DARYL K. GRANNER Chapter 5 Feedback Control in Endocrine Systems DANIEL N. DARLINGTON and MARY F. DALLMAN Chapter 6 Endocrine Rhythms EVE VAN CAUTER Chapter 7 Growth and Development in the Normal Infant and Child GILBERT P. AUGUST

PART II THE ENDOCRINE BRAIN AND PITUITARY GLAND

Gary L. Robertson, Editor
Chapter 8 Morphology of the Endocrine Brain, Hypothalamus, and Neurohypophysis JOHN R. SLADEK, JR., and CELIA D. SLADEK Chapter 9 Physiology and Pathophysiology of the Endocrine Brain and Hypothalamus PAUL E. COOPER Chapter 10 Pineal Gland RUSSEL J. REITER Chapter 11 Morphology of the Pituitary In Health and Disease KAMAL THAPAR, KALMAN KOVACS, and EVA HORVATH

SECTION A ADENOHYPOPHYSIS
Chapter 12 Growth Hormone and Its Disorders GERHARD BAUMANN Chapter 13 Prolactin and Its Disorders LAURENCE KATZNELSON and ANNE KLIBANSKI Chapter 14 Adrenocorticotropin: Physiology and Clinical Aspects DAVID J. TORPY and RICHARD V. JACKSON Chapter 15 Thyroid-Stimulating Hormone and Its Disorders JOSHUA L. COHEN Chapter 16 Pituitary Gonadotropins and Their Disorders WILLIAM J. BREMNER, ILPO HUHTANIEMI, and JOHN K. AMORY Chapter 17 Hypopituitarism JOSEPH J. PINZONE Chapter 18 Hypothalamic and Pituitary Disorders in Infancy and Childhood ALAN D. ROGOL Chapter 19 The Optic Chiasm in Endocrinologic Disorders R. MICHAEL SIATKOWSKI and JOEL S. GLASER Chapter 20 Diagnostic Imaging of the Sellar Region ERIC BOUREKAS, MARY OEHLER, and DONALD CHAKERES Chapter 21 Medical Treatment of Pituitary Tumors and Hypersecretory States DAVID H. SARNE Chapter 22 Radiotherapy of Pituitary-Hypothalamic Tumors MINESH P. MEHTA Chapter 23 Neurosurgical Management of Pituitary-Hypothalamic Neoplasms DAVID S. BASKIN Chapter 24 Pituitary Tumors: Overview of Therapeutic Options PHILIPPE CHANSON

SECTION B NEUROHYPOPHYSIAL SYSTEM

Chapter 25 Physiology of Vasopressin, Oxytocin, and Thirst GARY L. ROBERTSON Chapter 26 Diabetes Insipidus and Hyperosmolar Syndromes PETER H. BAYLIS and CHRISTOPHER J. THOMPSON Chapter 27 Inappropriate Antidiuresis and Other Hypoosmolar States JOSEPH G. VERBALIS

PART III THE THYROID GLAND

Leonard Wartofsky, Editor
Chapter 28 Approach to the Patient with Thyroid Disease LEONARD WARTOFSKY Chapter 29 Morphology of the Thyroid Gland VIRGINIA A. LIVOLSI Chapter 30 Thyroid Physiology: Synthesis and Release, Iodine Metabolism, Binding and Transport H. LESTER REED Chapter 31 Thyroid Physiology: Hormone Action, Receptors, and Postreceptor Events PAUL M. YEN Chapter 32 Thyroid Hormone Resistance Syndromes STEPHEN JON USALA Chapter 33 Thyroid Function Tests ROBERT C. SMALLRIDGE Chapter 34 Thyroid Uptake and Imaging SALIL D. SARKAR and DAVID V. BECKER Chapter 35 Thyroid Sonography, Computed Tomography, and Magnetic Resonance Imaging MANFRED BLUM Chapter 36 Abnormal Thyroid Function Test Results in Euthyroid Persons HENRY B. BURCH Chapter 37 Adverse Effects of Iodide JENNIFER A. NUOVO and LEONARD WARTOFSKY Chapter 38 Nontoxic Goiter PAUL J. DAVIS and FAITH B. DAVIS Chapter 39 The Thyroid Nodule LEONARD WARTOFSKY and ANDREW J. AHMANN Chapter 40 Thyroid Cancer ERNEST L. MAZZAFERRI Chapter 41 Unusual Thyroid Cancers MATTHEW D. RINGEL Chapter 42 Hyperthyroidism KENNETH D. BURMAN Chapter 43 Endocrine Ophthalmopathy MELVIN G. ALPER and LEONARD WARTOFSKY Chapter 44 Surgery of the Thyroid Gland EDWIN L. KAPLAN Chapter 45 Hypothyroidism LAWRENCE E. SHAPIRO and MARTIN I. SURKS Chapter 46 Thyroiditis IVOR M. D. JACKSON and JAMES V. HENNESSEY Chapter 47 Thyroid Disorders of Infancy and Childhood WELLINGTON HUNG

PART IV CALCIUM AND BONE METABOLISM

John P. Bilezikian, Editor
Chapter 48 Morphology of the Parathyroid Glands VIRGINIA A. LIVOLSI Chapter 49 Physiology of Calcium Metabolism EDWARD M. BROWN Chapter 50 Physiology of Bone LAWRENCE G. RAISZ Chapter 51 Parathyroid Hormone DAVID GOLTZMAN and GEOFFREY N. HENDY Chapter 52 Parathyroid HormoneRelated Protein GORDON J. STREWLER Chapter 53 Calcitonin Gene Family of Peptides KENNETH L. BECKER, BEAT MLLER, ERIC S. NYLN, RGIS COHEN, OMEGA L. SILVA, JON C. WHITE, and RICHARD H. SNIDER, JR. Chapter 54 Vitamin D THOMAS L. CLEMENS and JEFFREY L. H. ORIORDAN Chapter 55 Bone Quantification and Dynamics of Turnover DAVID W. DEMPSTER and ELIZABETH SHANE Chapter 56 Markers of Bone Metabolism MARKUS J. SEIBEL, SIMON P. ROBINS, and JOHN P. BILEZIKIAN Chapter 57 Clinical Application of Bone Mineral Density Measurements PAUL D. MILLER, ABBY ERICKSON, and CAROL ZAPALOWSKI Chapter 58 Primary Hyperparathyroidism SHONNI J. SILVERBERG and JOHN P. BILEZIKIAN Chapter 59 Nonparathyroid Hypercalcemia ANDREW F. STEWART Chapter 60 Hypoparathyroidism and Other Causes of Hypocalcemia SUZANNE M. JAN DE BEUR, ELIZABETH A. STREETEN, and MICHAEL A. LEVINE Chapter 61 Renal Osteodystrophy KEVIN J. MARTIN, ESTHER A. GONZALEZ, and EDUARDO SLATOPOLSKY Chapter 62 Surgery of the Parathyroid Glands GERARD M. DOHERTY and SAMUEL A. WELLS, JR. Chapter 63 Osteomalacia and Rickets NORMAN H. BELL Chapter 64 Osteoporosis ROBERT LINDSAY and FELICIA COSMAN Chapter 65 Paget Disease of Bone ETHEL S. SIRIS Chapter 66 Rare Disorders of Skeletal Formation and Homeostasis MICHAEL P. WHYTE Chapter 67 Diseases of Abnormal Phosphate Metabolism MARC K. DREZNER Chapter 68 Magnesium Metabolism ROBERT K. RUDE Chapter 69 Nephrolithiasis MURRAY J. FAVUS and FREDRIC L. COE Chapter 70 Disorders of Calcium and Bone Metabolism in Infancy and Childhood THOMAS O. CARPENTER

PART V THE ADRENAL GLANDS

D. Lynn Loriaux, Editor
Chapter 71 Morphology of the Adrenal Cortex and Medulla DONNA M. ARAB OBRIEN Chapter 72 Synthesis and Metabolism of Corticosteroids PERRIN C. WHITE Chapter 73 Corticosteroid Action PERRIN C. WHITE Chapter 74 Tests of Adrenocortical Function D. LYNN LORIAUX Chapter 75 Cushing Syndrome DAVID E. SCHTEINGART Chapter 76 Adrenocortical Insufficiency D. LYNN LORIAUX Chapter 77 Congenital Adrenal Hyperplasia PHYLLIS W. SPEISER Chapter 78 Corticosteroid Therapy LLOYD AXELROD Chapter 79 Renin-Angiotensin System and Aldosterone DALILA B. CORRY and MICHAEL L. TUCK Chapter 80 Hyperaldosteronism JOHN R. GILL, JR. Chapter 81 Hypoaldosteronism JAMES C. MELBY Chapter 82 Endocrine Aspects of Hypertension DALILA B. CORRY and MICHAEL L. TUCK Chapter 83 Adrenocortical Disorders in Infancy and Childhood ROBERT L. ROSENFIELD and KE-NAN QIN Chapter 84 The Incidental Adrenal Mass D. LYNN LORIAUX Chapter 85 Physiology of the Adrenal Medulla and the Sympathetic Nervous System DAVID S. GOLDSTEIN Chapter 86 Pheochromocytoma and Other Diseases of the Sympathetic Nervous System HARRY R. KEISER Chapter 87 Adrenomedullary Disorders of Infancy and Childhood WELLINGTON HUNG Chapter 88 Diagnostic Imaging of the Adrenal Glands DONALD L. MILLER Chapter 89 Surgery of the Adrenal Glands GARY R. PEPLINSKI and JEFFREY A. NORTON

PART VI SEX DETERMINATION AND DEVELOPMENT

Robert W. Rebar and William J. Bremner, Editors
Chapter 90 Normal and Abnormal Sexual Differentiation and Development JOE LEIGH SIMPSON and ROBERT W. REBAR Chapter 91 Physiology of Puberty PETER A. LEE Chapter 92 Precocious and Delayed Puberty EMILY C. WALVOORD, STEVEN G. WAGUESPACK, and ORA HIRSCH PESCOVITZ Chapter 93 Micropenis, Hypospadias, and Cryptorchidism in Infancy and Childhood WELLINGTON HUNG

PART VII ENDOCRINOLOGY OF THE FEMALE

Robert W. Rebar, Editor
Chapter 94 Morphology and Physiology of the Ovary GREGORY F. ERICKSON and JAMES R. SCHREIBER Chapter 95 The Normal Menstrual Cycle and the Control of Ovulation ROBERT W. REBAR, GARY D. HODGEN, and MICHAEL ZINGER Chapter 96 Disorders of Menstruation, Ovulation, and Sexual Response ROBERT W. REBAR Chapter 97 Ovulation Induction MICHAEL A. THOMAS Chapter 98 Endometriosis ROBERT L. BARBIERI Chapter 99 Premenstrual Syndrome ROBERT L. REID and RUTH C. FRETTS Chapter 100 Menopause BRIAN WALSH and ISAAC SCHIFF Chapter 101 Hirsutism, Alopecia, and Acne ENRICO CARMINA and ROGERIO A. LOBO Chapter 102 Functioning Tumors and Tumor-Like Conditions of the Ovary I-TIEN YEH, CHARLES ZALOUDEK, and ROBERT J. KURMAN Chapter 103 The Differential Diagnosis of Female Infertility STEVEN J. ORY and MARCELO J. BARRIONUEVO Chapter 104 Female Contraception ALISA B. GOLDBERG and PHILIP DARNEY Chapter 105 Complications and Side Effects of Steroidal Contraception ALISA B. GOLDBERG and PHILIP DARNEY Chapter 106 Morphology of the Normal Breast, Its Hormonal Control, and Pathophysiology RICHARD E. BLACKWELL Chapter 107 Conception, Implantation, and Early Development PHILIP M. IANNACCONE, DAVID O. WALTERHOUSE, and KRISTINA C. PFENDLER Chapter 108 The Maternal-Fetal-Placental Unit BRUCE R. CARR Chapter 109 Endocrinology of Parturition JOHN R. G. CHALLIS Chapter 110 Endocrine Disease in Pregnancy MARK E. MOLITCH Chapter 111 Trophoblastic Tissue and Its Abnormalities CYNTHIA G. KAPLAN Chapter 112 Endocrinology of Trophoblastic Tissue Z. M. LEI and CH. V. RAO

PART VIII ENDOCRINOLOGY OF THE MALE

William J. Bremner, Editor
Chapter 113 Morphology and Physiology of the Testis DAVID M. DE KRETSER Chapter 114 Evaluation of Testicular Function STEPHEN J. WINTERS Chapter 115 Male Hypogonadism STEPHEN R. PLYMATE Chapter 116 Testicular Dysfunction in Systemic Disease H. W. GORDON BAKER Chapter 117 Erectile Dysfunction GLENN R. CUNNINGHAM and MAX HIRSHKOWITZ Chapter 118 Male Infertility RICHARD V. CLARK Chapter 119 Clinical Use and Abuse of Androgens and Antiandrogens ALVIN M. MATSUMOTO Chapter 120 Gynecomastia ALLAN R. GLASS Chapter 121 Endocrine Aspects of Benign Prostatic Hyperplasia ELIZABETH A. MILLER and WILLIAM J. ELLIS Chapter 122 Testicular Tumors NIELS E. SKAKKEBAEK and MIKAEL RRTH Chapter 123 Male Contraception JOHN K. AMORY and WILLIAM J. BREMNER

PART IX DISORDERS OF FUEL METABOLISM

C. Ronald Kahn, Editor SECTION A FOOD AND ENERGY
Chapter 124 Principles of Nutritional Management ROBERTA P. DURSCHLAG and ROBERT J. SMITH Chapter 125 Appetite ANGELICA LINDN HIRSCHBERG Chapter 126 Obesity JULES HIRSCH, LESTER B. SALANS, and LOUIS J. ARONNE Chapter 127 Starvation RUTH S. MACDONALD and ROBERT J. SMITH Chapter 128 Anorexia Nervosa and Other Eating Disorders MICHELLE P. WARREN and REBECCA J. LOCKE Chapter 129 Fuel Homeostasis and Intermediary Metabolism of Carbohydrate, Fat, and Protein NEIL B. RUDERMAN, KEITH TORNHEIM, and MICHAEL N. GOODMAN Chapter 130 Vitamins: Hormonal and Metabolic Interrelationships ALAA ABOU-SAIF and TIMOTHY O. LIPMAN Chapter 131 Trace Minerals: Hormonal and Metabolic Interrelationships ROBERT D. LINDEMAN Chapter 132 Exercise: Endocrine and Metabolic Effects JACQUES LEBLANC

SECTION B DIABETES MELLITUS

Chapter 133 Morphology of the Endocrine Pancreas SUSAN BONNER-WEIR Chapter 134 Islet Cell Hormones: Production and Degradation GORDON C. WEIR and PHILIPPE A. HALBAN Chapter 135 Glucose Homeostasis and Insulin Action C. RONALD KAHN Chapter 136 Classification, Diagnostic Tests, and Pathogenesis of Type 1 Diabetes Mellitus GEORGE S. EISENBARTH Chapter 137 Etiology and Pathogenesis of Type 2 Diabetes Mellitus and Related Disorders C. RONALD KAHN Chapter 138 Natural History of Diabetes Mellitus ANDRZEJ S. KROLEWSKI and JAMES H. WARRAM Chapter 139 Secondary Forms of Diabetes Mellitus VERONICA M. CATANESE and C. RONALD KAHN Chapter 140 Evaluation of Metabolic Control in Diabetes ALLISON B. GOLDFINE Chapter 141 Diet and Exercise in Diabetes OM P. GANDA Chapter 142 Oral Agents for the Treatment of Type 2 Diabetes Mellitus ALLISON B. GOLDFINE and ELEFTHERIA MARATOS-FLIER Chapter 143 Insulin Therapy and Its Complications GORDON C. WEIR Chapter 144 Pancreas and Islet Transplantation GORDON C. WEIR Chapter 145 Syndrome X GERALD M. REAVEN Chapter 146 Syndromes of Extreme Insulin Resistance JEFFREY S. FLIER and CHRISTOS S. MANTZOROS Chapter 147 Cardiovascular Complications of Diabetes Mellitus KARIN HEHENBERGER and GEORGE L. KING Chapter 148 Diabetic Neuropathy EVA L. FELDMAN, MARTIN J. STEVENS, JAMES W. RUSSELL, and DOUGLAS A. GREENE Chapter 149 Gastrointestinal Complications of Diabetes FREDERIC D. GORDON and KENNETH R. FALCHUK Chapter 150 Diabetic Nephropathy RALPH A. DEFRONZO Chapter 151 Diabetes and the Eye LAWRENCE I. RAND Chapter 152 Diabetes and Infection GEORGE M. ELIOPOULOS Chapter 153 Diabetes and the Skin ROBERT J. TANENBERG and RICHARD C. EASTMAN Chapter 154 The Diabetic Foot GARY W. GIBBONS Chapter 155 Diabetic Acidosis, Hyperosmolar Coma, and Lactic Acidosis K. GEORGE M. M. ALBERTI Chapter 156 Diabetes Mellitus and Pregnancy LOIS JOVANOVIC Chapter 157 Diabetes Mellitus in the Infant and Child DOROTHY J. BECKER and ALLAN L. DRASH

SECTION C HYPOGLYCEMIA
Chapter 158 Hypoglycemic Disorders in the Adult

RICHARD J. COMI and PHILLIP GORDEN Chapter 159 Localization of Islet Cell Tumors DONALD L. MILLER Chapter 160 Surgery of the Endocrine Pancreas JON C. WHITE Chapter 161 Hypoglycemia of Infancy and Childhood JOSEPH I. WOLFSDORF and MARK KORSON

SECTION D LIPID METABOLISM

Chapter 162 Biochemistry and Physiology of Lipid and Lipoprotein Metabolism ROBERT W. MAHLEY Chapter 163 Lipoprotein Disorders ERNST J. SCHAEFER Chapter 164 Treatment of the Hyperlipoproteinemias JOHN C. LAROSA Chapter 165 Endocrine Effects on Lipids HENRY N. GINSBERG, IRA J. GOLDBERG, and CATHERINE TUCK Chapter 166 Lipid Abnormalities in Diabetes Mellitus ROBERT E. RATNER, BARBARA V. HOWARD, and WILLIAM JAMES HOWARD

PART X DIFFUSE HORMONAL SECRETION

Eric S. Nyln, Editor
Chapter 167 General Characteristics of Diffuse Peptide Hormone Systems JENS F. REHFELD Chapter 168 Endogenous Opioid Peptides BRIAN M. COX and GREGORY P. MUELLER Chapter 169 Somatostatin YOGESH C. PATEL Chapter 170 Kinins DOMENICO C. REGOLI Chapter 171 Substance P and the Tachykinins NEIL ARONIN Chapter 172 Prostaglandins, Thromboxanes, and Leukotrienes R. PAUL ROBERTSON Chapter 173 Growth Factors and Cytokines DEREK LEROITH and VICKY A. BLAKESLEY Chapter 174 Compendium of Growth Factors and Cytokines BHARAT B. AGGARWAL Chapter 175 The Diffuse Neuroendocrine System ERIC S. NYLN and KENNETH L. BECKER Chapter 176 The Endocrine Brain ABBA J. KASTIN, WEIHONG PAN, JAMES E. ZADINA, and WILLIAM A. BANKS Chapter 177 The Endocrine Lung KENNETH L. BECKER Chapter 178 The Endocrine Heart MIRIAM T. RADEMAKER and ERIC A. ESPINER Chapter 179 The Endocrine Endothelium FRANCESCO COSENTINO and THOMAS F. LSCHER Chapter 180 The Endocrine Blood Cells HARISH P. G. DAVE and BEAT MLLER Chapter 181 The Endocrine Mast Cell STEPHEN I. WASSERMAN Chapter 182 The Endocrine Enteric System JENS J. HOLST Chapter 183 The Endocrine Kidney ALAN DUBROW and LUCA DESIMONE Chapter 184 The Endocrine Genitourinary Tract JAN FAHRENKRUG and SREN GRS Chapter 185 The Endocrine Skin MARK R. PITTELKOW Chapter 186 The Endocrine Adipocyte REXFORD S. AHIMA and JEFFREY S. FLIER

PART XI HERITABLE ABNORMALITIES OF ENDOCRINOLOGY AND METABOLISM

Kenneth L. Becker, Editor
Chapter 187 Inheritance Patterns of Endocrinologic and Metabolic Disorders R. NEIL SCHIMKE Chapter 188 Multiple Endocrine Neoplasia GLEN W. SIZEMORE Chapter 189 Heritable Disorders of Collagen and Fibrillin PETER H. BYERS Chapter 190 Heritable Diseases of Lysosomal Storage WARREN E. COHEN Chapter 191 Heritable Diseases of Amino-Acid Metabolism HARVEY J. STERN and JAMES D. FINKELSTEIN Chapter 192 Heritable Diseases of Purine Metabolism EDWARD W. HOLMES and DAVID J. NASHEL

PART XII IMMUNOLOGIC BASIS OF ENDOCRINE DISORDERS

Leonard Wartofsky, Editor
Chapter 193 The Endocrine Thymus ALLAN L. GOLDSTEIN and NICHOLAS R. S. HALL Chapter 194 Immunogenetics, the Human Leukocyte Antigen System, and Endocrine Disease JAMES R. BAKER, JR. Chapter 195 T Cells in Endocrine Disease ANTHONY PETER WEETMAN Chapter 196 B Cells and Autoantibodies in Endocrine Disease ALAN M. MCGREGOR Chapter 197 The Immune System and Its Role in Endocrine Function ROBERT VOLP

PART XIII ENDOCRINE AND METABOLIC DYSFUNCTION IN THE GROWING CHILD AND IN THE AGED
Wellington Hung, Editor
Chapter 198 Short Stature and Slow Growth in the Young THOMAS ACETO, JR.,DAVID P. DEMPSHER, LUIGI GARIBALDI, SUSAN E. MYERS, NANCI BOBROW, and COLLEEN WEBER Chapter 199 Endocrinology and Aging DAVID A. GRUENEWALD and ALVIN M. MATSUMOTO

PART XIV INTERRELATIONSHIPS BETWEEN HORMONES AND THE BODY

Kenneth L. Becker, Editor
Chapter 200 Cerebral Effects of Endocrine Disease HOYLE LEIGH Chapter 201 Psychiatric-Hormonal Interrelationships MITCHEL A. KLING,MARIANNE HATLE, RAMESH K. THAPAR, and PHILIP W. GOLD Chapter 202 Respiration and Endocrinology PRASHANT K. ROHATGI and KENNETH L. BECKER Chapter 203 The Cardiovascular System and Endocrine Disease ELLEN W. SEELY and GORDON H. WILLIAMS Chapter 204 Gastrointestinal Manifestations of Endocrine Disease ALLAN G. HALLINE Chapter 205 The Liver and Endocrine Function NICOLA DE MARIA, ALESSANDRA COLANTONI, and DAVID H. VAN THIEL Chapter 206 Effects of Nonrenal Hormones on the Normal Kidney PAUL L. KIMMEL ANTONIO RIVERA, and PARVEZ KHATRI Chapter 207 Renal Metabolism of Hormones RALPH RABKIN and MICHAEL J. HAUSMANN Chapter 208 Effects of Endocrine Disease on the Kidney ELLIE KELEPOURIS and ZALMAN S. AGUS Chapter 209 Endocrine Dysfunction due to Renal Disease ARSHAG D. MOORADIAN Chapter 210 Neuromuscular Manifestations of Endocrine Disease ROBERT B. LAYZER and GARY M. ABRAMS Chapter 211 Rheumatic Manifestations of Endocrine Disease DAVID J. NASHEL Chapter 212 Hematologic Endocrinology HARVEY S. LUKSENBURG, STUART L. GOLDBERG, and CRAIG M. KESSLER Chapter 213 Infectious Diseases and Endocrinology CARMELITA U. TUAZON and STEPHEN A. MIGUELES Chapter 214 Endocrine Disorders in Human Immunodeficiency Virus Infection STEPHEN A. MIGUELES and CARMELITA U. TUAZON Chapter 215 The Eye in Endocrinology ROBERT A. OPPENHEIM and WILLIAM D. MATHERS Chapter 216 Otolaryngology and Endocrine Disease STEPHEN G. HARNER Chapter 217 Dental Aspects of Endocrinology ROBERT S. REDMAN Chapter 218 The Skin and Endocrine Disorders JO-DAVID FINE, ADNAN NASIR, and KENNETH L. BECKER

PART XV HORMONES AND CANCER

Kenneth L. Becker, Editor
Chapter 219 Paraneoplastic Endocrine Syndromes KENNETH L. BECKER and OMEGA L. SILVA Chapter 220 Endocrine Tumors of the Gastrointestinal Tract SHAHRAD TAHERI, KARIM MEERAN, and STEPHEN BLOOM Chapter 221 Carcinoid Tumor and the Carcinoid Syndrome PAUL N. MATON Chapter 222 Hormones and Carcinogenesis: Laboratory Studies JONATHAN J. LI and SARA ANTONIA LI Chapter 223 Sex Hormones and Human Carcinogenesis: Epidemiology ROBERT N. HOOVER Chapter 224 Endocrine Treatment of Breast Cancer GABRIEL N. HORTOBAGYI Chapter 225 Endocrine Aspects of Prostate Cancer CHULSO MOON and CHRISTOPHER J. LOGETHETIS Chapter 226 Endocrine Consequences of Cancer Therapy DAIVA R. BAJORUNAS

PART XVI ENDOCRINOLOGY OF CRITICAL ILLNESS

Eric S. Nyln, Editor
Chapter 227 Critical Illness and Systemic Inflammation GARY P. ZALOGA BANKIM BHATT, and PAUL MARIK Chapter 228 Endocrine Markers and Mediators in Critical Illness ABDULLAH A. ALARIFI, GREET H. VAN DEN BERGHE, RICHARD H. SNIDER, JR., KENNETH L. BECKER, BEAT MLLER, and ERIC S. NYLN Chapter 229 The HypothalamicPituitaryAdrenal Axis in Stress and Critical Illness STEFAN R. BORNSTEIN and GEORGE P. CHROUSOS Chapter 230 Neuroendocrine Response to Acute Versus Prolonged Critical Illness GREET H. VAN DEN BERGHE Chapter 231 Fuel Metabolism and Nutrient Delivery in Critical Illness THOMAS R. ZIEGLER Chapter 232 Endocrine Therapeutics in Critical Illness ERIC S. NYLN, GARY P. ZALOGA, KENNETH L. BECKER, KENNETH D. BURMAN, LEONARD WARTOFSKY, BEAT MLLER, JON C. WHITE, and ABDULLAH A. ALARIFI

PART XVII ENDOCRINE AND METABOLIC EFFECTS OF TOXIC AGENTS

Kenneth L. Becker, Editor
Chapter 233 Endocrine-Metabolic Effects of Alcohol ROBERT H. NOTH and ARTHUR L. M. SWISLOCKI Chapter 234 Metabolic Effects of Tobacco, Cannabis, and Cocaine OMEGA L. SILVA Chapter 235 Environmental Factors and Toxins and Endocrine Function LAURA S. WELCH

PART XVIII ENDOCRINE DRUGS AND VALUES

Kenneth L. Becker, Editor
Chapter 236 Compendium of Endocrine-Related Drugs DOLLY MISRA, MICHELLE FISCHMANN MAGEE, and ERIC S. NYLN Chapter 237 Reference Values in Endocrinology D. ROBERT DUFOUR Chapter 238 Techniques of Laboratory Testing D. ROBERT DUFOUR Chapter 239 Effects of Drugs on Endocrine Function and Values MEETA SHARMA Chapter 240 DNA Diagnosis of Endocrine Disease J. FIELDING HEJTMANCIK and HARRY OSTRER Chapter 241 Dynamic Procedures in Endocrinology D. ROBERT DUFOUR and WILLIAM A. JUBIZ

PREFACE
This third edition of Principles and Practice of Endocrinology and Metabolism has been substantially and systematically revised. All of the chapters have been updated, many have been entirely rewritten, and many deal with completely new topics. Furthermore, additional important information and references have been added up until the very date of printing. The new chapters covering topics that did not appear in depth in the prior edition include: Molecular Biology: Present and Future; Pituitary Tumors: Overview of Therapeutic Options; The Incidental Adrenal Mass; Appetite; Pancreas and Islet Transplantation; Syndrome X; Endocrine Effects on Lipids; Compendium of Growth Factors and Cytokines; The Endocrine Blood Cells; The Endocrine Adipocyte; and Endocrine Disorders in Human Immunodeficiency Virus Infection. We would like to welcome the authors of these chapters, and also the new authors who have updated, extensively revised, or have entirely rewritten chapters on topics that appeared in the last edition. A new section has been added to this textbook: Endocrinology of Critical Illness. The six chapters comprising this section address the multiple aspects of this condition in a manner that is unique. Critical illness, which to some extent afflicts the great bulk of humankind at some time in their lives, has enormous hormonal and metabolic dimensions that relate directly to the diagnosis of the illness, influence the response of the host and the consequent evolution of the condition, and play a role in its outcome. The specific chapters include Critical Illness and Systemic Inflammation, Endocrine Markers and Mediators in Critical Illness, The HypothalamicPituitaryAdrenal Axis in Stress and Critical Illness, Neuroendocrine Response to Acute versus Prolonged Critical Illness, Fuel Metabolism and Nutrient Delivery in Critical Illness, and Endocrine Therapeutics in Critical Illness. These subjects are of great importance to every endocrine clinician as well as many who are involved in fundamental endocrine research. Overall, the goal of this textbook is to continue to provide, in a readable, understandable, and well-illustrated format, the clinical and basic information on endocrinology and metabolism that will be useful to both clinicians and basic scientists. We also wish this book to be a useful source of information for internists, house staff, and medical students. We have attempted to cover the field thoroughly and broadly, to include most of the known endocrine and metabolic disorders and hormonal messenger molecules, to furnish appropriate and current references, and to be of practical benefit to our readers. A complete CD version of this entire textbook is available. It contains approximately 4000 self-assessment questions that have been assembled and edited by Dr. Meeta Sharma. I wish to acknowledge the very helpful library assistance of Joanne Bennett. I am very grateful for the indispensable editorial and pharmaceutical aid of my Editorial Assistant, Roberta L. Brown, Pharm. D. Kenneth L. Becker, MD , PhD

PREFACE TO THE FIRST EDITION

Although there are several excellent large textbooks of endocrinology, we have felt the need for a book which would aim at encompassing all aspects of the field, a book which would be disease-oriented, would have practical applicability to the care of the adult and pediatric patient, and could be consulted to obtain a broad range of pathophysiologic, diagnostic, and therapeutic information. To fulfill this goal we called upon not only eminent specialists in endocrinology but also upon experts in many fields of medicine and science. The first part of the book surveys general aspects of endocrinology. The eight succeeding parts deal with specific fields of endocrinology: The Endocrine Brain and Pituitary Gland, The Thyroid Gland, Calcium and Bone Metabolism, The Adrenal Glands, Sex Determination and Development, Endocrinology of the Female, Endocrinology of the Male, and Disorders of Fuel Metabolism. Each of these parts contains relevant anatomic, physiologic, diagnostic, and therapeutic information and, when indicated, pediatric coverage of the topic. Diffuse Hormonal Secretion expounds upon the fact that endocrine function is not confined to anatomically discrete endocrine glands but is also intrinsic to all tissues and organs. This part is divided in two; it first presents a discussion of hormones which have a diffuse distribution and are not reviewed elsewhere in the book, and subsequently it deals with body constituents which are important sites of hormonal secretion. Heritable Abnormalities of Endocrinology and Metabolism underlines the importance of genetics in the causation of many endocrine and metabolic abnormalities. Endocrine and metabolic dysfunction in the young and in the aged is the subject of a separate part, because in both of these age groups hormonal function as well as endocrine disorders differ profoundly from those of individuals in their middle decades. Interrelationships Between Hormones and the Body discusses the impact of hormones on the soma and addresses clinical aspects of the disorders they may engender. Hormones and Cancer examines the phenomenon of hormone-induced neoplasms, elaborating on the fact that all neoplasms secrete hormones, that several of these hormones can cause additional clinical disorders, and that some neoplasms respond therapeutically to hormonal manipulation. The ensuing part, entitled Endocrine and Metabolic Effects of Toxic Agents deals with the sometimes subtle, sometimes profound influence of four nearly omnipresent agents: medication, alcohol, tobacco, and cannabis; it also addresses the consequences of environmental toxins on the endocrine system. The last part deals with the therapeutic use of drugs in endocrinology and the proper interpretation of laboratory values. It offers an extensive table on the clinical use of endocrine-related drugs, a table on reference values, and an outline of the dynamic procedures used in endocrinology. The goal of these tabular chapters is to facilitate the day-to-day evaluation and therapy of the endocrine patient. As a rule, the emphasis of this textbook is on the endocrinology of the human being. Animal data are presented only when contributing to a better understanding of human physiology and pathology. To maximize current relevance, historical information is kept to a minimum. While efforts were made to avoid repetition, the coverage of certain topics may recur when viewed from different standpoints. It is hoped that this will provide a wider dimension of the understanding of endocrine and metabolic function and dysfunction. In order not to interrupt continuity, bibliographic references are grouped at the end of each part. Finally, with the interest of the reader in mind, particular attention was given to composing an index as detailed as possible. I wish to thank the associate editors of this text for their skill, their enthusiasm, and their hard work. We all are very grateful for the expertise of our many eminent contributors. During the preparation of the manuscripts, there was considerable inter-communication between these contributors and their respective editors concerning both content and presentation. I wish to acknowledge the participation of Richard H. Snider, PhD, and Eric S. Nyln, MD, who have provided outstanding editorial assistance throughout the preparation of the textbook. The field of endocrinology and metabolism is evolving rapidly. New data are being developed continuously, and with this in mind, all contributors were encouraged to add up-to-date information until nearly the date of publication. There are numerous matters upon which there is no current common agreement, and logical arguments can be marshaled to buttress diametrically different viewpoints. This textbook is written by many authors; though most of the beliefs and conclusions of the contributors tend to reflect those of the editors, no attempt was made to impose a uniformity of pathophysiologic, diagnostic, or therapeutic viewpoints, and the book does not lack for differences of opinion. We hope that the Principles and Practice of Endocrinology and Metabolism will be a relevant sourcebook for those interested in the science and the practice of this fascinating discipline, whether they be clinicians, basic scientists, allied health personnel, or students. Kenneth L. Becker, MD , PhD

CONTRIBUTING AUTHORS
Alaa Abou-Saif, MD Gastroenterology Fellow Department of Medicine Division of Gastroenterology Georgetown University School of Medicine Washington, DC Gary M. Abrams, MD Associate Professor of Clinical Neurology Department of Neurology University of California, San Francisco, School of Medicine San Francisco, California Thomas Aceto, Jr., MD Professor of Pediatrics Chairman Emeritus of Pediatrics Saint Louis University School of Medicine Cardinal Glennon Childrens Hospital St. Louis, Missouri Bharat B. Aggarwal, PhD Professor of Medicine and Biochemistry Department of Bioimmunotherapy Chief, Cytokine Research Section University of TexasHouston Medical School M. D. Anderson Cancer Center Houston, Texas Zalman S. Agus, MD Emeritus Professor of Medicine University of Pennsylvania School of Medicine Philadelphia, Pennsylvania Rexford S. Ahima, MD , PhD Assistant Professor of Medicine Division of Endocrinology, Diabetes and Metabolism University of Pennsylvania School of Medicine Philadelphia, Pennsylvania Andrew J. Ahmann, MD Assistant Professor of Medicine Director of Adult Diabetes Services Oregon Health Sciences University School of Medicine Portland, Oregon Abdullah A. Alarifi, MD Consultant Endocrinologist Department of Medicine King Faisal Specialist Hospital and Research Centre Riyadh, Kingdom of Saudi Arabia K. George M. M. Alberti, MD , DPhil, PRCP, FRCP Professor of Medicine Department of Diabetes and Metabolism University of Newcastle upon Tyne Faculty of Medicine Newcastle upon Tyne, England Melvin G. Alper, MD Private Practice, Ophthalmology Chevy Chase, Maryland John K. Amory, MD Assistant Professor Department of Medicine University of Washington School of Medicine Veterans Affairs Puget Sound Health Care System Seattle, Washington Neil Aronin, MD Professor of Medicine and Cell Biology Director, Division of Endocrinology and Metabolism University of Massachusetts Medical School Worcester, Massachusetts Louis J. Aronne, MD Clinical Associate Professor of Medicine Weill Medical College of Cornell University New York, New York Gilbert P. August, MD Professor of Pediatrics Department of Endocrinology George Washington University School of Medicine and Health Sciences Childrens National Medical Center Washington, DC Lloyd Axelrod, MD Associate Professor of Medicine Harvard Medical School Physician and Chief of the James Howard Means Firm Massachusetts General Hospital

Boston, Massachusetts Daiva R. Bajorunas, MD Senior Director, Clinical Research Global Project Team Leader, Metabolism Aventis Pharmaceuticals Bridgewater, New Jersey H. W. Gordon Baker, MD , PhD, FRACP Associate Professor Department of Obstetrics and Gynaecology University of Melbourne School of Medicine Royal Womens Hospital Victoria, Australia James R. Baker, Jr., MD Professor of Medicine Department of Internal MedicineAllergy and Immunology Chief, Division of Allergy University of Michigan Medical School Ann Arbor, Michigan William A. Banks, MD Professor of Internal Medicine Division of Geriatrics Saint Louis University School of Medicine St. Louis, Missouri Robert L. Barbieri, MD Kate Macy Ladd Professor of Obstetrics, Gynecology and Reproductive Biology Harvard Medical School Boston, Massachusetts Marcelo J. Barrionuevo, MD Assistant Professor Department of Obstetrics and Gynecology University of Miami School of Medicine Margate, Florida David S. Baskin, MD , FACS Professor of Neurosurgery and Anesthesiology Baylor College of Medicine Houston, Texas Gerhard Baumann, MD Professor of Medicine Northwestern University Medical Center Chicago, Illinois Peter H. Baylis, MD , FRCP, FAMS Professor of Experimental Medicine Dean, Department of Medicine The Medical School University of Newcastle upon Tyne Faculty of Medicine Newcastle upon Tyne, England David V. Becker, MD Professor of Radiology and Medicine Division of Nuclear Medicine and Endocrinology Weill Medical College of Cornell University New York Presbyterian Hospital New York, New York Dorothy J. Becker, MB, Bch Professor of Pediatrics University of Pittsburgh School of Medicine Childrens Hospital of Pittsburgh Pittsburgh, Pennsylvania Kenneth L. Becker, MD , PhD Professor of Medicine Professor of Physiology and Experimental Medicine Director of Endocrinology and Metabolism George Washington University School of Medicine and Health Sciences Veterans Affairs Medical Center Washington, DC Norman H. Bell, MD Distinguished University Professor of Medicine Medical University of South Carolina College of Medicine Charleston, South Carolina Bankim Bhatt, MD Medical Resident Department of Medicine Georgetown University School of Medicine Washington Hospital Center Washington, DC John P. Bilezikian, MD Professor of Medicine and Pharmacology Department of Medicine Columbia University College of Physicians and Surgeons New York, New York Richard E. Blackwell, MD , PhD

Professor of Obstetrics and Gynecology University of Alabama School of Medicine Birmingham, Alabama Vicky A. Blakesley, MD , PhD Director, Department of New Product Evaluation International Division Abbott Laboratories Abbott Park, Illinois Stephen Bloom, MA, MD , DSc, FRCPath, FRCP, FMedSci Professor of Medicine Department of Metabolic Medicine Division of Investigative Science University of London Imperial College School of Medicine London, England Manfred Blum, MD Professor of Clinical Medicine and Radiology Director, Nuclear Endocrine Laboratory New York University School of Medicine New York, New York Nanci Bobrow, PhD Assistant Clinical Professor of Pediatrics Cardinal Glennon Childrens Hospital Saint Louis University School of Medicine St. Louis, Missouri Susan Bonner-Weir, PhD Associate Professor of Medicine Harvard Medical School Senior Investigator Joslin Diabetes Center Boston, Massachusetts Stefan R. Bornstein, MD , PhD Assistant Professor and Research Scholar Pediatric and Reproductive Endocrinology Branch National Institute of Child Health and Human Development National Institutes of Health Bethesda, Maryland Eric Bourekas, MD Assistant Professor of Radiology Section of Diagnostic and Interventional Neuroradiology Ohio State University College of Medicine and Public Health Columbus, Ohio William J. Bremner, MD , PhD Robert G. Petersdorf Professor and Chairman Department of Medicine University of Washington School of Medicine Seattle, Washington Edward M. Brown, MD Professor of Medicine EndocrineHypertension Division Harvard Medical School Brigham and Womens Hospital Boston, Massachusetts Henry B. Burch, MD Associate Professor of Medicine Uniformed Services University of the Health Sciences F. Edward Hbert School of Medicine Bethesda, Maryland Department of Endocrine-Metabolic Service Walter Reed Army Medical Center Washington, DC Kenneth D. Burman, MD Professor of Medicine Uniformed Services University of the Health Sciences F. Edward Hbert School of Medicine Bethesda, Maryland Clinical Professor Department of Medicine George Washington University School of Medicine and Health Sciences Professor of Medicine Georgetown University School of Medicine Chief, Endocrine Section Washington Hospital Center Washington, DC Peter H. Byers, MD Professor of Pathology and Medicine University of Washington School of Medicine Seattle, Washington Enrico Carmina, MD Professor Department of Endocrinology University of Palermo Palermo, Italy Visiting Professor Department of Obstetrics and Gynecology Columbia University College of Physicians and Surgeons

New York, New York Thomas O. Carpenter, MD Professor of Pediatrics Yale University School of Medicine YaleNew Haven Hospital New Haven, Connecticut Bruce R. Carr, MD Professor Paul C. Macdonald Distinguished Chair in Obstetrics and Gynecology Director, Division of Reproductive Endocrinology University of Texas Southwestern Medical Center at Dallas Southwestern Medical School Dallas, Texas Veronica M. Catanese, MD Assistant Professor Department of Medicine and Cell Biology New York University School of Medicine New York, New York Donald Chakeres, MD Professor of Radiology Ohio State University College of Medicine and Public Health Columbus, Ohio John R. G. Challis, PhD, Dsc, FIBiol , FRCOG, FRSC Department of Physiology Medical Sciences Building University of Toronto Faculty of Medicine Toronto, Ontario Canada Philippe Chanson, MD Professor of Medicine Department of Endocrinology University Paris XI Bictre University Hospital Le Kremlin-Bictre France William W. Chin, MD Professor of Medicine Harvard Medical School Boston, Massachusetts Vice President, Lilly Research Laboratories Eli Lilly & Co. Lilly Corporate Center Indianapolis, Indiana George P. Chrousos, MD Chief, Pediatric and Reproductive Endocrinology Branch National Institutes of Health Bethesda, Maryland Richard V. Clark, MD , PhD Principal Clinical Research Physician Clinical PharmacologyExploratory Department Glaxo Wellcome Research and Development Research Triangle Park, North Carolina Thomas L. Clemens, MD , PhD Professor of Medicine and Molecular and Cellular Physiology Department of Internal Medicine/Endocrinology University of Cincinnati College of Medicine Cincinnati, Ohio Fredric L. Coe, MD Professor Departments of Medicine and Physiology University of Chicago Pritzker School of Medicine Chicago, Illinois Joshua L. Cohen, MD Associate Professor of Medicine Department of Endocrinology George Washington University School of Medicine and Health Sciences Washington, DC Rgis Cohen, MD , PhD Praticien Hospitalier Endocrine Staff Physician Avicenne Hospital Bobigny, France University of Leonardo Da Vinci Paris, France Warren E. Cohen, MD Associate Clinical Professor of Pediatrics and Neurology George Washington University School of Medicine and Health Sciences Washington, DC Medical Director, United Cerebral Palsy Nassau County, New York Alessandra Colantoni, MD Assistant Professor of Medicine

Department of Gastroenterology and Hepatology Loyola University of Chicago Stritch School of Medicine Loyola University Medical Center Maywood, Illinois Richard J. Comi, MD Associate Professor of Medicine Section of Endocrinology and Metabolism Dartmouth Medical School DartmouthHitchcock Medical Center Hanover, New Hampshire Paul E. Cooper, MD , FRCPC Associate Professor of Neurology Departments of Clinical Neurological Sciences and Medicine University of Western Ontario Faculty of Medicine and Dentistry Health Sciences Addition London, Ontario Canada Dalila B. Corry, MD Associate Clinical Professor of Medicine Department of Medicine University of California, Los Angeles, UCLA School of Medicine Los Angeles, California Chief, Nephrology Olive View Medical Center Sylmar, California Francesco Cosentino, MD , PhD Assistant Professor Department of Experimental Medicine and Pathology University La Sapienza Rome, Italy Senior Research Associate Cardiovascular Research Department of Cardiology University Hospital Zurich, Switzerland Felicia Cosman, MD Associate Professor of Clinical Medicine Department of Medicine Columbia University College of Physicians and Surgeons New York, New York Helen Hayes Hospital West Haverstraw, New York Brian M. Cox, PhD Professor of Pharmacology and Neuroscience Department of Pharmacology Uniformed Services University of the Health Sciences F. Edward Hbert School of Medicine Bethesda, Maryland Glenn R. Cunningham, MD Associate Chief of Staff Department of Medicine University of TexasHouston Medical School Veterans Affairs Medical Center Houston, Texas Mary F. Dallman, PhD Professor of Physiology University of California, San Francisco, School of Medicine San Francisco, California Daniel N. Darlington, PhD Associate Professor of Surgery Departments of Surgery and Physiology University of Maryland School of Medicine Baltimore, Maryland Philip Darney, MD , MSc Professor of Obstetrics, Gynecology, and Reproductive Sciences University of California, San Francisco, School of Medicine San Francisco General Hospital San Francisco, California Harish P. G. Dave, MB, ChB, MRCP(UK) Associate Professor of Medicine Department of Hematology George Washington University School of Medicine and Health Sciences Veterans Affairs Medical Center Washington, DC Faith B. Davis, MD Professor of Medicine and Cell Biology and Cancer Research Albany Medical College Staff Physician Stratton Veterans Affairs Medical Center Albany, New York Paul J. Davis, MD Professor of Medicine and Cell Biology and Cancer Research Senior Associate Dean for Clinical Research Albany Medical College Research Physician

Wadsworth Center, New York State Department of Health Staff Physician Stratton Veterans Affairs Medical Center Albany, New York Suzanne M. Jan De Beur, MD Assistant Professor of Medicine Johns Hopkins University School of Medicine Baltimore, Maryland Ralph A. DeFronzo, MD Professor of Medicine Chief, Diabetic Division Member, Nephrology Division University of Texas Medical School at San Antonio University Health Center San Antonio, Texas David M. De Kretser, MD , MBBS, FRACP Professor and Director Monash Institute of Reproduction and Development Monash University Monash Medical Centre, Clayton Clayton, Victoria Australia Nicola De Maria, MD Research Associate Liver Transplant Program Loyola University Medical Center Maywood, Illinois David P. Dempsher, MD , PhD Associate Professor of Pediatrics Cardinal Glennon Childrens Hospital Saint Louis University School of Medicine St. Louis, Missouri David W. Dempster, PhD Professor of Clinical Pathology Columbia University College of Physicians and Surgeons New York, New York Director, Regional Bone Center Helen Hayes Hospital West Haverstraw, New York Luca deSimone Nephrology Fellow Beth Israel Medical Center New York, New York Gerard M. Doherty, MD Associate Professor of Surgery Section of Surgical Oncology and Endocrinology Washington University School of Medicine St. Louis, Missouri Allan L. Drash, MD Emeritus Professor of Pediatrics University of Pittsburgh School of Medicine Pittsburgh, Pennsylvania Marc K. Drezner, MD Professor of Medicine Head, Section of Endocrinology, Diabetes, and Metabolism University of Wisconsin Medical School Madison, Wisconsin Alan Dubrow, MD Clinical Assistant Professor of Medicine Department of Nephrology Beth Israel Deaconess Medical Center New York, New York D. Robert Dufour, MD Clinical Professor of Pathology George Washington University School of Medicine and Health Sciences Washington, DC Uniformed Services University of the Health Sciences F. Edward Hbert School of Medicine Bethesda, Maryland Chief, Pathology and Laboratory Medicine Service Veterans Affairs Medical Center Washington, DC Roberta P. Durschlag, PhD, RD Clinical Assistant Professor Department of Health Sciences Boston University School of Medicine Boston, Massachusetts Richard C. Eastman, MD Cygnus, Inc. Redwood City, California George S. Eisenbarth, MD , PhD Professor of Pediatrics, Immunology, and Medicine

University of Colorado Health Sciences Center Barbara Davis Center for Childhood Diabetes Denver, Colorado George M. Eliopoulos, MD Associate Professor of Medicine Harvard Medical School Beth Israel Deaconess Medical Center Boston, Massachusetts William J. Ellis, MD Associate Professor and Clinic Director Department of Urology University of Washington School of Medicine Seattle, Washington Abby Erickson, BA Colorado Center for Bone Research Lakewood, Colorado Gregory F. Erickson, PhD Professor Department of Reproductive Medicine University of California, San Diego, School of Medicine La Jolla, California Eric A. Espiner, MD , FRACP, FRS(NZ) Professor Department of Endocrinology University of Otago Christchurch School of Medicine Christchurch Public Hospital Christchurch, New Zealand Jan Fahrenkrug, MD , DMSci Professor Department of Clinical Chemistry University of Copenhagen Faculty of Health Sciences Bispebjerg Hospital Copenhagen, Denmark Kenneth R. Falchuk, MD Associate Professor of Medicine Harvard Medical School Beth Israel Deaconess Medical Center Boston, Massachusetts Murray J. Favus, MD Professor of Medicine University of Chicago Pritzker School of Medicine Chicago, Illinois Eva L. Feldman, MD , PhD Professor of Neurology University of Michigan Medical School Ann Arbor, Michigan Jo-David Fine, MD , MPH Professor Department of Dermatology University of North Carolina at Chapel Hill School of Medicine Chapel Hill, North Carolina James D. Finkelstein, MD Senior Clinician Department of Medicine Veterans Affairs Medical Center Washington, DC Jeffrey S. Flier, MD George C. Reisman Professor of Medicine Harvard Medical School Beth Israel Deaconess Medical Center Boston, Massachusetts Ruth C. Fretts, MD , MPH Assistant Professor Department of Obstetrics and Gynecology Harvard Medical School Beth Israel Deaconess Medical Center Boston, Massachusetts Om P. Ganda, MD Associate Clinical Professor Department of Medicine Harvard Medical School Joslin Diabetes Center Beth Israel Deaconess Medical Center Boston, Massachusetts Luigi Garibaldi Beth Israel Medical Center Newark, New Jersey Gary W. Gibbons, MD Associate Clinical Professor of Surgery

Harvard Medical School Director, Quality Improvement Department of Surgery Beth Israel Deaconess Medical Center Boston, Massachusetts John R. Gill, Jr., MD Scientist, Emeritus Hypertension-Endocrine Branch National Heart, Lung, and Blood Institute National Institutes of Health Bethesda, Maryland Henry N. Ginsberg, MD Professor of Medicine Columbia University College of Physicians and Surgeons New York, New York Joel S. Glaser, MD Professor Departments of Ophthalmology and Neurology University of Miami School of Medicine Bascom Palmer Eye Institute Miami, Florida Department of Ophthalmology Cleveland Clinic of Florida Coral Gables, Florida Allan R. Glass, MD Adjunct Professor of Medicine Uniformed Services University of the Health Sciences F. Edward Hbert School of Medicine Bethesda, Maryland Philip W. Gold, MD Branch Chief Department of Intramural Research Programs National Institute of Mental Health National Institutes of Health Bethesda, Maryland Alisa B. Goldberg, MD Assistant Adjunct Professor Department of Obstetrics, Gynecology and Reproductive Sciences University of California, San Francisco, School of Medicine San Francisco General Hospital San Francisco, California Ira J. Goldberg, MD Professor of Medicine Columbia University College of Physicians and Surgeons New York, New York Stuart L. Goldberg, MD Assistant Director, Bone Marrow Transplantation Program Temple University School of Medicine Philadelphia, Pennsylvania Allison B. Goldfine, MD Instructor of Medicine Department of Cellular and Molecular Physiology Harvard University Joslin Diabetes Center Boston, Massachusetts Allan L. Goldstein, PhD Chair, Department of Biochemistry and Molecular Biology George Washington University School of Medicine and Health Sciences Washington, DC David S. Goldstein, MD , PhD Chief, Clinical Neurocardiology Section National Institutes of Health Bethesda, Maryland David Goltzman, MD Professor of Medicine and Physiology McGill University Faculty of Medicine Royal Victoria Hospital Montreal, Quebec Canada Esther A. Gonzalez, MD Assistant Professor Division of Nephrology Saint Louis University School of Medicine St. Louis, Missouri Michael N. Goodman, PhD Professor of Medicine Department of Internal Medicine University of California, Davis, School of Medicine Sacramento, California Phillip Gorden, MD Director Emeritus National Institute of Diabetes and Digestive and Kidney Diseases

National Institutes of Health Bethesda, Maryland Frederic D. Gordon, MD Assistant Professor of Medicine Department of Hepatobiliary Surgery and Liver Transplantation Tufts University School of Medicine Lahey Clinic Medical Center Boston, Massachusetts Daryl K. Granner, MD Joe C. Davis Professor of Biomedical Science Professor of Molecular Physiology, Biophysics, and Internal Medicine Vanderbilt University School of Medicine Director, Vanderbilt Diabetes Center Staff Physician Veterans Affairs Hospital Nashville, Tennessee Sren Grs, MD Senior Registrar Department of Obstetrics and Gynaecology Herlev University Hospital Herlev, Denmark Douglas A. Greene, MD Executive Vice President Department of Clinical Sciences and Product Development Merck & Co., Inc. Rahway, New Jersey David A. Gruenewald, MD , FACP Assistant Professor of Medicine University of Washington School of Medicine Veterans Affairs Puget Sound Health Care System Seattle, Washington Joel F. Habener, MD Professor of Medicine Laboratory of Molecular Endocrinology Harvard Medical School Massachusetts General Hospital Boston, Massachusetts Philippe A. Halban, PhD Professor of Medicine Louis-Jeantet Research Laboratories Geneva University Medical Center Geneva, Switzerland Nicholas R. S. Hall, PhD Health and Human Performance Orlando, Florida Allan G. Halline, MD Assistant Professor of Medicine Section of Digestive and Liver Diseases University of Illinois at Chicago College of Medicine Chicago, Illinois Stephen G. Harner, MD Professor of Otolaryngology Department of Otolaryngology Mayo Medical School Rochester, Minnesota Marianne Hatle, MD Resident University of Maryland School of Medicine Baltimore, Maryland Michael J. Hausmann, MD Professor Department of Nephrology Faculty of Health Sciences Ben Gurion University of the Negev Scroka Medical Center of Kupat Holim Beer Sheva, Israel Karin Hehenberger, MD , PhD Research Fellow Joslin Diabetes Center Harvard Medical School Boston, Massachusetts J. Fielding Hejtmancik, MD , PhD Medical Officer National Eye Institute National Institutes of Health Bethesda, Maryland Geoffrey N. Hendy, PhD Professor of Medicine McGill University Faculty of Medicine Royal Victoria Hospital Montreal, Quebec

Canada James V. Hennessey, MD Associate Professor of Medicine Division of Endocrinology Brown University School of Medicine Rhode Island Hospital Providence, Rhode Island Jules Hirsch, MD Professor Emeritus and Physician-in-Chief Emeritus Laboratory of Human Behavior and Metabolism Rockefeller University Rockefeller University Hospital New York, New York Angelica Lindn Hirschberg, MD , PhD Associate Professor of Obstetrics and Gynecology Karolinska Institute Karolinska Hospital Stockholm, Sweden Max Hirshkowitz, MD Associate Professor Department of Psychiatry Baylor College of Medicine Director, Sleep Center Houston Veterans Affairs Medical Center Houston, Texas Gary D. Hodgen, PhD Professor Department of Obstetrics and Gynecology Eastern Virginia Medical School Chair The Howard and Georgeanna Jones Institute for Reproductive Medicine Norfolk, Virginia Edward W. Holmes, MD Chairman, Department of Medicine University of Pennsylvania School of Medicine Philadelphia, Pennsylvania Jens J. Holst, MD Department of Medical Physiology University of Copenhagen Faculty of Health Sciences The Panum Institute Copenhagen, Denmark Robert N. Hoover, MD , ScD Director, Epidemiology and Biostatistics Program National Cancer Institute National Institutes of Health Bethesda, Maryland Gabriel N. Hortobagyi, MD Professor of Medicine, Chairman Department of Breast and Gynecologic Medical Oncology University of TexasHouston Medical School M. D. Anderson Cancer Center Houston, Texas Eva Horvath, PhD Associate Professor of Pathology Department of Laboratory Medicine Division of Pathology University of Toronto Faculty of Medicine St. Michaels Hospital Toronto, Ontario Canada Barbara V. Howard, PhD President, MedStar Clinical Research Institute Washington, DC William James Howard, MD Professor of Medicine George Washington University School of Medicine Senior Vice President and Medical Director Washington Hospital Center Washington, DC Ilpo Huhtaniemi, MD , PhD Professor of Physiology University of Turku Faculty of Medicine Turku, Finland Wellington Hung, MD , PhD Professor Emeritus of Pediatrics Georgetown University School of Medicine Professorial Lecturer in Pediatrics George Washington University School of Medicine and Health Sciences Washington, DC Mehboob A. Hussain, MD Department of Medicine

New York University School of Medicine New York, New York Philip M. Iannaccone, MD , PhD George M. Eisenberg Professor Department of Pediatrics Northwestern University Medical School Childrens Memorial Institute of Education and Research Chicago, Illinois Ivor M. D. Jackson, MB, ChB Professor of Medicine Division of Endocrinology Brown University School of Medicine Rhode Island Hospital Providence, Rhode Island Richard V. Jackson, MBBS, FRACP Associate Professor of Medicine University of Queensland Faculty of Health Sciences Greenslopes Private Hospital Queensland, Australia Lois Jovanovic, MD Clinical Professor of Medicine University of Southern California School of Medicine Los Angeles, California Director and Chief Scientific Officer Sansum Medical Research Institute Santa Barbara, California William A. Jubiz, MD Director Endocrinology Center Cali, Colombia C. Ronald Kahn, MD Mary K. Iacocca Professor of Medicine Harvard Medical School President and Director, Research Division Joslin Diabetes Center Boston, Massachusetts Cynthia G. Kaplan, MD Associate Professor of Pathology SUNY at Stony Brook School of Medicine Health Sciences Center Stony Brook, New York Edwin L. Kaplan, MD , FACS Professor of Surgery University of Chicago Pritzker School of Medicine Chicago, Illinois Abba J. Kastin, MD Chief of Endocrinology Departments of Medicine and Neuroscience Tulane University School of Medicine Veterans Affairs Medical Center New Orleans, Louisiana Laurence Katznelson, MD Assistant Professor of Medicine Harvard Medical School Massachusetts General Hospital Boston, Massachusetts Harry R. Keiser, MD Scientist Emeritus National Heart, Lung, and Blood Institute Clinical Center National Institutes of Health Bethesda, Maryland Ellie Kelepouris, MD Professor of Medicine Temple University School of Medicine Philadelphia, Pennsylvania Craig M. Kessler, MD Professor of Medicine and Pathology Chief, Division of Hematology-Oncology Georgetown University School of Medicine Lombardy Cancer Center Washington, DC Parvez Khatri, MD Fellow, Department of Medicine/Nephrology George Washington University School of Medicine and Health Sciences Washington, DC Paul L. Kimmel, MD Professor of Medicine George Washington University School of Medicine and Health Sciences Washington, DC Director, Diabetic Nephropathy Program Division of Kidney, Urologic, and Hematologic Diseases National Institute of Diabetes and Digestive and Kidney Diseases

National Institutes of Health Bethesda, Maryland George L. King, MD Professor of Medicine Acting Director of Research Joslin Diabetes Center Harvard Medical School Boston, Massachusetts Anne Klibanski, MD Professor of Medicine Harvard Medical School Chief, Neuroendocrine Unit Massachusetts General Hospital Boston, Massachusetts Mitchel A. Kling, MD Associate Professor of Psychiatry and Medicine University of Maryland School of Medicine Veterans Affairs Medical Center Baltimore, Maryland Mark Korson, MD Associate Professor of Pediatrics Division of Genetics Tufts University School of Medicine New England Medical Center Boston, Massachusetts Kalman Kovacs, MD , PhD Professor of Pathology Department of Laboratory Medicine Division of Pathology University of Toronto Faculty of Medicine Saint Michaels Hospital Toronto, Ontario Canada Andrzej S. Krolewski, MD , PhD Associate Professor of Medicine Chief, Section of Genetics and Epidemiology Harvard Medical School Research Division Joslin Diabetes Center Boston, Massachusetts Robert J. Kurman, MD Richard W. TeLinde Distinguished Professor of Gynecologic Pathology Departments of Gynecology, Obstetrics, and Pathology Johns Hopkins University School of Medicine Baltimore, Maryland John C. LaRosa, MD , FACP President SUNY Downstate Medical Center College of Medicine University Hospital of Brooklyn Brooklyn, New York Robert B. Layzer, MD Professor Emeritus of Neurology University of California, San Francisco, School of Medicine San Francisco, California Jacques LeBlanc, MD Professor Emeritus of Physiology Universit Laval Faculty of Medicine Quebec City, Canada Peter A. Lee, MD , PhD Professor of Pediatrics Pennsylvania State University College of Medicine The Milton S. Hershey Medical Center Hershey, Pennsylvania Z. M. Lei, MD , PhD Assistant Professor of Obstetrics and Gynecology University of Louisville School of Medicine Louisville, Kentucky Hoyle Leigh, MD Professor of Psychiatry University of California, San Francisco, School of Medicine San Francisco, California Derek LeRoith, MD , PhD Chief, Molecular and Cellular Endocrinology Branch National Institute of Diabetes and Digestive and Kidney Diseases National Institutes of Health Bethesda, Maryland Michael A. Levine, MD Professor of Pediatrics, Medicine, and Pathology Director, Pediatric Endocrinology Johns Hopkins University School of Medicine

Baltimore, Maryland Jonathan J. Li, PhD Director, Division of Etiology and Prevention of Hormone-Associated Cancers Professor of Pharmacology, Toxicology and Preventive Medicine University of Kansas School of Medicine Kansas Cancer Institute Kansas City, Kansas Sara Antonia Li, MD Associate Director Hormonal Carcinogenesis Laboratory University of Kansas School of Medicine Kansas Cancer Institute Kansas City, Kansas Robert D. Lindeman, MD Professor Emeritus of Medicine Department of Internal Medicine University of New Mexico School of Medicine University of New Mexico Hospital Albuquerque, New Mexico Robert Lindsay, MBChB, PhD, FRCP Professor of Clinical Medicine Columbia University College of Physicians and Surgeons New York, New York Chief of Internal Medicine Helen Hayes Hospital West Haverstraw, New York Timothy O. Lipman, MD Professor of Medicine Georgetown University School of Medicine Chief, GastroenterologyHepatology Nutrition Section Veterans Affairs Medical Center Washington, DC Virginia A. Livolsi, MD Professor of Pathology Department of Pathology and Laboratory Medicine University of Pennsylvania School of Medicine Philadelphia, Pennsylvania Rogerio A. Lobo, MD Willard C. Rappleye Professor of Obstetrics and Gynecology Chairman, Department of Obstetrics and Gynecology Columbia University College of Physicians and Surgeons Columbia Presbyterian Medical Center Director, Sloane Hospital for Women New York, New York Rebecca J. Locke Research Assistant Columbia University College of Physicians and Surgeons New York, New York Christopher J. Logethetis, MD Chairman and Professor Department of Genitourinary Medical Oncology University of TexasHouston Medical School M. D. Anderson Cancer Center Houston, Texas D. Lynn Loriaux, MD , PhD Professor and Chair Department of Medicine Oregon Health Sciences University School of Medicine Portland, Oregon Harvey S. Luksenburg, MD Assistant Professor of Medicine George Washington University School of Medicine and Health Sciences Washington, DC Thomas F. Lscher, MD Professor and Head of Cardiology Hospital Universitaire de Zurich Zurich, Switzerland Ruth S. MacDonald, RD, PhD Professor of Nutrition Department of Food Science and Human Nutrition University of MissouriColumbia School of Medicine Columbia, Missouri Michelle Fischmann Magee, MD , MB, BCh, BAO Medical Director, Diabetes Team MedStar Clinical Research Institute Washington Hospital Center Washington, DC Robert W. Mahley, MD , PhD Professor of Pathology and Medicine Director, Gladstone Institute of Cardiovascular Disease

University of California, San Francisco, School of Medicine San Francisco, California Christos S. Mantzoros, MD , Dsc Assistant Professor of Medicine Department of Internal Medicine Harvard Medical School Beth Israel Deaconess Medical Center Boston, Massachusetts Eleftheria Maratos-Flier, MD Associate Professor of Medicine Research Division Harvard Medical School Joslin Diabetes Center Boston, Massachusetts Paul Marik, MBBCh, FCP(SA), FRCP(C), FCCM, FCCP Department of Critical Care Mercy Hospital of Pittsburgh Pittsburgh, Pennsylvania Kevin J. Martin, MB, BCh, FACP Professor of Internal Medicine Department of Nephrology Director, Division of Nephrology Saint Louis University School of Medicine St. Louis, Missouri William D. Mathers, MD Professor of Ophthalmology Oregon Health Sciences University School of Medicine Casey Eye Institute Portland, Oregon Paul N. Maton, MD , FRCP, FACP, FACG Digestive Disease Specialists Incorporated Digestive Disease Research Institute Oklahoma City, Oklahoma Alvin M. Matsumoto, MD Professor Department of Medicine University of Washington School of Medicine Chief of Gerontology Veterans Affairs Puget Sound Health Care System Seattle, Washington Ernest L. Mazzaferri, MD , MACP Professor Emeritus and Chairman Department of Internal Medicine Ohio State University College of Medicine and Public Health Columbus, Ohio Alan M. McGregor, MA, MD , FRCP Professor of Medicine Kings College Guys, Kings and St. Thomas School of Medicine London, England Karim Meeran, MD , MRCP Senior Lecturer Division of Endocrinology and Metabolism University of London Imperial College School of Medicine Hammersmith Hospital London, England Minesh P. Mehta, MD , MB, ChB Associate Professor and Chairman Department of Human Oncology University of Wisconsin Medical School Madison, Wisconsin James C. Melby, MD Professor of Medicine and Physiology Boston University School of Medicine Boston Medical Center Boston, Massachusetts Stephen A. Migueles, MD Fellow, Infectious Diseases Laboratory of Immunoregulation National Institute of Allergy and Infectious Diseases National Institutes of Health Bethesda, Maryland Donald L. Miller, MD Professor of Radiology and Nuclear Medicine Uniformed Services University of the Health Sciences F. Edward Hbert School of Medicine Bethesda, Maryland Elizabeth A. Miller Urology Resident University of Washington School of Medicine Seattle, Washington

Paul D. Miller, MD Clinical Professor Department of Medicine University of Colorado Health Sciences Center Denver, Colorado Dolly Misra, MD Assistant Clinical Professor of Medicine Division of Endocrinology and Metabolism George Washington University School of Medicine and Health Sciences Washington, DC Diabetes and Endocrine Consultants Waldorf, Maryland Mark E. Molitch, MD Professor of Medicine Center for Endocrinology, Metabolism, and Molecular Medicine Northwestern University Medical School Chicago, Illinois Chulso Moon, MD , PhD Clinical Fellow Department of Medicine University of TexasHouston Medical School M. D. Anderson Cancer Center Houston, Texas Arshag D. Mooradian, MD Professor of Medicine Director of Endocrinology, Diabetes and Metabolism Saint Louis University School of Medicine St. Louis, Missouri Gregory P. Mueller, PhD Professor of Physiology Uniformed Services University of the Health Sciences F. Edward Hbert School of Medicine Bethesda, Maryland Beat Mller, MD Department of Internal Medicine Division of Endocrinology University Hospitals Basel, Switzerland Susan E. Myers, MD Assistant Professor of Pediatrics Saint Louis University School of Medicine Cardinal Glenn Childrens Hospital St. Louis, Missouri David J. Nashel, MD Professor of Medicine Georgetown University School of Medicine Chief of Medical Service Veterans Affairs Medical Center Washington, DC Adnan Nasir, MD , PhD Department of Dermatology University of North Carolina at Chapel Hill School of Medicine Chapel Hill, North Carolina Jeffrey A. Norton, MD Professor of Surgery Vice Chairman, Department of Surgery University of California, San Francisco, School of Medicine San Francisco Veterans Affairs Medical Center San Francisco, California Robert H. Noth, MD Associate Professor of Medicine Department of Internal Medicine University of California, Davis, School of Medicine Davis, California Veterans Affairs Outpatient Clinic Martinez, California Jennifer A. Nuovo Endocrinologist MedClinic of Sacramento Sacramento, California Eric S. Nyln, MD Associate Professor of Medicine Department of Endocrinology George Washington University School of Medicine and Health Sciences Veterans Affairs Medical Center Washington, DC Donna M. Arab OBrien, MD Department of Medicine Division of Endocrinology St. Josephs Health Centre Toronto, Ontario

Canada Mary Oehler, MD Staff Radiologist Mount Carmel East Hospital New Albany, Ohio Robert A. Oppenheim, MD Naperville Eye Associates Naperville, Illinois Jeffrey L. H. ORiordan Emeritus Professor of Metabolic Medicine University College London, United Kingdom Steven J. Ory, MD Clinical Associate Professor of Obstetrics and Gynecology University of Miami School of Medicine Miami, Florida Harry Ostrer, MD Associate Professor of Pediatrics and Pathology Human Genetics Program New York University School of Medicine New York, New York Weihong Pan, MD , PhD Assistant Professor of Medicine Tulane University School of Medicine New Orleans, Louisiana Yogesh C. Patel, MD , PhD, FACP, FRCP(C), FRACP, FRSC Professor of Medicine Director, Division of Endocrinology and Metabolism McGill University Faculty of Medicine Royal Victoria Hospital Montreal, Quebec Canada Gary R. Peplinski, MD Surgical Service San Francisco Veterans Affairs Medical Center San Francisco, California Ora Hirsch Pescovitz, MD Professor of Pediatrics, Physiology, and Biophysics Department of Pediatric Endocrinology Indiana University School of Medicine James Whitcomb Riley Hospital for Children Indianapolis, Indiana Kristina C. Pfendler, MD Postdoctoral Scholar Department of Obstetrics and Gynecology University of California, San Francisco, School of Medicine San Francisco, California Joseph J. Pinzone, MD Assistant Professor of Medicine Department of Internal Medicine George Washington University School of Medicine and Health Sciences Washington, DC Mark R. Pittelkow, MD Professor of Dermatology, Biochemistry, and Molecular Biology Mayo Medical School Consultant, Department of Dermatology Mayo Clinic Rochester, Minnesota Stephen R. Plymate, MD Research Professor of Medicine University of Washington School of Medicine Veterans Affairs Puget Sound Health Care System Seattle, Washington Ke-Nan Qin, MD Fellow of Pediatric Endocrinology Department of Pediatrics University of Chicago Pritzker School of Medicine University of Chicago Childrens Hospital Chicago, Illinois Ralph Rabkin, MB, Bch, MD Professor of Medicine and Nephrology Department of Medicine Stanford University School of Medicine Stanford, California Veterans Affairs Palo Alto Health Care System Palo Alto, California Miriam T. Rademaker, PhD Professor of Medicine University of Otago Christchurch School of Medicine

Christchurch, New Zealand Lawrence G. Raisz, MD Professor of Medicine Department of Endocrinology University of Connecticut School of Medicine University of Connecticut Health Center Farmington, Connecticut Lawrence I. Rand, MD Clinical Assistant Professor of Ophthalmology Harvard Medical School Boston, Massachusetts Ch. V. Rao, PhD Professor and Director Department of Obstetrics and Gynecology University of Louisville School of Medicine Louisville, Kentucky Robert E. Ratner, MD Associate Clinical Professor of Medicine George Washington University School of Medicine and Health Sciences Director, MedStar Clinical Research Institute Washington, DC Gerald M. Reaven, MD Professor of Medicine Stanford University School of Medicine Stanford, California Robert W. Rebar, MD Professor Department of Obstetrics and Gynecology University of Cincinnati College of Medicine Chief, Obstetrics and Gynecology University Hospital Cincinnati, Ohio Associate Executive Director American Society for Reproductive Medicine Birmingham, Alabama Robert S. Redman, DDS, MSD, PhD Chief, Oral Diagnosis Section, Dental Service Veterans Affairs Medical Center Washington, DC Clinical Associate Professor Department of Oral and Maxillofacial Pathology University of Maryland School of Medicine Baltimore College of Dental Surgery Baltimore, Maryland H. Lester Reed, MD Clinical Professor of Medicine University of Auckland Faculty of Medical and Health Sciences Middlemore Hospital Auckland, New Zealand Domenico C. Regoli, MD Professor Emeritus Department of Pharmacology Universite de Sherbrooke Faculte de Medecine Sherbrooke, Quebec Canada Jens F. Rehfeld, MD , DSc Professor of Clinical Biochemistry University of Copenhagen Faculty of Health Sciences Copenhagen University Hospital Copenhagen, Denmark Robert L. Reid, MD , FRCS(C) Professor Department of Obstetrics and Gynaecology Queens University School of Medicine Faculty of Health Sciences Kingston General Hospital Kingston, Ontario Canada Russel J. Reiter, PhD Professor of Neuroendocrinology Department of Cellular and Structural Biology University of Texas Medical School at San Antonio University Health Center San Antonio, Texas Matthew D. Ringel, MD Assistant Professor of Medicine Uniformed Services University of the Health Sciences F. Edward Hbert School of Medicine Bethesda, Maryland Assistant Clinical Professor of Medicine George Washington University School of Medicine and Health Sciences Section of Endocrinology Washington Hospital Center

Washington, DC Antonio Rivera, MD Fellow, Department of Medicine Section of Renal Diseases and Hypertension George Washington University Medical Center Washington, DC Gary L. Robertson, MD Professor of Medicine and Neurology Department of Endocrinology Northwestern University Medical School Chicago, Illinois R. Paul Robertson, MD Professor of Medicine and Pharmacology Scientific Director, Pacific Northwest Research Institute Seattle, Washington Simon P. Robins, PhD, Dsc Head, Skeletal Research Unit Rowett Research Institute Aberdeen, Scotland Alan D. Rogol, MD , PhD Professor of Clinical Pediatrics Department of Pediatrics University of Virginia School of Medicine University of Virginia Medical Center Charlottesville, Virginia Clinical Professor of Internal Medicine Virginia Commonwealth University School of Medicine Richmond, Virginia Prashant K. Rohatgi, MB, MD Professor of Medicine George Washington University School of Medicine and Health Sciences Veterans Affairs Medical Center Washington, DC Mikael Rrth, MD Professor of Clinical Oncology University of Copenhagen Faculty of Health Sciences Rigshospitalet Copenhagen, Denmark Robert L. Rosenfield, MD Professor of Pediatrics and Medicine Department of Pediatric Endocrinology University of Chicago Pritzker School of Medicine Chicago, Illinois Robert K. Rude, MD Professor of Medicine University of Southern California School of Medicine Los Angeles, California Neil B. Ruderman, MD , DPhil Professor Department of Medicine and Physiology Boston University School of Medicine Boston, Massachusetts James W. Russell, MD Assistant Professor Department of Neurology University of Michigan Medical School Ann Arbor GRECC Ann Arbor, Michigan Lester B. Salans, MD Adjunct Professor The Rockefeller University Clinical Professor of Medicine Mt. Sinai School of Medicine New York, New York Salil D. Sarkar Department of Radiology SUNY Health Sciences Center at Brooklyn College of Medicine Brooklyn, New York David H. Sarne, MD Associate Professor of Medicine Department of Internal Medicine University of Illinois at Chicago College of Medicine Chicago, Illinois Ernst J. Schaefer, MD Professor of Medicine Lipid Division Tufts University School of Medicine New England Medical Center Boston, Massachusetts Isaac Schiff, MD

Joe Vincent Meigs Professor of Gynecology Department of Obstetrics and Gynecology Harvard Medical School Massachusetts General Hospital Boston, Massachusetts R. Neil Schimke, MD Professor of Medicine and Pediatrics Chief, Division of Endocrinology and Genetics University of Kansas School of Medicine Kansas City, Kansas James R. Schreiber, MD Elaine and Mitchell Yanow Professor and Head Department of Obstetrics and Gynecology Washington University School of Medicine St. Louis, Missouri David E. Schteingart, MD Professor of Internal Medicine Division of Endocrinology and Metabolism University of Michigan Medical School Ann Arbor, Michigan Ellen W. Seely, MD Assistant Professor of Medicine Director of Clinical Research Endocrine-Hypertension Division Harvard Medical School Brigham and Womens Hospital Boston, Massachusetts Markus J. Seibel, MD , PD Associate Professor of Medicine Division of Endocrinology and Metabolism University of Heidelberg Medical School Heidelberg, Germany Elizabeth Shane, MD Professor of Medicine Columbia University College of Physicians and Surgeons New York, New York Lawrence E. Shapiro, MD Clinical Professor of Medicine SUNY at Stony Brook School of Medicine Health Sciences Center Stony Brook, New York Director, Division of Endocrinology Winthrop University Hospital Mineola, New York Meeta Sharma, MBBS, MD Assistant Director, Diabetes Team Division of Endocrinology Georgetown University School of Medicine MedStar Diabetes Institute Washington Hospital Center Washington, DC R. Michael Siatkowski, MD Associate Professor of Ophthalmology Dean A. McGee Eye Institute Oklahoma City, Oklahoma Omega L. Silva, MD Professor Emeritus of Medicine George Washington University School of Medicine and Health Sciences Washington, DC Shonni J. Silverberg, MD Associate Professor of Medicine Columbia University College of Physicians and Surgeons New York, New York Joe Leigh Simpson, MD Ernst W. Bertner Chairman and Professor Department of Obstetrics and Gynecology Baylor College of Medicine Houston, Texas Ethel S. Siris, MD Madeline C. Stabile Professor of Clinical Medicine Department of Medicine Columbia University College of Physicians and Surgeons New York, New York Glen W. Sizemore, MD Professor of Medicine Division of Endocrinology and Metabolism Loyola University of Chicago Stritch School of Medicine Maywood, Illinois Niels E. Skakkebaek, MD Professor of Growth and Reproduction University of Copenhagen Faculty of Health Sciences Rigshospitalet

Copenhagen, Denmark Celia D. Sladek, PhD Professor and Acting Chair Department of Physiology and Biophysics Finch University of Health Sciences Chicago Medical School North Chicago, Illinois John R. Sladek, Jr., PhD Professor and Chairman Department of Neuroscience Finch University of Health Sciences Chicago Medical School North Chicago, Illinois Eduardo Slatopolsky, MD Renal Division Washington University School of Medicine St. Louis, Missouri Robert C. Smallridge, MD Professor of Medicine Mayo Medical School Chair, Endocrine Division Mayo Clinic Jacksonville, Florida Robert J. Smith, MD Professor of Medicine Chief of Endocrinology Brown University School of Medicine Director, Hallett Diabetes Center Rhode Island Hospital Providence, Rhode Island Richard H. Snider, Jr., PhD Chief Chemist Endocrinology Research Laboratory Veterans Affairs Medical Center Washington, DC Phyllis W. Speiser, MD Professor of Clinical Pediatrics Department of Pediatrics New York University School of Medicine New York, New York North Shore University Hospital Manhasset, New York Harvey J. Stern, MD , PhD Genetics and IVF Institute Fairfax, Virginia Martin J. Stevens, MD Associate Professor of Internal Medicine University of Michigan Medical School Ann Arbor, Michigan Andrew F. Stewart, MD Professor of Medicine Chief, Division of Endocrinology University of Pittsburgh School of Medicine Pittsburgh, Pennsylvania Elizabeth A. Streeten, MD Clinical Assistant Professor of Medicine Department of Endocrinology, Diabetes, and Obesity University of Maryland School of Medicine Baltimore, Maryland Gordon J. Strewler, MD Professor of Medicine Department of Medical Service Harvard Medical School Boston, Massachusetts Veterans Affairs Boston Healthcare System West Roxbury, Massachusetts Martin I. Surks, MD Professor of Medicine and Pathology Department of Medicine Albert Einstein College of Medicine of Yeshiva University Montefiore Medical Center Bronx, New York Arthur L. M. Swislocki, MD Associate Professor of Medicine Department of Internal Medicine University of California, Davis, School of Medicine Davis, California Veterans Affairs Outpatient Clinic Martinez, California Shahrad Taheri, MSc, MB, MRCP Wellcome Trust Research Fellow Division of Endocrinology and Metabolism University of London Imperial College School of Medicine

Hammersmith Hospital London, England Robert J. Tanenberg, MD , FACP Professor of Medicine Section of Endocrinology and Metabolism Brody School of Medicine East Carolina University School of Medicine Greenville, North Carolina Kamal Thapar, MD Assistant Professor of Neurosurgery University of Toronto Faculty of Medicine Toronto Western Hospital, University Health Toronto, Ontario Canada Ramesh K. Thapar, MD Senior Resident Department of Psychiatry University of Maryland School of Medicine Baltimore, Maryland Michael A. Thomas, MD Associate Professor Department of Clinical Obstetrics and Gynecology University of Cincinnati College of Medicine Cincinnati, Ohio Christopher J. Thompson, MB, ChB, MD , FRCPI Consultant Physician and Endocrinologist Department of Endocrinology Royal College of Surgeons in Ireland Beaumont Hospital Dublin, Ireland Keith Tornheim, PhD Associate Professor of Biochemistry Boston University School of Medicine Boston, Massachusetts David J. Torpy, MBBS, PhD, FRACP Senior Lecturer Department of Medicine University of Queensland Faculty of Health Sciences Brisbane, Australia Carmelita U. Tuazon, MD , MPH Professor of Medicine George Washington University School of Medicine and Health Sciences Washington, DC Catherine Tuck, MD Assistant Professor of Medicine Columbia University College of Physicians and Surgeons New York, New York Michael L. Tuck, MD Professor of Medicine University of California, Los Angeles, UCLA School of Medicine Los Angeles, California Veterans Affairs Medical Center, Sepulveda Sepulveda, California Stephen Jon Usala, MD , PhD Clinical Associate Professor Department of Medicine Texas Tech University Health Sciences Center School of Medicine Amarillo, Texas Eve Van Cauter, PhD Professor of Medicine University of Chicago Pritzker School of Medicine Chicago, Illinois Greet H. Van Den Berghe, MD , PhD Associate Professor of Intensive Care Medicine Catholic University of Leuven Leuven, Belgium David H. Van Thiel, MD Director of Transplantation Loyola University of Chicago Stritch School of Medicine Loyola University Medical Center Liver Transplant Office Maywood, Illinois Joseph G. Verbalis, MD Professor of Medicine and Physiology Georgetown University School of Medicine Washington, DC Robert Volp, MD , FRCP(C), MACP, FRCP (Edin & Lord) Professor Emeritus Department of Medicine

University of Toronto Faculty of Medicine Toronto, Ontario Canada Steven G. Waguespack, MD Fellow, Adult and Pediatric Endocrinology Departments of Medicine and Pediatrics Division of Endocrinology Indiana University School of Medicine Riley Childrens Hospital Indianapolis, Indiana Brian Walsh, MD Director, Menopause Center Department of Obstetrics and Gynecology Harvard Medical School Brigham and Womens Hospital Boston, Massachusetts David O. Walterhouse, MD Assistant Professor of Pediatrics Northwestern University Medical School Childrens Memorial Hospital Chicago, Illinois Emily C. Walvoord, MD Senior Fellow Department of Pediatric Endocrinology and Diabetology Indiana University School of Medicine Riley Hospital for Children Indianapolis, Indiana James H. Warram, MD , ScD Senior Investigator Section on Genetics and Epidemiology Research Division Harvard Medical School Joslin Diabetes Center Boston, Massachusetts Michelle P. Warren, MD Professor of Obstetrics and Gynecology and Medicine Wyeth Ayerst Professor of Womens Health Columbia University College of Physicians and Surgeons New York, New York Leonard Wartofsky, MD , MPH, MACP Clinical Professor of Medicine Georgetown University School of Medicine Clinical Professor of Medicine George Washington University School of Medicine and Health Sciences Chair, Department of Medicine Washington Hospital Center Clinical Professor of Medicine Howard University College of Medicine Washington, DC Professor of Medicine and Physiology Uniformed Services University of the Health Sciences F. Edward Hbert School of Medicine Bethesda, Maryland Stephen I. Wasserman, MD Helen M. Ranney Professor of Medicine Chair, Department of Medicine University of California, San Diego, School of Medicine La Jolla, California Colleen Weber, RN Pediatric Endocrine Nurse Cardinal Glennon Childrens Hospital St. Louis, Missouri Anthony Peter Weetman, MD , FRCP, DSc Professor of Medicine University Department of Clinical Sciences University of Sheffield School of Medicine Northern General Hospital Sheffield, England Gordon C. Weir, MD Professor of Medicine Research Division Harvard Medical School Joslin Diabetes Center Boston, Massachusetts Laura S. Welch, MD Director, Occupational and Environmental Medicine Georgetown University School of Medicine Washington Hospital Center Washington, DC Samuel A. Wells, Jr., MD Professor of Surgery Washington University School of Medicine St. Louis, Missouri

Jon C. White, MD , FACS Director of Surgical Intensive Care Department of Surgery Veterans Affairs Medical Center Associate Professor of Surgery George Washington University School of Medicine and Health Sciences Washington, DC Perrin C. White, MD Professor of Pediatrics University of Texas Southwestern Medical Center at Dallas Southwestern Medical School Dallas, Texas Michael P. Whyte, MD Medical-Scientific Director Department of Metabolic and Molecular Research Professor of Medicine, Pediatrics, and Genetics Division of Bone and Mineral Diseases Washington University School of Medicine BarnesJewish Hospital St. Louis, Missouri Gordon H. Williams, MD Professor of Medicine Harvard Medical School Chair, Endocrine-Hypertension Division Brigham and Womens Hospital Boston, Massachusetts Stephen J. Winters, MD Professor of Medicine Chief, Division of Endocrinology and Metabolism University of Louisville School of Medicine Louisville, Kentucky Joseph I. Wolfsdorf, MB, BCh Associate Professor of Pediatrics Department of Medicine Division of Endocrinology Harvard Medical School Childrens Hospital National Medical Center Boston, Massachusetts I-Tien Yeh, MD Associate Professor Department of Pathology University of Texas Medical School at San Antonio University Health Center San Antonio, Texas Paul M. Yen, MD Chief, Molecular Regulation and Neuroendocrinology Clinical Endocrinology Branch National Institute of Diabetes and Digestive and Kidney Disease National Institutes of Health Bethesda, Maryland James E. Zadina, PhD Professor of Medicine Tulane University School of Medicine Director, Neuroscience Laboratory Department of Research Veterans Affairs Medical Center New Orleans, Louisiana Gary P. Zaloga, MA, MD Director of Critical Care Medicine Department of Medicine Georgetown University School of Medicine Washington Hospital Center Washington, DC Charles Zaloudek, MD Professor Department of Pathology University of California, San Francisco, School of Medicine San Francisco, California Carol Zapalowski, MD , PhD Colorado Center for Bone Research Lakewood, Colorado Thomas R. Ziegler, MD Associate Professor of Medicine Division of Endocrinology/Metabolism Emory University School of Medicine Atlanta, Georgia Michael Zinger, MD Clinical Instructor Department of Obstetrics and Gynecology Division of Reproductive Endocrinology University of Cincinnati College of Medicine Cincinnati, Ohio

EDITORS
EDITOR Kenneth L. Becker, MD , PhD Professor of Medicine Professor of Physiology and Experimental Medicine Director of Endocrinology and Metabolism George Washington University School of Medicine and Health Sciences Veterans Affairs Medical Center Washington, DC ASSOCIATE EDITORS John P. Bilezikian, MD Professor of Medicine and Pharmacology Department of Medicine Columbia University College of Physicians and Surgeons New York, New York William J. Bremner, MD , PhD Robert G. Petersdorf Professor and Chairman Department of Medicine University of Washington School of Medicine Seattle, Washington Wellington Hung, MD , PhD Professor Emeritus of Pediatrics Georgetown University School of Medicine Professorial Lecturer in Pediatrics George Washington University School of Medicine and Health Sciences Washington, DC C. Ronald Kahn, MD Mary K. Iacocca Professor of Medicine Harvard Medical School President and Director, Research Division Joslin Diabetes Center Boston, Massachusetts D. Lynn Loriaux, MD , PhD Professor and Chair Department of Medicine Oregon Health Sciences University School of Medicine Portland, Oregon Eric S. Nyln, MD Associate Professor of Medicine Department of Endocrinology George Washington University School of Medicine and Health Sciences Veterans Affairs Medical Center Washington, DC Robert W. Rebar, MD Professor Department of Obstetrics and Gynecology University of Cincinnati College of Medicine Chief, Obstetrics and Gynecology University Hospital Cincinnati, Ohio Associate Executive Director American Society for Reproductive Medicine Birmingham, Alabama Gary L. Robertson, MD Professor of Medicine and Neurology Department of Endocrinology Northwestern University Medical School Chicago, Illinois Richard H. Snider, Jr., PhD Chief Chemist Endocrinology Research Laboratory Veterans Affairs Medical Center Washington, DC Leonard Wartofsky, MD , MPH, MACP Clinical Professor of Medicine Georgetown University School of Medicine Clinical Professor of Medicine George Washington University School of Medicine and Health Sciences Chair, Department of Medicine Washington Hospital Center Clinical Professor of Medicine Howard University College of Medicine Washington, DC Professor of Medicine and Physiology Uniformed Services University of the Health Sciences F. Edward Hbert School of Medicine Bethesda, Maryland

CHAPTER 1 ENDOCRINOLOGY AND THE ENDOCRINE PATIENT Principles and Practice of Endocrinology and Metabolism

CHAPTER 1 ENDOCRINOLOGY AND THE ENDOCRINE PATIENT

KENNETH L. BECKER, ERIC S. NYLN, AND RICHARD H. SNIDER, J R. Definitions Role of the Endocrine System Hormones Chemical Classification Sources, Controls, and Functions Transport Types of Secretory Transport Overlap of Exocrine and Endocrine Types of Secretion Tyranny of Hormone Terminology Endocrine System Interaction with all Body Systems Genetics and Endocrinology Normal and Abnormal Expression or Modulation of the Hormonal Message and its Metabolic Effect The Endocrine Patient Frequency of Endocrine Disorders Cost of Endocrine Disorders Factors that Influence Test Results Reliability of the Laboratory Determination Determination of Abnormal Test Results Risks of Endocrine Testing Cost and Practicability of Endocrine Testing Conclusion Chapter References

DEFINITIONS
Endocrinology is the study of communication and control within a living organism by means of chemical messengers that are synthesized in whole or in part by that organism. Metabolism, which is an integral part of the science of endocrinology, is the study of the biochemical control mechanisms that occur within living organisms. The term includes such diverse activities as gene expression; biosynthetic pathways and their enzymatic catalysis; the modification, transformation, and degradation of biologic substances; the biochemical mediation of the actions and interactions of such substances; and the means for obtaining, storing, and mobilizing energy. The chemical messengers of endocrinology are the hormones, endogenous informational molecules that are involved in both intracellular and extracellular communication.

ROLE OF THE ENDOCRINE SYSTEM

The mammalian organism, including the human, is multicellular and highly specialized with regard to sustaining life and reproductive processes. Reproduction requires gametogenesis, fertilization, and implantation. Subsequently, the new intrauterine conception must undergo cell proliferation, organogenesis, and differentiation into a male or female. After parturition, the newborn must grow and mature sexually, so that the cycle may be repeated. To a considerable extent, the endocrine system influences or controls all of these processes. Hormones participate in all physiologic functions, such as muscular activity, respiration, digestion, hematopoiesis, sense organ function, thought, mood, and behavior. The overall purpose of the coordinating, regulating, integrating, stimulating, suppressing, and modulating effects of the many components of the endocrine system is homeostasis. The maintenance of a healthy optimal internal milieu in the presence of a continuously changing and sometimes threatening external environment is termed allostasis.

HORMONES
CHEMICAL CLASSIFICATION Most hormones can be classified into one of several chemical categories: amino-acid derivatives (e.g., tryptophan serotonin and melatonin; tyrosine dopamine, norepinephrine, epinephrine, triiodothyronine, and thyroxine; L -glutamic acid g-aminobutyric acid; histidine histamine), peptides or polypeptides (e.g., thyrotropin-releasing hormone, insulin, growth hormone, nerve growth factor), steroids (e.g., progesterone, androgens, estrogens, corticosteroids, vitamin D and its metabolites), and fatty acid derivatives (e.g., prostaglandins, leukotrienes, thromboxanes). SOURCES, CONTROLS, AND FUNCTIONS Previously hormones were thought to be synthesized and secreted predominantly by anatomically discrete and circumscribed glandular structures, called ductless glands (e.g., pituitary, thyroid, adrenals, gonads). However, many microscopic organoid-like groups of cells and innumerable other cells of the body contain and secrete hormones (see Chap. 175). The classic glands of endocrinology have lost their exclusivity, and although they are important on physiologic and pathologic levels, the widespread secretion of hormones throughout the body by nonglandular tissues is of equal importance. Most hormones are known to have multiple sources. Moreover, the physiologic stimuli that release these hormones are often found to differ according to their locale. The response to a secreted hormone is not stereotyped but varies according to the nature and location of the target cells or tissues. TRANSPORT TYPES OF SECRETORY TRANSPORT Hormones have various means of reaching target cells. In the early decades of the development of the field of endocrinology, hormones were conceived to be substances that traveled to distal sites through the blood. This is accomplished by release into the extracellular spaces and subsequent entrance into blood vessels by way of capillary fenestrations. The most appropriate term for such blood-bone communication is hemocrine (Fig. 1-1).

FIGURE 1-1. Different types of hormonal communication detailed in the chapter. The darkened areas on the cell membrane represent receptors. (H, hormone.) See text for explanations.

Several alternative means of hormonal communication exist, however. Paracrine communication involves the extrusion of hormonal contents into the surrounding interstitial spaces; the hormone then interacts with receptors on nearby cells (see Fig. 1-1 and Chap. 4 and Chap. 175).1 Direct paracrine transfer of cytoplasmic messenger molecules into adjacent cells may occur through specialized gap junctions (i.e., intercrine secretion).2 Unlike hemocrine secretion, in which the hormonal secretion is diluted within the circulatory system, paracrine secretion delivers a very high concentration of hormone to its target site. Juxtacrine communication occurs when the messenger molecule does not traverse a fluid phase to reach another cell, but, instead, remains associated with the plasma membrane of the signaling cell while acting directly on an immediately adjacent receptor cell (e.g., intercellular signaling that is adhesion dependent and occurs between endothelial cells and leukocytes and transforming growth factor-a in human endometrium).3,4 Hormones may be secreted and subsequently interact with the same cell that released the substance; this process is autocrine secretion (see Fig. 1-1).5 The secreted hormone stimulates, suppresses, or otherwise modulates the activity of the secreting cell. Autocrine secretion is a form of self-regulation of a cell by its own product. When peptide hormones or other neurotransmitters or neuromodulators are produced by neurons, the term neurocrine secretion is used (see Fig. 1-1).6 This specialized form of paracrine release may be synaptic (i.e., the messenger traverses a structured synaptic space) or nonsynaptic (i.e., the messenger is carried to its local or distal site of action by way of the extracellular fluid or the blood). Nonsynaptic neurocrine secretion has also been called neurosecretion. An example of neurosecretion is the release of vasopressin and oxytocin into the circulatory system by nervous tissue of the pituitary (see Chap. 25). Several peptides and amines are secreted into the luminal aspect of the gut (e.g., gastrin, somatostatin, luteinizing hormone releasing hormone, calcitonin, secretin, vasoactive intestinal peptide, serotonin, substance P).7 This process may be called solinocrine secretion (see Fig. 1-1), from the Greek word for a hollow tube. Solinocrine secretion also occurs into the bronchi, the urogenital tract, and other ductal structures.8 Commonly, the same hormone can be transported by more than one of these means.9 Extracellular transportation may not always be necessary for hormones to exert their effects. For example, some known hormonal secretions that are transported by one or more of these mechanisms are also found in extremely low concentrations within the cytoplasm of many cells. In such circumstances, these hormones do not appear to be localized to identifiable secretion granules and probably act primarily within the cell. This phenomenon may be called intracrine secretion. As shown in Figure 1-1, the process comprising uptake of a hormone precursor H1 and intracellular conversion into H2 (e.g., estrogens) or H3 (e.g., androgens) and subsequent binding and nuclear action is also a form of intracrine communication. OVERLAP OF EXOCRINE AND ENDOCRINE TYPES OF SECRETION Classically, an exocrine gland is a specialized structure that secretes its products at an external or internal surface (e.g., sweat glands, sebaceous glands, salivary glands, oxyntic or gastric glands, pancreatic exocrine glandular system, prostate gland). An exocrine gland may be unicellular (e.g., mucous or goblet cells of the epithelium of mucous membranes) or multicellular (e.g., salivary glands). Many multicellular exocrine glands possess a structured histologic organization that is suited to the production and delivery of secretions that are produced in relatively large quantities. A specialized excretory duct or system of ducts usually constitutes an intrinsic part of the gland. Some exocrine glandular cells secrete their substances by means of destruction of the cells themselves (i.e., holocrine secretion); an example is the sebaceous glands. Other exocrine glandular cells secrete their substances by way of the loss of a portion of the apical cytoplasm along with the material being secreted (i.e., apocrine secretion); an example is the apocrine sweat glands. Alternatively, in many forms of exocrine secretion, the secretory cells release their products through the cell membrane, and the cell remains intact (i.e., merocrine secretion); an example is the salivary glands. The constituents of some exocrine glands, particularly those opening on the external surface of the body, sometimes function as pheromones, which are chemical substances that act on other members of the species.10 Many exocrine glands contain cells of the diffuse neuroendocrine system (see Chap. 175) and neurons; both cell types secrete peptide hormones. Peptide hormones, steroids, and prostaglandins are found in all exocrine secretions (e.g., sweat, saliva, milk, bile, seminal fluid; see Chap. 106).11,12,13 and 14 Although they usually are not directly produced in such glands, thyroid and steroid hormones are found in exocrine secretions as well.15,16,17 and 18 The preferred approach is to view the term exocrine as a histologic-anatomic entity and not as a term that is meant to be antithetical to or to contrast with the term endocrine. Endocrinologists are concerned clinically and experimentally with all means of hormonal communication. The word endocrine is best used in a global sense, indicating any and all means of communication by messenger molecules.

TYRANNY OF HORMONE TERMINOLOGY

Hormones usually are named at the time of their discovery. Sometimes, the names are based on the locations where they were first found or on their presumed effects. However, with time, other locations and other effects are discovered, and these new locations or effects often are more physiologically relevant than the initial findings. Hormonal names are often overly restrictive, confusing, or misleading. In many instances, such hormonal names have become inappropriate. For example, atrial natriuretic hormone is present in the brain, hypothalamus, pituitary, autonomic ganglia, and lungs as well as atrium, and it has effects other than natriuresis (see Chap. 178). Gastrin-releasing peptide is found in semen, far from the site of gastrin release. Somatostatin, which was found in the hypothalamus and named for its inhibitory effect on growth hormone, occurs in many other locations and has multiple other functions (see Chap. 169). Calcitonin, which initially was thought to play an important role in regulating serum calcium and was named accordingly, appears to exert many other effects, and its influence on serum calcium may be quite minor (see Chap. 53). Growth hormonereleasing hormone and arginine vasopressin are found in the testis, where effects on growth hormone release or on the renal tubular reabsorption of water are most unlikely. Vasoactive intestinal peptide is found in multiple tissues other than the intestines (see Chap. 182). Insulin, named for the pancreatic islets, is found in the brain and elsewhere.19 Prostaglandins have effects that are far more widespread than those exerted in the secretions of the prostate, from which their name derives (see Chap. 172). The endocrine lexicon also contains substances called hormones that are not hormones. In the human, melanocyte-stimulating hormone (MSH) is not a functional hormone, but it comprises amino-acid sequences within the proopiomelanocortin (POMC) molecule: a-MSH within the adrenocorticotropic hormone (ACTH) moiety, b-MSH within d-lipotropin, and d-MSH within the N-terminal fragment of POMC (see Chap. 14). Numerous peptide hormones exist that, because of their effects on DNA synthesis, cell growth, and cell proliferation, have been called growth factors and cytokines (see Chap. 173 and Chap. 174). These substances, which act locally and at a distance, often do not have the sharply delineated target cell selectivity that was attributed to them when they first were discovered. Their terminology also is confusing and often misleading. Aside from occasional readjustments of hormonal nomenclature, no facile solution appears to exist to the quandary of terminology, other than an awareness of the pitfalls into which the terms may lead us.

ENDOCRINE SYSTEM INTERACTION WITH ALL BODY SYSTEMS

Although speaking in terms of the cardiovascular, respiratory, gastrointestinal, and nervous systems is convenient, the endocrine system anatomically and functionally overlaps with all body systems (see Part X). Extensive overlap is found between the endocrine system and the nervous system (see Chap. 175 and Chap. 176). Hormonal peptides are synthesized in the cell bodies of neurons, are transported along axons to nerve terminals, and are released at the nerve endings. Within these neurons, they coexist with classic neurotransmitters and often are coreleased with them. These substances play a role in neuromodulation or neurosecretion by means of the extracellular fluid. The nerves in which peptide hormones appear to play a role in the transmission of information are called peptidergic nerves.20 It is the ample similarity of ultrastructure, histochemistry, and hormonal contents of nerve cells and of many peptide-secreting endocrine cells that has led to the concept of the diffuse neuroendocrine system.

GENETICS AND ENDOCRINOLOGY

The rapid application of new discoveries and new techniques in genetics has revolutionized medicine, including the field of endocrinology. DNA probes have been targeted to selected genes, and the chromosomal locations of genes related to many hormones and their receptors have been determined. A complete map of the human genome is gradually emerging.21 This approach has led to new knowledge about hormone biosynthesis and has provided important information concerning species differences and evolution. The elucidation of the chromosomal loci for genes controlling the biosynthesis of hormone receptors should provide insights into the

physiologic effects of hormones. Clinically, these techniques have potential significance as a diagnostic aid in evaluating afflicted patients, a means of identifying asymptomatic heterozygotes, and a method for identification of unborn individuals at risk (i.e., prenatal diagnosis; see Chap. 240). Delineation of processes of genetic expression is revealing the mechanisms of hormonal disease (e.g., obesity22) and also may lead to gene therapy for some forms of endocrine illness or humoral-mediated conditions.23

NORMAL AND ABNORMAL EXPRESSION OR MODULATION OF THE HORMONAL MESSAGE AND ITS METABOLIC EFFECT
A sophisticated and faultless machinery is required for appropriate hormonal expression. The hormonal messenger is subject to modifications that may occur anywhere from its initial synthesis to its final arrival at its target site. Subsequently, the expression of the message at this site (i.e., its action) may also be modified (see Chap. 4). The modulations or alterations of the hormonal message or its final action may be physiologic or pathologic. Table 1-1 summarizes some of the normal and abnormal modulations of a hormonal message and its subsequent metabolic effects.

TABLE 1-1. Modulation of the Hormone Message and Its Subsequent Physiologic or Pathologic Metabolic Effects

On a physiologic level, the first steps in the genetic ordering of hormonal synthesis, the subsequent posttranslational processing of the hormone, the postsecretory extracellular transport, the receptor mediation of the hormone and subsequent transduction, and the inactivation and clearance of the hormone all contribute to expressing, diversifying, focalizing, and specifying the hormonal message and its ultimate action. On a pathologic level, all of these steps are subject to malfunction, causing disease. Our increased knowledge of endocrine systems has forced us to rethink many traditional concepts. To dispel some common misconceptions, listing several nots of endocrinology may be worthwhile (Table 1-2).

TABLE 1-2. Several Nots of Modern Endocrinology

THE ENDOCRINE PATIENT

FREQUENCY OF ENDOCRINE DISORDERS In a survey of the subspecialty problems seen by endocrinologists, the six most common, in order of frequency, were found to be diabetes mellitus, thyrotoxicosis, hypothyroidism, nontoxic nodular goiter, diseases of the pituitary gland, and diseases of the adrenal gland. Some conditions seen by endocrinologists are infrequent or rare (e.g., congenital adrenal hyperplasia, pseudohypoparathyroidism), whereas others are relatively common (e.g., Graves disease, Hashimoto thyroiditis), and some are among the most prevalent diseases in general practice (e.g., diabetes mellitus, obesity, hyperlipoproteinemia, osteoporosis, Paget disease). The third most common medical problem encountered by general practitioners is diabetes mellitus, and the tenth most frequent problem is obesity.66 Of the total deaths in the United States (i.e., both sexes, all races, and all ages combined), diabetes mellitus is the seventh most common cause. The most common cause of death (heart disease) and the third most common (cerebrovascular accidents) are greatly influenced by metabolic conditions such as diabetes mellitus and hyperlipemia.67 COST OF ENDOCRINE DISORDERS The frequency and morbidity of endocrine diseases such as osteoporosis, obesity, hypothyroidism, and hyperthyroidism, and the grave consequences of other endocrine disorders such as Cushing syndrome and Addison disease demonstrate that the expense to society is considerable. In the case of diabetes, the health care expenditure is staggering. Approximately 10.3 million people have diabetes in the U.S., and an estimated 5.4 million have undiagnosed diabetes. Direct medical expenses attributed to diabetes total $44.1 billion. The total annual medical expenses of people with diabetes average $10,071 per capita, as compared to $2,669 for persons without diabetes.68 Interestingly, these expenses may be less if the appropriate specialties are involved in the care.69 FACTORS THAT INFLUENCE TEST RESULTS In clinical medicine, hormonal concentrations usually are ascertained from two of the most easily obtained sources: blood and urine. The diagnosis of an endocrinopathy often depends on the demonstration of increased or decreased levels of these blood or urine constituents. However, several factors must be kept in mind when interpreting a result that appears to be abnormal. These may include age, gender, time of day, exercise, posture, emotional state, hepatic and renal status, presence of other illness, and concomitant drug therapy (see Chap. 237 and Chap. 239). RELIABILITY OF THE LABORATORY DETERMINATION The practice of clinical endocrinology far from a large medical center was previously hindered by the difficulty in obtaining blood and urine tests essential for appropriate diagnosis and follow-up care. However, accurate and rapid analyses now are provided by commercial laboratories. Nevertheless, wherever performed, some tests are unreliable because of methodologic difficulties. Other tests may be difficult to interpret because of a particular susceptibility to alteration by physiologic or pharmacologic factors (e.g., plasma catecholamines; see Chap. 86). Although many tests are sensitive and specific, they all have innate interassay and intraassay variations that may be particularly misleading when a given value is close to the clinical medical decision point (see Chap. 237). Some laboratory differences are due to hormone heterogeneity (e.g., growth hormone has several isoforms, which bind differently to growth hormonebinding proteins).70 DETERMINATION OF ABNORMAL TEST RESULTS Not uncommonly, the intellectual or commercial enthusiasm engendered by a new diagnostic procedure of presumed importance is found to be unjustified, because the

test was based on an invalid premise, because too few of ill patients were studied, because normative data to establish reference values were insufficient, or because subsequent studies were not confirmatory (see Chap. 237 and Chap. 241). The increased sophistication of medical testing has made the physician and the patient aware of the presence of abnormalities that may be harmless: physiologic deviations from that which is most common, or pathologic entities that commonly remain asymptomatic. Such findings may cause considerable worry, lead to the expense and risk of further diagnostic procedures, and even cause needless therapeutic intervention. Some abnormalities are the result of methods of imaging. For example, sonography of the thyroid may demonstrate the presence of small nodules within the thyroid gland of a person without any palpable abnormality of that region of the gland; most such microlesions are benign or behave as if they were. Another abnormality revealed by imaging is the occasional heterogeneous appearance of a normal pituitary gland on a computed tomography (CT) examination. Intermingled CT-lucent and CT-dense areas are seen on the scan, and such nonhomogeneous areas may be confused with a microadenoma.71,72 The increasing use of magnetic resonance imaging (MRI) of the brain may reveal a bona fide asymptomatic microadenoma of the pituitary gland, but extensive endocrine workup often reveals many such lesions to be nonfunctional. They occur in as much as 10% of the normal population.73 Rathke cleft cysts of the anterior sella turcica or the anterior suprasellar cistern often are seen by MRI.74 Although an occasional patient may have a large and symptomatic lesion,75 most of these lesions are small and asymptomatic. During MRI or CT examination of the brain, the examiner often incidentally encounters a primary empty sella, an extension of the subarachnoid space into the sella turcica with a resultant flattening of the pituitary gland in a patient without any pituitary lesion or any prior surgery of that region (see Chap. 11). Although some of these patients may be symptomatic, most have no associated symptoms or hormonal deficit. Another, albeit rare, lesion of the pituitary region seen on MRI is a sellar spine. This asymptomatic anatomic variant is an osseous spine arising in the midline from the dorsum sella that protrudes into the pituitary fossa; it may be an ossified remnant of the cephalic tip of the notochord.76 MRI or CT scanning of the abdomen may reveal the presence of harmless morphologic variations of the adrenal gland (i.e., incidentalomas) that sometimes leads to unnecessary surgery.77 (See Chap. 84.) RISKS OF ENDOCRINE TESTING Endocrine testing is not always benign. Many procedures can cause mild to marked side effects.78,79,80 and 81 Other diagnostic maneuvers, particularly angiography, may result in severe illness.82 The expected benefit of any procedure that is contemplated for a patient clearly should be greater than the risk. COST AND PRACTICABILITY OF ENDOCRINE TESTING In addition to being aware of the many factors that influence hormonal values, the limitations of laboratory determinations, and the potential risks of some of these procedures, the endocrinologist must be aware of their expense, particularly because medical costs have increased at an annual rate that is almost twice the rate of overall inflation during the last several years. A hypertensive patient with hypokalemia who is taking neither diuretics nor laxatives should undergo studies of the renin-angiotensin-aldosterone system, and appropriate pharmacologic or dietary manipulations of sodium balance should be instituted (see Chap. 90). But what should be done with the hypertensive patient who is normokalemic? Occasionally, such a person may have an aldosteronoma.83 Should such normokalemic patients be studied? Similarly, should the approximately 25 million hypertensive patients in the United States undergo urinary collections for determinations of catecholamine metabolites to find the rare patient with pheochromocytoma? In the context of the individual physician-patient relationship, the answers to such questions may not be difficult, but they become more controversial when placed within the framework of fiscal guidelines.

CONCLUSION
The complexity of the endocrine system presents a profound intellectual challenge. The macrosystem of endocrine glands secretes its hormones under the influence of other gland-based releasing factors or neural influences or both. The very act of secretion alters subsequent secretion by means of feedback controls (see Chap. 5). Superimposed on this already complex arrangement, the microsystem of dispersed, somewhat independent, but overlapping units throughout the body, as well as the continuous modulation of the receptors for the secreted hormones, allow general or focal actions that are coordinated with other body functions, tempered to the occasion, and appropriate to the needs of the individual. That such a complex system may go awry and that a dysfunction may have a considerable impact on the patient is not surprising. Because endocrinology and metabolism are broad subjects that incorporate much, if not all, of normal body functions and disease states, they defy easy categorization. However, these enormous complexities, rather than deterring the clinician, researcher, or student, should provide a stimulus to probe deeper into areas difficult to understand and should hasten the eventual application of new developments to patient care. CHAPTER REFERENCES
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. Hofbauer LC, Khosla S, Dunstan CR, et al. The roles of osteoprotegerin and osteoprotegerin ligand in the paracrine regulation of bone resorption. J Bone Mineral Res 2000; 15:2. Usadel H, Bornstein SR, Ehrhart-Bornstein M, et al. Gap junctions in the adrenal cortex. Horm Metab Res 1993; 25:653. Zimmerman GA, Lorant DE, McIntyre TM, Prescott SM. Juxtacrine intercellular signaling: another way to do it. Am J Respir Cell Mol Biol 1993; 9:573. Bush MR, Mele JM, Couchman GM, Walmer DK. Evidence of juxtacrine signaling for transforming growth factor alpha in human endometrium. Biol Reprod 1998; 59:1522. Hashimoto K, Higashiyama S, Asada H, et al. Heparin-binding epidermal growth factor-like growth factor is an autocrine growth factor for human keratinocytes. J Biol Chem 1994; 269:20060. Gouin FJ. Morphology, histology and evolution of Myriapoda and insects. 3. The nervous system and the neurocrine systems. Fortschr Zool 1965; 17:189. Uvnas-Wallensten K. Luminal secretion of gut peptides. Clin Gastroenterol 1980; 9:545. Van Minnen J. Production and exocrine secretion of LHRH-like material by the male rat reproductive tract. Peptides 1988; 9:515. Becker KL. The coming of age of a bronchial epithelial cell. Am J Respir Crit Care Med 1994; 149:183. Cohn BA. In search of human skin pheromones. Arch Dermatol 1994; 130:1048. Corps AN, Brown KD, Rees LH, et al. The insulin-like growth factor I content in human milk increases between early and full lactation. J Clin Endocrinol Metab 1988; 67:25. Hammami MM, Haq A, Al-Sedairy S. The level of endothelin-like immunoreactivity in seminal fluid correlates positively with semen volume and negatively with plasma gonadotropin levels. Clin Endocrinol 1994; 40:361. Schmidt NA. Salivary cortisol testing in children. Issues Compr Pediatr Nurs 1998; 20:183. Voss HF. Saliva as a fluid for measurement of estriol levels. Am J Obstet Gynecol 1999; 180:S226. Nizankowska B, Abramowicz T, Korezowski R, Rusin J. Triiodothyronine and thyroxine in human, cow's and formula milk. Exp Clin Endocrinol 1988; 91:116. Langer P, Moravec R, Ohradka B, Foldes O. Iodothyronines in human bile. Endocrinol Exp 1988; 22:35. Shieh CC, Chang SC, Tzeng CR, et al. Measurement of testosterone in seminal plasma, saliva and serum by solid-phase enzyme immunoassay. Andrologia 1987; 19:614. O'Rorke A, Kane MM, Gosling JP, et al. Development and validation of a monoclonal antibody enzyme assay for measuring progesterone in saliva. Clin Chem 1994; 40:454. Hendricke SA, Roth J, Rishi S, Becker KL. Insulin in the nervous system. In: Krieger DT, Bronnstein MJ, Martin J, eds. Brain peptides. New York: John Wiley & Sons, 1983:403. Bean AJ, Zhang X, Hokfelt T. Peptide secretion: what do we know? FASEB J 1994; 8:630. Zabarovsky ER, Allikmets R, Kholodnyuk I, et al. Construction of representative NOTI linking libraries specific for the total human genome and chromosome 3. Genomics 1994; 20:312. Perusse L, Chagnon YC, Weisnagel J, Bouchard C. The human obesity gene map: the 1998 update. Obes Res 1999; 7:111. Moldawer LL, Edwards PD, Minter RM, et al. Application of gene therapy to acute inflammatory disease. Shock 1999; 12:83. Atkinson J, Martin R. Mutations to nonsense codons in human genetic disease: implications for gene therapy by nonsense suppressor tRNAs. Nucleic Acids Res 1994; 22:1327. Miller WL. Molecular biology of steroid hormone synthesis. Endocr Rev 1988; 9:295. Oohoshi H, Ohgawara H, Nanjo K, et al. Familial hyperproinsulinemia associated with NIDDM. Diabetes Care 1993; 16:1340. Abraham MR, Jahingir A, Alekseev AE, Terzic A. Channelopathies of inwardly rectifying potassium channels. FASEB J 1999; 13:1901. Goksel DL, Fischbach K, Duggirala R, et al. Variant in sulfonylurea receptor-1 gene is associated with high insulin concentrations in nondiabetic Mexican Americans: SUR-1 gene variant and hyperinsulinemia. Hum Genet 1998; 103:280. Duquesnoy P, Sobrier ML, Duriez B, et al. A single amino acid substitution in the exoplasmic domain of the human growth hormone GH receptor confers familial GH resistance (Laron syndrome) with positive GH-binding activity by abolishing receptor homodimerization. EMBO J 1994; 13:1386. Rosenfeld RG, Rosenbloom AL, Guevara-Aguirre J. Growth hormone (GH) insensitivity due to primary GH receptor deficiency. Endocr Rev 1994; 15:369. Haas M, Forbush B III. The Na +-K+-Cl- cotransporters. J Bioenerg Biomembr 1998; 30:161. Torpy DJ, Chrousos GP. Hyper- and hypoaldosteronism. Vitam Horm 1999; 57:177. Glaser T, Lewis WH, Brung GH, et al. The beta subunit of follicle stimulating hormone is deleted in patients with aniridia and Wilms' tumor allowing a further definition of the WAGR locus. Nature 1986; 321:882. Sergi C, Serpi M, Mller-Navia J, et al. CATCH 22 syndrome: report of 7 infants with follow-up data and review of the recent advancements in the genetic knowledge of the locus 22q11. Pathologica 1999; 91:166. Stojdl DF, Bell JC. SR protein kinases: the splice of life. Biochem Cell Biol 1999; 77:293.

35a. Franklyn JA, Sheppard MC. Hormonal control of gene expression. Clin Endocrinol (Oxf) 1988; 29:337.

36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83.

Izumo S, Mahdavi V. Thyroid hormone receptor a isoforms generated by alternative splicing differentially activate myosin HC gene transcription. Nature 1988; 334:539. Koenig RJ, Lazar MA, Hodin RA, et al. Inhibition of thyroid hormone action by a non-hormone binding c-erbA protein generated by alternative mRNA splicing. Nature 1989; 337:659. Jung LJ, Kreiner T, Scheller RH. Prohormone structure governs proteolytic processing and sorting in the Golgi complex. Recent Prog Horm Res 1993; 48:415. Seidah NG, Day R, Marcinkiewicz M, Chrtien M. Mammalian paired basic amino acid convertases of prohormones and proproteins. Annals N Y Acad Sci 1993; 680:135. Belgorosky A, Rivarola MA. Progressive increase in non sex hormonebinding globulin-bound testosterone and estradiol from infancy to late puberty in girls. J Clin Endocrinol Metab 1988; 67:234. Tang KT, Yang HJ, Choo KB, et al. A point mutation in the albumin gene in a Chinese patient with familial dysalbuminemic hyperthyroxinemia. Eur J Endocrinol 1999; 141:374. Dahl KD, Bicsak TA, Hsueh AJW. Naturally occurring antihormones: secretion of FSH antagonists by women treated with a GnRH analog. Science 1988; 239:72. Wang C, Dahl KD, Leung A, et al. Serum bioactive follicle-stimulating hormone in men with idiopathic azoospermia and oligospermia. J Clin Endocrinol Metab 1987; 65:629. Bach JF. Antireceptor or antihormone autoimmunity and its relationship with the idiotype network. Adv Nephrol 1987; 16:251. Palmer JP, Asplin CM, Clemons P, et al. Insulin antibodies in insulindependent diabetics before insulin treatment. Science 1983; 222:1337. Yu L, Robles DT, Abiru N, et al. Early expression of antiinsulin autoantibodies of humans and the NOD mouse: evidence for early determination of subsequent diabetes. Proc Natl Acad Sci U S A 2000; 97:1701. Di Paolo S, Georgino R. Insulin resistance and hypoglycemia in a patient with systemic lupus erythematosus: description of antiinsulin receptor antibodies that enhance insulin binding and inhibit insulin action. J Clin Endocrinol Metab 1991; 73:650. Auclair M, Vigaroux C, Desbois-Mouthon C, et al. Antiinsulin receptor autoantibodies associate with insulin receptor substrate-1 and -2 and cause severe cell resistance to both insulin and insulin-like growth factor-I. J Clin Endocrinol Metab 1999; 84:3197. Chiovato L, Santini F, Vitti P, et al. Appearance of thyroid stimulating antibody and Graves disease after radioiodine therapy for toxic nodular goitre. Clin Endocrinol 1994; 40:803. Liberman UA, Eil C, Marx SJ. Clinical features of hereditary resistance to 1,25-dihydroxyvitamin D (hereditary hypocalcemic vitamin D resistant rickets type II). Adv Exp Med Biol 1986; 196:391. Chrousos GP, Detera-Wadleigh SD, Karl M. Syndromes of glucocorticoid resistance. Ann Intern Med 1993; 119:1113. Zennoro MC, Borensztein P, Seubrier F, et al. The enigma of pseudohypoaldosteronism. Steroids 1994; 59:96. Zoppi S, Wilson CM, Harbison MD, et al. Complete testicular feminization caused by an amino-terminal truncation of the androgen receptor with downstream initiation. J Clin Invest 1993; 91:1105. Talbert LM, Raj MH, Hammond MG, Greer T. Endocrine and immunologic ovary syndrome. Fertil Steril 1984; 42:7411. Fraser IS, Russell P, Greco S, Robertson DM. Resistant ovary syndrome and premature ovarian failure in young women with galactosemia. Clin Reprod Fertil 1986; 4:133. Suzuki Y, Hashimoto N, Shimada F, et al. Defects in insulin binding and receptor kinase in cells from a human with type A insulin resistance and from her family. Diabetologia 1991; 34:86. Bichet DG, Razi M, Lonergan M, et al. Hemodynamic and coagulation responses to 1-desamino (8-D-arginine) vasopressin in patients with congenital nephrogenic diabetes insipidus. N Engl J Med 1988; 318:881. Singer I, Forrest JN Jr. Drug-induced states of nephrogenic diabetes insipidus. Kidney Int 1976; 10:82. Jung Cy, Lee W. Glucose transporters and insulin action: some insights into diabetes management. Arch Pharm Res 1999; 22:329. Visser TJ, Kaptein E, Terpstra OT, Krenning EP. Deiodination of thyroid hormone by human liver. J Clin Endocrinol Metab 1988; 67:17. Bunnett NW. Postsecretory metabolism of peptides. Am Rev Respir Dis 1987; 136:S27. Roupas P, Herington AC. Receptor-mediated endocytosis and degradative processing of growth hormone by rat adipocytes in primary culture. Endocrinology 1987; 120:2158. Benzi L, Ceechetti P, Ciccarone A, et al. Insulin degradation in vitro and in vivo: a comparative study in men. Evidence that immunoprecipitable, partially rebindable degradation products are released from cells and circulate in blood. Diabetes 1994; 43:297. Yamaguchi T, Fukase M, Kido H, et al. Meprin is predominantly involved in parathyroid hormone degradation by the microvillar membranes of rat kidney. Life Sci 1994; 54:381. Ferrari P, Lovati E, Frey FJ. The role of the II beta-hydroxysteroid dehydrogenase type 2 in human hypertension. J Hypertens 2000; 18:241. Hum DW, Belanger A, Levesque E, et al. Characterization of UDP-glucuro-nyltransferases active on steroid hormones. J Steroid Biochem Mol Biol 1999; 69:413. National Center for Health Statistics. Monthly Vital Statistics Report 1999; 47(19):1. American Diabetes Association. Economic consequences of diabetes mellitus in the U.S. in 1977. Diabetes Care 1998; 21:296. Levetan CS, Passaro MD, Jablonski KA, Ratner RE. Effect of physician specialty on outcomes in diabetic ketoacidosis. Diabetes Care 1999; 22:1790. Baumann G. Growth hormone heterogeneity in human pituitary and plasma. Horm Res 1999; 51(Suppl 1):2. Roppolo HMN, Latchaw RE, Meyer JD, Curtin HD. Normal pituitary gland: 1. Macroscopic anatomyCT correlation. Am J Neuroradiol 1983; 4:927. Tihansky DP, Crossen J, Markowitz H. Pseudotumor artifact of the dorsum sella in CT scanning. Comput Radiol 1987; 11:241. Hall WA, Luciano MG, Doppman JL, et al. Pituitary magnetic resonance imaging in normal human volunteers: occult adenomas in the general population. Ann Intern Med 1994; 120:817. Kucharczyk W, Peck WW, Kelly WM, et al. Rathke cleft cysts: CT, MR imaging, and pathologic features. Radiology 1987; 165:491. Hama S, Arita K, Tominago A. Symptomatic Rathke's cleft cyst coexisting with central diabetes insipidus and hypophysitis: case report. Endocr J 1999; 46:187. Fujisawa I, Asato R, Togashi K, et al. MR imaging of the sellar spine. J Comput Assist Tomogr 1988; 12:644. McGrath PC, Sloan DA, Schwartz RW, Kenady DE. Advances in the diagnosis and therapy of adrenal tumors. Curr Opin Oncol 1998; 10:52. Ratzmann GW, Zollner H. Hypomagnesemia and hypokalemia in the insulin hypoglycemia test. Z Gesamte Inn Med 1985; 40:567. Read RC, Doherty JE. Cardiovascular effects of induced insulin hypoglycemia in man during the Hollander test. Am J Surg 1972; 104:573. Sobel RJ, Ariad S. Adverse cardiovascular responses to thyrotropin-releasing hormone (200 micrograms) in cardiac patients. Isr J Med Sci 1987; 23:1107. Boice JD Jr. The danger of x-raysreal or apparent. N Engl J Med 1986; 315:828. Hash RB. Intravascular radiographic contrast media: issues for family physicians. J Am Board Fam Pract 1999; 12(1):32. Bravo EL, Tarazi RC, Dustan HP, et al. The changing clinical spectrum of primary aldosteronism. Am J Med 1983; 74:641.

CHAPTER 2 MOLECULAR BIOLOGY: PRESENT AND FUTURE Principles and Practice of Endocrinology and Metabolism

CHAPTER 2 MOLECULAR BIOLOGY: PRESENT AND FUTURE

MEHBOOB A. HUSSAIN AND JOEL F. HABENER Cloning of Genes Genomic Libraries and Gene Isolation Gene Amplification by Polymerase Chain Reaction Variations of Polymerase Chain Reaction Approaches to the Quantitative Assessment of Gene Expression Transcription Assays Messenger RNA Assays Protein Expression Assays DNA-Protein Interaction Assays Genetic Manipulations in Animals in Vivo Transgenic Approaches Conditional (Developmental) Interruption of Gene Expression Prospects for the Future for Conditional Transgene Expression Expressed Sequence Tags DNA Arrays for the Profiling of Gene Expression Oligonucleotide Arrays (Genomic and Expressed Sequence Tags) Complementary DNA Arrays (Specific Tissues) Strategies for Mapping Genes on Chromosomes Genetic Linkage Maps and Quantitative Trait Loci Restriction Enzymes and Chromosomal Mapping Physical Maps Separating Chromosomes Somatic Cell Hybridization Chromosome Walking (Positional Cloning) Candidate Gene Approach Future Prospects Human Genome Project Stem Cells Somatic Cell Cloning in Vivo Gene Knock-Out Libraries Gene Therapy: Vectors and Problems Chapter References

The beginnings of molecular biology as a distinct discipline occurred in the late 1940s and early 1950s with the recognition that polynucleotides were the repository of genetic information in the form of DNA and the transmitters of genetic information in the form of messenger RNA (mRNA), and that transfer RNAs are fundamental for the assembly of amino acids into proteins. Detailed descriptions of the historical developments of this modern era of molecular biology are provided in several books.1,2,3 and 4 These were exciting times, as understanding progressed rapidly from the discovery by Avery and Brundage that DNA was a genetic substance; Chargaff established that DNA is composed of four different deoxyribonucleotides (dATP, dGRP, dTTP, dCTP); Watson and Crick elucidated the double-helical structure of DNA; Jacob and Monod identified mRNA as the intermediary in the transfer of information encoded in DNA to the assembly of amino acids into proteins; Holly discovered transfer RNAs; and Nirenberg et al. discovered the genetic code (i.e., each of the 21 amino acids is specified by a triplet of nucleotides, or codons, within the mRNA to be translated into a protein). In the 1960s, several major discoveries paved the way for the development of recombinant DNA technology and genetic engineering. Two of the major breakthroughs that made this possible were the discoveries of reverse transcriptase5 and restriction endonucleases,6,7 and 8 and techniques for determining the precise sequence of nucleotides in DNA.9,10 Reverse transcriptase, which is found encoded in the RNA of certain tumor viruses, is the means by which the virus makes DNA copies of its RNA templates. It allows molecular biologists to copy mRNA into complementary DNA (cDNA), which is an essential step in the preparation of recombinant DNA for purposes of cloning. Another fundamental discovery was that of restriction endonucleases, enzymes that cut DNA at specific sequences, typically of 4 to 10 base pairs. The application of specific restriction endonucleases allows for the cleavage of DNAs at precise locations, a property that is critical for the engineering of DNA segments. A most critical and important discovery was the technologic methodology to determine the sequential order of nucleotides in DNA. Both chemical and enzymatic approaches were developed. Currently, the nucleotide sequences of DNAs are determined by sophisticated automated instruments using random enzymatic cleavages of DNAs labeled with fluorescent markers. By fortunate coincidence, research into the mechanisms by which bacteria become resistant to certain antibiotics led to the discovery of bacterial plasmids, which are viruses that live within bacteria and lend genetic information to the bacteria to ensure their survival. Plasmids faithfully replicate within bacteria. Importantly, plasmid DNA is relatively simple in structure and is amenable to genetic engineering by excision of DNA sequences and insertion of foreign DNA sequences, which will replicate within bacteria without interference by the host bacterium. These plasmids have become useful vehicles in which to express and amplify foreign DNA sequences.

CLONING OF GENES
Complementary DNA Libraries. The cloning of a particular expressed gene begins with the preparation and cloning of cDNAs from mRNAs of a particular cell (Fig. 2-1; Table 2-1) (for a more comprehensive description, see reference 11 and reference 12). The cDNAs are prepared by priming the reverse transcription of mRNAs, using reverse transcriptase and short oligonucleotide fragments of oligodeoxyribothymidine, which preferentially bind to the 3'-polyadenylate, or poly(A), tract that is characteristic of cellular mRNAs. Alternatively, random oligonucleotides of different base compositions may be used. Double-stranded DNA is then prepared from the single-stranded cDNA by using DNA polymerase, and the cDNAs are inserted into bacterial plasmids that have been cleaved at a single site with a restriction endonuclease. To ensure a reasonably high efficiency of insertion of the foreign DNA into the plasmids, cohesive, or sticky, ends are first prepared by adding short DNA sequences to the ends of the foreign DNA and to the plasmids. Vectors that are commonly used are derivatives of the plasmid pBR322, which was engineered specifically for the purposes of cloning DNA fragments (see Fig. 2-1). Foreign DNA is inserted into a unique site that is prepared by endonuclease cleaving of a desired site within a polylinker, multiple cloning site engineered into the plasmid. This site is often located within the gene that codes for bacterial b-galactosidase. The backbone plasmid also carries a gene for resistance to ampicillin or tetracycline. Thus, bacteria containing the plasmids can be selected by their resistance to ampicillin or tetracycline; those specifically containing DNA inserts can be selected by their inability to express b-galactosidase and to cleave b-galactopyranoside (blue-white screening).

FIGURE 2-1. An approach used in construction and molecular cloning of recombinant DNA. A, Preparation of double-stranded DNA from an mRNA template. The enzyme reverse transcriptase is used to reversetranscribe a single-stranded DNA copy complementary to the mRNA primed with an oligonucleotide of

polydeoxythymidylic acid hybridized to the poly(A) tract at the 3' end of mRNA. A complementary copy of the DNA strand is then prepared with DNA polymerase. Ends of double-stranded DNA are made flush by cleavage with the enzyme S1 nuclease, and homopolymer extensions of deoxycytidine are synthesized on 3' ends of DNA with the enzyme terminal transferase. Oligo(dC) homopolymer extensions form sticky ends for purposes of insertion of DNA into a linearized bacterial plasmid on which complementary oligo(dG) homopolymer extensions have been synthesized. B, Insertion of foreign DNA into a bacterial plasmid for molecular cloning. A bacterial plasmid, typically pBR322, that has been specifically engineered for purposes of cloning DNA is linearized by cleavage with restriction endonuclease Pst I. Poly(dG) homopolymer extensions are synthesized onto 3' ends of plasmid DNA. Foreign DNA with complementary poly(dC) homopolymer extensions is hybridized to and inserted into the plasmid. Recombinant plasmid DNA is transfected into susceptible host strains of bacteria, in which plasmid replicates apart from bacterial chromosomal DNA. Bacteria are then grown on a plate containing tetracycline. Colonies that are resistant to tetracycline are tested for sensitivity to ampicillin. Because native plasmids contain genes encoding resistance to both tetracycline and ampicillin and the gene encoding resistance to ampicillin is inactivated by insertion of a foreign DNA at the Pst I site, bacterial colonies harboring plasmids with DNA inserts are resistant to tetracycline and sensitive to ampicillin. Subsequent screening of tetracycline-resistant, ampicillin-sensitive clones containing specific DNA-inserted sequences is carried out by either DNA hybridization with labeled DNA probes or by other techniques such as hybridization arrest and cell-free translation.

TABLE 2-1. Approaches for the Selection of Cloned Complementary DNAs (cDNAs)

Hybridization Screening. The recombinant plasmids containing DNA sequences that are complementary to the specific mRNAs of interest are identified by hybridizing recombinant plasmids to the initial mRNA preparations used in the cloning. The hybrid-selected mRNA is subsequently eluted and translated in a cell-free system appropriate for the protein under study. Alternatively, specific inhibition of the translation of an mRNA can be used to identify the DNA of interest: DNA that is complementary to the mRNA being translated will bind the RNA, thus precluding translation and reducing the amount of the protein being synthesized. The initial techniques of hybridization selection and hybridization arrest, in which cell-free translation is used as the assay system, are now supplanted by hybridization of the bacterial colonies with synthetic oligonucleotide probes that are labeled with phosphorus-32 (32P). Mixtures of oligonucleotides in the range of 14 to 17 bases are prepared that are complementary to the nucleotide sequences predicted from the known amino-acid sequences of segments of the protein encoded by mRNA. Because of the degeneracy in the genetic code (there are 61 amino-acid codons and 20 amino acids), mixtures of from 24 to 48 oligonucleotides ordinarily represent all possible sequences complementary to a particular 14- to 17-base region of mRNA. Expression Screening. Later-generation cDNA libraries have been prepared in bacterial phages (l gt-11) or hybrids between plasmids and phages (phagemids), which have been engineered to allow the bacteria infected with the recombinant phages to translate mRNAs expressed from the cDNAs, and thereby to produce the protein products encoded by the cDNAs. The desired sequence of interest can be selected at the protein level by screening the library of bacterial clones with an antiserum directed to the protein. When the desired product is a DNA-binding protein, the library can be screened with a labeled DNA duplex containing copies of the target sequence to which the protein binds. Yeast Two-Site Interaction Trap. The cloning of cDNAs encoding proteins that interact with other known proteins can be accomplished using the yeast two-site interaction trap, which functions much as a bait and fish system. The bait is a cDNA encoding a known protein that is engineered to bind to an enhancer in the promoter of a gene that encodes a factor essential for the survival of a yeast cell. The sequences (fish) in the cDNA library are engineered with a strong transcriptional transactivation domain, such as that from the herpes simplex virus and yeast transcription factors VP16 or Gal-4, respectively. The occurrence of proteinprotein interactions between the bait and one of the fish activates the expression of the yeast survival gene, which thereby allows for the selection and cloning of the yeast cell that harbors the described cDNA sequence from the cDNA library. Rapid Amplification of Complementary DNA Ends. Most often cDNAs isolated by one or more of the approaches described above lack the complete sequence and are deficient in the 5' ends. The 5' sequences are determined by using the rapid amplification of cDNA ends (RACE) technique.

GENOMIC LIBRARIES AND GENE ISOLATION

Southern Blots and Hybridization Screening. The techniques used in the cloning of genomic DNA are similar to those used for cloning cDNA, except that the genomic sequences are longer than the cDNA sequences and different cloning vectors are required. The common vectors are derivatives of the bacteriophage l that can accommodate DNA fragments of 10 to 20 kilobases (kb). Certain hybrids of bacteriophages and plasmids, called cosmids, can accommodate inserts of DNA of up to 40 to 50 kb. Even larger segments of DNA up to 1 to 2 megabases (Mb) can be cloned and propagated in yeast and are called yeast artificial chromosomes (YACs). In the cloning of genomic DNA, restriction fragments are prepared by partial digestion of unsheared DNA with a restriction endonuclease that cleaves the DNA into many fragments. DNA fragments of proper size are prepared by fractionation on agarose gels and are ligated to the bacteriophage DNA. The fragments of DNA containing the desired sequences can be detected by hybridization of a membrane blot prepared from the gel with a 32P-labeled cDNA, a Southern blot. The recombinant DNA is mixed with bacteriophage proteins, which results in the production of viable phage particles. The recombinant bacteriophages are grown on agar plates covered with growing bacteria. Then the bacteria are infected by a phage particle, which lyses the bacteria to form visible plaques. Specific phage colonies are transferred by nitrocellulose filters and are hybridized by cDNA probes labeled with 32P, similar to a Southern blot. Libraries of genomic DNA fragments and tissue-specific cDNAs from various animal species cloned in plasmids and bacteriophages are available from a number of commercial laboratories. The development of yeast chromosomal libraries that harbor large segments (several megabases) of chromosomal DNA has markedly accelerated the generation of gene linkage maps. Enhancer Traps. One approach to identifying novel genes imbedded in the genome is to randomly insert a transcriptional reporter gene into chromosomal DNA that has been cleaved into 1- to 2-kb fragments by digestion with a restriction endonuclease. The family of ligated hybrid fragments is then cloned into plasmids that are individually introduced (transfected) into host cell lines (e.g., NIH or BHK fibroblasts). After the transfected cell lines are incubated with the cloned DNA fragments for 1 to 2 days, extracts are prepared from the cells and assayed for expression of the transcriptional reporter gene. Typical transcriptional reporter genes used are firefly luciferase, bacterial chloramphenicol acetyl transferase, or bacterial alkaline phosphatase. When, by chance, a transcriptional enhancer is encountered, as determined by the activation of the reporter gene, the particular cloned DNA fragment is sequenced and searched for transcribed exonic and/or intronic sequences of genes, many of which typically reside 100 to 1000 base pairs from the enhancer sequence. The transcribed sequences of genes usually, but not always, reside 3' (downstream) from enhancer sequences. Rapid Amplification of Genomic DNA Ends. The principle of rapid amplification of genomic DNA ends (RAGE) is similar to that of RACE previously described and allows for the identification of unknown DNA sequences in genomic DNA. Oligonucleotide primers (amplimers) are annealed to the test genomic DNA sample and extended on the genomic DNA template with DNA polymerase, and a second set of oligonucleotide primers is ligated to the extended ends. The extended DNA fragments are then amplified by polymerase chain reaction (see next section), isolated by electrophoresis on agarose gels, and sequenced.

GENE AMPLIFICATION BY POLYMERASE CHAIN REACTION

The development of the polymerase chain reaction PCR, a technique for the rapid amplification of specific DNA sequences, constituted a major technological breakthrough.13,14,15 and 16 This procedure relies on the unique properties of a thermally stable DNA polymerase (Taq polymerase) to allow for sequential annealing of small oligonucleotide primers that bracket a DNA sequence of interest; the result is successive synthesis of the DNA strands. Specific DNA sequences as short as 50 and as long as several thousand base pairs can be amplified over a million-fold in just a few hours by using an automated thermal cycler. The technique is so sensitive that DNA (genomic DNA or cDNA reverse-transcribed from RNA) from a single cell can be so amplified. Indeed, a sample containing only a single target DNA molecule

can be amplified. The applications of this technique are diverse. Not only is it possible to amplify and to clone rare sequences for detailed studies, but also the technique has applications in the fields of medical diagnosis and forensics. Scarce viruses can be detected in a drop of serum or urine or a single white blood cell. Genotyping can be done from a blood or semen stain, saliva, or a single hair. Paradoxically, a major drawback of PCR is its exquisite sensitivity, which leaves open the possibility of false-positive results because of minute contaminations of the samples being tested. Thus, extreme precautions must be taken to avoid the introduction of contaminants. PCR is carried out using DNA polymerase and oligonucleotide primers complementary to the two 3' borders of the duplex segment to be amplified. The objective of PCR is to copy the sequence of each strand between the regions at which the oligonucleotide primers anneal. Thus, after the primers are annealed to a denatured DNA containing the segment to be amplified, the primers are extended using DNA polymerase and the four deoxynucleotide triphosphates. Each primer is extended toward the other primer. The result is a double-stranded DNA (which itself is then denatured and annealed again with primer, and the DNA polymerase reaction is repeated). This cycle of steps (denaturation, annealing, and synthesis) may be repeated 60 times. At each cycle, the amount of duplex DNA segment doubles, because both new and old DNA molecules anneal to the primers and are copied. In principle (and virtually in practice), 2n copies (where n = number of cycles) of the duplex segment bordered by the primers are produced. The heat-stable polymerase isolated from thermophilic bacteria (Thermophilus aquaticus), Taq polymerase, allows multiple cycles to be carried out after a single addition of enzyme. The DNA, an excess of primer molecules, the deoxynucleotide triphosphates, and the polymerase are mixed together at the start. Cycle 1 is initiated by heating to a temperature adequate to assure DNA denaturation, followed by cooling to a temperature appropriate for primer annealing to the now-single strands of the template DNA. Thereafter, the temperature is adjusted for DNA synthesis (elongation) to occur. The subsequent cycles are initiated by again heating to the denaturation temperature. Thus, cycling can be automated by using a computer-controlled variable-temperature heating block. In addition to permitting automation, the use of the DNA polymerase of T. aquaticus has another advantage. The enzyme is most active between 70 and 75C. Base pairing between the oligonucleotide primers and the DNA is more specific at this temperature than at 37C, the optimal functioning temperature of Escherichia coli DNA polymerase. Consequently, the primers are less likely to anneal nonspecifically to unwanted DNA segments, especially when the entire genome is present in the target DNA. VARIATIONS OF POLYMERASE CHAIN REACTION Simple modifications of the PCR conditions can expand the opportunities of the PCR. For example, synthesizing oligonucleotide primers that recognize domains (motifs) shared by cDNAs and their respective protein products, and choosing less stringent annealing conditions for the primers, permit new sequences of yet unknown DNAs to be generated with PCR, ultimately resulting in the discovery of new cDNAs belonging to the same family. For example, the pancreatic B-cell transcription factor IDX-1 was identified by PCR using oligonucleotide primers that would anneal to sequences shared by the homeodomain transcription factor family. PCR primers can be modified in their sequence and thus are not completely complementary to the template DNA. The amplified PCR product then carries the sequence of the primer and not the original DNA sequence. This strategy can be used to insert mutations site-specifically into known DNA sequences.

APPROACHES TO THE QUANTITATIVE ASSESSMENT OF GENE EXPRESSION

TRANSCRIPTION ASSAYS Nuclear Run-On Assays. Several assays are available that provide an index of relative rates of gene transcription (Fig. 2-2). A simple, straightforward assay is the nuclear run-on assay in which nuclei are isolated from tissue culture cells and nascent RNA chains are allowed to continue to polymerize in the presence of radiolabeled deoxyribonucleotides in vitro. This assay has the advantage that it surveys the density of nascent transcripts made from the endogenous genes of cells and, on average, is a good measure of gene transcription rates in response to the existing environmental conditions in which the cultured cells are maintained. Newly synthesized RNA is applied (hybridized) to a nylon membrane on which a cDNA target complementary to the desired RNA has been adsorbed. Radiolabeled RNA hybridized to the cDNA is determined in a radiation counter.

FIGURE 2-2. Approaches to the quantitative assessment of gene expression. Shown are the various types of assays that can be used to examine regulation of gene expression at various levels. (mRNA, messenger RNA; RNase, ribonuclease; RT-PCR, reverse transcription polymerase chain reaction.)

Cell-Free In Vitro Systems. Rates of RNA synthesis can also be determined in broken cell or cell-free lysates to assess the relative strengths of different promoters. To restrict the newly synthesized radiolabeled RNA to a single size and, thus, to enable more ready detection by electrophoresis, a DNA template is used that does not contain guanine bases, called a G-free cassette. RNA synthesis is carried out in the absence of the guanine nucleotide. After synthesis of a specified length of RNA at the end of which guanine bases are encountered, RNA synthesis is terminated. Transfection of Promoter-Reporters in In Vivo Cell Culture. Many of the currently used assays of gene transcription employ promoter sequences fused to genes encoding proteins that can be quantitated by bioassays (e.g., bacterial chloramphenicol acetyl transferase, firefly luciferase, alkaline phosphatase, or green fluorescent protein). The hybrid DNAs, so called promoter-reporter DNAs, are introduced into tissue culture cells by one of several chemical methods (i.e., DNA adsorbed to calcium phosphate precipitates, diethylaminoethyl (DEAE)-dextran incorporated into liposomes, or human artificial chromosomes [Table 2-2]); or physical methods (i.e., electroporation, direct microinjection of DNA, or ballistic injection using a gene gun [Table 2-3]). After introduction of the reporter DNA into the cells, the transfected cells are incubated for a specified time under the desired experimental conditions, the cells are harvested, and extracts are prepared for assays of the reporter-specific enzymatic activity. By these transfection methods, cell-type specificity for the expression of genepromoter sequences can be determined by comparing promoter-reporter efficiencies in cells of different phenotypes. In addition, important transcriptional control sequences in the promoter can be mapped by DNA mutagenesis studies.

TABLE 2-2. Chemical Methods for Introducing Genes into Mammalian Cells

TABLE 2-3. Physical Methods for Introducing Genes into Mammalian Cells

Transfection of Transcription Factor Expression Vectors. An extension of the promoter-reporter transfection approach is to cotransfect recombinant expression plasmids encoding transcription factors that bind to control sequences in the promoter DNA and activate transcription of the reporter. By this approach, critical functional components of transcription factors and critical bases in DNA control sequences can be examined experimentally. Transgenic In Vivo Mouse Models. A method developed for examining specificity of tissue expression and efficiency of expression of promoter-reporter genes is their introduction into mice in vivo, using transgenic technology (see the section Genetic Manipulations in Animals In Vivo). Recombinant promoter-reporter genes are injected into the pronucleus of fertilized mouse ova and implanted into surrogate females. The tissues of transgenic neonatal mice are examined for the tissue distribution and relative strength of the expression of the reporter function. Commonly used reporter functions are the genes encoding either b-galactosidase or green fluorescent protein. MESSENGER RNA ASSAYS Northern Blot Hybridization. RNA blotting (Northern blotting) is analogous to DNA blotting (Southern blotting). RNA is separated according to size by electrophoresis through agarose gels. Generally, the electrophoresis is performed under conditions that denature the RNA so that the effects of RNA secondary structure on the electrophoretic mobility of the RNAs can be minimized. Alkaline conditions are unsuitable; therefore, agents such as glyoxal, formaldehyde, or urea are used. The size-separated RNA is transferred by blotting to an immobilizing membrane without disturbing the RNA distribution along the gel. A labeled DNA is then used as a probe to find the position on the blot of RNA molecules corresponding to the probe. The immobilized RNA is incubated with DNA under conditions allowing annealing of the DNA to the RNA on the immobilized matrix. After washing away excess and unspecifically annealed DNA, the matrix is exposed to an x-ray film to detect the position of the probe. RNA blotting allows the estimation of the size of the RNA that is being detected. In addition, the intensity of the band on an x-ray film indicates the abundance of the RNA in the cell or tissue from which the RNA was extracted. Solution Hybridization Ribonuclease Protection. To obtain more precise information on the amount of a specific RNA species in a certain cell or tissue, a single-stranded radioactive probe is generated that is complementary to a portion of the RNA being studied. An excess amount of this single-stranded probe is then mixed in solution with the total RNA of the cells or tissue being investigated. Digestion with ribonuclease of all single-stranded nucleic acids present after hybridization leaves the double-stranded species, consisting of the labeled probe annealed to its complementary RNA, in the solution. The contents of the solution are then size-separated on an electrophoretic gel, which is exposed to an x-ray film. Knowing the amount of input labeled single-stranded probe allows a quantification of the specific RNA present in the total RNA of the cells or tissue. In Situ Hybridization. In situ hybridization with labeled single-stranded probes onto tissues is, in principle, similar to the ribonuclease protection assay. Detection and determination of the location of a certain species of RNA within a tissue is possible. Reverse Transcription Polymerase Chain Reaction. Reverse transcription polymerase chain reaction (RT-PCR) can be used to quantitate the abundance of a specific RNA. This method is particularly practical when small amounts of tissue or cells are available to be analyzed. The RNA is reverse-transcribed to DNA with reverse transcriptase. The cDNA population is then subjected to PCR amplification with specific primers that recognize the cDNA in question. By choosing the number of PCR cycles within the linear range of product generated after each cycle (i.e., enough primers, nucleotides, and DNA polymerase in the reaction mixture for none of them to be the limiting factor of the reaction) and adding to the PCR reaction a defined amount of an artificial DNA template that is also recognized by the primer oli gonucleotides but yields a differentsized product, one can detect differences in abundance of cDNA (and hence RNA in the original sample) among two or more samples. Newer methods allow for an on-line monitoring of each PCR reaction of the product generated. This is achieved by using primer oligonucleotides that can be monitored during the PCR reaction cycles by external optical devices. Such on-line continuous monitoring allows the performance of PCR reactions without prior determination of the number of cycles required to keep the PCR reaction within the linear range of amplification. Continuous PCR monitoring provides immediate information on abundance of a given cDNA species in PCR reactions. Knowledge of the absolute amount of labeled oligonucleotide primer added to the PCR reaction at the start can be used to determine the exact amount of the PCR product generated. PROTEIN EXPRESSION ASSAYS Cell-Free Translation. A commonly used method to analyze proteins encoded by mRNA is to translate the mRNA in cell-free translation systems in vitro. By this method, proteins can be radioactively labeled to a high specific activity. The cell-free translation also provides the primary protein product, such as a proprotein or prohormone, encoded by the mRNA. Pulse and Pulse-Chase Labeling. Studies of protein syntheses can also be carried out in vivo by incubation of cultured cells or tissues with radioactive amino acids (pulse labeling). Posttranslational processing (e.g., enzymatic cleavages of prohormones) can be assessed by first incubating the cells or tissues for a short time with radioactive amino acids and then incubating them for an additional period with unlabeled amino acids (pulse-chase labeling). Western Immunoblot. Another approach to the analyses of particular cellular proteins is the Western immunoblot technique. Proteins in cell extracts are separated by electrophoresis on polyacrylamide or agarose gels and transferred to a nylon or nitrocellulose membrane, which is then treated with a solution containing specific antibodies to the protein of interest. The antibodies that are bound to the protein fixed to the membrane are detected by any one of several methods, such as secondary antibodies tagged with radioisotopes, fluorophores, or enzymes. Immunocytochemistry. A refinement of the Western immunoblot technique is the detection of specific proteins within cells by immunocytochemistry (immunohistochemistry). Cultured cells or tissue sections are fixed on microscope slides and treated with solutions containing specific antibodies. The antibodies that are bound to the proteins within the cells are detected with fluorescently tagged secondary antibodies or by an avidin-biotin complex. Immunocytochemistry is a powerful technique when used for the simultaneous detection of two or even three different proteins with examination by confocal microscopy. DNA-PROTEIN INTERACTION ASSAYS ELECTROPHORETIC MOBILITY GEL SHIFT AND SOUTHWESTERN BLOTS The binding of proteins such as transcription factors to DNA sequences is commonly done by two approaches: electrophoretic mobility shift assay (EMSA) and Southwestern blotting. Typically, EMSA consists of incubation of protein extracts with a radiolabeled DNA sequence or probe. The mixture is then analyzed by electrophoresis on a nondenaturing polyacrylamide gel, followed by autoradiography or autofluorography to evaluate the distribution of the radioactivity or fluorescence in the gel. Interactions of specific proteins with the DNA probe are manifested by a retardation of the electrophoretic migration of the labeled probe, or band shift. The EMSA technique can be extended to include antibodies to specific proteins in the incubation mixture. The interaction of a specific antibody with a protein bound to the DNA probe causes a further retardation of migration of the DNA-protein complex, leading to a super shift. PROTEIN-PROTEIN INTERACTION ASSAYS

A number of different assays are used to determine and evaluate protein-protein interactions. Two in vitro assays are coimmunoprecipitation and polyhistidine-tagged glutathione sulfonyl transferase (GST) pull-down. Two in vivo assays are the yeast and mammalian two-site interaction assays. Coimmunoprecipitation. The commonly used coimmunoprecipitation assay makes use of antisera to specific proteins. In circumstances in which two different proteins, A and B, associate with each other, an antiserum to protein A will immunoprecipitate not only protein A, but also protein B. Likewise, an antiserum to protein B will coimmunoprecipitate proteins B and A. In practice, the proteins under investigation are radiolabeled by synthesis in the presence of radioactive amino acids, either in cell-free transcriptiontranslation systems in vitro, or in cell culture systems in vivo. Coimmunoprecipitated proteins are detected by gel electrophoresis and autoradiography. Alternatively, the proteins so immunoprecipitated or coimmunoprecipitated can be assayed by Western immunoblot techniques. Glutathione Sulfonyl Transferase Pull-Down. GST is an enzyme that has a high affinity for its substrate, glutathione. This property of high-affinity interactions has been exploited to develop a cloning vector plasmid encoding GST and containing a polylinker site that allows for the insertion of coding sequences for any protein of interest. Thus, if protein A is believed to interact with protein B, the coding sequence for either protein A or protein B can be inserted into the GST vector. The GSTprotein A or B fusion protein is synthesized in large amounts by multiplication and expression of the plasmid vector in bacteria. The GST-fusion protein is then incubated with either labeled or unlabeled proteins in extracts of cells or nuclei. Proteins in the extracts bound to protein A or B in the GST-fusion protein are pulled down from the extracts by capturing the GST on glutathione-agarose beads. Proteins are released from the beads and analyzed by either gel electrophoresis and autoradiography (labeled proteins) or by Western immunoblot (unlabeled proteins). Similar methods using polyhistidine tag in place of GST are also used for pull-down experiments. Far Western Protein Blots. A variation on the Western blotting technique is the Far Western blot. In this technique, a radio-labeled or fluorescence-labeled known protein (instead of an antibody) is applied to a membrane to which proteins from an electrophoretic gel have been transferred. If the known protein binds to any one or more proteins on the membrane, it is detected as a labeled band by autoradiograph or autofluorography. Relatively strong protein-protein interactions are required for this approach to succeed. Yeast and Mammalian Cell Two-Site Interaction Traps. The two-site interaction traps are useful for demonstrating protein-protein interactions in vivo. The principle of the approach is that, when a specific protein-protein interaction occurs, it reconstitutes an active transcription factor which then activates the transcription of a reporter gene. The cells (yeast or mammalian) are programmed to constitutively express a strong DNA-binding domain, such as Gal-4, fused to the expression sequence of the selected protein, protein A (the bait). The cells are also programmed to express a transcriptional reporter (e.g., CAT or luciferase linked to a promoter) that has binding sites for Gal-4. Thus, protein A anchors to the DNA-binding site of the reporter promoter via the Gal-4 binding domain but does not activate transcription of the reporter gene, and no reporter function is expressed. Protein B, however, is expressed as a fusion protein with a strong transcriptional activator sequence (e.g., the transcriptional transactivation domain of Gal-4 or of VP16). This transcriptional activation domainprotein B fusion protein does not bind DNA, but when, or if, protein B physically interacts with (binds to) protein A, a fully active transcription factor is reconstituted, the promoter reporter gene is transcribed, and the reported function is expressed. The yeast two-hybrid system can be used to clone proteins that interact with a bait protein such as protein A. In this instance, a cDNA protein expression library is prepared or obtained that has all of the cDNAs of a given tissue fused to a coding sequence for a transactivation domain (e.g., VP16). Further, the reporter consists of a survival factor essential for the growth of the yeast cell. Thus, when a cDNA encodes a protein B (fish) that interacts with the bait protein A, the yeast cell expresses the survival protein and survives, whereas the other yeast cells die.

GENETIC MANIPULATIONS IN ANIMALS IN VIVO

TRANSGENIC APPROACHES To create transgenic mice, DNA is injected into the male pronucleus of one-cell mammalian embryos (fertilized ova) that are then allowed to develop by insertion into the reproductive tract of pseudopregnant foster mothers (Fig. 2-3A). The transgenic animals that develop from this procedure contain the foreign DNA integrated into one or more of the host chromosomes at an early stage of embryo development. As a consequence, the foreign DNA is generally transmitted to the germline, and, in a number of instances, the foreign genes are expressed. Because the foreign DNA is injected at the one-cell stage, a good chance exists that the DNA will be distributed among all the progeny cells as development proceeds. This situation provides an opportunity to analyze and compare the qualitative and quantitative efficiencies of expression of the genes among various organs. The technique is quite efficient; >50% of postinjection ova produce viable offspring, and, of these, ~10% efficiently carry the foreign genes. In the transgenic animals, the foreign genes can be passed on and expressed at high levels in subsequent generations of progeny.

FIGURE 2-3. Approaches for (A) the integration of a foreign gene into the germline of mice, and (B) disruption or knock-out of a specific gene. A, DNA containing a specific foreign gene is microinjected into the male pronucleus of fertilized ova obtained from the oviduct of a mouse. Ova are then implanted into the uterus of pseudopregnant surrogate mothers. Progeny are analyzed for the presence of foreign genes by hybridization with 32P-labeled DNA probe and DNA prepared from a piece of tail from a mouse, which has been immobilized on a nitrocellulose filter (tail blots). B, To create a knock-out of a gene, pluripotential embryonic stem (Es) cells are used in vitro to introduce an engineered plasmid DNA sequence that will recombine with a homologous gene that is targeted. The recombination excises a portion of the gene in the ES cells, rendering it inactive (no longer expressible). ES cells in which the homologous recombination occurred successfully are selected by a combined positive-negative drug selection. The engineered ES cells are injected into the blastocoele of 3.5-day blastocysts that are then implanted into the uterus of pseudopregnant mice. The offspring are both chimeric and germline for expression of the knock-out gene and must be cross-bred to homozygosity for the genotype of a complete knock-out of the gene that is targeted for disruption.

Transgenic approaches can also be used to prevent the development of the lineage of a particular cell phenotype or to impair the expression of a selected gene. A cell lineage can be ablated by targeting a microinjected DNA containing a subunit of the diphtheria toxin to a particular cell type, using a promoter sequence specifically expressed in that cell type. The diphtheria toxin subunit inhibits protein synthesis when expressed in a cell, thereby killing the cell. The expression of a particular gene can be impaired by similar cell promoter specific targeting of a DNA expression vector to a cell that produces an antisense mRNA to the mRNA expressed by the gene of interest. The antisense mRNA hybridizes to nuclear transcripts and processed mRNAs; this results in their degradation by double-stranded RNAspecific nucleases, thereby effectively attenuating the functional expression of the gene. The efficacy of the impairment of the mRNA can be enhanced by incorporating a ribozyme hammerhead sequence in the expressed antisense mRNA so as physically to cleave the mRNA to which it hybridizes. Another approach to producing a particular gene loss of function is to direct expression of a dominant negative protein (e.g., a receptor made deficient in intracellular signaling by an appropriate mutation, or a mutant transcription factor deficient in transactivation functions but sufficient for DNA binding). These dominant negative proteins compete for the essential functions of the wild-type proteins, resulting in a net loss of function. Another approach, termed targeted transgenesis, combines targeted homologous recombination in embryonic stem (ES) cells with gain-of-function transgenic approaches.11,12,17 This method allows for targeted integration of a single-copy transgene to a single desired locus in the genome and thereby avoids problems of random and multiple-copy integrations, which may compromise faithful expression of the transgene in the conventional approach. GENE ABLATION (KNOCK-OUTS) A major advance beyond the gain-of-function transgenic mouse technique has been the development of methods for producing loss of function by targeted disruption or replacement of genes. This approach uses the techniques of homologous recombination in cultured pluripotential ES cells, which are then injected into mouse

blastocysts and implanted into the uteri of pseudopregnant mice (Fig. 2-3B). The targeting vector contains a core replacement sequence consisting of an expressed-cell lethal-drug resistance gene (selectable marker) (e.g., neomycin [Pgk-neo]) flanked by sequences homologous to the targeted cellular gene, and a second selectable marker gene (e.g., thymidine kinase [pgk-tk]). The ES cells are transfected with the gene-specific targeting vector. Cells that take up vector DNA and in which homologous recombination occurs are selected by their resistance to neomycin (positive selection). To select against random integration, a susceptibility to killing by thymidine kinase (negative selection) is used; only homologous recombination in which the thymidine kinase gene has been lost will confer survival benefit. Because the ES cells are injected into multicellular 3.5-day blastocysts, many of the offspring are mosaics, but some are germline heterozygous for the recombined gene. F1 generation mice are then bred to homozygosity so as to manifest the phenotype of the gene knock-out. Using this approach of targeted gene disruption, literally thousands of knock-out mice have been created. Many of these knock-out mice are models for human genetic disorders (e.g., those of endocrine systems such as pancreatic agenesis [homeodomain protein IDX-1], familial hypocalciuric hypercalcemia [calcium receptor], intrauterine growth retardation [insulinlike growth factor-II receptor], salt-sensitive hypertension [atrial natriuretic peptide], and obesity [a3-adrenergic receptor]). CONDITIONAL (DEVELOPMENTAL) INTERRUPTION OF GENE EXPRESSION Although targeted transgenesis using chosen site integration and targeted disruption of genes has proven helpful in analyses of the functions of genes, conditionally to induce expression of transgenes or conditionally to inactivate a specific gene is useful. Early on, randomly integrated vectors for the expression of transgenes used the metallothionein promoter that is readily inducible by the administration of heavy metals to transgenic mice. Now techniques have been developed to conditionally inactivate targeted genes in a defined spatial and temporal pattern. Several approaches to achieve conditional gene inactivation have been developed. Two of these approaches are (a) the Cre recombinaseloxP system (Fig. 2-4)18 and (b) the tetracycline-inducible transactivator vector (tTA) system (Fig. 2-5).19 Occasionally, both of these systems have been used effectively to knock out (Cre-loxP) or to attenuate (reverse tTA) the expression of specific genes. Both the Cre-loxP and reverse tTA systems require the creation of two independent strains of transgenic mice, which are then crossed to produce double transgenic mice.

FIGURE 2-4. Schema of the Cre-loxP approach to conditionally knock out a specifically targeted gene in mice. A, The approach requires the creation of two separate strains of transgenic mice that are crossed to produce double transgenic mice to effect the conditional gene knock-outs. One mouse strain is created so as to replace the gene of interest by one that has been flanked by loxP recombination sequences (floxed), using targeted recombinational gene replacement in embryonic stem cells as illustrated in Figure 2-3B. The other mouse strain is a transgenic mouse in which the Cre recombinase enzyme expression vector is targeted to the tissue of interest using a tissue-specific promoter, such as the proinsulin gene promoter, to target and restrict expression to pancreatic B cells. B, A more detailed depiction of the strategy for preparation of the gene replacement by homologous recombination to generate mice with a floxed gene. This approach is similar to that described in Figure 2-3B to create knock-out mice.18

FIGURE 2-5. Diagram showing the approach to reversible conditional expression of a gene in mice, using a tetracycline-inducible gene expression system. A, As in the Cre-loxP system (see Fig. 2-4A), the tetracyclineinducible gene system requires the creation of two independent strains of transgenic mice. One strain of mice targets the expression of a specially engineered transcription factor (rtTA) to the tissue of interest, using a tissue-specific promoter (TSP). B, the rtTA transcription factor consists of a modification of the bacterial tetracycline-responsive repressor that has been genetically engineered so as to convert it into a transcriptional transactivator when exposed to tetracycline or one of its analogs. The other mouse strain is one in which a gene of interest is introduced, usually driven by a ubiquitous promoter such as a viral promoter (CMV, RSV) or an actin promoter. The gene of interest could be one encoding an antisense RNA to a messenger RNA of a protein that is to be knocked out. The creation of double transgenic mice then allows for the expression of the gene of interest in a specific tissue under the control of the induced tetracycline. (See text for more detailed description.57) (tet op, tetracycline resistance operon; P, promoter; As, antisense; TPE, tissue promoter element.)

CRE RECOMBINASELOXP SYSTEM The Cre-loxP approach is based on the Cre-loxP recombination system of bacteriophage P1 (see Fig. 2-4). This system is capable of mediating loxP sitespecific recombination in embryonal stem cells and in transgenic mice. Conditional targeting requires the generation of two mouse strains. One transgenic strain expresses the Cre recombinase under control of a promoter that is cell-type specific or developmental stage specific. The other strain is prepared by using ES cells to effect a replacement of the targeted gene with an exact copy that is flanked by loxP sequences required for recombination by the Cre recombinase. The recombined gene is said to be floxed. The presence of the loxP sites does not interfere with the functional expression of the gene and will be normally expressed in all of its usual tissues not coexpressing the Cre recombinase. In those tissues in which the Cre is expressed by virtue of its tissue-specific promoter, the target gene will be deleted by homologous recombination. Thus, the Cre-loxP system acts like a timer in which the events that are to take place are predetermined by the prior reprogramming of the genes: the target gene will be ablated during development where and when the promoter chosen to drive the expression of Cre is activated. Thus, a disadvantage of the Cre-loxP system is the lack of control over when the gene knock-out will take place, because it is preprogrammed in the system. Newer genetically engineered Cre derivatives allow for pharmacologic activation of the recombinant event. A potential advantage of the Cre-loxP system is that one can theoretically generate extensive collections of mice expressing the Cre recombinase specifically and individually in many different tissues so that these mice could be made commercially available to investigators. CONDITIONAL TETRACYCLINE-INDUCIBLE FORWARD AND REVERSE TRANSACTIVATOR VECTOR SYSTEMS The Cre-loxP system leads to the irreversible targeted disruption of a particular gene at the time that the promoter encoding the Cre recombinase is activated during development. Having available a system that can be reversibly activated at any time would be desirable. A system that holds promise in this regard is the tetracycline-inducible transactivator vector (forward or reverse tTA), which, in response to tetracycline, switches on a specific gene bearing a promoter containing the tetracycline-responsive operon (see Fig. 2-5). This system allows any recombinant gene marked by the presence of the tet operon to be turned on or off at will simply by the administration of a potent tetracycline analog to the transgenic mice. The vectors were engineered from the sequences of the E. coli bacterial tetracycline resistance operon (tet op), in which a repressor sits on the operon, keeping the resistance gene off. When tetracycline binds to the repressor, it is deactivated, falls off of the operon, and turns on the gene. First, the repressor was converted into an activator by fusing the DNA-binding domain to the potent activator sequence (VP16) of the herpes simplex virus. In this system, tetracycline turned off the activator (tet-off) and thereby caused failure of expression of target genes containing the tet operon binding sites for the repressor turned into an activator. This tTA system required the continued presence of tetracycline to keep the gene off and withdrawal of the tetracycline to turn on the gene, raising problems of long and variable clearance times for the drugs. Turning the gene on by administration of tetracycline (tet-on) would be preferable. Therefore, the tTA vector was reengineered to reverse the action of tetracycline: in the current vector system, the binding of tetracycline to the reverse

tTA enhances its binding to the tet operon. Theoretically, as the reverse tTA system now works, any gene can be reversibly turned on by the administration of tetracycline or one if its more potent analogs in the double transgenic mouse, which consists of a cross between a mouse that has the reverse tTA targeted to express in a specific tissue and a mouse that has a ubiquitously expressed transgene for any gene X under the control of the tet operon. The equivalent of gene knock-outs can be accomplished by constructing gene X in a context to express an antisense RNA containing a ribozyme sequence. When induced by tetracycline, antisense-ribozyme RNA binds to the mRNA expressed by gene X, cleaves it, and thereby functionally inactivates the gene. PROSPECTS FOR THE FUTURE FOR CONDITIONAL TRANSGENE EXPRESSION The availability of the Cre-loxP and the forward and reverse tTA systems now makes it feasible to combine their key features in the creation of triple transgenic mice so that a targeted recombinational disruption of a gene can be accomplished by the administration of tetracycline. The Cre recombinase could be placed under the control of a tissue-specific promoter containing the tet operon uniquely responsive to the presence of tTA and targeted to a specific tissue by standard pronuclear injection targeted transgenesis. A second transgenic mouse is created with a ubiquitously expressed promoter during the expression of the reverse tTA. In the third mouse, the gene desired to be deleted would be replaced with an appropriately floxed gene. The latter mouse would be prepared by implantation of recombinantly engineered ES cells into blastocysts. The administration of tetracycline to the triple transgenic mouse would induce the Cre recombinase in a tissue-specific manner, thus allowing temporal and spatial control of gene knock-outs. EXPRESSED SEQUENCE TAGS A very informative database of expressed sequence tags (ESTs) is being generated and placed in GenBank. Expressed sequence tags are prepared by random, single-pass sequencing of mRNAs from a repertoire of different tissues, mostly embryonic tissues (e.g., brain, eye, placenta, liver). Currently, the EST database contains ~50% of the estimated expressed genes in humans and mammals (70,00080,000). The EST database will become extremely valuable when the sequences of the human, rat, and mouse genomes are completed.

DNA ARRAYS FOR THE PROFILING OF GENE EXPRESSION

Two variants of DNA-array chip design exist.20,21 The first consists of cDNA (sequences unknown) immobilized to a solid surface such as glass and exposed to a set of labeled probes of known sequences, either separately or in a mixture of the probes. The second is an array of oligonucleotide probes (sequences known, based on either known genes in GenBank or ESTs) that are synthesized either in situ or by conventional synthesis followed by on-chip immobilization (Fig. 2-6). The array is exposed to labeled sample DNA (unknown sequence) and hybridized, and complementary sequences are determined.

FIGURE 2-6. Sample preparation and hybridization for oligonucleotide assay. A complementary DNA (cDNA) is transcribed in vitro to RNA, and then reverse-transcribed to cRNA. This material is fragmented and tagged with a fluorescent tag molecule (F). The fragments are hybridized to an array of oligonucleotides representing portions of DNA sequences of interest. After washing, hybridization of the cRNA probe is detected by localization of the fluorescent signals. (PCR, polymerase chain reaction.)

In cDNA chips, immobilized targets of single-stranded cDNAs prepared from a specific tissue are hybridized to single-stranded DNA fluorescent probes produced from total mRNAs to evaluate the expression levels of target genes. OLIGONUCLEOTIDE ARRAYS (GENOMIC AND EXPRESSED SEQUENCE TAGS) The oligonucleotide gene chip (1.28 1.28 cm2) consists of a solid-phase template (glass wafer) to which high-density arrays of oligonucleotides (distance between oligonucleotides of 100 ) are attached, with each probe in a predefined position in the array. Each gene EST is represented by 20 pairs of 25 base oligonucleotides from different parts of the gene (5' end, middle, and 3' end). The specificity of the detection method is controlled by the presence of single-base mismatch probes. Pairs of perfect and single-base mismatch probes corresponding to each target gene are synthesized on adjacent areas on the arrays. This is done to identify and subtract nonspecific background signals. The gene chip is sensitive enough to detect one to five transcripts per cell and is much more sensitive than the Northern blot technique. COMPLEMENTARY DNA ARRAYS (SPECIFIC TISSUES) Poly (A) mRNA is isolated from cells or tissue of interest, and synthesis of double-stranded cDNA is accomplished by reverse transcription of cDNA, followed by synthesis of double-stranded cDNA using DNA polymerase I. In vitro transcription of double-stranded cDNA to cRNA is accomplished using biotin-16-UTP and biotin-11-CTP for labeling and a T7 RNA polymerase as enzyme. This cRNA is used for hybridization with the gene chip. The gene chip is stained with R-physoerythrin streptavidin to detect biotin-labeled nucleotides, and different wash cycles are performed. Thereafter the gene chip is scanned digitally and analyzed by special software. (A grid is automatically placed over the scanned image of the probe array chip to demarcate the probe cells.) After grid alignment, the average intensity of each probe cell is calculated by the software, which then analyzes the patterns and generates a report. The applications of the gene chip include: 1. 2. 3. 4. Simultaneous analysis of temporal changes in gene expression of all known genes and ESTs. Sequencing of DNA. Large-scale detection of mutations and polymorphisms in specific genes (i.e., BRCA1, HIV-1, cystic fibrosis CFTR, b-thalassemia). Gene mapping by determining the order of overlapping clones.

Expensive equipment for generating and analyzing the data using genechips is required. When the cloning of all genes is completed (Human Genome Project), the gene chip will allow monitoring of the expression of all known genes in various situations.

STRATEGIES FOR MAPPING GENES ON CHROMOSOMES

GENETIC LINKAGE MAPS AND QUANTITATIVE TRAIT LOCI A genetic linkage map shows the relative locations of specific DNA markers along the chromosome.22,23,24,25,26 and 27 Any inherited physical or molecular characteristic that differs among individuals and is easily detectable in the laboratory is a potential genetic marker. Markers can be expressed DNA regions or DNA segments that have no known coding function, but whose inheritance pattern can be followed. DNA sequence differences (polymorphisms; i.e., nucleotide differences) are especially useful markers because they are plentiful and easy to characterize precisely. Markers must be polymorphic to be useful in mapping. Alternative DNA polymorphisms exist among individuals, even among members of a single family, so that they are detectable among different families. Polymorphisms are variations in DNA sequence in the genome that occur every 300 to 500 base pairs. Variations within protein-encoding exon sequences can lead to observable phenotypic changes (e.g., differences in eye color, blood type, and disease susceptibility). Most variations occur within introns and have little or no effect on the phenotype (unless they alter exonic splicing patterns), yet these polymorphisms in DNA sequence are detectable and can be used as markers. Examples of these types of markers are: (a) restriction fragment

length polymorphisms (RFLPs), which reflect sequence variations in DNA sites that are cleaved by specific DNA restriction enzymes; and (b) variable number of tandem repeat sequences (VNTRs), which are short repeated sequences that vary in the number of repeated units and, therefore, in length. The human genetic linkage map is constructed by observing how frequently any two polymorphic markers are inherited together. Two genetic markers that are in close proximity tend to be passed together from mother to child. During gametogenesis, homologous recombination events take place in the metaphase of the first meiotic step (meiotic recombination crossing-over). This may result in the separation of two markers that originally resided on the same chromosome. The closer the markers are to each other, the more tightly linked they are and the less likely that a recombination event will fall between and separate them. Recombination frequency provides an estimate of the distance between two markers. On the genetic map, distances between markers are measured in terms of centimorgans (cM), named after the American geneticist Thomas Hunt Morgan. Two markers are said to be 1 cM apart if they are separated by recombination 1% of the time. A genetic distance of 1 cM is roughly equal to a physical distance of 1 million base pairs of DNA (1 Mb). The current resolution of most human genetic map regions is approximately 10 Mb. An inherited disease can be located on the map by following the inheritance of a DNA marker present in affected individuals but absent in unaffected individuals, although the molecular basis of a disease or a trait may be unknown. Linkage studies have been used to identify the exact chromosomal location of several important genes associated with diseases, including cystic fibrosis, sickle cell disease, Tay-Sachs disease, fragile X syndrome, and myotonic dystrophy. RESTRICTION ENZYMES AND CHROMOSOMAL MAPPING The restriction endonucleases, which have been isolated from various bacteria, recognize short DNA sequences and cut DNA molecules at those specific sites. A natural biofunction of restriction endonucleases is to protect bacteria from viral infection or foreign DNA by destroying the alien DNA. Some restriction enzymes cut DNA very infrequently, generating a small number of very large fragments, whereas other restriction enzymes cut DNA more frequently, yielding many smaller fragments. Because hundreds of different restriction enzymes have been characterized, DNA can be cut into many differentsized fragments. PHYSICAL MAPS Different types of physical maps vary in their degree of resolution. The lowest resolution physical map is the chromosomal (cytogenetic) map, which is based on the distinctive banding pattern observed by light microscopy of stained chromosomes. A cDNA map shows the locations of expressed DNA (exons) on the chromosomal map. The more detailed cosmid contiguous DNA block (contig) map depicts the order of overlapping DNA fragments spanning the genome (see the section High-Resolution Physical Mapping). A macrorestriction map describes the order and distance between restriction enzyme cleavage sites. The highest resolution physical map will be the complete elucidation of the DNA base-pair sequence of each chromosome in the human genome. LOW-RESOLUTION PHYSICAL MAPPING Chromosomal Map. In a chromosomal map, genes or other identifiable DNA fragments are assigned to their respective chromosomes, with distances measured in base pairs. These markers can be physically associated with particular bands (identified by cytogenetic staining) primarily by in situ hybridization, a technique that involves tagging the DNA marker with an observable label. The location of the labeled probe can be detected after it binds to its complementary DNA strand in an intact chromosome. As with genetic linkage mapping, chromosomal mapping can be used to locate genetic markers defined by traits observable only in whole organisms. Because chromosomal maps are based on estimates of physical distance, they are considered to be physical maps. The number of base pairs within a band can only be estimated. Fluorescence In Situ Hybridization.28,29 A fluorescently labeled DNA probe locates a DNA sequence detected on a specific chromosome. The fluorescence in situ hybridization (FISH) method allows for the orientation of DNA sequences that lie as close as 2 to 5 Mb. Modifications to the in situ hybridization methods, using chromosomes at a stage in cell division (interphase) when they are less compact, increase map resolution by an additional 100,000 base pairs. A cDNA map shows the positions of expressed DNA regions (exons) relative to particular chromosomal regions or bands. (Expressed DNA regions are those transcribed into mRNA.) The cDNA is synthesized in the laboratory using the mRNA molecule as the template. This cDNA can be used to map the genomic region of the respective molecule. A cDNA map can provide the chromosomal location for genes whose functions are currently unknown (ESTs). For hunters of disease genes, the map can also suggest a set of candidate genes to test when the approximate location of a disease gene has been mapped by genetic linkage analysis. HIGH-RESOLUTION PHYSICAL MAPPING Two current approaches to high-resolution mapping are termed top-down (producing a macrorestriction map) and bottom-up (resulting in a contig map). With either strategy, the maps represent ordered sets of DNA fragments that are generated by cutting genomic DNA with restriction enzymes. The fragments are then amplified by cloning or by PCR methods. Electrophoretic techniques are used to separate the fragments (according to size) into different bands, which are visualized by staining or by hybridization with DNA probes of interest. The use of purified chromosomes, separated either by fluorescence-activated flow sorting from human cell lines or in hybrid cell lines, allows a single chromosome to be mapped. A number of strategies can be used to reconstruct the original order of the DNA fragments in the genome. Many approaches make use of the ability of single strands of DNA and/or RNA to hybridize to form double-stranded segments. The extent of sequence homology between the two strands can be inferred from the length of the double-stranded segment. Fingerprinting uses restriction enzyme cleavage map data to determine which fragments have a specific sequence (finger-print) in common and, therefore, overlap. Another approach uses linking clones as probes for hybridization to chromosomal DNA cut with the same restriction enzyme. In top-down mapping, a single chromosome is cut (using rare-cutter restriction enzymes) into large pieces, which are ordered and subdivided; the smaller pieces are then mapped further. The resulting macrorestriction maps depict the order of and distance between locations at which rare-cutter restriction sites are found in the chromosome. This approach yields maps with more continuity and fewer gaps between fragments than contig maps, but map resolution is lower and the map may not be useful in finding particular genes. In addition, this strategy generally does not produce long stretches of mapped sites. Currently, this approach allows DNA pieces to be located in regions measuring ~100 kb to 1 Mb. The development of pulsed-field gel (PFG) electrophoretic methods has improved the mapping and cloning of large DNA molecules. Whereas conventional gel electrophoretic methods separate pieces of DNA <40 kb in size, PFG separates molecules up to 10 Mb, allowing application of both conventional and new mapping methods to large genomic regions. The bottom-up (contig) approach involves cutting the chromosome into small pieces, each of which is cloned and ordered. The ordered fragments form contigs, or contiguous DNA blocks. Currently, the resulting library of clones varies in size from 10 kb to 1 Mb. An advantage of this approach is the accessibility of these stable clones to other researchers. Contig construction can be verified by FISH, which localizes cosmids to specific regions within chromosomal bands. Contig maps consist of a linked library of small overlapping clones of DNA representing a complete chromosomal segment. Although useful for finding genes localized in a small area, contig maps are difficult to extend over large stretches of a chromosome because all regions are not clonable. DNA probe techniques can be used to fill in the gaps, but they are time consuming. Technological improvements now make possible the cloning of large DNA pieces using artificially constructed chromosome vectors that carry human DNA fragments as large as 1 Mb. These vectors are maintained in yeast cells (i.e., YACs).30,31 Before YACs were developed, the largest cosmids carried inserts of only 20 to 40 kb. YAC methodology drastically reduces the number of clones to be ordered; many YACs span entire human genes. A more detailed map of a large YAC insert can be produced by subcloning (a process in which fragments of the original insert are cloned into smaller insert vectors). Because some YAC regions are unstable, large-capacity bacterial vectors have also been developed (bacterial artificial chromosome, BAC). SEPARATING CHROMOSOMES Flow-sorting uses flow cytometry to separate, according to size, chromosomes isolated from cells during cell division, when they are condensed and stable.32,33 As the chromosomes flow singly past a laser beam, they are differentiated by analyzing the amount of DNA present, and individual chromosomes are directed to specific

collection tubes. SOMATIC CELL HYBRIDIZATION In somatic cell hybridization, human cells and rodent tumor cells are fused (hybridized); over time, after the chromosomes mix, human chromosomes are preferentially lost from the hybrid cell until only one or a few remain. Those individual hybrid cells are then propagated and maintained as cell lines containing specific human chromosomes. Improvements to this technique have generated a number of hybrid cell lines, each with a specific single human chromosome. Starting maps and sequences is relatively simple. Finishing them requires either development of new strategies or use of a combination of existing methods. After a sequence is determined, the task that remains is to fill in the many large gaps left by current mapping methods. One approach is a single chromosome microdissection, in which a piece of DNA is physically cut from a chromosomal region of particular interest and then broken up into smaller pieces and amplified by PCR or cloning. These fragments of DNA can then be mapped and sequenced by conventional methods. CHROMOSOME WALKING (POSITIONAL CLONING) Chromosome walking, one strategy for cloning genes and filling gaps, involves hybridizing a primer of known sequence to a clone from an unordered genomic library and synthesizing a short complementary strand (called walking on the chromosome). The complementary strand is then sequenced and its end used as the next primer for further walking. In this way, the adjacent (previously unknown) region is identified and sequenced. The chromosome is thus systematically sequenced from one end to the other. Because primers must be synthesized chemically, a disadvantage of this technique is the need to construct a large number of different primers for walking large distances. Chromosome walking is also used to locate specific genes by sequencing the chromosomal segments between markers that flank the gene of interest (i.e., positional cloning). The degree of difficulty encountered in finding a disease gene of interest depends largely on what information is already known about the gene and especially on what kind of DNA alterations cause the disease. Spotting the disease gene is difficult when disease results from a single altered DNA base; sickle cell anemia is an example of such a case, as are probably most major human inherited diseases. When disease results from a large DNA rearrangement, the anomaly can be detected as alterations in the physical map of the region or even by direct microscopic examination of the chromosome. The location of these alterations pinpoints the site of the gene. CANDIDATE GENE APPROACH Another approach to identifying a disease-causing or diseaserelated gene is the candidate gene approach. Here, a gene known to be important in the development or function of a certain organ or organ system is analyzed for mutations in a cohort with the disease and compared with the gene in a healthy control group. This approach has been valuable in identifying single gene mutations in well-defined, relatively small pedigrees. With the possibility for large-scale genome-wide screening using DNA-array technology, this approach may be applied to large cohorts in a widespread manner in the future.

FUTURE PROSPECTS
HUMAN GENOME PROJECT The Human Genome Project was begun in 1990 by the U.S. government and is coordinated by the U.S. Department of Energy and the National Institutes of Health. Although it was originally conceived as a 15-year program, rapid technological advances as well as collaboration between genome projects of several countries (United Kingdom, Germany, Japan, France) and privately funded endeavors to sequence the human genome have accelerated the project to an expected completion date of 2001, with 90% of the genome available in the spring of 2000.34 The Human Genome Project initiated by the U.S. government has set out to sequence the entire genome of one human individual (identity strictly secret), whereas the privately initiated effort is sequencing the entire genome of five different individuals. The latter approach is addressing the issue of some, although limited, variability in the genome of humans. The goals of the Human Genome Project are to identify the estimated 80,000 expressed genes in human DNA; to determine the sequences of the ~3 billion chemical bases that make up the human DNA; to store the information in databases that will be made accessible to the public; to develop tools for data analysis; and to address the ethical, legal, and societal issues that may arise from the project. Furthermore, to allow comparison of genetic information among species and to address questions related to the interaction of humans with infectious pathogens, researchers are studying the genetic makeup of several nonhuman organisms, including Drosophila melanogaster, Plasmodium falciparum, and the laboratory mouse. The genomes of several organisms such as E. coli, Saccharomyces cerevisiae, Caenorhabditis elegans, Bacillus subtilis, Synechtosis species, A. fulgidus, P. aerophilum, Haemophilus influenzae, M. thermoautrophicum, M. jannaschii, A. aolicolus, Borrelia burgdorferi, Treponema pallidum, Mycoplasma pneumoniae, M. genitalum, Chlamydia trachomatis, Rickettsia prowazekii, Helicobacter pylori, and Mycobacterium tuberculosis are already known and accessible. During progress of the Human Genome Project, several new goals have been formulated. For example, one goal is to identify regions of the human genome that differ from person to person. Although the majority of individuals' DNA sequences are the same, estimates are that humans are only ~99% identical genetically. These DNA sequence variations can have a major impact on how people's bodies respond to disease (i.e., to environmental insults, such as bacteria, viruses, and toxins; and to drugs and other therapies). Methods have been developed to rapidly detect different types of variation, particularly the most common type, called single-nucleotide polymorphisms (SNPs), which occur approximately once every 100 to 300 bases. SNP maps will ultimately help identify multiple genes associated with complex diseases such as cancer, diabetes, vascular disease, and some forms of mental illness. These associations are difficult to establish with conventional gene-hunting methods. because a single altered gene may make only a small contribution to disease risk. The atlas of the human genome will revolutionize medical practice and biologic research into the twenty-first century and beyond. All human genes will be found, and accurate diagnostics will be developed for most inherited diseases. In addition, animal models for human disease research will be more easily developed, facilitating the understanding of gene function in health and disease. Single genes associated with a number of diseases (e.g., cystic fibrosis, Duchenne muscular dystrophy, myotonic dystrophy, neurofibromatosis, diabetes mellitus, and retinoblastoma) have already been identified. Diseases caused by several genes or by a gene interacting with environmental factors can be studied more efficiently. Genetic susceptibilities have been implicated in many major disabling and fatal diseases, including heart disease, stroke, diabetes, and several kinds of cancer. The identification of these genes and their proteins will pave the way to more effective therapies and preventive measures. The potential benefits of the Human Genome Project are manifold. Only some of the important potential applications are discussed here. Molecular medicine (a term lending more significance to medicine and the understanding of human disease and its potential cure at the molecular level) will gain increasing importance in medical practice. Improved diagnosis and earlier detection of genetic predisposition to disease are already commonplace for certain conditions (cystic fibrosis, trisomy, fragile X syndrome, myotonic dystrophy, and neurofibromatosis). The design of drugs that are tailored to the demands imposed by genetic predisposition and the choice of treatment based on genetic information will allow a more rational approach to treatment. Microbial genomics, which yields the sequence and function of microbial genes, can potentially identify new energy sources (biofuels) and identify bacteria useful in environmental remediation, toxic waste reduction, and industrial processing. New tools for risk assessment for exposure to radiation and toxic agents can be made more reliably at the genetic level with information on the human genome. Furthermore, comparison of genomic sequences can increase the capabilities to study evolution through germline mutations in lineages; examine migration of different population groups based on female genetic inheritance; investigate mutations on the Y chromosome to trace lineage and migration of males; and compare breakpoints in the evolution of mutations with ages of populations and historical events. DNA forensic analysis aimed at proper identification of individuals involved in crimes and catastrophic events, establishment of paternity and family relationships, and matching of organ donors is already commonplace with the limited possibilities available today. The expansion of genetic information with the Human Genome Project should allow for improving the reliability of such tests. Agriculture, livestock breeding, and bioprocessing are additional areas affected directly by the genome project. Understanding plant and animal genomes can potentially allow the creation of stronger, more resistant plants and animals. Bioengineered seeds are already being used to grow disease-resistant crops, thereby reducing costs in agriculture. Some additional applications that are already being tested experimentally are the design of pesticides, development of edible vaccines incorporated into food products, development of new environmental cleanup uses for plants like tobacco, and generation of animals and plants producing molecules to be used for the treatment of human conditions (i.e., clotting factors, hemoglobin). Although human genome research itself does not pose any new ethical dilemma, the use of data arising from these studies presents challenges that need to be addressed before the data accumulate significantly. Some of the issues pertinent to the question are summarized in Table 2-4. To assist in policy development, the ethics component of the Human Genome Project is funding conferences and research

projects to identify and consider relevant societal issues, as well as activities to promote public awareness of these topics.

TABLE 2-4. Implications and Issues Arising from the Human Genome Project

STEM CELLS EMBRYONIC STEM CELLS ES cells are pluripotent cells that give rise to all adult cell types.35 They can be derived from the blastocyst, a preimplantation-stage embryo,36 or from primordial germ cells, cells of the early embryo that eventually differentiate into sperm and oocytes.37 Mouse ES cells have been used for approximately a decade to generate genetically altered mice (knock-out, knock-in, Cre recombinants). Mouse ES cells are pluripotent in that they can differentiate into many cell types. ES cells, unlike fertilized eggs, cannot develop completely into individual organisms. ES cells that are placed into a uterus will never develop into an embryo. After desired manipulations, ES cells are injected into the cavity of a developing blastocyst, which ultimately develops into a chimera harboring all tissue cells from the original embryo as well as filiae of the artificially introduced ES cells. ES cells cannot develop into an embryo on their own; they must be placed into an artificial environment, one in which the host cells provide the placental tissues of the conceptus. The derivation of human ES cells from developing fetuses has opened exciting new possibilities for therapy as well as raising legal and ethical issues.36,38 To be able to generate tissue for potential transplantation or organ replacement, ES cells that are immunologically identical to the prospective host would need to be generated. This could possibly be achieved by somatic nuclear transfer, with a nucleus of a prospective host inserted into a recipient enucleated oocyte of either the same or a different species. From the embryo that developed, the inner cell mass of the blastocyst would need to be isolated and grown in vitro to yield ES cells, which then would need to be differentiated into the desired tissue. To date, not all of these steps have been mastered. The use of ES cells therapeutically has dangers, however. Mouse ES cells are tumorigenic, growing into teratomas or teratocarcinomas when injected anywhere in the adult mouse. Human ES cells might behave similarly. PLURIPOTENT STEM CELLS IN ADULT ORGANS Renewable tissues (blood, intestinal epithelium, epidermis) are considered to harbor stem cells that can, by multiplication, yield cells which can then further differentiate into cells of the respective host tissue. Cells in the nervous system have been found to have the capacity to generate new neurons and glial cells (astrocytes and oligodendrocytes), and, because of this, they are considered to be neuronal stem cells.39,40,41 These cells can be isolated from the wall of the lateral ventricle of the brain.42 These cells, which constitute ~0.1% to 1% of the cells in the ependymal lining, express immature neural markers consistent with a stem-cell function. In vitro, these stem cells divide in response to epidermal growth factor and fibroblast growth factor-2.43 On dividing, they are thought to give rise to both neural and glial progeny. If these cells are transplanted into the intact brain, they integrate and differentiate into a range of neuronal and glial cells.44,45 Moreover, these cells can regenerate blood cells when transplanted into lethally radiated host mice. Human embryonic stem cells, which can be cloned46 and grown for extended periods,47 reside in the adult brain (Table 2-5).48

TABLE 2-5. Human Stem Cells Isolated

Another example of pluripotent stem cells is mesenchymal stem cells (MSCs), which can be isolated from bone marrow.49 These cells can be expanded in culture and still differentiate into osteoblasts, osteocytes, chondrocytes, adipocytes, tendonassociated cells, and myotubules50; they have been transplanted into children with osteogenesis imperfecta and into women who have undergone high-dose myelotoxic chemotherapy,51,52 and used for Achilles tendon repair.53 Several ethical and biologic questions need to be addressed before ES cells can be used widely for human applications (Table 2-6).

TABLE 2-6. Biologic and Ethical Questions Pertinent to Use of Embryonic Stem (ES) Cells

SOMATIC CELL CLONING IN VIVO BLASTOMERE SEPARATION Cloning by blastomere separation (called twinning) (Table 2-7) involves splitting a developing embryo soon after fertilization of the egg by a sperm to give rise to two or more embryos. The resulting organisms are identical twins (clones) containing DNA from both the mother and the father. This technique has been used for approximately one decade for generating transgenic or knock-out embryos and animals. Twinning in the sense of generating two or more genetically identical organisms has been practiced for thousands of years with plants.

TABLE 2-7. Animals Cloned by Somatic Nuclear Transfer

SOMATIC NUCLEUS TRANSFER Cloning by somatic nucleus transfer involves the insertion of a nucleus of a cell from one single adult into an egg from which the original nucleus has been removed. This allows for the generation of an animal with the genetic information coming from one single adult (as opposed to cloning by blastomere separation). The technique of somatic nucleus transfer was attempted in the 1950s in frogs, and early attempts were conducted in mammals in the 1980s. In 1996, lambs were successfully cloned via somatic nucleus transfer from embryonic cells that had been cultured in vitro for several months. Dolly, the lamb that was cloned by transferring the nucleus of a cell from the udder of a 6-year-old sheep, gained broad attention. The theoretical possibility was created of generating an unlimited number of genetically identical individuals from one single parent. However, the clones generated through somatic nucleus transfer are not completely genetically identical to the donor animal. The mitochondrial DNA in the host oocyte originates from the mother. Dolly appears to be a healthy sheep and has reportedly given birth to a lamb named Bonnie per vias naturalis. In 1997, a second lamb named Polly was generated by an additional cellular manipulation. Nuclei from fetal fibroblasts were modified in vitro to incorporate the cDNA for human factor IX under control of the b-lactoglobin promoter. These transgenic fibroblasts were used as donor nuclei for somatic transfer. Polly, the lamb generated by this method, produces up to 40 g/L of human factor IX in its milk. Factor IX can now be isolated without much effort and used in therapeutic applications. Since these advances were reported, successful cloning of calves and mice has also been accomplished. The majority of clones fail sometime during development, and some fail after birth. One report suggests that the immune defense system of animals cloned by somatic nucleus transfer may be impaired. Further, telomeres of cloned animals appear to be shorter, suggesting that the cells may have a reduced life span and the animals may manifest symptoms of premature aging. However, the offspring of Dolly appears to have telomeres of normal length. Some accomplishments of somatic nuclear transfer are summarized in Table 2-7. GENE KNOCK-OUT LIBRARIES A main area of investigation when the entire genome of humans as well as that of laboratory animals (i.e., mice and rats) is known and large-scale mutation screening of the entire genome can be performed rapidly will be the study of the function of genes identified in the context of diseases and the consequences of mutations in these genes. For this purpose, collections of knock-out mice strains are made available to researchers for analysis in various assay systems. Furthermore, tissue-specific ablation of certain genes with the Cre-loxP methodology are being provided to researchers. The effects of gene ablation in defined physiologic situations and the effects of drugs in an animal lacking the function of a defined gene can then be studied more efficiently. GENE THERAPY: VECTORS AND PROBLEMS54 The concept of gene therapy is the transfer of genetic information to a host organism or a specific tissue within an organism. The product of the delivered gene either may be missing or of insufficient quantity in the host (replacement therapy) or may be of pharmacologic/therapeutic value (e.g., immune modulation, vaccination). Agents carrying the DNA to target cells are called vectors. The requirements for an ideal vector are that it should accommodate an unlimited amount of inserted DNA, lack the ability of autonomous replication of its own DNA, be easily manufactured, and be available in a concentrated form. For most applications, it should have the ability to target specific cell types or limit its gene expression in the long term or in a controlled fashion. It should not be toxic or immunogenic. Such vectors do not exist, and none of the DNA delivery systems so far available for in vivo gene transfer is perfect. Thus, the advancement of gene therapy lies in the development and improvement of new gene vector systems. Different ways of introducing genes into mammalian cells and tissues have been devised, the simplest being the inoculation of naked DNA by means of microinjection, electroporation, or bioballistics (known as the gene gun technique55). More elaborate and efficient ways include the use of self-assembling complexes of lipid-DNA (e.g., liposomes) protein-DNA, or lipid-protein-DNA, and viral vectors. The physicochemical methods of gene delivery are summarized in Table 2-2 and Table 2-3. Viral vectors can be fragments of viral DNA containing the DNA to be transferred or the viral particle itself.56 The viral particle is rendered replication defective through the manipulation of the viral DNA, and the end product is a nonpathogenic viral vector carrying the genetic information of therapeutic interest. Many different types of virus have been used for genetransfer purposes. The viruses holding most promise for future application in biomedicine (retroviruses, adenoviruses, and adenoassociated viruses) are discussed in the following sections. Retroviral Vectors. The Moloney murine leukemia virus (MoMuLV) was the first virus used in developing a vector system for gene transfer. Parts of the retroviral genome that are involved in the replication of the MoMuLVgag, pol, and envare removed and replaced by a gene of interest. What remains of the retrovirus are the long terminal repeats harboring regulatory elements, integration signals, and transcription promoters, and a packaging signal. To produce retroviral vectors carrying the gene of interest, one must use a packaging cell line harboring the gag, pol, and env genes, which have previously been incorporated into the genome of such packaging cells. The vectors are produced in high titers in the packaging cells, purified, and injected into the organism (in vivo gene therapy) or put in contact with cells collected from the patient and maintained in cell culture (ex vivo gene therapy). Retroviral vectors have the ability to integrate their genomic material into the host genome as a result of the presence of the remaining retroviral regulatory sequences in the form of the long terminal repeats. Once integrated, the inserted gene can be expressed to produce the desired protein. Advantages of retroviral vectors are that they are well characterized, they can be produced in high titers, and they have a high efficiency of infection. Disadvantages are the limited size of DNA (78 kb) that retrovirus vectors can accommodate, the requirement that the target cells be in cell division to allow for the integration of the vectors, and the potential for insertional mutagenesis due to retroviral vector integration at random in the genome. The latter feature has the potential to interrupt important genes in the cell, with serious consequences that include oncogenesis through the activation of protooncogenes or inactivation of tumorsuppressor genes. Lentiviruses are a group of retroviruses that have the ability to infect and integrate their genome even in cells that are not dividing. Thus, the use of lentiviruses as gene therapy vectors is broader than that of other retroviruses and is under investigation.

Adenoviral Vectors. The human adenoviruses are nonenveloped DNA viruses with a linear double-stranded DNA of ~36 kb, encapsulated in an icosahedral capsoid measuring 70 to 100 nm in diameter. Adenoviruses are known pathogens in humans, and most, if not all, adult humans have been exposed to adenoviruses and have antibodies against adenovirus antigens. An adenovirus does not depend on host-cell division for its replication, and the chromosome rarely integrates into the genome, remaining episomal in most cases. Integration seems to occur mainly in the presence of high levels of infection in dividing cells, but this event does not contribute significantly to the utility of these viruses as vectors. Adenoviral vectors have a broad spectrum of cell infectivity that includes virtually all postmitotic and mitotic cells; they also can be produced in high titers. The genome of the adenovirus encodes ~15 proteins. Viral gene expression occurs in a coordinated fashion and is mainly directed by the E1A and E1B genes, localized within the 5' region of the adenoviral genome. These genes have transactivation functions for the transcription of various viral and host-cell genes. Because E1 genes are involved in adenoviral replication, their removal renders the virus replication incompetent. The removal also creates room for insertion of a foreign gene of therapeutic interest. An exogenous DNA can also replace the E3 region, which produces a product enabling the virus to evade the immune system. Packaging cells carrying adenoviral genes that provide transcomplementation functions are required to produce defective adenoviral vectors. Packaging cells of the NIH-293 cell lineage are human embryo kidney (HEK) cells that have been previously transformed with type 5 adenovirus. These cells retain the E1A and E1B regions of the viral genome covalently linked to their genomic DNA. Disadvantages of the adenoviral vectors include the short duration of transgene expression, because the vector usually does not integrate stably into the host-cell genome. Further, the size of the inserted foreign DNA sequence is limited, and cellular and humoral immune responses are typically triggered against the adenoviral particles or against the host cell that eventually expresses adenoviral proteins, thus limiting the longevity of the adenovirus vector. Adenoassociated Viral Vectors. Adenoassociated virus (AAV) is a small nonenveloped, nonpathogenic DNA virus belonging to the Parvoviridae family. The AAV genome is a single-strand DNA molecule of 4681 bases, including two inverted long terminal repeats (ITRs). ITRs are 145-base-long palindromic sequences involved in the regulation of the AAV cell cycle. They are located in the 5' and 3' terminal portions of the viral genome and serve as origins and initiators for DNA replication. Flanked by the ITRs, two large open-reading frames code for a regulatory protein and a structural protein, called rep and cap, respectively. The protein coding sequence located in the 5' region (rep gene) encodes four nonstructural proteins involved in the genomic replication. The 3' region contains the cap gene, which encodes three structural proteins required for the formation of the viral capsid. AAV is capable of replication in a cell only in the presence of a helper virus (adenovirus or herpes virus) that provides by transcomplementation the helper factors that are essential for its replication. In the absence of a helper virus, the AAV genome preferentially integrates into a specific site on the short arm of chromosome 19, between q13.3 and qter, called AAVS1. The ITRs as well as a rep transcript play an important role in this process, which results in a latent infection in mitotic as well as in postmitotic cells. Episomal virus and insertion in nonspecific sites has been documented. Among the advantages of AAV as a potential gene vector in human gene therapy are the lack of relation to human diseases, broad infectivity spectrum, and ability to stably integrate into the host genome. This integration can occur in cells that are not dividing, although at a lower frequency than in dividing cells. The site-directed integration is also a most favorable property of AAV. DNA-BASED VACCINATION Immunization with transfer of genetic material represents a novel approach to vaccination.57,58 The technology involves transferal of a gene (encoding an antigenic protein cloned in an expression vector) to a host, leading to the induction of an immune response. Direct gene transfer may be undertaken using either viral vectors or recombinant plasmid DNA. Viral vectors have the disadvantages of being derived from pathogens (like traditional vaccines based on attenuated virus), and, therefore, are of limited interest for the purpose of immunization. In contrast, DNA plasmids encoding antigens are more frequently used because they do not have the inconvenience of classic vaccines: they are safe, inexpensive, easy to produce, heat stable, and amenable to genetic manipulation. Currently, two main delivery systems are available for gene vaccination. Plasmid DNA is injected intramuscularly, or DNA is coated onto gold beads and transferred into the epidermis or dermis by a bioballistic process (gene gun). The intramuscular injection is the most widely used method for immunization, and it consists of direct injection of naked DNA into skeletal muscle. Plasmid DNA in some instances is injected into muscle directly in saline solution or after injection of toxins or a local anesthetic to cause necrosis and regeneration of the injected muscle, thereby increasing the expression of the encoded antigen and amplifying the immunologic response. It is unclear whether the elevated antigen in regenerating muscle cells is due to an increased expression of the antigen gene or to the antigen contained within antigen-presenting cells that are recruited to the site of tissue damage. Humoral and cell-mediated immune responses have been induced by the direct intramuscular injection of plasmid DNA endocrine immunogens. An antibody response was first reported against an influenza virus protein in mice, and specific cytotoxic T-cell responses were also detected in different systems (i.e., human immunodeficiency virus infection and hepatitis B) after genetic immunization. Protective immunity was first demonstrated in mice injected intramuscularly with nucleoprotein DNA of influenza virus. In this model, researchers have indicated that both CD4+ and CD8+ T cells contributed to the protection. Protective immune responses have also been demonstrated in mice against Leishmania major, Plasmodium yoelii, Mycobacterium tuberculosis, dengue virus, and herpes simplex virus. The bioballistic (gene gun) method uses a helium gas pressuredriven device to deliver gold particles coated with plasmid directly into the skin. When gene vaccines are administered by gene gun technology, most of the plasmid DNA is taken up by keratinocytes and some dermal fibroblasts; they become transfected and produce the encoded antigen. Humoral responses using bioballistic approaches were demonstrated using plasmids encoding human growth hormone and human a-antitrypsin. The nature of the immune response elicited by these DNA vaccination approaches is not clearly understood. In an initial study, the suggestion was made that DNA vaccination elicits a cell-mediated immune response, because passive transfer of serum from immune mice did not engender protective immunity. Depletion experiments demonstrated that both CD4+ and CD8+ cell populations were involved in host protection against infection. However, some investigators have also reported an induction of immunoglobulins after gene gunmediated immunization. CHAPTER REFERENCES
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. Angier N. Natural obsessions: the search for the oncogene. Boston: Houghton Mifflin, 1988. Judson HF. The eighth day of creation: makers of the revolution in biology. ColdSpring Harbor: Cold Spring Harbor Press, 1996. Hall SS. Invisible frontier: the race to synthesize a human gene. New York: Atlantic Monthly Press, 1987. Morange MA. History of molecular biology. Cambridge, MA: Harvard University Press, 1998. Baltimore D. RNA-dependent DNA polymerase in virions of RNA tumour viruses. Nature 1970; 226:1209. Kelly TJ Jr, Smith HO. A restriction enzyme from Hemophilus influenzae. II. J Mol Biol 1970; 51:393. Smith HO, Wilcox KW. A restriction enzyme from Hemophilus influenzae. I. Purification and general properties. J Mol Biol 1970; 51:379. Danna K, Nathans D. Specific cleavage of simian virus 40 DNA by restriction endonuclease of Hemophilus influenzae. Proc Natl Acad Sci U S A 1971; 68:2913. Maxam AM, Gilbert W. A new method for sequencing DNA. Proc Natl Acad Sci U S A 1977; 74:560. Sanger F, Nicklen S, Coulson AR. DNA sequencing with chainterminating inhibitors. Proc Natl Acad Sci U S A 1977; 74:5463. Sambrook J, Fritsch EF, Maniatis T. Molecular cloning: a laboratory manual, 2nd ed. Cold Spring Harbor: Cold Spring Harbor Press, 1989. Ausubel FM, Brend R, Kingston RE, et al. Current protocols in molecular biology. New York: John Wiley and Sons, 1993. Saiki RK, Scharf S, Faloona F, et al. Enzymatic amplification of betaglobin genomic sequences and restriction site analysis for diagnosis of sickle cell anemia. Science 1985; 230:1350. Freeman WM, Walker SJ, Vrana KE. Quantitative RT-PCR: pitfalls and potential. Biotechniques 1999; 26:112. Wright PA, Wynford-Thomas D. The polymerase chain reaction: miracle or mirage: a critical review of its uses and limitations in diagnosis and research. J Pathol 1990; 162:99. Mullis KB, Ferre F, Gibbs RA. The polymerase chain reaction. New York: Springer-Verlag, 1994. Bronson SK, Smithies O. Altering mice by homologous recombination using embryonic stem cells. J Biol Chem 1994; 269:27155. Rajewsky K, Gu H, Kuhn R, et al. Conditional gene targeting. J Clin Invest 1996; 98:600. Gossen M, Freundlieb S, Bender G, et al. Transcriptional activation by tetracyclines in mammalian cells. Science 1995; 268:1766. Ramsay G. DNA chips: state-of-the art. Nat Biotechnol 1998; 16:40. Kurian KM, Watson CJ, Wyllie AH. DNA chip technology. J Pathol 1999; 187:267. Donis-Keller H, Green P, Helms C, et al. A genetic linkage map of the human genome. Cell 1987; 51:319. Grompe M, Gibbs RA, Chamberlain JS, Caskey CT. Detection of new mutation disease in man and mouse. Mol Biol Med 1989; 6:511. Rikke BA, Johnson TE. Towards the cloning of genes underlying murine QTLs. Mamm Genome 1998; 9:963. Leach RJ, O'Connell P. Mapping of mammalian genomes with radiation (Goss and Harris) hybrids. Adv Genet 1995; 33:63. Martin N, Boomsma D, Machin G. A twin-pronged attack on complex traits. Nat Genet 1997; 17:387. Farral M. Affected sibpair linkage tests for multiple linked susceptibility genes. Genet Epidemiol 1997; 14:103. Chang SS, Mark HF. Emerging molecular cytogenetic technologies. Cytobios 1997; 90:7. Luke S, Shepelsky M. FISH: recent advances and diagnostic aspects. Cell Vis 1998; 5:49. Coulson A, Waterston R, Kiff J, et al. Genome linking with yeast artificial chromosomes. Nature 1988; 335:184. Vollrath D, Davis RW, Connelly C, Hieter P. Physical mapping of large DNA by chromosome fragmentation. Proc Natl Acad Sci U S A 1988; 85:6027. Roslaniec MC, Bell-Prince CS, Crissman HA, et al. New flow cytometric technologies for the 21st century. Hum Cell 1997; 10:3. Ferguson-Smith MA. Genetic analysis by chromosome sorting and painting: phylogenetic and diagnostic applications. Eur J Hum Genet 1997; 5:253. Collins FS. Shattuck lecturemedical and societal consequences of the Human Genome Project. N Engl J Med 1999; 341:28. Vogel G. Harnessing the power of stem cells. Science 1999; 283:1432. Thomson JA, Itskovitz-Eldor J, Shapiro SS, et al. Embryonic stem cell lines derived from human blastocysts [see comments] [published erratum appears in Science 1998; 282(5395):1827].

37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58.

Science 1998; 282:1145. Shamblott MJ, Axelman J, Wang S, et al. Derivation of pluripotent stem cells from cultured human primordial germ cells. Proc Natl Acad Sci U S A 1998; 95:13726. Marshall E. A versatile cell line raises scientific hopes, legal questions. (News; comment). Science 1998; 282:1014. McKay R. Stem cells in the central nervous system. Science 1997; 276:66. Weiss S, van der Kooy D. CNS stem cells: where's the biology (a.k.a. beef)? J Neurobiol 1998; 36:307. Alvarez-Buylla A, Temple S. Stem cells in the developing and adult nervous system. J Neurobiol 1998; 36:105. Johansson CB, Momma S, Clarke DL, et al. Identification of a neural stem cell in the adult mammalian central nervous system. Cell 1999; 96:25. Cattaneo E, McKay R. Proliferation and differentiation of neuronal stem cells regulated by nerve growth factor. Nature 1990; 347:762. Gage FH, Kempermann G, Palmer TD, et al. Multipotent progenitor cells in the adult dentate gyrus. J Neurobiol 1998; 36:249. Takahashi M, Palmer TD, Takahashi J, Gage FH. Widespread integration and survival of adult-derived neural progenitor cells in the developing optic retina. Mol Cell Neurosci 1998; 12:340. Flax JD, Aurora S, Yang C, et al. Engraftable human neural stem cells respond to developmental cues, replace neurons, and express foreign genes. Nat Biotechnol 1998; 16:1033. Svendsen CN, ter Borg MG, Armstrong RJ, et al. A new method for the rapid and long term growth of human neural precursor cells. J Neurosci Methods 1998; 85:141. Eriksson PS, Perfilieva E, Bjork-Eriksson T, et al. Neurogenesis in the adult human hippocampus [see comments]. Nat Med 1998; 4:1313. Prockop DJ. Marrow stromal cells as stem cells for nonhematopoietic tissues. Science 1997; 276:71. Horwitz EM, Prockop DJ, Fitzpatrick LA, et al. Transplantability and therapeutic effects of bone marrow-derived mesenchymal cells in children with osteogenesis imperfecta [see comments]. Nat Med 1999; 5:309. Koc H, Gurman G, Arslan O, et al. Is there an increased risk of graft-versus-host disease after allogeneic peripheral blood stem cell transplantation? Blood 1996; 15:2362. Lazarus HM, Haynesworth SE, Gerson SL, et al. Ex vivo expansion and subsequent infusion of human bone marrowderived stromal progenitor cells (mesenchymal progenitor cells): implications for therapeutic use. Bone Marrow Transplant 1995; 16:557. Young RG, Butler DL, Weber W, et al. Use of mesenchymal stem cells in a collagen matrix for Achilles tendon repair. J Orthop Res 1998; 16:406. Dani SU. The challenge of vector development in gene therapy. (In Process Citation). Braz J Med Biol Res 1999; 32:133. Robinson HL. DNA vaccines: basic mechanisms and immune responses. Int J Mol Med 1999; 4:549. Stone D, David A, Bolognani F, et al. Viral vectors for gene delivery and gene therapy within the endocrine system. J Endocrinol 2000; 164:103. Davis HL. Plasmid DNA expression systems for the purpose of immunization. Curr Opin Biotechnol 1997; 8:635. Seder RA, Gurunathan S. DNA vaccines: designer vaccines for the 21st century. N Engl J Med 1999; 341:277.

CHAPTER 3 BIOSYNTHESIS AND SECRETION OF PEPTIDE HORMONES Principles and Practice of Endocrinology and Metabolism

CHAPTER 3 BIOSYNTHESIS AND SECRETION OF PEPTIDE HORMONES

WILLIAM W. CHIN Overview of Peptide Hormone Synthesis and Secretion Gene Structure Messenger RNA Processing RNA Transport Translation Posttranslation Secretory Granule Secretion Regulation of Polypeptide Hormone Synthesis Generation of Diversity Chapter References

In the endocrine system, hormones are factors produced by groups of cells clustered in specific tissues, commonly known as glands, and released into the general circulation to affect the function of distant target cells. Because hormones are responsible for the control of a complex metabolic milieu, they, along with the hormone-producing and target cells, participate in intricate regulatory networks (see Chap. 4 and Chap. 5). An important feature is positive regulation of hormone synthesis and secretion. For example, gonadotropin-releasing hormone (GnRH)from the hypothalamus stimulates production and release of the pituitary gonadotropins. Another common theme is regulation by negative feedback, by which a trophic hormone stimulates the production and secretion of a second hormone in a target cell that acts on the original gland to decrease secretion of the trophic hormone. For example, thyroid-stimulating hormone (TSH) is produced and secreted from the thyrotrope in the anterior pituitary gland. It stimulates the thyroid gland to synthesize and secrete thyroid hormones, which act on the thyrotrope to decrease further production and release of TSH. Thus, hormones may regulate the biosynthesis and release of other hormones. Moreover, hormones may control other cellular activities by determining the amount and activity of other proteins. This regulation occurs largely at the gene transcriptional level, although some regulation occurs at post-transcriptional, translational, and posttranslational levels. In addition to operating at the broader level of the endocrine system, factors secreted from a given cell can influence cellular activities in adjacent or neighboring cells (i.e., paracrine effects) or within itself (i.e., autocrine effects; see Chap. 1). Intracellular communication effected by hormonal factors is critical for the integrated function of an organism. The pivotal and ubiquitous nature of hormones in these tasks makes understanding their synthesis and release essential. This chapter describes the events that occur in the biosynthesis of polypeptide hormones in hormone-secretory cells, from the gene to the final, bioactive protein hormone, including the intracellular structures involved in this process. Insights about the regulation of the hormone-producing cell derived from studies involving recombinant DNA technology and molecular and cellular biology are highlighted.

OVERVIEW OF PEPTIDE HORMONE SYNTHESIS AND SECRETION

Proteins are important as the backbones of polypeptide hormones and as integral components of enzymes that participate in the biosynthetic pathways of steroid and thyroid hormones as well as enzymes that participate in intracellular synthetic and degradative actions and in energy generation. They are critical membrane, receptor, and cytoskeletal molecules, and an appreciation of the pathways for polypeptide synthesis and their associated cellular structures is important. This section describes the general pathways of polypeptide synthesis, with an emphasis on the informational flow from the gene to the final functional protein and the cell structures involved in each of the steps in this highway of biochemical events.1,2 and 3 The production of a functional protein hormone requires numerous steps, each one involving modifications or processing of precursor molecules. The eukaryotic cell consists of two major compartments, the nucleus and cytoplasm (Fig. 3-1 and Fig. 3-2), which are delimited by plasma membranes that are topologically contiguous with one another (see Fig. 3-2). The nucleus is surrounded by a nuclear envelope consisting of an outer and inner membrane encompassing a cisternal space.4 The nuclear envelope is perforated by nuclear pore complexes that permit communication of the nucleoplasm with the cytoplasm.

FIGURE 3-1. Electron micrograph of a rat lactotrope. The rat lactotrope in the anterior pituitary gland synthesizes and secretes prolactin. This electron micrograph reveals a portion of a lactotrope, including the two major compartments, nucleus (N) and cytoplasm. The nucleus is delineated by a double membrane, the nuclear envelope. The key organelles in the synthesis and processing of polypeptide hormones include therough endoplasmic reticulum (r), Golgi stack (G), secretory vesicle (v), and secretory granule (sg). Mitochondria (m), the major source of cellular energy, are shown. The secretory granules are characterized by an electron-dense material representing condensed polypeptide hormone. (Courtesy of Gwen V. Childs, University of Texas at Galveston.)

FIGURE 3-2. Diagram of the secretory cell. The secretory cell contains two major compartments: nucleus and cytoplasm. The nucleus is delimited by a nucleoplasmic membrane that is perforated by nuclear pore complexes. The nuclear membrane is contiguous with the endoplasmic reticulum (ER). Transport of polypeptides from the ER to the next organelle, the Golgi stack, is accomplished by way of transport vesicles or transitional elements. Polypeptide hormones exit from the Golgi stack by formation of secretory vesicles and granules. The stored polypeptide hormone in the secretory granule is released in the process of emiocytosis or exocytosis on receipt of the appropriate extracellular stimuli. The shaded areas represent topologically extracellular spaces. The lumen of the ER and Golgi are contiguous with the extracellular space.

The nucleus contains much of the cellular nucleic acid or genetic material in the form of genes, which harbor the information necessary for the initial production of precursor RNAs and hence functional proteins. The cytoplasm contains multiple organelles that are involved in the synthesis of proteins and their processing. These

events occur through an assembly-line arrangement, by which the initial protein precursors are altered by changes in their primary structures and by glycosylation and other chemical modifications. These organelles include the endoplasmic reticulum (ER), where initial protein synthesis occurs, and a complex membranous structure known as the Golgi stack, where further protein processing and posttranslational modifications, sorting, and translocation take place. The secretory cell is differentiated from other cell types by the presence of secretory granules that emerge from the Golgi stack. These granules are specialized, membrane-bound organelles that contain polypeptide hormones in high concentration that may be stored for long periods. Stimulus-secretion coupling allows release of the hormone from the granule on physiologic demand. The informational flow from the gene to the final protein is shown in Figure 3-3. Each protein produced by a cell is encoded by a gene. In a typical mammalian haploid genome, ~100,000 genes are grouped together into clusters called chromosomes. However, only a subset of these genes is expressed in a given cell. The genes in chromosomes are organized as chromatin with its DNA bound to basic histone and acidic nonhistone proteins. Specifically, the DNA is wrapped nearly twice around an histone-octamer core at regular intervals (i.e., every 140 nucleotides) to yield nucleosomes and a structure resembling beads on a string. These protein-DNA interactions provide the gene with vital secondary and tertiary structures. The DNA in this condensed form probably is transcriptionally inactive. However, modifications in this structure, dictated by developmental or regulated patterns, may determine whether a particular gene may be transcribed at all and, if so, at what rate. The role of chromatin structure in determining the breadth of cell-specific gene transcription remains to be clarified.

FIGURE 3-3. Informational flow from the polypeptide hormone gene to the bioactive secreted hormone. The gene, containing information in the form of DNA, is transcribed into heterogeneous nuclear RNA. This heterogeneous nuclear RNA is rapidly processed to form mature messenger RNA (mRNA). These events occur in the nucleus. The mRNA enters the cytoplasm, where it interacts with the protein synthetic machinery and undergoes translation, by which the protein hormone precursor is synthesized. Then numerous cotranslational and posttranslational processes occur in the rough endoplasmic reticulum and Golgi stack to yield the mature protein hormone. Secretory vesicles bud and emerge from the trans-Golgi to produce the secretory granule. In this state, the polypeptide hormone is stored and released on stimulation by the appropriate extracellular signals, whereupon the hormone enters the extracellular space. The secreted hormone may be acted on further by extracellular processes to yield other active hormone species and may be subjected to peripheral degradation.

The information in a gene appears in the form of a double-stranded polymer of deoxyribonucleotides (DNA). The protein sequence is embedded in triplets of nucleotides in a tandem array dictated by the genetic code. The information present as DNA in a gene must be transferred to another molecule, RNA, which transfers the original information from the nucleus to the cytoplasm. The informational transfer from the gene to RNA is known as transcription. Initially, a precursor RNA, called heterogeneous nuclear RNA (hnRNA), is synthesized using the original DNA in the gene as a template. This precursor molecule is then rapidly processed into the mature messenger RNA (mRNA). The mRNA exits the nucleus to enter the cytoplasm. The mRNA molecule is a version of the gene with its data represented as a linear sequence of ribonucleic acids. The mRNA quickly interacts with the protein synthetic machinery, the ribosomes of the cytoplasm. The ribosome is a complex structure that contains ribosomal RNAs and its associated proteins. Within this structure, the information present in the mRNA is eventually transferred to protein information. The availability of several adapter molecules known as transfer RNA (tRNA) allows the information in the triplets of nucleotides (i.e., codons) in the mRNA to be converted to amino acids. Because one of four different nucleotides can occupy each position in a triplet codon, 43 or 64 sequence possibilities exist. More codons are available than are necessary to encode the 20 essential amino acids. The genetic code contains redundancy or degeneracy so that a single amino acid may be represented by more than one codon. Because a protein molecule has a beginning and an end, the mRNA must contain information for the start and stop of translation. Important codons include AUG, or initiator methionine, which is the first amino acid of all newly synthesized polypeptides, and UAG, UAA, and UGA, which are termination codons. Proteins destined for secretion are produced on ribosomes and possess a NH2-terminal signal or leader peptide that interacts with cytoplasmic and ER receptors to mediate rapid ribosome-ER membrane association. This new complex allows newly synthesized proteins destined for secretion to enter the lumen of the ER, which is topologically located outside the cell. Soon after this occurs, cotranslational events take place, and the newly synthesized polypeptide is then sequestered within the cisternal space of the ER. The polypeptide then migrates from the ER through the Golgi stack, where further processing occurs. After the transfer from the cis to the medial to the trans regions of the Golgi complex, the maturing polypeptide hormone is sorted and transferred to the secretory granule. Polypeptides in secretory vesicles and granules are released into the extracellular space by fusion of the vesicle with the plasma membrane and by exocytosis of the contained material. Hormones in secretory granules are stored until the appropriate extracellular signal, generally a calcium flux, is received to prompt the release of the contents of the secretory granule. In response to the chemical signal, the membranes of the secretory granule fuse with the plasma membrane, and emiocytosis (i.e., regulated exocytosis) occurs. In a series of alterations in precursor RNA and protein molecules, the eventually mature and bioactive polypeptide hormone is synthesized and secreted.

GENE STRUCTURE
Polypeptide hormones may be encoded by single or multiple genes. Frequently, a hormone requires only intramolecular folding and formation of disulfide linkages in a single protein backbone to form the bioactive molecule. Sometimes, the bioactive hormone is formed by the covalent or noncovalent association of two or more subunits derived from a single gene or multiple genes. An example of the former is insulin, which is initially synthesized as a precursor with polypeptide subunits A and B interrupted by peptide C. However, during its intracellular processing, disulfide linkages are formed between subunits A and B, with the proteolytic cleavage and removal of peptide C. Two subunits are associated in a covalent manner to yield the bioactive insulin molecule. Major examples of the latter case are glycoprotein hormones (i.e., TSH and gonadotropins) (see Chap. 15 and Chap. 16). In this family of hormones, each member consists of two subunits encoded by separate genes located on separate chromosomes. The subunits become associated in a noncovalent manner to form the bioactive dimer.5 Several hormones require proteolytic cleavage of the precursor molecule before the formation of the bioactive product. The major example is adrenocorticotropic hormone (ACTH) and b-lipotropin produced from the precursor preproopiomelanocortin by trypsin-like proteolytic cleavage at dibasic residues. Each polypeptide may require covalent modifications of its polypeptide backbone. In the glycoprotein hormones, each subunit contains several N-linked carbohydrate moieties, and the b subunit of human chorionic gonadotropin contains additional O-linked oligosaccharides. In yet other molecules, the addition of sulfate, phosphate, acetyl, and COOH-amide groups is necessary for full bioactivity. Bioactive peptides found in the brain-gut axis require a COOH-terminal amide group for full activity. This conversion is catalyzed by an a-amidation enzyme on substrate hormones that possess COOH-terminal sequences: X-Y-Gly X-Y-NH2.6,7 The gene that encodes the polypeptide hormone is part of a simple or complex transcriptional unit. A simple transcriptional unit (Fig. 3-4) is composed of two major components: structural and regulatory. A simple unit produces a single mRNA, whereas a complex unit may yield multiple mRNAs, some of which may encode different proteins. The structural region encodes information that is ultimately found in mRNA (Fig. 3-5). However, in most eukaryotic genes, the coding region is not contiguous with that in the mRNA. Intervening or extraneous segments of DNA are placed between regions that eventually are found in mature mRNA. The coding regions in genes that are ultimately found in mRNA are known as exons, and the intervening sequences are known as introns. The role and function of introns are unknown, although introns separate functional domains in many polypeptides.8 A similar exon encoding an epidermal growth factor (EGF)like domain has been detected in genes encoding the low-density lipoprotein receptor, the precursor of EGF, and clotting factors IX and X.9 These and other data suggest that introns may play important roles in the evolution of protein families. Moreover, introns may participate in alternate splicing of exons, leading to increased mRNA and polypeptide diversity.

FIGURE 3-4. The transcriptional unit. Each polypeptide hormone or subunit is encoded by a transcriptional unit. This diagram shows the transcriptional unit that contains structural and regulatory regions. The regulatory region is shown at the 5'-flanking portion of the transcriptional unit. However, such regulatory elements may occur in other parts of the gene, including introns or 3&3039;-flanking regions. The structural region is bounded by the transcriptional initiation or cap site at the 5' end and the polyadenylation site at the 3' end. The signals for transcription termination are more than 50 to 200 nucleotides downstream of the polyadenylation site.

FIGURE 3-5. The structural region of the transcriptional unit. The structural region contains DNA information that is completely copied and transcribed into the heterogeneous nuclear RNA or RNA precursor. The important feature of eukaryotic structural regions of the gene is the presence of exons and introns. The exon contains sequences that are retained in the mature messenger RNA; the intron sequences are removed during RNA splicing in the nucleus. The first nucleotide of the structural region is known as the cap site, which is the point at which transcription begins in the first exon. The structural region terminates at the polyadenylation site, which is determined in part by the presence of a polyadenylation signal, AATAAA, located 15 to 20 nucleotides upstream of the polyadenylation site.

The regulatory region contains elements that determine whether a gene is transcribed and, if so, in what quantity (Fig. 3-6). A structural region alone is ineffective in this informational flow. The presence of a regulatory region is obligatory for the expression of its associated gene. The DNA in a gene is not found devoid of associated proteins in nature. Marked secondary, tertiary, and quaternary structure is found in genes located in chromatin, and covalent modifications of nucleotide residues are found within genes. In particular, methylation of cytosine residues may play an essential part in determining whether regulatory regions are open for transcription. The regulatory region plays an important role in moment-to-moment regulation of gene expression and in the tissue-specific and developmental programs of gene expression.

FIGURE 3-6. The regulatory region of the transcriptional unit. This diagram shows the important elements of the regulatory region. Only part of the structural region, including its cap site, is shown. The TATA box is located ~25 to 30 nucleotides upstream of the cap site. It is fixed in position and orientation and binds important binding proteins that allow interaction with RNA polymerase II. The upstream promoter element is located 40 to 110 nucleotides upstream of the cap site and binds critical proteins that also interact with the RNA polymerase II. The interplay of the upstream promoter element and TATA box are crucial for basal expression of a given structural region. Another element is the enhancer, which is located a variable distance from the cap site, including locations downstream of the cap site, and is independent of orientation. The enhancer element also binds to DNA-binding proteins that further augment or decrease transcription. Enhancer interactions determine regulation of gene expression above or below basal levels. A special subset of enhancers includes the hormone regulatory element, which mediates the effects of steroid and thyroid hormones and second messengers induced by polypeptide hormones.

The process of transcription, whereby information in the gene as DNA is transformed into information as RNA, produces a large RNA precursor molecule (see Fig. 3-3).10 The initial RNA transcript commences at the 5' end of the first exon, continues through the other exons and introns in the structural region, and terminates at the 3' end of the last exon. This hnRNA contains more information than in the mature mRNA because of the inclusion of intron sequences. The enzyme RNA polymerase II performs the transcription reaction using the DNA in the gene as the template. The first nucleotide in the RNA precursor transcript, generally a purine (A or G), is defined as the transcriptional start, also called the cap site. Subsequent ribonucleotides are polymerized during RNA elongation until the transcript is completed in the process of termination. The initiation of transcription by RNA polymerase II often is the rate-limiting step in gene expression and is determined by the interaction of the enzyme with a number of nuclear factors and DNA elements in the regulatory region of the transcription unit.11,12 Two types of elements participate in regulation of transcription of the typical eukaryotic gene. The first component is the cis-acting element, which is a DNA segment located in the regulatory region of a transcriptional unit. These cis elements are located near the structural region of a gene. Another component is the trans-acting element or factor, which usually consists of a DNA-binding protein present in the nucleus that is encoded by a gene separate from the transcriptional unit that is being regulated and interacts with a cis-regulatory element. These nuclear factors are said to act in trans to regulate gene expression. Mutations that affect trans-activation occur in distant genes but not in the gene that is being regulated. Interactions of transacting factors with cis-DNA elements determine tissue-specific, developmental, and regulated expression of genes. Three important DNA elementsthe TATA box, the upstream promoter element, and enhancersoccur in the regulatory region, which is commonly located at the 5'-flanking region of the transcriptional unit (see Fig. 3-6).12,13 However, regulatory regions or elements thereof may occur in other sections of the transcriptional unit, including introns and 3'-flanking regions. The key elements include the TATA box, an AT-rich region located in a fixed position and orientation 25 to 30 nucleotides upstream from the cap site. This DNA element binds several TATA-binding proteins and associated factors found in the nucleus that allow for efficient interaction of RNA polymerase II with the transcriptional unit. The TATA box interactions, although not strictly required, are important for accurate and efficient initiation of transcription. The nature of the transcription initiation complex has been clarified. The components include at least seven transcription factors (i.e., TFIIA, TFIIB, TFIID, TFIIE, TFIIF, TFIIH, TFIIJ) and RNA polymerase II. A major constituent is TFIID, which is also a large multisubunit (>700-kDa) complex with at least eight proteins. One of these, the TATA-binding protein (TBP), allows the binding of TFIID to the TATA box or related Inr (initiator) sequences. The other components of TFIID have also been partially characterized and are known as coactivators or TBP-associated proteins (TAFs). They are essential for the communication of enhancer-binding protein signals to the basal transcriptional machinery and the subsequent regulation of gene expression. Basal transcription depends on the formation of a preinitiation

complex involving TFIID-TFIIA-TFIIB, followed by the rapid entry of RNA polymerase II to facilitate the establishment of the transcriptional machinery.14,15 and 16 The second DNA element is the upstream promoter element (UPE), which is located 60 to 110 nucleotides upstream from the cap site.17 It includes elements such as the CCAAT and GC-rich (GGGCGG) boxes that bind to CAAT-binding proteins and SP1 (a cellular DNA-binding protein that interacts with the SV40 genome), respectively. These elements also associate with DNA-binding proteins that augment the efficiency of transcription by RNA polymerase II. These UPEs may or may not require specific TATA boxes to perform their function most efficiently. Together, the TATA box and UPEs are components close to the structural region and are essential for maintenance of basal levels of gene transcription (Fig. 3-7).

FIGURE 3-7. Regulation of transcriptional rates by interactions of transacting factors. Various permutations of interactions of nuclear binding proteins with various DNA elements within regulatory regions determine rates of transcription at the basal and regulated levels. Such proteins include the TATA box, the upstream promoter element (UPE), and enhancer or hormone regulatory element (HRE)binding proteins.

Enhancers are located in variable positions and may act independently of orientation. They may be located more distal than the promoter elements and are found up to several thousand nucleotides upstream or downstream of the transcriptional unit. These elements also bind proteins that enhance transcriptional rates or diminish them (i.e., silencers) in an ill-defined manner and constitute the foci of regulated transcription (see Fig. 3-7). Several proposed mechanisms include the cooperative interaction of a number of DNA-binding proteins to effect efficient formation of the transcription-initiation complex of RNA polymerase II with the regulatory or promoter region.18 Another hypothesis suggests that the interaction of proteins with these elements opens up the configuration of DNA, perhaps by bending to allow access of the gene to the transcription machinery. With recombinant DNA techniques, a reporter gene construct can be produced that may be transfected into foreign cells by gene transfer.19,20 This allows the expression of the reporter gene with enhanced production of an enzyme or polypeptide product that is not normally produced in eukary-otic cells. The synthesis of such products may be detected by sensitive enzyme assays or radioimmunoassays. DNA constructs in which a structural region corresponding to the enzyme alone is transfected into cells are not expressed in the absence of regulatory regions. However, if a promoter element is placed 5' to the reporter gene, then expression may occur. Using such approaches, structural analysis of various portions of the 5'-regulatory regions of genes, including enhancer and upstream regulatory elements, may be performed. After the RNA transcript is initiated, the RNA polymerase II continues the process of template reading by elongation of the transcript until termination occurs. The actual site of transcription termination is variable, located 50 to 200 or more nucleotides downstream from the 3' end of the last exon or polyadenylation site.21 Although potential weak consensus sequences have been discerned that may determine the site at which the initial RNA transcript is terminated, the polyadenylation site appears to be obtained only by virtue of endonucle-olytic cleavage of longer heterogeneous 3' ends of the hnRNA. The well-conserved consensus polyadenylation site sequence AAUAAA, which is located 15 to 20 nucleotides upstream from the polyadenylation site, and other proximal downstream sequences 10 to 12 nucleotides from the polyadenylation site, may serve as points of recognition for this processing event. Although the mechanisms of polyadenylation are not well known, the presence of the consensus sequences suggests a requirement for stem-loop formation and involvement of small nuclear ribonucleoproteins (snRNPs).

MESSENGER RNA PROCESSING

The hnRNA product of gene transcription is rapidly processed in the nucleus with a half-time (t1/2) of 5 to 20 minutes (Fig.3-8).22 Three major events transform the large heterogeneous RNA precursors into the mature RNA. First, at the 5' end of hnRNA, a 7methylguanosine residue is added to the first nucleotide of the transcript by means of a 5'-5' triphosphate bond after 20 to 30 nucleotides have been polymerized. This reaction is rapid (t1/2 < 1 minute) and is catalyzed by the 5'-capping enzyme, including guanylyl and methyl transferases. The 5'-methyl cap associates with 5'-cap binding proteins, which favors the formation of a stable 40S translation-initiation complex and increases the stability and efficiency of translation of the eventual mature mRNA.23

FIGURE 3-8. Gene transcription and RNA processing. The initial RNA transcript is known as heterogeneous nuclear RNA (hnRNA I). It contains exons and introns of the structural region and rapidly undergoes 5' capping with 7methylguanosine (7meG) and 3' polyadenylation (An) (hnRNA II). Little heterogeneous nuclear RNA has been detected without 5' cap or 3' polyadenylated (poly[A]) tails. In a slower process, introns are removed by RNA splicing followed by religation of exon sequences. The mature messenger RNA (mRNA) is composed of fused exon sequences and contains a 5' cap and a 3' poly(A) tail.

The second modification occurs at the 3' end and involves the addition of a polyadenylate, or poly(A), tail. Polyadenylation includes the addition of 250 to 300 adenylate (A) residues at the polyadenylation site located at the 3' end of the RNA. This poly(A) tail, which is reduced to 30 to 250 residues during nuclear processing and export, may also be important for increased RNA stability. These two additions, capping and polyadenylation, occur within minutes after the synthesis of hnRNA and generally before RNA splicing; almost all isolated hnRNA contains both modifications. The third major processing step involved in mRNA maturation is the removal of introns during RNA splicing.24,25 and 26 This process includes endonucleolytic cleavage of introns and religation of exons. The 5' and 3' ends of introns have consensus sequences, as shown in Figure 3-9.

FIGURE 3-9. RNA splicing: consensus intron sequences and mechanisms for intron removal. A consensus sequence has been determined for the 5' and 3' ends of intron sequences. Data suggest the potential mechanism of intron removal by means of lariat formation preceded by interaction with nuclear RNAs. (nt, nucleotides.)

These consensus sequences may be necessary for the appropriate interaction of U1 snRNP species present in the nucleus to serve as a splicing adapter for the splicing process. Moreover, a polypyrimidine tract is located adjacent to the 3' AG residues and a critical adenylate residue in a branch sequence, 30 nucleotides upstream of the 3' end of the intron. The first step in the splicing process involves the formation of the spliceosome, which includes the hnRNA, U1 snRNP, and other factors. The initial event is endonucleolytic cleavage at the 5' splice site, followed by the formation of a 5'-2' phosphodiester bond between the 5' G and the downstream A located in a branch sequence. This lariat intermediate is then cleaved at the 3' end and degraded, and the exons are ligated. The removal of introns from hnRNA must be precise; errors can change the exon or mRNA-coding regions. The sequence of removal of multiple introns within a gene is generally nonrandom, although the mechanism is unknown. Variations in the splicing pattern in a given hnRNA transcript can occur, and tissue-specific interactions of RNA splicing-modification proteins may dictate alternate patterns of intron-RNA splicing, causing altered mRNA forms.27 In a complex transcriptional unit, an alternate exon choice, including alternative internal acceptor and donor site use, may yield different mRNA products. A complex transcriptional unit may also possess alternate transcriptional start sites in the same contiguous segment of DNA (i.e., in the same exon) or in multiple transcriptional start sites in different exons contributed by alternate exon choice. Another possible mechanism for diversity in the complex transcriptional unit is alternative final exon choice (i.e., differences in polyadenylation sites). The splicing process is another rate-limiting step and takes place over 5 to 30 minutes; it is much slower than the capping and polyadenylation reactions. What role RNA splicing plays in the informational flow is unclear. However, the potential contribution of RNA diversity by RNA splicing has been discussed. There are mRNAs that lack poly(A) tails (e.g., histone mRNAs), mRNAs that lack a 5' cap (e.g., poliovirus mRNAs), and eukaryotic genes that lack introns. Such modifications are not essential for RNA maturation.

RNA TRANSPORT
The newly synthesized mature mRNA is actively transferred from the nuclear to the cytoplasmic compartments by way of the nuclear pore complex (NPC). The NPC is a large multiple-component structure that is located in the nuclear envelope and serves as a channel for the movement of macromolecules such as RNAs. The mRNA and other RNAs subject to transport are closely associated with proteins and exist as ribonucleoproteins. Each RNA likely possesses distinct protein-targeting sequences that permit its export and import. This shuttling of mRNA from the nucleus to the cytoplasm is mediated by a large family of transport factors known collectively as exportins and importins. However, the precise nature of the interactions of these shuttling proteins, mRNAs, and the NPC is not well understood.28

TRANSLATION
The structure of mRNA is shown in Figure 3-10. The exons encode two major regions of the mRNA: translated and untranslated. The translated or coding region contains the open reading frame, beginning from the initiation methionine codon to the termination codon. The untranslated regions flank the coding region and are known as 5' or 3' untranslated regions. The functions of the untranslated regions are not well established, but data indicate that the 5' untranslated region may be important in determining the efficiency of translation of the mRNA.29,30 The 3' untranslated region may contain important RNA elements, especially several AU-rich sequences that determine the stability of mRNA in the cytoplasm.31 Each of these regions may mediate its effects by binding to specific RNA-binding proteins.32,33

FIGURE 3-10. Structure and translation of messenger RNA (mRNA). In most cases, the mature mRNA represents the fusion of multiple exons. These sequences encode two major regions: translated and untranslated. The translated or coding region is delimited by the translation initiator codon, AUG, at its 5' end and the termination codon, UGA, UAA, UAG, at its 3' end. This coding region represents a series of codons in an open-reading frame that determines the amino-acid sequence of its encoded polypeptide. The 5' and 3' untranslated regions are shown. The mRNA enters the cytoplasm to interact with the ribosome. There, protein synthesis is initiated, and by way of a series of several cotranslational events, secretory polypeptide hormone precursors are processed. The steps involve cleavage of the signal or leader peptide, followed by addition of asparagine-linked carbohydrate moieties in glycoprotein subunits or hormones, and intramolecular folding with the formation of disulfide linkages. These events occur within the lumen of the rough endoplasmic reticulum. These partially processed polypeptide hormones are then shuttled to the Golgi stack, where these molecules are transported, sorted, and further processed posttranslationally to yield the bioactive hormone located in secretory granules or vesicles.

The mRNA in the cytoplasm rapidly interacts with the ribosome (see Fig. 3-10). The ribosome is a complex ribonuclear particle that contains 28S, 18S, and 5S RNAs, along with a group of ribosomal proteins. Among these proteins are factors responsible for the initiation, elongation, and termination of mRNA translation. For the typical mRNA, 3 to 15 ribosomes may be attached at any given time. As the ribosome reads the mRNA in the process of translation, amino acids are brought to the translation complex by way of adapter tRNA molecules. These molecules are differentiated by the presence of anticodon structures (i.e., RNA sequences complementary to a particular codon) at one end and attachment sites for specific amino-acid residues at the other end of the L-shaped molecule. The reading of successive codons causes the alignment of the appropriate amino acids and polymerization to yield the polypeptide chain. Translation initiation occurs at the initiator codon or AUG, which represents the amino residue methionine. Translation generally begins at the first AUG codon located at the 5' end of the mRNA. This initiation methionine codon is normally followed by an open reading frame of codons encoding amino acids until a termination codon is reached. When a UAG, UAA, or UGA is encountered, protein synthesis stops, and the nascent polypeptide chain is released from the ribosome complex. The context of the methionine codon that is used for translation initiation has been characterized further to include a consensus sequence: 5'-CCACCAUGG-3'. This sequence nest presents the AUG as the most favorable initiation codon.34 However, examples have been found in which the AUG is located 5' of the authentic start site. In these instances, the context may not be ideal or may be quickly followed by a termination codon in frame. Whether peptides encoded by these short-reading frames are eventually expressed is unknown.35,36 All polypeptide hormones and almost all other proteins destined for membrane, lysosome, ER, and Golgi stack locations or for secretion are encoded by a larger

polypeptide precursor. All polypeptide hormones possess a signal or leader peptide that is a characteristic segment of protein located at the N-terminal end37,38 (Table 3-1). Although no consensus primary sequence has been obtained for this signal peptide, it generally possesses a hydrophobic core preceded by basic amino-acid residues in its 16- to 30-amino-acid residue extent.

TABLE 3-1. Polypeptide Hormones: Some of Their Precursor Proteins*

Several events occur before the entire polypeptide chain is synthesized (Fig. 3-11). After the synthesis of ~70 amino acids, the signal recognition particle (SRP), a group of six proteins and a small RNA (7S), interacts with the signal peptide to momentarily halt translation elongation in the RNAribosome-nascent protein complex.39,40 The 7S RNA contains a signal peptide recognition and an elongation arrest domain. This complex then interacts with the SRP receptor, an integral membrane protein located on the cytoplasmic face of the ER. In this process, poly-ribosomes are attached to membranous structures associated with the endoplasmic reticulum to form the rough ER (RER). After this interaction occurs, translational arrest is relieved, and translation proceeds as usual. At this point, the signal peptide is vectorially transported through the membrane into the cisternal aspect of the ER. The newly synthesized protein has been translocated from the inside to the outside of the cell in a topologic sense.

FIGURE 3-11. Details of translational and cotranslational processes. The messenger RNA (mRNA) interacts with the ribosome where protein synthesis is initiated. In the case of polypeptide hormones, the first segment of protein synthesized is the N-terminal signal or leader peptide. As soon as the signal peptide emerges from the ribosomal complex, a protein-RNA particle known as the signal recognition particle (SRP) associates with the signal peptide. This interaction allows the ribosomal-mRNAnascent polypeptide complex to interact with the SRP receptor located on the cytoplasmic face of the endoplasmic reticulum (ER) membrane and brings the ribosome in close apposition to the ER to form the rough ER. The momentary translational arrest that occurs on interaction of the complex with SRP is released to allow further protein synthesis. Cleavage of the signal peptide from the apoprotein by signal peptidase and other modifications, including addition of asparagine-linked carbohydrates (CHO), intramolecular folding, and disulfide linkage formation, occurs coincidentally with release of ribosomes from the ER. In this manner, the partially processed protein, although initially synthesized in the cytoplasmic space, enters the luminal space.

As protein synthesis continues, the signal peptide is transiently immobilized in the membrane by virtue of its hydrophobic nature or its binding to a putative signal peptide receptor.41,42 Although the nascent protein chain is transferred to the cisterna by way of an unknown, energy-dependent translocation process, a luminal surface enzyme, signal peptidase, rapidly performs proteolytic cleavage to remove the signal peptide. The transmembrane transport of the protein does not require signal peptide cleavage and may take place by way of a protein channel or, less likely, through lipid. If the protein is to be N-glycosylated (i.e., to contain asparagine-linked carbohydrate moieties), other enzymes and the dolichol-lipid oligosaccharide carrier provide core glycosylation in this cotranslational process. Moreover, protein folding and oxidation of cysteine residues in disulfide formation occur. At the completion of protein synthesis and complete transfer of the protein to the luminal space, the SRP complex dissociates from its receptor and is recycled into the cytoplasm. Polyribosomes also are disaggregated to form free ribosomes and RNA. The translation of the polypeptide hormone causes the synthesis of a polypeptide core derived from the initial protein precursor, which is already modified, in some instances, by the addition of carbohydrate moieties and by folding and formation of intramolecular disulfide linkages. The precursor polypeptide encoded by mRNA is not found in vivo, because the signal peptide is removed before the completion of the polypeptide chain. The exit from the ER probably depends on appropriate protein assembly or conformation, but glycosylation is not required.

POSTTRANSLATION
Up to the translational and cotranslational steps in the ER, all secretory, membrane, lysosome, endogenous ER, and Golgi proteins have traversed the same biosynthetic path. After this point, the major task of sorting and transferring the proteins to the correct intracellular destinations must be completed. This complex process occurs in the Golgi stack and requires sorting signals among the proteins and sorting mechanisms in this organelle. A polypeptide hormone destined for regulated secretion must exit the ER, traverse the Golgi stack, and arrive properly in the secretory granule (Fig. 3-12).

FIGURE 3-12. The polypeptide hormone highway. Protein hormone synthesis is initiated in the cytoplasm on polyribosomes. The partially processed hormone, with the signal peptide removed and N-linked carbohydrate moieties attached and with appropriate folding, enters the lumen of the rough endoplasmic reticulum (RER). By way of transport vesicletransitional elements, these partially processed products are transferred to the Golgi stack on fusion and release. In a serial process of budding formation of secretory vesicles and fusion, processed products are transferred through the Golgi stack, from which they exit as secretory vesicles or granules after sorting in the trans and trans-Golgi network compartments of the Golgi. Materials are then released from granules by the fusion of vesicles or granules with the plasma membrane.

The Golgi stack comprises a series of flattened, saccular membranous compartments that encompass four histologically and functionally distinct regions: the cis, medial, and trans regions of the Golgi complex and the trans-Golgi network (TGN)43,44 (Fig. 3-13). The cis-Golgi region is most proximal to the transitional elements of the RER, and the TGN is most distal. The maintenance of distinct Golgi-specific antigens, unique enzyme markers, and different lectin-binding characteristics suggest that the compartments are not contiguous.

FIGURE 3-13. The Golgi stack. The Golgi stack consists of numerous membranous compartments, including cis, medial, and trans-Golgi elements. These compartments may be differentiated by the presence of specific enzymes. Partially processed protein hormones traverse this system by way of intermediate secretory vesicles in a budding-fusion reiterative process. In addition to transport, protein processing occurs. Sorting with routing to ultimate destinations in cellular sites is accomplished in the trans-Golgi network (TGN). Secretory peptides may be sorted to constitutive or regulated secretory pathways. Constitutive secretory pathways are equivalent to the pathways taken by membrane proteins, whereby nonclathrin-coated membrane segments are used. The regulated secretory-secretory granule pathway involves a clathrin-coated pit among membrane segments. This is similar to the pathway taken by lysosomal components. (Adapted from Griffiths G, Simons K. The trans Golgi network: sorting at the exit site of the Golgi complex. Science 1986; 234:438.)

A vesicle transfer model has been proposed to account for transport of materials from the RER to the TGN. In this model, membrane vesicles form from the upstream compartment by budding at the rims of the Golgi plates and rejoin the adjacent downstream compartment by vesicle fusion and the interaction of microfilaments. The reiterative process of budding and fusion of secretory or transport vesicles causes vectorial transfer of proteins from the RER to the TGN in a unidirectional and energy-dependent process. The newly synthesized polypeptide in the lumen of the RER is first translocated to the cis-Golgi region (see Fig. 3-12). From this point, the protein is transported and processed in the Golgi stack. This organelle may be appropriately considered an assembly line for posttranslational processing. It is here that N-linked carbohydrate cores are further modified among glycoproteins45 (Fig. 3-14). This process involves digestion of the high-mannose peripheral sugars in the N-linked carbohydrate cores by multiple glycosidases and subsequent addition of distal or terminal sugars by way of numerous glucosyltrans-ferases. The steps in this process of carbohydrate maturation occur in different Golgi compartments. Other processes also occur, including phosphorylation, acetylation, sulfation, acylation, -amidation of COOH termini, addition of ubiquitin, other modifications, and degradation.46

FIGURE 3-14. Proximal and distal glycosylation. The pathway of glycosylation in the rough endoplasmic reticulum (RER) and Golgi is shown. Core carbohydrate moieties are added cotranslationally by way of a dolichol-sugar intermediate (Dol-) to Asn residues in the protein backbone in the RER. Several glycosidases (steps 14) remove distal sugars in this compartment. Distal glycosylation occurs by the actions of mannosidases (steps 57) and glycosyl transferases (steps 6, 81) in the Golgi. Phosphorylation (I, II) of N-acetyl glucosamines in carbohydrate moieties in the cis Golgi occurs in proteins destined for lysosome localization. (From Kornfeld R, Kornfeld S. Assembly of asparagine-linked oligosaccharides. Annu Rev Biochem 1985; 54:631.)

Another important function of the Golgi stack is the delivery of nascent polypeptides to the appropriate targets within the cell, which occurs in the trans-Golgi region or TGN.47 The proteins destined for lysosomal sites are targeted to those organelles by way of the mannose-6-phosphate receptor.48 In a similar manner, receptor and secretory proteins are targeted to membrane and secretory granule sites, respectively.49,50,51,52 and 53 The nearly mature polypeptide emerges from the Golgi stack in the TGN, where transport organelles, known as secretory vesicles or granules, are formed. These vesicles allow the exit of the nearly mature protein hormone from the Golgi stack. Secretory proteins are released from a cell by way of two pathways: the constitutive pathway and the regulated pathway.54,55 The constitutive pathway is thought to be mediated by a passive aggregation sorting mechanism whereby peptide hormones form aggregates in the TGN, an action that is facilitated by acidic pH and high calcium concentrations in this compartment. The polarity of the secretory faces of epithelial cells enables proteins that are released in a nonregulated or constitutive manner to be released on the apical surface and regulated release to be performed at the basolateral surface. Whether such polarity of secretion exists in endocrine cells is unknown. Constitutive release generally involves the rapid exocytosis of newly synthesized peptides, but regulated secretion involves the classic secretory granule and signaled degranulation, causing hormone- or factor-regulated release of hormones. Secretory peptides must be segregated into one pathway or the other. Regulated secretion involves the formation of secretory residues and granules composed of clathrin-containing membrane segments, as found in lysosomes. Proteins destined for regulated secretion must end up in a reservoir known as the secretory granule, where the polypeptide hormones are concentrated and stored. This pathway is now considered to operate by active sorting via a signal ligand receptor. Proteins destined for secretion in this manner clearly contain sorting signals in their precursor molecules. For instance, the precursors to proopiomelanocortin (POMC) and proenkephalin have a stretch of aliphatic hydrophobic and acidic amino-acid residues at the N termini that are necessary and sufficient for efficient sorting into secretory granules. Further, carboxypeptidase E (Cpe) appears to serve as a sorting receptor for these peptide signals as determined by biochemical and genetic approaches. In particular, the Cpefat, which harbors a mutant and ineffective Cpe, is obese, diabetic, and infertile. It has elevated levels of proinsulin in pancreatic B cells and of POMC in the anterior pituitary, and decreased insulin and ACTH release.56 Three types of vesicles are formed in the TGN. One is the secretory vesicle, which is not clathrin coated and mediates nonreceptor-dependent transport of membrane proteins and protein to be secreted in the constitutive pathway. The other two are the secretory granule, which is partially clathrin coated and mediates the receptor-dependent transfer of regulated secretory peptides, and the lysosome, which is predominantly clathrin coated and mediates transport of lysosomal enzymes and proteins.57 The secretory vesicle participates in the default, bulk-flow sorting system, but the others require the presence of sorting patches or sorting signals based on secondary and tertiary, but not primary, structures.55 Although the secretory granules are derived from immature granules with clathrin-coated pits, the precise nature of the receptor-mediated sorting of peptide hormones is unknown. Evidence exists for pH-regulated, receptor-dependent sorting in the trans-Golgi and TGN. The pH of the compartments decreases as the Golgi stack is traversed from cis to trans regions. Such gradients in pH may participate in the molecular aggregation of polypeptide hormones. Possibly, these aggregates formed in the process of hormone concentration may initiate the budding of secretory granules. Chloroquine, which prevents Golgi acidification, may inhibit granule formation by preventing aggregation in neutralized Golgi stacks. Proteolytic processing of protein precursors (i.e., proproteins or polyproteins) to yield smaller bioactive peptides (see Table 3-1) also occurs in acidic Golgi and

secretory vesicles.58 Such proteolysis, however, is not required for packaging.

SECRETORY GRANULE
Much has been learned about the nature of polypeptide hormones and secretory granules.55,59,60 The hormones in this organelle are highly concentrated. In particular, a number of polypeptide hormones are condensed in a crystal lattice formation to increase the amount of hormone (up to 200-fold) in this organelle. Secretory granules allow cells to store enough hormone to be released on demand by extracellular signals at a level not possible by de novo synthesis. The t1/2 of stored hormones may be days, whereas the t1/2 of similar proteins in secretory vesicles may be minutes. The size of secretory granules varies greatly, depending on the nature of the stored hormone. The condensation of hormone is demonstrated by the presence of electron-opaque or dense cores. The granule core is quite stable and is often visible even after exocytosis or in vitro enzymatic digestion of the granule membrane. It is osmotically inert yet sensitive to pH levels higher than 7.0.55 The formation of the secretory granule proceeds in stages, beginning in the trans-Golgi, where the initial hormone concentration may be observed. This aggregation process60a is facilitated by changes in pH, calcium concentration, and possible presence of other proteins such as secretogranins, chromogranins, and sulfated proteoglycans. Aggregates may form in different regions of the secretory granule. The colocalization of two or more polypeptide hormones in a granule may be observed. Within a cell, the relative distribution of two hormones is constant from granule to granule; however, variability in overall distribution is achieved from cell to cell. The mechanism by which the gonadotrope, a cell that generally produces luteinizing hormone (LH) and follicle-stimulating hormone (FSH), may be regulated to release LH and FSH differentially remains unclear.61

SECRETION
Secretory granules release their contents by cytoskeletal protein-mediated movement of the granule toward the cellular surface.61a There, secretory granule membranes fuse with the plasma membrane and allow eversion or exocytosis of stored hormone.62 This process of emiocytosis causes secretion of hormone. The mechanisms involved in stimulus-secretion coupling are not well known, although responses to cellular signals causing changes in intracellular calcium, ion currents, or intra-cellular pH may lead to these events. In the unstimulated cell, a web of actin-associated microfilaments on the cytoplasmic face of the plasma membrane may act as a physical barrier to secretory granule fusion. However, changes in intracellular calcium, ion currents, or intracellular pH may cause differences in actin-binding protein interactions and alterations in the secretory barrier and permit exocytosis to occur. Secretion and rapid membrane fusion of multiple secretory granules require an endocytotic pathway to retrieve the extra membranes resulting from exocytosis in the plasma membrane and to return them to the Golgi stack and lysosome.

REGULATION OF POLYPEPTIDE HORMONE SYNTHESIS

Regulation of the biosynthesis of polypeptides may occur at any of the biosynthetic levels in the pathway (Table 3-2 and Fig. 3-4). Of major interest is the regulation of peptide hormone synthesis at the transcriptional level.

TABLE 3-2. Loci of Genetic Regulation of Polypeptide Hormone Synthesis*

Studies using gene transfer and structure-function analysis have established that specific DNA elements in the regulatory region of the transcriptional unit are critical for determining transcriptional rates of various structural regions.63,64 In particular, hormone-regulatory elements (HREs) have been characterized for glucocorticoid, estrogen, androgen progesterone, vitamin D, mineralocorticoid, retinoic acid, and thyroid hormone receptors. In each case, a DNA element 8 to 20 nucleotides long may be necessary and sufficient for conferring hormonal regulation to its associated structural region. Several factors, including the steroid and thyroid hormones, interact with nuclear receptor proteins, which interact with DNA elements directly to modulate gene transcription.65,66,67 and 68 For the glucocorticoid receptor, the glucocorticoid ligand binds to the inactive glucocorticoid receptor in the cytoplasm, present in a complex with heat shock proteins, hsp 90 and hsp 70, and others. The activated receptor-ligand complex interacts as a transacting factor to bind the DNA element corresponding to the glucocorticoid regulatory element (GRE). Studies have been performed on GREs in genes for mouse mammary tumor virus (MMTV) and murine sarcoma viruses, human metallothionein IIa, tyrosine aminotransferase, tryptophan oxygenase, and growth hormone, and in other genes. The long terminal repeat region of MMTV contains five GREs.64,69 A consensus sequence for the putative GRE is shown by the sequence 5'-GGTA-CANNNTGTTCT-3', inwhich N = A, C, G, or T. The structures of the steroid and thyroid hormone receptors are better known. These hormone receptors are encoded by genes related to a viral oncogene, v-erbA.70,71 The thyroid hormone receptor is encoded by the protooncogene c-erbA. Each receptor contains a stereotypic structure, including a protein that is ~45 to 60 kDa, with a central DNA-binding domain and a carboxyl-terminal ligand-binding domain. These and other regions mediate trans-activation, dimerization, and nuclear localization. The DNA-binding region consists of multiple cysteine and histidine residues that are critical for the formation of Zn2+ fingers first described in the DNA-binding protein TFIIIa, a transcription regulatory factor for the 5S ribosomal gene in Xenopus.72 This Zn2+finger interaction is a common motif for the binding of many eukaryotic proteins to DNA.73,74 and 75 The steroidthyroid hormone receptors represent the first major examples of trans-acting factors well described in mammalian systems. The motif found in prokaryotic systems, particularly the interactions of cro and lambda repressor proteins with their target DNA elements in bacteriophage lambda, occurs with a homopolymeric dimer of subunits containing alpha helixturnalpha helix structure. The binding generally involves protein dimers; it requires a twofold axis of symmetry in the DNA sequence and involves the major groove of the target DNA over several helical turns. Data indicate that the thyroid hormone, retinoic acid, and vitamin D receptors are active only in the heterodimeric state, with other nuclear factors such as retinoid X receptors as their partners. Hormones that act by way of surface membrane receptors may induce the production of second messengers that may directly or indirectly interact with DNA elements within the gene.76,77,78 and 79 HREs may not be restricted to interactions observed with steroidthyroid hormone receptor complexes, but they may involve other protein-DNA interactions. Advances in the isolation of such trans-acting factors and the identification of cis-acting HREs will probably speed an understanding of the molecular mechanisms of the hormonal regulation of gene expression at the transcriptional level.80 The presence of multiple enhancer elements or HREs in the regulatory regions of genes allows fine tuning of transcriptional efficiency and influences the rate of production of the initial RNA transcript81,82,83,84 and 85 (Fig. 3-15).

FIGURE 3-15. Thyroid hormone action. This diagram depicts the mechanism of action of thyroid hormones in the regulation of a thyroid hormoneresponsive gene. Thyroxine (T4) or triiodothyronine (T3) enters the cell. T4 is converted to T3 intracellularly in many cells by means of 5'-deiodinase activity. T3 then enters the nucleus, where it binds to the nuclear thyroid hormone receptor, which is encoded by c-erbA. This hormone nuclear receptor complex then serves as a transacting factor for binding to a thyroid hormone regulatory element (TRE), which may then positively or negatively regulate gene expression, with resultant production of RNA and protein derived from the thyroid hormoneregulated gene. (mRNA, messenger RNA.)

Other loci for regulation in this biosynthetic pathway include elongation and termination of transcription86 (see Fig. 3-9). The various steps of RNA maturation, most notably RNA splicing, may also change mRNA levels encoding a particular polypeptide hormone, which ultimately determines the amount of polypeptide produced. The nuclear stability of the hnRNA and transport of the RNA from the nucleus to the cytoplasm also may be regulated. A major determinant of the steady-state levels of mRNA is cytoplasmic mRNA stability. Examples include the estrogen regulation of chicken liver vitellogenin mRNA, prolactin regulation of breast casein mRNA stabilities, and thyroid hormone control of the TSH b subunit.87,88 and 89 The interaction of mRNA with the protein synthetic machinery in the process of translation may be regulated. Several examples of translational control have been observed, including glucose regulation of the translational efficiency of insulin mRNA. Moreover, a number of the posttranslational processing events that occur in the RER and Golgi stack and the control of secretory granule formation and release may also be loci for regulation. Even after proteins are released from the secretory cell, the bioactive peptide may be further acted on by degradative processes and proteolytic events that may activate proteins in extracellular steps to determine the bioactivity of a particular polypeptide hormone. A major example involves the cascade of the extracellular enzymatic conversion of the precursors of angiotensin II (see Chap. 79). Another example of postsecretion proteolytic processing of precursor polypeptides involves the conversion of iodinated thyroglobulin to the iodinated thyronines, thyroxine and triiodothyronine, in the follicular cell of the thyroid. Plasma stability of a polypeptide is a major determinant of the activity of the hormone in its eventual interaction with target cells.

GENERATION OF DIVERSITY
A major example of the generation of diversity is the calcitonin and calcitonin generelated peptide (CGRP) system. In this system, the C cell of the thyroid expresses a calcitonin-CGRP transcript that initially contains six exons. In the C cell, tissue-specific factors determine the use of the polyadenylation site in the fourth exon, but in the brain, transcription through the sixth exon, which encodes CGRP and the alternative polyadenylation site present in that exon, provides the alternative splicing and deletion of the fourth exon, which encodes calcitonin. The C cells express mostly calcitonin and not much CGRP; conversely, the hypothalamus produces mostly CGRP but not much calcitonin (see Chap. 53 and Fig. 53-1). Other examples of alternative splicing yielding different polypeptides include the synthesis of the alternate human growth hormone form, substance P, substance K, and protooncogenes.90,91 Alternative processing of polypeptides in a posttranslational process is important for the generation of polypeptide diversity.92,93 A major example of this is the production of ACTH and b-lipotropin from the POMC precursor (Fig. 3-16). Using the same mRNA transcript, the anterior pituitary gland produces ACTH and b-lipotropin, and the intermediate lobe of the pituitary gland performs further alternate proteolytic processing and produces b-endorphin, corticotropin-like intermediate lobe peptide (CLIP), a-melanocyte-stimulating hormone, and other products (see Chap. 16).

FIGURE 3-16. Alternative protein processing of the preproopiomelano-cortin (POMC) precursor. In the anterior pituitary gland, the single POMC precursor is processed posttranslationally to produce adreno-corticotropic hormone (ACTH) and b-lipotropin (b-LPH). However, the intermediate lobe further processes these peptides to a-melanocyte-stimulating hormone (a-MSH), corticotropin-like intermediate lobe peptide (CLIP), g-lipotropin (g-LPH), and b-endorphin. (From Douglass J, Civielli O, Herbert E. Polyprotein gene expression: generation of diversity of neuroendocrine peptides. Annu Rev Biochem 1984; 53:665.)

CHAPTER REFERENCES
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. Alberts B, Watson JD, Bray D, et al. Molecular biology of the cell. New York: Garland Publishing, 1994. Matsudaira P, Berk A, Zipursky L, et al. Molecular cell biology. New York: WH Freeman Co, 1999. Lewin BM. Genes VII. New York: Oxford University Press, 1999. Newport JW, Douglass JF. The nucleus: structure, function, and dynamics. Annu Rev Biochem 1987; 56:535. Chin WW. Organization and expression of glycoprotein hormone genes. In: Imura H, ed. The pituitary gland. New York: Raven Press, 1985:164. Eipper BA, Mains RE, Glembotski CC. Identification in pituitary tissue of a peptide alpha-amidation activity that acts on glycine-extended peptides and requires molecular oxygen, copper, and ascorbic acid. Proc Natl Acad Sci U S A 1983; 80:5144. Bradbury AF, Finnie MDA, Smyth DG. Mechanism of C-terminal amide formation by pituitary enzymes. Nature 1982; 298:686. Gilbert W. Why genes in pieces? Nature 1978; 271:501. Sudhoff TC, Russell DW, Goldstein JL, et al. Cassette of eight exons shared by genes for LDL receptor and EGF receptor. Science 1985; 228:893. Nevins JR. The pathway of eukaryotic mRNA formation. Annu Rev Biochem 1983; 52:441. Darnell JE Jr. Variety in the level of gene control in eukaryotic cells. Nature 1982; 297:365. Rowley A, Dowell SJ, Diffley FX. Recent developments in the initiation of chromosomal DNA replication: a complex picture emerges. Biochim Biophys Acta 1994; 1217:239. Albright SR, Tjian R. TAFs revisited: more data reveal new twists and confirm old ideas. Gene 2000; 242:1. Zawel L, Reinberg D. Advances in RNA polymerase II transcription. Curr Opin Cell Biol 1992; 4:488. Pugh BF, Tjian R. Diverse transcriptional functions of the multisubunit eukaryotic TFIID complex. J Biol Chem 1992; 267:679. Conaway RC, Conaway JW. General initiation factors for RNA polymerase II. Annu Rev Biochem 1993; 62:161. Busby S, Ebright RH. Promotor structure, promotor recognition, and transcription activation in prokaryocytes. Cell 1994; 79:743. Ptashne M. Gene regulation by proteins acting nearby and at a distance. Nature 1986; 322:697. Walker MD, Edlund T, Boulet AM, Rutter WJ. Cell-specific expression controlled by the 5'-flanking region of insulin and chymotrypsin genes. Nature 1983; 306:557. Edlund T, Walker MD, Barr PJ, Rutter WJ. Cell-specific expression of the rat insulin gene: evidence for role of two distinct 5'-flanking elements. Science 1985; 230:912. Platt T. Transcription termination and the regulation of gene expression. Annu Rev Biochem 1986; 55:339. Brawerman G. Determinants of messenger RNA stability. Cell 1987; 48:5. Shatkin AJ. mRNA cap binding proteins: essential factors for initiating translocation. Cell 1985; 40:223. Padgett RA, Grabowski PJ, Konarska MM, et al. Splicing of messenger RNA precursors. Annu Rev Biochem 1986; 55:1119. Sharp PA. Splicing of messenger RNA precursors. Science 1987; 253:766. Keller W. The RNA lariat: a new ring to the splicing of mRNA precursors. Cell 1984; 39:423.

27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60.

Andreadis A, Gallego ME, Nadal-Ginard B. Generation of protein isoform diversity by alternative splicing: mechanistic and biological implications. Annu Rev Cell Biol 1987; 3:207. Weis K. Importins and exportins: how to get in and out of the nucleus. Trends Biol Sci 1998; 23:185. Pelletier J, Sonenberg N. Insertion mutagenesis to increase secondary structure within the 5'-noncoding region of a eucaryotic mRNA reduces translational efficiency. Cell 1985; 40:515. Darveau A, Pelletier J, Sonenberg A. Differential efficiencies of in vitro translation of mouse cmyc transcript differing in the 5'-untranslated region. Proc Natl Acad Sci U S A 1985; 82:2315. Shaw G, Kamen R. A conserved AU sequence from the 3'-untranslated region of GM-CSF mRNA mediates selective mRNA degradation. Cell 1986; 46:659. Nielsen DA, Shapiro DJ. Insights into hormonal control of messenger RNA stability. Mol Endocrinol 1990; 4:953. Atwater JA, Wisdom R, Verma IM. Regulated mRNA stability. Annu Rev Genet 1990; 24:519. Kozak M. Compilation and analysis of sequences upstream from the translational start site in eukaryotic mRNAs. Nucleic Acids Res 1984; 12:857. Kozak M. Selection of initiation sites by eucaryotic ribosomes: effect of inserting AUG triplets upstream from the coding sequence for preproinsulin. Nucleic Acids Res 1984; 12:3873. Kozak M. Bifunctional messenger RNAs in eukaryotes. Cell 1986; 47:481. Von Heijne G. A new method for predicting signal sequence cleavage sites. Nucleic Acids Res 1986; 14:4683. Gierasch LM. Signal sequences. Biochemistry 1989; 28:1. Walter P, Gilmore R, Blobel G. Protein translocation across the endoplasmic reticulum. Cell 1984; 38:5. Wickner WT, Lodish HF. Multiple mechanisms of protein insertion into and across membranes. Science 1985; 230:400. Wiedmann M, Kurzchalia TV, Hartmann E, Rapoport TA. A signal sequence receptor in the endoplasmic reticulum membrane. Nature 1987; 328:830. Dunphy WG, Rothman JE. Compartmental organization of the Golgi stack. Cell 1985; 42:13. Mellman I, Warren G. The road taken: past and future foundations of membrane traffic. Cell 2000; 100:99. Griffiths G, Simons K. The trans Golgi network: sorting at the exit site of the Golgi complex. Science 1986; 234:438. Kornfeld R, Kornfeld S. Assembly of asparagine-linked oligosaccharides. Annu Rev Biochem 1985; 54:631. Dice JF. Molecular determinants of protein half-lives in eukaryotic cells. FASEB J 1987; 1:349. Sanders SL, Schekman R. Polypeptide translocation across the endoplasmic reticulum membrane. J Biol Chem 1992; 267:13791. Kornfeld S. Trafficking of lysosomal enzymes in normal and disease states. J Clin Invest 1986; 77:1. Munro S. Pelham HRB. A C-terminal signal prevents secretion of lumenal ER proteins. Cell 1987; 48:899. Munro S, Pelham HRB. An HSP70-like protein in the ER: identity with the 78 kd glucose-regulated protein and immunoglobulin heavy chain binding protein. Cell 1986; 46:291. Johnson LM, Bankaitis VA, Emr SD. Distinct sequence determinants direct intracellular sorting and modification for a yeast vacuolar protease. Cell 1987; 48:875. Valls LA, Hunter CP, Rothman JH, Stevens TH. Protein sorting in yeast: the localization determinant of yeast vacuolar carboxypeptide Y residues in the propeptide. Cell 1987; 48:887. Moore HH, Kelly RB. Re-routing of a secretory protein by fusion with human growth hormone sequences. Nature 1986; 321:443. Kelly RB, Grote E. Protein targeting in the neuron. Annu Rev Neurosci 1993; 16:95. Burgess TL, Kelly RB. Constitutive and regulated secretion of proteins. Annu Rev Cell Biol 1987; 3:243. Loh YP, Snell CR, Cool DR. Receptor-mediated targeting of hormones to secretory granules. Role of carboxypeptidase E. Trends Endocrinol Metab 1997; 8:130. Tooze J, Tooze SA. Clathrin-coated vesicular transport of secretory proteins during the formation of ACTH-containing secretory granules in AtT-20 cells. J Cell Biol 1986; 103:839. Orci L, Ravazzola M, Storch M-J, et al. Proteolytic maturation of insulin is a post-Golgi event which occurs in acidifying clathrin-coated secretory vesicles. Cell 1987; 49:865. Mellman I, Fuchs R, Helenius A. Acidification of the endocytic and exocytic pathways. Annu Rev Biochem 1986; 55:663. Hong W, Tang BL. Protein trafficking along the exocytotic pathway. Bioas-says 1993; 15:231.

60a. Gerdes HH, Glombik MM. Signal-mediated sorting to the regulated pathway of protein secretion. Anat Anz 1999; 181:447. 61. Inoue K, Kurosumi K. Ultrastructural immunocytochemical localization of LH and FSH in the pituitary of the untreated male rat. Cell Tissue Res 1984; 235:77. 61a. Gullberg U, Bengtsson N, Bulow E, et al. Processing and targeting of granule proteins in human neutrophils. J Immunol Methods 1999; 232:201. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. 91. 92. 93. DeLisle RC, Williams IA. Regulation of membrane fusion in secretory exocytosis. Annu Rev Physiol 1986; 48:225. Thomas G, Thorne BA, Hruby DE. Gene transfer technique to study neuropeptide processing. Annu Rev Physiol 1988; 50:323. Yamamoto KR. Steroid receptor regulated transcription of specific genes and gene networks. Annu Rev Genet 1985; 19:209. Shupnik MA, Chin WW, Habener JF, Ridgway EC. Transcriptional regulation of the thyrotropin subunit genes by thyroid hormone. J Biol Chem 1985; 260:2900. Larsen PR, Harney JW, Moore DD. Sequences required for cell type specific thyroid hormone regulation of rat growth hormone promoter activity. J Biol Chem 1986; 261:14373. Wright PA, Crew MD, Spindler SR. Discrete positive and negative thyroid hormone-responsive transcription regulatory elements of the rat growth hormone gene. J Biol Chem 1987; 262:5659. Flug F, Copp RP, Casanova J, et al. Cis-acting elements of the rat growth hormone gene which mediate basal and regulated expression by thyroid hormone. J Biol Chem 1987; 262:6373. Jantzen HM, Strahle U, Gloss B, et al. Cooperativity of glucocorticoid response elements located far upstream of the tyrosine aminotransferase gene. Nature 1987; 49:29. Weinberger C, Thompson CC, Ong ES, et al. The c-erb-A gene encodes a thyroid hormone receptor. Nature 1986; 324:64 1. Green S, Chambon P. A super family of potentially oncogenic hormone receptors. Nature 1986; 324:615. Brown DD. The role of stable complexes that repress and activate eucaryotic genes. Cell 1984; 37:359. von Hippel PH, Bear DG, Morgan WD, McSwiggen JA. Protein-nucleic acid interactions in transcription: a molecular analysis. Annu Rev Biochem 1984; 53:389. Harrison SC. A structural taxonomy of DNA-binding domains. Nature 1991; 353:715. Pabo CO. Transcription factors: structural families and principles of DNA recognition. Annu Rev Biochem 1992; 61:1053. Murdoch GH, Franco R, Evans RM, Rosenfeld RG. Polypeptide hormone regulation of gene expression. Thyrotropin-releasing hormone rapidly stimulates both transcription of the prolactin and the phosphorylation of a specific nuclear protein. J Biol Chem 1983; 258:15329. Montminy MR, Sevarino KA, Wagner JA, et al. Identification of a cyclic AMP responsive element within the rat somatostatin gene. Proc Natl Acad Sci U S A 1986; 83:6682. Hunter T, Karin M. The regulation of transcription by phosphorylation. Cell 1992; 70:375. Habener JF. Cyclic AMP response element binding proteins: a cornucopia of transcription factors. Mol Endocrinol 1990; 4:1087. Kadonaga JT, Tjian R. Affinity purification of sequence-specific DNA-binding proteins. Proc Natl Acad Sci U S A 1986; 83:5889. Brent R. Repression of transcription in yeast. Cell 1985; 42:3. Diamond MI, Miner JN, Yoshinaga SK, Yamamoto KR. Transcriptional factor interactions: selectors of positive and negative regulation from a single DNA element. Science 1990; 249:1266. Guarente L. Yeast promoters: positive and negative elements. Cell 1984; 36:799. Jones NC. Negative regulation of enhancers. Nature 1986; 321:202. Maniatis T, Goodbourn S, Fischer JA. Regulation of inducible and tissue-specific gene expression. Science 1987; 236:1237. Yanofsky C. Transcription attenuation. J Biol Chem 1988; 263:609. Brock ML, Shapiro DJ. Estrogen stabilizes vitellogenin mRNA against cytoplasmic degradation. Cell 1983; 34:207. Guyette WA, Matusik RJ, Rosen JM. Prolactin-mediated transcriptional and post-transcriptional control of casein gene expression. Cell 1979; 17:1013. Krane IM, Spindel ER, Chin WW. Thyroid hormone decreases the stability and the poly(A) tract length of rat thyrotropin -subunit messenger RNA. Mol Endocrinol 1991; 5:469. Koenig RJ, Lazar MA, Hodin RA, et al. Inhibition of thyroid hormone action by a non-hormone binding cerbA protein generated by alternative mRNA splicing. Nature 1989; 337:659. Chew SL. Alternative splicing of mRNA as a mode of endocrine regulation. Trends Endocrinol Metab 1997; 8:405. Douglass J, Civelli O, Herbert E. Polyprotein gene expression: generation of diversity of neuroendocrine peptides. Annu Rev Biochem 1984; 53:665. Wilson HE, White A. Prohormones: their clinical relevance. Trends Endocrinol Metab 1998; 9:396.

CHAPTER 4 HORMONAL ACTION Principles and Practice of Endocrinology and Metabolism

CHAPTER 4 HORMONAL ACTION

DARYL K. GRANNER General Features of Hormone Systems and Historical Perspective Target Cell Concept Hormone Receptors General Features Recognition and Coupling Domains of Receptors Receptor Occupancy and Bioeffect Agonist-Antagonist Concept Regulation of Receptors Structure of Receptors Classification of Hormones Mechanism of Action of Group I Hormones Mechanism of Action of Group II Hormones Chapter References

GENERAL FEATURES OF HORMONE SYSTEMS AND HISTORICAL PERSPECTIVE

Multicellular organisms use intercellular communication mechanisms to ensure their survival by coordinating the responses necessary for adjusting to constantly changing external and internal environments. Two systems comprising several highly differentiated tissues have evolved to serve these functions. One is the nervous system, and the other is the endocrine system, which classically has been viewed as using mobile hormonal messages that are secreted from one gland or tissue to act on a distant tissue. There is an exquisite convergence of these regulatory systems. For example, neural regulation of the endocrine system is important; many neurotransmitters resemble hormones in their synthesis, release, transport, and mechanism of action; and many hormones are synthesized in the nervous system (see Chap. 175). The focus of this chapter is the endocrine system and how hormones work. The word hormone is derived from a Greek term that means to arouse to activity. Classically defined, a hormone is a substance that is synthesized in one organ and transported by the circulatory system to act on another tissue. However, this original description is too restrictive, because hormones can act on adjacent cells (i.e., paracrine action) and on the cell in which they were synthesized (i.e., autocrine action) without entering the circulation. Early studies concentrated on defining the endocrine action of hormones by removing or ablating an organ to localize the site of production. An extract of the tissue was then used to restore the function, and this served as a bioassay for subsequent purification of the hormone and the elucidation of physiologic and biochemical actions. This classic era of the study of hormonal action was descriptive. During this period, many hormones were discovered, and their major effects were defined. Because it was assumed that hormones had a unique source and a single or predominant action, they were named for the tissue of origin (e.g., thyroid hormone) or for the action (e.g., growth hormone). The next era of investigation of hormonal action was characterized by the discovery of many more hormones and by a more detailed analysis of how hormones work. The investigation of their functions was aided by methods and ideas previously exploited by endocrinologists, including the use of radioisotopes, the concept of turnover, improved means of purifying molecules, and the availability of sophisticated analytic machinery. Such studies changed the direction of research in hormonal action from a descriptive (i.e., organ or tissue) to a mechanistic (i.e., molecule or function) approach. Where a molecule worked was no longer as important as how it acted. A single hormone could have hemocrine (i.e., transportation through the blood), paracrine, or autocrine actions but affect the different target cells in a similar way, and some effects could be produced by a variety of hormones. For example, naming a single molecule the growth hormone was incorrect, because this hormone is but one of severalincluding the thyroid hormones, sex hormones, glucocorticoids, insulin, and various growth-promoting polypeptidesthat are involved in growth, and growth promotion is only one of the actions of the so-called growth hormone. The principles of hormone synthesis, storage, secretion, transport, metabolism, and feedback control were established during this period. A major contribution was the elaboration of the concept of hormone receptors, and of the properties of specificity and selectivity of response, how target cells are defined, how responses are modulated, how signals are transduced from the outside of a cell to its interior, and how hormones can be classified according to their mechanism of action. The techniques of molecular biology and recombinant DNA have been applied to hormonal action with remarkable success. It is now possible to analyze hormonal effects on gene expression and to study which few nucleotides of the 3109 in each haploid genome confer the response. Another exciting area is the overlapping spectrum of activity of components of hormonal action systems with nonhormonal proteins. Consider the similar features of the guanosine triphosphate (GTP)binding proteins involved in the hormone-sensitive adenylate cyclase system with the transforming RAS oncogene family of proteins or with transducin, which is the protein that couples photoactivation to the visual response.1,2 The homology of platelet-derived growth factor (PDGF) gene and the v-sis transforming gene is remarkable, as is the similarity between the insulin and epidermal growth factor receptors, both of which have intrinsic tyrosine kinase activity.3,4,5 and 6 Researchers are exploring the molecular bases of endocrine diseases, such as pseudohypoparathyroidism, several types of dwarfism, Graves disease, certain types of extreme insulin resistance, testicular feminization, acromegaly, vitamin D resistance, and hereditary nephrogenic diabetes insipidus, to name a few.7,8,9,10,11,12,13 and 14 This knowledge has challenged many of the earlier concepts of hormonal action and endocrine disease.

TARGET CELL CONCEPT

There are ~200 types of differentiated cells in humans. Only a few produce hormones, but virtually all of the 75 trillion cells in a human body are targets of one or more of the ~50 known hormones. The concept of target cells is undergoing redefinition. It was thought that hormones affected a single cell type, or only a few kinds of cells, and that a hormone elicited a unique biochemical or physiologic action. For example, it was presumed that thyroid-stimulating hormone (TSH) stimulated thyroid growth and thyroid hormonogenesis; adrenocorticotropic hormone (ACTH, also called corticotropin) enhanced growth and function of the adrenal cortex; glucagon increased hepatic glucose production; and luteinizing hormone (LH) stimulated gonadal steroidogenesis. However, these same hormones also stimulate lipolysis in adipose cells.15 Although the physiologic importance of this effect is unclear, the concept of unique sites of actions of these hormones is untenable. A more relevant example is that of insulin, which effects various responses in different cells and occasionally influences different processes within the same cell. It enhances glucose uptake and oxidation in muscle, lipogenesis in fat, amino acid transport in liver and lymphocytes, and protein synthesis in liver and muscle. These and other examples necessitated a reevaluation of the target cell concept. With the delineation of specific cell-surface and intracellular hormone receptors, the definition of a target has been expanded to include any cell in which the hormone binds to its receptor, whether or not a biochemical or physiologic response has been determined. This definition also is imperfect, but it has heuristic merit, because it presumes that not all actions of hormones have been elucidated. The response of a target cell is determined by the differentiated state of the cell, and a cell can have several responses to a single hormone. Cells can respond to a hormone in a hemocrine, paracrine, or autocrine manner. An example is the hormone gastrin-releasing peptide (also called mammalian bombesin). Gastrin-releasing peptide has hemocrine and paracrine actions in the gut but is produced by and stimulates the growth of small cell carcinoma cells of the lung.16 Several factors determine the overall response of a target cell to a hormone. The concentration of a hormone around the target cell depends on the rate of synthesis and secretion of the hormone, the proximity of target and source, the association-dissociation constants of the hormone with specific plasma carrier proteins, the rate of conversion of an inactive or suboptimally active form of the hormone into the active form, and the rate of clearance of the hormone from blood by other tissues or by degradation or excretion. The actual response to the hormone depends on the relative activity and state of occupancy, or both, of the specific hormone receptors on the plasma membrane or within the cytoplasm or nucleus; the metabolism of the hormone within the target cell; the presence of other factors within the target cell that are necessary for the hormone response; and postreceptor desensitization of the cell. Alterations of any of these processes can change the hormonal effect on a given target cell and must be considered in addition to the classic feedback loops.

HORMONE RECEPTORS
GENERAL FEATURES

One of the major challenges in making the hormone-based communication system work is depicted in Figure 4-1. Hormone concentrations are very low in the extracellular fluid, generally in the range of 10-15 to 10-9 M. This is much lower than that of the many structurally similar molecules (e.g., sterols, amino acids, peptides) and other molecules that circulate at concentrations in the 10-5 to 10-3 M range. Target cells must identify the various hormones present in small amounts and differentiate a given hormone from the 106- to 109-fold excess of other, often closely related, molecules. This high degree of discrimination is provided by cell-associated recognition molecules called receptors. Hormones initiate their bioeffects by binding to specific receptors, and because any effective control system must provide a means of stopping a response, hormone-induced actions usually terminate after the effector dissociates from the receptor.

FIGURE 4-1. Specificity and selectivity of hormone receptors. Many different molecules circulate in the extracellular fluid (ECF), but only a few are recognized by hormone receptors. Receptors must select these molecules from among high concentrations of the other molecules. This simplified drawing shows that a cell may have no hormone receptors, have one receptor, have a receptor but no hormone in the vicinity, or have receptors for several hormones.

A target cell is defined by its ability to bind a given hormone selectively by means of a receptor, an interaction that is often quantitated using radioactive ligands that mimic hormone binding. Several features of this interaction are important. The radioactivity must not alter the bioactivity of the ligand. The binding should be specific, in which case the ligand is displaceable by unlabeled agonist or antagonist. Binding should be saturable. Binding should occur within the concentration range of the expected biologic response. RECOGNITION AND COUPLING DOMAINS OF RECEPTORS All receptors, whether for polypeptides or steroids, have at least two functional domains, and most have several more. A recognition domain binds the hormone, and a second region, the coupling domain, generates a signal that links hormone recognition to some intracellular function. The binding of hormone by receptor implies that some region of the hormone molecule has a conformation that is complementary to a region of the receptor molecule. The degree of similarity, or fit, determines the tightness of the association; this is measured as the affinity of binding. If the native hormone has a relative affinity of 1, other natural molecules range between 0 and 1. In absolute terms, this actually spans a binding affinity range of more than a trillion. Ligands with a relative affinity of more than 1 for some receptors have been synthesized and are used to study receptor biology. Coupling (i.e., signal transduction) occurs in two ways. Polypeptide and protein hormones, and the catecholamines, bind to receptors located in the plasma membrane, and thereby generate signals that regulate various intracellular functions. Steroids, thyroid hormones, retinoids, and other hormones of this class interact with intracellular receptors, and this complex provides the initial signal. The amino acid sequences of the recognition and coupling domains have been identified in many polypeptide hormone receptors. Hormone analogues with specific amino acid substitutions were used to change binding and alter the bioactivity of the hormone. Steroid hormone receptors also have these two functional domains; one site binds the hormone and the other binds to specific DNA regions. They also have other domains important for their function, which are described later. Several receptors have been characterized by recombinant DNA techniques, and structural analysis shows that these domains are highly homologous. This homology has been used to isolate cDNAs encoding several receptors that had not been obtained through classic protein purification procedures. The investigations have shown that these nuclear receptors are part of a large family of related proteins.17 This family of proteins is thought to regulate gene transcription, often in association with other transcription factors and coregulatory molecules. The ligands for these are called orphan receptors. The dual functions of binding and coupling ultimately define a receptor, and it is the coupling of hormone binding to signal transduction, called receptor-effector coupling, that provides the first step in the amplification of the hormonal response. This dual purpose also differentiates the target cell receptor from the plasma carrier proteins that bind hormone without generating a signal. It is important to differentiate the binding of hormones to receptors from the association that hormones have with various transport or carrier proteins. Table 4-1 lists several features of these functionally different classes of proteins.

TABLE 4-1. A Comparison of Hormone Receptors with Transport Proteins

RECEPTOR OCCUPANCY AND BIOEFFECT The concentrations of hormone required for occupancy of the receptor and for elicitation of a specific biologic response often are similar (Fig. 4-2A). This is especially true for steroid hormones, but some polypeptide hormones also exhibit this characteristic. This tight coupling is remarkable, considering the many steps that must occur between hormone binding and complex responses, such as transport, enzyme induction, cell lysis, or cell replication. When receptor occupancy and bioeffect are tightly coupled, significant changes in the latter occur when receptor occupancy changes. This happens when fewer receptors are available (Fig. 4-3A) or the affinity of the receptor changes but hormone concentration remains constant (see Fig. 4-3B). Otherwise, there is a marked dissociation of binding and effect, and a maximal bioeffect occurs when only a small percentage of the receptors are occupied (see effect 2 in Fig. 4-2B).

FIGURE 4-2. Hormone binding and biologic effect are compared in the absence (A) and presence (B, effect 2) of spare receptors. Some biologic effects in a tissue may be tightly coupled to binding, but others demonstrate the spare receptor phenomenon (e.g., compare effects 1 and 2 in B). (From Granner DK. Characteristics of hormone systems. In: Martin DW Jr, Mayer PA, Rodwell VW, Granner DK, eds. Harper's review of biochemistry, 20th ed. Los Altos, CA: Lange Medical Publications, 1985:501.)

FIGURE 4-3. Changes of receptor occupancy have large effects on the biologic response when effector and receptor occupancy are tightly coupled. This can occur when the receptor number changes (A) or when the affinity of the receptor for the hormone changes (B). In the hypothetical case shown in (A), a decrease from 20,000 receptors per cell to 10,000 results in a 50% decrease of the maximal response, a Vmax effect. A decrease in affinity (i.e., solid to interrupted line in [B], or rightward shift) means that more hormone is required for a given effect, but the same maximal response can be obtained. This is a Km effect. (From Granner DK. Characteristics of hormone systems. In: Murray RK, Granner DK, Mayer PA, Rodwell VW, eds. Harper's biochemistry, 21st ed. Norwalk, CT: Appleton & Lange, 1988.)

Receptors not involved in the elicitation of the response are called spare receptors. They are observed in the response of several polypeptide hormones and are thought to provide a means of increasing the sensitivity of a target cell to activation by low concentrations of hormone and to provide a reservoir of receptors. The concept of spare receptors is operational and may depend on which aspect of the response is examined and which tissue is involved. For example, there is excellent agreement between LH binding and cyclic adenosine monophosphate (cAMP) production in rat testis and ovarian granulosa cells (there generally are no spare receptors when any hormone activates adenylate cyclase), but steroidogenesis in these tissues, which is cAMP dependent, occurs when fewer than 1% of the receptors are occupied (see effects 1 and 2 in Fig. 4-2).18 Transcription of the phosphoenolpyruvate carboxykinase gene is repressed when far fewer than 1% of hepatoma cell insulin receptors are occupied, but there is a high correlation between insulin binding and amino acid transport in thymocytes.19 Other examples of the dissociation of receptor binding and biologic effects include the effects of catecholamines on muscle contraction, lipolysis, and ion transport.20 These end-responses presumably reflect a cascade or multiplier effect of the hormone. Different responses within the same cell can require various degrees of receptor occupancy. For example, successively greater degrees of occupancy of the adipose cell insulin receptor increase, in sequence, lipolysis, glucose oxidation, amino acid transport, and protein synthesis.21 AGONIST-ANTAGONIST CONCEPT Molecules can be divided into four groups according to their ability to elicit a hormone receptormediated response. These classes are agonists, partial agonists, antagonists, and inactive agents (Table 4-2).

TABLE 4-2. Classification of Steroids According to Their Action as Glucocorticoids

Agonists elicit the maximal response, although different concentrations may be required. In the example of Figure 4-4,1,2 and 3 could be porcine insulin, porcine proinsulin, and guinea pig insulin, respectively. In all systems tested, these insulins have the same rank order of potency, but each elicits a maximal response if present in sufficient concentration. Likewise, 1, 2, and 3 could be dexamethasone, cortisol, and corticosterone (see Table 4-2).Partial agonists evoke an incomplete response even when very large concentrations of the hormone are used, as shown by line B of Figure 4-5. Antagonists generally have no effects themselves, but they competitively inhibit the action of agonists or partial agonists (see lines A through C in Fig. 4-5). Many structurally similar compounds elicit no effect and have no effect on the action of the agonists or antagonists. These are classified as inactive agents and are represented as line D in Figure 4-5.

FIGURE 4-4. Within a class of hormonesglucocorticoids, for exampledifferent molecules may have different potencies. In this case, hormones 1, 2, and 3 are all agonists, but very different concentrations are required to achieve a given biologic response. The binding of steroid to receptor would parallel each of these curves. (From Granner DK. Characteristics of hormone systems. In: Murray RK, Granner DK, Mayer PA, Rodwell VW, eds. Harper's biochemistry, 21st ed. Norwalk, CT: Appleton & Lange, 1988.)

FIGURE 4-5. Classification of hormones according to their biologic activity. Steroids, for example, can be classified as agonists (line A), partial agonists (line B), antagonists (C in A+C or B+C), or inactive agents (dotted line D). This drawing represents induction of the enzyme tyrosine aminotransferase. (From Granner DK. Characteristics of hormone systems. In: Murray RK, Granner DK, Mayer PA, Rodwell VW, eds. Harper's biochemistry, 21st ed. Norwalk, CT: Appleton & Lange, 1988.)

Partial agonists also compete with agonists for binding to and activation of the receptor, when they become partial antagonists. The extent of the inhibition of agonist activity caused by partial or complete antagonists depends on the relative concentration of the various steroids. Generally, much higher concentrations of the antagonist are required to inhibit an agonist than are necessary for the latter to exert its maximal effect. Because these concentrations are rarely achieved in vivo, this phenomenon is used for studies of the mechanism of action of hormones in vitro. The binding of a ligand to the receptor must facilitate a change in this molecule so that it can bind to DNA. This phenomenon was first suggested in studies that used the steroids in Table 4-2.22 The hypothesis assumes that agonists bind to and fully activate the receptor and elicit the maximal biologic response; that partial agonists fully occupy the receptor but afford incomplete activation and therefore a partial response; and that antagonists fully occupy the receptor, but because this complex is unable to bind to DNA, it elicits no intrinsic response but does inhibit the action of agonists. REGULATION OF RECEPTORS The number of hormone receptors on or in a cell is in a dynamic state and can be regulated physiologically or be influenced by diseases or therapeutic measures. The receptor concentration and affinity of hormone binding can be regulated. Some changes can be acute and can significantly affect hormone responsiveness of the cell. For instance, cells exposed to b-adrenergic agonists for minutes to hours no longer activate adenylate cyclase in response to more agonist, and the biologic response is lost. This desensitization occurs by two mechanisms.23 The loss of receptors, called down-regulation, involves the internal sequestration of receptors, segregating them from the other components of the response system, including the regulatory and catalytic subunits of adenylate cyclase. Removal of the agonist results in the return of receptors to the cell surface and restoration of hormonal sensitivity.23 An example of a second form of desensitization of the a-adrenergic system involves the covalent modification of receptor by phosphorylation.24 This cAMP-dependent process entails no change in receptor number and no translocation. Reconstitution experiments show that because the phosphorylated receptor is unable to activate adenylate cyclase, the activation and hormone binding domains are uncoupled.23 Other examples of physiologic adaptation that is accomplished through down-regulation of receptor number by the homologous hormone include insulin, glucagon, thyrotropin-releasing hormone, growth hormone, LH, follicle-stimulating hormone, and catecholamines. A few hormones, such as angiotensin II and prolactin, up-regulate their receptors. The changes in receptor number can occur over a period of minutes to hours and are probably an important means of regulating biologic responses. How the loss of receptor affects the biologic response elicited at a given hormone concentration depends on whether there are spare receptors (Fig. 4-6). Suppose there is a fivefold reduction in receptor number in a cell. With no spare receptors (see Fig. 4-6A), the maximal response obtained is 20% that of control, hence, the effect is on the Vmax. With spare receptors (see Fig. 4-6B), the maximal response is obtained, but at five times the originally effective hormone concentration, analogous to a Km effect.

FIGURE 4-6. The effect a five-fold loss of receptors has on a biologic system that lacks (A) or has (B) spare receptors. (From Granner DK. Characteristics of hormone systems. In: Martin DW Jr, Mayer PA, Rodwell VW, Granner DK, eds. Harper's review of biochemistry, 20th ed. Los Altos, CA: Lange Medical Publications, 1985:502.)

STRUCTURE OF RECEPTORS The acetylcholine receptor (AChR), which exists in relatively large amounts in the electric organ of Torpedo californica, was the first plasma membraneassociated receptor to be studied in detail. The AChR consists of four subunits: a2, b, d and g.25 The two a subunits bind acetylcholine.26 The technique of site-directed mutagenesis has been used to show which regions of this subunit participate in the formation of the transmembrane ion channel, which is the major function of the AChR.25 Other receptors occur in very small amounts, and recombinant DNA techniques have been used to deduce many of the structures and to find and characterize new receptors. The insulin receptor is a heterotetramer (a2b2) linked by multiple disulfide bonds, in which the extramembrane a subunit binds insulin and the membrane-spanning b subunit transduces the signal through the tyrosine kinase component of the cytoplasmic portion of this polypeptide27 (Fig.4-7). The insulin-like growth factor-I (IGF-I) receptor has a similar structure, and the epidermal growth factor (EGF) and low-density lipoprotein receptors are similar in many respects28,29 and 30 (see Fig. 4-7). Receptors that couple ligand binding to signal transduction through G-protein intermediaries characteristically have seven membrane-spanning domains.31

FIGURE 4-7. Schematic representation of the structures of the low-density lipoprotein (LDL), epidermal growth factor (EGF), and insulin receptors. The amino terminus (NH2) of each is in the extracellular portion of the molecule. The carboxyterminus (COOH) is in the cytoplasm. The open boxes represent cysteine-rich regions that are thought to be involved in ligand binding. Each receptor has a short domain (~25 amino acids) that traverses the plasma membrane (hatched line) and an intracellular domain of variable length. The EGF and insulin receptors have tyrosine kinase activity associated with the cytoplasmic domain ( ) and have autophosphorylation sites in this region. The insulin receptor is a heterotetramer connected by disulfide bridges (vertical bars).

Members of the nuclear receptor superfamily have several functional domains: a ligand-binding domain in the carboxyl-terminal region, an adjacent DNA-binding domain, and one or more trans-activation domains. There may also be dimerization, nuclear translocation, and heat shock protein domains, and regions that allow for interactions with a number of other accessory factor and coregulatory proteins17,32 (Fig. 4-8). The amino acid sequence homology is particularly strong in the various DNA-binding domains, and it was this feature that led to the elucidation of the nuclear receptor superfamily.17

FIGURE 4-8. Nuclear receptor family members have several general domains. The amino-terminal region is most variable and often contains a trans -activating domain (TAD1). The DNA-binding domain (DBD) is most conserved, and this feature led to the discovery that these receptors are part of a large family of DNA-binding proteins. The hormone-or ligand-binding domain (LBD), which affords specificity, is located in the carboxyl-terminal (COOH) region of the molecule and contains a second trans -activating domain (TAD2). Also shown are regions that allow for nuclear translocation, dimerization, and interaction with heat shock protein (Hsp90). Members of this family that have no known ligand are called orphan receptors.

CLASSIFICATION OF HORMONES
A classification based on the location of receptors and the nature of the signal used to mediate hormonal action within the cell appears in Table 4-3, and general features of each group are listed in Table 4-4.

TABLE 4-3. Hormones and Their Actions: Classification According to Mechanism of Action

TABLE 4-4. General Features of Hormone Groups

The hormones in group I are lipophilic. After secretion, these hormones associate with transport proteins, a process that circumvents the solubility problem while prolonging the plasma half-life by preventing the hormone from being metabolized and excreted. These hormones readily traverse the plasma membrane of all cells and encounter receptors in the cytosol or the nucleus of target cells. The ligand-receptor complex is assumed to be the intracellular messenger in this group. The second major group consists of water-soluble hormones that bind to the plasma membrane of the target cell. These hormones regulate intracellular metabolic processes through intermediary molecules, called second messengers (the hormone itself is the first messenger), which are generated because of the ligand-receptor interaction. The second-messenger concept arose from the observation of Sutherland33 that epinephrine binds to the plasma membrane of pigeon erythrocytes and increases intracellular cAMP. This was followed by a series of experiments in which cAMP was found to mediate the metabolic effects of many hormones. Hormones that use this mechanism are shown in group IIA. Several hormones, some of which were previously thought to affect cAMP, appear to use cyclic guanosine

mono-phosphate (cGMP) (group IIB) or calcium or phosphatidylinositide metabolites (or both) as the intracellular signal (group IIC). The intracellular messenger has been identified as a protein kinase/phosphatase cascade for the hormones listed in group D. A few hormones fit in more than one category (i.e., some hormones act through cAMP and Ca2+), and assignments change with new information. MECHANISM OF ACTION OF GROUP I HORMONES A schematic representation of the mechanism of action of group I hormones (see Table 4-3) is shown in Figure 4-9. These lipophilic molecules probably diffuse through the plasma membrane of all cells but encounter their specific, high-affinity receptor only within target cells. The hormone-receptor complex then undergoes an activation reaction that causes size, conformation, and surface charge changes that render it able to bind to chromatin. In some caseswith the glucocorticoid receptor, for examplethis process involves the disruption of a receptorheat shock protein complex. Whether the association and activation processes occur in the cytoplasm or nucleus appears to depend on the specific hormone. The hormone-receptor complex binds to specific regions of DNA and activates or inactivates specific genes.34,35 By selectively affecting gene transcription and the production of the respective messenger RNAs (mRNAs), the amounts of specific proteins are changed, and metabolic processes are influenced. The effect of each of these hormones is specific; generally, the hormone affects <1% of the proteins or mRNAs in a target cell.

FIGURE 4-9. A general model of group I hormone action. The hormone binds to intracellular receptors in the cytoplasm or the nucleus and causes a conformational change. The hormone-receptor complex then binds to a specific region on DNA called the hormone response element. This interaction, with the help of various accessory factors and coregulators, results in the activation or repression of a restricted number of genes. The hormone response elements and associated factor elements are called hormone response units. The bubble indicates that this region of DNA is open or accessible to the transcription complex. These regions of DNA are often found to be sensitive to digestion by the enzyme DNase I.

The nuclear actions of steroid hormones predominate and are well defined, but direct actions of these hormones in the cytoplasm and on various organelles and membranes also have been described.36 An effect of estrogens, cAMP, and glucocorticoids on mRNA degradation rates has been demonstrated.37,38,39 and 40 Glucocorticoids also affect posttranslational processing of some proteins.41 Although the biochemistry of gene transcription in mammalian cells is not completely understood, a general model of the structural requirements of steroid regulation of gene transcription can be drawn (Fig. 4-10). Steroid-regulated genes must be in regions of open or transcriptionally active chromatin (depicted as the bubble in Fig. 4-9), as defined by their susceptibility to digestion by the enzyme DNase I.42 The open or closed conformation of chromatin may be regulated by the extent of acetylation of the histones that combine with DNA to form chromatin (discussed later this section). Genes have at least two separate regulatory regions in the DNA sequence immediately 5' of the transcription initiation site. The first of these, the basal promoter element (BPE), is generic, because it is present in some form in all genes.43 This is depicted as containing the consensus sequence GTATA (A/T)A(A/T), called the TATA box (see Chap. 3), because this is the structure found most frequently. Another common component of the BPE is the CAAT box; this sequence or some equivalent structure usually is present. The BPE appears to specify the site of RNA polymerase II attachment to DNA and therefore the accuracy of transcript initiation.44

FIGURE 4-10. Structural requirements for hormonal regulation of gene transcription. Transcription starts at the arrow, where 1+ signifies the first nucleotide of the gene that is transcribed or copied. Immediately adjacent, on the 5' upstream side, is the basal promoter element, which generally consists of a TATA box and other components, such as a CAAT box. Hormone response elements (which bind the liganded receptor) can be anywhere in the 5' region or in the gene itself. Other DNA elements (accessory factor elements) that cooperate with the hormone response elements to regulate transcription can also be located at various sites. Together these form a hormone response unit.

A second regulatory region is located slightly farther upstream than the BPE, and this may also consist of several discrete elements. This region modulates the frequency of transcript initiation and is less dependent on position and orientation. In these respects, it resembles the transcription enhancer elements found in other genes.45,46 The regulatory region consists of two types of DNA elements in genes that respond to hormones. Hormone response elements (HREs) are short segments of DNA that bind a specific hormone receptor/ligand complex.32,34,35 HREs are often capable of regulating transcription from test promoter/reporter gene constructs, but in most physiologic circumstances, other DNA element/protein complexes are required. The HRE usually is found within 250 nucleotides of the transcription initiation site, but the precise location of the HRE varies from gene to gene. Identification of an HRE requires that it bind the hormone-receptor complex more avidly than does surrounding DNA or DNA from another source. The HRE also must confer hormone responsiveness. Putative regulatory sequence DNA can be ligated to reporter genes to assess this point. Usually, these fusion genes contain reporter genes not ordinarily influenced by the hormone, and these genes often are not expressed in the tissue being tested. Commonly used reporter genes are firefly luciferase or bacterial chloramphenicol acetyltransferase. The fusion gene is transfected into a target cell, and if the hormone now regulates the transcription of the reporter gene, an HRE is functionally defined. Position, orientation, and base-substitution effects can be precisely described using this technique. Although HREs do transmit a hormone response in simple promoter test conditions, the situation is much more complex in most naturally occurring genes. The HRE must interact with other elements (and associated binding proteins) to function optimally. Such assemblies of cis-acting DNA elements and trans acting factors are called hormone response units (HRUs).45 An HRU, therefore, consists of one or more HREs and one or more DNA elements with associated accessory factors (Fig. 4-11). In complex promotersregulated by a variety of hormonescertain accessory factor components of one HRU (glucocorticoid) may be part of that for another (retinoic acid). This arrangement may provide for the hormonal integration of complex metabolic responses.

FIGURE 4-11. The hormone response unit (HRU). The HRU is an assembly of DNA elements and bound proteins. An essential component is the hormone response element with ligand-bound receptor (R). Also important are the accessory factor (AF) elements with bound transcription factors. More than two dozen of these accessory factors have been linked to hormone effects on transcription. The AFs can interact with each other or with the nuclear receptors. The components of the HRU communicate with the basal transcription machinery through a coregulator complex. The components of the coregulator complex, some of which are described in the text, provide for the direction and specificity of the hormone response. (From Granner DK. Hormone action. In: Murray RK, Granner DK, Mayes PA, Rodwell VW, eds. Harper's biochemistry, 25th ed. Nor-walk, CT: Appleton & Lange, 1999.)

The communication between an HRU and the basal transcription apparatus is accomplished by one or more of a class of coregulator molecules (see Fig. 4-11). The first of these described was the cAMP response element binding (CREB) protein, so-called CBP. CBP, through an amino terminal domain, binds to phosphorylated serine 137 of CREB and facilitates transactivation in response to cAMP. It thus is described as a coactivator. CBP and its close relative, p300, interact with a number of signaling molecules, including activator protein-1, signal transducers and activators of transcription, nuclear receptors, and CREB.46 CBP/p300 also binds to the p160 family of coactivatorsdescribed in the next paragraphand to a number of other proteins. It is important to note that CBP/p300 also has intrinsic histone acetyltransferase (HAT) activity. The importance of this is described later in this section. Some of the many actions of CBP/p300 appear to depend on intrinsic enzyme activities and the ability of this protein to serve as a scaffold for the binding of other proteins. Three other families of coactivator molecules, all of ~160 kDa, have been described. These members of the p160 family of coactivators include (a) SRC-1 and NCoA-1; (b) GRIP 1, TIF2, and NCoA-2; and (c) p/CIP, ACTR, AIB1, RAC3, and TRAM-1.47 The different names for members within a subfamily often represent species variations or minor splice variants. There is ~35% amino acid identity between members of the different subfamilies. The role of these many coactivators is still evolving. It appears that certain combinations are responsible for specific ligand-induced actions through various receptors. The role of HAT is particularly interesting. Mutations of the HAT domain disable many of these transcription factors. Current thinking holds that these HAT activities acetylate histones and result in the remodeling of chromatin into a transcription-efficient environment.48 In keeping with this hypothesis, histone deacetylation is associated with the inactivation of transcription. In certain instances, the removal of a corepressor complex through a ligand-receptor interaction results in the activation of transcription. For example, in the absence of hormone, the thyroid or retinoic acid receptors are bound to a corepressor complex containing N-CoR or SMRT and associated proteins, some of which have histone deacetylase activity.47 The target gene is repressed until the binding of hormone to the thyroid receptor results in the dissociation of this complex, and gene activation then ensues. MECHANISM OF ACTION OF GROUP II HORMONES Most group II hormones are water soluble, have a short plasma half-life and no transport proteins, and initiate a response by binding to a receptor located in the plasma membrane (see Table 4-3 and Table 4-4). CYCLIC ADENOSINE MONOPHOSPHATE AS THE SECOND MESSENGER Cyclic AMP (3',5' adenylic acid), a ubiquitous nucleotide derived from adenosine triphosphate (ATP) through the action of the enzyme adenylate cyclase (Fig. 4-12), plays a crucial role in the action of several hormones. The intracellular level of cAMP is increased or decreased by various hormones (see Table 4-3). This effect varies from tissue to tissue and sometimes within a given tissue, depending on which specific hormone-receptor interactions occur. For example, epinephrine causes large increases of cAMP in muscle and relatively small changes in liver; the opposite is true of glucagon. Tissues that respond to several hormones of this group do so through unique receptors converging on a single class of adenylate cyclase molecules. The best example of this is the adipose cell in which epinephrine, ACTH, TSH, glucagon, LH, melanocyte-stimulating hormone, and vasopressin stimulate adenylate cyclase and increase cAMP.15 Combinations of maximally effective concentrations are not additive, and treatments that destroy one receptor response have no effect on the response of other hormones. Some actions of cAMP occur outside the nucleus, presumably by mediating changes in the degree of phosphorylation of critical enzymes involved in metabolic processes, such as glycogenolysis, gluconeogenesis, and lipolysis.49 Effects of cAMP on the transcription of many genes have been described, and this clearly is a major action of this molecule.50

FIGURE 4-12. The formation and metabolism of cyclic adenosine monophosphate (AMP). (ATP, adenosine triphosphate.)

ADENYLATE CYCLASE SYSTEM The components of this system in mammalian cells are shown in Figure 4-13. The interaction of the hormone with its receptor causes the activation or inactivation of adenylate cyclase. This process is mediated by two GTP-dependent regulatory protein complexes, designated Gs (stimulatory) and Gi (inhibitory), each of which is composed of three subunits: a, b, and g (see Fig. 4-13). These G proteins are part of a large family with more than 20 members that mediate many biologic processes in addition to hormone responses51 (Table 4-5). Adenylate cyclase, located on the inner surface of the plasma membrane, catalyzes the formation of cAMP from ATP in the presence of magnesium (see Fig. 4-12).

FIGURE 4-13. Components of the hormone receptorG-protein effector system. Receptors that couple to effectors through G proteins typically have seven membrane-spanning domains. The amino (N) and carboxy (C) termini are shown. In the absence of hormone (left), the heterotrimeric G-protein complex (i.e., a, b, g) is in an inactive, guanosine diphosphate (GDP)bound form and is probably not associated with the receptor. This complex is anchored to the plasma membrane through prenylated groups on the bg subunits (wavy lines) and perhaps by myristoylated groups on a subunits. On binding of hormone to the receptor, there is a presumed conformational change of the receptor and activation of the G-protein complex. This results from the exchange of GDP and guanosine triphosphate (GTP) on the a subunit, after which a and bg dissociate. The a subunit binds to and activates the effector (E). E can be adenylate cyclase (as), a K+ channel (a1 a0), phospholipase C-b(aq) or other molecule. The bg subunit can also have direct actions on E.

TABLE 4-5. Classes and Functions of G Proteins

What was originally considered a single protein with hormone binding and catalytic domains was found to be a system of extraordinary complexity. Biochemical and genetic studies have established the biochemical uniqueness of the hormone-receptor, GTP-regulatory, and catalytic domains of the adenylate cyclase complex. Two parallel systems, stimulatory (s) and inhibitory (i), converge on the catalytic molecule (C). Each system consists of a receptor, Rs or Ri, and a regulatory complex, Gs or Gi. Gs and Gi are trimers composed of a, b, and g subunits (see Fig. 4-13) and are so called because they bind and hydrolyze GTP. The as and aiare unique proteins of 39 to 46 kDa; the b subunits are 37-kDa proteins, and the g subunits are 8-kDa proteins.51 The binding of a hormone to Rs or Ri yields a receptor-mediated activation of G, which requires Mg2+-dependent guanosine diphosphate (GDP)/GTP exchange by a and the subsequent dissociation of b and g from a. The as subunit has GTPase activity; the active form, as GTP, is inactivated on hydrolysis of the GTP to GDP, and the trimeric Gs complex is reformed.52 Cholera toxin, an irreversible activator of cyclase, causes adenosine diphosphate (ADP) ribo can stimulate adenylate cyclase directly, and in some instances, the bg complex, which is tightly associated and always acts as a unit, augments this action. Less is known about how hormones inhibit adenylate cyclase activity. Direct inhibitory effects of Gi protein on adenysylation of as on arginine at position 201 in the protein, and in so doing, inactivates the GTPase; as remains in the active form.53,54 The ai also has GTPase activity; however, because GDP does not freely dissociate from ai the latter is reactivated by an exchange of GTP for GDP. Pertussis toxin irreversibly activates adenylate cyclase by promoting the ADP ribosylation ofai on a cysteine residue, which prevents this subunit from being activated by ligand-bound receptor.55 The exact mechanism of activation and inactivation of the adenylate cyclase moiety has not been established.55a The as form can stimulate adenylate cyclase directly, and in some instances, the bg complex, is tightly associated and always acts as a unit, augments this action. Less is known about how hormones inhibit adenylate cyclase activity. Direct inhibitory effects of Gi protien on adenylate cyclase have been difficult to detect. One hypothesis is that the bg complex that is liberated from Gi, which exists in most cells in great abundance over Gs, could bind to and inactivate as. This would effectively shut off the stimulatory signals. Alternatively,bg may directly inhibit some forms of adenylate cyclase. The a subunits and the bg complex have actions independent of those on adenylate cyclase. Some forms of a s stimulate Ca2+ channels and inhibit K+ channels. Similarly, some forms of ai stimulate K+ channels and inhibit Ca2+ channels. Other subunits, particularly those of the Gq family (see Table 4-5), are involved in the activation of members of the phospholipase C group of enzymes. This is important in the generation of the intracellular signals of inositol triphosphate and diacylglycerol. The bg complex also can stimulate K+ channels and activate certain isoforms of phospholipase C.51 Members of the family of G proteins mediate a variety of important processes (see Table 4-5). Transducin, the protein that is important in coupling light to photoactivation in the retina, is a member of the G-protein family, as are the proteins involved in smell and taste.56 The products of the RAS oncogenes, which are involved in regulating cell growth, are members of the larger superfamily.57 The G proteins are themselves a family within the large superfamily of GTPases and can be classified according to sequence homology into at least four subfamilies, as illustrated in Table 4-5. There are more than 20 a subunits, and there are at least four b subunits and six g subunits. There are at least five adenylate cyclase molecules.58 There may well be more members of each group. Pseudohypoparathyroidism, an experiment of nature, is a syndrome characterized by hypocalcemia and hyperphos-phatemia, the biochemical hallmarks of hypoparathyroidism, and by several congenital defects (see Chap. 60). Individuals with pseudohypoparathyroidism do not have defective parathyroid function; they secrete large amounts of bioactive parathyroid hormone. Some have target organ resistance on the basis of a postreceptor defect. They are partially deficient in G protein (probably only the as subunit) and fail to couple binding to adenylate cyclase stimulation.58,59 The observation that patients with pseudohypoparathyroidism often have defective responses to other hormones, including TSH, glucagon, and b-adrenergic agents, is not surprising.60 Many G-proteinlinked endocrinopa-thies have now been described (see Chap. 60). (See Chap ref 60a.) Approximately 40% of persons with acromegaly have a G-proteinlinked disease. These individuals have one of two mutations in the as subunit that affects the intrinsic GTPase activity of this protein. One mutation, at arginine position 201, affects the same site that is ADP ribosylated by cholera toxin.12 The constitutive overproduction of cAMP results in excessive release of growth hormone and somatotropic cell adenomas. In this way, as-subunit functions as an oncogene. Other a-subunit mutations that affect the GTPase domain result in tumors of the adrenal cortex and ovary.61 CYCLIC ADENOSINE MONOPHOSPHATEDEPENDENT PROTEIN KINASE In prokaryotic cells, cAMP binds to a specific protein, catabolite regulatory protein, which binds directly to DNA and influences gene expression.62 The analogy of this to steroid hormone action is apparent. In eukaryotic cells, cAMP binds to protein kinase A (PKA), which is a heterotetrameric molecule consisting of two regulatory subunits (R) and two catalytic subunits (C).63 Cyclic AMP binds to the regulatory subunits and yields the following reaction:

The R2C2 complex has no enzymatic activity, but the binding of cAMP by R dissociates R from C, thereby activating the latter. The active C subunit catalyzes the transfer of the g-phosphate of ATP (Mg2+) to a serine or threonine residue in various proteins. The consensus phosphorylation sites of PKA are -Arg-Arg-X-Ser- and -Lys-Arg-X-X-Ser-, in which X can be any amino acid.64 Protein kinase activities originally were described as cAMP dependent or cAMP independent. This area has also become considerably more complex, because protein phosphorylation is recognized as a general regulatory mechanism. Dozens of protein kinases have been described. All are unique molecules and show considerable variability with respect to subunit composition, molecular weight, autophosphorylation, Km for ATP, and substrate specificity.65 PHOSPHOPROTEINS The effects of cAMP in eukaryotic cells are thought to be mediated by protein phosphorylation and dephosphorylation.66 The effect of cAMP, including such diverse processes as steroido-genesis, secretion, ion transport, carbohydrate and fat metabolism, enzyme induction, gene regulation, and cell growth and replication, could be conferred by a specific protein kinase, a specific phosphatase, or by specific substrates for phosphorylation. Many substrates have been identified. For example, the transcription factor, CREB, mediates many of the effects of cAMP on gene transcription.67,68 CREB binds to the cAMP response element in the unphosphorylated state but is much more active in stimulating transcription after it has been phosphorylated by PKA.69 Phosphorylation on serine 133 allows CREB to bind to the coactivator CBP.46 This complex is associated with enhanced rates of transcription of target genes. PHOSPHODIESTERASES AND PHOSPHOPROTEIN PHOSPHATASES Reactions caused by hormones in class IIA can be terminated in several ways, including the hydrolysis of cAMP by phosphodesterases.70 The presence of these hydrolytic enzymes ensures a rapid turnover of the signal (i.e., cAMP), and there is a rapid termination of the biologic process after the hormonal stimulus is removed. The cAMP phosphodiesterases exist in low- and high-Km forms and are themselves subject to regulation by hormones and by intracellular messengers such as calcium, probably acting through calmodulin.71,72 and 73 Inhibitors of phosphodiesterase, most notably xanthine derivatives, increase intracellular cAMP and mimic or prolong the actions of hormones. Another means of controlling hormonal action is the regulation of the protein dephosphorylation reaction. There are two classes of phosphoprotein phosphatases. One group attacks phosphate on threonine or serine residues, another attacks phosphotyrosine. The threonine or serine protein phosphatases have been divided into two groups. Type 1 enzymes dephosphorylate the b subunit of phosphorylase kinase and are inhibited by small heat- and acid-stable proteins (i.e., inhibitors 1 and 2). The type 2 enzymes dephosphorylate the a subunit of phosphorylase kinase preferentially and are insensitive to the inhibitor proteins.74 There are a few members of each group, but there are many more protein kinases than there are phosphatases. The regulatory role of phosphatases has been best worked out in the case of glycogen metabolism.75 Certain phosphatases may attack specific residues; for example, some phosphatases remove phosphate from tyrosine residues.76 EXTRACELLULAR CYCLIC ADENOSINE MONOPHOSPHATE Some cAMP leaves cells and can be readily detected in extracellular fluids. The actions of glucagon on liver and of vasopressin or parathyroid hormone on the kidney are reflected in elevated levels of cAMP in plasma and urine, respectively.77 This led to diagnostic tests of target organ responsiveness. Extracellular cAMP has little or no bioactivity in mammals, but it is an extremely important intercellular messenger in lower eukaryotes and prokaryotes. HORMONES THAT ACT THROUGH CYCLIC GUANOSINE MONOPHOSPHATE Cyclic GMP is made from GTP by the enzyme guanylate cyclase, which exists in soluble and membrane-bound forms.78,79 and 80 Each of these isozymes has unique kinetic, physiochemical, and antigenic properties. For some time, cGMP was thought to be the functional counterpart of cAMP, but it became apparent that cGMP had a unique place in hormonal action. The atriopeptins, a family of peptides produced in cardiac atrial tissues, cause natriuresis, diuresis, vasodilation, and inhibition of aldosterone secretion (see Chap. 178). These peptides, such as atrial natriuretic factor, bind to and activate the membrane-bound form of guanylate cyclase. This causes an increase of cGMP, as much as 50-fold in some cases, which is thought to mediate these effects. Other evidence links cGMP to vasodilation. A series of compounds, including nitric oxide, nitroprusside, nitroglycerin, sodium nitrite, and sodium azide, cause smooth muscle relaxation and are potent vasodilators. These agents increase cGMP by activating nitric oxide synthase, which in turn produces nitric oxide. Nitric oxide activates the soluble form of guanylate cyclase. Inhibitors of cGMP phosphodiesterase enhance and prolong these responses. The increased cGMP activates cGMP-dependent protein kinase, which phosphorylates several smooth muscle proteins, including the myosin light chain. Presumably, this is involved in relaxation of smooth muscle and vasodilation. A cGMP phosphodiesterase attenuates these responses. Sildenafil (Viagra) is a potent phosphodiesterase inhibitor. The use of this compound in erectile dysfunction is based on the fact that it prolongs the accumulation and action of cGMP on penile smooth musculature. HORMONES THAT ACT THROUGH CALCIUM AND PHOSPHATIDYLINOSITIDES Ionized calcium is an important regulator of various cellular processes, including muscle contraction, stimulus-secretion coupling, the blood clotting cascade, enzyme activity, and membrane excitability. It also is an intracellular messenger of hormonal action.81,82 Calcium Metabolism. The extracellular calcium concentration is ~1.2 mM and is rigidly controlled. The intracellular free concentration of this ion (Ca2+) is much lower, ~100 to 200 nM, and the concentration associated with intracellular organelles is in the range of 1 to 20 M. Despite this 5000- to 10,000-fold concentration gradient and a favorable transmembrane electrical gradient, Ca2+ is restrained from entering the cell. Resting (basal) Ca2+ concentrations are determined by the activities of several Ca2+ exchangers and/or Ca2+ pumps located in the plasma membrane and endoplasmic reticulum membrane. Hormones can raise cytosolic Ca2+ via three mechanisms: (a) activation of ligand-gated Ca2+ channels in plasma membrane; (b) depolarization of the plasma membrane that activates inward-directed voltage-gated Ca2+ channels in the plasma membrane; and (c) activation of phospholipase c (PLC) through a Gq (G protein) mechanism, or by tyrosine phosphorylation, to generate inositol trisphosphate (IP3), which stimulates release from intracellular Ca2+ reservoirs. Ca2+ concentrations are rapidly returned to the basal concentrations by a reversal of one or more of these mechanisms, as a sustained elevation of Ca2+ is toxic to cells. Two observations led to the current understanding of how Ca2+ serves as an intracellular messenger of hormonal action. The first was the ability to quantitate the rapid changes of intracellular Ca2+ concentration that are implicit in a role for Ca2+ as an intracellular messenger. Such evidence was provided by various techniques, including the use of quin-2 or fura-2, fluorescent Ca2+ chelators.83 Rapid changes of Ca2+ in the submicroolar range can be quantitated using these compounds. The second important observation linking Ca2+ to hormonal action involved the definition of the intracellular targets of Ca2+ action. The discovery of a Ca2+-dependent regulator of phosphodiesterase activity provided the basis for understanding how Ca2+and cAMP interact within cells.73 Calmodulin. The major calcium-dependent regulatory protein is calmodulin, a 17-kDa protein that is homologous to the muscle protein troponin C in structure and function.84,85 Calmodulin has four Ca2+-binding sites, and full occupancy of these leads to a marked conformational change of the protein. This conformational change is linked to the ability of calmodulin to activate enzymes. Calmodulin can be a constituent subunit of complex proteins. For example, it is the d subunit of phosphorylase b kinase.86 The interaction of Ca2+ with calmodulin and the resultant change of activity of the latter are similar conceptually to the binding of cAMP to protein kinase and the subsequent activation of this molecule. Calmodulin participates in regulating various protein kinases and enzymes of cyclic nucleotide generation and degradation.86a As shown in Figure 4-14, the Ca2+-calmodulin complex activates specific calmodulin-dependent protein kinases and a multifunctional calmodulin-dependent protein kinase. Other enzymes are regulated directly (adenylate cyclase, cyclic nucleotide phosphodiesterase, phosphorylase kinase) or indirectly (Ca2+/Mg2+-ATPase, glycerol-3-phosphate dehydrogenase, glycogen synthase, guanylate cyclase, myosin kinase, nicotinamide-adenine dinucleotide (NAD) kinase, phospholipase A2, pyruvate carboxylase, pyruvate dehydrogenase, and pyruvate kinase) by Ca2+, probably through calmodulin.

FIGURE 4-14. Many hormone effects are mediated by phosphatidy-linositide metabolites and by ionic calcium (Ca2+). Certain hormone-receptor interactions are coupled through a G-protein complex to a membrane-associated enzyme, phospholipase C. This enzyme catalyzes the hydrolysis of phosphatidylinositol bisphosphate (PIP2)into diacylglyceride (DAG) and inositol triphosphate (IP3).These intracellular messengers activate specific enzymes. DAG directly activates protein kinase C. IP33 combines with a specific receptor on Ca2+ cellular organelles (e.g., the endoplasmic reticulum [ER]), releasing Ca2+into the cytoplasm. Ca2+, in combination with calmodulin (Cam), activates enzymes such as the specific Cam-kinase and a multifunctional Cam-kinase. Ca2+-Cam also binds to and changes the activity of a number of other proteins. (Prot, protein.) (Courtesy of John Exton.)

Along with its effects on enzymes and ion transport, Ca2+-calmodulin regulates the activity of many structural elements in cells. These include the actin-myosin complex of smooth muscle, which is under b-adrenergic control, and various microfila-ment-mediated processes in noncontractile cells, including cell motility, conformation changes, the mitotic apparatus, granule release, and endocytosis. Calcium regulates the transcription of the FOS gene by phosphorylating CREB, probably through the mechanism described in Figure 4-14. This is an interesting example of the convergence of two different signal transduction pathways: the cAMP system and the Ca2+ system.87 Calcium as a Mediator of Hormonal Action. A role for ionized calcium in hormonal action is suggested by observations that the effect of many hormones is blunted in Ca2+-free media or when intracellular calcium is depleted; can be mimicked by agents that increase cytosolic Ca2+, such as the Ca2+ ionophore A23187; and involves changes of cellular calcium flux. These processes have been studied in some detail in the pituitary, smooth muscle, platelets, and salivary gland, but most is known about how vasopressin and a-adrenergic catecholamines regulate glycogen metabolism in the liver. Phosphorylase activation results from the conversion of phosphorylase b to phosphorylase a through the action of the enzyme phosphorylase b kinase. This enzyme contains calmodulin as its d subunit, and its activity is increased through Ca2+ concentration ranges of 0.1 to 1.0 M. Addition of a1 agonists or vasopressin to isolated hepatocytes causes a threefold increase of cytosolic Ca2+ (from 0.2 to 0.6 M) within a few seconds. This change precedes and equals the increase in phosphorylase a activity, and the hormone concentrations required for both processes are comparable.88,89 This effect on Ca2+ is inhibited by a1-antagonists, and removal of the hormone causes a prompt decline of cytosolic Ca2+ and phosphorylase a. The initial source of the Ca2+ appears to be intracellular organelle reservoirs, which seem to be sufficient for the early effects of the hormones. More prolonged action appears to require enhanced influx, inhibition of Ca2+ efflux, or both through the Ca2+ pump. The latter may depend on concomitant increases of cAMP. Role of Phosphatidylinositide Metabolism in Ca2+-Dependent Hormonal Action. Some signal must provide communication between the hormone receptor on the plasma membrane and the intracellular Ca2+ reservoirs. This is accomplished by the products of phosphatidylinositide metabolism.82 Phosphatidylinositol 4,5-bisphosphate is hydrolyzed to 1,4,5-triphosphate and diacylglycerol through the action of (IP3) phospholipase C (see Fig. 4-14). These two signals activate different pathways. IP3 binds to a receptor on the surface of intracellular organelles that serve as Ca2+ repositories. The binding of IP3 to this receptor (which is similar to the ryanodine receptor) opens these Ca2+ channels, and cytosolic free Ca2+increases.82 Ca2+ enters into cells from the extracellular fluid. Diacylglycerol activates protein kinase C, and this enzyme alters metabolic processes by phosphorylating various substrate proteins. Steroidogenic agents, including ACTH and cAMP in the adrenal cortex; angiotensin II, K+, serotonin, ACTH, and dibutyryl cAMP in the zona glomerulosa of the adrenal; LH in the ovary; and LH and cAMP in the Leydig cells of the testes have been associated with increased amounts of phosphatidic acid, phosphatidylinositol, and polyphosphatidylinositides in the respective target tissues.90 Other examples include the addition of thyrotropin-releasing hormone to pituitary cells, which is followed within 15 seconds by a marked increase of inositol degradation by phospholipase C. The intracellular levels of inositol diphosphate and triphosphate increase markedly, mobilizing intracellular calcium. This activates calcium-dependent protein kinase, which phosphorylates several proteins, one of which appears to be involved in TSH release.91 Calcium also appears to be the intracellular mediator of gonadotropin-releasing hormone action on LH release, which probably involves calmodulin.92 The roles that Ca2+ and phosphoinositide breakdown products may play in hormonal action are shown in Figure 4-14. In this scheme, the phosphoinositide products are the second messengers, and Ca2+ is a tertiary messenger. Hormones that couple through G proteins generate signals by this mechanism, as do some hormones that initiate signal transduction by the activation of an intrinsic tyrosine kinase activity in the receptor. It is likely that several examples of the complex networking of intracellular messengers will be discovered. HORMONES THAT USE A KINASE OR PHOSPHATASE CASCADE AS THE INTRACELLULAR MESSENGER This important group of hormones had been listed under the category of intracellular mediator unknown. A major breakthrough came with the discovery that the EGF receptor contained an intrinsic tyrosine kinase activity that was activated on the binding of the ligand, EGF.93 Shortly thereafter, the insulin and IGF-I receptors were also found to contain intrinsic, ligand-activated tyrosine kinase activity.27,94 Several receptors, generally those involved in binding ligands involved in growth control, have intrinsic tyrosine kinase activity or associate with proteins that are tyrosine kinases.95 Another distinguishing feature of this class of hormone action is that these kinases preferentially phosphorylate tyrosine residues, and tyrosine phosphorylation is infrequent (<0.03% of total amino acid phosphorylation) in mammalian cells.4 Some of the hormones, such as EGF, that activate tyrosine kinases activate phospholipase C and exert effects through Ca2+ and IP3 or diacylglycerol. Investigators have considered that tyrosine kinase activation could also initiate a phosphorylation and dephosphorylation cascade that involved the action of one or several other protein kinases and the counterbalancing actions of phosphatases. Considerable evidence supports this hypothesis. Two mechanisms are used to initiate this cascade. Some hormonessuch as growth hormone, prolactin, erythropoietin, and the cytokinesinitiate their action by activating tyrosine kinase, but this activity is not an integral part of the hormone receptor. The hormone-receptor interaction somehow activates cytoplasmic protein tyrosine kinases (PTKs), such as Tyk-2, Jak-1, or Jak-2.96 These PTKs phosphorylate one or more cytoplasmic proteins, which then associate with other docking proteins through binding to Src homology 2 domains, ~100 amino acidlong segments that are referred to as SH2 domains. Further steps must be elucidated, but it is presumed that hormones of this class mediate their effects, at least in part, through a kinase cascade that is initially activated by PTK. Activation of the intrinsic tyrosine kinase activity of the insulin receptor results in the phosphorylation of a substrate, called the insulin receptor substrate (IRS), on tyrosine residues. At least four IRS proteins have been identified. Phosphorylated IRS proteins bind to the SH2-domains of a variety of proteins that presumably are directly or indirectly involved in mediating different effects of insulin. For example, the binding of IRS-2 to PI-3 kinase results in the activation of the latter, and this links insulin action to phosphoinositide metabolism and thereby to many physiologic processes, including translocation of the glucose transporter and the regulation of genes involved in metabolism. The growth-promoting effects of insulin, and of IGF-I, appear to result from the phosphorylation of IRS-1 by the insulin receptor. Phosphorylated IRS-1 binds to another SH2 domaincontaining protein, GRB-2, which, in turn, activates the MAP kinase pathway. The result of this interaction is the activation of a cascade of threonine or serine kinases.97 The exact role of the many docking proteins, kinases, and phosphatases in insulin action remains to be established. It is particularly important to link these various pathways to the well-established physiologic and biochemical actions of this hormone, and to decipher how the specificity of hormone action is achieved by the many hormones that use one or more of the components of this complex array of signaling proteins. CHAPTER REFERENCES
1. 2. 3. 4. 5. Robishaw JD, Russell DW, Harris BA, et al. Deduced primary structure of the a-subunit of the GTP-binding stimulatory protein of adenylate cyclase. Proc Natl Acad Sci U S A 1986; 83:1251. Obin M, Nowell T, Taylor A. The photoreceptor G-protein transducin (Gt) is a substrate for ubiquitin-dependent proteolysis. Biochem Biophys Res Commun 1994; 200:1169. Waterfield MD, Scrace GT, Whittle N, et al. Platelet-derived growth factor is structurally related to the putative transforming protein p28 sis of simian sarcoma virus. Nature 1983; 304:35. Goldstein J, Brown M, Anderson RGW, et al. Receptor-mediated endocytosis. Annu Rev Cell Biol 1985; 1:1. Hunter GK, Cooper JA. Protein-tyrosine kinases. Annu Rev Biochem 1985; 54:897.

6. Sahal D, Ramachandvan J, Fujita-Yamaguchi Y. Specificity of tyrosine protein kinases of the structurally related receptors for insulin and insulin-like growth factor I: Tyr-containing synthetic polymers as specific inhibitors or substrates. Arch Biochem Biophys 1988; 260:416. 7. Spiegel AM, Levine MA, Aurbach GD, et al. Deficiency of hormone receptor-adenylate cyclase coupling protein: basis for hormone resistance in pseudo-hypoparathyroidism. Am J Physiol 1982; 243:E37. 8. Laron Z, Kowadlo-Silbergeld A, Eshet R, Pertzelan A. Growth hormone resistance. Ann Clin Res 1980; 12:269. 9. Davies TF. Diseases of the TSH receptor. Clin Endocrinol Metab 1983; 12:79. 10. Kahn CR, Flier JS, Bar RS, et al. The syndromes of insulin resistance and acanthosis nigricans. N Engl J Med 1976; 294:739. 11. Griffin JE, Wilson JD. The syndromes of androgen resistance. N Engl J Med 1980; 302:198. 12. Landis CA, Masters SB, Spada A, et al. GTPase inhibiting mutations activate the alpha chain of Gs and stimulate adenyl cyclase in human pituitary tumors. Nature 1989; 340:692. 13. Kristjansson K, Rut AR, Hewison M, et al. Two mutations in the hormone binding domain of the vitamin D receptor cause tissue resistance to 1,25-dihydroxyvitamin D. J Clin Invest 1993; 92:12. 14. Rosenthal W, Antaramian A, Gilbert S, Birnbaumer M. Nephrogenic diabetes insipidus: a V 2 vasopressin receptor unable to stimulate adenylyl cyclase. J Biol Chem 1993; 268:13030. 15. Butcher RW, Baird CE. Effects of prostaglandins on adenosine 3,5 mono-phosphate levels in fat and other tissues. J Biol Chem 1968; 243:1713. 16. Cuttitta F, Carney DN, Mulshine J, et al. Bombesin like peptides can function as autocrine growth factors in human small cell lung cancer. Nature 1985; 316:823. 17. Evans R. The steroid and thyroid hormone receptor superfamily.Science 1988; 240:889. 18. Dufau ML, Catt KJ. Gonadotropin receptors and regulation of steroidogenesis in the testes and ovary. Vitam Horm 1978; 36:461. 19. Andreone TL, Beale EG, Bar R, Granner DK. Insulin decreases phosphoenolpyruvate carboxykinase (GTP) mRNA activity by a receptor mediated process. J Biol Chem 1982; 257:35. 20. Macquire MD, Ross EM, Gilman AG. b-Adrenergic receptor: ligand binding properties and the interaction with adenylyl cyclase. Adv Cyclic Nucleotide Res 1977; 8:1. 21. Kono T, Barham FW. The relationship between the insulin-binding capacity of fat cells and the cellular response to insulin: studies with intact and trypsin-treated cells. J Biol Chem 1971; 246:6210. 22. Samuels HH, Tomkins GM. Relation of steroid structure to enzyme induction in hepatoma tissue culture cells. J Mol Biol 1970; 52:57. 23. Benovic JL, Pike LJ, Cerione RA, et al. Phosphorylation of the mammalian b-adrenergic receptor by cyclic AMP-dependent protein kinase. J Biol Chem 1985; 260:7094. 24. Sibley DR, Benovic JL, Caron MG, Lefkowitz RJ. Phosphorylation of cell surface receptors: a mechanism for regulating signal transduction pathways. Endocr Rev 1988; 9:38. 25. Mishina M, Tobimatsu T, Imoto K, et al. Location of functional regions of acetylcholine receptor a-subunit by site-directed mutagenesis. Nature 1985; 313:364. 26. Gotti C, Frigerio F, Bolognesi M, et al. Nicotinic acetylcholine receptor: a structural model for alpha-subunit peptide 18820, the putative binding site for cholinergic agents. FEBS Lett 1988; 228:118. 27. Kasuga M, Karlsson FA, Kahn CR. Insulin stimulates phosphorylation of the 95,000 dalton subunit of its own receptor. Science 1982; 215:185. 28. Massague J, Czech MP. The subunit structures of two distinct receptors for insulin-like growth factors I and II and their relationship to the insulin receptor. J Biol Chem 1982; 257:5038. 29. Ullrich A, Coussens L, Hayflick JS, et al. Human epidermal growth factor cDNA sequence and aberrant expression of the amplified gene in A431 epidermoid carcinoma cells. Nature 1984; 309:418. 30. Russell DW, Schneider WJ, Yamamoto T, et al. Domain map of the LDL receptor: sequence homology with the epidermal growth factor precursor. Cell 1984; 37:577. 31. Dohlman HG, Thorner J, Caron MG, Lefkowitz RJ. Model systems for the study of seven-transmembrane-segment receptors. Annu Rev Biochem 1991; 60:653. 32. Yamamoto KR. Steroid receptor regulated transcription of specific genes and gene networks. Annu Rev Genet 1985; 19:209. 33. Sutherland EW. An introduction. In: Robison GA, Butcher RW, Sutherland EW, eds. Cyclic AMP. New York: Academic Press, 1971:1. 34. Beato M. Gene regulation by steroid hormones. Cell 1989; 56:335. 35. Lucas PC, Granner DK. Hormone response domains in gene transcription. Annu Rev Biochem 1992; 61:1131. 36. Reynhout JK, Smith LD. Studies on the appearance and nature of a maturation-inducing factor in the cytoplasm of amphibian oocytes exposed to progesterone. Dev Biol 1974; 38:394. 37. Brock ML, Shapiro DJ. Estrogen stabilizes vitellogenin mRNA against cytoplasmic degradation. Cell 1983; 34:207. 38. Diamond DJ, Goodman HM. Regulation of growth hormone mRNA synthesis by dexamethasone and triiodothyronine. J Mol Biol 1985; 181:41. 39. Petersen DD, Koch SR, Granner DK. 3' noncoding region of phospho-enolpyruvate carboxykinase mRNA contains a glucocorticoid-responsive mRNA stabilizing element. Proc Natl Acad Sci U S A 1989; 86:7800. 40. Hod Y, Hanson R. Cyclic AMP stabilizes the messenger RNA for phospho-enolpyruvate carboxykinase (GTP) against degradation. J Biol Chem 1988; 263:7747. 41. Firestone GL, Payvar F, Yamamoto KR. Glucocorticoid regulation of protein processing and compartmentalization. Nature 1982; 300:221. 42. Weintraub H, Groudine M. Chromosomal subunits in active genes have an altered conformation. Science 1976; 93:848. 43. Baker C, Ziff E. Promoters and heterogeneous 5' termini of the mRNAs of adenovirus serotype 2. J Mol Biol 1981; 149:189. 44. McKnight SL, Kingsbury R. Transcriptional control signals of a eucaryotic protein coding gene. Science 1982; 217:316. 45. Wang J-Ch, Stromstedt P-E, Sugiyama T, Granner DK. The phosphoe-nolpyruvate carboxykinase gene glucocorticoid response unit. Identification of the functional domains of accessory factors HNF3 and HNF4 and the necessity of proper alignment of their cognate binding sites. Mol Endocrinol 1999; in press. 13:604. 46. Montminy M. Transcriptional regulation by cyclic AMP. Annu Rev Biochem 1997; 66:807. 47. Torchia J, Glass C, Rosenfeld MG. Co-activators and co-repressors in the investigation of transcriptional responses. Curr Opin Cell Biol 1998; 10:373. 48. Grunstein M. Histone acetylation in chromatin structure and transcription. Nature 1997; 389:349. 49. Denton RM. Early events in insulin actions. Adv Cyclic Nucleotide Protein Phosphorylation Res 1986; 20:293. 50. Lee KAW, Masson N. Transcriptional regulation by CREB and its relatives. Biochim Biophys Acta 1993; 1174:221 51. Hepler JR, Gilman AG. G proteins. Trends Biochem Sci 1992; 17:383. 52. Brandt DR, Ross EM. GTPase activity of the stimulatory GTP-binding regulatory protein of adenylate cyclase, Gs : accumulation and turnover of enzyme-nucleotide intermediates. J Biol Chem 1985; 260:266. 53. Gill DM, Meren R. ADP-ribosylation of membrane proteins catalyzed by cholera toxin: basis of the activation of adenylate cyclase. Proc Natl Acad Sci U S A 1978; 15:3050. 54. Vaughan M. Choleragen, adenylate cyclase, and ADP-ribosylation. Harvey Lect 1983; 77:43. 55. Katada T, Ui M. Direct modification of the membrane adenylate cyclase system by isletactivating protein due to ADP-ribosylation of a membrane protein. Proc Natl Acad Sci U S A 1982; 79:3129. 55a. Gether U. Uncovering molecular mechanisms involved in activation of G protein-coupled receptors. Endocr Rev 2000; 21:90. 56. 57. 58. 59. 60. Nukada T, Tanabe T, Takahashi H, et al. Primary structure of the a-subunit of bovine adenylate cyclase-inhibiting G-protein deduced from the cDNA sequence. FEBS Lett 1986; 197:305. Hurley JB, Simon MI, Teplow DB, et al. Homologies between signal transducing G proteins and RAS gene products. Science 1984; 226:860. Farfel Z, Brickman AS, Kaslow HR, et al. Defect of receptor-cyclase coupling protein in pseudohypoparathyroidism. N Engl J Med 1980; 303:237. Spiegel AM, Weinstein LS, Shenker A. Abnormalities in G protein-coupled signal transduction pathways in human disease. J Clin Invest 1993; 92:1119. Levine MA, Downs RW Jr, Moses AM, et al. Resistance to multiple hormones in patients with Aurbach pseudohypoparathyroidism. Am J Med 1983; 74:545.

60a. Farfel Z, Bourne HR, Iiri T. The expanding spectrum of G protein diseases. N Engl J Med 1999; 340:1012. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. Lyons J, Landis CA, Harsh G, et al. Two G protein oncogenes in human endocrine tumors. Science 1990; 249:655. De Crombrugghe B, Busby S, Buc H. Cyclic AMP receptor protein: role in transcriptional activation. Science 1984; 224:831. Beavo JA, Bechtel PJ, Krebs EG. Mechanisms of control for cAMP-dependent protein kinase from skeletal muscle. Adv Cyclic Nucleotide Res 1975; 5:241. Kemp BE, Graves DJ, Benjamin E, Krebs EG. Role of multiple basic residues in determining the substrate specificity of cyclic AMP-dependent protein kinase. J Biol Chem 1977; 252:4888. Hunter T. A thousand and one protein kinases. Cell 1987; 50:823. Kuo JF, Greengard P. Cyclic nucleotide-dependent protein kinase: IV. Widespread occurrence of adenosine 3',5'-monophosphate-dependent protein kinase in various tissues and phyla of the animal kingdom. Proc Natl Acad Sci U S A 1969; 64:1349. Montminy MR, Bilezikjian ML. Binding of a nuclear protein to the cyclic-AMP response element of the somatostatin gene. Nature 1987; 328:175. Hoeffler JP, Meyer TE, Yun Y, et al. Cyclic AMP-responsive DNA-binding protein: structure based on a cloned placental cDNA. Science 1988; 242:1430. Quinn PG, Granner DK. Cyclic AMP-dependent protein kinase regulates transcription of the phosphoenolpyruvate carboxykinase gene but not binding of nuclear factors to the cyclic AMP regulatory element. Mol Cell Biol 1990; 10:3357. Wells JN, Hardman JG. Cyclic nucleotide phosphodiesterases. Adv Cyclic Nucleotide Res 1977; 8:119. Strada SJ, Thompson JW. Multiple forms of cyclic nucleotide phosphodiesterases: anomalies or biologic regulators. Adv Cyclic Nucleotide Res 1977; 9:265. Francis SH, Kono T. Hormone-sensitive cAMP phosphodiesterase in liver and fat cells. Mol Cell Biol 1982; 42:109. Cheung WY, Lynch TJ, Wallace RW. An endogenous Ca2+-dependent activator protein of brain adenylate cyclase and cyclic nucleotide phosphodiesterase. Adv Cyclic Nucleotide Res 1977; 9:233. Cohen P. The structure and regulation of protein phosphatases. Annu Rev Biochem 1989; 58:453. Dent P, Lavoinne A, Nakielny S, et al. The molecular mechanism by which insulin stimulates glycogen synthesis in mammalian skeletal muscle. Nature 1990; 348:302. Walton KM, Dixon JE. Protein tyrosine phosphatases. Annu Rev Biochem 1993; 62:101. Chase LR, Aurbach GD. Parathyroid function and the renal excretion of 3' adenylic acid. Proc Natl Acad Sci U S A 1967; 58:518. Murad F, Mittal CK, Arnold WP, et al. Guanylate cyclase: activation by azide, nitro compounds, nitric oxide, hydroxyl radical and inhibition by hemoglobin and myoglobin. Adv Cyclic Nucleotide Res 1978; 9:145. Takio K, Wade RD, Smith SF, et al. Guanosine cyclic 3,5 phosphate dependent protein kinase: a chimeric protein homologous with two separate protein families. Biochemistry 1984; 23:4207. Chinkers M, Garbers DL. Signal transduction by guanylyl cyclases. Annu Rev Biochem 1991; 60:553. Rasmussen H. The calcium messenger system. N Engl J Med 1986; 314:1094. Berridge M. Inositol triphosphate and calcium signaling. Nature 1993; 361:315. Tsien RY, Rink TJ, Poenie M. Measurement of cytosolic free Ca2+ in individual cells using fluorescence microscopy with dual excitation wavelengths. Cell Calcium 1985; 6:145. Watterson M, Sharief F, Vanaman TC. The complete amino acid sequence of the Ca 2+-dependent modulator protein (calmodulin) of bovine brain. J Biol Chem 1980; 255:962. Means AR, Chafouleas JG. Calmodulin in endocrine cells. Annu Rev Physiol 1982; 44:667. Cohen P, Burchell A, Foulkes JG, et al. Identification of the Ca2+-dependent modulatory protein as the fourth subunit of rabbit skeletal muscle phosphorylase kinase. FEBS Lett 1978; 92:287.

86a. Means AR. Regulatory cascades involving calmodulin-dependent protein kinases. Mol Endocrinol 2000; 14:4. 87. Sheng M, Thompson MA, Greenberg ME. CREB: a Ca(2+)-regulated transcription factor phosphorylated by calmodulin-dependent kinases. Science 1991; 252:1427. 88. Williamson JR, Cooper RH, Hoek JB. Role of calcium in the hormonal regulation of liver metabolism. Biochim Biophys Acta 1981; 639:243. 89. Blackmore PF, Hughes BP, Charest R, et al. Time course of a1 -adrenergic and vasopressin actions on phosphorylase activation: calcium efflux, pyridine nucleotide reduction, and respiration in hepatocytes. J Biol Chem 1982; 258:10488. 90. Farese RV. The role of the phosphatidate-phosphoinositide cycle in the action of steroidogenic hormones. Endocr Res 1985; 10:515. 91. Kolesnick RN, Musacchio I, Thaw C, Gershengorn MD. Thyrotropin (TSH)-releasing hormone decreases phosphatidylinositol and increases unesterified arachidonic acid in thyrotropic cells: possible early events in stimulation of TSH secretion. Endocrinology 1984; 114:671. 92. Conn PM, Staley D, Jinnah H, Bates M. Molecular mechanism of gonado-tropin releasing hormone action. J Steroid Biochem 1985; 23:703. 93. Staros JV, Cohen S, Russo MW. Epidermal growth factor receptor: characterization of its protein kinase activity. In: Cohen P, Housley MD, eds. Molecular mechanism of transmembrane signaling New York: Elsevier Science Publishing, 1985:253. 94. Ullrich A, Bell JA, Chen EY, et al. Human insulin receptor and its relationship to the tyrosine kinase family of oncogenes. Nature 1985; 313:756. 95. Fantl WJ, Johnson DE, Williams LT. Signaling by receptor tyrosine kinases. Annu Rev Biochem 1993; 62:453.

96. Argetsinger LS, Campbell GS, Yang X, et al. Identification of JAK2 as a growth hormone receptor-associated tyrosine kinase. Cell 1993; 74:237. 97. Cheatum B, Kahn CR. Insulin action and the insulin signaling network. Endocr Rev 1995; 16:117.

CHAPTER 5 FEEDBACK CONTROL IN ENDOCRINE SYSTEMS Principles and Practice of Endocrinology and Metabolism

CHAPTER 5 FEEDBACK CONTROL IN ENDOCRINE SYSTEMS

DANIEL N. DARLINGTON AND MARY F. DALLMAN Feedback Definition Components Localization of Components Amplification of Output Time Constants Loads on Endocrine Systems Examples of Feedback in Endocrine Systems HypothalamicPituitaryTarget Endocrine Feedback Feedback Effects on Amplitude and Pulse Frequency: The Gonadotropin System Parturition and Suckling Regulated Variables Endocrine Cells as Receptors and Effectors Receptors in Proximity to the Endocrine Gland Receptors that can Initiate Multisystem Responses Conclusion Chapter References

The function of endocrine systems is to maintain homeostasis of the organism and to perpetuate the species. Homeostasis is maintained by the continual adjustment of nervous and endocrine systems in response to a changing environment. Because most endocrine systems are intimately related to the central nervous system and the autonomic nervous system (see Chap. 8,Chap. 9 and Chap. 85), a continual bidirectional flow of informationone relaying the conditions in the organism, and the other directing neural and endocrine responsesis present and functions to maintain appropriate operating conditions. To produce stable systems, it has been useful to apply the knowledge gained from the engineering discipline to physiologic control systems. The quantitative relations among system components in engineering and the mathematical description of these have been useful in describing biologic control systems.1 The notion of a setpoint reference signal, which can be explicitly delineated in an engineered system, is, in most biologic systems, an apparent setpoint, deduced by the behavior of the system, and probably normally arises from the interaction of a hormone or metabolite with receptors and cellular function.

FEEDBACK
DEFINITION Feedback is a process within a system that occurs when the product or result of activity in the system modifies the factors that produce that product or result. The result can modify these factors in either of two ways: (a) by stimulating the factors to generate more product (positive feedback), or (b) by inhibiting the factors so that fewer products are generated (negative feedback). Mammalian organisms are composed of many feedback systems. Each system has its own specific function and is self-regulating. However, all of the systems are interconnected, making the overall system extremely complex. Usually, negative feedback is used to maintain a variable within a range that is advantageous for optimal cellular function. When an internal or external stimulus perturbs a regulated variable, neural or endocrine responses (or both) occur that counteract the disturbance and return the variable to within its normal range. Such performance requires a means to determine the level of activity or the concentration of the variable that is regulated. COMPONENTS The components of a simple endocrine feedback system are shown in Figure 5-1. A variable is monitored by a receptor that is sensitive to changes in that variable. The receptor is connected to a structure or structures that process and integrate the signal generated by alterations in the variable. The signal processor includes a reference setpoint and a comparator element that compares the receptor input with the reference signal. The difference between input from the receptor and the setpoint is generated by the comparator, and the resulting error signal then modifies the existing activity of the effector endocrine gland. If the error signal is positive, the effector gland is stimulated to secrete, and the concentration of hormone in the circulation increases to levels that are adequate to effect the normalization of the perturbed regulated variable. Because many hormones are secreted into the circulation, the concentration of hormone achieved is affected by the rate of clearance and by the amount of hormone that is bound in the plasma compartment by a specific carrier protein (i.e., only the free, unbound hormone is available to cellular receptors).

FIGURE 5-1. Model for a simple endocrine feedback system consisting of a variable being monitored by a receptor, a signal processor that analyzes the signal, and an endocrine gland with a secretion rate that is regulated by the processor. The concentration of free hormone in the circulation depends on the clearance of the hormone and the presence and concentration of specific binding protein in the circulation. Any change in the variable is sensed, and a response (hormone) is elicited to bring the variable back within its normal range.

LOCALIZATION OF COMPONENTS The receptor, processing unit, and endocrine gland may all reside within a single cell or in a group of similar cells functioning in concert, or they may involve entirely separate and widely dispersed units. For example, the B cell (b cell, insulin-producing) of the pancreas is directly sensitive to changes in extracellular glucose concentration (see Chap. 134). An increase in blood glucose is thus sensed by the B cell and increases the rate of insulin synthesis and secretion. The increased concentration of insulin causes increased glucose transport into those cells that contain insulin receptors and a reduction in blood glucose concentration. The normalization of glucose concentration eliminates the stimulus to the B cell. Arterial blood pressure is monitored by mechanoreceptors in the carotid artery and aortic arch. The central nervous system processes and integrates afferent information from the mechano-receptors. The autonomic nervous and various endocrine systems represent the gland that secretes norepinephrine and various hormones; when there is a marked decrease in arterial pressure, the release of these hormones effectively restores arterial blood pressure to normal. AMPLIFICATION OF OUTPUT All feedback systems amplify their outputs (glandular secretions) to correct for the effects of perturbations. The effectiveness of this amplification in maintaining a regulated variable within the normal range is termed the gain of the system. The gain of a system is determined by comparing the magnitude of excursion of a

regulated variable under conditions in which the feedback loop is abolished with the magnitude of excursion under conditions in which the feedback loop is intact. Figure 5-2 shows a theoretical situation in which the control of glucose is examined. At the moment of onset of the glucose infusion, the pancreatic B cells are either removed from the system or left itact. In the absence of the B cells, insulin secretion would not occur, and the blood glucose concentration would increase because of the glucose infusion in proportion to the infusion rate. The steady-state glucose concentration would be determined by the volume into which glucose is distributed and the rate at which glucose is lost from that volume of distribution in the absence of insulin. With the B cells of the pancreas in the system (i.e., under closed-loop conditions), the infusion-induced increase in blood glucose concentration would increase the concentration of glucose bathing the pancreatic B cell and cause secretion of insulin. The elevated insulin concentration activates increased target cell uptake of glucose, lowering the circulating glucose levels. Glucose concentration in the open-loop condition increases well above the normal range; whereas under closed-loop control, glucose concentration is maintained close to the normal range.

FIGURE 5-2. Gain in an endocrine feedback system. The diagram shows the (hypothetical) blood glucose (top) and insulin (bottom) concentrations before and during infusion of glucose. The gain is defined as the ratio of the difference between the responses (perturbed glucose levels) of the open-loop and closed-loop conditions (arrow b) and the difference between the response of the closed-loop glucose and the preinfusion levels (arrow a). The gain is an index of feedback loop effectiveness. (Redrawn from Guyton AC. Textbook of medical physiology, 5th ed. Philadelphia: WB Saunders, 1976.).

The error in the regulation of blood glucose is the difference between the preinfusion value (see Fig. 5-2, dotted line) and the new level (see Fig. 5-2, arrow a). The degree by which the variable (glucose) is controlled is the difference between the open-loop glucose concentration and the closed-loop glucose concentration (see Fig. 5-2, arrow b). The gain of the system can then be calculated as the ratio between arrow b and arrow a.2 In this example, the difference between the open-loop and closed-loop response is 10 units, and the difference between the closed-loop response and preinfusion levels is 2 units. The gain of this system is, therefore, 10/2, or 5 (usually denoted as a negative number to show normalization). Thus, for each unit increase of the controlled variable under closed-loop conditions, there would be a five-fold change if the feedback loop were open. TIME CONSTANTS The effectiveness of feedback can also be described in terms of the time elapsed between the onset of the stimulus (the change in the regulated variable) and the effect of the system response. This is a function of the time constant of each component of the feedback loop. In a multicomponent feedback system, the component with the slowest time constant determines the delay time between stimulus and response (time constant of the system). Figure 5-3 compares the imagined performance of two systems, one with a fast-time constant (left) and the other with a slow-time constant (right). Time constants may clarify the requirement for the hierarchical feedback loops that are so commonly encountered in biologic systems.

FIGURE 5-3. Comparison of the effectiveness of feedback systems (hypothetical) with fast- and slow-time constants. Notice that for the same stimulus, the feedback system with the fast-time constant regulates the variable near normal levels and does not overshoot or under-shoot as does the feedback system with the slow-time constant.

The time constants for transmission of neural information are in the milliseconds to seconds range. Peptide and protein hormones, acting on receptors in the plasma membrane to alter intracellular concentrations of a second messenger, exert many of their primary effects on cellular function over the range of seconds to minutes. Still slower, in hours, are the effects of steroid and thyroid hormones, which are mediated by alterations in the rate of gene transcription and translation, and of protein synthesis. Interestingly, those hormones with the longest time constants (adrenal and gonadal steroids, hepatic somatomedins) are secreted by glands controlled by other hormones with faster time constants, and these systems exhibit high-gain feedback control. Moreover, the feedback inhibition exerted by these hormones with characteristically slow target effects may be mediated by mechanisms with faster time constants than those of their target effects. Thus, the overall performance of a relatively slow system may appear tighter than anticipated and more like that of a system with an overall fast-time constant. For instance, the secretion of glucocorticoids is regulated by a cascade system in which corticotropin-releasing hormone (CRH), secreted from the median eminence of the hypothalamus in response to appropriate stimulation, acts on corticotropes in the anterior pituitary to cause synthesis and secretion of adrenocorticotropic hormone (ACTH). ACTH, likewise, acts on the zona fasciculata cells of the adrenal cortex to stimulate the synthesis and secretion of cortisol and corticosterone (see Chap. 14). The initiation of a stimulus to CRH until a rise in circulating concentration of cortisol or corticosterone occurs requires <5 minutes. The minimum time required for cortisol to increase gluconeogenesis, enabling increased hepatic glucose secretion, may be ~2 hours. If insulin-induced hypoglycemia were the stimulus that provoked CRH secretion, return of the extracellular glucose concentration to normal would relieve the stimulus to CRH. However, in the absence of a more rapid cortisol feedback effect on the hypothalamic CRH-secreting cells and on the corticotrope cells that secrete ACTH, there would clearly be an overshoot in the amount of cortisol secreted in response to the stimulus of hypoglycemia because of the lag time between the elevated concentrations of cortisol in the circulation and the effect of these elevated levels on peripheral target cells. The direct feedback effects of cortisol, the other steroid hormones, and thyroid hormones on their trophic hormonal controllers allow tighter control of these systems. LOADS ON ENDOCRINE SYSTEMS Endocrine systems have the ability to restore homeostasis in response to varying degrees of perturbation. The load on the system is the degree or strength of the perturbation. In response to the perturbation, the endocrine system adjusts the amount of free circulating hormone in an attempt to regain homeostasis. The amount of free circulating hormone (or hormonal load) is dependent on the secretion and clearance rates of the hormone, and the production and clearance rates of the specific hormone-binding proteins. Newly secreted free hormone is immediately accessible to clearance from the circulation by metabolism in liver, kidney, and other tissues, and by filtration and excretion by the kidney. Moreover, many hormones are bound in the circulation by high-affinity specific binding globulins. The bound hormone is unavailable for diffusion into the interstitial fluid and cannot affect the target organ or tissue. However, the bound and free hormone exists in an equilibrium in plasma such that as free hormone is cleared, bound hormone becomes free. Bound hormone acts as a storage of free hormone. If, for any reason, the rate of metabolism or clearance changes, or if the production or clearance rates of its specific binding hormone changes, then the input signal to the endocrine gland that is producing the hormone must change to accommodate the change in load of free hormone.

EXAMPLES OF FEEDBACK IN ENDOCRINE SYSTEMS

Importantly, our current state of knowledge does not allow us to apply all these principles to most endocrine systems. Nonetheless, there is plentiful evidence for the existence of feedback regulation in all endocrine systems. Knowledge of this feed-back has been used in the design and interpretation of many clinical tests for the determination of endocrine disorders. HYPOTHALAMICPITUITARYTARGET ENDOCRINE FEEDBACK LONG-LOOP FEEDBACK One of the most clinically obvious and simplest forms of negative feedback control in endocrine systems involves suppression of the secretion of a trophic factor (or hormone) by the hormone it stimulates. For example, hormone A stimulates the secretion of hormone B, which in turn suppresses the secretion of hormone A. Hormone B may suppress the secretion of hormone A by acting directly on the cells that secrete A, or indirectly, by acting on the cells (or neurons) that stimulate the secretion of A. This type of control is exemplified in the relations between the hypothalamus, anterior pituitary gland, and peripheral endocrine glands controlled by pituitary hormones. The hypothalamus secretes neurohormones that stimulate (or inhibit) the secretion of specific anterior pituitary hormones, which in turn stimulate a peripheral target gland to secrete hormone and, with sufficient stimulation, to grow (see Chap. 8 and Chap. 9). In Figure 5-4, CRH-containing neurons in the hypothalamus release CRH into the hypophysial portal system.3 ACTH released from corticotropes in response to CRH stimulates cortisol synthesis and secretion from the adrenal cortex. Cortisol acts to inhibit the secretion of ACTH from the corticotrope and to inhibit CRH secretion from the hypothalamic neuron, and it may also act on extrahypothalamic sites that regulate CRH synthesis and secretion.4,5 Clinically, the feedback effects of cortisol are important. Long-term therapy with pharmacologic amounts of glucocorticoids suppresses ACTH secretion to the extent that the adrenal cortices atrophy and become unresponsive to ACTH. The atrophic adrenal cortex does not secrete normal quantities of cortisol, and the abrupt discontinuation of exogenous glucocorticoids may lead to a patient who displays all the signs of cortisol deficiency (see Chap. 76 and Chap. 78).

FIGURE 5-4. Simplified feedback loop in the hypothalamicpituitaryadrenal cortical system. (ACTH, adrenocorticotropin hormone; CRH, corticotropin-releasing hormone; pit., pituitary.) The dotted arrows show the sites of feedback by cortisol.

The feedback effects of cortisol on ACTH release is an example of long-loop feedback: secretion of the peripheral gland affecting the secretion of the pituitary trophic hormone. Long-loop feedback occurs in most of the anterior pituitary hormone systems and is most apparent when the capacity for hormone synthesis or secretion in the peripheral target gland is compromised or abolished. The magnitude of the effects of inhibiting or removing the long-loop feedback signal on circulating concentrations of the appropriate pituitary trophic hormone (i.e., the effects of opening the feedback loop) is large. The dramatic increases in the circulating concentrations of the pituitary trophic hormone that occur under conditions in which there is an abnormally low feedback signal from the target endocrine gland are useful clinically in distinguishing between a primary and a secondary disturbance in an endocrine system. For instance, hypothyroidism could arise from a primary disturbance in the synthesis of thyroxine in the thyroid gland, or it could result from lack of stimulation of the thyroid gland by thyrotropin (thyroid-stimulating hormone [TSH]). In both cases, circulating thyroxine concentrations would be low; if the defect were due to a primary thyroidal disturbance, TSH concentrations would be high, whereas if the defect were due to lack of TSH secretion, circulating concentrations would be low. SHORT-LOOP FEEDBACK Evidence also exists for short-loop inhibition of secretion of hypothalamic-releasing hormones by the trophic hormones of the anterior pituitary. For example, growth hormone secreted by somatotropes in the anterior pituitary stimulates secretion of insulin-like growth factors from the liver and other peripheral tissues (see Chap. 12 and Chap. 173). The somatomedins exhibit long-loop feedback on growth hormone secretion. However, growth hormone has an effect on the hypothalamic-releasing and release-inhibiting hormones that regulate its secretion (Fig. 5-5). There is evidence suggesting that this feedback can occur either by inhibiting the secretion of growth hormone-releasing hormone or by stimulating the secretion of growth hormone-inhibiting hormone (somatostatin) into the hypophysial portal circulation.6,7 There is growing evidence that similar short-loop feedback circuits exist for the other hypophysial hormones. Quantitatively, these appear to be less important than the long-loop feedback, and they may serve as a fine-tuning system within the central nervous system.

FIGURE 5-5. Long- and short-loop feedback. Neurons in the hypothalamus release somatostatin and growth hormone-releasing hormone (GHRH) into the circulation that bathe the growth hormone-secreting cell (somatotrope). Growth hormone stimulates somatomedin C formation in peripheral tissues. Both growth hormone and somatomedins act at the level of the hypothalamus to inhibit growth hormone release.

ULTRASHORT-LOOP FEEDBACK Ultrashort-loop feedback is suspected to occur when a hormone acts on its own cell type to inhibit further secretion of itself. There is little evidence for this phenomenon; however, it could be of considerable importance in the regulation of secretion of posterior pituitary hormones. In the hypothalamus, collateral axons of oxytocin and vasopressin neurons make contact with other oxytocin and vasopressin neurons.8,9 These cellcell interactions could help to sharpen and synchronize the secretion of hormone in a given cell type. In the anterior pituitary, vasoactive intestinal peptide (VIP) stimulates secretion of prolactin from lactotropes. However, this effect may be autocrine/paracrine in nature, because lacto-tropes express and secrete both VIP and prolactin into the extra-cellular fluid (Fig. 5-6). Thus, VIP can then stimulate secretion of VIP/prolactin from its own or neighboring lactotropes.10,11 and 12

FIGURE 5-6. Ultrashort-loop feedback. Vasoactive intestinal peptide (VIP) is secreted from lactotropes with prolactin in the anterior pituitary. VIP can stimulate further secretion of VIP/prolactin from the same cell or from neighboring lactotropes (autocrine and paracrine).

FEEDBACK EFFECTS ON AMPLITUDE AND PULSE FREQUENCY: THE GONADOTROPIN SYSTEM Many, and perhaps most, anterior pituitary hormones are secreted in a pulsatile manner (i.e., surges in secretion occur at regular intervals). Evidence suggests that this arises from synchronous bursts of activity in hypothalamic neurons that secrete the appropriate releasing factors. The gonadal endocrine system has been best studied from this point of view. The gonadotropin-releasing hormone (GnRH) is released in bursts that appear to drive secretion of luteinizing hormone (LH) in episodes (see Chap. 16). In humans and in subhuman primates, GnRH and LH concentrations in hypophysial and systemic blood demonstrate cycles with periods of 2 to 4 hours. LH stimulates testosterone secretion in men and ovulation, luteal formation, and estrogen and progesterone secretion in women. NEGATIVE FEEDBACK BY SEX STEROIDS Testosterone, estrogen, and progesterone exert long-loop feed-back on the cycle amplitude and frequency of GnRH and LH. Testosterone and progesterone decrease the frequency and amplitude of the GnRH cycles and thus lower the mean concentration of LH in the circulation. The absence of both hormones leads to acceleration of the cycle frequency and amplitude. The effects of estradiol are complex. Low concentrations of estradiol or high concentrations of estradiol in the presence of progesterone inhibit both the frequency and amplitude of LH secretory pulses, thus decreasing both minimum and maximum concentrations of LH in the circulation.13,13a,14 POSITIVE FEEDBACK BY SEX STEROIDS Conversely, during the late follicular phase of the menstrual cycle, high concentrations of unopposed estradiol increase pulse frequency and pulse amplitude of LH.13 Thus, a single hormone (estradiol) exerts either negative or positive effects on its trophic hormone, depending on the duration, concentration, and conditions of exposure. The unstable behavior of the gonadal system in this example of positive feedback is characteristic of positive feedback systems. The positive effect of estradiol leads to increasingly frequent and large pulses of LH that in turn lead to increasing secretion of estradiol until the system becomes too unstable to exist, and ovulation occurs. Because the ovum-containing follicle was the primary source of estradiol-secreting granulosa cells, ovulation disrupts the system, and the whole process can begin again (see Chap. 94 and Chap. 95). PARTURITION AND SUCKLING The female reproductive system provides two other good examples of positive feedback: parturition- and suckling- induced oxytocin secretion. In both of these examples, there is a neuroendocrine control loop. In the former, stretch receptors in the uterine cervix are stimulated by presentation of the infant's head; the receptors, in turn, activate neural pathways in the spinal cord and the brainstem that cause secretion of oxytocin from the posterior pituitary. Oxytocin interacts with receptors on the uterine myometrium to cause uterine contraction and further stretch of the uterine cervix. The positive feedback cycle is ended with delivery of the baby and cessation of uterine stretch (see Chap. 25 and Chap. 109). With suckling, stimulation of the nipple activates stretch receptors in the nipples that stimulate spinal cord and brainstem pathways to oxytocin secretion. Oxytocin then causes myoepithelial cell contraction in milk ducts of the breast and milk letdown. The cycle is interrupted when suckling stops (see Chap. 25 and Chap. 106). REGULATED VARIABLES All controlled variables of an organism are monitored by receptors and processors that ensure that the variables are maintained within a relatively narrow normal range. When an internal or external stimulus changes or perturbs a regulated variable, the change is sensed and corrected by the nervous and endocrine systems. Some examples of variables that are regulated largely by activity in the autonomic nervous system and by hormones include blood pressure, blood oxygen and carbon dioxide tension, extracellular fluid volume, tissue substrates, metabolites and ions, the composition of filtrate in the kidney tubules, blood flow to different vascular beds, and enzyme production by exocrine glands. All of these variables are regulated by the feedback arrangement described earlier (see Fig. 5-1). Each variable is monitored by a sense organ that is connected to an information processor that controls neural or glandular secretion. The sense organ can be a modified structure that is (a) connected to the central nervous system, (b) connected to the gland, or (c) an integral part of the endocrine effector cell. The effector organ can be an endocrine gland (adrenal cortex, pancreas) or an autonomic postganglionic nerve terminal. When the concentration or level of function of a regulated variable is perturbed, a neural or hormonal response occurs that rectifies the perturbation. ENDOCRINE CELLS AS RECEPTORS AND EFFECTORS Some blood-borne agents (nonhormonal) have profound effects on the secretion of certain endocrine gland cells. The action of glucose on the pancreatic B cell was mentioned earlier. The overall effect of glucose-stimulated insulin secretion is to reduce elevated blood glucose levels to the normal range by removing glucose from the blood and facilitating storage in other tissues. Clearly, the action of insulin reduces the stimulus to its secretion. Another example of this kind is the direct effect of potassium on aldosterone secretion from cells of the adrenal zona glomerulosa (see Chap. 79). Small increases in the extra-cellular fluid concentration of potassium increase aldosterone synthesis and secretion. Aldosterone facilitates secretion of potassium by cells of the distal tubule of the kidney, thus reducing the extracellular fluid concentration of this cation.15 This type of feedback regulation also exists for gastrointestinal secretions. The gut hormones are secreted from polarized cells in the intestine that are stimulated by gut contents on the luminal surface and that secrete hormones from the basal surface. Examples also can be given of direct stimulation of one endocrine gland that causes recruitment of another hormonal system. The chief cells of the parathyroid glands are directly sensitive to the concentration of ionized calcium in the extracellular fluid (Fig. 5-7; see Chap. 51). If the plasma concentration of ionized calcium decreases, the chief cells respond to the decrease in calcium with an increased rate of secretion of parathyroid hormone. Parathyroid hormone facilitates calcium reabsorption by the renal tubules and increases the release of calcium from bone. This hormone also facilitates the formation of 1,25-dihydroxycholecalciferol.16 This vitamin D metabolite further facilitates calcium release from bone and increases intestinal absorption of calcium (see Chap. 54). Although the regulation of the ionized calcium concentration in plasma is more complex than described here (see Chap. 49), this exemplifies how hormonal systems can be recruited for regulation of a variable.

FIGURE 5-7. Feedback regulation of plasma calcium levels. A decrease in extracellular fluid ionized calcium concentration (Ca2+) stimulates the release of parathyroid

hormone phosphate acts to increase plasma calcium by facilitating calcium reabsorption from kidney, calcium release from bone, and formation of 1,25-dihydroxycholecalciferol (1,25[OH]2D) from 25-hydroxycholecalciferol (25[OH]D). (It also attenuates formation of the inactive form of the hormone, 24,25-dihydroxy-cholecalciferol [24,25(OH)2D].) 1,25(OH)2D increases calcium release from bone and the intestinal absorption of calcium. The dotted lines show the restitution of plasma calcium concentration.

RECEPTORS IN PROXIMITY TO THE ENDOCRINE GLAND Endocrine tissue may modify its hormonal secretory rate as a regulated variable changes but may not actually sense the change itself. For example, the macula densa comprises a specialized cell type that senses sodium or chloride transport from the distal tubular filtrate of the kidney (Fig. 5-8). The macula densa is adjacent to the juxtaglomerular cells in the afferent arterioles entering the Bowman capsule. A fall in plasma sodium concentration is reflected in the sodium concentration in the tubular ultrafiltrate. A decrease in sodium delivery to the distal tubule is sensed by the macula densa cells. The macula densa stimulates the juxtaglomerular cells to secrete renin.17 Renin, released into the general circulation, cleaves the terminal 10 amino acids from angiotensinogen, producing angiotensin I. Angiotensin I is converted in the circulation (particularly as it passes through the pulmonary vascular bed) to angiotensin II. Angiotensin II stimulates the secretion of aldosterone from the cells of the adrenal zona glomerulosa, and the effect is the stimulation of sodium reabsorption and potassium and hydrogen excretion in the renal distal tubule.15 The net movement of sodium into the extracellular fluid compartment increases the plasma sodium concentration and relieves the stimulus to renin secretion.

FIGURE 5-8. Feedback regulation of plasma sodium levels and the existence of two feedback loops in the system (dotted lines). As plasma sodium [Na+] falls, the macula densa in the renal distal tubule senses the sodium fall in the filtrate and stimulates renin release from the juxtaglomerular (JG) cells. Renin enzymatically converts angiotensinogen (in the blood) into angiotensin I. Angiotensin I is enzymatically converted to angiotensin II, which stimulates aldosterone release. Aldosterone facilitates sodium reabsorption from the distal tubules, which leads to an increase in plasma sodium. (c.e., converting enzyme; stim., stimulates.)

The regulation of renin secretion, however, involves two feedback loops. Aldosterone facilitates sodium reabsorption and thus removes the initial stimulus to renin secretion; however, angiotensin II also directly inhibits renin secretion from the juxtaglomerular cells (see Fig. 5-8). Again, as with the hypothalamic pituitary target gland cascades, the hormonal feed-back may serve to inhibit the system more rapidly than would occur with the full restitution of plasma sodium concentration, thus preventing excessive secretion of aldosterone and gaining greater efficiency of the system (see Chap. 79 and Chap. 183). RECEPTORS THAT CAN INITIATE MULTISYSTEM RESPONSES Multisystem control involves specialized receptor organs that are directly connected to or embedded in central components of the autonomic nervous system. Organizationally, the autonomic pathways in the medulla, pons, midbrain, and hypothalamus constitute this central system. Afferent input is provided from chemoreceptors, high- and low-pressure baroreceptors, and other peripheral receptors by way of the spinal cord and cranial nerves. Chemosensitive structures exist in the brain and include the circumventricular organs, glucose receptors and osmoreceptors in the hypothalamus, medullary cells that are sensitive to pH, and cells throughout this region of the brain that contain receptors for steroid hormones4,18,19 and insulin.20 The types of response generated by alterations in input from these receptors tend to involve multiple systems (see Chap. 8 and Chap. 9). The efferent sympathetic and parasympathetic components of the autonomic nervous system innervate all endocrine tissue and can modify the secretion of glands. For example, the well-known cephalic phase of digestion includes vagal stimulation of insulin secretion, which occurs well in advance of substrate (glucose) stimulation of the pancreatic B cell. Insulin secretion, however, is just one of a set of autonomically mediated responses that occurs to the sight, smell, or taste of food. Also included in this phase of digestion is increased salivation, increased gastric motility and acid secretion, and increased gastrin secretion. Stimulation of both central and peripheral receptors occurs with signals that regulate energy balance. In addition to the neural and hormonal feedback information from the alimentary system that occurs with ingestion of food, occupancy of central receptors for glucocorticoids and insulin appears to be critical for overall regulation of energy balance.20,21 Glucocorti-coids and insulin, apparently through their actions on the central nervous system,20,21 are, respectively, stimulatory and inhibitory to food intake; by contrast, their peripheral effects are, respectively, inhibitory and stimulatory to energy storage. It appears likely that the reciprocal effects of glucocorticoids and insulin on food intake are mediated, in part, through their opposing actions on the orexigenic peptide that is synthesized in the arcuate nuclei of the hypothalamus, neuropeptide Y.20,21 Insulin inhibits20 and glucocorticoids stimulate21 neuropeptide Y synthesis and secretion. Neuropeptide Y stimulates food intake in satiated animals and, over the long term, causes obesity. Axons containing neuropeptide Y, in turn, innervate nearby hypothalamic cell groups that are known to constitute a neural network that determines food intake. The metabolic effects of glucocorticoids and insulin on energy balance (see Chap. 72 and Chap. 135) are antagonistic to each other and are opposite in direction to their central, antagonistic effects on energy acquisition. Thus, this bihormonal effector and signaling system serves as a relatively simple overall regulator of energy stores. Stimulation of peripheral receptors occurs with severe hemorrhage (Fig. 5-9). A sudden decrease in arterial blood pressure and venous return occasioned by a rapid hemorrhage of 20% of the blood volume is registered by the peripheral stretch receptors in the great veins and atria and in the aortic arch and carotid arteries. The information leads to increased autonomic sympathetic activity and the secretion of epinephrine from the adrenal medulla. Secreted catecholamines act to increase heart rate and stroke volume (by way of chronotropic and inotropic actions on cardiac muscle) and to increase peripheral vascular resistance by constriction of vascular smooth muscle22,23 (see Chap. 85).

FIGURE 5-9. Example of complex feedback regulation of blood pressure and blood volume after hemorrhage. (Ang, angiotensin; Symp. Act., sympathetic activity.)

Increased sympathetic outflow to the juxtaglomerular cells of the kidney and decreased renal perfusion pressure or flow cause renin secretion24,25 and, thus, an increase in circulating angiotensin II concentration. Angiotensin II acts directly on arterial smooth muscle to cause vasoconstriction and, therefore, increased arterial blood pressure. Angiotensin II, by way of an action on the subfornical organ, also stimulates the conscious desire to drink fluids and further augment sympathetic activity.19,26 Moreover, angiotensin II acts on the adrenal glomerulosal cells to stimulate the secretion of aldosterone,27 which causes increased reabsorption of sodium by the kidney and aids in the restoration of the extracellular fluid volume. The hemorrhage-induced alteration in atrial baroreceptor input to the brain also causes stimulation of vasopressin secretion, which causes direct constriction of the vascular smooth muscle and, through its action on kidney, antidiuresis (see Chap. 25 and Chap. 206). All of these hormonal actions contribute to the maintenance of cardiac function and adequate perfusion pressure.28,29 In addition, the behavioral and renal actions of the hormones tend to restore extracellular fluid volume to normal. Clearly, in some situations, if one hormonal system is removed, the other systems take over to maintain arterial blood pressure within its normal range. However, all the systems are necessary to combat severe changes in the system. For example, resting arterial pressure is not affected by pharmacologic blockade of the receptors for vasopressin or angiotensin II. However, the administration of these same blocking agents or blockade of the sympathetic nervous system seriously impairs the restitution of arterial pressure after hemorrhage.30 The endocrine responses to hemorrhage enumerated earlier and in Figure 5-9 do not represent a complete list of the hormonal changes elicited by hemorrhage-induced changes in baroreceptor input to hypothalamic control systems. Increased secretion of ACTH, growth hormone, and pancreatic glucagon occurs, and pancreatic insulin secretion decreases. After the surgical removal of baroreceptor input to the central autonomic pathways, hemorrhage no longer elicits either vasopressin or ACTH secretion, strongly suggesting that it is this information that drives the responses.

CONCLUSION
Feedback information allows the control systems to adjust appropriately to the internal conditions and to diminish over-shoots in endocrine responses that are occasioned by long delays between hormonal secretion and target tissue response. CHAPTER REFERENCES
1. Houk JC. Control strategies in physiological systems. FASEB J 1988; 2:97. 2. Guyton AC. Textbook of medical physiology, 9th ed. Philadelphia: WB Saunders, 1996:7. 3. Engler D, Redei E, Kola I. The corticotropin-release inhibitory factor hypothesis: a review of the evidence for the existence of inhibitory as well as stimulatory hypophysiotropic regulation of adrenocorticotropin secretion and biosynthesis. Endocrine Rev 1999; 20:460. 4. Keller-Wood ME, Dallman MF. Corticosterone inhibition of ACTH secretion. Endocr Rev 1984; 5:1. 5. Dallman MF, Akana SF, Levin N, et al. Corticosteroids and the control of function in the hypothalamic-pituitary-adrenal (HPA) axis. Ann NY Acad Sci 1994; 746:22. 6. Frohman LA, Jansson J. Growth-hormone releasing hormone. Endocr Rev 1986; 7:223. 7. Muller EE. Neural control of somatotropic function. Physiol Rev 1987; 67:962. 8. Dreifuss JJ, Kelly JS. Recurrent inhibition of antidromically identified rat supraoptic neurons. J Physiol (Lond) 1972; 200:87. 9. Moos F, Freund-Merier MJ, Guerne Y, et al. Release of oxytocin and vasopressin by magnocellular nuclei in vitro: specific facilitatory effects of oxytocin on its own release. J Endocrinol 1984; 102:63. 10. Denef C. Paracrine interactions in the anterior pituitary. Clin Endocrinol Metab 1986; 15:1. 11. Schwartz J, Cherny R. Intercellular communication within the anterior pituitary influencing the secretion of hypophysial hormones. Endocr Rev 1992; 13:453. 12. Lambertz SW, Macleod RM. Regulation of prolactin secretion at the level of the lactotroph. Physiol Rev 1990; 70:279. 13. Plant TM. Gonadal regulation of hypothalamic gonadotropin-releasing hormone release in primates. Endocr Rev 1986; 7:75. 13a. Evans JJ. Modulation of gonadotropin levels by peptides acting at the anterior pituitary gland. Endocr Rev 1999; 20:46. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. Vitale ML, Chiocchio SR. Serotonin, a neurotransmitter involved in the regulation of luteinizing hormone release. Endocr Rev 1993; 14:480. Carey RM, Sen S. Recent progress in the control of aldosterone secretion. Recent Prog Horm Res 1986; 42:251. MacDonald BR. Parathyroid hormone, prostaglandins and bone resorption. World Rev Nutr Diet 1986; 47:163. Schnermann J. Juxtaglomerular cell complex and the regulation of renal salt. Am J Physiol 1998; 274:2263. Dallman MF. Viewing the ventromedial hypothalamus from the adrenal gland. Am J Physiol 1984; 246:R1. Reid IA. Actions of angiotensin II on the brain: mechanisms and physiologic role. Am J Physiol 1984; 246:F533. Schwartz MW, Figlewicz DP, Baskin DG, et al. Insulin in the brain: a hormonal regulator of energy balance. Endocr Rev 1992; 13:387. Dallman MF, Strack AM, Akana SF, et al. Feast and famine: critical role of glucocorticoids with insulin in daily energy flow. Front Neuroendocrinol 1993; 14:303. Darlington DN, Tehrani MJ. Blood flow, vascular resistance, and blood volume after hemorrhage in conscious adrenalectomized rat. J Appl Physiol 1997; 83:1648. Darlington DN, Jones RO, Marzella L, Gann DS. Changes in regional vascular resistance and blood volume after hemorrhage in fed and fasted awake rats. J Appl Physiol 1995; 78:2025. Darlington DN, Keil LC, Dallman MF. Potentiation of hormonal responses to hemorrhage and fasting, but not hypoglycemia in conscious adrenalectomized rats. Endocrinology 1989; 125:1398. Darlington DN, Chew G, Ha T, et al. Corticosterone, but not glucose, treatment enables fasted adrenalectomized rats to survive moderate hemorrhage. Endocrinology 1990; 127:766. Thrasher TN, Keenan CR, Ramsay DJ. Cardiovascular afferent signals and drinking in response to hypotension in dogs. Am J Physiol 1999; 277:R795. Burnay MM, Python CP, Vallotton MB, et al. Role of the capacitative calcium influx in the activation of steroidogenesis by angiotensin-II in adrenal glomerulosa cells. Endocrinology 1994; 135:751. 28. Dampney RA. Functional organization of central pathways regulating the cardiovascular system. Physiol Rev 1994; 74:323. 29. Herd JA. Cardiovascular response to stress. Physiol Rev 1991; 71:305. 30. Rossi NF, Shrier RW. Role of arginine vasopressin in regulation of systemic arterial blood pressure. Annu Rev Med 1986; 37:13.

CHAPTER 6 ENDOCRINE RHYTHMS Principles and Practice of Endocrinology and Metabolism

CHAPTER 6 ENDOCRINE RHYTHMS

EVE VAN CAUTER Circadian Rhythms Basic Mechanisms Medical Implications Ultradian Rhythms Period Range and Origin Physiologic Significance Methodologic Aspects of the Study of Endocrine Rhythms Diurnal Variations Pulsatile Variations Rhythms in the Somatotropic Axis Rhythms in the Pituitary-Adrenal Axis Rhythms in Prolactin Secretion Rhythms in the Gonadotropic Axis Rhythms in the Thyrotropic Axis Rhythms in Glucose Regulation Chapter References

Pronounced temporal oscillations ranging in period from a few minutes to a year occur throughout the endocrine system. These oscillations are part of the wide variety of rhythms that exist in all living organisms. Rhythms with a period of ~24 hours have been called circadian, from the Latin circa diem, meaning around a day. According to the terminology of circadian biology, 24-hour rhythms should only be referred to as circadian if they are primarily driven by an endogenous clock system. However, it has become increasingly apparent that nearly all 24-hour rhythms partly reflect the influence of an endogenous circadian pacemaker located in the hypothalamus and partly the influence of other events occurring with a 24-hour periodicity, such as the alternation of waking and sleeping, light and dark, food intake, and postural changes. Thus, the use of the term circadian is often extended to all diurnal variations recurring regularly at a time interval of ~24 hours. Circadian rhythms have been used to separate two further classes of biologic rhythms: ultradian rhythms, with periods shorter than 24 hours, and infradian rhythms, with periods longer than 24 hours. The pulsatile release of pituitary hormones belongs to the ultradian range. The menstrual cycle and the seasonal variations in the reproductive system are infradian rhythms. Reproducible circadian changes occur in the secretory pattern of the vast majority of hormones.

CIRCADIAN RHYTHMS
BASIC MECHANISMS There is virtually neither a tissue nor a function in the human organism that does not exhibit regular changes from day to night. Distinct patterns of secretion recur at 24-hour intervals for essentially all hormones. Much of the temporal variability and organization of endocrine release ultimately results from the activity of two interacting time-keeping mechanisms in the central nervous system: endogenous circadian rhythmicity and sleep-wake homeostasis (Fig. 6-1). In mammals, endogenous circadian rhythmicity is generated by a pacemaker or group of oscillators located in the paired suprachiasmatic nucleus (SCN) of the hypothalamus.1 Sleep-wake homeostasis is an hourglass-like mechanism relating the amount and quality of sleep to the duration of prior wakefulness.2

FIGURE 6-1. Schematic representation of the central mechanisms involved in the control of temporal variations in pituitary hormone secretions over the 24-hour cycle. Sleep-wake homeostasis is in an hourglass-like mechanism, which relates the propensity for deep non-rapid eye movement (non-REM) sleep to the amount of prior wakefulness. Circadian rhythmicity is an endogenous near-24-hour oscillation generated in the suprachiasmatic nuclei (SCN) of the hypothalamus and transmitted via neural as well as humoral mechanisms. (ACTH, adreno-corticotropic hormone; FSH, follicle-stimulating hormone; GH, growth hormone; LH, luteinizing hormone; PRL, prolactin; TSH, thyrotropin-stimulating hormone.)

The endogenous nature of human circadian rhythms has been established by experiments in which subjects were isolated in a soundproof chamber with no time cues available. Under those conditions, the subjects show free-running rhythms with periods, which deviate slightly from 24 hours and vary from one individual to another. In humans, the endogenous free-running period is usually slightly longer than 24 hours. In the past few years, enormous progress has been made in elucidating the molecular and genetic mechanisms underlying the generation of circadian rhythms in mammals. After the discovery of the first mammalian circadian clock gene, called Clock,3,4 a number of other mammalian clock genes were rapidly identified and sequenced.5 Intense efforts are now under way to understand how these genes and their protein products interact with each other through a series of activational and inhibitory pathways to produce the precise periodicity of the circadian signal generated in the SCN. Environmental agents, primarily the light-dark cycle (i.e., photic cues), affect the expression and properties of the circadian signal. Light exposure in the later part of the subjective night and in the early part of the subjective day (e.g., around dawn) generally advances the phase of the circadian oscillation. In contrast, light exposure in the later part of the subjective day and during the early part of the subjective night (e.g., around dusk) delays circadian phase. Light-dark information reaches the SCN via a direct afferent pathway from the retina as well as an indirect projection via the intergeniculate leaflet of the thalamus.1 There also is good evidence to indicate that nonphotic cues are also capable of altering circadian function. Many of the non-photic stimuli that induce phase shifts in the circadian clock appear to do so by either stimulating physical activity or activating the pathways by which information about the activity-rest state of the organism reaches the clock.1 Phase-shifts may occur when activity pathways are stimulated during the normal rest period or, vice-versa, when rest is enforced during the normal active period. As is the case for light, the magnitude and direction of the phase-shifts depend on their timing relative to the internal circadian clock. There is now substantial evidence to indicate that the Raphe nuclei of the brainstem are involved in mediating the effects of activity on the circadian clock. Sleep-wake homeostasis is thought to involve a putative neural sleep factor (factor S) that rises during waking and decays exponentially during sleep,2 and regulates the timing, amount, and intensity of the deeper stages of sleep. These stages of deep sleep correspond to stages III and IV of non rapid-eye-movement (REM) sleep, usually referred to as slow-wave (SW) sleep because of the appearance of well-defined waves in the frequency range 0.5 to 4.0 Hz in the electroencephalogram. The neuroanatomical and neurochemical basis of sleep-wake homeostasis has not been entirely elucidated but appears to involve the basal forebrain cholinergic region and adenosine, a neuromodulator of which the extracellular concentrations in this region increase during sustained wakefulness and decrease during sleep.6 The mechanisms controlling sleep propensity and maintenance in the human appear to be somewhat different from those occurring in most other mammalian species. Indeed, human sleep is generally consolidated in a single 7-to 9-hour period, whereas fragmentation of the sleep period in several bouts is the rule in most other mammals. Possibly as a result of this consolidation of the sleep period, the wake-sleep and sleep-wake transitions in the human are associated with physiologic changesand, in particular, endocrine changesthat are more marked than those observed in animals. Humans are also unique in their capacity to maintain wakefulness despite an increased pressure to sleep. While decrements in mood and cognitive function associated with such behaviors have long been recognized, these behaviors have been found to be associated with endocrine and metabolic alterations that may result in long-term adverse health consequences.6a The pathways by which circadian rhythmicity, sleep-wake homeostasis, and their interactions modulate hormonal release are largely unknown. At the level of the

central nervous system (see Fig. 6-1), humoral and/or neural signals originating from the hypothalamic circadian pacemaker and from brain regions involved in sleep regulation affect the activity of the hypothalamic structures responsible for the pulsatile release of neuroendocrine factors (e.g., corticotropin-releasing hormone [CRH], growth hormonereleasing hormone [GHRH], etc.), which stimulate or inhibit intermittent secretion of pituitary hormones. It appears that stimulatory or inhibitory effects of sleep on endocrine release are primarily associated with SW sleep, rather than REM sleep. Theoretically, the modulation of neuroendocrine release by sleep and circadian rhythmicity could be achieved by modulation of pulse amplitude, modulation of pulse frequency, or a combination of both. The data available so far seem to indicate that circadian rhythmicity of pituitary hormonal release is achieved primarily by modulation of pulse amplitude without changes in pulse frequency, whereas sleep-wake and REM/non-REM transitions affect pulse frequency. Pituitary hormones that influence endocrine systems not directly controlled by hypothalamic factors probably mediate, at least partially, the modulatory effects of sleep and circadian rhythmicity on these systems (e.g., counterregulatory effects of growth hormone [GH] and cortisol on glucose regulation). Figure 6-2 shows examples of 24-hour patterns of plasma cortisol, thyrotropin (thyroid-stimulating hormone [TSH]), prolactin (PRL), and GH levels observed in normal young men in the presence (left) and in the absence (right) of sleep. To eliminate the effects of feeding, fasting, and postural changes, the subjects remained recumbent throughout the study, and the normal meal schedule was replaced by intravenous glucose infusion at a constant rate. These profiles exemplify the high degree of temporal organization provided by circadian rhythms in the endocrine system. As is the case for the majority of hormones, the plasma levels of these four pituitary-dependent hormones follow a pattern that repeats itself day after day. The nocturnal rise of TSH starts in the evening, at a time when cortisol secretion is quiescent, and ends at the beginning of the sleep period, when GH and PRL concentrations surge. The early morning period is associated with low TSH, low PRL, and low GH concentrations but high cortisol levels. Thus, the release of these four hormones follows a highly coordinated temporal program that results from interactions between circadian rhythmicity, sleep, and pulsatile release. It is apparent that the presence or absence of sleep had only modest effects on the wave shape of the cortisol profile but markedly affected the profilers of TSH, PRL, and GH. Thus, the relative contributions of circadian rhythmicity and sleep-wake homeostasis vary from one endocrine axis to the other, but, as will be shown in subsequent sections, inputs from both central nervous system processes can be recognized in the majority of 24-hour hormonal patterns.

FIGURE 6-2. From top to bottom: Temporal profiles of plasma cortisol (COR), thyrotropin (thyroid-stimulating hormone [TSH]), prolactin (PRL), and growth hormone (GH) in a healthy male adult studied over a 24-hour cycle with a normal sleep period (left) or continuously awake (right). The black bars denote the sleep periods; the open bars the periods of nocturnal sleep deprivation.

MEDICAL IMPLICATIONS Among the medical implications of circadian rhythms, those relating to clinical diagnosis are the most obvious. Indeed, because of the wide amplitude of certain rhythms, the estimation of the mean level of a parameter from a single measurement may involve an error exceeding 100%. For example, a plasma cortisol level of 15 g/dL (i.e., 414 nmol/L) at 8:00 a.m. is perfectly normal, whereas the same value obtained at 8:00 p.m. is suggestive of some form of hypercortisolism. Differentiation between normal and pathologic levels may be greatly improved by adequately selecting the time of sample collection. To illustrate this concept, Figure 6-3 shows the mean of 56 profiles of plasma cortisol from patients with Cushing syndrome as compared with the mean of 60 profiles obtained in normal subjects. Overlap between individual values from the two groups of subjects should be expected at all times except between 10:00 p.m. and 2:00 a.m., when the mean cortisol level allows for the discrimination between healthy controls and patients with hypercortisolism in 114 out of 116 cases. As discussed later in this section, it is also important to find out whether the patient has been recently involved in night work or has just returned from transmeridian travel when interpreting certain hormonal data for diagnostic purposes.

FIGURE 6-3. Mean, across individuals, of the 24-hour profile of plasma cortisol in 60 normal subjects (closed circles) and 56 patients with Cushing syndrome (open circles). The hatched and shaded areas represent one standard error of the mean above and below the mean for normal subjects or patients with Cushing syndrome, respectively.

Another medical implication of circadian rhythms is based on the fact that the response of the organism to many stimuli depends on the time at which that stimuli is applied. This finding forms the theoretical basis of chronopharmacology (i.e., the investigation of drug effects as a function of their time of administration) and of chronotherapy (i.e., the design of better protocols of treatment that take into account the chronobiologic characteristics of the system).7 Abnormal 24-hour regulation of endocrine and other physiologic functions occur in a variety of highly prevalent conditions, including shift work, jet lag (see Chap. 10), blindness, major depressive illness (see Chap. 201), and sleep disorders. The health symptoms of shift workers and the general feeling of discomfort experienced after a transmeridian flight are well-known consequences of the desynchronization between internal and external time. Both are associated with a variety of physical and performance deficits. The jet lag syndrome usually includes symptoms of fatigue, subjective discomfort, sleep disturbances, reduced mental and psychomotor performance, and gastrointestinal tract disorders. The malaise partly reflects a state of internal desynchrony, because different physiologic systems adapt to abrupt shifts of environmental time at different rates. The syndrome subsides as adaptation to local time is achieved. Unless the condition of jet lag is repeated at frequent time intervals for prolonged periods of time (as in the case of air transportation professionals), this transient syndrome is not thought to be associated with long-term adverse effects on physical and mental health. In contrast, shift work, which is voluntarily accepted by millions of workers, is a major health hazard, involving an increased risk of cardiovascular illness, gastrointestinal disorders, psychosocial symptoms, sleep disturbances, substance abuse, reduced immune function, and infertility.8,9 and 10 Shift workers are generally in a condition of chronic sleep debt,9 which could in itself result in endocrine and metabolic disturbances. Workers on permanent or rotating night shifts do not fully adapt to these schedules, even after several years,11,12 and live in a chronic state of internal desynchrony of the endocrine system. An example is shown in Figure 6-4, which compares mean 24-hour profiles of plasma cortisol and PRL in permanent night workers as compared with day workers.13,14 and 15 The components of these endocrine rhythms, which are primarily controlled by sleep-wake homeostasis (e.g., sleep-related PRL secretion), partly adapted to the night schedule, whereas components reflecting circadian timing (i.e., onset of the early morning elevation of cortisol secretion) showed little, if any, adaptation. As a result, the night workers had to initiate sleep when cortisol levels were already high and maintain wakefulness despite elevated concentrations of PRL, a hormone thought to be involved in sleep regulation.

FIGURE 6-4. Mean 24-hour profiles of plasma cortisol (left) and prolactin (right), either in healthy subjects who work during the daytime or in healthy subjects who have been working during the night for at least 2 years. The vertical lines at each time point represent the standard error of the mean. The black bars represent the sleep periods. (Data from Weibel L, Follenius M, Spiegel K, et al. Comparative effect of night and daytime sleep on the 24-hour cortisol secretory profile. Sleep 1995; 18:549; Weibel L, Spiegel K, Follenius M, et al. Internal dissociation of the circadian markers of the cortisol rhythm in night workers. Am J Physiol 1996; 270:E608; Spiegel K, Weibel L, Gronfier C, et al. Twenty-four hour prolactin profiles in night workers. Chronobiol Int 1996; 13:283.)

Abnormal synchronization of circadian rhythms, including hormonal rhythms, have also been observed in totally blind subjects. These disturbances are thought to reflect a lack of entrainment to the 24-hour environmental periodicity that is due to the absence of photic synchronization. In severely depressed subjects, early timing of a number of circadian rhythms, including hormonal rhythms, has been observed. These findings provide the basis for the phase-advance hypothesis for affective illness, which proposes that abnormalities in circadian time-keeping are involved in the pathophysiology of depression.16 Patients experiencing sleep apnea, a condition in which frequent awakenings interrupt sleep, have abnormalities of sleep-related endocrine release. When the condition is untreated, nocturnal PRL and GH levels do not increase to the same extent as in healthy subjects with normal sleep. Studies that have examined the impact of treatment with continuous positive airway pressure have demonstrated that treatment of the sleep disorder partly normalizes the endocrine alteration.17,17a This is illustrated in the profiles shown in Figure 6-5 in the case of GH. After 3 months of treatment, nasal continuous positive airway pressure therapy increases plasma insulin-like growth factor-I levels in men with severe obstructive sleep apnea.18 In children, surgical correction of obstructive sleep apnea may restore GH secretion and a normal growth rate.19

FIGURE 6-5. Mean profiles of plasma growth hormone (GH) in patients with sleep apnea studied before (top) and after (bottom) treatment with continuous positive airway pressure. The vertical line at each time point represents the standard error of the mean. (Data from Saini J, Krieger J, Brandenberger G, et al. Continuous positive airway pressure treatment: effects on growth hormone, insulin and glucose profiles in obstructive sleep apnea patients. Horm Metab Res 1993; 25:375. Reproduced from Van Cauter E, Spiegel K. Circadian and sleep control of hormonal secretions. In: Turek FW, Zee PC, eds. Regulation of sleep and circadian rhythms. New York: Marcel Dekker, Inc., 1999:397.)

Age-related changes in endocrine, metabolic, and behavioral circadian rhythms have been well described.20,21 One of the most prominent changes is a reduction in rhythm amplitude. The overall findings of a study that examined age-related differences in 24-hour endocrine rhythms and sleep in healthy subjects are shown in Figure 6-6. The cortisol and TSH rhythms were dampened in the older group. A decrease by at least 50% in the nocturnal release of both GH and melatonin was observed in the older volunteers, and SW sleep was drastically diminished. Deficits in the maintenance and depth of nocturnal sleep are generally paralleled by decreased alertness during the daytime. Many circadian rhythms are also phase advanced in older subjects such that specific phase points of the rhythms occur earlier than in young subjects.21 The alterations in circadian regulation are closely associated with changes in sleep-wake habits (i.e., earlier bedtimes and waketimes).

FIGURE 6-6. From top to bottom, mean 24-hour profiles of plasma cortisol, thyrotropin (thyroid-stimulating hormone [TSH]), melatonin, prolactin, and growth hormone (GH) levels, and distribution of slow-wave (SW) and rapid-eye-movement (REM) stages in sleep in old and young subjects. The distribution of SW and REM is expressed in minutes of each 15-minute sampling interval spent in the corresponding stage. Vertical lines at each time point represent the standard error of the mean. The black bars correspond to the mean sleep period. (From van Coevorden A, Mockel J, Laurent E, et al. Neuroendocrine rhythms and sleep in aging. Am J Physiol 1991; 260:E651.)

ULTRADIAN RHYTHMS
PERIOD RANGE AND ORIGIN The term ultradian is primarily used to designate rhythmicities with periods ranging from fractions of hours to several hours. In the hourly range, the most prominent ultradian rhythms are the alteration of REM and non-REM stages in sleep and the pulsatility of hormonal secretions. Oscillations in the 80- to 120-minute range also occur at the rate of urine flow and gastric motility. Pulsatile hormonal release is ubiquitous in the endocrine system. Indeed, the plasma levels of most hormones undergo episodic fluctuations of variable duration and magnitude, referred to as secretory episodes or pulses. These pulses recur at 1- to 4- hour intervals (see Fig. 6-2). Pulsatility has been observed for anterior and posterior pituitary hormones; for hormones directly under their control; and for other endocrine variables, such as insulin, glucagon, and renin. The long-disputed theory that ultradian variations of brain activitysimilar to those occurring during sleepwere also present during wake (constituting a basic rest-activity cycle) has received some experimental support. One study demonstrated the existence of an ultradian rhythm of brain electrical activity in the frequency range of 13 to 35 Hz, an index of central alertness, during waking.22 Furthermore, it appeared that pulses of cortisol release were significantly

associated with increases in this marker of alertness. In discussing the origin of ultradian hormonal fluctuations, one must distinguish two general cases: that of the hormones under direct hypothalamic control and that of hormones that are part of more peripheral endocrine systems. In the first case, periodic release ultimately reflects phasic neural activation, whereas in the second case, oscillatory behavior is a property of the dynamics of local regulatory networks. The episodic pulses of anterior pituitary hormones, which are representative of the first case, result from secretion in discrete bursts in response to pulsa-tile stimulation and/or inhibition by hypothalamic factors. The specific hypothalamic mechanism controlling pulsatile release appears to be different for each pituitary axis, but common stimuli may operate in different axes. The state of knowledge is most advanced in the case of the gonadotropins, for which pulsatile release is caused by intermittent discharges of gonadotropin-releasing hormone (GnRH) into the pituitary portal circulation. The GnRH pulses are obtained by synchronous discharges of GnRH-containing neurons in the arcuate of the mediobasal hypothalamus.23 Sharp increases in the frequency of hypothalamic multiunit electrical activity are associated with each pulse of peripheral luteinizing hormone (LH) level.23 The pulsatile behavior is intrinsic to the GnRH neurons, as attested by their coordinate secretion in tissue cultures.24 Intermittent stimulation by pituitary hormones is then in turn responsible for the episodic release of hormones under their control (see Chap. 16). For hormones other than those controlled by the hypo-thalamo-pituitary axis, the mechanisms causing episodic variations in plasma levels are generally less well understood. Notable exceptions are the ultradian 1- to 2-hour oscillations of insulin secretion that occur after meal ingestion, as well as during constant glucose infusion or continuous enteral nutrition.25,26 Theoretical and experimental evidence suggests that in conditions of normal glucose tolerance, these oscillations arise from the negative feedback interactions linking insulin and glucose.26 PHYSIOLOGIC SIGNIFICANCE The physiologic significance of hormonal pulsatility has been first demonstrated by experiments showing that normal LH and follicle-stimulating hormone (FSH) levels may be restored by pulsatile, but not continuous, administration of exogenous GnRH to primates with lesions in the hypothalamus that had abolished endogenous GnRH production.23 These findings were rapidly applied to the treatment of a variety of disorders of the pituitary-gonadal axis, using either pulsatile GnRH administration to correct a deficient production of endogenous GnRH or long-acting GnRH analogs to induce pituitary desensitization.27 Similar approaches have been investigated for other hypothalamo-pituitary axes.

METHODOLOGIC ASPECTS OF THE STUDY OF ENDOCRINE RHYTHMS

DIURNAL VARIATIONS To investigate circadian hormonal rhythms, a group of individuals is usually studied for a minimum of 24 hours each, following the same experimental protocol. The demonstration of circadian rhythmicity is then based on the observation of consistently reproducible characteristics in the set of 24-hour profiles obtained. To validate such an approach, the group of subjects should be as homogeneous as possible, not only in terms of physical parameters, such as age and gender, but also in terms of living habits, such as bedtimes and meal schedules. To maximize interindividual synchronization, the volunteers should comply with a standardized schedule of meals and bedtimes for several days before the investigation. Blood sampling is usually done at regular intervals through a catheter inserted into a forearm vein. During bedtime hours, the catheter is connected to plastic tubing extending to an adjacent room to collect blood samples without disturbing the subject. Because of the major modulatory effects exerted by sleep stages on hormonal release, it is important to obtain polygraphic sleep recordings using standardized methods for recording and scoring. If polygraphic monitoring is not possible, estimates of sleep onset and awakenings should be carefully recorded. Daytime naps should be avoided. To obtain valid estimations of the circadian parameters, it is necessary to sample at intervals not exceeding 1 hour. Indeed, the pulsatile variations may bias the estimation of the characteristics of the circadian rhythm if sampling is less frequent. Procedures to analyze 24-hour hormonal profiles usually involve the fitting of a smooth curve to the data. The times of occurrence of the maximum and the minimum of the best-fit curve are often referred to as, respectively, the acrophase and the nadir. The amplitude of the circadian rhythm may be estimated as 50% of the difference between the maximum and the minimum of the best-fit curve. PULSATILE VARIATIONS The definition of an optimal sampling protocol to study pulsatile hormonal fluctuations depends on the type of phenomenon under study. Presently, sampling at 1-minute intervals represents probably the fastest rate technically achievable with reasonable precision. Sampling every few minutes uncovers high-frequency, low-amplitude variations superimposed on the slower pulsatile release occurring at hourly intervals. These intensified rates of venous sampling may not allow for the estimation of the characteristics of major secretory bursts, because the total duration of sampling compatible with this rate of blood withdrawal limits the observed number of large peaks. Sampling rates of 20 and 30 minutes only detect major pulses lasting >1 hour. The analysis of pulsatile variations may be considered at two levels. The researcher may wish to define and characterize significant variations in peripheral levels based on estimations on the size of measurement error (i.e., primarily assay error). However, under certain circumstances, it is possible to mathematically derive secretory rates from the peripheral concentrations. This procedure, often referred to as deconvolution, often reveals more pulses of secretion than the analysis of peripheral concentrations. It also more accurately defines the temporal limits of each pulse. The association between pulsatile GH secretion and sleep stages in a single individual is shown in Figure 6-7.28 The profile shown on the top represents the plasma levels of GH measured at 15-minute intervals. Three pulses were found significant using a pulse detection algorithm (i.e., ULTRA28). The corresponding profile of GH secretory rates calculated by deconvolution is shown in the second profile from the top. A single compartment model for GH disappearance with a half-life of 19 minutes and a volume of distribution of 7% of the body weight was used in this calculation. Pulse analysis of the secretory rates now reveals the occurrence of three additional pulses of GH secretion. The three lower profiles show the percentages of each 15-minute sampling interval spent in stages of wake, SW, and REM, respectively. When the profile of plasma concentrations is compared with the SW profile, it appears that subsequent to its initiation in concomitance with the beginning of the first SW period, the sleep-onset GH pulse spanned the first 3 hours of sleep, without apparent modulation by non-REM and REM stages. However, the profile of secretory rates clearly reveals that GH was preferentially secreted during the SW stage, with interruptions of secretory activity coinciding with the intervening REM or wake stages. Deconvolution demonstrated a closer association between SW stages and active GH secretion than the analysis of plasma levels, because the temporal limits of each pulse were more accurately defined and additional pulses were revealed. The validity of the deconvolution procedure is critically dependent on the knowledge of the clearance kinetics of the hormone under study; extra caution in interpreting the data must be exerted, because this procedure involves an amplification of measurement error, with increased risk of false-positive error.

FIGURE 6-7. 24-hour profile of plasma growth hormone (GH; top) sampled at 15-minute intervals in a normal man. The black bar indicates the sleep period. The second panel from the top shows the profile of GH secretory rates derived by deconvolution from the profile of plasma concentrations. Significant pulses of plasma levels and secretory rates are indicated by arrows. The three lower panels represent the temporal distribution of slow-wave (SW) stages (III + IV), wake, and rapid eye movement (REM) during sleep. Vertical lines indicate the temporal association between pulses of GH secretion and SW stages. (From Van Cauter E. Computer-assisted analysis of endocrine rhythms. In: Rodbard D, Forti G, eds. Computers in endocrinology. New York: Raven Press, 1990:59.)

Whether examining peripheral concentrations or secretory rates, there are two major approaches to analyzing the episodic fluctuations. The first, and most commonly used, is the time domain analysis in which the data are plotted against time and pulses are detected and identified. The second is the analysis in the frequency domain in which amplitude is plotted against frequency or period. These two approaches differ fundamentally both in the mathematical treatment of the data and in the questions they may help to resolve; therefore, they should be viewed as being complementary. The regularity of pulsatile behavior may be quantified by both approaches (i.e., by examining the distribution of interpulse intervals derived from a time domain analysis or by examining the distribution of spectral power in a

frequency domain analysis). Additionally, another analytical tool, the approximate entropy, has been introduced to quantify regularity of oscillatory behavior in endocrine and other physiologic time series.29,30

RHYTHMS IN THE SOMATOTROPIC AXIS

In normal subjects, the 24-hour profile of plasma GH consists of stable low levels abruptly interrupted by bursts of secretion (see Fig. 6-2, Fig. 6-5, Fig. 6-6 and Fig. 6-7). The most reproducible secretory pulse occurs shortly after sleep onset, in association with the first phase of SW sleep.31 Other secretory pulses may occur in later sleep and during wakefulness in the absence of any identifiable stimulus. In women, daytime GH pulses are more frequent than in men, and the sleep-associated pulse, although still present, does not generally account for most of the 24-hour GH release. The secretory profile is less regular in women than in men.32 Circulating estradiol concentrations play an important role in determining overall levels of spontaneous GH secretion.33 Sleep onset elicits a pulse in GH secretion whether sleep is advanced, delayed, interrupted, or fragmented. Delta wave electroencephalographic activity consistently precedes the elevation in plasma GH levels. While SW sleep is clearly a major determinant of the 24-hour profile of GH secretion in humans, there is also evidence for the existence of a circadian modulation of the occurrence and height of GH pulses, reflecting decreased somatostatin inhibition in the evening and nocturnal hours.34 Administration of a specific GHRH antagonist results in a near total suppression of sleep-related GH release, indicating an important role for GHRH in the control of nocturnal GH secretion.35 Two studies involving pharmacologic stimulation of SW sleep have provided evidence for a common mechanism in the control of SW sleep and GH secretion and have indicated that compounds, which increase SW sleep, could represent a novel class of GH secretagogues. Indeed, enhancement of SW sleep by oral administration of low g-hydroxybutyrate (a naturally occurring metabolite of g-aminobutyric acid used in the treatment of narcolepsy) or ritanserin (a selective 5HT2 receptor antagonist) results in simultaneous and highly correlated increases of nocturnal GH release.36,37 The total amount and the temporal distribution of GH release are strongly dependent on age.31 A pulsatile pattern of GH release, with increased pulse amplitude during sleep, is present in prepubertal boys and girls. During puberty, the amplitude of the pulses, but not the frequency, is increased, particularly at night. Maximal overall GH concentrations are reached in early puberty in girls and in late puberty in boys. Age-related decreases in GH secretion have been well documented in both men and women and are illustrated in Figure 6-6. There is a marked suppression of GH levels throughout the 24-hour span in obese subjects. In normal-weight subjects, fasting, even for only 1 day, enhances GH secretion via an increase in both pulse amplitude and pulse frequency.38 Nonobese juvenile or maturity-onset diabetic patients hypersecrete GH during wakefulness as well as during sleep, primarily because of an increase in the amplitude of pulses.39 This abnormality may disappear when blood sugar levels are strictly controlled. In acromegaly, GH is hypersecreted throughout the 24-hour span, with a pulsatile pattern superimposed on elevated basal levels, indicative of the presence of tonic secretion.40,41 After trans-sphenoidal surgery, a normal circadian pattern of GH release can be restored.41 In contrast, bromocriptine therapy lowers the overall GH secretion but does not lead to the resumption of normal 24-hour profiles.

RHYTHMS IN THE PITUITARY-ADRENAL AXIS

Twenty-four-hour profiles of cortisol typical of normal subjects are shown in Figure 6-2, Figure 6-3, Figure 6-4 and Figure 6-6. The patterns of plasma adrenocorticotropic hormone (ACTH) and cortisol variations show an early morning maximum, declining levels throughout daytime, a quiescent period of minimal secretory activity, and an abrupt elevation during late sleep. With a 15-minute sampling rate, ~15 pulses of ACTH and cortisol can be detected in a 24-hour span. The cortisol profiles shown in the upper panels of Figure 6-2 illustrate the remarkable persistence of the wave shape of the rhythm in the absence of sleep and support the notion that the 24-hour periodicity of corticotropic activity is primarily controlled by circadian rhythmicity. Nevertheless, modulatory effects of the sleep or wake condition have been clearly demonstrated. Sleep onset is reliably associated with a short-term inhibition of cortisol secretion,42,43 and 44 although this effect (which appears to be related to slow-wave stages45) may not be detectable when sleep is initiated at the peak of corticotropic activity (i.e., in the morning13). Conversely, awakening at the end of the sleep period is consistently followed by a pulse of cortisol secretion.46,47 During sleep deprivation, these rapid effects of sleep onset and sleep offset on corticotropic activity are obviously absent, and the amplitude of the rhythm is reduced as compared with normal conditions (see Fig. 6-2). In addition to the immediate modulatory effects of sleep-wake transitions on cortisol levels, nocturnal sleep deprivationeven partialresults in higher-than-normal cortisol concentrations on the following evening.48 Sleep loss, thus, appears to delay the normal return to evening quiescence of the corticotropic axis. This suggests that sleep loss may slow down the rate of recovery of the hypothalamicpituitaryadrenal axis response after a challenge. A circadian variation parallel to that of cortisol occurs for the plasma levels of adrenal steroids.49 Figure 6-8 shows the mean 24-hour profiles of cortisol, 11-hydroxyandrostenedione, dehydroepiandrosterone, and androstenedione (AD) for 10 normal young men. The amplitude of the circadian variation in 11-hydroxyandrostenedione levels, a steroid derived from adrenal AD, is essentially similar to that of cortisol, whereas the amplitude of the variations in dehydroepiandrosterone and AD levels, two steroids of partially gonadal origin, is much lower.49 Pulses of the plasma concentrations of adrenal steroids occur in remarkable synchrony with bursts of cortisol secretion, indicating that pulsatile ACTH release is reflected in the temporal organization of all adrenal secretions.

FIGURE 6-8. Mean 24-hour profiles of plasma cortisol, 11-hydroxyan-drostenedione (11-OAD), dehydroepiandrosterone (DHEA), and andro-stenedione (AD) obtained at 15-minute intervals in 10 normal young men. For each subject and each hormone, the data were expressed as a percentage of the individual 24-hour mean before calculating group values. The vertical bars at each time point represent the standard error of the mean. (Data from Lejeune-Lenain C, Van Cauter E, Desir D, et al. Control of circadian and episodic variations of adrenal androgens secretion in man. J Endocrinol Invest 1987; 10:267.)

A distinct circadian rhythm of serum cortisol levels emerges at ~6 months of age. Once this periodicity has been established, it persists throughout adulthood and has been observed through the ninth decade.21 The overall pattern of the rhythm remains unchanged (see Fig. 6-6), but in older subjects, the nadir is advanced by 1 to 2 hours and the amplitude is decreased.21 In women, oral contraceptive therapy results in a large increase of the mean cortisol level and of the amplitude of the rhythm resulting from an estrogen-induced elevation of transcortin-binding capacity. Thus, when hypercortisolism is suspected in a female patient, it is essential to know whether the patient is receiving estrogen treatment. The 24-hour profile of pituitary-adrenal secretion remains unaltered in a wide variety of pathologic states. Disease states in which alterations of the cortisol rhythm have been observed50 include primarily (a) disorders involving abnormalities in binding and/or metabolism of cortisol, (b) the various forms of Cushing syndrome, and (c) severe depression. The relative amplitude of the circadian rhythm and of the episodic fluctuations of cortisol is blunted in patients with liver disease and in patients with anorexia nervosa, primarily because of the decreased metabolic clearance of cortisol. In contrast, in hyperthyroidism, for which cortisol production and peripheral metabolism are increased, episodic pulses are enhanced. In hypothyroid patients, there is diminished cortisol clearance, the mean level is markedly elevated, and the relative amplitude of the rhythm is, therefore, dampened. Figure 6-9 illustrates typical 24-hour cortisol patterns of plasma cortisol and cortisol secretory rates in a patient with pituitary Cushing disease and a patient with major endogenous depression as compared with a normal subject. In patients with Cushing syndrome secondary to adrenal adenoma or ectopic ACTH secretion, the circadian variation of plasma cortisol is invariably absent. However, in pituitary-dependent Cushing disease, a low-amplitude circadian variation may persist, suggesting that there is no defect in the neural clock generating the periodicity. Cortisol pulsatility is blunted in ~70% of patients with Cushing disease, suggesting autonomous tonic secretion of ACTH by a pituitary tumor. However, in ~30% of these patients, the magnitude of the pulses is instead enhanced. These hyperpul-satile patterns could be caused by enhanced hypothalamic release of CRH or persistent pituitary responsiveness to CRH. Hypercortisolism with persistent circadian rhythmicity and increased pulsatility is found in a majority of acutely depressed patients. In these patients, who do not

develop the clinical signs of Cushing syndrome despite the high cortisol levels, the quiescent period often occurs earlier than in normal subjects of comparable age. This phase-advance could reflect an alteration in the regulation of the circadian pacemaker system. When a clinical remission is obtained, the hypercortisolism and the abnormal timing of the quiescent period disappear, indicating that these disturbances are state rather than trait dependent. Contrasting with the increased cortisol pulsatility that characterizes many patients with major depression, a few studies have suggested that pulsatile variations are of lower amplitude in post-traumatic stress disorder and chronic fatigue syndrome.51,52

FIGURE 6-9. Twenty-four-hour profiles of cortisol secretory rate (top) and plasma cortisol (bottom) in a normal subject (left), a patient with pituitary Cushing disease (middle), and a patient with major endogenous depression of the unipolar subtype (right). Cortisol secretory rates were derived from plasma levels using a two-compartment model for cortisol distribution and metabolism. Note that circadian rhythmicity is markedly attenuated in the subject with Cushing disease, but it is preserved in the depressed patient. Cortisol secretion is entirely intermittent in the normal subject and the depressed patient, but the secretory pattern of the patient with Cushing disease shows evidence of tonic cortisol release.

RHYTHMS IN PROLACTIN SECRETION

Under normal conditions, the 24-hour profile of plasma PRL levels (see Chap. 13) follows a bimodal pattern, with minimal concentrations around noon, an afternoon phase of augmented secretion, and a major nocturnal elevation starting shortly after sleep onset and culminating around midsleep. Episodic pulses occur throughout the 24-hour span. Morning awakening is consistently associated with a brief PRL pulse.53,54 Studies on PRL during daytime naps or after shifts of the sleep period have demonstrated that sleep onset is invariably associated with an increase in PRL secretion. This is well illustrated by the profiles shown in Figure 6-2 in the presence and in the absence of sleep and in Figure 6-4, which compares the profiles of day and night workers. However, a sleep-independent circadian component of PRL secretion may be observed in some individuals, particularly in women.54 An example may be seen in the PRL profiles of night workers (see Fig. 6-4) in whom a nocturnal elevation occurred despite nocturnal activity.15 When sleep structure is characterized by power spectral analysis of the electroencephalogram, a close temporal association between increased PRL secretion and SW activity is clearly apparent.55 Conversely, prolonged awakenings, which interrupt sleep, are consistently associated with decreasing PRL concentrations. Thus, shallow and fragmented sleep is generally associated with lower nocturnal PRL levels. This is indeed what is observed in elderly subjects (see Fig. 6-6),21 who have an increased number of awakenings and markedly decreased amounts of SW sleep, and in whom a dampening of the nocturnal PRL rise is evident. Benzodiazepine hypnotics taken at bedtime often cause an increase in the nocturnal PRL rise, resulting in concentrations in the pathologic range for part of the night.56 Absence or blunting of the nocturnal increase of plasma PRL has been reported in a variety of pathologic states, including uremia, breast cancer in postmenopausal women, and Cushing disease. In subjects with insulin-dependent diabetes, the circadian and sleep modulation of PRL secretion is preserved, but overall levels are markedly diminished.57 In hyperprolactinemia associated with prolactinomas, the nocturnal elevation of PRL is preserved in patients with microadenomas but altered in patients with macroadenomas.58 Selective removal of PRL-secreting microadenomas can result in the normalization of the PRL pattern.

RHYTHMS IN THE GONADOTROPIC AXIS

Rhythms in the gonadotropic axis cover a wide range of frequencies, from episodic release in the ultradian range to diurnal rhythmicity and monthly and seasonal cycles. These various rhythms interact to provide a coordinated temporal program governing the development of the reproductive axis and its operation at every stage of maturation. The following summary is centered on 24-hour rhythms and their interaction with pulsatile release at the various stages of maturation of the human reproductive system. More detailed reviews may be found elsewhere.50,59,60 Before puberty, LH levels are very low. Both LH and FSH appear to be secreted in a pulsatile pattern but with a very low amplitude, which is insufficient to activate the gonad. The onset of puberty is associated with an augmentation of pulsatile activity in a majority of both girls and boys. In pubertal children, the magnitude of the nocturnal pulses of LH and FSH is consistently increased during sleep. As the pubescent child enters adulthood, the daytime pulse amplitude increases as well, eliminating or diminishing the diurnal rhythm. In pubertal girls, there is a diurnal variation of circulating estradiol levels, with higher concentrations during the daytime than during the nighttime. The lack of parallelism between gonadotropin and estradiol levels reflects a 6- to 8-hour delay between gonadotropin stimulation and the subsequent ovarian response. In pubertal boys, the nocturnal rise of testosterone coincides with the elevation of gonadotropins. Patterns of LH release in adult men exhibit large interindi-vidual variability. The diurnal variation is dampened and may become undetectable. A marked diurnal rhythm in circulating testosterone levels in young adults, with minimal levels in the late evening and maximal levels in the early morning, has been well demonstrated. In young male adults, the amplitude of the circadian variation averages 25% of the 24-hour mean.49 In older men, the amplitude of LH pulses is decreased, and no significant diurnal pattern can be detected. However, the circadian rhythm in testosterone remains apparent, although markedly dampened. In adult women, the 24-hour variation in plasma LH is modulated by the menstrual cycle. In the early follicular phase, LH pulses are large and infrequent, and a slowing of the frequency of secretory pulses occurs during sleep. In the midfollicular phase, pulse amplitude is decreased, pulse frequency is increased, and frequency modulation of LH pulsatility by sleep is less apparent. Pulse amplitude increases again by the late follicular phase. No modulation by sleep is apparent until the early luteal phase, when nocturnal slowing of pulsatility is again evident. During the luteal-follicular transition, there is a four- to five-fold increase in LH pulse frequency, which accompanies the selective FSH rise necessary for normal folliculogenesis. Toward menopause, gonadotropin levels are elevated but show no consistent circadian pattern. Abnormal ultradian and/or circadian hormonal profiles have been found in a wide variety of reproductive disorders. The findings pertaining to disorders of female and male reproduction for which an abnormal function of the hypothalamic pulse generator and/or its modulation by diurnal rhythmicity seem to be primarily involved have been reviewed elsewhere.50,59,60

RHYTHMS IN THE THYROTROPIC AXIS

In normal adult men and women, TSH levels are low throughout the daytime and begin to increase in the late afternoon or early evening.61 Maximal levels occur shortly before sleep. During sleep, TSH levels generally decline slowly. A further decrease occurs in the morning hours (see Fig. 6-2 and Fig. 6-6). Studies involving sleep deprivation and shifts of the sleep-wake cycle have consistently indicated that an inhibitory influence is exerted on TSH secretion during sleep. Interestingly, when sleep occurs during daytime hours, TSH secretion is not suppressed significantly below normal daytime levels. When the depth of sleep is increased by prior sleep deprivation, the nocturnal TSH rise is even further blunted. There is a consistent association between descending slopes of TSH concentrations and SW stages.62 The pronounced enhancing effect of sleep deprivation on the nighttime TSH rise is illustrated in Figure 6-2. The timing of the evening rise seems to be controlled by circadian rhythmicity. The temporal pattern of TSH secretion seems to reflect both tonic and pulsatile release, with both the frequency and the amplitude of the pulses increasing during the nighttime. Pulses of TSH secretion persist during somatostatin or dopamine treatment, suggesting that the control of pulsatility is largely thyrotropin-releasing hormone dependent.61 Because triiodothyronine and thyroxine are largely bound to serum protein, the existence of a circadian rhythm independent of postural changes has been difficult to establish. However, under conditions of sleep deprivation, the increased amplitude of the TSH rhythm results in an increased amplitude of the tri-iodothyronine rhythm, which becomes detectable in a majority of subjects.

The fact that the inhibitory effects of sleep on TSH secretion are time dependent may cause, under certain circumstances, elevations of plasma TSH levels, which reflect the misalignment of sleep and circadian timing. Figure 6-10 shows the mean profiles of plasma TSH levels observed in a group of normal young men in the course of adaptation to simulated jet lag. 63 After a 24-hour baseline period, the sleep-wake cycle and the dark period were abruptly advanced by 8 hours. In the course of adaptation, TSH levels increased progressively, because daytime sleep failed to inhibit TSH, and nighttime wakefulness was associated with large circadian-dependent TSH elevations. As a result, mean TSH levels after awakening from the second shifted sleep period were more than two-fold higher than during the same time interval after normal nocturnal sleep. This study demonstrates that the subjective discomfort and fatigue associated with jet lag may involve a prolonged elevation of a hormonal concentration in the peripheral circulation.

FIGURE 6-10. Mean (and standard error of the mean) profiles of plasma thyrotropin (thyroid-stimulating hormone [TSH]) from eight normal young men who were subjected to an 8-hour advance of the sleep-wake and dark-light cycles. Black bars indicate bedtime periods. (Data from Hirschfeld U, Moreno-Reyes R, Akseki E, et al. Progressive elevation of plasma thyrotropin during adaptation to simulated jet lag: effects of treatment with bright light or zolpidem. J Clin Endocrinol Metab 1996; 81:3270.)

In older adults, there is an overall decrease of TSH levels (see Fig. 6-6), but the rhythm persists, albeit slightly dampened and with an earlier evening rise than in younger adults.64 Fasting decreases overall TSH levels by decreasing pulse amplitude, resulting in a dampening of the nocturnal surge.65 A decreased or absent nocturnal rise of TSH has been observed in a wide variety of nonthyroidal illnesses, suggesting that hypothalamic dysregulation generally affects the circadian TSH surge. The nocturnal TSH surge is diminished or absent in hyperthyroidism, central hypothyroidism, and in various conditions of hypercortisolism. The lack of normal nocturnal elevation of TSH levels also appears to be a sensitive index of preclinical hyperthyroidism. In poorly controlled diabetic states, whether type 1 or type 2, the surge also disappears.66 Correction of hyperglycemia is associated with a reappearance of the nocturnal elevation.66

RHYTHMS IN GLUCOSE REGULATION

In normal humans, glucose tolerance varies with the time of day.67 Figure 6-11 illustrates circadian variations in glucose tolerance to oral glucose, identical meals, constant glucose infusion, and enteral nutrition in normal subjects. Plasma glucose responses to oral glucose, intravenous glucose, or meals are markedly higher in the evening than in the morning. Overnight studies of subjects sleeping in the laboratory have consistently observed that despite the prolonged fasting condition, glucose levels remain stable or fall only minimally across the night, contrasting with the clear decrease that is associated with daytime fasting. Thus, a number of mechanisms operative during nocturnal sleep must intervene to maintain stable glucose levels during the overnight fast. Experimental protocols using intravenous glucose infusion or enteral nutrition (the two experimental conditions allowing for the study of nighttime glucose tolerance during sleep without awakening the subjects) have shown that glucose tolerance deteriorates further as the evening progresses, reaches a minimum around midsleep, and then improves to return to morning levels (see Fig. 6-11).67 There is evidence to indicate that this diurnal variation in glucose tolerance is partly driven by the wide and highly reproducible 24-hour rhythm of circulating levels of cortisol, an important counterregulatory hormone.68 Diminished insulin sensitivity and decreased insulin secretion in relation to elevated glucose levels are both involved in causing reduced glucose tolerance later in the day. During the first part of the night, decreased glucose tolerance is due to decreased glucose utilization both by peripheral tissues (relaxed muscles and rapid insulin-like effects of sleep-onset GH secretion) and by the brain (imaging studies have demonstrated a reduction in glucose uptake during SW sleep). During the second part of the night, these effects subside, as sleep becomes more shallow and fragmented. Thus, complex interactions of circadian and sleep effects result in a consistent and predictable pattern of changes of the setpoint of glucose regulation over the 24-hour cycle.

FIGURE 6-11. Twenty-four-hour pattern of blood glucose changes in response to oral glucose (top panel; 50 g glucose every 3 hours), identical meals, constant glucose, and continuous enteral nutrition in normal young adults. At each time point, the mean glucose level is shown with the standard error of the mean. (Reproduced from Van Cauter E, Polon-sky KS, Scheen AJ. Roles of circadian rhythmicity and sleep in human glucose regulation. Endocr Rev 1997; 18:716.)

Human insulin secretion is a complex oscillatory process including rapid pulses recurring every 10 to 15 minutes superimposed on slower, ultradian oscillations with periods in the 90- to 120-minute range.26 The ultradian oscillations are tightly coupled to glucose, and the periodicity of the insulin secretory oscillations can be entrained to the period of an oscillatory glucose infusion, supporting the concept that these ultradian oscillations are generated by the glucose-insulin feedback mechanism.69 Stimulatory effects of sleep on insulin secretion are mediated by an increase in the amplitude of the oscillation.70 The rapid 10- to 15-minute pulsations seem to have a different origin from that of the ultradian oscillations.26 Indeed, they may appear independently of glucose, because they have been observed in the isolated perfused pancreas and in isolated islets. Thus, the existence of an intrapancreatic pacemaker generating rapid oscillations has been postulated. In obese and diabetic subjects, the diurnal and ultradian variations in glucose regulation are abnormal. In obesity, the morning versus evening difference in glucose tolerance observed in normal subjects is abolished. In type 1 diabetic patients, the increase in glucose levels and/or insulin requirements, which occurs in a prebreakfast period ranging from 5:00 a.m. to 9:00 a.m., has been called the dawn phenomenon.71 A role for nocturnal GH secretion in the pathogenesis of the dawn phenomenon has been demonstrated in some, but not all, studies. The observation of a dawn phenomenon in type 2 diabetes patients under normal dietary conditions has been less consistent. Prominent diurnal variations in glucose levels and insulin secretion in both normal subjects and diabetic patients become apparent during prolonged fasting.72 Figure 6-12 illustrates these variations in diabetic patients and age-, sex-, and weight-matched controls studied during a 24-hour fast following an overnight fast.72 As expected, glucose levels initially declined as a result of the fasting condition, but started rising again in the late evening to reach a morning maximum. The nocturnal rise of glucose during prolonged fasting could represent a normal diurnal variation in the set-point of glucose regulation amplified by counterregulatory mechanisms activated by the fasting condition.

FIGURE 6-12. Individual 24-hour glucose profiles from four subjects with type 2 diabetes who remained fasted throughout the study period. The two upper profiles were obtained from 8 a.m. the first day until 8 a.m. the next day. The two lower profiles were obtained from 2 p.m. the first day until 2 p.m. the next day. Irrespective of the timing of the beginning of the fasting period, glucose concentrations started increasing in the evening and peaked in the morning. The dashed lines represent the best-fit curve. (From Shapiro ET, Polonsky KS, Copinschi G, et al. Nocturnal elevation of glucose levels during fasting in noninsulin-dependent diabetes. J Clin Endocrinol Metab 1991; 72:444.)

The rapid and ultradian oscillations of insulin secretion are perturbed in type 2 diabetes and in impaired glucose tolerance without hyperglycemia. The rapid pulses appear to be less regular and of shorter duration than in normal subjects.73,74 A less regular oscillatory pattern may already be detected in relatives of patients with type 2 diabetes. The ultradian oscillations, which have an exaggerated amplitude in obese subjects without apparent changes in frequency or pattern of recurrence, are irregular and of lower amplitude in subjects with established type 2 diabetes.75,76 and 77 Disturbances in the pattern of entrainment of insulin secretion to oscillatory glucose infusions are evident in type 2 diabetes patients, in nondiabetic subjects with impaired glucose tolerance,77 and in nondiabetic first-degree relatives of subjects with type 2 diabetes.26 CHAPTER REFERENCES
1. 2. 3. 4. 5. 6. Turek FW. Circadian rhythm. Horm Res 1998; 49:109. Borbely AA. Processes underlying sleep regulation. Horm Res 1998; 49:114. Vitaterna MH, King DP, Chang AM, et al. Mutagenesis and mapping of a mouse gene, Clock, essential for circadian behavior. Science 1994; 264:719. King DP, Zhao Y, Sangoram AM, et al. Positional cloning of the mouse circadian Clock gene. Cell 1997; 89:641. Dunlap JC. Molecular bases for circadian clocks. Cell 1999; 96:271. Porkka-Heiskanen T, Strecker RE, Thakkar M, et al. Adenosine: a mediator of the sleep-inducing effects of prolonged wakefulness. Science 1997; 276:1255.

6a. Spiegel K, Leproult R, Van Cauter E. Impact of sleep debt on metabolic and endocrine function. Lancet 1999; 9:1435. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. Reinberg AE. Concepts in chronopharmacology. Annu Rev Pharmacol Toxicol 1992; 32:51. Czeisler CA, Johnson MP, Duffy JF, et al. Exposure to bright light and darkness to treat physiologic maladaptation to night work. N Engl J Med 1990; 322:1253. Tepas DI, Carvalhais AB. Sleep patterns of shiftworkers. Occup Med 1990; 5:199. Van Reeth O. Sleep and circadian disturbances in shift work: strategies for their management. Horm Res 1998; 49:158. Roden M, Koller M, Pirich K, et al. The circadian melatonin and cortisol secretion pattern in permanent night shift workers. Am J Physiol 1993; 265:R261. Folkard S, Minors DS, Waterhouse JM. Chronobiology and shift work: current issues and trends. Chronobiologia 1985; 12:31. Weibel L, Follenius M, Spiegel K, et al. Comparative effect of night and daytime sleep on the 24-hour cortisol secretory profile. Sleep 1995; 18:549. Weibel L, Spiegel K, Follenius M, et al. Internal dissociation of the circadian markers of the cortisol rhythm in night workers. Am J Physiol 1996; 270:E608. Spiegel K, Weibel L, Gronfier C, et al. Twenty-four hour prolactin profiles in night workers. Chronobiol Int 1996; 13:283. Wehr TA, Goodwin FK. Biological rhythms in manic-depressive illness. In: Wehr TA, Goodwin FK, eds. Circadian rhythms in psychiatry. Pacific Grove, CA: Boxwood Press, 1983:129. Van Cauter E, Spiegel K. Circadian and sleep control of hormonal secretions. In: Turek FW, Zee PC, eds. Regulation of sleep and circadian rhythms. New York: Marcel Dekker, Inc., 1999:397.

17a. Saini J, Krieger J, Brandenberger G, et al. Continuous positive airway pressure treatment: effects on growth hormone, insulin and glucose profiles in obstructive sleep apnea patients. Horm Metab Res 1993; 25:375. 18. Grunstein RR, Handelsman DJ, Lawrence SJ, et al. Neuroendocrine dysfunction in sleep apnea: reversal by continuous positive airways pressure therapy. J Clin Endocrinol Metab 1989; 68:352. 19. Goldstein SJ, Wu RHK, Thorpy MJ, et al. Reversibility of deficient sleep entrained growth hormone secretion in a boy with achondroplasia and obstructive sleep apnea. Acta Endocrinol 1987; 116:95. 20. Czeisler CA, Chiasera AJ, Duffy JF. Research on sleep, circadian rhythms and aging: applications to manned spaceflight. Exp Gerontol 1991; 26:217. 21. van Coevorden A, Mockel J, Laurent E, et al. Neuroendocrine rhythms and sleep in aging. Am J Physiol 1991; 260:E651. 22. Chapotot F, Gronfier C, Jouny C, et al. Cortisol secretion is related to electroencephalographic alertness in human subjects during daytime wakefulness. J Clin Endocrinol Metab 1998; 83:4263. 23. Hotchkiss J, Knobil E. The menstrual cycle and its neuroendocrine control. In: Knobil E, Neil JD, eds. The physiology of reproduction, 2nd ed. New York: Raven Press, 1994:711. 24. Wetsel WC, Valenca MM, Merchenthaler I, et al. Intrinsic pulsatile secretory activity of immortalized luteinizing hormone-releasing hormone-secreting neurons. Proc Natl Acad Sci U S A 1992; 89:4149. 25. Simon C, Brandenberger G, Follenius M. Ultradian oscillations of plasma glucose, insulin, and C-peptide in man during continuous enteral nutrition. J Clin Endocrinol Metab 1987; 64:669. 26. Polonsky KS, Sturis J, Van Cauter E. Temporal profiles and clinical significance of pulsatile insulin secretion. Horm Res 1998; 49:178. 27. Conn PM, Crowley WF. Gonadotropin-releasing hormone and its analogues. N Engl J Med 1991; 324:93. 28. Van Cauter E. Computer-assisted analysis of endocrine rhythms. In: Rod-bard D, Forti G, eds. Computers in endocrinology. New York: Raven Press, 1990:59. 29. Pincus SM, Keefe DL. Quantification of hormone pulsatility via an approximate entropy algorithm. Am J Physiol 1992; 262:E741. 30. Pincus SM, Goldberger AL. Physiological time-series analysis: what does regularity quantify? Am J Physiol 1994; 266:H1643. 31. Van Cauter E, Plat L, Copinschi G. Interrelations between sleep and the somatotropic axis. Sleep 1998; 21:553. 32. Pincus SM, Gevers EF, Robinson IC, et al. Females secrete growth hormone with more process irregularity than males in both humans and rats. Am J Physiol Physiol 1996; 270:E107. 33. Ho KY, Evans WS, Blizzard RM, et al. Effects of sex and age on the 24-hour profile of growth hormone secretion in man: importance of endogenous estradiol concentrations. J Clin Endocrinol Metab 1987; 64:51. 34. Jaffe C, Turgeon D, DeMott Friberg R, et al. Nocturnal augmentation of growth hormone (GH) secretion is preserved during repetitive bolus administration of GH-releasing hormone: potential involvement of endogenous somatostatin: a clinical research center study. J Clin Endocrinol Metab 1995; 80:3321. 35. Ocampo-Lim B, Guo W, DeMott Friberg R, et al. Nocturnal growth hormone (GH) secretion is eliminated by infusion of GH-releasing hormone antagonist. J Clin Endocrinol Metab 1996; 81:4396. 36. Gronfier C, Luthringer R, Follenius M, et al. A quantitative evaluation of the relationships between growth hormone secretion and delta wave electroencephalographic activity during normal sleep and after enrichment in delta waves. Sleep 1996; 19:817. 37. Van Cauter E, Plat L, Scharf M, et al. Simultaneous stimulation of slow-wave sleep and growth hormone secretion by g-hydroxybutyrate in normal young men. J Clin Invest 1997; 100:745. 38. Hartman ML, Veldhuis JD, Johnson ML, et al. Augmented growth hormone (GH) secretory burst frequency and amplitude mediate enhanced GH secretion during a two-day fast in normal men. J Clin Endocrinol Metab 1992; 74:757. 39. Edge JA, Dunger DB, Matthews DR, et al. Increased overnight growth hormone concentrations in diabetic compared with normal adolescents. J Clin Endocrinol Metab 1990; 71:1356. 40. Gelato MC, Oldfield E, Loriaux DL, Merriam GR. Pulsatile growth hormone secretion in patients with acromegaly and normal men: the effects of growth hormone-releasing hormone infusion. J Clin Endocrinol Metab 1990; 71:585. 41. Hartman ML, Veldhuis JD, Vance ML, et al. Somatotropin pulse frequency and basal concentrations are increased in acromegaly and are reduced by successful therapy. J Clin Endocrinol Metab 1990; 70:1375. 42. Weitzman ED, Zimmerman JC, Czeisler CA, Ronda JM. Cortisol secretion is inhibited during sleep in normal man. J Clin Endocrinol Metab 1983; 56:352. 43. Born J, Muth S, Fehm HL. The significance of sleep onset and slow wave sleep for nocturnal release of growth hormone (GH) and cortisol. Psycho-neuroendocrinology 1988; 13:233. 44. Van Cauter E, Blackman JD, Roland D, et al. Modulation of glucose regulation and insulin secretion by circadian rhythmicity and sleep. J Clin Invest 1991; 88:934. 45. Follenius M, Brandenberger G, Bardasept J, et al. Nocturnal cortisol release in relation to sleep structure. Sleep 1992; 15:21. 46. Spath-Schwalbe E, Gofferje M, Kern W, et al. Sleep disruption alters nocturnal ACTH and cortisol secretory patterns. Biol Psychiatry 1991; 29:575. 47. Pruessner JC, Wolf OT, Hellhammer DH, et al. Free cortisol levels after awakening: a reliable biological marker for the assessment of adrenocortical activity. Life Sci 1997; 61:2539. 48. Leproult R, Copinschi G, Buxton O, Van Cauter E. Sleep loss results in an elevation of cortisol levels the next evening. Sleep 1997; 20:865. 49. Lejeune-Lenain C, Van Cauter E, Desir D, et al. Control of circadian and episodic variations of adrenal androgens secretion in man. J Endocrinol Invest 1987; 10:267. 50. Van Cauter E, Turek FW. Endocrine and other biological rhythms. In: DeGroot LJ, ed. Endocrinology. Philadelphia: WB Saunders, 1995:2487. 51. Yehuda R, Teicher MH, Trestman RL, et al. Cortisol regulation in posttraumatic stress disorder and major depression: a chronobiological analysis. Biol Psychiatry 1996; 40:79. 52. MacHale SM, Cavanaugh JTO, Bennie J, et al. Diurnal variation of adreno-cortical activity in chronic fatigue syndrome. Neuropsychobiology 1998; 38:213. 53. Spiegel K, Follenius M, Simon C, et al. Prolactin secretion and sleep. Sleep 1994; 17:20. 54. Waldstreicher J, Duffy JF, Brown EN, et al. Gender differences in the temporal organization of prolactin (PRL) secretion: evidence for a sleep-independent circadian rhythm of circulating PRL levels: a clinical research center study. J Clin Endocrinol Metab 1996; 81:1483. 55. Spiegel K, Luthringer R, Follenius M, et al. Temporal relationship between prolactin secretion and slow-wave electroencephalographic activity during sleep. Sleep 1995; 18:543. 56. Copinschi G, Van Onderbergen A, L'Hermite-Balriaux M, et al. Effects of the short-acting benzodiazepine triazolam, taken at bedtime, on circadian and sleep-related hormonal profiles in normal men. Sleep 1990; 13:232. 57. Iranmanesh A, Veldhuis JD, Carlsen EC, et al. Attenuated pulsatile release of prolactin in men with insulin-dependent diabetes mellitus. J Clin Endo-crinol Metab 1990; 71:73. 58. Seki K, Uesato T, Kato K, Shima K. Twenty-four hour secretory pattern of prolactin in hyperprolactinaemic patients with pituitary micro- and macroadenomas. Acta Endocrinol 1984; 106:433. 59. Hayes FJ, Crowley WFJ. Gonadotropin pulsations across development. Horm Res 1998; 49:163. 60. Filicori M, Tabarelli C, Casadio P, et al. Interaction between menstrual cyclicity and gonadotropin pulsatility. Horm Res 1998; 49:169. 61. Behrends J, Prank K, Dogu E, Brabant G. Central nervous system control of thyrotropin secretion during sleep and wakefulness. Horm Res 1998; 49:173. 62. Goichot B, Brandenberger G, Saini J, et al. Nocturnal plasma thyrotropin variations are related to slow-wave sleep. J Sleep Res 1992; 1:186.

63. Hirschfeld U, Moreno-Reyes R, Akseki E, et al. Progressive elevation of plasma thyrotropin during adaptation to simulated jet lag: effects of treatment with bright light or zolpidem. J Clin Endocrinol Metab 1996; 81:3270. 64. van Coevorden A, Laurent E, Decoster C, et al. Decreased basal and stimulated thyrotropin secretion in healthy elderly men. J Clin Endocrinol Metab 1989; 69:177. 65. Romijn JA, Adriaanse R, Brabant G, et al. Pulsatile secretion of thyrotropin during fasting: a decrease of thyrotropin pulse amplitude. J Clin Endo-crinol Metab 1990; 70:1631. 66. Bartalena L, Cossu E, Grasso L, et al. Relationship between nocturnal serum thyrotropin peak and metabolic control in diabetic patients. J Clin Endocrinol Metab 1993; 76:983. 67. Van Cauter E, Polonsky KS, Scheen AJ. Roles of circadian rhythmicity and sleep in human glucose regulation. Endocr Rev 1997; 18:716. 68. Plat L, Byrne MM, Sturis J, et al. Effects of morning cortisol elevation on insulin secretion and glucose regulation in humans. Am J Physiol 1996; 270:E36. 69. Sturis J, Polonsky KS, Mosekilde E, Van Cauter E. Computer model for mechanisms underlying ultradian oscillations of insulin and glucose. Am J Physiol 1991; 260:E801. 70. Simon C. Ultradian pulsatility of plasma glucose and insulin secretion rates: circadian and sleep modulation. Horm Res 1998; 49:185. 71. Bolli GB, Gerich JE. The dawn phenomenon: a common occurrence in both non-insulin-dependent and insulin-dependent diabetes mellitus. N Engl J Med 1984; 310:746. 72. Shapiro ET, Polonsky KS, Copinschi G, et al. Nocturnal elevation of glucose levels during fasting in noninsulin-dependent diabetes. J Clin Endo-crinol Metab 1991; 72:444. 73. Lang DA, Matthews DR, Burnett M, Turner RC. Brief, irregular oscillations of basal plasma insulin and glucose concentrations in diabetic man. Diabetes 1981; 30:435. 74. O'Rahilly S, Turner R, Matthews D. Impaired pulsatile secretion of insulin in relatives of patients with non-insulin-dependent diabetes. N Engl J Med 1988; 318:1225. 75. Polonsky KS, Given BD, Hirsch LJ, et al. Abnormal patterns of insulin secretion in non-insulin-dependent diabetes mellitus. N Engl J Med 1988; 318:1231. 76. Simon C, Brandenberger G, Follenius M, Schlienger JL. Alteration in the temporal organisation of insulin secretion in Type 2 (non-insulin-dependent) diabetic patients under continuous enteral nutrition. Diabetologia 1991; 34:435. 77. O'Meara NM, Sturis J, Van Cauter E, Polonsky KS. Lack of control by glucose of ultradian insulin secretory oscillations in impaired glucose tolerance and in non-insulin-dependent diabetes mellitus. J Clin Invest 1993; 92:262.

CHAPTER 7 GROWTH AND DEVELOPMENT IN THE NORMAL INFANT AND CHILD Principles and Practice of Endocrinology and Metabolism

CHAPTER 7 GROWTH AND DEVELOPMENT IN THE NORMAL INFANT AND CHILD

GILBERT P. AUGUST Height and Weight Growth Velocity Skinfold Thickness Body Mass Index Bone Age Height Prediction Body Proportions Genital Development Developmental Endocrinology Growth Hormone and Insulin-Like Growth Factors Adrenocorticotropic Hormone and Adrenocortical Steroids Thyroid-Stimulating Hormone and Thyroid Hormones Luteinizing Hormone, Follicle-Stimulating Hormone, and the Sex Steroids Renin and Aldosterone Chapter References

The pediatric population is composed of continually changing individuals. Thus, a knowledge of normal developmental changes is required for the clinician to recognize deviant growth and development and abnormalities in the hormonal milieu.

HEIGHT AND WEIGHT

Height and weight standards derived from one ethnic group cannot always be applied to other ethnic groups. Furthermore, use of current standards is important, because growth data obtained from a previous generation may not apply to the present generation. The National Health Survey collected growth and anthropometric data on American children from 1963 to 1974 (Table 7-1).1,2,3,4,5,6,7,8 and 9 These data provide standards that can be correlated with sex, race, and socioeconomic status. The growth charts distributed by pharmaceutical companies are derived from these National Health Survey data. Because the growth standards of American white and black children do not differ significantly, a single growth standard can be used. Significant differences do exist, however, in the growth standards of American children of Asian descent.7

TABLE 7-1. Height and Weight of Children, Body Proportions, and Error of Prediction of Adult Height (United States)

Figure 7-1, Figure 7-2, Figure 7-3 and Figure 7-4 show the current growth data from the National Health Survey. In Table 7-1, the mean heights, mean weights, and standard deviations (SDs) for children 2 to 18 years of age are detailed. The standard deviation is useful in evaluating extreme deviations in growth (e.g., heights and weights below the standard curves).10 Children whose heights or weights are below the standard curves constitute a significant proportion of the population, because the commonly used growth charts provide data from only the 5th through the 95th percentiles.

FIGURE 7-1. Length and weight of boys. Birth to 36 months of age. Centers for Disease Control and Prevention, National Center for Health Statistics. CDC growth charts: United States. Full size charts are available on the internet at http://www.cdc.gov/nchs/about/major/nhanes/growthcharts/charts.htm May 30, 2000.

FIGURE 7-2. Height and weight of boys. 2 to 20 years of age. Centers for Disease Control and Prevention, National Center for Health Statistics. CDC growth charts: United States. Full size charts are available on the internet at http://www.cdc.gov/nchs/about/major/nhanes/growthcharts/charts.htm May 30, 2000.

FIGURE 7-3. Length and weight of girls. Birth to 36 months of age. Centers for Disease Control and Prevention, National Center for Health Statistics. CDC growth charts: United States. Full size charts are available on the internet at http://www.cdc.gov/nchs/about/major/nhanes/growthcharts/charts.htm May30, 2000.

FIGURE 7-4. Height and weight of girls. 2 to 20 years of age. Centers for Disease Control and Prevention, National Center for Health Statistics. CDC growth charts: United States. Full size charts are available on the internet at http://www.cdc.gov/nchs/about/major/nhanes/growthcharts/charts.htm May 30, 2000.

For children aged 2 to 18 years, growth curves for as low as 2 SD below the mean are available. These curves are more in keeping with the usual definition of normal as comprising 95% of the population. Only 2.5% of children would be considered as short by these standards (Fig. 7-1, Fig. 7-2, Fig. 7-3 and Fig. 7-4). The National Center for Health Statistics has revised the current growth charts to reflect a more contemporaneous standard. The new charts contain data from the 3rd to 97th percentiles. The new growth charts are available on the internet at http://www.cdc.gov/growthcharts. Socioeconomic status was found to play a small but significant adverse role in the growth of children with very low family incomes.6 Beyond age 2 to 3 years, children grow throughout childhood until puberty along a particular height percentile channel that is determined by genetic factors. The normal pubertal growth spurt is reflected by an upward shift in height percentile. As growth decelerates later in puberty with impending epiphyseal closure, a shift downward in height percentile occurs, so that adult height more closely approximates the prepubertal percentile ranking. The reason for these shifts lies in the cross-sectional character of the standard growth charts. Cross-sectional data mask normal longitudinal growth patterns. The growth charts derived from studies in Britain have superim-posed longitudinal growth lines that represent temporal variations in the onset of puberty.11 Similar data are available for North American children.12 A shift in height percentile can also occur during infancy when growth velocities are adjusted in children whose birth lengths are either too long or too short in comparison with their parents' heights. The shift upward in percentile generally takes place within the first 6 months of life and is usually completed by 12 months of age. The shift downward in percentile begins later, generally after 3 months of age, and is completed by 18 months of age.13 The influence of midparental stature on the interpretation of a child's growth status is considered to be of such importance by some workers that conversion graphs and adjustment formulas have been developed. Tanner and coworkers14 developed such curves for British children, and adjustment tables are now available for American children.15 Parental heights can also be used to predict target adult height.16 This can be valuable when the effects of a treatment are analyzed. The standard growth charts derived from the National Health Survey data are not as applicable for tracking the growth of premature infants or children small for gestational age. However, growth standards are available for such children through 4 years of age.17,18 These growth charts must be adjusted for the degree of low birth weight.19 Further, infants small for gestational age may not truly catch up when their heights are compared with those of their siblings.20,21 Whenever sufficient data exist to construct growth charts for children with constitutional diseases, those standards should be used for comparison of current growth status and for expected adult height. Such growth standards are available for children with achondroplasia, Down syndrome, diastrophic dysplasia, Duchenne muscular dystrophy, Klinefelter syndrome, Noonan syndrome, pseudoachondroplasia, Russell-Silver syndrome, spondyloepiphyseal dysplasia, and William syndrome.22 Similar information is available for children with gonadal dysgenesis.23,24 Careful attention to detail is important to produce accurate height measurements that can be used to track a child's growth (Fig. 7-5). Length should be measured on a horizontal board with the head held firmly against the upright at one end and the length determined by sliding the movable upright against the child's heels. Length should be determined until 3 years of age. The growth curve from birth until 3 years of age is standardized on length. Height must be measured using a wall-mounted instrument or tape. This permits the child to stand fully upright. The feet should be together, the back and heels should be pressed against the wall or vertical part of the measuring instrument, and the head should be held in the Frankfort horizontal plane with upward pressure exerted under the child's mastoid process to hold the correct position and to measure maximal height. The measurement of height is taken by sliding a right-angle block downward until it touches the child's head; the height can then be read from the vertically mounted rule. Accurate, reproducible height measurements cannot be obtained from the commonly used height-measuring devices attached to weighing scales. The importance of adhering to the correct technique is documented in the National Health Survey data.1,2,3,4,5 and 6

FIGURE 7-5. Technique of accurate measurement of standing height.

GROWTH VELOCITY The standard growth charts with their height percentiles are growth attainment charts, which represent how much height the child has attained by a particular age. A concept that is often more useful, especially when evaluating the longitudinal growth of an individual child, is growth velocity, which depicts growth during a given

period. The difference in height at the beginning and end of a given period of time is annualized and then expressed as centimeters per year.11,12 The X-axis on a growth velocity chart is the chronologic age; the Y-axis is the growth velocity in centimeters per year. These charts take into consideration the variability in height velocity caused by the normal variability in age of onset of puberty. An advantage of height velocity charts is their ability to demonstrate more quickly an aberration of growth or the effects of treatment. Short-term growth velocity data, however, may be subject to error because of marked seasonal variation; children generally grow faster in the spring through early summer.24 The growth velocity charts clearly demonstrate that a child's growth rate is most rapid during infancy and puberty and is relatively stable during the elementary school years. Although these charts present height velocity percentiles ranging from the 3rd to 97th percentiles, these may not represent the biologic normal range. Longitudinal growth that is consistently below the 10th percentile may not be sufficient to maintain a child in a single height attainment percentile; instead, that child may demonstrate a downward shift in height percentile.25 SKINFOLD THICKNESS Skinfold thickness is used as a measure of total body fat.26 In research studies, more than one site needs to be measured, but for clinical use, the more readily accessible triceps skin-fold is commonly used. The National Health Survey used a carefully standardized technique to measure the triceps skin-fold. The Lange caliper (Fig. 7-6) was used on a skinfold parallel to the long axis of the arm over the triceps muscle halfway between the elbow and acromial process of the scapula; care was taken to apply the caliper so that the pressure plates remained parallel to each other. Table 7-2 presents the normal values.27,28 and 29

FIGURE 7-6. Measurement of skinfold thickness with the Lange caliper.

TABLE 7-2. Percentiles for Triceps Skinfold Thickness (in Millimeters) by Year of Age27,28

BODY MASS INDEX Body mass index (BMI) is another convenient measure of body fat. It is derived by calculating the weight in kilograms divided by the the height in meters squared. Population ranges for American children are now available according to ethnic group for ages 5 to 17 years.30 The 15th, 50th, and 85th percentiles for each ethnic group appear in Table 7-3. The BMI will assume greater applicability once the new BMI curves are available from the National Center for Health Statistics. For adults, grade 1 obesity is defined as a BMI of over 25 and grade 2 obesity as a BMI over 30. Similar cutoff levels have been suggested for late adolescence. These BMI values are close to the 80th and 95th percentiles, respectively, on the National Center for Health Statistics charts.31

TABLE 7-3. Body Mass Index Centiles for Children 517 Years of Age

BONE AGE The concept of bone age is based on the skeletal changes that occur with the physical growth and maturation of the child. These skeletal changes include the calcification, growth, and shaping of the epiphyseal centers of the bones and their eventual fusion with the diaphyses. The fact that these changes occur in a regular sequence in the different bones as determined by the radiographic examination of normally developing children permits one to estimate a bone age for a particular child. Any of a number of techniques can be used, as well as any part of the skeleton. The most popular method in the United States compares a single anteroposterior radiograph of the left hand and wrist with the series of standard films in the atlas compiled by Greulich and Pyle.32 In Europe, the Tanner method is used to calculate the bone age of the left hand and wrist by using maturity indicators for each bone, so that a composite score is derived.33 The Tanner-Whitehouse method has been standardized for American children aged 8 to 16 years.34 As with any other laboratory test, there is a mean population bone age that corresponds to the particular best fit in the Greulich and Pyle atlas and a standard deviation or range of expected values. The normal range of expected bone ages for a child younger than 1 year of age is 3 to 6 months; for a child of 6 to 11 years of age, the range is 2 years; and for a child of 12 to 14 years of age, the range is 2 years.35 The National Health Survey compared the bone ages of contemporary children with those derived from the original study of Greulich and Pyle between 1917 and 1942. The original study population was composed of upper-middle-class children in Cleveland. In children aged 6 to 11 years, good congruence was seen except that

contemporary 10- and 11-year-old children showed a significant retardation in bone age of 0.2 to 0.65 years.36 The assumption that bone age advances 1 year for each calendar year may not be correct; bone age advances more rapidly than chronologic age during the onset of puberty and during peak height velocity of puberty. At these times, the average advancement in bone age over chronologic age is 1.73 years, with a range of 0.8 to 2.8 years.37 Bone age advancement should not be attributed falsely to therapeutic intervention. These results serve as another example of the difference between normal longitudinal growth of a single child and the cross-sectional standards that are usually available for tracking a child's growth and development.37 The determination of bone age can be useful if one remains aware of its limitations. The particular standard used must be verified for the population being studied. The person interpreting the bone age radiograph must be consistent. Separate male and female standards exist, and there is a range of normal (see Chap. 18 and Chap. 217). HEIGHT PREDICTION One of the most common applications of a bone age determination is the prediction of eventual adult height. Bayley and Pinneau38 published prediction tables based on the percentage of adult height attained at various bone ages. The tables were further refined by separating the children whose bone ages were>1 year advanced or retarded compared with chronologic age. Separate prediction tables are provided for these children as well as for males and females. More mathematically refined systems have been proposed that use regression formulas with the addition of variables such as weight, midparental stature, growth velocity, and menarcheal status. The Roche-Wainer-Thissen (RWT) method uses the child's recumbent length rather than standing height and is standardized for American children.39 The Tanner-Whitehouse system (TW2 Mark 2) is standardized for British children.33 It includes a larger number of children at the extremes of the standardizing group and thus tries to counter one criticism of height prediction methodsthat they are most accurate for the prediction of the final adult height of normal children whose bone ages are not significantly retarded. In addition, the TW2 Mark 2 system no longer takes into account midparental stature in the prediction regression formulas, as did the previous formulas. Regardless of which system is used for the prediction of adult height, the prediction always has an error. For the RWT method, the 90% confidence range for a prediction is given in Table 7-1. For the TW2 Mark 2 system, the error is stated as a residual standard deviation. For comparison purposes, in Table 7-1, the residual standard deviation has been multiplied by 1.645 to provide the 90% confidence range for a predicted adult height. To obtain a 95% confidence range, the residual standard deviation is multiplied by 1.96. For normal children, these two methods are comparable and appear more accurate than the older Bailey and Pinneau tables39; the stated error in the latter method is 5.1 cm for 2 SD. In a comparative study of the three methods that examined Finnish children, the RWT method was slightly more accurate.40,41 The important issue is not how well the various height prediction systems work in normal children but rather how well they work in children with abnormal growth patterns. In children with familial tall stature, the Tanner method was slightly more accurate than the other two methods.41 Both the RWT and Tanner methods, however, overestimated adult height in children with precocious puberty or gonadal dysgenesis; the Bailey and Pinneau tables were more accurate in these conditions. The overestimation for children with precocious puberty ranged from 30 cm at 5 years of age to just 5 cm at 13 years of age. The overestimation of adult height was 10 cm at the earlier ages in children with gonadal dysgenesis. The greater accuracy of height prediction with increasing age of the patient also holds true for children with constitutional tall stature.42 In addition, the RWT and Tanner methods both overestimated the adult heights of children with primordial short stature.41 These studies used the older Tanner methods. Tanner and coworkers33 point out several limitations to the accuracy of their new formulas that should apply to other height prediction systems. Although the formulas take into account several variables, some unpredictability still exists, as reflected in the residual standard deviation. Tanner and co-workers33 point out that no particular reason exists that the equations should be valid in cases of precocious puberty or achondroplasia. In part, the error in height prediction, measured by whatever means, may be due to the variability in total height gained during puberty. In a review43 of four studies of American children, from the onset of puberty, girls had an average gain in height of 25.4 to 31 cm, and boys had an average gain in height of 28.2 to 31 cm. The standard deviations ranged from 4 to 6.8 cm for girls and 4.3 to 5.5 cm for boys. Thus, although considerable height was gained during puberty, considerable variability was seen as well. Similar results were reported in European studies.43 Confounding this further is the inverse relationship between height obtained during puberty and the age of onset of puberty.44 This phenomenon may be related to the shortened duration of puberty, at least in girls, with the later onset of puberty.45 Finally, an additional caution arises from a study that compared the final adult height achieved by short children with normal bone ages and by short children with significantly retarded bone ages with predictions of the children's height using the Bailey and Pinneau tables.46 Although only 1 of 17 children with normal bone age had an overpredicted adult height, 5 of 10 children with a retarded bone age had an over-prediction of their adult heightthey failed to reach an adult height within 5.1 cm of the predicted height.46 This is the group of children for whom the physician is most often called on to use height prediction tables and provide reassurance to the patient and family. A comparison of predicted adult height, based on bone age and current height, with the predicted genetic target adult height based on parental heights is often useful.16 This is of importance when determining if the child's current height is inconsistent with genetic potential or when one is evaluating the effects of a therapeutic intervention. BODY PROPORTIONS Significant racial differences exist in regard to body proportions. The sitting/standing height ratio was determined during the National Health Survey.4,5 These data are presented in Table 7-1. Careful attention to detail is required. Height measurement using a wall-mounted instrument is required. Sitting height is measured on a sitting height table, with the child sitting erect in a standardized manner and with the head held in the Frankfort plane4,5 (Fig. 7-7).

FIGURE 7-7. Technique of accurate measurement of sitting height.

GENITAL DEVELOPMENT Table 7-4 presents the normal standards for penile and testicular size.47,48 and 49 Penile length is a stretched length; traction is applied to the penis and the rule is applied firmly to the root of the penis. Too often, the suprapubic fat pad is not compressed enough so that one obtains a penile length that is artifactually too short (see Chap. 93).

TABLE 7-4. Penile and Testicular Size During Childhood

DEVELOPMENTAL ENDOCRINOLOGY
The pediatric patient undergoes continuous change that is reflected not only in growth rates but also in the secretory activity of the endocrine system. The perinatal period is one of rapid change in the endocrine system when the child is adjusting to extrauterine life (Table 7-6). This is further complicated by the differences that may exist in hormonal levels in the full-term versus preterm infant. This section outlines some of these hormonal variations. As a general rule, one can divide the analysis of hormonal secretory activity into the following developmental periods: fetal, perinatal, childhood, and pubertal.50,51

TABLE 7-6. Blood Concentrations of Gonadotropins, Gonadal Steroids, Renin, and Aldosterone in Infancy and Childhood

GROWTH HORMONE AND INSULIN-LIKE GROWTH FACTORS Immunoreactive growth hormone is found in human anterior pituitary glands as early as the seventh week of gestation.49 Pituitary growth hormone increases from 0.44g per gland at 10 to 14 weeks of gestation to a mean of 675g per gland at term. However, pituitary growth hormone level is fairly constant after 25 weeks of gestation. Growth hormone is also detectable in fetal serum in high concentration. These high concentrations of growth hormone may be a reflection of hypothalamic-pituitary neuroendocrine immaturity. At 10 to 14 weeks of gestation, the mean growth hormone level in fetal serum is 65 ng/mL, and it rises rapidly to a peak fetal concentration of 132 ng/mL at 20 to 25 weeks of gestation. Thereafter, serum growth hormone declines to 35 ng/mL at 35 to 40 weeks of gestation. The postulated immature neuroendocrine control of growth hormone secretion persists into the neonatal period. Growth hormone is not suppressed by glucose until 1 month of age, and the normal sleep-induced rise in serum growth hormone levels does not appear until 3 months of age.51 Insulin-like growth factor-I (IGF-I) plasma concentrations show a marked age-related pattern. The normal ranges, as reported by the Quest Diagnostic Nichols Institute, are shown in Figure 7-8 and Figure 7-9. The range is large, and a marked overlap of values is seen when one compares the lower range of normal with values typically associated with growth hormone deficiency. Plasma IGF-I levels increase with age; during puberty concentrations are reached that would be associated with acromegaly if encountered in an adult. After puberty, the levels decline to the normal adult range. Females generally have IGF-I levels ~20% higher than those of males. The normal level of IGF-I may correlate better with the child's physiologic development. Such a relation is seen during puberty (see Fig. 7-9). For prepubertal boys, a similar relation with bone age may be valid as well.52,53 Insulin-like growth factorbinding protein-3 (IGFBP-3) has been suggested as another serum protein reflective of growth hormone secretion and function.54 In a manner similar to IGF-I, normal values for IGFBP-3 vary according to age and stage of puberty (Fig. 7-10).

FIGURE 7-8. Variation of insulin-like growth factor-I with age and sex. (Data derived from the normal values provided by Quest Diagnostic Nichols Institute, San Juan Capistrano, California.)

FIGURE 7-9. Variation of insulin-like growth factor-I with stage of puberty and sex. (Data derived from the normal values provided by Quest Diagnostic Nichols Institute, San Juan Capistrano, California.)

FIGURE 7-10. Variation of insulin-like growth factorbinding protein-3 with age. (Data derived from the normal values provided by Quest Diagnostic Nichols Institute, San Juan Capistrano, California.)

Measurement of insulin-like growth factorbinding protein-2 (IGFBP-2) has been proposed as another aid in the diagnosis of growth hormone deficiency55 (see Table 7-5). Interestingly, IGFBP-2 is increased in growth hormone deficiency, rather than being decreased as is the case with both IGF-I and IGFBP-3.

TABLE 7-5. Normal Values for Insulin-Like Binding Protein-2 (IGFBP-2)

ADRENOCORTICOTROPIC HORMONE AND ADRENOCORTICAL STEROIDS Adrenocorticotropic hormone (ACTH) is detectable in fetal serum as early as 10 to 14 weeks of gestation. The concentration of ACTH declines toward term, falling from 249 pg/mL 65 SE (standard error) to 143 pg/mL 9 SE at term. By 1 week of age, the serum ACTH levels are similar to those found in older children. The pattern of adrenal corticosteroids secreted varies from those found in older children. This is due to the presence of a fetal zone in the adrenal gland that constitutes 70% to 85% of the total weight of the adrenal gland at term. In both full-term and premature infants, the fetal zone undergoes rapid involution after birth, so that it constitutes just 10% of the adrenal gland by 4 to 5 months of age.56 Plasma concentrations of adrenal steroids are shown in Figure 7-11.

FIGURE 7-11. Variation of selected adrenal steroids with age and onset of puberty. (DHEA, dehydroepiandrosterone.) (Data derived from the normal values provided by Endocrine Sciences, Calabasas Hills, California.)

THYROID-STIMULATING HORMONE AND THYROID HORMONES As with the previously discussed hormone systems, thyroid function varies greatly depending on gestational age and perinatal circumstances (see Chap. 47). Thyroid-stimulating hormone (TSH) is detected in fetal serum as early as 10 weeks of gestation; however, the levels are lowonly 2.4U/mL 0.14 SE at 10 to 20 weeks of gestation. This increases to 9.6U/mL 0.93 SE at 25 to 30 weeks of gestation. The increase in TSH is accompanied by increases in mean free thyroxine (T4) and total T4 serum concentrations. Free T4 increases from 1.8 ng/dL at 10 to 20 weeks of gestation, to 2.6 ng/dL at 22 weeks, to 3 ng/dL at 30 weeks, and to 4 ng/dL at term. The rise in mean T 4-binding globulin during this period helps to produce an increase in serum T4 levels, which rise from 3g/dL at 20 weeks of gestation to 9.4g/dL (range, 5.715.6) at 30 weeks of gestation. Without any further rise in T4-binding globulin, the fetal T4 rises to 10.1g/mL at 35 weeks of gestation (range, 6.116.8) and to 10.9g/dL by term (range, 6.618.1).57,58 Between 30 and 40 weeks of gestation, some further maturation seems to take place of the hypothalamic systems responsible for the feedback inhibition of TSH by circulating thyroid hormones. This is suggested by the fall in serum TSH concentration to 8.9U/mL 0.93 SE at term as the serum free T4 rises.59 New second- and third-generation TSH assays have been introduced, and the absolute values listed should be viewed relative to the newer assays. The perinatal period is associated with marked changes in thyroid hormone levels. Within a half hour of birth, the serum TSH level rises to a mean of slightly over 18U/mL and then rapidly returns to normal within a day. The rise in serum TSH is paralleled within 24 hours by rises in serum T4 and triiodothyronine (T3) levels; T3 levels then decline in a few days, and T4 levels decline in a few weeks. Similar but less marked changes are seen in premature infants.59 The rise in serum T3 is not completely dependent on the rise in TSH but rather is also due to a change in the peripheral conversion of T4 to T3, as opposed to reverse T3, which occurs in the prenatal period59 (see Chap. 15, Chap. 33 and Chap. 47). The changes in thyroid hormone levels during the neonatal period in premature infants is also dependent on the degree of prematurity and the infant's health status.60 LUTEINIZING HORMONE, FOLLICLE-STIMULATING HORMONE, AND THE SEX STEROIDS As early as the 10th week of gestation, luteinizing hormone (LH) and follicle-stimulating hormone (FSH) are found in fetal pituitary glands. Serum LH and FSH levels in the fetus are high, well within the adult castrate range. Higher fetal serum FSH levels are found in females. Toward term, serum LH and FSH levels decline to the low levels typically found at term, but shortly after birth, serum levels of LH and FSH rise again, with LH levels higher in the male and FSH levels higher in the female. Serum LH and FSH levels may not decline to the concentrations deemed typical of childhood for 6 to 12 months (Table 7-6; see also Chap. 16 and Chap. 91). Marked differences appear to exist in the patterns of immunoreactive LH and FSH as compared with bioactive LH and FSH. The ratios also appear to vary with age and stage of development.61,62 and 63 The fetal serum estradiol level does not differ in males and females, but males have higher serum levels of testosterone before the 20th week of gestation,

corresponding to the period of development of the male external genitalia. The serum testosterone level ranges from 100 to 600 ng/dL.64 The fetal serum testosterone level declines toward term, but in cord blood, it remains higher in males than in females. Coincident with the postnatal surge of LH, serum testosterone levels in males rise again to mid-pubertal levels, remaining higher than concentrations thought appropriate for childhood until 4 to 9 months of age.65,66,67 and 68 The normal concentrations of these hormones and some other adrenal hormones are presented in Table 7-6 and Figure 7-11. Although the serum androstenedione level is in the typical prepubertal range by 1 year of age, it begins to rise again before any manifestations of puberty. Androstenedione begins to rise at 8 years of age in boys and at 7 years of age in girls. A similar pattern can be seen for dehydroepiandrosterone and for dehydroepiandrosterone sulfate. The rise in these hormones starts at 7 years of age in boys and at 6 years of age in girls. Similar to other adrenal androgens, dehydroepiandrosterone sulfate also manifests serum concentrations that vary with the child's age. In premature infants, the mean serum concentration is 263g/dL 40.1 SD; in full-term infants, 58.9g/dL 4.5 SD; in male children 1 to 6 years of age, 15.4g/dL 6.8 SD; in female children 1 to 6 years of age, 24.7g/dL 11.1 SD; in male children 6 to 8 years of age, 18.8g/dL 4.1 SD; in female children 6 to 8 years of age, 30.4g/dL 7.6 SD; in male children 8 to 10 years of age, 58.6g/dL 10.1 SD; and in female children 8 to 10 years of age, 117.3g/dL 41.7 SD69 (see Chap. 91 and Chap. 92). RENIN AND ALDOSTERONE The plasma concentrations of renin and aldosterone are markedly elevated in the newborn and decline slowly to accepted normal levels70 (see Table 7-6). CHAPTER REFERENCES
1. National Center for Health Statistics. Height and weight of children, United States. Vital and Health Statistics, PHS Pub. No. 1000series 11, No. 104. Public Health Service. Washington: Government Printing Office, September 1970. 2. National Center for Health Statistics. Height and weight of youths 1217 years, United States. Vital and Health Statistics, PHS Pub. No. 1000series 11, No. 124. Public Health Service. Washington: Government Printing Office, January 1973. 3. National Center for Health Statistics. NCHS growth curves for children birth18 years, United States. Vital and Health Statistics, PHS Pub. No. 1000series 11, No. 165. Public Health Service. Washington: Government Printing Office, November 1977. 4. National Center for Health Statistics. Body dimensions and proportions, white and Negro children 611 years, United States. Vital and Health Statistics, PHS Pub. No. 1000series 11, No. 143. Public Health Service. Washington: Government Printing Office, December 1974. 5. National Center for Health Statistics. Body weight, stature, and sitting height, white and Negro youths 1217 years, United States. Vital and Health Statistics, PHS Pub. No. 1000series 11, No. 126. Public Health Service. Washington: Government Printing Office, August 1973. 6. National Center for Health Statistics: Height and weight of children: socioeconomic status. Vital and Health Statistics, PHS Pub. No. 1000series 11, No. 119. Public Health Service. Washington: Government Printing Office, October 1972. 7. Barr GD, Allen CM, Shinefield HR. Height and weight of 7500 children of three skin colors. Am J Dis Child 1972; 124:866. 8. Wingerd J, Schoen EJ, Solomon IL. Growth standards in the first two years of life based on measurements of white and black children in a prepaid healthcare program. Pediatrics 1971; 47:818. 9. Wingerd J, Solomon IL, Schoen EJ. Parent-specific height standards for preadolescent children of three racial groups, with method for rapid determination. Pediatrics 1973; 52:555. 10. Lippe BM. Short stature in children: evaluation and management. J Pediatr Health Care 1987; 1:313. 11. Tanner JM, Whitehouse RM. Clinical longitudinal standards for height, weight, height velocity, weight velocity, and stages of puberty. Arch Dis Child 1976; 51:170. 12. Tanner JM, Davies PSW. Clinical longitudinal standards for height and weight velocity for North American children. J Pediatr 1985; 107:317. 13. Smith DW, Truog W, Rogers JE, et al. Shifting linear growth during infancy: illustration of genetic factors in growth from fetal life through infancy. J Pediatr 1976; 89:225. 14. Tanner JM, Goldstein H, Whitehouse RH. Standards for children's heights at ages 29 years allowing for heights of parents. Arch Dis Child 1970; 45:755. 15. Himes JH, Roche AF, Thissen D, Moore WM. Parent-specific adjustments for evaluation of recumbent length and stature of children. Pediatrics 1985; 75:304. 16. Luo ZC, Albertson-Wikland K, Karlberg J. Target height as predicted by parental heights in a population-based study. Pediatr Res 1998; 44:563. 17. Cruise MO. A longitudinal study of the growth of low birth weight infants. I. Velocity and distance growth, birth to 3 years. Pediatrics 1973; 51:620. 18. Fritzhardinge PM, Steven EM. The small-for-date infant. I. Later growth patterns. Pediatrics 1972; 49:671. 19. Guo SS, Roche AE, Chumlea W, et al. Adjustments to the observed growth of preterm low birth weight infants for application to infants who are small for gestational age at birth. Acta Med Auxol 1998; 30:71. 20. Strauss RS, Dietz WH. Growth and development of term infants born with low birth weight: effects of genetic and environmental factors. J Pediatr 1998; 133:67. 21. Leger J, Limoni C, Collin D, Czernichow P. Prediction factors in the determination of final height in subjects born small for gestational age. Pediatr Res 1998; 43:808. 22. Ranke MB. Disease-specific growth charts: do we need them? Acta Paediatr Scand 1989; 356(Suppl):17. 23. Park E, Bailey JD, Cowell CA. Growth and maturation of patients with Turner's syndrome. Pediatr Res 1983; 17:1. 24. Lyon AJ, Preece MA, Grant DB. Growth curves for girls with Turner syndrome. Arch Dis Child 1985; 60:932. 25. Marshall WA. Evaluation of growth rate in height over periods of less than one year. Arch Dis Child 1971; 46:414. 26. Cureton KJ, Boileau RA, Lohman TG. A comparison of densitometric, potassium-40 and skinfold estimates of body composition in prepubescent boys. Hum Biol 1975; 47:321. 27. National Center for Health Statistics. Skinfold thickness of children 611 years. Vital and Health Statistics, PHS Pub. No. 1000series 11, No. 120. Public Health Service, Washington: Government Printing Office, October 1972. 28. National Center for Health Statistics. Skinfold thickness of youths 1217 years, United States. Vital and Health Statistics. PHS Pub. No. 1000series 11, No. 132. Public Health Service, Washington: Government Printing Office, January 1974. 29. Sann L, Durand M, Picard J, et al. Arm fat and muscle areas in infancy. Arch Dis Child 1988; 63:256. 30. Rosner B, Prineas R, Loggie J, Daniels SR. Percentiles for body mass index in U.S. children 5 to 17 years of age. J Pediatr 1998; 132:211. 31. Dietz WH, Robinson TN. Use of the body mass index (BMI) as a measure of overweight in children and adolescents. J Pediatr 1998; 132:191. 32. Greulich W, Pyle SL. A radiographic standard of reference for the growing hand and wrist. Chicago: Year Book Medical Publishers, 1971. 33. Tanner JM, Whitehouse RH, Cameron N, et al. Assessment of skeletal maturity and prediction of adult height (TW2 method), 2nd ed. London: Academic Press, 1983. 34. Tanner J, Oshman D, Bahhage F, Healy M. Tanner-Whitehouse bone age reference values for North American children. J Pediatr 1997; 131:34. 35. Graham CB. Assessment of bone maturation: methods and pitfalls. Radiol Clin North Am 1972; 10:185. 36. National Center for Health Statistics. Skeletal maturity of children 611 years. Vital and Health Statistics, PHS Pub. No. 1000series 11, No. 140. Public Health Service. Washington: Government Printing Office, November 1974. 37. Buckler JMH. Skeletal age changes in puberty. Arch Dis Child 1984; 59:115. 38. Bailey N, Pinneau SR. Tables for predicting adult height from skeletal age: revised for use with the Greulich-Pyle hand standards. J Pediatr 1952; 14:423. 39. Roche AF, Wainer H, Thissen D. The RWT method for the prediction of adult stature. Pediatrics 1975; 56:1026. 40. Lenko HL. Prediction of adult height with various methods in Finnish children. Acta Paediatr Scand 1979; 68:85. 41. Zachmann M, Sobradillo B, Frank H, Prader A. Bayley-Pinneau, Roche-Wainer-Thissen and Tanner height predictions in normal children and in patients with various pathologic conditions. J Pediatr 1978; 93:749. 42. de Waal WJ, Greyn-Fokker MH, Stijnen TH, et al. Accuracy of final height prediction and effect of growth-reductive therapy in 362 constitutionally tall children. J Clin Endocrinol Metab 1996; 81:1206. 43. Abbassi V. Growth and normal puberty. Pediatrics 1998; 102:507. 44. August GP, Julius JR, Blethen SL. Adult height in children with growth hormone deficiency who are treated with biosynthetic growth hormone: the National Cooperative Growth Study experience. Pediatrics 1998; 102:512. 45. Marti-Henneberg C, Vizmanos B. The duration of puberty in girls is related to the timing of its onset. J Pediatr 1997; 131:618. 46. Blethen SL, Gaines S, Weldon V. Comparison of predicted and adult heights in short boys: effect of androgen therapy. Pediatr Res 1984; 18:467. 47. Winter JSD, Faiman C. Pituitary-gonadal relations in male children and adolescents. Pediatr Res 1972; 6:126. 48. Zachmann M, Prader A, Kind HP, et al. Testicular volume during adolescence: cross sectional and longitudinal studies. Helv Paediatr Acta 1974; 29:61. 49. Schonfeld WA. Primary and secondary sexual characteristics: study of their development in males from birth through maturity, with biometric study of penis and testes. Am J Dis Child 1943; 65:535. 50. Belisle S, Tulchinsky D. Amniotic fluid hormones. In: Tulchinsky D, Ryan KJ, eds. Maternal-fetal endocrinology. Philadelphia: WB Saunders, 1980:169. 51. Gluckman PD, Grumbach MM, Kaplan SL. The human fetal hypothalamus and pituitary gland: the maturation of neuroendocrine mechanisms controlling the secretion of fetal pituitary growth hormone, prolactin, gonadotropin, and adrenocorticotropin-related peptides. In: Tulchinsky D, Ryan KJ, eds. Maternal-fetal endocrinology. Philadelphia: WB Saunders, 1980:196. 52. Cacciari E, Cicognani A, Pirazzoli P, et al. Differences in somatomedin-C between short-normal subjects and those of normal height. J Pediatr 1985; 106:891. 53. Juul A, Bang P, Hertel NT, et al. Serum insulin-like growth factor-I in 1030 healthy children, adolescents, and adults: relation to age, sex, stage of puberty, testicular size, and body mass index. J Clin Endocrinol Metab 1994; 78:744. 54. Blum WF, Albertsson-Wikland K, Rosberg S, Ranke MB. Serum levels of insulin-like growth factor I (IGF-I) and IGF binding protein 3 reflect spontaneous growth hormone secretion. J Clin Endocrinol Metab 1993; 76:1610. 55. Smith WJ, Nam TJ, Underwood LE, et al. Use of Insulin-like growth factor-binding protein-2 (IGFBP-2), IGFBP-3, and IGF-I for assessing growth hormone status in short children. J Clin Endocrinol Metab 1993; 77:1294. 56. Sperling MA. Newborn adaptation: adrenocortical hormones and ACTH. In: Tulchinsky D, Ryan KJ, eds. Maternal-fetal endocrinology. Philadelphia: WB Saunders, 1980:387. 57. Oddie TH, Fisher DA, Bernard B, Lam RW. Thyroid function at birth in infants of 30 to 45 weeks gestation. J Pediatr 1977; 90:803. 58. Burrow GN, Fisher DA, Larsen PR. Maternal and fetal thyroid function. N Engl J Med 1994; 331:1072. 59. Fisher DA, Klein AH. The ontogenesis of thyroid function and its relationship to neonatal thermogenesis. In: Tulchinsky D, Ryan KJ, eds. Maternal-fetal endocrinology. Philadelphia: WB Saunders, 1980:281. 60. Van Wassenaer AG, Kok JH, Dekker FW, De Vulder JJM. Thyroid function in very preterm infants: influences of gestational age and disease. Pediatr Res 1997; 42:604. 61. Beitins IZ, Padmanabhan V. Bioactivity of gonadotropins. Endocrinol Metab Clin North Am 1991; 20:85. 62. Beck-Peccoz P, Padman V, Baggiani AM, et al. Maturation of hypotha-lamic-pituitary-gonadal function in normal human fetuses: circulating levels of gonadotropins, their common a-subunits and free testosterone, and discrepancy between immunological and biological activities of circulating follicle-stimulating hormone. J Clin Endocrinol Metab 1991; 73:525. 63. Kletter GB, Padmanabhan V, Brown MB, et al. Serum bioactive gonadotropin during male puberty: a longitudinal study. J Clin Endocrinol Metab 1993; 76:432. 64. Reyes FI, Boroditsky RS, Winter JSD, Faiman C. Studies on human sexual development. II. Fetal and maternal serum gonadotropin and sex steroid concentrations. J Clin Endocrinol Metab 1974; 38:612. 65. Forest MC, Cathiard AM, Bertrand JA. Evidence of testicular activity in early infancy. J Clin Endocrinol Metab 1973; 37:148. 66. Forest M, Cathiard AM. Pattern of plasma testosterone and D-4-andro-stenedione in normal newborns: evidence for testicular activity at birth. J Clin Endocrinol Metab 1975; 41:977. 67. Winter JSD, Faiman C, Hobson WC, et al. Patterns of serum gonadotropin concentrations from birth to four years of age in man and chimpanzee. J Clin Endocrinol Metab 1975; 40:545. 68. Winter JSD, Hughes IA, Reyes FI, et al. Pituitary-gonadal relations in infancy. II. Patterns of serum gonadal steroid concentration in man from birth to two years of age. J Clin Endocrinol Metab 1976; 42:679. 69. Reiter ED, Fuldaurer VG, Root AW. Secretion of the adrenal androgen, dehydroepiandrosterone sulfate, during normal infancy, childhood, in sick infants and in children with endocrinologic abnormalities. J Pediatr 1977; 90:766.

70. Fiselier T, Monnens L, van Munster P, et al. The renin-angiotensin-aldoster-one system in infancy and childhood in basal conditions and after stimulation. Eur J Pediatr 1984; 143:18. 82blank

CHAPTER 8 MORPHOLOGY OF THE ENDOCRINE BRAIN, HYPOTHALAMUS, AND NEUROHYPOPHYSIS Principles and Practice of Endocrinology and Metabolism

CHAPTER 8 MORPHOLOGY OF THE ENDOCRINE BRAIN, HYPOTHALAMUS, AND NEUROHYPOPHYSIS

JOHN R. SLADEK, JR., AND CELIA D. SLADEK Overview of the Hypothalamus Nuclear Topography Hypothalamic Zones Preoptic Area Classic Pathways Afferent Pathways Efferent Pathways Hypothalamo-Neurohypophysial System Chemical Neuroanatomy Hormones as Chemical Neurotransmitters Vasopressin Release as a Model of Interactive Chemical Circuitry Chapter References

OVERVIEW OF THE HYPOTHALAMUS

The human hypothalamus is a tiny wedged-shaped mass of tissue, composed primarily of gray matter, that subserves widespread functions, ranging from those considered somewhat automatic (i.e., autonomic) to more complex behaviors requiring a high level of integration. The hypothalamus is situated in the ventralmost diencephalon (Fig. 8-1). It is bounded laterally by portions of the subthalamus; medially by the vertically oriented, slitlike third ventricle; rostrally by the lamina terminalis; caudally by the mesencephalon; and dorsally by the thalamus. Ventrally, the hypothalamus is contiguous with the infundibulum and pituitary stalk; the latter serves to transmit axons en route to the neurohypophysis. On the ventral surface of the brain, the hypothalamus appears as a prominent set of protuberances, the paired mammillary bodies caudally, and the midline infundibular eminence. The boundaries of the hypothalamus are easily identified by the optic chiasm rostrally and by the caudal edge of the mammillary bodies as they are situated at the rostral limit of the interpeduncular fossa. Here, the caudal limits of the hypothalamus merge with the midbrain.

FIGURE 8-1. Human hypothalamus in sagittal section. The hypothalamus is bounded anteriorly by the lamina terminalis, the tissue that bridges the anterior commissure and optic chiasm; posteriorly by the brainstem; and dorsally by the overlying thalamus (adjacent to the label for the third ventricle). The hypothalamus is relatively small in comparison to other brain regions, yet is intimately involved in mediating or moderating many brain and endocrine functions. (From Krieger DT, Hughes JC, eds. Neuroendocrinology. Sunderland, MA: Sinauer Associates, 1980:3.)

Classic neuroanatomic techniques have been used to identify hypothalamic nuclei cytoarchitectonically. Some nuclei are relatively easy to identify microscopically, based on size or packing density of neurons; others benefit from placement close to readily identified landmarks such as the third ventricle or the optic chiasm. Prominent myelinated fiber bundles punctuate the relatively nondescript nuclear divisions of the hypothalamus, indicating the diverse reciprocal connections that exist between the hypothalamus and numerous brain regions. The fornix and mammillary fasciculus interconnect the hypothalamus with, for example, the hippocampus, the thalamus, and the brainstem. Other brain regions influence hypothalamic function through pathways that are seen to advantage with current histochemical methods. These systems, which include the medial forebrain bundle and the dorsal longitudinal fasciculus, interconnect the hypothalamus with olfactory areas rostrally and with autonomic centers of the brainstem and spinal cord caudally. Thus, the hypothalamus is an important link between the forebrain and brainstem, integrating visceral, endocrine, and behavioral functions through a highly ordered set of reciprocating circuits.

NUCLEAR TOPOGRAPHY
HYPOTHALAMIC ZONES Because the hypothalamus is so rich in perikaryal groups and so poor in landmarks (e.g., myelinated fiber bundles), the prudent approach is to consider easily identified hypothalamic zones as geographic points with respect to the more than two dozen classically identified hypothalamic nuclei, as detailed by Nauta and Haymaker.1 Zones in both longitudinal and coronal planes are illustrated in Figure 8-2.

FIGURE 8-2. Diagram showing anatomic relationship of hypothalamic nuclei, by zones. Here, the hypothalamus is considered as a hexagon, bordered anteriorly by the preoptic area and posteriorly by the midbrain. The third ventricle (wide black line) forms its midline. Most hypothalamic nuclei can be found within three longitudinal zones, designated as midline, medial, and lateral in relation to the third ventricle and fornix. These zones can be further subdivided into three anteroposterior regions or levels (supraoptic, tuberal, and mammillary), allowing identification of hypothalamic nuclei by position, an approach that is much simpler than distinguishing the nuclei by the classic cytoarchitectonic method. The three levels, shown in the coronal plane in Figure 8-3, depict the essential hypothalamic nuclei.

Longitudinal Plane. The longitudinal zones are based on the phylogenetically primitive organization of lower vertebrates, in which the hypothalamus is characterized by a relatively cell-rich medial zone. A somewhat acellular lateral zone is separated from the medial zone by a prominent fiber bundle, the fornix. The medial zone can be subdivided further into a midline zone that is immediately adjacent to the third ventricle and contains a relatively homogeneous nuclear mass called the periventricular stratum. The more laterally placed portions of the medial hypothalamic zone contain reasonably differentiated cell clusters, including the medial preoptic,

anterior hypothalamic, ventromedial, dorsomedial, paraventricular, posterior, and premammillary nuclei. Coronal Plane. For convenience, the hypothalamus in the coronal plane can be further subdivided anteroposteriorly into three regions: the supraoptic, tuberal, and mammillary regions (Fig. 8-3 and Fig. 8-4). The supraoptic region includes that portion of the hypothalamus situated between the optic chiasm and tuber cinereum. The tuberal zone extends caudally to the most anterior portion of the mammillary bodies. The mammillary region includes the most caudal hypothalamus to the mesodiencephalic junction.

FIGURE 8-3. The hypothalamus in the coronal plane, demonstrating the relative positions of hypothalamic nuclei. At the supraoptic level, the midline zone (dashed line) consists of the periventricular nucleus (Peri) and its ventral expansion, the suprachiasmatic nucleus (SCN). The periventricular nucleus continues through the more caudal levels. At tuberal levels, the SCN is replaced by the arcuate nucleus (Arc). The medial zone, defined as all the tissue between the fornix laterally and the periventricular nucleus medially, contains five major hypothalamic nuclei that essentially replace each other positionally at each level. Thus, the paraventricular (PVN) and anterior (AN) nuclei, which occupy the supraoptic level, are replaced by the dorsomedial (DM) and ventromedial (VM) nuclei at tuberal levels; these nuclei in turn are replaced by the posterior nucleus (POST) at the most caudal, mammillary level. The lateral zone consists almost exclusively of the lateral hypothalamic area (LHA), with the exception of the supraoptic nucleus (SON), which is seen at supraoptic levels. The third ventricle appears as the dark structure in the midline. (F, fornix; ME, median eminence; MM, mammillary body; OC, optic chiasm; OT, optic tract.)

FIGURE 8-4. The hypothalamus in a parasagittal plane, showing the relative positions of major nuclei. These relationships can be better appreciated by comparing this figure with Figure 8-1 and Figure 8-3. For example, three nuclear groups, the paraventricular, dorsomedial, and posterior, appear to replace one another as one proceeds anatomically, rostrally to caudally, through the three major coronal levelssupraoptic, tuberal, and mammillary (see Fig. 8-3). (From Carpenter MB, Sutin J, eds. Human neuroanatomy. Baltimore: Williams & Wilkins, 1983:553.)

The midline zone accordingly contains the periventricular stratum throughout most of its rostrocaudal extent. At tuberal regions, the ventral periventricular stratum expands laterally to accommodate the arcuate (infundibular) nucleus. At supraoptic levels, a ventral region of the periventricular stratum is identified as the suprachiasmatic nucleus. Although numerous other fine points of periventricular zone anatomy exist, one could consider the two subcomponents of this region as the suprachiasmatic and arcuate nuclei. The medial hypothalamic zone, defined as all remaining tissue lateral to the periventricular stratum and medial to the descending limb of the fornix, contains the paraventricular and anterior hypothalamic nuclei at supraoptic levels, the dorsomedial and ventromedial nuclei at tuberal levels, and the posterior hypothalamic nucleus at mammillary levels (see Fig. 8-3). Thus, as one proceeds rostrally to caudally through the three hypothalamic regions, a gradual replacement of these well-differentiated nuclei is seen. The lateral hypothalamic zone is relatively undistinguished, with the lateral hypothalamic area occupying most of this zone in all three regions. The lone exception is the supraoptic nucleus, which is seen at the supraoptic level. Because of the complex nature of the paraventricular nucleus, it is sometimes considered as part of the periventricular zone, but the alarlike lateral extensions of this nucleus clearly place a large part of it within the medial zone. PREOPTIC AREA The preoptic area, located rostral to the strict limits of the hypothalamus, contains nuclei in all three (medial to lateral) zones, including the preoptic periventricular, medial preoptic, and lateral preoptic nuclei (see Fig. 8-4). Functionally, the preoptic region is integrated with the remainder of the hypothalamus because it includes neurons that regulate anterior pituitary function as well as structures that are essential for fluid and electrolyte balance. A full description of these hypothalamic regions and nuclei may be found in two classic accounts.1,2

CLASSIC PATHWAYS
AFFERENT PATHWAYS Afferent pathways to hypothalamic nuclei arise primarily from the brainstem, thalamus, basal ganglia, cerebral cortex, and olfactory areas. The detailed anatomy of these connections and hypothalamic interconnections is described in several fine reviews.1,3,4 Briefly, the reticular formation and visceral centers of the brainstem connect with the hypothalamus through two prominent pathways, the mammillary peduncle and the dorsal longitudinal fasciculus. Visceral and somatic information also reaches the hypothalamus from the locus ceruleus, vagal nuclei, periaqueductal gray area, and nuclei of the solitary tract (Fig. 8-5). The fornix transmits fibers from the hippocampus by direct projections to the mammillary bodies. Additional afferents from the piriform cortex and the amygdala also reach the hypothalamus, probably by other routes. Olfactory information through the medial forebrain bundle and stria terminalis reaches the hypothalamus either directly or indirectly through the previously mentioned cortical regions, the stria terminalis being the primary pathway from the amygdala. Most of the afferent pathways are reciprocal; thus, the hypothalamic efferent connections are extensive.

FIGURE 8-5. Major hypothalamic pathways. Pathways to the hypothalamus arise from several areas, including olfactory, limbic, and brainstem regions; reciprocal

circuits exist, particularly to autonomic centers of the caudal medulla and spinal cord. The dorsal longitudinal fasciculus conveys most of the descending information to brainstem nuclei and is identified by the parallel, opposite arrows in this drawing. Another major pathway, the fornix, carries higher cortical information from the hippocampus (Hipp.) to the mammillary nucleus. (AV, nucleus ventralis anterior of thalamus; BL, Ce., and Co., basolateral, central, and cortical amygdaloid nuclei; Bulb. olf., olfactory bulb; Coll. sup., superior colliculus; H, hypothalamus; M, mammillary body; MD, dorsomedial thalamic nucleus; N. Dors. n. X, dorsal motor vagal nucleus; N. loc. coer., locus coeruleus; N. raphe, nuclei of the raphe; N. tr. solit., nucleus tractus solitarius; Olf. Cort., olfactory cortex; Optic ch., optic chiasma; Periaq. gr., periaqueductal gray; Prefr. cort., prefrontal cortex; RF, reticular formation; S, septum; VM, ventromedial hypothalamic nucleus.) (From Brodal A, ed. Neurological anatomy in relation to clinical medicine, 3rd ed. New York: Oxford University Press, 1981:698.)

EFFERENT PATHWAYS The dorsal longitudinal fasciculus transmits information to brainstem reticular centers as well as to visceral and somatic efferent nuclei. This system descends caudally to innervate preganglionic autonomic centers of the spinal cord, particularly the intermediolateral gray column.5,6 Although the hypothalamus was once thought to interconnect with autonomic centers by multisynaptic pathways, it is now clear that the dorsal motor nucleus of the vagus, the nuclei of the solitary tract, and the nucleus ambiguus receive direct projections from the paraventricular nucleus of the hypothalamus. Thus, reciprocal connections between these regions form an integrated autonomic circuit. Other fibers reach the brainstem through the medial forebrain bundle. Hypothalamic efferents projecting to the thalamus traverse the mammillothalamic tract to the anterior nucleus of the thalamus, where they are relayed to the cerebral cortex. Other efferent fibers ascend over the more diffuse periventricular system. The hypothalamus is connected with the neurohypophysis by the hypothalamo-neurohypophysial tract, a system of unmyelinated fibers that terminates in the neurohypophysis for the purpose of storing and then delivering vasopressin, oxytocin, and possibly other endogenous peptides such as dynorphin to the pituitary blood. HYPOTHALAMO-NEUROHYPOPHYSIAL SYSTEM In contrast to other hypothalamic nuclei, the paraventricular and supraoptic nuclei are easy to identify (Fig. 8-6). Each nucleus contains the largest nerve cells in the hypothalamus (Fig. 8-7). The cells stain densely with common dyes that reveal the abundant ribonucleoprotein (i.e., Nissl substance) in their perikaryal cytoplasm. Ultrastructurally, the protein-synthesizing machinery of these neurons also is evident and reflects their high level of activity with respect to the manufacture of vasopressin, oxytocin, and coexistent peptides. The neurons are revealed in excellent detail when stained immunohistochemically with antibodies directed against specific hypothalamic peptides (see Fig. 8-7).7,8 Such analysis has revealed that oxytocin neurons also contain cholecystokinin9 and that vasopressin neurons also contain dynorphin.10 Axons from the supraoptic and paraventricular nuclei as well as associated accessory nuclei of the hypothalamus project ventrally and caudally (see Fig. 8-6) to the underlying neurohypophysis, where they release oxytocin, vasopressin, and associated peptides into the peripheral circulation to regulate fluid and electrolyte balance and to initiate the smooth muscle contraction that is associated with lactation and parturition.

FIGURE 8-6. Immunohistochemical staining for neurophysins associated with vasopressin and oxytocin reveals the extent of the neurohypophysial system at the supraoptic level of the rat hypothalamus. Dense staining is seen within neurons of the paraventricular (PVN) and supraoptic (SON) nuclei. A ventral flow of axons from each nucleus (arrows) represents the origin of the hypothalamo-neurohypophysial tract as it courses to the posterior pituitary. Stained neurons also are seen within the suprachiasmatic nucleus (SCN); however, this nucleus does not contribute fibers to the neural lobe. (OC, optic chiasm; V, third ventricle.) Original magnification 30

FIGURE 8-7. Appearance of neurons of the paraventricular nucleus after immunohistochemical staining for vasopressin. This set of magnocellular neurons occupies a subnucleus of the paraventricular nucleus that contains primarily vasopressin neurons that project to the neural lobe.30 The neurons are large, multipolar, and possess beaded processes. The position of the third ventricle is indicated (V). Original magnification 350

Other neurons of the paraventricular nucleus are global, with widespread connections to autonomic centers of the brainstem and spinal cord, and to the forebrain and cortical areas, including the septum, cingulum, and hippocampus.11a These far-reaching interconnections focus attention on oxytocin and vasopressin as more than simply neurohypophysial hormones. The demonstration of these peptides in presynaptic nerve terminals suggests that they may function as neurotransmitters.

CHEMICAL NEUROANATOMY
HORMONES AS CHEMICAL NEUROTRANSMITTERS The hypothalamus is one of the most complicated areas of the brain with respect to chemical neurotransmitters because of the small amount of tissue occupied by the hypothalamus and the great number of transmitter substances located within hypothalamic nerve cells and associated fiber systems.12 The classic studies of Bargmann13 and of Scharrer and Scharrer,14 who are credited with describing the principles of neurosecretion, brought to light the unique chemical characteristics of neurons of the supraoptic and paraventricular nuclei. The introduction in the early 1960s of the Falck-Hillarp histofluorescence method15 allowed identification of the catecholamines dopamine and norepinephrine within the hypothalamus. Most notable were the dopaminergic neurons of the tuberoinfundibular system, which are involved in regulating the release of anterior pituitary substances.16 A decade later, immunohistochemical methods17 increased the list of known hypothalamic chemical neurotransmitters and modulators to include substances such as luteinizing hormonereleasing hormone, corticotropin-releasing hormone, vasoactive intestinal peptide, neurotensin, somatostatin, enkephalin, endorphin, cholecystokinin, galanin, and several others.18,19 Although the discovery of luteinizing hormonereleasing hormone20 and somatostatin within hypothalamic, preoptic, and adjacent regions of the endocrine hypothalamus was not surprising, the finding of gut peptides and the discovery of a complex system of opioid neurons21 have redefined the hypot halamus, based on the chemical cytoarchitecture of transmitters and hormones.18,21a The application of in situ hybridization techniques to localize messenger RNA for these peptides indicates that this wide array of peptides is synthesized in the

hypothalamus and that neurons have the capacity to synthesize multiple regulatory peptides simultaneously.22 The specific peptides produced by a given neuron are not static but depend on the stimuli received by that cell.22,23 For example, hypophysectomy dramatically increases the expression of galanin in the vasopressin neurons of the supraoptic nucleus and of cholecystokinin in the oxytocin neurons, but salt loading induces the expression of tyrosine hydroxylase in the vasopressin neurons and of corticotropin-releasing hormone in the oxytocin neurons.24 The role of these simultaneously released peptides is not completely defined, but in at least some instances, they interact to regulate hormone release from the anterior pituitary or to modulate hormone release from the posterior pituitary. VASOPRESSIN RELEASE AS A MODEL OF INTERACTIVE CHEMICAL CIRCUITRY NOREPINEPHRINE REGULATION OF VASOPRESSIN AND OXYTOCIN The role of afferents to the paraventricular and supraoptic nuclei is considered relative to the regulation of vasopressin release as exemplary of the kind of functionally interactive chemical circuitry that is being revealed with respect to hypothalamic neuroanatomy. The supraoptic and paraventricular nuclei receive dense, diverse afferent inputs, many arising from the brainstem reticular formation. One of these is a well-defined system of noradrenergic afferents to vasopressin neurons of the supraoptic and paraventricular nuclei (Fig. 8-8). The paraventricular nuclei in turn send reciprocal peptidergic fibers to the reticular core of the brainstem. Norepinephrine-containing perikaryal groups in the brainstem have been designated A1 to A7.25 These noradrenergic neurons originally were seen primarily within the reticular formation of the pons and medulla. Later, attention focused on groups A1, A2, A5, and A7 as projecting to the hypothalamus through a ventral pathway that ascends within the dorsal portion of the reticular formation of the brainstem, entering the medial forebrain bundle at hypothalamic levels in association with serotonergic and dopaminergic systems from the brainstem. 26 On reaching the hypothalamus, these fibers exit the medial forebrain bundle to supply the supraoptic and paraventricular nuclei. The densest patterns of noradrenergic fibersperhaps the densest patterns in the entire brainare seen in the mammalian hypothalamus27,28 (Fig. 8-9). These fibers appear in contact with the cell bodies of the magnocellular neurons, lending further support to the concept that norepinephrine plays a role in the regulation of neurohypophysial peptides.29

FIGURE 8-8. Ascending noradrenergic axons reach the magnocellular nuclei of the hypothalamus through the ventral norepinephrine pathway (VNE) of the brainstem, which continues rostrally in the medial forebrain bundle (MFB) of the diencephalon. The neurons of origin of this system (A1, A2) are located in the lateral reticular formation of the medulla near the lateral reticular nucleus (LRN) and in an area of the dorsomedial medulla that is important in cardiovascular regulationthe nucleus solitarius (SOL) and the dorsal motor vagal nucleus (DMX), respectively. This pathway, which is probably reciprocal from the paraventricular nucleus (PVN), supplies a dense noradrenergic innervation to the magnocellular nuclei, as shown for the monkey (also see Fig. 8-9). (CC, corpus callosum; MLF, medial longitudinal fasciculus; OC, optic chiasm; PY, pyramidal tract; SON, supraoptic nucleus.)

FIGURE 8-9. Patterns of catecholamine innervation of hypothalamic nuclei at a supraoptic level in the rhesus monkey. Exceptionally dense patterns exist in the paraventricular (Pa) and supraoptic (So) nuclei; less impressive patterns are seen in other nuclei at this level. Depictions such as this led to intense examination of the role of catecholamines in the release of neurohy-pophysial peptides. (An, anterior nucleus; op ch, optic chiasma; Prl, lateral preoptic nucleus; Sc, suprachiasmatic nucleus; v, third ventricle.) (From Hoffman GE, Felten DL, Sladek JR Jr. Monoamine distribution in primate brain. III. Catecholamine-containing varicosities in the hypothalamus of Macaca mulatta. Am J Anat 1976; 147:501.)

In subsequent studies, the simultaneous demonstration of catecholamine fluorescence and peptide immunohistochemistry30 revealed a consistent juxtaposition between catecholamine varicosities and magnocellular perikarya in the supraoptic and paraventricular nuclei in both rodents31 and primates.32 Specifically, vasopressin-containing neurons in ventral portions of the supraoptic nucleus and the major vasopressin subcomponent of the paraventricular nucleus were seen to be studded with brightly fluorescent catecholamine-containing varicosities. In contrast, oxytocin neurons received far fewer fluorescent fibers on their cell bodies and proximal dendrites. The preferential innervation of vasopressin neurons primarily reflects innervation by the A1 noradrenergic cells of the ventrolateral medulla.33,34 and 35 The sparse innervation of the oxytocin neurons reflects innervation of the nuclei by the A2 noradrenergic cell group that does not differentiate between oxytocin and vasopressin neurons.35 CATECHOLAMINES, VASOPRESSIN, AND AUTONOMIC REGULATION The noradrenergic cell groups in the brainstem that innervate the vasopressin neurons are intimately involved in autonomic regulation. The A2 catecholamine neurons are located in the nucleus of the tractus solitarius, which receives baroreceptor information from the carotid sinus and aortic arch. This information is also transmitted to the A1 cells in the ventrolateral medulla that project to the vasopressin neurons.36 The paraventricular nucleus also is innervated by the locus ceruleus (A6), an important autonomic relay center.37 Thus, the catecholamine input to the supraoptic and paraventricular nuclei represents a source of autonomic regulation of neurohypophysial function. These pathways are important because blood pressure, blood volume, and the partial pressure of oxygen are potent regulators of vasopressin release from the neural lobe. Stimulation of the A1 region results in an increase in blood pressure and enhanced vasopressin release, and inhibition of this region prevents baroreceptor-initiated secretion of vasopressin.38 This response is prevented by destruction of the catecholamine terminals in the supraoptic and paraventricular nuclei39 but is not prevented by administration of adrenergic antagonists.40 This can be explained by the finding that, in addition to norepinephrine, the A1 neurons produce neuropeptide Y, adenosine triphosphate (ATP), and substance P.41,42 Pharmacologic evidence indicates that these neuroactive substances participate in the excitation of vasopressin neurons by the A1 pathway.43,44 The central role of the paraventricular nucleus and vasopressin in blood pressure regulation and other autonomic functions was underscored by the finding of reciprocal pathways containing vasopressin and oxytocin in the same brainstem regions that give origin to the ascending noradrenergic fibers to the supraoptic and paraventricular nuclei. The finding of neurophysin-positive varicosities in juxtaposition to norepinephrine perikarya in the A1, A2, A5, and A7 brainstem groups suggests a functionally interactive reciprocal circuit.45 Moreover, microinjection of vasopressin into these regions has profound effects on cardiovascular function. Thus, a reciprocal pathway involving catecholamine afferents and vasopressin efferents is one mechanism involved in hypothalamic modulation of the autonomic nervous system.

OTHER CHEMICALLY DEFINED AFFERENTS AND VASOPRESSIN RELEASE The paraventricular and supraoptic nuclei receive chemically defined afferents from other central nervous system sites.46,47 These include serotonergic projections from the midbrain raphe nuclei, GABAergic (transmitting or secreting g-aminobutyric acid) projections from the region of the nucleus accumbens, and several hypothalamic projections. The hypothalamic afferents include cholinergic48 and GABAergic projections arising from cells near the paraventricular and supraoptic nuclei, b-endorphinergic afferents from the arcuate nucleus, histaminergic afferents from the tuberomammillary nuclei,49,50 dopaminergic afferents from neurons in the A11, A12, and A13 groups,51 and numerous projections from the preoptic region, including projections from the subfornical organ and the organum vasculosum of the lamina terminalis. The chemical nature of these last projections includes glutamatergic, GABAergic, angiotensinergic, and atriopeptidergic fibers. The functional role of many of these afferents is unknown, but the afferents from the preoptic region and the circumventricular organs clearly are intimately involved in the osmotic regulation of vasopressin release as part of their role in fluid and electrolyte balance. A particularly important role for the classic excitatory amino acid neurotransmitter glutamate in regulating hypothalamic activity has been suggested by the presence of numerous immunocytochemically identified glutamatergic synapses in the hypothalamus and the finding that glutamate antagonists virtually eliminate excitatory postsynaptic potentials in the paraventricular, supraoptic, and arcuate nuclei. Thus, the hypothalamus continues to emerge as a complex brain region in which a wide variety of classic and more recently discovered transmitters serve numerous endocrine and other functions related to autonomic, limbic, and pituitary activity.52 CHAPTER REFERENCES
1. Nauta WJH, Haymaker W. Hypothalamic nuclei and fiber connections. In: Haymaker W, Anderson E, Nauta WJH, eds. The hypothalamus. Springfield, IL: Charles C Thomas, 1969:136. 2. Crosby EC, Woodburne RT. The comparative anatomy of the preoptic area and the hypothalamus. Research PublicationsAssociation for Research in Nervous and Mental Disease. New York: Raven Press, 1939; 20:52. 3. Brodal A. The autonomic nervous system: the hypothalamus. In: Brodal A, ed. Neurological anatomy in relation to clinical medicine. New York: Oxford University Press, 1981:698. 4. Palkovits M, Zaborszky L. Neural connections of the hypothalamus. In: Morgane PJ, Panksepp J, eds. Handbook of the hypothalamus, vol 1. Anatomy of the hypothalamus. New York: Marcel Dekker, 1979:379. 5. Swanson LW, Kuypers HBJM. The paraventricular nucleus of the hypothalamus: cytoarchitectonic subdivision and organization of projections to the pituitary, dorsal vagal complex, and spinal cord as demonstrated by retrograde fluorescence double-labeling methods. J Comp Neurol 1980; 194:555. 6. Nilaver G, Zimmerman EA, Wilkins J, et al. Magnocellular hypothalamic projections to the lower brainstem and spinal cord of the rat. Neuroendocrinology 1980; 30:150. 7. Swaab D, Pool C, Nijveldt F. Immunofluorescence of vasopressin and oxytocin in the rat hypothalamo-neurohypophyseal system. J Neural Transm 1975; 36:195. 8. Silverman AJ, Zimmerman EA. Magnocellular neurosecretory system. Annu Rev Neurosci 1983; 6:357. 9. Vanderhagen JJ, Lotstra F, Vandesand F, Dierickx K. Coexistence of cholecystokinin and oxytocin-neurophysin in some magnocellular hypothalamohypophyseal neurons. Cell Tissue Res 1981; 221:227. 10. Watson SJ, Akil H, Fischli W, et al. Dynorphin and vasopressin: common localization in magnocellular neurons. Science 1982; 216:85. 11. Buijs RA. Intra- and extrahypothalamic vasopressin and oxytocin pathways in the rat: pathways to the limbic system, medulla oblongata and spinal cord. Cell Tissue Res 1978; 192:423. 11a. Jordan J, Shannon JR, Black BK, et al. The pressor response to water drinking in humans: a sympathetic reflex? Circulation 2000; 101:504. 12. Hoffman GH, Phelps CJ, Khachaturian H, Sladek JR Jr. Neuroendocrine projections to the median eminence. In: Pfaff DW, Ganten D, eds. Current topics in neuroendocrinology, vol 7. Morphology of hypothalamus and its connections. New York: Springer-Verlag, 1986:161. 13. Bargmann W. ber der neurosekretorische Vernupfung von Hypothalamus und Neurohypophyse. Z Zellforsch 1949; 34:610. 14. Scharrer E, Scharrer B. Hormones produced by neurosecretory cell. Recent Prog Horm Res 1954; 10:183. 15. Falck B, Hillarp N-A, Thieme G, Torp A. Fluorescence of catecholamines and related compounds condensed with formaldehyde. J Histochem Cytochem 1962; 10:348. 16. Fuxe K, Hkfelt T. The influence of central catecholamine neurons on the hormone secretion from the anterior and posterior pituitary. In: Stutinsky F, ed. Neurosecretion. Berlin: Springer-Verlag, 1967:166. 17. Sternberger L, Hardy P, Cuculis J, Meyer H. The unlabeled antibody enzyme method in immunohistochemistry: preparation and properties of soluble antigen-antibody complex (horseradish peroxidase-antihorseradish peroxidase) and its use in identification of spirochetes. J Histochem Cytochem 1969; 18:315. 18. Hkfelt T, Elde R, Fuxe K, et al. Aminergic and peptidergic pathways in the nervous system with special reference to the hypothalamus. In: Reichlin S, Baldessarini RJ, Martin BJ, eds. The hypothalamus. Research PublicationsAssociation for Research in Nervous and Mental Disease. New York: Raven Press, 1978:69. 19. Gai WP, Geffen LB, Blessing WW. Galanin immunoreactive neurons in the human hypothalamus: colocalization with vasopressin-containing neurons. J Comp Neurol 1990; 298:265. 20. Hoffman GE, Melnyk V, Hayes T, et al. Immunocytology of LHRH neurons. In: Scott DE, Kizlowski GP, Weindl A, eds. Brain-endocrine interactions, vol III. Neural hormones and reproduction. Basel: S Karger, 1978:67. 21. Watson SJ, Khachaturian H, Akil H, et al. Comparison of the distribution of dynorphin systems and enkephalin systems in brain. Science 1982; 218:1134. 21a. Patrick RL. Synaptic clefts are made to be crossed: neurotransmitter signaling in the central nervous system. Toxicol Pathol 2000; 28:31. 22. Lightman SL, Young WS III. Vasopressin, oxytocin, dynorphin, enkephalin and corticotrophin releasing factor mRNA stimulation in the rat. J Physiol (Lond) 1987; 394:23. 23. Meister B, Villar MJ, Ceccatelli S, Hkfelt T. Localization of chemical messengers in magnocellular neurons of the hypothalamic supraoptic and paraventricular nuclei: an immunohistochemical study using experimental manipulations. Neuroscience 1990; 37:603. 24. Meister B. Gene expression and chemical diversity in hypothalamic neuro-secretory neurons. Mol Neurobiol 1993; 7:87. 25. Dahlstrm A, Fuxe E. Evidence for the existence of monoamine-containing neurons in the central nervous system. I. Demonstration of monoamines in the cell bodies of brainstem neurons. Acta Physiol Scand 1965; 62:1. 26. Ungerstedt U. Stereotaxic mapping of the monoamine pathways in the rat brain. Acta Physiol Scand 1971; 367:1. 27. Cheung Y, Sladek JR Jr. Catecholamine distribution in feline hypothalamus. J Comp Neurol 1975; 164:339. 28. Hoffman GE, Felten DL, Sladek JR Jr. Monoamine distribution in primate brain. III. Catecholamine-containing varicosities in the hypothalamus of Macaca mulatta. Am J Anat 1976; 147:501. 29. Sladek JR, Sladek CD. Neurological control of vasopressin release. Fed Proc 1985; 44:66. 30. Sladek JR Jr, McNeill TH. Simultaneous monoamine histofluorescence and neuropeptide immunocytochemistry. IV. Verification of catecholamine-neurophysin interactions through single section analysis. Cell Tissue Res 1980; 210:181. 31. McNeill TH, Sladek JR Jr. Simultaneous monoamine histofluorescence and neuropeptide immunocytochemistry. II. Correlative distribution of cate-cholamine varicosities and magnocellular neurosecretory neurons in the rat supraoptic and paraventricular nuclei. J Comp Neurol 1980; 193:1023. 32. Sladek JR Jr, Zimmerman EA. Simultaneous monoamine histofluorescence and neuropeptide immunocytochemistry. VI. Catecholamine innervation of vasopressin and oxytocin neurons in the rhesus monkey hypothalamus. Brain Res Bull 1983; 9:431. 33. Sawchenko PE, Swanson LW. The organization of noradrenergic pathways from the brainstem to the paraventricular and supraoptic nuclei in the rat. Brain Res Rev 1982; 4:275. 34. McKellar S, Loewy AD. Organization of some brainstem afferents to the paraventricular nucleus of the hypothalamus in the rat. Brain Res 1981; 217:351. 35. Cunningham ET Jr, Sawchenko PE. Reflex control of magnocellular vasopressin and oxytocin secretion. Trends Neurosci 1991; 14:406. 36. Day TA, Sibbald JR. A1 cell group mediates solitary nucleus excitation of supraoptic vasopressin cells. Am J Physiol 1989; 257:R1020. 37. Sladek JR Jr. Central catecholamine pathways to vasopressin neurons. In: Schrier RW, ed. Vasopressin. New York: Raven Press, 1985:343. 38. Blessing WW, Willoughby JO. Inhibiting the rabbit caudal ventrolateral medulla prevents baroreceptor-initiated secretion of vasopressin. J Physiol (Lond) 1985; 367:253. 39. Lightman SL, Todd K, Everitt BJ. Ascending noradrenergic projections from the brainstem: evidence for a major role in the regulation of blood pressure and vasopressin secretion. Exp Brain Res 1984; 55: 145. 40. Day TA, Renaud LP, Sibbald JR. Excitation of supraoptic vasopressin cells by stimulation of the A1 noradrenaline cell group: failure to demonstrate role for established adrenergic or amino acid receptors. Brain Res 1990; 516:91. 41. Bittencourt JC, Benoit R, Sawchenko PE. Distribution and origins of substance P-immunoreactive projections to the paraventricular and supraoptic nuclei: partial overlap with ascending catecholaminergic projections. J Chem Neuroanat 1991; 4:63. 42. Lundberg JM, Pernow J, Lacroix JS. Neuropeptide Y: sympathetic cotransmitter and modulator? News Physiol Sci 1989; 4:13. 43. Day TA, Sibbald JR, Khanna S. ATP mediates an excitatory noradrenergic neuron input to supraoptic vasopressin cells. Brain Res 1993; 607:341. 44. Sibbald JR, Wilson BKJ, Day TA. Neuropeptide Y potentiates excitation of supraoptic neurosecretory cells by noradrenaline. Brain Res 1989;499: 164. 45. Sladek JR Jr, Sladek CD. Anatomical reciprocity between magnocellular peptides and noradrenaline in putative cardiovascular pathways. Prog Brain Res 1983; 60:437. 46. Sladek CD, Armstrong WE. Effect of neurotransmitters and neuropeptides on vasopressin release. In: Gash DM, Boer GJ, eds. Vasopressin. New York: Plenum Publishing, 1987:275. 47. Renaud LP, Bourque CW. Neurophysiology and neuropharmacology of hypothalamic magnocellular neurons secreting vasopressin and oxytocin. Prog Neurobiol 1991; 36:131. 48. Mason WT, Ho YW, Eckenstein F, Hatton GI. Mapping of cholinergic neurons associated with rat supraoptic nucleus: combined immunocytochemical and histochemical identification. Brain Res Bull 1983; 11:617. 49. Kjaer A, Larsen PJ, Knigge U, et al. Histamine stimulates c-fos expression in hypothalamic vasopressin, oxytocin, and corticotropin releasing hormone-containing neurons. Endocrinology 1994; 134:482. 50. Atkins VJ, Bealer SL. Hypothalamic histamine release, neuroendocrine and cardiovascular responses during tuberomammillary nucleus stimulation in the conscious rat. Neuroendocrinology 1993; 57:849. 51. Lindvall O, Bjorklund A, Skagerberg G. Selective histochemical demonstration of dopamine terminal systems in rat di- and telencephalon: new evidence for dopaminergic innervation of hypothalamic neurosecretory nuclei. Brain Res 1984; 306:19. 52. Levey AI. Molecules of the brain. Hosp Pract (Off Ed) 2000; 35(2):41 and 51.

CHAPTER 9 PHYSIOLOGY AND PATHOPHYSIOLOGY OF THE ENDOCRINE BRAIN AND HYPOTHALAMUS Principles and Practice of Endocrinology and Metabolism

CHAPTER 9 PHYSIOLOGY AND PATHOPHYSIOLOGY OF THE ENDOCRINE BRAIN AND HYPOTHALAMUS

PAUL E. COOPER Hypothalamic-Pituitary Unit Embryology Functional Neuroanatomy Hypophysiotropic Hormones Functions of the Nonendocrine Hypothalamus Temperature Regulation Appetite Regulation Emotion and Libido Autonomic Functions Biologic Rhythms Memory Sleep Hypothalamic Syndromes Chapter References

The brain and endocrine system have been linked since the 1940s, when the hypothalamus was first assigned a central role in the control of anterior pituitary secretion.1,2 This chapter discusses the functional neuroanatomy of the hypothalamic-pituitary unit, as well as the important nonendocrine functions of the hypothalamus.

HYPOTHALAMIC-PITUITARY UNIT
EMBRYOLOGY Between the third and fourth weeks of embryonic development, a longitudinal groove, the sulcus limitans, appears in the lumen of the neural tube. This sulcus divides the alar (dorsal) plate from the basal (ventral) plate. The basal plate plays no part in the development of the hypothalamus or pituitary, participating only in the formation of nervous tissue caudal to the diencephalon. At ~5.5 weeks, the alar plate, in the region that gives rise to the diencephalon, develops a longitudinal groove contiguous with the sulcus limitans. This groove, the hypothalamic sulcus, divides the alar plate into a dorsal portion, which gives rise to the thalamus, and a ventral portion, which gives rise to the hypothalamus3 (Fig. 9-1A and Fig. 9-1B).

FIGURE 9-1. A, Cross section of a human embryo (3.5 weeks) through the first cervical somite. B, Cross section of a human embryo (5.5 weeks) in the region of the diencephalon. C, Midsagittal section of the hypothalamus of a human embryo (11 weeks).

During the fifth week of embryonic life, the anterior pituitary begins to form as a diverticulum (Rathke pouch) of the buccal cavity. It expands dorsally to join the basal portion of the forebrain (see Fig. 9-1C) so that by the eighth week, the posterior pituitary, which begins as a diverticulum of the floor of the third ventricle, has contacted the Rathke pouch and been invested by it. By the 11th week, the cavity of the Rathke pouch becomes flattened and loses its connection with the buccal cavity.3 Remnants of its attachment to the buccal cavity form the craniopharyngeal duct, residual cells of which persist in the posterior lobe of the pituitary, the hypophysial stalk, and the basisphenoid. Traditionally, these residual cells are thought to be the origin of craniopharyngioma; however, because craniopharyngiomas occasionally develop in the sella or along the craniopharyngeal duct itself, this idea on the origin of these tumors has been questioned. Careful examination of the superior pharynx in infants and adults almost always reveals a pharyngeal hypophysis at the buccal end of the craniopharyngeal duct. It is composed mainly of undifferentiated epithelial cells with a few chromophobes and chromophils, although rarely this residual pharyngeal hypophysial tissue has been reported to produce excessive pituitary hormone secretion. FUNCTIONAL NEUROANATOMY Neuroanatomically, the traditional borders of the hypothalamus are the lamina terminalis (rostrally); the posterior edge of the mammillary body (caudally); the hypothalamic sulcus (dorsally); the floor of the third ventricle (ventrally); and the internal capsule, basis pedunculi, and subthalamus (laterally). The hypothalamus usually is divided into three regions: the chiasmatic (preoptic region), the tuberal region, and the mammillary complex. Of these three, the first two contain the neuronal groups and tracts that are of most significance in neuroendocrine regulation (Table 9-1).

TABLE 9-1. Hypothalamic Regions and Nuclei of Neuroendocrine Interest

Histologically, the hypothalamus contains two types of neurons: large (magnocellular) and small (parvicellular). Magno-cellular neurons are of two classes: those that secrete arginine vasopressin (AVP) and neurophysin II, and those that secrete oxytocin and neurophysin I. AVP often is colocalized with dynorphin or angiotensin II, whereas oxytocin frequently is colocalized with cholecystokinin (CCK), corticotropin-releasing hormone (CRH), metenkephalin, or proenkephalin. Parvicellular neurons

contain a variety of neuropeptides or biogenic amines, many of which are colocalized. The parvicellular neurons of the hypothalamus that are found in the immediate periventricular region of the third ventricle, from the preoptic area to the mammillary bodies, project to the median eminence, where their neurosecretory products are released to influence anterior pituitary function. Functionally discrete groups of parvicellular neurons have been located in the anterior preoptic region, the paraventricular nucleus, and the arcuate nucleus. The more medial area of the hypothalamus has collections of neurons that receive inputs from the brainstem and the limbic system. They, in turn, project to the periventricular zone, the limbic system, the brainstem, and the spinal cord in a circuit that is suited to the coordination of endocrine, autonomic nervous, and more complex behavioral responses to ensure homeostasis. In the lateral hypothalamic area is found the medial fore-brain bundle, a neural pathway that connects the brainstem reticular formation with the septum and the limbic system and also has input into the periventricular secretory neurons. HYPOPHYSIOTROPIC HORMONES The anterior pituitary is known to produce six peptide hormones (Table 9-2). The secretion of each of these is under the control of one or more hypothalamic neuropeptides and several other classic neurotransmitters. A comprehensive review of these intricate interactions has been published elsewhere.4

TABLE 9-2. Anterior Pituitary and Hypophysiotropic Hormones

In primates, a few neurons containing gonadotropin-releasing hormone (GnRH) are found in clusters above the optic chiasm, medial to the supraoptic nuclei, but most are found in the medial basal hypothalamus.5 Their axons project to the median eminence, where GnRH is released in a pulsatile fashion. In humans, GnRH-positive neurons are found in the arcuate nucleus.4 The GnRH-containing neurons of the medial basal hypothalamus have an intrinsic firing pattern that results in the pulsatile release of GnRH. Norepinephrine, epinephrine, serotonin, acetylcholine, and N-methyl D-aspartic acid all stimulate GnRH release, as do angiotensin II (probably through its action on norepinephrine) and neuropeptide Y.4 Opioid peptides and g-aminobutyric acid (GABA) inhibit GnRH release.4 CRH-containing neurons are found in the medial portion of the paraventricular nucleus. From there, they project to the median eminence, where they release CRH in a pulsatile fashion (see Chap. 8). CRH release is stimulated by norepinephrine, epinephrine, serotonin, glutamine, aspartamine, acetylcholine, angiotensin II, and neuropeptide Y, and is inhibited by GABA, AVP, opioid peptides, and substance P.4 Neurons containing thyrotropin-releasing hormone (TRH) are found in the periventricular portion of the paraventricular nucleus and in a similar location in the anterior hypothalamus. They project to the median eminence through the lateral retrochiasmatic area.5 TRH release is responsive to triiodothyronine (T3) and thyroxine (T4) levels. It also is stimulated by norepi-nephrine, dopamine, and serotonin, and is inhibited by GABA and opioid peptides.4 Prolactin release is inhibited by dopamine and stimulated by TRH and vasoactive intestinal peptide (VIP). Dopamine-containing neurons are found in the arcuate nucleus and the preoptic ventricular nucleus.6 They project into hypothalamic regions rich in TRH and somatostatin. VIP-containing neurons are found in the paraventricular nucleus and project to the median eminence. They also are found in the suprachiasmatic nucleus. Neurons containing growth hormonereleasing hormone are found in the arcuate nucleus, just lateral to the median eminence. Their nerve terminals project to the median eminence and pituitary stalk.7,8 The release of growth hormonereleasing hormone is stimulated by norepinephrine and serotonin.4 Neurons containing somatostatin (growth hormone release inhibiting hormone) are distributed widely in the central nervous system (see Chap. 169). In the hypothalamus, they are localized to neurons in the paraventricular nucleus. Their axons travel laterally and ventrally toward the optic chiasm and then caudally to the median eminence. The arcuate and ventromedial nuclei receive somatostatinergic innervation from other sources. Many somatostatin-containing nerve terminals are found in the ventromedial arcuate complex, the suprachiasmatic nucleus, the ventral premammillary nuclei, and the organum vasculosum of the lamina terminalis.5,9 HYPOTHALAMIC NEURAL CONNECTIONS Hypothalamic afferent pathways carry important information on emotion and visceral function. Ascending visceral afferents convey data to the hypothalamus from baroreceptors, volume receptors, and taste receptors. These inputs reach the hypothalamus over poorly defined pathways that relay in the reticular formation and midline thalamic nuclei. Somatosensory information also reaches the hypothalamus. The limbic system has rich afferent connections to the hypothalamus. The medial fore-brain bundle serves as an important integrative pathway between the brainstem and the limbic system. The amygdala sends fibers to the hypothalamus through the stria terminalis, and the fornices relay information from the hippocampal formation to the mammillary bodies of the hypothalamus. The mammillary bodies also receive input from and provide input to the anterior thalamic nuclei through the mammillothalamic tract and, therefore, are indirectly connected to the cingulate cortex. Finally, afferent hypothalamic connections from the dorsomedial thalamic nucleus and direct corticohypothalamic connections from the orbital surface of the frontal lobe are found.10 Two major hypothalamic efferent pathways are the dorsal longitudinal fasciculus and the mammillotegmental fasciculus. The fibers of the dorsal longitudinal fasciculus terminate in the dorsal motor nucleus of the vagus, the salivatory and lacrimal nuclei, the intermediolateral cell column of the thoracolumbar cord, and the sacral autonomic area. In addition, efferents from this tract go to various brainstem motor nuclei connected with eating and drinking, and to spinal cord motor neurons that participate in the shivering that raises body temperature. The fibers of the mammillotegmental fasciculus end in the raphe nuclei of the midbrain and pons. Finally, the mammillothalamic fasciculus contains two-way connections with the anterior nuclei of the thalamus.10 Through these rich connections, the hypothalamus monitors and influences body functions, preserving the constancy of the internal milieu. SUPRACHIASMATIC NUCLEUS The suprachiasmatic nucleus is responsible for generating circadian rhythms in all mammals, including humans. It is composed of neurons that contain AVP, VIP, neuropeptide Y, and neurotensin. A marked seasonal variation exists in suprachiasmatic nucleus cell numbers and volumes in humans, with summer values being approximately half those found in the fall. Hypothalamic tumors that involve the anterior hypothalamus disturb circadian rhythms in humans.11,12 The volume of the suprachiasmatic nucleus and the number of cells is similar in men and women, although the shape of the nucleus is different.13 Homosexual males have been reported to have a suprachiasmatic nucleus that is 1.7 times larger than that in heterosexual males and contains 2.1 times as many cells.14 The functional implication of this finding, if confirmed, remains to be determined. SEXUALLY DIMORPHIC NUCLEUS The sexually dimorphic nucleus, or intermediate nucleus, is found in the preoptic area. It contains twice as many cells in young adult males as in comparable adult

females.15 Cell numbers in the nucleus are similar in males and females at birth and not until ~4 years of age can any difference be detected. No difference in cell number in this nucleus is observed when the brains of homosexual and heterosexual males are compared.14,16 The third nucleus of the interstitial nuclei of the anterior hypothalamus has been found to be larger in males than in females, and is approximately half as large in homosexual males as in heterosexual males.17 This finding, which has not been confirmed, suggests yet another hypothalamic area that is dimorphic with respect to gender as well as sexual orientation. BLOOD SUPPLY OF THE HYPOTHALAMIC-PITUITARY UNIT The hypothalamus receives its arterial blood supply from the circle of Willis, the internal carotid and posterior cerebral arteries. Apart from blood that drains into the pituitary gland, most blood from the hypothalamus enters the basal vein of Rosenthal through numerous small venous plexuses in and around the hypothalamus. The blood supply of the pituitary gland is more complex (see Chap. 11). The posterior pituitary (neurohypophysis) is supplied by blood directly from the inferior hypophysial artery and drains into the inferior hypophysial veins. At its most rostral portion, in the median eminence, the neurohypophysis is supplied by the superior hypophysial arteries; free communication exists between the blood supply of the median eminence and that of the posterior pituitary. Apart from the most superficial layers of the gland, which are supplied by small capsular arteries, the anterior pituitary gland lacks a direct arterial blood supply. Blood is supplied to the median eminence by the superior hypophysial arteries, and the major venous drainage of the median eminence is through the portal veins to the anterior pituitary. Blood then drains from the anterior pituitary into the posterior pituitary, from which it enters the cavernous sinus or returns to the median eminence. This anatomic configuration means that blood from the median eminence, rich in hypothalamic factors and neurotransmitters, is directed toward the anterior pituitary, where it influences the secretion of pituitary hormones. Whether blood from the anterior pituitary, rich in pituitary hormones, returns to the median eminence to participate in feedback control and other possible effects on the brain is debated.18,19 ANATOMY OF THE PARAPITUITARY AREA Disorders of the hypothalamic-pituitary unit may be seen clinically because of neuroendocrine dysfunction or because of symptoms caused by compression of local structures. To fully understand the symptoms produced by compression, one must understand the anatomy of the parapituitary area (Fig. 9-2).

FIGURE 9-2. Anatomic relationships of the pituitary fossa and cavernous sinus. The lateral wall of the sella turcica is formed by the cavernous sinus. The sinus contains the carotid artery, two branches of the fifth cranial nerve (ophthalmic [Ophth.] and maxillary [Max.]), the third nerve (oculomotor), the fourth nerve (trochlear), and the sixth nerve (abducens). The optic chiasm and optic tract are located superior and lateral, respectively, to the pituitary. (A, artery; N, nerve.) (From Martin JB, Reichlin S, Brown GM, eds. Clinical neuroendocrinology. Philadelphia: FA Davis, 1987:447.)

The pituitary gland rests in the sella turcica. Inferiorly, it is bounded by the sphenoid sinus. Superiorly, it is separated from the cranial cavity by a double layer of dura mater, the diaphragma sellae; through this passes the pituitary stalk. Normally, the dura tightly surrounds the stalk. If the opening is larger or if the intracranial pressure is increased, however, the arachnoid membrane may herniate into the sella, displacing the pituitary gland peripherally and enlarging the sellathe empty sella syndrome (see Chap. 11 and Chap. 17). The diaphragma sellae is pain sensitive; stretching of this tissue by pituitary enlargement may give rise to frontotemporal headaches. The optic chiasm lies 3 to 10 mm above the pituitary fossa. In 90% of persons, the chiasm is partly or completely above the diaphragma. Of the remaining 10% of persons, approximately half have an anteriorly placed chiasm (prefixed) and half have a posteriorly placed chiasm (postfixed).20 This anatomic variability accounts for the lack of visual field abnormalities in some patients with a large suprasellar extension of a pituitary tumor. Compression of the optic chiasm by a pituitary tumor usually affects the crossing fibers in the chiasm. These come from the nasal portions of the retina, which serve the temporal fields. Thus, the typical visual field abnormality that occurs in chiasmatic compression is a bitemporal hemianopsia. Because the most inferiorly placed fibers in the optic chiasm carry information from the superior visual fields, tumors pushing on the chiasm from below tend to cause superior quadrantanopsias of the bitemporal variety, whereas lesions pushing on the chiasm from above cause inferior bitemporal quadrantanopsias. Unfortunately, several exceptions to these rules do occur. Lesions in and around the optic chiasm tend to produce incongruous visual field defects (i.e., of a different configuration in each eye; see Chap. 19). Immediately lateral to the pituitary gland are the cavernous sinuses. Within each is found the carotid artery and its sympathetic plexus; cranial nerves III, IV, and VI; and the ophthalmic and maxillary divisions of the trigeminal nerve. Sudden expansion of a pituitary tumor into the cavernous sinus may produce cranial nerve palsy. Conversely, an aneurysm of the cavernous portion of the carotid artery can erode laterally and mimic the plain skull radiographic appearance of a pituitary tumor and even lead to mild hypopituitarism.

FUNCTIONS OF THE NONENDOCRINE HYPOTHALAMUS

TEMPERATURE REGULATION For the body to function efficiently, its temperature must be maintained within narrow limits. Wide variations beyond this range can result in serious metabolic derangement and death. The hypothalamus ensures that the heat gained by the body from metabolic activity, and in some circumstances from the environment, is balanced by the heat lost.21,22 The preoptic anterior hypothalamus contains thermosensitive neurons that monitor the temperature of blood (Fig. 9-3). Experiments have shown that serotonin (5-hydroxytryptamine) released in this area stimulates hypothalamic heat production centers. This effect is blocked by norepinephrine or epinephrine.

FIGURE 9-3. Hypothalamic temperature regulation mechanisms. The preoptic anterior hypothalamic area functions as a thermostat and contains mechanisms for regulation of heat loss. The posterior hypothalamus integrates heat production mechanisms. Lesions of the preoptic anterior hypothalamic area cause hyperthermia; lesions of the posterior hypothalamus cause hypothermia or poikilothermia. (Modified from Myers RD. Ionic concepts of the set-point for body temperature. In: Lederis K, Cooper KE, eds. International symposium on recent studies of hypothalamic function, Calgary, 1973. Basel: S Karger, 1974:371; and from Cooper PE, Martin JB.

Neuroendocrinologic diseases. In: Rosenberg R, ed. Comprehensive neurology. New York: Raven Press, 1991:608.)

The caudolateral portion of the hypothalamus is insensitive to changes in body temperature, but the regulatory center that determines the normal setpoint of 37C is located here. The injection of acetylcholine-like substances into this region causes profound and long-lasting hypothermia. It also is through this area that the main pathways controlling heat loss and heat conservation travel to the midbrain, pons, medulla, and spinal cord. Intraventricular injection of the neuropeptides mammalian bombesin, neurotensin, TRH, somatostatin, and b-endorphin decreases body temperature and thus implicates these substances in thermoregulation. CRH appears to be an important mediator of thermogenesis in response to serotonin and its agonists and to cytokines. No peptide or group of peptides, however, has been singled out as the physiologic regulator of body temperature.20 In some experiments, prostaglandin E2 has increased body temperature. Hypothalamic injury after head trauma or cerebral infarction can produce prolonged hypothermia resulting either from a change in the setpoint or from an impairment of heat production mechanisms. Paroxysmal hypothermia has been described in a few patients. These persons appear to have a temporary alteration in setpoint, with the body temperature falling to 32C or lower. The hypothermia may last for minutes to days and is associated with fatigue, decreased alertness, hypoventilation, and even cardiac arrhythmia. The loss of body heat is caused by increased sweating and vasodilation. Although the paroxysmal nature of these episodes has suggested an epileptic etiology, the attacks are not prevented by use of anticonvulsants.21 Paroxysmal hyperthermia can occur in some conditions. Acute damage to the preoptic anterior hypothalamus from surgery, subarachnoid hemorrhage, or cerebral infarction can lead to profound impairment of heat loss mechanisms, and the resulting hyperthermia can be lethal. Cyclic hyperthermia has been seen in some patients, but the neuropathologic substrate for this condition is unknown. Some have responded to therapy with phenytoin. Sustained hyperthermia probably is not seen in hypothalamic dysfunction. Reported case studies of prolonged hyperthermia attributed to hypothalamic dysfunction have not excluded an underlying malignancy or unrecognized infection. Cyclic hypothermia, responsive to treatment with anticonvulsant agents, such as clonidine or cyproheptadine, has been described.23 Large lesions in the posterior hypothalamus or lesions in the brainstem that damage the hypothalamic outflow tracts may result in poikilothermia, a condition in which the body temperature varies with environmental temperature. Most affected patients have hypothermia, although in hot, humid conditions, hyperthermia may be a problem.21 During fever, the body's temperature setpoint is elevated, although the ability to regulate temperature around the new setpoint is normal.24 In response to an infection or other cause of inflammation, the body's inflammatory cellsprimarily monocytesrelease cytokines,25 which act at the hypothalamus to cause fever.24 Interleukin-1 (IL-1) releases phospholipases in the hypothalamus that, in turn, release arachidonic acid from plasma membranes. Arachidonic acid causes a rise in prostaglandin E, which raises the body temperature setpoint. Treatment with acetylsalicylic acid or acetaminophen to reduce fever probably affects this process. Animal studies suggest that the action of IL-1 occurs in the preoptic anterior hypothalamus through the reduction of the sensitivity of warm-sensitive neurons, allowing the body to tolerate a higher temperature. Once this new setpoint has been established, the hypothalamus uses normal physiologic mechanisms to maintain body temperature by peripheral vasoconstriction, reduced sweating, and, if necessary, increases in heat production through shivering. Tumor necrosis factor (TNF) is another cytokine that alters the setpoint in the hypothalamus and increases the production of IL-1 locally, in the hypothalamus. Most TNF seems to be produced in macrophages stimulated by bacterial endotoxin. Inter-leukin-6 (IL-6) and interferon-g also act directly at the hypothalamus to raise the setpoint. The exact role of the cytokines in regulating body temperature and in causing fever is unclear because complex interactions exist among these compounds. IL-1, for example, stimulates its own production and interferon-g stimulates IL-1 production, whereas interleukin-4 suppresses the production of IL-1, TNF, and IL-6. IL-1 production also is inhibited by glucocorticoids and prostaglandin E. The fulminant hyperthermia (malignant hyperthermia) that can occur during anesthesia is not hypothalamic in origin. Rather it results from excessive muscle contraction caused by an abnormality of the muscle membrane. Another syndrome of hyperthermia is the neuroleptic malignant syndrome. This condition, characterized clinically by hyperthermia, rigidity of skeletal muscles, autonomic instability, and fluctuating levels of consciousness, has been associated with the use of major tranquilizers, rapid withdrawal from treatment with dopaminergic agents (e.g., L -dopa or bromocriptine), and, less commonly, the use of tricyclic antidepressants. The common denominator in the syndrome seems to be an alteration in dopamine function in the hypothalamus. Treatment consists of discontinuing the use of neuroleptics, providing general support, and administering anticholinergics for mild cases, bromocriptine (5 mg orally or nasogastrically four times daily) for more severe cases, and benzodiazepines or dantrolene (23 mg/kg per day intravenously to a maximum of 10 mg/kg per day) for resistant cases.26,27 APPETITE REGULATION In most animals, the body is able to balance its intake of food and output of energy to maintain body weight. It is the hypothalamus that receives inputs from the periphery that either stimulate or inhibit the intake of food, and it is the hypothalamus that, likewise, sends signals to other parts of the brain to influence endocrine, autonomic, and motor nervous system function.28 The interaction between the limbic system and the hypothalamus is critical in translating the need for food into behaviors such as hunting and stalking. The destruction of both ventromedial nuclei in the rat or the cat markedly increases food intake for a few days. As the animal becomes obese, the overeating decreases; if it is then fasted back to ideal weight and given free access to food, the animal again increases its food intake until it becomes obese. Lesions of the ventromedial nuclei, however, do not produce a pure syndrome of obesity. During the phase of overeating, the animals often are irritable and aggressive, becoming lethargic and passive when the increased weight is achieved. The pituitary gland is not necessary for the weight gain because hypophysectomized animals also become obese. If hyperphagia is prevented by tube feeding, the animals still become obese. Hyperinsulinemia has been observed in lesioned animals within minutes of surgery; if this is prevented by lesioning the pancreatic B cells with streptozocin, the obesity and hyperphagia are prevented. Classic neurophysiology explained the hyperphagia in animals with a lesion of the ventromedial nuclei on the basis of unopposed activity of a hypothalamic feeding center in the lateral hypothalamus. Implantation of electrodes in this lateral area causes, after stimulation, marked increases in food intake, whereas destruction of this area causes aphagia, even in animals with concomitant ventromedial nuclei lesions. The limbic system plays an important role in appetite. One suggested interaction is that the urge to eat arises in the hypothalamus and the limbic structures modify food intake through a discriminative function. Bilateral lesions in the amygdala cause prolonged aphagia and adipsia. Stimulation of this same area in the fed animal does not cause increased eating, but intake increases if the stimulation is performed while the animal is eating. Cells in the ventromedial hypothalamus also monitor blood glucose levels and coordinate the hypothalamic response to hypoglycemia. Although hypothalamic lesion studies have shown that blood glucose is decreased on stimulation of the anterior and tuberal regions medially, acute hypothalamic damage tends to produce hyperglycemia by activation of the sympathetic nervous system and a resulting glycogenolysis in the liver. Another important effect is the elevation of growth hormone, a contrainsulin hormone, which causes glucose levels to rise. In patients deficient in growth hormone, on either a hypothalamic or a pituitary basis, profound hypoglycemia can occur and may even be a presenting symptom, especially in children. Many of the features of lateral hypothalamic lesions in animals have been thought to result from damage to the nigrostriatal bundle, a dopamine-containing tract connecting the substantia nigra to the basal ganglia. Lesions in this tract cause anorexia and weight loss, whereas lesions in the ventral adrenergic bundle, a norepinephrine-containing tract originating in the locus coeruleus, cause hyperphagia and obesity. Serotonin is thought to inhibit eating, whereas GABA agonists have the opposite effect. Of interest has been the putative role of neuropeptides in the regulation of appetite. The gastrointestinal peptides CCK, bombesin, and glucagon can inhibit feeding behavior through an action on the vagal nucleus. Neuropeptide Y can cause hypothalamic obesity by inhibiting sympathetic drive and stimulating insulin release.29 An important putative stimulator of appetite is b-endorphin. One potential mechanism for this action is its ability to stimulate the release of insulin. Other neurotransmitters have been assumed to play a role in the control of appetite. Experimentally, the following substances decrease appetite: norepinephrine, serotonin, CCK, neurotensin, bombesin, TRH, naloxone, somatostatin, and VIP. The following substances increase appetite: dopamine, GABA, b-endorphin, enkephalin, and neuropeptide Y.30 (Also see Chap. 125.)

Increased oxidation of fatty acids raises levels of 3-hydroxy-butyrate, which can reduce food intake through an action at the level of the hypothalamus. The ob, db, and fa genes have all been implicated in the regulation of feeding in animals through production of a circulating factor. In the case of the ob gene, this circulating factor is leptin. Genetically obese mice have a leptin deficiency that, when corrected, leads to a reduction in food intake with a resulting fall in body weight. Leptin levels have been found to be high in obese humans. This suggests an insensitivity to endogenous leptin in these individuals.31 A discussion of the role of other peptides such as uncoupling proteins, agouti protein, melanocortin receptor isoforms, melanin-concentrating hormone, and the proteins responsible for the tub and fat monogenic mouse models of obesity is beyond the scope of this chapter but has been reviewed.32 Hypothalamic tumors that cause hyperphagia, aggressive behavior, and the development of marked obesity have been described. Similar clinical syndromes can be seen in patients with hypothalamic damage caused by radiation therapy or encephalitis. For many years, the syndromes of anorexia nervosa and bulimia nervosa have been considered by many to be purely psychiatric; however, the finding of reduced serotonin levels in the cerebrospinal fluid of patients with bulimia nervosa, the low cerebrospinal fluid levels of norepinephrine in patients with anorexia nervosa, and the enhanced secretion of CCK-8-S in patients with anorexia nervosa suggest that neurotransmitter or neuropeptide abnormalities could be responsible for at least some of the clinical features of these syndromes.33 The question remains whether these changes are a result of the condition or its cause (see Chap. 128). EMOTION AND LIBIDO The anatomic substrate for emotion is widespread and not confined to any single area of the brain; the frontal and temporal lobes, the limbic system, and the hypothalamus all participate in emotion. The hypothalamus is thought to play an important role in the integration and expression of emotion, especially sexual and aggressive behaviors.34 The hypothalamus is necessary for angry behavior in cats. If all cerebral tissue rostral to the tuberal region is removed, angry behavior still can be induced by minor stimuli, but if the remaining hypothalamus then is removed, this activity no longer can be provoked. Electrical stimulation of the caudal hypothalamus in the cat can elicit rage reactions. Bilateral anteromedial hypothalamic lesions cause normally friendly cats to become aggressive, as do bilateral lesions in the ventromedial nuclei, or electrical stimulation in the perifornical region or the periaqueductal gray area between the third and fourth ventricles. The limbic system appears to exert a tonic inhibition on the perifornical region of the hypothalamus. Damage to these inhibitory pathways or stimulation in this area results in angry behavior. Ablation of certain areas in the hypothalamus produces fearful behavior, and stimulation of certain hypothalamic areas in the dorsal hippocampal formation and the septal region produces pleasure reactions. For such reactions to occur, the basal telencephalon and thalamus must be intact. Some humans report a pleasurable or glowing feeling with electrical stimulation in the septal area; others report a feeling of sexual gratification. Primates from which the amygdala, piriform cortex, and part of the hippocampal formation have been removed bilaterally exhibit several behavioral disturbances, including hyper-sexuality, loss of discrimination of taste for thirst-quenching liquids, hyperoral behavior, and loss of awareness of harmful or painful objects (Klver-Bucy syndrome). Rage attacks have occurred in patients with lesions in the caudal hypothalamus, the subthalamus, and even the midbrain. They also have been induced by manipulation of the hypothalamus during surgery. Fear and rage in patients with hypothalamic disorders usually occur in situations that normally would be associated with these same emotions, but in lesser degrees. In those cases examined pathologically, lesions of the basal portion of the brain, especially the descending pathways to the hypothalamus from the cerebral cortex, or the ventromedial nuclei, usually are found. Stimulation of the posterior portion of the hypothalamus in humans can arouse feelings of fear and horror. Lesions of the mammillary bodies and vicinity are associated with drowsiness, somnolence, apathy, and indifference. Psychosurgery directed at the medial posterior hypothalamus or the caudolateral region has caused apathy in previously aggressive persons. Euphoria rarely is seen in adults with hypothalamic disease but is seen in children as part of the diencephalic syndrome. This syndrome most commonly manifests in infancy as failure to thrive; a glioma of the anterior hypothalamus usually is found. Children with this condition are emaciated despite normal or excessive food intake, and are described as jovial and as having excessive energy. Approximately 50% of affected children have nystagmus. A tendency to hypoglycemia also may bring these children to medical attention. Despite growth failure, growth hormone levels usually are elevated. Computed tomography demonstrates abnormalities in most cases (see Chap. 18). Normal libido is the product of an interaction between hypothalamic and extrahypothalamic sites. Usually, hypothalamic damage leads to decreased GnRH levels and reduced libido. Hypersexuality is a rare accompaniment of hypothalamic disease and may be seen with or without increased libido.35 Paroxysmal hypersexuality consisting of sexual urges, genital sensations, and orgasm has been observed in association with temporal lobe tumors or epilepsy but not with primary hypothalamic disease. Altered sexual preference has been described in association with hypothalamic lesions.36 Epileptiform activity in the temporal lobe has been linked to violent behavior. This is extremely rare, however. In a review of 5400 patients with epilepsy, violent behavior during a seizure was found in only 19 cases.37 Evidence is accumulating to suggest that the hypothalamus plays a role in the symptoms of fibromyalgia and chronic fatigue syndrome.38 AUTONOMIC FUNCTIONS The hypothalamus plays an important role in the integration of the functions of the autonomic nervous system.39 CARDIOVASCULAR FUNCTIONS Signals of cardiovascular status are sent to the brain from baroreceptors and chemoreceptors at the carotid sinus and aortic arch, and from pressure/volume receptors in the atrium. This information is fed into the nucleus of the tractus solitarius, as well as into cells of the paramedian reticular nuclei in the medulla. In the medulla, these afferent signals are modulated by inputs from higher centers, particularly the hypothalamus. The main sympathetic outflow to the heart begins in the paraventricular nucleus of the hypothalamus. Some of the paraventricular neurons project to the intermediolateral cell column of the spinal cord (the site of origin of preganglionic sympathetic neurons), whereas others project to the dorsal motor nucleus of the vagus where they can influence parasympathetic output to the heart.40 Stimulation of the anterior hypothalamus, particularly the preoptic area, causes bradycardia, hypotension, and decreased baroreflex activity. Such stimulation also can lower the threshold for ventricular fibrillation and cause a variety of electrocardiographic changes. Such changes are common in patients with subarachnoid hemorrhage. Hypothalamic disease also may cause hypertension, and the disrupted autonomic function that occurs in subarachnoid hemorrhage can cause myocardial necrosis, an effect blocked by sympathetic blockade. The hypothalamus receives input from the nucleus of the tractus solitarius and relays information from higher cortical centers to brainstem vasomotor centers. The many and varied cardiovascular phenomena that occur in response to emotion, especially hypertension and cardiac arrhythmias, probably are mediated by the hypothalamus.40 The anterior portion of the ventral third ventricle has been shown, by lesion studies, to be a site at which body fluid homeostasis and arterial blood pressure are regulated. Angiotensin II levels reflect changes in serum osmolality or sodium concentration. In addition, the area receives sensory neural input from the kidney, carotid sinus, and other baroreceptors. Output from this area can induce vasoconstriction (through the sympathetic nervous system), and abnormal function of this region has been implicated in the pathogenesis of hypertension.41

RESPIRATION Acute pulmonary edema has been described in association with numerous injuries to the central nervous system, such as increased intracranial pressure, epilepsy, and hypothalamic injury.41a Classically, neurogenic pulmonary edema was defined as normal pulmonary capillary wedge pressures with increased protein content of edema fluid in the lung. One hypothesis has been that noncardiac pulmonary edema can be caused by excessive sympathetic outflow from the hypothalamus. Virtually all cases of this type of pulmonary edema, however, are accompanied by system hypertension; therefore, when the experimental physiologic data are reviewed, separation of cardiac from noncardiac causes of edema is impossible. Most of the data support centrally induced systemic hypertension as the cause of the syndrome.42 GASTROINTESTINAL FUNCTIONS Under normal resting conditions, gastric motility is relatively autonomous. Higher cortical centers, including the neocortex and limbic lobe, may influence gastric activity through their connections with the hypothalamus. Afferent signals go to the gut through the vagus nerves. Electrical stimulation of areas of the rostral hypothalamus causes a prompt fall in gastric pH, an effect that is prevented by vagotomy. Stimulation of the tuberal and caudolateral areas of the hypothalamus causes reduced secretion of gastric acid that is of slower onset. This is unaffected by vagotomy but is prevented by bilateral adrenalectomy. Stimulation and ablation studies in the monkey have shown that cortical areas affect peristalsis, the volume of gastric secretion, and its enzyme and acid content. Similar effects can be seen with stimulation or lesioning of the amygdala. Bilateral lesions in the anterolateral hypothalamus increase basal gastric acid secretion, an effect that is abolished by interruption of the fibers of the fornix and medial forebrain bundle. This observation suggests that the limbic lobe and frontal neocortex normally exert a tonic, inhibitory effect on basal gastric secretion. Prolonged hypothalamic stimulation often results in hemorrhage and ulceration in the gastric mucosa of monkeys and dogs, whereas sympathectomy prevents the hemorrhage but not the ulceration. Lesions anywhere from the rostral hypothalamus to the region of the vagal nuclei in the medulla oblongata may cause acute hemorrhagic erosions of the gastric mucosa; extensive ulceration of the lower esophagus and stomach; and acute perforation of the esophagus, stomach, and duodenum. Such lesions are most likely to occur after damage to or pressure on the hypothalamus, particularly its tuberal region. BIOLOGIC RHYTHMS Biologic rhythms are ubiquitous in the animal and plant kingdoms, occurring in single cells, tissues, organs, individual animals, and populations (see Chap. 6). Their periods range from milliseconds to years. Many endocrine rhythms are circadian, having a period ~24 hours long.43 Ultradian rhythms, with periods shorter than 24 hours, and infradian rhythms, with periods longer than 24 hours, certainly exist, but the neural mechanisms responsible for these rhythms are debated. Circadian rhythms have been studied extensively. Integrity of the retinohypothalamic projection, which terminates in the suprachiasmatic nucleus, appears to be essential for the light entrainment of circadian rhythms, and the suprachiasmatic nucleus appears essential for circadian rhythmicity. Ablation of the suprachiasmatic nucleus is associated with the loss of all circadian rhythms. The exact functioning of the suprachiasmatic nucleus in the generation of circadian rhythms is unknown, but at least two coupled oscillators appear to exist. These oscillators cause the circadian variations seen in brain monoamines, plasma amino acids, corticotropin and cortisol, growth hormone, prolactin, vasopressin, aldosterone, insulin, glucose, and sex steroids, as well as pineal activity, body temperature, and autonomic function.44 MEMORY The hypothalamus is closely linked to nervous structures associated with memory function: the reticular formation, hippocampal formation, and other limbic structures.45,46 Although lesions in the mammillary bodies are found in virtually all cases of Wernicke-Korsakoff syndrome (retrograde amnesia, confabulation, apathy),47 the memory deficit in this condition has been thought to correlate best with lesions in the dorsomedial nucleus of the thalamus. Nonetheless, a case report of a brain-injured patient who was studied using magnetic resonance imaging suggests that, at least in trauma, hypothalamic injury alone, without accompanying thalamic injury, can cause marked, relatively focal, memory deficits.48 Many patients with hypothalamic lesions exhibit disturbances of short-term memory, with sparing of immediate recall and long-term memory. Bilateral hippocampal lesions are associated with severe memory disturbance, although some controversy exists about whether involvement of the fornices, the major outflow tract of the hippocampal formation, produces permanent memory loss. Evidence suggests that fornix transection can cause wide-ranging memory disturbance in humans (see Chap. 176).49 SLEEP The region of the hypothalamic-midbrain junction appears to play an important role in sleep and wakefulness. Two types of patients were seen in the encephalitis pandemics50 of 1917 and 1920: those with prolonged somnolence and ophthalmoplegia, and those with agitation and hyperkinesia. Pathologically, somnolence correlated with damage to the tegmentum of the midbrain, and agitation was seen in patients with anterior hypothalamic lesions. Damage to the hypothalamic-midbrain junction by multiple sclerosis, abscess, or infarction has been described as causing hypersomnia or inversion of the sleep-waking cycles. Some studies support the concept of a sleep center in the anterior hypothalamus that, if damaged, causes hyperactivity and sleeplessness, whereas the posterior hypothalamus appears to be involved in the production of rapid eye movement sleep. The ascending reticular formation participates in normal wakefulness. Damage to it results in coma (complete unresponsiveness). Alternatively, the hypothalamus seems to be the generator of normal sleep-wake cycles. Damage to the hypothalamus causes either excessive wakefulness or somnolence (i.e., an unresponsive state from which the patient can be aroused, at least temporarily).

HYPOTHALAMIC SYNDROMES
Several clinical syndromes have been attributed to hypothalamic dysfunction.51 Previously, patients have been described who had autonomic overactivity associated with tumors in the region of the hypothalamus and third ventricle. This was termed diencephalic epilepsy, although its epileptic nature is in doubt because electroencephalograms do not show seizure activity during spells and the condition does not respond to administration of anticonvulsants. Glioma of the anterior hypothalamus in early childhood produces a clinical syndrome characterized by profound emaciation despite normal or excessive food intake, excess energy, and euphoriathe diencephalic syndrome of infancy (see Chap. 18). Kleine-Levin syndrome is characterized by episodes of somnolence followed by hyperactivity, irritability, and increased appetite. Adolescent boys are most commonly affected. The attacks occur every 3 to 6 months and last days to weeks. The cause is unknown, and the specific hypothalamic pathology has not been determined.52 CHAPTER REFERENCES
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. Fortier C. Obituary: Geoffrey Wingfield Harris. Endocrinology 1972; 90:851. Bajusz E, Ernst A. Scharrer. Editorial. Neuroendocrinology 1965; 1:65. Hamilton WJ, Boyd JD, Mossman HW. Human embryology: prenatal development of form and function, 3rd ed. Cambridge: W Heffer & Sons, 1962:315. Palkovits M. Neuropeptides in the brain. In: Martini L, Ganong WF, eds. Frontiers in neuroendocrinology, vol 10. New York: Raven, 1988:1. Page RB. Neuropharmacology of anterior pituitary control. In: Barrow DL, Selman WR, eds. Neuroendocrinology, vol 5. Concepts in neurosurgery. Baltimore: Williams & Wilkins, 1992:31. Mller EE, NisticD G. Brain messengers and the pituitary. San Diego, CA: Academic Press, 1989:1. Jacobowitz DM, Schulte H, Chrousos GP, Loriaux DL. Localization of GRF-like immunoreactive neurons in rat brain. Peptides 1984; 4:521. Merchenthaler I, Thomas CR, Arimura A. Immunocytochemical localization of growth hormone releasing factor (GHRF)-containing structures in the rat brain using anti-rat GHRF serum. Peptides 1984; 5:1071. Beal MF, Mazurek MF, Martin JB. A comparison of somatostatin and neuropeptide Y distribution in monkey brain. Brain Res 1987; 10:405. Barr M, Kiernan JA. The human nervous system: an anatomical viewpoint, 6th ed. Philadelphia: JB Lippincott, 1993:181. Schwartz WJ, Bosis NA, Hedley-Whyte ET. A discrete lesion of ventral hypothalamus and optic chiasm that disturbed the daily temperature rhythm. J Neurol 1986; 233:1. Cohen RA, Albers HE. Disruption of human circadian and cognitive regulation following a discrete hypothalamic lesion: a case study. Neurology 1991; 41:726. Swaab DF, Hofman MA, Lucassen PJ, et al. Functional neuroanatomy and neuropathology of the human hypothalamus. Anat Embryol (Berl) 1993; 187:317. Swaab DF, Fliers E, Partiman TS. The suprachiasmatic nucleus of the human brain in relation to sex, age and senile dementia. Brain Res 1985; 342:37.

15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41.

Swaab DF, Hofman MA. An enlarged suprachiasmatic nucleus of the human brain in homosexual men. Brain Res 1990; 537:141. Swaab DF, Hofman MA. Sexual differentiation of the human hypothalamus: ontogeny of the sexually dimorphic nucleus of the preoptic area. Dev Brain Res 1988; 44:314. LeVay S. A difference in hypothalamic structure between heterosexual and homosexual men. Science 1991; 253:1034. Bergland RM, Page RB. Can the pituitary secrete directly to the brain? (Affirmative anatomical evidence.) Endocrinology 1978; 102:1325. Page RB. Pituitary blood flow. Am J Physiol 1982; 243:E427. Banna M. Terminology, embryology and anatomy. In: Hankinson J, Banna M, eds. Major problems in neurology, vol 6. Pituitary and parapituitary tumors. London: WB Saunders, 1976:1. Cooper PE. Disorders of the hypothalamus and pituitary gland. In: Joynt RJ, ed. Clinical neurology, vol 3. Philadelphia: JB Lippincott, 1993:1. Rothwell NJ. CNS regulation of thermogenesis. Crit Rev Neurobiol 1994; 8:1. Kloos RT. Spontaneous periodic hypothermia. Medicine 1995; 74:268. Cooper KE. The neurobiology of fever: thoughts on recent developments. Annu Rev Neurosci 1987; 10:297. Reichlin S. Neuroendocrine-immune interactions. N Engl J Med 1993; 329:1246. Rosenberg MR, Green M. Neuroleptic malignant syndromereview of response to therapy. Arch Intern Med 1989; 149:1927. Gratz SS, Levinson DF, Simpson GM. The treatment and management of neuroleptic malignant syndrome. Prog Neuropsychopharmacol Biol Psychiatry 1992; 16:425. Bray GA, Fisler J, York DA. Neuroendocrine control of the development of obesity: understanding gained from studies of experimental animal models. Front Neuroendocrinol 1990; 11:128. York DA. Lessons from animal models of obesity. Endocrinol Metab Clin North Am 1996; 25:781. Cezayirli RC, Robertson JT. Pharmacologic modulation of hypothalamic control of appetite. In: Givens JR, ed. The hypothalamus. Chicago: Year Book Medical Publishers, 1984:115. Considine RV, Sinha MK, Heiman ML, et al. Serum immunoreactive-leptin concentrations in normal-weight and obese humans. New Engl J Med 1996; 334:292. Bessesen DH, Faggioni R. Recently identified peptides involved in the regulation of body weight. Semin Oncol 1998; 25(Suppl 6):28. Leibowitz SF. Brain monoamine projections and receptor systems in relation to food intake, diet preference, meal patterns, and body weight. In: Brown GM, Koslow SH, Reichlin S, eds. Neuroendocrinology and psychiatric disorders. New York: Raven Press, 1983:383. Van de Poll NE, Van Goozen SH. Hypothalamic involvement in sexuality and hostility: comparative psychological aspects. Prog Brain Res 1992; 93:343. Plum F, VanUitert R. Nonendocrine diseases and disorders of the hypothalamus. In: Reichlin S, Baldessarini RJ, Martin JB, eds. The hypothalamus. New York: Raven Press, 1978:415. Miller BL, Cummings JL, McIntyre H, et al. Hypersexuality or altered sexual preference following brain injury. J Neurol Neurosurg Psychiatry 1986; 49:867. Delgado-Escuetta AV, Mattson RH, King L, et al. The nature of aggression during epileptic seizures. N Engl J Med 1981; 305:711. Crofford LJ, Demitrack MA. Evidence that abnormalities of central neurohormonal systems are key to understanding fibromyalgia and chronic fatigue syndrome. Rheum Dis Clin North Am 1996; 22:267. Grossman A, ed. Neuroendocrinology of stress. Baillieres' clinical endocrinology and metabolism, vol 1. London: Baillieres Tindall, 1987:247. Valeriano J, Elson J. Electrocardiographic changes in central nervous system disease. Neurol Clin 1993; 11:257. Samuels MA. Neurally induced cardiac damage. Neurol Clin 1993; 11:273.

41a. Keegan MT, Lanier WL. Pulmonary edema after resection of a fourth ventricle tumor: possible evidence for a medulla-mediated mechanism. Mayo Clin Proc 1999; 74:264. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. Simon RP. Neurogenic pulmonary edema. Neurol Clin 1993; 11:309. Aschoff J. Circadian rhythms: general features and endocrinological aspects. In: Kreiger DT, ed. Endocrine rhythms. New York: Raven Press, 1979:1. Moore RY. The anatomy of central neural mechanisms regulating endocrine rhythms. In: Kreiger DT, ed. Endocrine rhythms. New York: Raven Press, 1979:63. Zola-Morgan S, Squire LR. Neuroanatomy of memory. Annu Rev Neurosci 1993; 16:547. Wilson MA, McNaughton BL. Reactivation of hippocampal ensemble memories during sleep. Science 1994; 265:676. Pitella JE, Giannetti AV. Morphometric study of the neurons in the medial mammillary nucleus in acute and chronic Wernicke's encephalopathy. Clin Neuropathol 1994; 13:26. Dusoir H, Kapur N, Byrnes DP, et al. The role of diencephalic pathology in human memory disorder: evidence from a penetrating paranasal brain injury. Brain 1990; 113:1695. Gaffan EA, Gaffan D, Hodges JR. Amnesia following damage to the left fornix and to other sites: a comparative study. Brain 1991; 114:1297. Creisler CA, Klerman EB. Circadian and sleep-dependent regulation of hormone release in humans. Recent Prog Horm Res 1999; 54:97. Martin JB. Neurologic manifestations of hypothalamic disease. Prog Brain Res 1992; 93:31. Mukaddes NM, Alyanak B, Kora ME, Polvan O. The psychiatric symptomatology in Kleine-Levin syndrome. Child Psychiatry Hum Dev 1999; 29:253.

CHAPTER 10 PINEAL GLAND Principles and Practice of Endocrinology and Metabolism

CHAPTER 10 PINEAL GLAND

RUSSEL J. REITER Morphology and Innervation Gross Anatomy Innervation Histology Indoleamine Metabolism Melatonin Serotonin Influence of Hormones and Drugs on Melatonin Secretion Plasma Levels of Melatonin and Its Metabolism Factors Influencing Plasma Melatonin Rhythm LightDark Cycle Age and Sexual Maturation Aversive Stimuli Possible Sex Steroid Feedback Clinical Implications Sexual Maturation Ovulatory Relationships Psychological Interrelationships Sleep Jet Lag and other Biologic Rhythm Disorders Tumor Growth Free Radical Scavenging Conclusion Chapter References

MORPHOLOGY AND INNERVATION

GROSS ANATOMY The pineal gland, so called because of its resemblance to a pine cone (Latin pineas), is a multifaceted endocrine organ whose secretory products directly or indirectly affect every organ and cell in the body. Embryologically, it is derived from neuroectoderm as an outgrowth of the diencephalon. In adults, it is attached to the posterodorsal aspect of the diencephalon, and its proximal portion is invaginated by the pineal recess of the third ventricle (Fig. 10-1). The gland weighs ~130 mg, with great individual variation seen, and is ~1 cm long. The blood supply to the gland is derived from the posterior choroidal branches of the posterior cerebral arteries. The pineal gland has a copious blood supply. In advanced age, it may acquire calcium deposits (corpora arenacea, acervuli), which are visible on skull radiographs.

FIGURE 10-1. Midsagittal view of the human brain showing the relationship of the pineal gland to other neural structures.

INNERVATION The pineal gland is innervated by sympathetic axons and by axons coming directly from the brain.1 Postganglionic sympathetic axons arrive in the pineal from perikarya located in the superior cervical ganglia. The cell bodies of the preganglionic fibers that end in the ganglia are located in the intermediolateral cell column of the upper thoracic cord. The neurons in the intermediolateral cell columns receive terminals from, among other sources, perikarya in the hypothalamus, possibly in the paraventricular nuclei; these nuclei are connected to the supra-chiasmatic nuclei, which receive a prominent input from the ganglion cells of the retinas through the retinohypothalamic tract. Through this complex series of neurons, the retinas are functionally related to the pineal gland. Sympathetic innervation of the pineal is essential for the rhythmic metabolism of indoleamines and for the pineal's endocrine functions. The sympathetic neurotransmitter mediating the cyclic production of pineal indoleamines is norepinephrine. Besides sympathetic innervation, the pineal gland receives axons directly from the brain that enter through the stalk. The function of these fibers relative to the physiology of the pineal is unknown. HISTOLOGY The major cellular elements within the pineal gland are the pinealocytes, which are arranged into cords or follicles separated by connective tissue septa. With advancing age, the septa become more prominent. Ultrastructurally, the pinealocytes have one to several cytoplasmic processes that typically terminate near the many capillaries perfusing the gland. The nerve endings in the pineal gland usually lie close to the pinealocyte processes in the perivascular spaces, but true morphologic synapses are not obvious.

INDOLEAMINE METABOLISM
MELATONIN Within the pinealocyte, tryptophan, an amino acid taken up from the blood, is metabolized to a variety of potential hormones (Fig. 10-2).2 The compound that has been most thoroughly investigated is N-acetyl-5-methoxytryptamine, or melatonin. Tryptophan is initially metabolized to serotonin (5-hydroxytryptamine); the concentration of this monoamine in the pineal gland exceeds that in any other organ. Especially during the daily dark period, serotonin is N-acetylated by the enzyme serotonin N-acetyltransferase to form N-acetylserotonin.3 The activity of N-acetyltransferase increases many-fold during darkness (Fig. 10-3). N-acetylserotonin is converted to melatonin by the action of hydroxyindole-O-methyltransferase. Once formed, melatonin is rapidly secreted into the blood vascular system and, as a consequence, plasma melatonin levels typically are highest at night, when pineal melatonin production is greatest.

FIGURE 10-2. Interactions of postganglionic sympathetic neurons with the endocrine unit of the pineal gland, the pinealocyte, and the conversion of tryptophan to a variety of indole products. Within the pinealocyte, cyclic adenosine monophosphate (AMP) acts as a second messenger to stimulate the activity of serotonin N-acetyltransferase (NAT), primarily during darkness. Serotonin also is acted on by hydroxyindole-O-methyl-transferase (HIOMT) to form 5-methoxytryptamine, and by monoamine oxidase to form 5-hydroxyindole-3-acetylaldehyde. Melatonin is a major secretory product of the pineal gland. (Modified from Reiter RJ, MacLeod RM, Mller EE, eds. Neuroendocrine perspectives, vol 3. New York: Elsevier, 1984:350.)

FIGURE 10-3. Pineal, blood, and urinary rhythms of various constituents. Methods are available for estimating blood and urinary levels of melatonin and its metabolites. The shaded area in the right half of the figure represents the daily dark period. (Ac Co A, Acetyl coenzyme A; CoA, coenzyme A; S Ad M, S-adenosyl methionine; S Ad H, S-adenosyl homocysteine; CNS, central nervous system.) (From Reiter RJ. Pineal indoles: production, secretion and actions. In: Muller CE, MacLeod RM, eds. Neuroendocrine perspectives, vol 13. Amsterdam: Elsevier, 1984:358.).

SEROTONIN Hydroxyindole-O-methyltransferase also acts directly on serotonin, converting it to 5-methoxytryptamine (see Fig. 10-2), a compound that has been proposed as a pineal hormone. Serotonin also is metabolized by monoamine oxidase; the product is eventually converted to 5-methoxytryptophol and 5-methoxyindoleacetic acid, the former of which possibly is released from the pineal.

INFLUENCE OF HORMONES AND DRUGS ON MELATONIN SECRETION

Most of what is known about the interactions of the postganglionic neurotransmitter norepinephrine with the pinealocyte comes from observations in animals, although the relationships probably are similar in humans. Norepinephrine is released into the synaptic cleft, primarily during the daily dark period, after which it acts primarily on b-adrenergic receptors on the pinealocyte membrane, where it stimulates cyclic adenosine monophosphate production and, eventually, N-acetyltrans-ferase activity and melatonin synthesis (see Fig. 10-2).4 Alpha receptors on the pinealocyte membrane also may be involved in mediating the nocturnal rise in melatonin. In humans, the peripheral infusion of isoproterenol, a b-receptor agonist, does not increase blood levels of melatonin as readily as it does in some animals. In addition, orciprenaline and L -dopa are ineffective in altering circulating melatonin in humans. The inability of isoproterenol to stimulate human pineal melatonin production may relate to the relative insensitivity of the pinealocyte b-receptors to the agonist throughout most of the day; animal studies have shown that either the number of receptors or their affinity for the ligand is greatly increased at night.5 Circulating norepinephrine may be relatively ineffective in promoting pineal melatonin production because the sympathetic nerve endings in the pineal gland have an active uptake mechanism for the catecholamine, thereby protecting the pinealocytes from circulating norepinephrine. Other drugs that seem incapable of altering blood melatonin levels in humans include scopolamine, amphetamine, thyrotropin-releasing hormone, luteinizing hormonereleasing hormone, and desaminocys-D-arg-vasopressin.5 Both clonidine and dexamethasone have been reported to lower plasma levels of melatonin. In general, pineal melatonin production seems to be out of the usual endocrine feedback loop.

PLASMA LEVELS OF MELATONIN AND ITS METABOLISM

Plasma levels of melatonin seem to reflect closely the amount being synthesized and secreted by the pineal gland (see Fig. 10-3).6 When pineal melatonin production increases at night, plasma levels rise accordingly. In the few individuals in whom the pineal has been surgically removed because of tumors in the epithalamic region, plasma levels of melatonin either are severely depressed or are undetectable.7 In animals, pinealectomy is associated with very low or nonmeasurable amounts of melatonin in plasma. The nocturnal rise in plasma melatonin levels has been measured by bioassay, radioimmunoassay, and gas chromatographymass spectrometry.7a Depending on the technique used, daytime levels of plasma melatonin range from undetectable to 20 pg/mL. At night, values may increase to 300 pg/mL; however, with the most specific methods, nighttime levels typically are <50 pg/mL. Short-term rises may be superimposed on the 24-hour cycle of melatonin production because of the episodic secretion of melatonin by the pineal gland. Although the amplitude of the circadian melatonin varies widely between individuals, the rhythm is highly reproducible within each person. Most melatonin in the plasma may be bound to proteins, especially albumin. Melatonin readily penetrates cell membranes and tissue barriers (e.g., the bloodbrain barrier). Some of its primary sites of action are undoubtedly in the central nervous system (see ref. 7b). Because of its rapid metabolism, usually <1% of exogenously administered melatonin escapes into the urine in unmetabolized form. Melatonin is enzymatically degraded in at least two sites, the liver and the brain (see Fig. 10-3).2 In animals, melatonin is rapidly metabolized in the liver. Humans with hepatic cirrhosis reportedly have higher than normal daytime levels of plasma melatonin, which implicates the liver as the primary site of its metabolism. In addition, the half-life of plasma melatonin is longer in humans with impaired liver function than in those with normal liver function. In the liver, melatonin is chiefly metabolized to 6-hydroxymelatonin; this is conjugated to sulfuric acid and, to a much lesser extent, to glucuronic acid. A small amount of melatonin is converted to N-acetylserotonin in the liver and appears in the blood as a sulfate or glucuronide conjugate. Melatonin taken up by the central nervous system is converted in part to N-acetyl-5-methoxykynurenamine, which, along with the hepatic metabolites of melatonin, is excreted in the urine. Besides its enzymatic degradation, melatonin is nonenzymatically metabolized when it scavenges free radicals. Melatonin's ability to scavenge the highly toxic hydroxyl radical is one of the newly discovered functions of this widely acting hormone.

FACTORS INFLUENCING PLASMA MELATONIN RHYTHM

LIGHTDARK CYCLE The primary factor determining plasma melatonin levels is the prevailing lightdark environment.8 The rhythmic production and secretion of melatonin is synchronized by the lightdark cycle. Typically, daytime is associated with low plasma melatonin levels and nighttime is associated with high levels (see Fig. 10-3).6 Blind humans with no retinal light perception exhibit free-running melatonin cycles, with periods of ~24.7 hours. In these persons, the highest plasma melatonin values can occur either at night or during the day.9 The rise in plasma melatonin levels in sighted humans may actually precede the removal of artificial light at bedtime.5 Short-term dark exposure during the day typically is not associated with increased circulating melatonin levels. When persons are exposed to light during the normal dark period, however, plasma melatonin levels drop rapidly if the light has a brightness of 2500 lux or more (2500 lux is roughly four to five times the intensity of normal room light but considerably less than the intensity of sunlight).10 Individuals awakened at night and exposed to darkness or low light intensities (<300 lux) typically do not exhibit a marked alteration in the rhythmic production of melatonin. When the lightdark cycle is shifted, such as in transmeridian travel or with a new restactivity schedule, the melatonin rhythm also is shifted; however, the change is not immediate.11 Thus, if humans are phase-shifted by 12 hours, elevated levels of blood melatonin and its urinary metabolites may not occur for several days. Within 7 to 12 days, depending somewhat on age (see later), the melatonin cycle readjusts to the prevailing lightdark environment so that elevated levels coincide with the period of darkness. The interval required for reentrainment of the melatonin cycle also depends on the magnitude of the phase shift: the greater the phase shift, the greater the interval required for the melatonin rhythm to reentrain. Nocturnal secretion of melatonin seems unrelated to sleep stage.12 Maximal plasma melatonin levels usually appear before the maximal rapid eye movement sleep peak. In animals that experience natural photoperiodic and temperature conditions throughout the year, season has a significant influence on the plasma melatonin rhythm, affecting both the duration and the magnitude of the nocturnal rise.2 In humans, who live under more controlled environmental conditions, the influence of season is less obvious but may be measurable.13

AGE AND SEXUAL MATURATION

Age substantially influences the pattern of melatonin production (Fig. 10-4). Whereas the daynight periodicity in plasma melatonin levels probably is not present at birth, by the end of the first year of life children exhibit a robust rhythm.14 The highest nighttime levels of melatonin occur in children 1 to 5 years of age. Several reports document a significant attenuation of the nocturnal rise in plasma melatonin levels between 5 and 15 years of age (Fig. 10-5).15 This decrease is of particular interest because of its possible relationship to pubertal development. In animals, the activated pineal gland, or exogenous melatonin administration, can retard sexual maturation.16 Hence, the gradual decrease in nocturnal plasma melatonin levels as humans mature sexually may permit the establishment of adult neuroendocrinegonadal relationships.17

FIGURE 10-4. Changes in blood melatonin levels as a function of age. During the infantile period (age 0 to 1 year), the melatonin rhythm matures; after the first year of age, a decrease is seen in nocturnal melatonin levels. The drop in nighttime melatonin may be rapid (A and C) or gradual (B) from 5 to 15 years of age (immature period). Low nocturnal melatonin titers in advanced age may be a consequence of a gradual reduction (A and B) or associated with some critical event late in life (C).

FIGURE 10-5. Nocturnal serum melatonin values in male and female subjects aged 1 to 35 years. Values are plotted against chronologic age and state of sexual development (Tanner stages I to VI). Between 5 and 15 years of age, as individuals pass through puberty, nocturnal melatonin levels decrease by roughly 75%. (From Waldhauser F, Dietzel M. Daily and annual rhythms in human melatonin secretion: role in puberty control. Ann N Y Acad Sci 1985; 435:205.)

The decrease in nighttime melatonin levels in the decade beginning at 5 years of age has not been universally demonstrated. Even if a reduction in the ability of the pineal to secrete melatonin is correlated with pubertal onset, this does not mean that the indoleamine normally limits the maturation of the reproductive system. The case could be that gonadotropins or sex steroids secreted by the pituitary and gonads, respectively, inhibit the conversion of serotonin to melatonin in the pineal gland. Finally, the decrease in plasma melatonin may be merely a function of age and not causally related to the status of the reproductive organs. In one case report, however, a strong correlation was found between the decrease in circulating melatonin and pubertal onset. After adolescence, most individuals continue to exhibit a 24-hour rhythm in plasma melatonin production and the urinary excretion of metabolites. In advanced age, however, the ability of the pineal to produce melatonin is severely limited, especially at night. In elderly persons, a nocturnal rise in melatonin is barely detectable. AVERSIVE STIMULI Aversive stimuli, which cause the release of large amounts of catecholamines into the systemic circulation, have been investigated for their effects on the discharge of melatonin from the pineal gland. In general, procedures such as electroconvulsive therapy, pneumoencephalography, and lumbar puncture, and conditions such as excessive short-term exercise, insulin-induced hypoglycemia, and psychologic conflict have a minor influence on day-time plasma melatonin levels.5 In one study, however, strenuous exercise in women significantly increased daytime circulating melatonin levels.18 The consequences of short-term rises in plasma melatonin during the day are unknown. POSSIBLE SEX STEROID FEEDBACK Several studies have described plasma melatonin levels as a function of the human menstrual cycle. The rationale for these studies was that, if melatonin has an

antigonadotropic action in humans, as it has in some other mammalian species, then perhaps ovarian secretory products may limit pineal melatonin secretion at the time of ovulation. In some studies, nocturnal plasma melatonin levels were found to be lower during the midmenstrual cycle (ovulation) than during the postmenstrual or premenstrual periods.19 These results suggest that the midcycle lessening of the nocturnal melatonin increase may permit the release of pituitary gonadotropins required for ovulation. Not all investigators have reported menstrual cycle variations in the 24-hour melatonin rhythms, however. In acyclic athletic women, higher than normal circulating melatonin values have been reported. Melatonin secretion has not been found to be linked to prolactin, growth hormone, adrenocorticotropin, testosterone, or luteinizing hormone.5

CLINICAL IMPLICATIONS
SEXUAL MATURATION The pineal gland and melatonin have been implicated in several clinical entities.5 Besides the possible link between gradually decreasing nocturnal melatonin levels and pubertal development (see Fig. 10-5), pineal tumors may alter sexual maturation (also see Chap. 92).20,21 Tumors of the pineal gland are more prevalent in men than in women and may either retard or advance sexual development.20 The opposite responses to space-occupying lesions of the pineal region are explained on the basis of cellular origin (i.e., parenchymal or nonparenchymal) and the consequential endocrine capabilities of the tumorous mass. In addition, these tumors can cause increased intracranial pressure and, because of associated hydrocephalic dilation of the third ventricle, Parinaud syndrome (paralysis of upward gaze and slightly dilated pupils that react normally on accommodation but not to light) (Fig. 10-6). Hypermelatoninism with an enlarged pineal gland has been shown to be associated with delayed sexual development.

FIGURE 10-6. Computed tomographs (CTs) of a 67-year-old man with Parinaud syndrome and deteriorating mental status. Four years previously, the CT scan shown in A was obtained because of disorientation and dementia thought to be secondary to increased intracranial pressure. A 2-cm mass was seen in the pineal region (arrowheads) with triventricular dilation (arrow) (third and lateral ventricles). The fourth ventricle was not involved. The patient refused definitive surgery, and a ventriculoperitoneal shunt was performed. In the CT scan shown in B, the lesion has grown to 7 cm; the ventricles decreased in size considerably after shunt placement. (Courtesy of Dr. Frederick T. Borts.)

OVULATORY RELATIONSHIPS The observation that nocturnal melatonin levels may be lowest at the time of ovulation in women suggests that a decrease in the indoleamine concentration may permit ovulation.19 In animals, exogenously injected melatonin can inhibit both the release of ova and the surge of ovulatory hormones associated with this process. Melatonin may have a similar antiovulatory capability in humans when given in pharmacologic doses.22 Higher than normal nocturnal melatonin secretion is associated with hypothalamic amenorrhea in women, and with delayed puberty.23 PSYCHOLOGICAL INTERRELATIONSHIPS A possible link between various types of depression and pineal function has been proposed. Several psychiatric diseases have a strong seasonal component; one of these is seasonal affective disorder.24 This condition is characterized by recurring periods of depression during the winter months, when the natural day length is shortest. Typical symptoms are hypersomnia, excessive eating, a craving for carbohydrates, and sadness. Phototherapy for this disorder entails exposing the patient to bright (2500 lux), full-spectrum artificial light, usually early in the morning. The symptoms normally associated with seasonal depression are greatly reduced by phototherapy but are partially reinstated if melatonin is given orally during the period of phototherapy. Bright light early in the morning prevents the phase delay in the melatonin rhythm that may be typical of patients with seasonal affective disorder. SLEEP Persons given exogenous melatonin report feeling sleepy. Under usual conditions, high plasma melatonin levels are associated with the nightly sleep interval.25 Melatonin has been reported to affect brain levels of serotonin, a compound that may be important in the initiation or maintenance of sleep. In various studies, melatonin given orally or sprayed intranasally facilitated the onset of sleep or led to tiredness, and generally had a sedative effect. Melatonin is being touted as a sleep-enhancing agent, although the data is contradictory.25a JET LAG AND OTHER BIOLOGIC RHYTHM DISORDERS Jet lag has been related to elevated melatonin levels during times that the concentration of this indoleamine should be low.11 Flight across time zones is associated with a phase shift in the sleepactivity cycle. The circulating melatonin rhythm requires time to readjust, and during this interval melatonin levels are elevated or depressed at unusual times. This lack of synchrony is believed to be related to the fatigue associated with jet lag.26 Melatonin administration is being tested as a therapy for jet lag. The ability of melatonin to influence jet lag relates to its influence on the biologic clock (i.e., the suprachiasmatic nuclei). The neurons in these nuclei contain numerous receptors for melatonin.27 The membrane receptors on which melatonin acts are part of the superfamily of receptors that possess seven transmembrane domains. The receptors are pertussis toxin sensitive and G-protein linked. Cyclic adenosine monophosphate (cAMP) functions as the intracellular messenger.28 TUMOR GROWTH The possibility that the pineal gland and melatonin may be related to tumor growth has received much attention.29 The interactions of the pineal gland with tumor promotion are most obvious in the case of malignant growths that are dependent on sex steroids or prolactin. In addition, because of its free radical scavenging ability (see later), melatonin may prevent DNA damage that precedes tumor initiation. Depressed nocturnal plasma melatonin levels in patients with mammary cancer may contribute to tumor growth. In pre-menopausal and postmenopausal women with clinical stage I and II mammary carcinogenesis, the nocturnal melatonin rise was attenuated in those with estrogen receptorpositive tumors, compared to those with estrogen receptornegative mammary growths. 30 In the patients with estrogen receptorpositive tumors, a strong negative correlation was seen between plasma melatonin concentrations and estrogen receptors in the tumor. Whether the lower levels of melatonin in these women were a cause or an effect of the tumor, or an unrelated temporal association, was impossible to determine. Melatonin is known to suppress cellular proliferation and tissue growth.29 FREE RADICAL SCAVENGING Melatonin has been shown to be a highly effective scavenger of the hydroxyl radical, singlet oxygen, and the peroxynitrite anion, and it stimulates the activity of several antioxidative enzymes.31 Because of its multiple antioxidative actions, melatonin has been tested for its ability to reduce oxidative damage due to ionizing radiation and a variety of xenobiotics; in these tests pharmacologic levels of melatonin have been shown to efficiently abate oxidative damage.32 These findings have generated interest in the potential use of melatonin in the treatment of neurodegenerative diseases of the aged that may involve free radical damage to neurons.31,33 Examples of

such conditions include Alzheimer, Parkinson, and Huntington diseases.

CONCLUSION
The human pineal gland is a highly active organ that secretes at least one indole hormone, melatonin, and possibly several other active substances, either indoleamines or peptides. Assays are available to measure melatonin in bodily fluids34 and its metabolites in the urine. The cyclic production of melatonin is highly characteristic, with highest levels occurring during the dark/sleep phase. The consequences of altered melatonin rhythms are incompletely understood but likely relate to several clinical entities. In addition, some persons probably have an excess (hypermelatoninism)17 or deficiency (hypomelatoninism)35 in melatonin production and secretion. CHAPTER REFERENCES
1. 2. 3. 4. 5. 6. 7. Vollrath L. Functional anatomy of the human pineal gland. In: Reiter RJ, ed. The pineal gland. New York: Raven Press, 1984:285. Reiter RJ. Pineal melatonin: cell biology of its synthesis and of its physiological interactions. Endocr Rev 1991; 12:151. Reiter RJ. Pineal gland: interface between the photoperiodic environment and the endocrine system. Trends Endocrinol Metab 1991; 2:13. Pangerl B, Pangerl A, Reiter RJ. Circadian variations of adrenergic receptors in the mammalian pineal gland: a review. J Neural Transm 1990; 81:17. Vaughan GM. Melatonin in humans. Pineal Res Rev 1984; 2:141. Arendt J. Mammalian pineal rhythms. Pineal Res Rev 1985; 3:161. Neuwelt EA, Lewy AJ. Disappearance of plasma melatonin after removal of neoplastic pineal gland. N Engl J Med 1983; 308:1132.

7a. Cook MR, Graham C, Kavet R, et al. Morning assessment of nocturnal melatonin secretion in older women. J Pineal Res 2000; 28:41. 7b. Witt-Enderby PA, Li PK. Melatonin receptors and ligands. Vitamin Horm 2000; 58:321. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. Reiter RJ. The mammalian pineal gland as an end organ of the visual system. In: Wetterberg L, ed. Light and biological rhythms in man. Oxford: Pergamon, 1993:145. Lewy AJ, Newsome DA. Different types of melatonin circadian secretory rhythms in some blind subjects. J Clin Endocrinol Metab 1983; 56:1103. Lewy AJ, Wehr TA, Goodwin FK, et al. Light suppresses melatonin secretion in humans. Science 1980; 210:1267. Fevre-Montage M, Van Cauter E, Retetoff S, et al. Effects of jet lag on hormonal pattern. II. Adaptation of melatonin circadian periodicity. J Clin Endocrinol Metab 1978; 52:642. Vaughan GM, Allen JP, de la Pena A. Rapid melatonin transients. Waking Sleeping 1979; 3:169. Stokkan KA, Reiter RJ. Melatonin rhythms in Arctic urban residents. J Pineal Res 1994; 16:33. Attanasio A, Borelli P, di Rocco E, et al. Clinical significance of melatonin in children. In: Gupta D, Borelli P, Attanasio A, eds. Pediatric neuroendocrinology. London: Croom Helm, 1985:203. Waldhauser F, Dietzel M. Daily and annual rhythms in human melatonin secretion: role in puberty control. Ann N Y Acad Sci 1985; 453:205. Reiter RJ. The pineal and its hormones in the control of reproduction. Endocr Rev 1980; 1:109. Puig-Domingo M, Webb SM, Serrano J, et al. Melatonin-related hypogonadotropic hypogonadism. N Engl J Med 1992; 17:81. Carr DB, Reppert SM, Mullen B, et al. Plasma melatonin increases during exercise in women. J Clin Endocrinol Metab 1981; 53:224. Hariharasubramanian N, Nair NPV, Pilapel C, et al. Plasma melatonin levels during menstrual cycle: changes with age. In: Gupta D, Reiter RJ, eds. The pineal gland and puberty. London: Croom Helm, 1986:166. Vaughan GM, Meyer GG, Reiter RJ. Evidence for a pineal-gonadal relationship in the human. In: Reiter RJ, ed. The pineal and reproduction. Basel: Karger, 1978:191. Herrick MK. Pathology of pineal tumors. In: Neuwelt EA, ed. Diagnosis and treatment of pineal region tumors. Baltimore: Williams & Wilkins, 1984:31. Voordouw BCG, Euser R, Verdonk RER, et al. Melatonin and melatonin-progestin combinations alter pituitary-ovarian function in women and can inhibit ovulation. J Clin Endocrinol Metab 1992; 74:108. Berga SL, Mortola JF, Yen SSC. Amplification of nocturnal melatonin secretion in women with functional hypothalamic amenorrhea. J Clin Endocrinol Metab 1988; 66:242. Rosenthal NE, Sack DA, James SP, et al. Seasonal affective disorder and phototherapy. Ann N Y Acad Sci 1985; 453:260. Wurtman RJ, Lieberman HR. Melatonin secretion as a mediator of circadian variations in sleep and sleepiness. J Pineal Res 1985; 2:301.

25a. Spitzer RL, Terman M, Williams JB, et al. Jetlag: clinical features, validation of a new syndrome-specific scale, and lack of response to melatonin in a randomized, double-blind trial. Am J Psychiatry 1999; 156:1392. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. Daan S, Lewy AJ. Scheduled exposure to daylight: a potential strategy to reduce jet lag following transmeridian flight. Psychopharmacol Bull 1984; 20:566. Liu C, Weaver DR, Jin X, et al. Molecular dissection of two distinct actions of melatonin on the suprachiasmatic circadian clock. Neuron 1997; 19:91. Delagrange P, Guardiola-Lemaitre B. Melatonin, its receptors, and its relationships with biological rhythm disorders. Clin Neuropharmacol 1997; 20:482. Blask DE. Melatonin in oncology. In: Yu HS, Reiter RJ, eds. Melatonin. Boca Raton: CRC Press, 1993:447. Tamarkin L, Danforth D, Lichter A, et al. Decreased nocturnal plasma melatonin peak in patients with estrogen positive breast cancer. Science 1982; 216:1003. Reiter RJ. Oxidative damage in the central nervous system: protection by melatonin. Prog Neurobiol 1998; 56:1. Reiter RJ, Tang L, Garcia JJ, Muoz-Hoyos A. Pharmacological actions of melatonin in oxygen radical pathophysiology. Life Sci 1997; 60:2255. Brusco LI, Marquez M, Cardinali DP. Monozygotic twins with Alzheimer's disease treated with melatonin: case report. J Pineal Res 1998; 25:260. Miles A. Melatonin: perspectives in the life sciences. Life Sci 1989; 44:375. Li Y, Jiang H, Wang ML, et al. Rhythms of serum melatonin in patients with spinal lesions at the cervical, thoracic, or lumbar region. Clin Endocrinol (Oxf) 1989; 30:47.

CHAPTER 11 MORPHOLOGY OF THE PITUITARY IN HEALTH AND DISEASE Principles and Practice of Endocrinology and Metabolism

CHAPTER 11 MORPHOLOGY OF THE PITUITARY IN HEALTH AND DISEASE

KAMAL THAPAR, KALMAN KOVACS, AND EVA HORVATH The Normal Pituitary Embryology Anatomy Blood Supply Innervation Cytologic Features The Abnormal Adenohypophysis Developmental Abnormalities Circulatory Disturbances Inflammatory Disorders of the Pituitary Miscellaneous Conditions Adenohypophysial Cell Hyperplasia1 Pituitary Neoplasms Pituitary Adenomas Classification of Pituitary Adenomas by Hormone Production Malignant Pituitary Lesions Craniopharyngiomas Pathology of the Neurohypophysis and Hypothalamus Inappropriate Secretion of Vasopressin Diabetes Insipidus Basophilic Cell Invasion Miscellaneous Findings Interruption of the Hypophysial Stalk Neoplasms Chapter References

In 1886, the French neurologist Pierre Marie proposed that the pituitary gland plays a fundamental role in the development of acromegaly. Since then, remarkable progress has been made in understanding hypophysial structure and function, the bio-chemistry of pituitary hormones, the regulation of pituitary hormone synthesis and release, and the morphologic and clinical manifestations of pituitary abnormalities. This chapter focuses on pituitary morphology in health and disease. Because the hypothalamus is closely related to the pituitary, diseases of the hypothalamus also are summarized.

THE NORMAL PITUITARY 1 ,2

EMBRYOLOGY The pituitary gland is derived from two sources. The epithelial part, which includes the pars distalis, pars intermedia, and pars tuberalis, originates from an evagination of the stomodeal ectoderm called the Rathke pouch. The neural portion, which includes the pars infundibularis or infundibulum, the neural stalk, and the pars posterior or pars nervosa, arises from the floor of the diencephalon. The Rathke pouch is detectable at approximately the third week of gestation as a small, thin-walled vesicle in the roof of the stomodeum, which is the primitive buccal cavity. After increasing in size, it adheres to the infundibulum. Its distal end becomes narrower and forms the craniopharyngeal canal, which subsequently is obliterated, although in some cases it may remain patent until the end of intrauterine life or even after birth. The anterior wall of its proximal portion, where cell replication is faster, gives rise to the pars distalis; the posterior wall develops to become the pars intermedia. The anterolateral part of the Rathke pouch grows upward on both sides, in front of the infundibulum, forming the pars tuberalis (Fig. 11-1).

FIGURE 11-1. Embryogenesis of pituitary gland. A, Early invagination of primitive stomodeum and infundibular process. B, Growth of mesoderm constricts Rathke pouch. C, Further development pinches off Rathke pouch from oral cavity. D, Rathke cleft components develop into the pars distalis, pars tuberalis, and possibly the pars intermedia. Infundibular process develops into infundibular stalk and pars nervosa. (From Tindall GT, Barrow DL. Disorders of the Pituitary. St Louis: Mosby, 1986:10, with permission.)

By the end of the third month of intrauterine life, the gross features of the pituitary gland are clearly recognizable. The infundibulum becomes elongated, and the pituitary is embedded deeper in the sella turcica. The neurohypophysis differentiates into the proximal median eminence and the distal posterior lobe, which are connected by the hypophysial stalk. Pituitary hormones are synthesized early in embryonic life. In humans, growth hormone (GH) and adrenocorticotropic hormone (ACTH) can be demonstrated by immunocytology and radioimmunoassay at approximately the ninth week of gestation. These two hormones are soon followed by the appearance of the a and then the b subunits of glycoprotein hormones: thyroid-stimulating hormone (TSH), follicle-stimulating hormone (FSH), and luteinizing hormone (LH). Prolactin is the last adenohypophysial hormone to be produced; it can be detected at approximately the 20th week of intrauterine life. Vasopressin and oxytocin are found at ~10 weeks of gestation. Histologic differentiation also takes place early. Acidophilic cells are noticeable at approximately the third month of gestation; basophilic cells appear a little later. In approximately the eighth week of embryonic life, a large connective tissue mass carrying blood vessels to the developing anterior lobe becomes visible. The neurosecretory material in the posterior lobe can be recognized at approximately the fifth month of gestation. ANATOMY The pituitary lies in the sella turcica, or hypophysial fossa, at the base of the brain; it is surrounded by the sphenoid bone. The pituitary gland is an oval, bean-shaped, bilaterally symmetric organ measuring ~13 mm transversely, 9 mm anteroposteriorly, and 6 mm vertically. The average weight of the pituitary is 0.6 g; it ranges from 0.4 g to 0.8 g in adults, and averages 0.1 g at birth. A reduction in weight is evident in old age, and an increase occurs during pregnancy and lactation. The pituitary gland weighs somewhat more in multiparous women than in nulliparous women or in men. The anterior lobe is larger than the posterior lobe, constituting ~80% of the organ. The cut surface of the adenohypophysis is brownish red and can be distinguished from the

sharply demarcated grayish neurohypophysis. The pituitary is covered by the dura, a dense layer of connective tissue that lines the sella. The sella diaphragm, the connective tissue dura covering the superior surface of the sella, has a small central opening that is penetrated by the hypophysial stalk. The diameter of the opening is ~5 mm. Well protected in the bony sella, the pituitary is located in the vicinity of several structures. The lateral walls of the sella on both sides are close to the cavernous sinuses, the internal carotid arteries, and the oculomotor, trochlear, and abducent nerves. Below and in front of the sella lies the sphenoid sinus, which is separated from the sella by a thin layer of bone. Above the sellar diaphragm and in front of the hypophysial stalk is the optic chiasm. Above the roof of the sella is the median eminence, the hypothalamus, and the third ventricle of the brain. Anatomically, the pituitary is divided into two different structures: the adenohypophysis, which consists of the pars distalis, the pars intermedia, and the pars tuberalis; and the neurohypophysis, which consists of the median eminence, the hypophysial stalk, and the pars posterior or pars nervosa. The pars distalis, the largest part of the adenohypophysis, is the main site of adenohypophysial hormone synthesis and discharge. In humans, the pars intermedia is rudimentary, and its functional significance is unknown. The pars tuberalis, the upward extension of the adenohypophysis, surrounds two sides of the hypophysial stalk and consists of adenohypophysial cells, primarily gonadotropes and thyrotropes. The pars nervosa, the downward extension of the brain, is connected to the hypothalamus by the hypophysial stalk (Fig. 11-2).

FIGURE 11-2. Sagittal diagram of the pituitary and its important anatomic features, including blood supply. (From Tindall GT, Barrow DL. Disorders of the Pituitary. St Louis: Mosby, 1986:11, with permission.)

BLOOD SUPPLY Blood is supplied to the pituitary by the superior and inferior hypophysial arteries, which arise from the internal carotid arteries. The superior hypophysial arteries penetrate the infundibulum and terminate in the surrounding capillary network. The hypothalamic hormones are synthesized in different structural parts and are transported along the nerve fibers to the infundibulum, where they permeate through the capillary walls into the blood. The larger parallel veins deriving from these capillaries are the long portal vessels. They extend downward in the hypophysial stalk and terminate in adenohypophysial capillaries, which carry high concentrations of hypothalamic hormones. The short portal vessels, originating in the distal part of the hypophysial stalk and posterior lobe, also run to the adenohypophysis. Approximately 70% to 90% of the adenohypophysial blood supply is carried by long portal vessels and 10% to 30% is carried by short portal vessels. A descending branch of the superior hypophysial artery known as the loral artery provides some direct arterial blood supply to the anterior lobe without passing through the infundibulum. The capsular arteries, arising from the inferior hypophysial arteries, transport additional arterial blood to the pituitary capsule and a few rows of adenohypophysial cells under the capsule. The inferior hypophysial arteries carry blood to the neurohypophysis. Venous blood is transported from the pituitary by neighboring venous sinuses to the jugular veins. It appears that blood flow may be reversed, and some blood may flow from the adenohypophysis to the brain. The neurohypophysis has an important role in directing blood either to the adenohypophysis or to the hypothalamus. Electron microscopic studies show that the adenohypophysial capillaries are lined by fenestrated endothelium. A subendothelial space and a distinct basement membrane can be seen under the endothelial layer. INNERVATION Despite its close proximity to the nervous system, the adenohypophysis has no direct nerve supply, except for a few sympathetic nerve fibers that penetrate the anterior lobe along the vessels. The nerve fibers may affect adenohypophysial blood flow but play no direct role in the regulation of adenohypophysial hormone secretion. The regulatory role of the hypothalamus is neurohumoral; it is manifested by stimulating and inhibiting hormones produced in the hypothalamus and transported by the portal vessels to the adenohypophysis. The posterior lobe is richly innervated through the hypophysial stalk by the supraopticohypophysial and tuberohypophysial tracts. The former originates in the supraoptic and paraventricular nuclei, the two magnocellular nuclei of the anterior hypothalamus, and transports the neurosecretory material along the unmyelinated nerve fibers from the hypothalamus to the posterior lobe. The latter arises in the central and posterior hypothalamus. CYTOLOGIC FEATURES ADENOHYPOPHYSIS Although cytologic details have been studied extensively, many questions remain unanswered. Immunocytologic and electron microscopic studies have helped investigators to define various cell types and to develop a functional cell classification that allows correlation of structural features with hormone production and endocrine activity. The long-accepted notion that the adenohypophysis consists of three cell categoriesacidophilic, basophilic, and chromophobicno longer is tenable. However, because of its convenience and simplicity, and the weight of tradition, this concept continues to influence terminology, especially that related to pathology. An alternative functional nomenclature based on immunocytologic and ultrastructural findings has been developed and is gaining widespread acceptance. This nomenclature recognizes five different cell types that produce the six known adenohypophysial hormones. Of the five cell types, twoGH cells and prolactin cellsbelong to the acidophilic series. The three other cell types belong to the basophilic series: corticotropes produce ACTH and other fractions of the pro-opiomelanocortin molecule, thyrotropes synthesize TSH, and gonadotropes make FSH and LH. Chromophobic cells are insufficiently granulated to be stained with acidic or basic dyes. Ultrastructurally, however, cells classed as chromophobes on the basis of light microscopic findings contain enough secretory granules and other characteristic fine structural features to be identified as distinct cell types. The cellular composition of the human adenohypophysis probably results from competition among various inducers acting on pluripotential precursor cells. Although the cell population in the glandular acini is not homogeneous, a general pattern of distribution of various cell types usually can be discerned in the normal adenohypophysis. This pattern may be altered in various diseases, especially diseases of the target glandsthe adrenal cortex, thyroid, and gonads. Somatotropes. Growth hormone cells, or somatotropes, are stained by acid dyes. They usually are abundant, accounting for ~50% of the adenohypophysial cells, and are located mainly in the lateral wings. The association of gigantism and acromegaly with acidophilic tumors first suggested that GH is produced in acidophilic cells. On electron microscopic examination, GH cells are seen to contain well-developed rough-surfaced endoplasmic reticulum, prominent Golgi complexes, and numerous secretory granules measuring 300 to 600 nm. The relative numeric proportions, distribution, and morphologic features of GH cells are remarkably constant in the human adenohypophysis and are not noticeably affected by age, sex, or various diseases. Pituitaries of cretins may show a reduced number of GH cells, but this is not a common finding in adult primary hypothyroidism. Prepubertal GH deficiency is associated with dwarfism, but the number, size, and morphologic appearance of GH cells often are normal. This finding is consistent with the fact that growth hormonereleasing hormone (GHRH) administration increases blood GH levels in these dwarfs and accelerates growth. As might be expected, in the adenohypophyses of patients with tumors that produce GHRH, GH-cell hyperplasia, and, less frequently, adenoma formation may be evident. Lactotropes. Prolactin cells, or lactotropes (or mammotropes), constitute ~15% to 20% of adenohypophysial cells. They are acidophilic or chromophobic and stain with erythrosin and carmoisin. However, these stains are not reliable and should be replaced by immunocytologic techniques, which demonstrate prolactin in the Golgi

apparatus and secretory granules. Prolactin cells are randomly scattered throughout the adenohypophysis, showing a concentration at the posterolateral edges, close to the neural lobe. On electron microscopic examination, prolactin cells appear either densely granulated, containing large secretory granules measuring up to 700 nm, or sparsely granulated, possessing prominent rough-surfaced endoplasmic reticulum, conspicuous Golgi complexes, and sparse, spherical, oval, or irregular secretory granules measuring 150 to 300 nm. The densely granulated cells are thought to be resting cells; sparsely granulated cells are assumed to be engaged in hormone secretion. Granule extrusion, regarded as a morphologic sign of hormone secretion, occurs on the capillary side of prolactin cells. A characteristic ultrastructural feature of prolactin cells is misplaced exocytosis, or extrusion of secretory granules from the cell on the lateral cell surface distant from capillaries, and intracellular extensions of the basement membranes. The number of prolactin cells varies considerably under certain conditions. Prolactin cells are the last to appear in the fetal pituitary. Because of the effect of maternal estrogen, they are numerous in the adenohypophyses of newborns. With the cessation of the estrogen effect, their numbers soon decrease and remain low during childhood. The number of prolactin cells increases spectacularly during pregnancy and lactation. This is a true hyperplasia and may explain the greater weight of the pituitary found in multiparous women. Nonetheless, there are no significant differences in the number of prolactin cells between men and nulliparous women, and no regression of prolactin cells is found in old age. Estrogen treatment and primary hypothyroidism of long duration may result in hyperplasia of prolactin cells. Corticotropes. Corticotropes are ACTH-producing cells that make up ~15% to 20% of adenohypophysial cells. They stain positively with periodic acidSchiff (PAS) stain and lead hematoxylin, stain with basic dyes, and are located mainly in the central mucoid wedge, where they are the predominant cell type. Corticotropes produce ACTH and other fragments of the pro-opiomelanocortin molecule, such as b-lipotropin (b-LPH) and endorphins. Positivity to PAS stain is explained by the carbohydrate moiety of the prohormone. Corticotropes are also seen lining the cystic cavities of the pars intermedia. Cells containing immunoreactive ACTH frequently spread into the posterior lobe. The endocrinologic significance of this change, called basophilic cell invasion, is unknown. On electron microscopic examination, corticotropes contain a widely dispersed, moderately developed, rough-surfaced endoplasmic reticulum, a prominent Golgi apparatus, a few bundles of microfilaments, and numerous spherical or irregular secretory granules that vary in electron density, often line up along the cell membrane, and measure 300 to 600 nm. In anencephaly, corticotropes fail to show a normal development and are nearly absent from the adenohypophysis. In long-standing hypocortisolism, corticotropes may increase in number and size and become vacuolated. In patients with tumors that produce corticotropin-releasing hormone (CRH), corticotropes may undergo hyperplasia, indicating that CRH induces their proliferation. The best-known morphologic abnormality of corticotropes in postnatal life is Crooke hyaline change,1 the deposition of a homogeneous, glassy, PAS-negative material in the cytoplasm that contains no ACTH. It corresponds to an accumulation of microfilaments seen in electron microscopic studies (Fig. 11-3). Crooke hyalinization, which is caused by cortisol excess, occurs in patients with Cushing disease, glucocorticoid hormone producing adrenal tumors, and paraneoplastic (ectopic) ACTH syndrome, and after protracted treatment with pharmacologic doses of cortisol or its derivatives. The changes are reversible: the hyaline material disappears with the removal of cortisol- or ACTH-producing tumors, or with the discontinuation of glucocorticoid therapy.

FIGURE 11-3. Crooke cells. Note accumulation of microfilaments (arrow) in the nontumorous portion of the adenohypophysis harboring an adrenocorticotropic hormonesecreting adenoma. 7800

Thyrotropes. Thyrotropes are TSH-producing cells that constitute ~5% of adenohypophysial cells, making them the least numerous in this structure. They are basophilic, stain positively with PAS, aldehyde fuchsin, and aldehyde thionin, and are located mainly in the anteromedial portion of the pars distalis in the mucoid wedge. On electron microscopic examination, they usually have large cytoplasmic processes and contain short, rough-surfaced endoplasmic reticulum membranes, well-developed Golgi complexes, numerous microtubules, and spherical secretory granules that often line up along the cell membrane and measure 100 to 250 nm. Thyrotropes often increase in number and size in patients with long-standing primary hypothyroidism and transform. into so-called thyroidectomy cells. These cells are large thyrotropes with a vacuolated cytoplasm, dilated endoplasmic reticulum cisternae, and large PAS-positive lysosomal globules. Diffuse or nodular thyrotrope hyperplasia may be marked in untreated long-standing primary hypothyroidism; adenomas composed of thyrotropes may be noted. In patients with hyperthyroidism, thyrotropes are sparse, small, and dense. Gonadotropes. Gonadotropes are FSH/LH-producing cells that constitute about 10% of adenohypophysial cells. Many gonadotropes are found in the central mucoid wedge, but they are also randomly distributed in the lateral wings, often close to prolactin cells. Gonadotropes are basophilic, PAS-positive cells. Most produce both FSH and LH, which can be demonstrated immunocytologically in the cytoplasm. Some gonadotropes, however, contain only FSH or LH. On electron microscopic examination, gonadotropes are characterized by a spherical nucleus and an abundant cytoplasm with a well-developed rough-surfaced endoplasmic reticulum, a conspicuous Golgi complex, and two populations of secretory granules, one with a mean diameter of 250 nm and the other with a diameter of 400 nm or more. Other gonadotropes contain only secretory granules averaging 250 nm in diameter. There are several transitional forms, indicating that gonadotropes derive from the same precursor with the ability to produce both FSH and LH. After removal of the gonads, castration cells (gonadectomy cells) are apparent in the pituitary. These cells are large gonadotropes with pale vacuolated cytoplasm and a peripherally located nucleus. In some castration cells, the entire cytoplasm appears to be transformed into one large vacuole, endowing the cell with a signet-ring appearance. On electron microscopic examination, the most prominent finding is marked dilation of the endoplasmic reticulum membrane network and a reduced number of secretory granules. The effect of gonadal steroids on the morphology of human gonadotropes has not been sufficiently documented. POSTERIOR LOBE AND PITUITARY STALK The posterior lobe, the downward extension of the central nervous system, is composed of nerve fibers, axon terminals, glial cells called pituicytes, and neurosecretory material stored in nerve endings in the form of granules that stain with Gomori chromalum-hematoxylin, aldehyde fuchsin, and aldehyde thionin. Immunocytologic techniques conclusively reveal the presence of vasopressin, oxytocin, and their carrier protein, neurophysin, in the neurosecretory material. The pituitary stalk, which connects the hypothalamus and median eminence with the pituitary, contains unmyelinated nerve fibers that terminate in the posterior lobe and portal vessels, and transport releasing and inhibiting hormones to the anterior lobe.

THE ABNORMAL ADENOHYPOPHYSIS 1a,2

Diseases of the adenohypophysis can be broadly divided into (a) developmental abnormalities; (b) vascular disorders; (c) inflammatory conditions; (d) miscellaneous alterations, including deposition of various substances; (e) hyperplasias; and (f) neoplasms. DEVELOPMENTAL ABNORMALITIES Pituitary Aplasia. Pituitary aplasia, the congenital absence of the hypophysis, is a rare abnormality that is often accompanied by other malformations. The agenesis may involve the entire pituitary gland or the adenohypophysis and is caused by defective formation of the Rathke pouch. If the affected new-born survives, severe hypopituitarism develops. Pituitary hypoplasia is the milder form of the same defect. Anencephaly. In anencephaly, the brain, including the hypothalamus, is missing; thus, no neurohumoral regulation is exerted on the pituitary. The posterior lobe is present in some cases and absent in others. The anterior lobe is reduced in size and contains decreased numbers of corticotropes. The other adenohypophysial cell

types are well developed and show no major abnormalities. Persistent Remnants of the Rathke Pouch. Remnants of the Rathke pouch persist in 20% to 50% of human pituitaries in the form of squamous cell nests. The nests vary in size and are located at the distal end of the stalk close to the anterior lobe. Persistent Cleft of the Rathke Pouch. Persistence of the cleft of the Rathke pouch is a harmless congenital defect and a common autopsy finding. The cleft fails to close and a distended, colloid-filled space is seen between the anterior and posterior lobes. Although this cleft is generally microscopic in proportion, it rarely may accumulate sufficient colloidal material to become an expansile and clinically significant intrasellar and suprasellar mass. These are known as Rathke cleft cysts, as discussed later (see the section on neoplasms). Although these lesions are most certainly not neoplastic in nature, they are discussed under the heading of neoplasms both for convenience and because they frequently mimic clinically and radiologically true cystic neoplasms of the sellar region. Pituitary Dystopia. Pituitary dystopia is a rare condition characterized by a failure of union of the neurohypophysis and adenohypophysis during early development. The pituitary stalk is foreshortened, resulting in an extrasellar location of the neural lobe and a failure of the latter to descend into the sella. Usually, there is no physical attachment between the neurohypophyses and the adenohypophyses, although, occasionally, the two may be tenuously attached by strands of tissue. The anomaly is generally inconsequential, most cases being incidental autopsy findings. Rarely, it may be accompanied by other abnormalities such as hypogonadism, growth retardation, or other congenital anomalies. Septo-Optic Dysplasia. Septo-optic dysplasia is a complex developmental disorder characterized by variable and often partial expression of midline structural abnormalities of the brain, hypoplasia of the optic nerves, and hypothalamic dysfunction. The last may manifest as anterior and posterior pituitary failure on a hypothalamic basis. Additional features of hypothalamic dysfunction may also be present, including alterations of temperature regulation, hyperphagia, and precocious puberty. The full syndrome is expressed in only a few cases; most patients come to medical attention during early childhood with pituitary insufficiency and visual dysfunction. Septo-optic dysplasia is a medically treatable condition that is wholly compatible with life. Anatomic Variations. Anatomic variations in the shape and position of the pituitary and its relationships to neighboring tissues are common. Although these differences have no major clinical significance, they may be important to radiologists and neurosurgeons. Pharyngeal Pituitary. In virtually all persons, a small ectopic focus of anterior pituitary tissue persists throughout life, and can usually be identified as a minute, oval, midline nodule embedded within the sphenoid bone. Known as the pharyngeal pituitary, this remnant of the Rathke pouch is usually <5 mm in size and is most frequently located deep within the mucosa or periosteum, beneath or near the vomerosphenoidal articulation; less often, it may be found in the nasopharynx or even within the nasal cavity. It is surrounded by a thin connective tissue capsule and consists of small clusters of chromophobic cells mixed with a few acidophilic and basophilic cells. In contrast to the pars distalis, the pharyngeal pituitary is richly innervated but has no portal blood supply; thus, it receives no hypothalamic hormones directly that might otherwise affect its secretory activity. Although immunocytologic techniques have disclosed various adenohypophysial hormones in the pharyngeal pituitary, this structure has no major endocrinologic significance and shows no marked histologic changes in patients with endocrine disorders. It cannot take over the function of the adenohypophysis after hypophysectomy or destructive adenohypophysial disease. The only clinically relevant feature of the pharyngeal pituitary is that it rarely may be the site for pituitary adenoma development.1,2 and 3 Most such tumors have been situated within the sphenoid sinus, and both nonfunctioning and hormonally active tumors have been reported. With respect to the latter, GH-producing tumors have been reported most commonly, followed in frequency by prolactin-producing and ACTH-producing adenomas. In the true ectopic pituitary adenoma, the intrasellar pituitary should be normal, although rarely, simultaneous development of an intrasellar and noncontiguous ectopic pituitary tumor has been reported. Another rare site for ectopic pituitary adenomas is the suprasellar region. Such ectopic tumors presumably arise from adenohypophysial cells of the pars tuberalis situated on the supradiaphragmatic portion of the pituitary stalk. Empty Sella Syndrome. The term empty sella refers to the anatomic state resulting from the intrasellar herniation of the subarachnoid space through a defective and enlarged diaphragmatic aperture. The result is compression and posterior displacement of the pituitary gland, enlargement of the sella, and a seemingly empty appearance of the sella on both gross and radiologic examination. It is of clinical and pathophysiologic importance to distinguish those cases of empty sella occurring without identifiable cause (i.e., primary empty sella) from those resulting from a loss of intrasellar volume, such as would occur after an infarction, surgery, or the radionecrosis of an intrasellar neoplasm (i.e., secondary empty sella).
PRIMARY EMPTY SELLA SYNDROME.4 Anatomic

defects in the diaphragma sellae of 5 mm or more have been demonstrated in ~40% of consecutive autopsies, with >20% exhibiting intrasellar extension of the subarachnoid space5 and 5% showing a fully developed empty sella.6 Whether such abnormalities alone are the cause of primary empty sella syndrome, a predisposing factor to it, or simply the result of some other process remains uncertain. Because most of these features have been incidental autopsy findings in persons without neurologic or endocrine symptoms, it is likely that additional factors contribute to the clinical syndrome. Elevated intracranial pressure is a potentially important contributing factor because it has been documented in patients with primary empty sella syndrome. Ten percent of patients with benign intracranial hypertension have a coexisting empty sella. This latter relationship is especially intriguing because both conditions share overlapping clinical profiles.7 Most cases of primary empty sella are discovered incidentally in patients who do not have symptoms. In the few patients who do have symptoms, the clinical profile is characteristic. Eighty percent of cases occur in middle-aged women, many of whom are obese and hypertensive. Spontaneous cerebrospinal fluid rhinorrhea, usually through a markedly thinned and eroded sellar floor, may complicate the primary empty sella in up to 10% of cases involving symptoms. Clinically evident pituitary dysfunction is unusual. Subtle abnormalities of the GH axis, appreciable only on dynamic endocrine testing, have been reported, as have rare accounts of panhypopituitarism.8 A modest hyperprolactinemia on the basis of stalk distortion occurs in fewer than 10% of patients. (The occasional occurrence of a pituitary microadenoma, most often a prolactin-producing adenoma, in association with primary empty sellar syndrome is purely coincidental.) In the most exceptional instances, intrasellar prolapse of the optic chiasm may be a source of visual dysfunction. In most cases, objective ophthalmologic findings, although rarely present, are the result of coexisting benign intracranial hypertension and not of the empty sella per se. The gross pathologic features of the condition include an enlarged and thin-walled, thinned-floor sella, the diaphragm of which consists of a narrow rim. A markedly flattened pituitary gland can be seen displaced against the posterior sellar wall. Despite marked distortion of the gland, its histologic appearance and immunochemical integrity remain largely intact. A secondary empty sella most commonly occurs after surgical extirpation or radiotherapy of a pituitary adenoma. The diaphragm may be developmentally deficient, eroded by the primary tumor, or affected by its treatment, permitting the descent of both the chiasm and the chiasmatic cistern into the sella. Because the latter may become entrapped and kinked by arachnoid adhesions and scar tissue, visual dysfunction is a common mode of presentation in patients with secondary empty sella syndrome. Secondary empty sella syndrome also may occur in the setting of atrophy of a nontumorous pituitary or of pituitary adenomas that have previously undergone massive hemorrhage or infarction, as in Sheehan syndrome and pituitary apoplexy, respectively.
SECONDARY EMPTY SELLA SYNDROME.

CIRCULATORY DISTURBANCES Pituitary Hemorrhage. Hemorrhages of the pituitary are rare. They may develop in patients with head trauma, various hematologic abnormalities, or increased intracranial tension. In rapidly growing pituitary adenomas, intrahypophysial pressure may increase, leading to compression of intrahypophysial or extrahypophysial portal vessels and the arrest of portal circulation. Vessels may undergo a subtle, hypoxic injury noticeable on electron microscopic examination. If the damage is severe, the vascular walls cannot withstand the elevations in blood pressure and they rupture, resulting in hemorrhage. Pituitary apoplexy is the extreme variant of this process (see later in this chapter). Pituitary Infarction. Pituitary infarction is a noninflammatory, coagulative necrosis caused by ischemia secondary to interruption of the blood supply. Small adenohypophysial infarcts are common, being found in ~1% to 6% of autopsies of unselected adult subjects. The lesions remain unrecognized clinically and can be detected only by histologic examination. A loss of 75% of adenohypophysial tissue produces no clinical symptoms of hypopituitarism and no biochemical abnormalities. Pituitary necrosis can be associated with several diseases. Postpartum pituitary necrosis (Sheehan syndrome) occurs in women who experienced severe blood loss and were in hypovolemic shock about the time of delivery. During shock, the. pituitary circulation may be interrupted and the anterior lobe undergoes ischemic infarction. Adenohypophysial necrosis also may occur in nonobstetric shock, but less frequently than in women with severe circulatory failure secondary to obstetric hemorrhage. This suggests that pregnancy predisposes women to pituitary necrosis, but neither the site of sensitization nor the mechanism leading to necrosis is known (see Chap. 17). Necrotic foci of varying sizes can be found in the pituitaries of patients with diabetes mellitus, head trauma, cerebrovascular accidents, increased intracranial pressure, and epidemic hemorrhagic fever. Pituitary infarction develops after disruption of the pituitary stalk, which causes an arrest of the adenohypophysial circulation. Adenohypophysial infarcts often can be seen in patients who were maintained on mechanical respirators before they died. The lesions represent coagulative infarcts and often are accompanied by severe hypoxic lesions of the brain.

The pathogenesis of pituitary infarction is unclear, and the mechanism of arrest of the adenohypophysial circulation is not known. Proposed causes include embolism, thrombosis, disseminated intravascular coagulation, vascular compression, vasospasm, and primary capillary damage. In postpartum pituitary necrosis, Sheehan postulated that severe spasm develops in those arterioles from which portal vessels arise.2 Vasospasm is followed by hypophysial ischemia and secondary thrombosis, resulting in coagulative infarcts that usually spare the posterior lobe and hypophysial stalk because these areas have a rich arterial blood supply. In cases of postpartum pituitary necrosis, infarcted areas may be large, involving more than 90% of the anterior lobe. Adenohypophysial cells are not capable of sufficient regeneration. Thus, when there is extensive infarction, permanent hypopituitarism develops. Because modern obstetric care usually prevents blood loss and obstetric shock in pregnant women, Sheehan syndrome has become rarer. Fibrous atrophy is the final phase of ischemic necrosis of the anterior lobe. The necrotic areas are replaced by fibrous tissue. The sequence of events is identical to that occurring in infarcts of other organs. Necrotic foci may occur in the posterior lobe and hypophysial stalk in association with head injuries, increased intracranial pressure, and obstetric and nonobstetric shock. These patients may develop diabetes insipidus. Pituitary Apoplexy. Classically defined, pituitary apoplexy9,10 refers to the abrupt and occasionally catastrophic occurrence of acute hemorrhagic infarction of a pituitary adenoma. The clinical syndrome is easily recognized, consisting of acute headache, meningismus, visual impairment, ophthalmoplegia, and alterations in consciousness. Without timely intervention, patients may die of subarachnoid hemorrhage or acute, life-threatening hypopituitarism. As defined herein, pituitary apoplexy is a complication in 1% to 2% of all pituitary adenomas. Silent, or subclinical, hemorrhage into a pituitary adenoma is considerably more common, as evidenced by the finding of hemorrhage, necrosis, or cystic change in up to 10% of all surgical specimens. There is little consensus as to which tumor types, if any, are most susceptible to apoplectic hemorrhage. Some have suggested that hormonally active tumors associated with acromegaly and Cushing disease are especially prone to apoplexy, whereas others have found large nonfunctioning tumors to bear the greatest risk. In the experience of the authors, large nonfunctioning pituitary tumors, particularly silent corticotrope adenomas, appear to have the highest inherent tendency to undergo apoplectic hemorrhage. The pathophysiologic basis of pituitary apoplexy remains speculative. Ischemic necrosis of a rapidly growing tumor, intrinsic vascular abnormalities peculiar to pituitary tumors, and compression of the superior hypophysial artery against the sellar diaphragm have all been suggested as mechanisms contributing to apoplectic hemorrhage.11 Predisposing factors loosely associated with apoplexy include bromocriptine therapy, anticoagulation, diabetic ketoacidosis, head trauma, estrogen therapy, and pituitary irradiation. Most cases, however, occur in the absence of any known predisposing condition. Chronologically, apoplexy begins with infarction of the tumor and the surrounding gland and is followed by hemorrhage and edema. This sudden increase in both pressure and volume within the tumor causes precipitous expansion of the tumor followed by mechanical compression of the optic apparatus and of structures within the cavernous sinus. The bulk of the hemorrhage is generally contained within a tense tumor capsule, although an extravasation of blood into the subarachnoid space frequently occurs. Obstructive hydrocephalus may further complicate apoplexy in large macroadenomas, particularly those having a significant suprasellar component. Glandular destruction of varying degree is a regular pathologic feature of apoplexy; this results in partial or total and transient or permanent hypopituitarism. Fortunately, the anterior pituitary has an astonishing reserve capacity; at least 75% to 90% of the gland must be destroyed before permanent endocrine deficits develop. The posterior pituitary, which has its own blood supply, generally escapes injury. Accordingly, diabetes insipidus only rarely is a complication of pituitary apoplexy (see Chap. 26). Histologically, most pituitary apoplexy specimens have a fairly uniform appearance, consisting primarily of blood and necrotic tumor. Often, only ghosts of neoplastic cells remain. A typical adenoma pattern still may be demonstrated with reticulin stains, thus confirming the presence of an underlying adenoma. Rare instances of pituitary apoplexy causing complete autoinfarction of a pituitary adenoma and resulting in spontaneous endocrinologic cure have been known to occur. In addition to the acute and subacute complications noted previously, a late manifestation of pituitary apoplexy is the secondary empty sella syndrome. INFLAMMATORY DISORDERS OF THE PITUITARY Inflammation. The pituitary gland rarely may be subject to a variety of inflammatory disorders, the pathogenesis of which ranges from acute suppuration to chronic granulomatous conditions to autoimmune processes. Despite the diversity of pathologic processes represented, their somewhat generic clinical presentation as nonfunctioning sellar masses frequently prompts a preoperative diagnosis of pituitary adenoma; their inflammatory nature is revealed only after pathologic examination of the surgical specimen. Infectious Diseases of the Pituitary Gland.4,12 With the availability of effective antimicrobial therapy, acute bacterial infection of the pituitary gland has become an exceedingly rare event that periodically surfaces in the form of isolated case reports. In many cases, neither a cause nor a predisposing condition can be found. Those of known etiology, however, appear to arise in one of two clinical settings. The first, and perhaps most important, mechanism involves secondary extension from a preexisting, anatomically contiguous purulent focus. An acute sphenoid sinusitis is most often implicated; less commonly, osteomyelitis of the sphenoid bone, cavernous sinus thrombophlebitis, peritonsillar abscess, mastoiditis, purulent otitis media, or bacterial meningitis serves as the inciting focus. The second mechanism underlying purulent hypophysitis is septicemia, with pituitary infection being a complication of hematogenous dissemination from any of numerous distant septic foci (pneumonia, osteomyelitis, endocarditis, septic abortion, retroperitoneal abscess). Microabscesses, particularly of the posterior lobe, are occasional autopsy findings in patients succumbing to overwhelming sepsis. In this context, they likely represent a clinically insignificant preterminal complication. Of note, postoperative pituitary abscess complicating transsphenoidal surgery is exceedingly rare. The symptoms of pituitary abscess often are indistinguishable from those of other sellar masses (headache, chiasmatic syndrome, hypopituitarism). When they are accompanied by symptoms of meningitis, however, the possibility of pituitary abscess should always be strongly considered.13 The bacteriology of pituitary abscess is diverse. When an organism can be isolated, Staphylococcus aureus, Diplococcus pneumoniae, group A Streptococcus, Klebsiella sp., and Citrobacter diversus have been reported most often. 12 A surprising number of pituitary abscesses has developed in the setting of a preexisting sellar lesion, such as pituitary adenoma, craniopharyngioma, or Rathke cleft cyst.12,13 Why such lesions should be especially vulnerable to abscess formation remains speculative, but it may be related to poor circulation or areas of necrosis present in some such lesions. The histologic picture of pituitary abscess is remarkable for extensive tissue destruction, as evidenced by necrosis and a dense polymorphonuclear inflammatory infiltrate. The mortality associated with pituitary abscess is high, approaching 28% in the absence of meningitis, and 45% when associated with meningitis.13 Pituitary infections also may be caused by a variety of other agents. Tuberculosis, still endemic in certain areas, was historically an important cause of hypopituitarism.12 In most instances, parapituitary involvement has been secondary to dense, plaque-like basilar meningitis. Secondary arteritis, a frequent accompaniment, can result in infarction of the pituitary stalk. Intrasellar tuberculomas usually are associated with widespread destruction of the pituitary, frequently are calcified, and almost always are associated with active tuberculosis elsewhere. When it causes symptoms, tuberculous involvement of the pituitary or its stalk manifests as anterior pituitary failure or diabetes insipidus. Syphilis, now uncommon in its consummate forms, was historically an important cause of destructive granulomatous inflammation of the pituitary. Manifesting as a discrete gummatous lesion or as diffuse scarring, the eventual result of syphilitic infection is massive destruction of the gland and, in some cases, hypopituitarism.8 Mycotic infection, notably aspergillosis, also has been reported to involve the sellar, parasellar, and orbital regions, presenting as an inflammatory mass. Parasitic infiltration of the sellar and parapituitary regions by Cysticercus and Echinococcus organisms has been known to produce a mass in this region. Finally, in the context of the acquired immunodeficiency syndrome and other immunosuppressed states, an additional spectrum of pituitary infection has emerged, including agents such as Pneumocystis carinii, Toxoplasma gondii, and cytomegalovirus.14 The clinical significance of pituitary involvement by these agents is still poorly understood. Lymphocytic Hypophysitis. Lymphocytic hypophysitis is a destructive, inflammatory disorder of the anterior pituitary; it is presumed to be autoimmune in nature.1,2,15,16 and 17 Occurring almost exclusively in women, the disease is often temporally related to pregnancy. Almost 70% of reported cases have occurred either during pregnancy or within the first postpartum year; only occasionally are women beyond reproductive age affected. Rarely, lymphocytic hypophysitis may occur in men.17 Aside from the fact that women are generally more prone to autoimmune conditions, no explanations exist for the strong female preponderance in lymphocytic hypophysitis. Frequently, concurrent or prior autoimmune disease of other endocrine glands (i.e., Hashimoto thyroiditis, adrenalitis) can be demonstrated, suggesting that lymphocytic hypophysitis is but one component of a generalized polyglandular autoimmune syndrome. That antipituitary antibodies, most often directed against the prolactin cell, have been detected in the serum of some affected patients strongly supports an autoimmune etiology. Whether the process is generated by humoral as well as cell-mediated mechanisms remains uncertain. The clinical presentation includes headache and visual defects caused by an expanding sellar mass, amenorrhea

and galactorrhea as the result of moderate hyperprolactinemia, and varying degrees of hypopituitarism. Grossly, the inflammatory response underlying lymphocytic hypophysitis produces an enlarged, firm pituitary gland, which often extends into the suprasellar space and may be accompanied by enlargement of the sella. Microscopically, the process is restricted to the anterior pituitary, where normal glandular architecture is disrupted by an extensive infiltration of lymphocytes, plasma cells, eosinophils, and macrophages15 (Fig. 11-4). Both the structural and the immunohistochemical integrity of involved cells are lost. In especially extensive cases, loosely organized lymphoid follicles and germinal centers may be present. The chronicity of the process is evidenced by varying degrees of fibrotic change. At the microscopic level, tuberculosis, syphilis, sarcoidosis, giant-cell granuloma, and postpartum pituitary infarction all may be considered in the differential diagnosis. Because the posterior lobe escapes injury, diabetes insipidus is not a feature of this condition.

FIGURE 11-4. Lymphocytic hypophysitis. Note massive mononuclear cell infiltration and extensive destruction of adenohypophysial parenchyma. Hematoxylineosin stain; 100.

The potentially lethal nature of lymphocytic hypophysitis is illustrated by the fact that many early descriptions of the condition stemmed from necropsy studies. Improved recognition of this condition coupled with hormone replacement therapy has rendered lymphocytic hypophysitis a curable disease. Granulomas. In addition to tuberculosis and syphilis, which are the two principal granulomatous infections affecting the pituitary, two additional noninfectious granulomatous processes may present as pituitary granulomas: sarcoidosis and giant-cell granuloma. Considered one of the great imitators, sarcoidosis is known for its tendency to involve pituitary-hypothalamic structures.1,2,4,18,19 Therefore, it serves as a diagnosis of exclusion for masses and other destructive inflammatory processes occurring in this region. Sarcoidosis is a relatively common, multisystem inflammatory disorder of unknown origin. It most commonly affects the lungs, lymph nodes, skin, and eyes, but virtually no organ system is spared. Clinically apparent nervous system involvement occurs in ~5% of cases and may involve the cranial, spinal, or peripheral compartments.
SARCOIDOSIS.

Within the cranium, sarcoidosis has a predilection for the base of the brain, where it entraps cranial nerves and hypothalamic-pituitary structures in an adhesive and infiltrative arachnoiditis. Discrete parenchymal masses occur much less frequently. Involvement of the pituitary, infundibulum, and hypothalamic structures occurs in ~1% of patients with established systemic disease. Rarely, isolated involvement of these structures may be the initial or sole feature of sarcoidosis. The clinical manifestations of sellar region sarcoidosis are variable and generally reflect hypothalamic or infundibular damage. Diabetes insipidus, somnolence, obesity, abnormal temperature regulation, hyperprolactinemia, and hypopituitarism have all been reported. With respect to the last of these, gonadotropic, thyrotropic, and adrenocorticotropic function is most commonly impaired. Although pathologic involvement of the adenohypophysis may be demonstrated, hypopituitarism in the setting of sarcoidosis is generally considered to be the result of injury to the hypophysiotropic areas of the hypothalamus or to the pituitary stalk. That many patients with diminished pituitary function retain responsiveness to exogenously administered hypothalamic releasing factors argues against excessive functional damage to hormone-producing cells of the anterior pituitary. The histologic appearance of sarcoidosis in the pituitary-hypothalamic region is similar to that of sarcoidosis in other organs. Noncaseating granulomas, consisting of giant cells, lymphocytes, and macrophages, can be seen in the anterior and posterior pituitary, the infundibulum, and the hypothalamus. Blood vessels often are involved by the inflammatory process. Depending on the stage of the disease, scar formation of varying degrees may be evident. Although necrosisthe histologic hallmark of tuberculosis and other infectious processesis absent in sarcoidosis, noncaseating granulomas may be a feature of numerous infectious processes. Thus, special stains and cultures for fungi and tubercle bacilli must be performed to exclude an infectious cause, particularly in the absence of known systemic sarcoidosis. Giant-cell granuloma of the pituitary1,2,19 is a rare condition, the earliest description of which stemmed from autopsy studies. Although historically the disease was seldom diagnosed during life, increasing awareness of giant-cell granuloma as a distinct clinicopathologic entity has led to its periodic detection among surgical specimens from patients harboring mass lesions of the sella. Like that of sarcoidosis, the etiology of giant-cell granuloma remains obscure. Because some histopathologic features are common to both conditions, immunemediated mechanisms, perhaps similar to those postulated for sarcoidosis, have been invoked as underlying the development of giant-cell granuloma as well. In contrast to sarcoidosis, however, giant-cell granuloma is a disease virtually exclusive to the pituitary gland; exceptional accounts of histologically similar granulomas have been reported in the adrenals.
GIANT-CELL GRANULOMA.

Topologically, the anterior lobe of the pituitary is hardest hit by giant-cell granuloma; neurohypophysial involvement occurs less frequently, and hypothalamic involvement is distinctly unusual. Histologically, the appearance is that of a noncaseating granulomatous inflammation, with giant cells bearing Schumann bodies, abundant histiocytes, and occasional lymphocytes. Extensive parenchymal destruction is the rule, eventually accompanied by fibrosis and scar formation. The clinical picture generally is dominated by hypopituitarism, which also may be accompanied by diabetes insipidus and moderate hyperprolactinemia, depending on the degree of damage to the pituitary stalk. The inflammatory process often is a source of considerable glandular enlargement, so much so that most surgically treated cases have masqueraded preoperatively as non-functioning pituitary macroadenomas. MISCELLANEOUS CONDITIONS Langerhans Cell Histiocytosis (Histiocytosis X). Classified under the rubric Langerhans cell histiocytosis1,2,4,19,20 are several related but poorly understood nonneoplastic processes that are unified pathologically by their content of highly characteristic histiocyte-like cells. The extent and nature of organ involvement as well as the clinical course in Langerhans cell histiocytosis is variable. Ranging from the fulminant, generalized, and frequently lethal Letterer-Siwe disease, to the multifocal eosinophilic granulomas of Hand-Schller-Christian disease, to the relatively innocent solitary eosinophilic granuloma of bone, hypothalamic-pituitary involvement may be a feature of each. Central nervous system involvement is a common feature of Langerhans cell histiocytosis, although only rarely do central nervous system lesions occur in the absence of disease elsewhere.20 There is an apparent predilection for involvement of the hypothalamus, infundibulum, and posterior pituitary.19,20 The anterior lobe is affected far less often. In most cases, involvement of these structures represents extension from an adjacent bony lesion, although occasionally, in the context of disseminated disease, parapituitary and meningeal involvement occurs in the absence of bony disease. Diabetes insipidus generally is the earliest and most prominent feature of pituitary involvement, reflecting posterior pituitary, infundibular, or hypothalamic infiltration. Perturbations of anterior pituitary function (typically GH deficiency) may occur; less often there may be a deficiency of other hormones. These are considered to be secondary to disease of the hypothalamus or pituitary stalk. Sometimes, damage to the stalk may be so extreme as to render the anterior lobe functionally, if not physically, disconnected from the hypothalamus. Moderate degrees of hyperprolactinemia also may be a feature of the condition, again reflecting hypothalamic or stalk injury. The histologic picture of Langerhans cell histiocytosis is typical of disease elsewhere and is characterized by infiltrates of histiocytes (often foamy in appearance), eosinophils, and lymphocytes. The essential component of the infiltrate is the Langerhans cell, a large mononuclear cell resembling a histiocyte and expressing S-100 protein as well as HLA-DR and CD1 antigens. The ultrastructural presence of Birbeck granules, which are pentalaminar tubular structures found in the cytoplasm of the Langerhans cell, is pathognomonic. A definitive diagnosis of Langerhans cell histiocytosis rests on the characteristic cytologic features of the Langerhans cell and the accompanying infiltrate. The identification of Birbeck granules and the antigenic determinant CD1 is confirmatory. Idiopathic Medical Conditions. Dysfunction of the hypothalamic-pituitary axis remains a poorly characterized component of many idiopathic medical conditions.1,2,9 These include the Laurence-Moon-Biedl syndrome, cerebral gigantism or Sotos syndrome, the Prader-Willi syndrome, and anorexia nervosa. Even at a purely clinical level, the endocrinologic aberrations accompanying these disorders are, at best, incompletely understood. Correspondingly, pathologic studies of hypothalamic and

pituitary tissues in each of these conditions have been few and the findings inconsistent. In all cases, if pathologic correlates of hypothalamic-pituitary dysfunction can be identified, it is usually the hypothalamus that shows a more consistent pattern of pathologic change; primary pathologic features in the pituitary are inconspicuous or absent in these disorders.19 Deposits. Deposits of various substances may be accompanied by anterior hypopituitarism. Amyloid deposits in the pituitary occur outside the cell in the walls of blood vessels and the interstitium and are part of the amyloidosis of other organs, mainly the kidneys, liver, spleen, and intestines. Pituitary amyloid is regarded as immune amyloid and has the same staining characteristics and ultrastructural appearance as amyloid in other organs. In some pituitary adenomas, especially prolactin-producing adenomas, the adenoma cells may produce amyloid that is different ultra-structurally from immune amyloid. In cells with massive amyloid accumulation, the plasmalemma becomes disrupted, and amyloid can be identified in the extracellular space. In hemochromatosis and hemosiderosis, iron pigment may be deposited in the cytoplasm of adenohypophysial cells (see Chap. 131). Iron storage is uneven, it occurs most extensively in the cytoplasm of gonadotropes compared with other cell types. Preferential iron deposits in gonadotropes may explain the occurrence of hypogonadism with iron overload. Electron microscopic examination demonstrates hemosiderin and ferritin particles in the cytoplasm of the adenohypophysial cells; this is incorporated into lysosomal dense bodies. Prussian blue staining combined with an immunoperoxidase technique discloses iron and adenohypophysial hormones in the cytoplasm of the same cells, which is consistent with the assumption that iron uptake does not block hormone production. Iron accumulation may be accompanied by fibrosis, which accounts for the development of adenohypophysial insufficiency. In some pituitary tumors, transferrin has been shown to exhibit stimulatory growth factor-like properties.21 Calcium deposits can be demonstrated at the site of necrosis or in pituitary tumors, especially craniopharyngiomas and prolactin-producing adenomas. Calcification is not so marked as to cause hypopituitarism. In Hurler syndrome, or gargoylism, mucopolysaccharides accumulate in the pituitary. The adenohypophysial cells, mainly acidophilic cells, exhibit marked cytoplasmic vacuolization. Electron microscopic studies show granular membranous bodies and prominent lipidladen lysosomes in some basophilic cells. ADENOHYPOPHYSIAL CELL HYPERPLASIA1,2,10,22 By definition, hyperplasia refers to a nonneoplastic increase in cell number. Although it is generally accepted that physiologic hyper-plasia regularly affects the pituitary gland (e.g., prolactin cell hyperplasia of pregnancy), the occurrence of pathologic forms of pituitary cell hyperplasia has long been questioned. It is now certain that pathologic forms of hyperplasia, although rare, do occur, and occasionally can be the source of both pituitary enlargement and a hypersecretory state in the absence of pituitary adenoma formation.22 Although a small increase in prolactin, ACTH, and TSH adenomas or tumorlets occurs in the setting of protracted estrogen administration, Addison disease, and hypothyroidism, respectively, there is little evidence to suggest that pituitary cell hyperplasia is a common precursor of adenoma formation in humans.10,23 Moreover, pituitary adenomas are rarely surrounded by a zone of hyperplastic cells of similar type. Beyond the conceptual uncertainties surrounding adenohypophysial cell hyperplasia are the real practical difficulties confounding its detection. Even when present, pituitary hyperplasia is difficult to diagnose, even by experienced pituitary pathologists. The small and frequently fragmented nature of surgical specimens coupled with the normal acinar, and sometimes nodular, histologic pattern of normal pituitary cells significantly complicates the identification of hyperplastic foci. Hyperplasia is best recognized on reticulin-stained specimens and by immunohis-tochemical techniques. The essential pathologic feature is expansion of acini with retention of the overall acinar morphology. Morphologically, pituitary cell hyperplasia can be focal, nodular, or diffuse, and generally involves cells of a single type; rarely, several cell types may be affected simultaneously. Prolactin Cell Hyperplasia. The most common form of prolactin cell hyperplasia is physiologic and occurs during pregnancy and lactation. In this context, proliferation of prolactin cells results in a doubling of the size of the gland. Another common form of prolactin cell hyperplasia is that which occurs as the result of the stalk section effect (i.e., interruption of dopamine delivery to the anterior lobe caused by any of several sellar and suprasellar lesions). The presence of prolactin hyperplasia adjacent to some corticotrope adenomas remains unexplained. In cases of long-standing primary hypothyroidism, prolactin cell hyperplasia is an occasional accompaniment, one that presumably reflects the trophic effects of thyrotropin-releasing hormone (TRH) on pituitary lactotropes. Furthermore, stores of hypothalamic dopamine are known to be diminished in untreated hypothyroidism. Accordingly, loss of tonic dopaminergic inhibition provides an additional neuroendocrine mechanism for lactotrope hyperplasia in this circumstance. Isolated lactotrope hyperplasia as the primary cause of hyperprolactinemia is exquisitely rare, as is the coexistence of prolactin cell hyperplasia with a prolactinoma. Growth Hormone Cell Hyperplasia. Growth hormone cell hyperplasia is a rare phenomenon. In virtually all instances, somatotrope hyperplasia occurs as the result of an extrapituitary GHRH-producing tumor (e.g., pancreatic islet cell tumor, pheochromocytoma, bronchial and intestinal carcinoids, small-cell carcinoma of the lung).24 In response to the stimulatory effect of GHRH, pituitary somatotropes enlarge, proliferate, and produce excess GH; acromegaly is the clinical result. Despite persistent stimulation by GHRH and the development of somatotrope hyperplasia, adenomatous transformation rarely occurs in this setting, even after protracted periods of stimulation. Although GHRH-producing tumors are rare causes of acromegaly, they should always be considered in the differential diagnosis. Idiopathic GH cell hyperplasia has yet to be conclusively demonstrated as a cause of acromegaly (see Chap. 219). Corticotrope Cell Hyperplasia. Considerable debate surrounds the role of idiopathic corticotrope hyperplasia as a cause of Cushing disease. In the experience of the authors, corticotrope hyperplasia alone, or in combination with a corticotrope adenoma, is responsible for up to 10% of all cases of pituitary-dependent Cushing disease.22 Theoretically, such cases of Cushing disease should be more refractory to cure by all but total hypophysectomy because hyperplastic foci either remain or are newly induced from the ongoing hyperplastic stimulus. Less controversial, and generally acknowledged, is the occurrence of corticotrope hyperplasia in response to a variety of extrapituitary CRH-producing tumors (e.g., neuroendocrine neoplasms, hypothalamic or adenohypophysial gangliocytomas; see Chap. 75 and Chap. 219). In this setting, hyperplasia may result in considerable glandular enlargement, at times sufficient to mimic a pituitary adenoma. Not surprisingly, nodular hyperplasia of pituitary corticotropes is regularly seen in untreated Addison disease.25 Thyrotrope Hyperplasia. Hyperplasia of TSH-producing cells occurs exclusively in the context of long-standing primary hypothyroidism.26 Frequently, such hyperplasia can be so pronounced as to enlarge the pituitary and simulate an adenoma. Pituitary surgeons should be aware of this lesion because numerous cases of thyrotrope hyperplasia have inadvertently been treated with surgical resection without the benefit of medical therapy. Thyroid hormone replacement alone is curative in many cases. Given the trophic effect of TRH on lactotropes, prolactin cell hyperplasia may be found to coexist with thyrotrope hyperplasia. Gonadotrope Hyperplasia. Hyperplasia of gonadotropes is a rare occurrence that often is difficult to recognize, even in pronounced cases. It has been found in the pituitaries of patients in whom primary hypogonadism commenced at a young age, although it is not encountered among autopsied pituitaries of postmenopausal women. Gonadotrope hyperplasia is not a cause of pituitary enlargement, nor is it an accompanying feature or a likely predecessor to gonadotrope adenoma formation.

PITUITARY NEOPLASMS
Numerous tumor types occur in the sellar region; they can be epithelial or mesenchymal and benign or malignant4,10 (Table 11-1). Some tumors also occur elsewhere; their histologic appearance is the same in the pituitary as in other organs. Some small tumors cause no clinical or biochemical abnormalities, whereas others destroy large areas of the pituitary, resulting in anterior hypopituitarism, diabetes insipidus, and hyperprolactinemia. Other tumors cause syndromes of hormonal hyperfunction (see Chap. 12, Chap. 13, Chap. 14, Chap. 15 and Chap. 16).

TABLE 11-1. Tumors and Nontumorous Lesions of the Sellar Region

PITUITARY ADENOMAS EPIDEMIOLOGY Pituitary adenomas are benign epithelial neoplasms derived from and composed of adenohypophysial cells. They account for ~10% to 15% of all intracranial neoplasms. Depending on the population surveyed, their reported annual incidence varies from 1.0 to 7.6 per 100,000 population.1,10 By this measure, pituitary tumors not only are the dominant form of neoplasia arising in the sellar region, but also are among the most frequent primary intracranial tumors encountered in clinical practice. These figures, derived primarily from neurosurgical series, may even underestimate the true incidence of pituitary adenomas, because their frequency in unselected autopsy cases approaches 25%.10 Thus, neoplastic transformation in the pituitary can be considered an exceedingly common event, albeit one that may not always manifest itself clinically. Although no age group is exempt from the development of pituitary tumors, there is a clear tendency for these lesions to become more common with age, with the highest incidence occurring between the third and sixth decades of life. Only rarely are they diagnosed in prepubertal patients. Based on surgical series, a female preponderance appears to exist, with women of child-bearing age being at greatest risk for tumor development. The basis of this increased susceptibility in women is uncertain. It may be that the susceptibility is more apparent than real, because manifestations of pituitary dysfunction are generally more conspicuous in premenopausal women, prompting earlier diagnosis by both patients and physicians. Moreover, the incidence of pituitary adenomas in autopsy series is equally distributed between the sexes.10 PATHOGENESIS AND MOLECULAR BIOLOGY2,10,23 Accumulating evidence indicates the development of pituitary adenomas to be a multistep and multicausal process that, in its most abbreviated form, consists of an irreversible tumor initiation phase followed by a tumor promotion phase (Fig. 11-5). The events necessary to accomplish the process are only superficially understood. Nonetheless, it is known that hereditary predisposition, endocrine and hypothalamic factors, and specific genomic mutations all appear to have some pathophysiologic role in the initiation or progression of pituitary adenomas. Before considering the relative contribution of each of these to pituitary tumor development, it is important to acknowledge that tumor development in the pituitary is a monoclonal process. This observation is of considerable conceptual relevance and provides the background on which other pathophysiologic events must be integrated.

FIGURE 11-5. Molecular events that are considered important in the development and progression of pituitary adenomas. Conceptually, the events contributing to pituitary tumor development can be distinguished as tumor induction events (right panel) and tumor promotion events (left panel). Tumor induction events represent specific genomic mutations that may be early transforming events (gsp, multiple endocrine neoplasia type 1 [MEN1]). Tumor promotion events represent growth-promoting events, which include additional mutations (PKC-a, H-ras), stimulation by hypothalamic hormones, or modulation by the endocrine status of the patient (i.e., Nelson syndrome).

CLONAL ORIGINS10,23 One of the most fundamental and historically contentious issues surrounding pituitary tumorigenesis relates to whether transformation in the pituitary is primarily the product of hypothalamic dysfunction or simply the result of an acquired transforming mutation intrinsic to an isolated adenohypophysial cell. The hypothalamic hypothesis suggests that pituitary adenomas arise as the eventual, downstream, and seemingly passive consequence of excessive trophic influences, emanating from a dysfunctional hypothalamus. Alternatively, the pituitary hypothesis suggests that pituitary adenomas arise as the direct result of an intrinsic pituitary defect, with neoplastic transformation occurring in relative autonomy of hypothalamic trophic influence. Whereas substantial evidence exists in support of both possibilities, the latter concept has been especially favored in view of the lack of peritumoral hyperplasia in association with pituitary adenomas and because many pituitary tumors can be definitively cured when completely removed. Neither of these would be expected were hypothalamic overstimulation the dominant tumorigenic mechanism. Further strengthening the idea that pituitary adenomas result from somatic mutations that occur at the level of a single, susceptible, adenohypophysial cell have been reports concerning their clonal composition. Using the strategy of allelic X-chromosome inactivation analysis, which assesses restriction fragment length polymorphisms and differential methylation patterns in various X-linked genes, several independent laboratories have confirmed a monoclonal composition for virtually all pituitary adenomas.27,28 Validation of the monoclonal nature of pituitary adenomas has been an important conceptual advance because it has established pituitary adenomas as monoclonal expansions of a single, somatically mutated, and transformed adenohypophysial cell. Were hypothalamic overstimulation the dominant initiating event, then a population of anterior pituitary cells should simultaneously be affected and a polyclonal tumor would be the expected result; this has not been the case. HYPOTHALAMIC FACTORS AND PITUITARY TUMORIGENESIS Whereas the demonstration that pituitary adenomas are monoclonal derivatives of a single transformed adenohypophysial cell does conform well to existing paradigms of human tumorigenesis, it should not be interpreted as somehow exonerating hypothalamic influences of a role in pituitary tumor development. On the contrary, the culpability of hypothalamic hormones in pituitary tumorigenesis continues to gain strength, and there has been renewed interest in integrating a role for these hormones in the current multistep monoclonal model.29 The class of hypothalamic hormones at issue are the hypothalamic hypophysiotropic hormones, which primarily include GHRH, somatostatin (SRIF), CRH, TRH, gonadotropin-releasing hormone (GnRH), and dopamine. Produced in hypothalamic nuclei, descending via the portal circulation, and binding to specific membrane receptors on their respective adenohypophysial target cells, these hormones govern the secretory and proliferative activity of each of the principal pituitary cell types. In logical extension of their physiologic trophic activities has been the implication that aberrant activity of these regulatory hormones in the form of excess stimulation or deficient inhibition may contribute to the genesis and/or progression of pituitary adenomas. For example, in states of pathologic GHRH excess, as occurs with rare GHRH-producing tumors (pancreatic endocrine tumors, carcinoids, pheochromocytomas, and hypothalamic hamartomas/gangliocytomas), chronic GHRH stimulation leads to hyperplasia of pituitary somatotropes, GH hypersecretion, and clinical acromegaly. Depending on the duration of exposure to the excess GHRH, progression from somatotrope hyperplasia to adenomatous transformation has been documented in some, but not all instances.24 A parallel phenomenon has been demonstrated in transgenic mice bearing the human GHRH transgene. That these animals develop gigantism, elevated GH levels, somatotrope hyperplasia, and, eventually, GH-producing pituitary adenomas, provides compelling and conclusive evidence of the tumor-promoting properties of GHRH.30 To date, analogous data implicating other hypophysiotropic hormones have been comparatively few; however, the convincing precedence provided by these GHRH data lends, at the very least, some plausibility to the idea that other hypophysiotropic hormones may also share similar tumorigenic potential. Whereas a specific somatic mutation of an adenohypophysial cell is requisite to adenomatous transformation, hypophysiotropic hormones may modify the cell's susceptibility for such a mutation to occur. Some hypophysiotropic hormones (GHRH, CRH, TRH) are known to induce early-response genes in their respective adenohypophysial target cells, mustering a potent, yet physiologic mitogenic response. Were it to occur, not only would the aberrant overactivity of these regulatory hormones be accompanied by increased proliferative activity of relevant adenohypophysial cells, but the possibility for a transforming mutation also would be correspondingly increased. The extent to which this actually does occur among pituitary adenomas is unknown, although abnormal hypothalamic and other neuroendocrine responses have been demonstrated in patients with pituitary adenomas, including those having undergone successful tumor removal.31 Tumor initiation aside, a more persuasive role for hypothalamic hormones can be envisaged during the tumor progression phase of pituitary tumorigenesis. Given that pituitary adenomas frequently express and retain responsiveness to hypothalamic hormones, the latter have been implicated in facilitating proliferation of the transformed clone. Of particular interest have been a number of reports wherein pituitary tumors themselves were shown to express at the protein and/or mRNA level various hypophysiotropic hormones together with their corresponding receptors. Somatotrope adenomas have been shown to express both GHRH-mRNA transcripts

and protein,32,33 experimental corticotrope adenomas have been shown to express CRH-mRNA transcripts,34 thyrotrope adenomas have been shown to express TRH mRNA,35 and gonadotrope adenomas have been shown to express GnRH mRNA and protein.36 These data strongly suggest the possibility that pituitary adenomas may be subject to autocrine and/or paracrine regulation by endogenously produced hypophysiotropic hormones. In the case of GH-secreting pituitary adenomas, tumoral expression of GHRH mRNA was shown to be an adverse eventone associated with higher proliferative activity, invasiveness, increased secretory activity, and a reduced likelihood of postoperative remission.37 In view of the foregoing, it should be clear that despite the monoclonal constitution of pituitary adenomas, a potential role of hypothalamic hormones in their genesis and/or progression is not easily dismissed (see Fig. 11-5). ENDOCRINE FACTORS A recurring theme, one borne from both experimental study and clinical observation, relates to the possible predisposing, promoting, or perhaps even the inductive effect of certain altered endocrine states to the development of pituitary adenomas. Of particular relevance are those states of target-gland failure wherein the pituitary is no longer subject to negative feedback effects imposed by target-gland hormones. For example, within the pituitary glands of patients with Addison disease and primary hypothyroidism of long duration, the respective frequencies of corticotrope and thyrotrope tumorlets was higher than that observed in control individuals.25,26 Admittedly, only a loose correlation, but stronger still, and of greater clinical concern is the behavior of corticotrope adenomas and thyrotrope adenomas in the setting of bilateral adrenalectomy (Nelson syndrome) and prior thyroidectomy, respectively.38 That such tumors tend to be notoriously more aggressive than those having an intact pituitary-target gland axis emphasizes the potential importance of negative-feedback inhibition in modulating the behavior and progression of these neoplasms. A final endocrine issue, one having been repeatedly implicated in pituitary adenoma development, particularly PRL-producing adenomas, concerns the role of estrogen as a contributor to transformation and/or neoplastic progression in the pituitary. The tumor-promoting properties of this sex steroid are mediated by specific estrogen receptors which, when ligand activated, dimerize and bind to specific DNA-addressing sites to induce transcription of various target genes governing cell proliferation. Whereas chronic estrogen administration routinely induces prolactinomas in rodents, evidence in favor of a similar relationship in humans has been less convincing. The increasing frequency with which prolactinomas were being diagnosed in the 1970s paralleled the use of oral contraceptives in women, a phenomenon that once invited speculation about a possible causal or predisposing effect of the latter in the development of the former. Several case-controlled studies, however, have failed to substantiate such a relationship.39 Still, estrogens are known to alter the morphology and secretory activity of human adenohypophysial cells, indicating that the anterior pituitary is an important target tissue for estrogen action.40 The observation that the hormonal milieu of pregnancy can, in some instances, stimulate prolactinoma growth indicates a potential responsiveness of these tumors to estrogenic stimulation. That human pituitary adenomas express estrogen receptors has been recognized for some time. In addition, estrogen receptor mRNA transcripts are present in all types of pituitary adenomas and in all cell types of the normal anterior pituitary gland.41 In at least a single but noteworthy instance, one involving a transsexual patient, high-dose estrogen therapy. correlated with the development of a human PRL-producing pituitary adenoma.42 Thus, the link between estrogens and human pituitary tumor formation remains somewhat circumstantial, but sufficiently so, that it cannot be entirely dismissed. Genomic Alterations: Oncogene Activation. Accompanying the realization that the somatic mutation of an isolated adenohypophysial cell is an event requisite to pituitary tumorigenesis has been a vigorous attempt to identify and characterize the responsible mutation. Of the genomic and cellular alterations known to occur in pituitary adenomas, relatively few appear to involve activating mutations of known oncogenes. To date, activating mutations of only two oncogenes have been reported in pituitary adenomas: gsp and H-ras. Whereas the former is encountered with some regularity in somatotrope adenomas and periodically in other pituitary adenoma types, the latter has been identified in only isolated instances. The only consistent evidence favoring oncogene activation as a transforming mechanism in the pituitary stems from the discovery and characterization of the gsp oncogene, an oncogene first identified in GH-producing pituitary adenomas.43,44 The signal transduction cascades governing the secretory and proliferative functions of pituitary somatotropes converge on the adenylate cyclase second messenger system. In the normal state, the hypothalamic hypophysiotropic hormone (GHRH) is the principal positive regulator of somatotrope function. After binding to its membrane receptor on the somatotrope cell surface, the GHRH-proliferative signal is coupled to a stimulatory, heterotrimeric G-protein, termed Gs, which binds GTP and activates adenylate cyclase. This results in elevations of cAMP levels that, through a series of poorly characterized downstream events, ultimately lead to GH secretion and somatotrope proliferation. Adenylate cyclase activation is normally a self-limiting, transient, and tightly regulated event. This is because one structural component of Gs, known as thea chain, maintains intrinsic GTPase activity that, after transducing the signal, hydrolyzes GTP and returns Gs to its inactive state, thus terminating the trophic signal. Activating mutations of gsp are the result of point mutations in the a chain gene of Gs. The mutant alpha chain has deficient GTPase activity and is therefore incapable of hydrolyzing GTP and turning off the proliferative signal. Therefore, such mutant forms of the a chain stabilize Gs in its active configuration, thus mimicking the trophic effects of persistent GHRH action. Bypassing the tight regulatory control normally provided by GHRH, somatotropes bearing the mutant a chain of Gs constitutively activate adenylate cyclase, providing an autonomous and unrestrained capacity for cell proliferation and GH secretion. Whereas in North American and European studies, activating mutations of gsp have been reported in ~40% of somatotrope adenomas,44 in Japan, such mutations are rare events.45 In neither geographic setting, however, does their presence confer any significantly distinctive clinical, behavioral, biochemical, or radiologic characteristics to the tumor. In one report, tumors exhibiting gsp mutations occurred in older patients, were smaller, and had lower basal GH levels than wild-type tumors, although this has not been uniformly observed.31,44 Activating mutations of the Gs a chain have also been identified in 10% of clinically nonfunctioning pituitary adenomas and very recently, in 5% of corticotrope adenomas.46,47 and 48 That such mutations should occur in tumors not somatotropic in nature suggests that stimulatory G proteins may underlie intracellular signaling in other cell types as well. As in the case of somatotrope adenomas, nonfunctioning and corticotrope adenomas bearing mutations of gsp do not appear to differ clinically from those tumors lacking this mutation. Aside from gsp mutations that have been identified with some regularity, the search for additional activating mutations involving other candidate oncogenes has not proved particularly informative. The only relevant finding concerns the very occasional demonstration of mutations involving the H-ras oncogene among isolated pituitary tumors. The first dedicated study of ras mutations in the pituitary involved the screening of 19 pituitary adenomas wherein an activating mutation of ras was identified in only a single instance.49 The noteworthy feature of this case, a prolactinoma, pertained to the unusual aggressiveness of the tumor; the tumor was remarkable for a rather early age of onset, multiple recurrences, very high prolactin levels, resistance to dopamine agonist therapy, and unrelenting invasiveness that ultimately proved fatal. In a subsequent genomic screening of 44 pituitary adenomas, none were found to have mutations of ras.50 In a third study, one focusing on pituitary carcinomas, mutations of H-ras were demonstrated in three of five distant metastases but in none of the primary pituitary lesions, nor in any invasive pituitary adenoma.51 Of these, two were corticotrope carcinomas and the third was a lactotrope carcinoma. Collectively, these data suggest that mutations of ras, while uncommon events in pituitary tumorigenesis, do appear to be associated with an unusually aggressive phenotype and are likely late events in pituitary tumor progression. Furthermore, the presence of H-ras mutations in secondary deposits but not in the primary lesion of pituitary carcinomas is especially intriguing for it suggests that this mutation may play a role in initiating and/or sustaining distant pituitary metastases (see Fig. 11-5). GENOMIC ALTERATIONS: TUMOR-SUPPRESSOR GENE INACTIVATION MEN1 Tumor-Suppressor Gene. Genetic predisposition to pituitary tumor development is restricted to a single and uncommon condition, the MEN1 syndrome. This autosomal-dominant condition is characterized by the simultaneous development of tumors involving the parathyroid glands, pancreatic islet cells, and the pituitary. A variably penetrant condition, only 25% of patients develop pituitary tumors, the majority of which are macroadenomas associated primarily with GH and/or PRL hypersecretion.1,10 Approximately 3% of all pituitary adenomas occur in the context of MEN1. The nature of the genetic defect in MEN1 has recently been identified and involves allelic loss of a putative tumor-suppressor gene at the 11q13 locus.52,53 In its recessive behavior, the MEN1 gene is typical of a tumor-suppressor gene, with susceptible individuals inheriting a germline mutation of one of the two 11q13 alleles. Subsequent spontaneous mutation, inactivation, or deletion of the remaining normal 11q13 locus in susceptible endocrine tissues ultimately leads to tumor formation in the involved tissue. Once believed to be a genetic defect that accounted for pituitary adenomas occurring exclusively in the context of MEN1, several studies have also demonstrated loss of the 11q13 locus in seemingly sporadic pituitary adenomas. In the earliest of these, allelic deletions of 11q13 were found in two of three sporadic prolactinomas.52 Subsequently, 4 of 12 sporadic GH cell adenomas were found to have deletions involving the 11q13 locus.54 More recently, allelic deletions of chromosome 11 were found in 18% of pituitary adenomas of all major types.55 Collectively, these data suggest that the 11q13 locus is the site of an important tumor-suppressor gene, the inactivation of which may be of pathogenic relevance to the development of both sporadic and MEN1related pituitary adenomas. p53 Tumor-Suppressor Gene. Mutations of the p53 gene represent the single most common nonrandom genomic alteration encountered in human cancer.56 The contribution of p53 gene mutations to pituitary tumorigenesis remains unsettled. Much of the difficulty in implicating or excluding a role for p53 mutations in these tumors has centered on the apparent incongruity between data obtained by conventional genomic screening from that provided by p53 immunohistochemistry studies.57 On the one hand, in all of three studies wherein the p53 gene was screened at the usual mutational hot spots, not a single pituitary tumor was identified as having a p53 mutation.50,58 Alternatively, in several reports wherein pituitary. tumors were studied immunohistochemically, conclusive nuclear accumulation of p53

protein was demonstrated in some pituitary tumors.57,59 In one report, nuclear accumulation of p53 was selectively present in none, 15%, and 100% of noninvasive adenomas, invasive adenomas, and pituitary carcinomas, respectively. Moreover, the growth fractions of p53 immunopositive tumors were significantly higher than those immunonegative for p53.59 In reconciling the genomic screening and immunohis-tochemical data, it is important to acknowledge the different inferences permissible by each methodology. Whereas p53 immunohistochemistry analysis was once regarded as a surrogate means of detecting underlying gene mutations, the concordance between these two methods has not proved as strong as previously believed, and the former cannot be regarded as unequivocal proof of the latter. Therefore, the weight of existing evidence argues against p53 gene mutations as events that contribute to pituitary tumor development and/or their progression. Still, the inability to invoke an underlying gene mutation as the basis for the observed accumulation of p53 protein among aggressive pituitary tumors should not diminish the practical utility of this protein as an immunohistochemical marker of aggressive behavior in this tumor system. Moreover, alternative and equally relevant pathophysiologic mechanisms other than p53 gene mutations may account for p53 immuno-positivity. Precedence in support of this has been provided by other human tumors, such as sarcomas and cervical carcinomas, wherein immunohistochemically apparent p53 accumulation reflects complex formation, sequestration, and resultant inactivation of wild-type p53 by other proteins, producing a state functionally equivalent to that occurring with a gene mutation; similar mechanisms may be operative in pituitary tumors.56 13q14. The first of the tumor-suppressor genes to be identified, the retinoblastoma gene (Rb) remains the prototypical example of this class of genes. Beyond its role in the development of familial retinoblastoma and various other malignancies and its overall contribution to cell-cycle regulatory control, studies of the Rb gene added a new dimension to the very concept of human cancer, illuminating recessive aspects of the process and the oncogenic consequences that accompany loss of protective genomic elements. The implication that the Rb gene might be involved in pituitary tumorigenesis had a somewhat serendipitous beginning. Transgenic mice in which one of the two germline Rb alleles had been deactivated failed to develop retinoblastomas as anticipated; instead they developed large, high-grade, invasive pituitary tumors.60 On further analysis, these tumors were shown to have lost the remaining normal Rb allele, convincingly implicating a second Rb hit as the basis for pituitary tumor development in this model. All of these tumors were found to be corticotropic in nature, immunoreactive for ACTH, and of pars intermedia origin (unpublished observations). Prompted by the provocative nature of these findings, a number of recent studies have sought to determine the relevance of Rb mutations in human pituitary tumors. In the first report, none of 18 informative pituitary tumors exhibited allelic Rb loss.61 This was further confirmed in a study of 30 informative pituitary adenomas wherein none was found to exhibit loss of heterozygosity (LOH) at the Rb gene locus.62 In another study, however, LOH was found at the Rb locus in all of seven pituitary carcinomas, including their metastatic deposits, and in all of six highly invasive pituitary adenomas.63 The significance of these latter observations vis--vis Rb gene mutations, however, was undermined by the finding of Rb protein in tumors exhibiting LOH at the Rb locus. In reconciling the apparent discordance between Rb gene and protein status, together with the two previous studies that excluded Rb mutations within pituitary tumors, the conclusion was made that another putative tumor-suppressor gene, one present on 13q14 but distinct from Rb, must be involved in pituitary tumor progression (see Fig. 11-5).63 OTHER GENOMIC ALTERATIONS Protein Kinase C Gene. Another genomic alteration identified in pituitary adenomas, specifically invasive ones, relates to the protein kinase C (PKC) family of second messengers. PKC family members are ubiquitous, membrane-bound, intra-cellular kinases whose function is to phosphorylate serine or threonine residues on important substrate proteins. Such kinase activity is thought to govern several important cellular processes, including the transmembrane signaling underlying cell proliferation and differentiation. Altered or aberrant PKC activity has been demonstrated in several human tumors, including pituitary adenomas. In comparison with normal pituitary tissue, increased PKC protein expression was identified in pituitary adenomas. More recently, it has been demonstrated that the specific PKC isoform that is overexpressed in pituitary tumors is PKC-a. Interestingly, this particular PKC isoform has been favored as the isoform that mediates the mitogenic functions of PKC. Not only was PKC-a found to be overexpressed in pituitary adenomas, but invasive adenomas also exhibited a conserved point mutation of the PKC-a gene.64 Cytogenetic Changes. Perhaps the most graphic evidence in support of the concept that neoplasia represents the successive accumulation of genetic alterations is derived from cytogenetic studies of tumor cells. In the case of pituitary adenomas, given their low intrinsic mitotic activity and overall benign constitution, metaphase preparations are not easily procured, and few direct cytogenetic analyses have been successfully performed. Of those that have, cytogenetic aberrations appeared more commonly among functioning pituitary adenomas than in endocrine-inactive ones.65,66 Overall, rearrangements involving chromosomes 17 and 19 were most commonly observed; trisomy of chromosome 7 as well as structural abnormalities of chromosomes 18p, 1, and 4 were observed less often. Given the limited number of cytogenetic analyses performed on pituitary adenomas to date, it is impossible to determine the pathophysiologic significance of the chromosomal aberrations so far cataloged. Whether such alterations are nonrandom events that contribute to tumor initiation and/or progression remains uncertain. Alternatively, such alterations may simply represent genetic instability inherent to neoplastic cells, being neither important nor informative to the evolution or progression of pituitary adenomas. GENERAL FEATURES OF PITUITARY ADENOMAS1,10 From clinical, pathologic, and biologic standpoints, tumors of the pituitary gland are a heterogeneous group. Underlying much of this heterogeneity is the fact that the normal adenohy-pophysis is composed of five principal cell types, each of which is susceptible to neoplastic transformation. The resulting adenoma tends to maintain the secretory activity nomenclature and some of the morphologic features of the cell of origin. Up to 70% of pituitary adenomas are hormonally active lesions that produce distinctive alterations in the endocrine equilibrium. The remainder are incapable of secreting a biologically active hormonal product and are referred to as nonfunctional. Further contributing to their overall heterogeneity are their variable and often unpredictable growth characteristics. Whereas some tumors exist largely as microadenomas, maintain a well-defined margin, and show little appreciable growth over time, others exhibit rapid proliferation rates and are markedly invasive or compressive of adjacent bony, neural, or vascular structures. This tremendous variation in biologic behavior cannot be predicted on morphologic grounds alone. Even the exceptional pituitary carcinoma, so defined on the basis of its metastatic capability, may appear entirely benign histologically. The clinical presentation of pituitary adenomas also is variable, although three principal patterns are recognized. The first involves pituitary hyperfunction, in the form of several characteristic hypersecretory states. Hypersecretion of prolactin, GH, ACTH, and, rarely, TSH, produces their corresponding clinical phenotypes: amenorrhea-galactorrhea syndrome, acromegaly or gigantism, Cushing disease, and secondary hyperthyroidism, respectively.1,4,67 By contrast, some pituitary adenomas produce symptoms of anterior pituitary deficiency, resulting from compression of the nontumorous pituitary or damage to the pituitary stalk or hypophysiotropic areas of the hypothalamus. This usually occurs insidiously in the context of large, nonfunctioning pituitary adenomas, but occasionally it may occur acutely in the setting of pituitary apoplexy. In the face of chronic compression, the various secretory cells of the pituitary differ in their functional reserve. Because gonadotropes are the most vulnerable, they usually are affected first, followed sequentially by thyrotropes and somatotropes. Corticotropes demonstrate the greatest functional resilience and generally are the last to be affected. Finally, the clinical presentation of a pituitary adenoma may be a neurologic one, with or without coexisting endocrinopathy. Given its strategic location at the skull base, a growing adenoma produces a predictable array of neurologic signs and symptoms. Suprasellar extension with compression of the optic chiasm results in a characteristic bitemporal hemianopic pattern of vision loss (see Chap. 20). Encroachment on the hypothalamus causes alterations of sleep alertness and behavior (see Chap. 9). Transgression of the lamina terminalis can bring these tumors into the region of the third ventricle, where obstruction to the flow of cerebrospinal fluid results in obstructive hydrocephalus. Lateral penetration of the tumor into the cavernous sinus occurs commonly, occasionally ensheathing the cavernous segment of the carotid artery and functionally compromising cranial nerves transiting the sinus. Some tumors extend in other directions and, if sufficiently large, can involve the anterior, middle, and, occasionally, posterior cranial fossae to produce a full spectrum of neurologic deficits. A final endocrinologic feature of pituitary adenomas, one that is common to many neoplastic, compressive, or infiltrative lesions affecting the sellar region, is moderate hyperprolactinemia. The blood prolactin levels are not extremely high, usually below 200 ng/mL. In addition, TRH stimulation, although not reliable as a test, may further increase the blood prolactin levels, whereas patients with prolactin-producing pituitary adenomas may exhibit no response or a blunted response to TRH. In patients with nonprolactin-producing tumors, the mechanism of hyperprolactinemia is explained by the so-called stalk section effect. Because prolactin secretion is suppressed by hypothalamic dopamine, interference with the synthesis, discharge, or adenohypophysial transport of dopamine may relieve pituitary prolactin cells from dopaminergic inhibition, resulting in increased prolactin release and hyperprolactinemia (see Chap. 13). As a rule, prolactin elevations in excess of 200 ng/mL are virtually always indicative of autonomous prolactin secretion from a neoplasm, either a prolactinoma or a plurihormonal pituitary tumor with a lactotropic component. MEANS OF CLASSIFICATION Since pituitary adenomas were first recognized, substantial effort has been devoted to classifying them into distinct groups. Endocrinologic classifications emphasize clinical manifestations of the disease or the dominant endocrine abnormality as assessed biochemically, most often with radioimmunoassay of blood hormone levels. Pathologic classifications hinge on structural features of adenoma cells and emphasize morphologic tumor markers. Modern pathologic classifications are based on sophisticated morphologic methods, such as immunocy-tochemistry analysis, ultrastructural analysis, immunoelectron microscopy, and morphometry. The conventional pathologic methods have been integrated with the powerful applications of molecular science. Methodologies such as in situ hybridization and Northern blot analysis permit the classification of pituitary adenomas on the basis of specific gene expression patterns (Fig. 11-6). Although these methods as well as those that catalog pituitary adenomas on the basis of specific genomic alterations are currently of research interest only, they will assume increasing importance in both the classification

of pituitary adenomas and the prediction of their biologic behavior.

FIGURE 11-6. Cellular distribution of pro-opiomelanocortin (POMC) mRNA in a corticotrope adenoma from a patient with Nelson syndrome as demonstrated by the method of in situ hybridization. The distribution of the black silver grains corresponds to the distribution of POMC mRNA. Original magnification 400.

Useful classifications are those that permit conclusions based on biologic behavior, prognosis, and response to various treatment modalities. Thus, pituitary adenomas are classified according to the staining affinities of the cell cytoplasm, size, growth pattern, endocrine activity, histologic features, hormone synthesis and content, granularity of cell cytoplasm, cellular composition and cytogenesis, or electron microscopic characteristics. On the basis of tinctorial features of the cell cytoplasm, acidophilic, basophilic, and chromophobic adenomas can be distinguished. Acidophilic adenomas were assumed to cause GH overproduction and, consequently, gigantism and acromegaly. Basophilic adenomas were thought to secrete ACTH and were linked to Cushing disease. Chromophobic adenomas were considered endocrinologically inactive tumors not associated with hormone excess. However, it has been clearly shown that acido-philic adenomas can synthesize hormones other than GH, such as prolactin, or can be endocrinologically inactive. Some basophilic adenomas represent silent corticotrope cell adenomas unaccompanied by ACTH excess or they produce hormones other than ACTH. Chromophobic adenomas may not be associated with hormone overproduction; however, this is not a general rule because more than half of all chromophobic adenomas are endocrinologically active tumors, secreting GH, prolactin, ACTH, TSH, FSH, LH, or the a subunit. The classification of pituitary adenomas on the grounds of the staining affinities of the cell cytoplasm has limited value because it does not consider hormone production, cellular composition, or cytogenesis, and because it fails to correlate structural features with secretory activity. GROWTH PATTERNS Based on size, pituitary adenomas can be divided into microadenomas and macroadenomas. Microadenomas measure <10 mm in greatest diameter and macroadenomas measure >10 mm. This classification is useful because a prognosis can be based on tumor size. Large tumors usually are difficult to remove completely, and recurrences are more frequent than with small, well-demarcated tumors confined to the sella. There are no light or electron microscopic differences between microadenomas and macroadenomas. Pituitary tumors grow either by expansion or by invasion and can be divided on the basis of growth patterns as follows: Grade 0: Intrapituitary microadenoma; normal sellar appearance. Grade 1: Intrapituitary microadenoma; focal bulging of sellar wall. Grade 2: Intrasellar macroadenoma; diffusely enlarged sella; no invasion. Grade 3: Macroadenoma; localized sellar invasion and/or destruction. Grade 4: Macroadenoma; extensive sellar invasion and/or destruction. Tumors are further subclassified on the basis of their extrasellar extensions, whether suprasellar or parasellar. Pituitary carcinomas (i.e., pituitary tumors with demonstrated metastatic spread) are sometimes designated as Grade 5. Intrapituitary and Intrasellar Adenomas. Intrapituitary adenomas are confined to the substance of the pituitary; intrasellar adenomas are located in the sella. These adenomas neither erode the sella nor invade parasellar tissue. They grow slowly, are well circumscribed, and are surrounded by a pseudocapsule consisting of condensed reticulin fibers and compressed rows of adenohypophysial cells; they have no fibrous capsule such as benign tumors may have elsewhere. Diffuse Adenomas. Diffuse adenomas are large and expansive; they may fill the entire sella and cause focal erosion of the sellar wall. Invasive Adenomas. Invasive adenomas grow more rapidly than do expansive adenomas. They have no sharp borders, they erode the sella and spread into neighboring tissue, they may infiltrate the sphenoid bone and invade the posterior lobe and cavernous sinus, and they may compress the optic nerve and other cranial nerves in the vicinity. Invasive adenomas may spread into the brain, penetrate into the third ventricle, and compress or destroy the hypothalamus and pituitary stalk. The aggressive behavior of invasive adenomas, although frequently translated into diminished surgical cure rates, is generally not reflected in their histologic features. Invasive adenomas exhibiting extreme local aggressiveness frequently maintain a deceptively innocent histologic appearance. In general, the usual morphologic markers of tumor aggressiveness (i.e., pleomorphism, nuclear atypia, increased cellularity, mitotic activity) correlate poorly with the invasive tendency, proliferative capacity, potential for recurrence, or overall biologic behavior of an adenoma. The incidence of local invasion depends not only on the different pituitary tumor types, but also on how the phenomenon is defined. Depending on the criteria used, invasion can be demonstrated radiologically, intraoperatively, and microscopically in 10%, 35%, and 90% of all pituitary adenomas, respectively.68 Not surprisingly, microscopic evidence of dural invasion increases with tumor size, being evident in 66% of microadenomas, 87% of macroadenomas, and 94% of macroadenomas with suprasellar extension. Because microscopic evidence of dural invasion is so ubiquitous a finding, it is a poor index of tumor aggressiveness. Therefore, it has become common practice to designate adenomas as invasive on the basis of radiologic or intraoperative evidence of gross invasion. The frequency of invasion among the various tumor types is presented in Table 11-2.

TABLE 11-2. Size and Frequency of Gross Local Invasion among the Major Pituitary Tumor Types

Because the histopathologic features of pituitary adenomas correlate poorly with their clinical aggressiveness, the development of a reliable and informative strategy to

predict their biologic behavior has remained one of the most inscrutable aspects of pituitary tumor biology. During the past decade, numerous strategies have been applied to the problem of predicting the proliferative potential of pituitary adenomas.10 Previously considered to be of research interest only, some of these methods are beginning to reach levels of clinical applicability. In attempting to distinguish mitotically active pituitary adenomas, which are prone to recurrence, from the more indolent ones, which are amenable to surgical cure, the presence of several cell-cyclespecific proliferation markers has been studied in pituitary adenomas. Two of these, Ki67 and proliferating cell nuclear antigen (PCNA), have been particularly informative from a clinical standpoint. The former is a cell-cyclespecific, nuclear-associated antigen of uncertain function whose expression is restricted to proliferating cells in the G1-to-M phase of the cell cycle. Thus, its immunochemical presence and relative abundance in tumor tissue provides some measure of the proliferative activity of a tumor. In one study, invasive adenomas expressed twice the amount of Ki67 protein as did noninvasive adenomas.69 Subsequent studies of Ki67 expression using MIB-1, an antibody that recognizes the Ki67 on better preserved and formalin-fixed preparations, have expanded our understanding of the cell kinetics of pituitary tumors.70 Specifically, (a) the growth fractions of pituitary tumors are low, most indices being <5%; (b) mean growth fractions are significantly higher among invasive adenomas and highest among pituitary carcinomas when compared with noninvasive adenomas (4.7%, 11.9%, and 1.4%, respectively); (c) virtually all noninvasive adenomas had growth fractions <3%, whereas those of most invasive adenomas and pituitary carcinomas were higher; and (d) the growth fraction of hormonally active pituitary tumors was significantly higher than that of endocrinologically inactive pituitary adenomas. Although these data suggest that a significant correlation does exist between Ki67 expression and pituitary tumor behavior, all reports to date have been retrospective in nature. Accordingly, prospective studies will be required to determine the long-term prognostic significance of this marker. PCNA is a critical accessory protein of the enzyme DNA polymerase-d. The former is an essential component of the replicon, the multimeric complex that polymerizes DNA nucleotides during leading-strand DNA replication just before cell division. Therefore, the nuclear expression of PCNA is a requirement for all replicating cells. PCNA expression has been studied in a large group of pituitary adenomas of all sizes and various immunotypes, including nonrecurrent and recurrent tumors.71 The expression of PCNA was significantly higher among recurrent pituitary adenomas than among nonrecurrent ones. Furthermore, the PCNA index was higher in larger tumors, particularly those exhibiting extrasellar extension. Among recurrent tumors, a higher PCNA index tended to be accompanied by a shorter disease-free interval. Flow cytometric analysis, although prognostically informative for many human epithelial malignancies, has been less than enlightening when applied to pituitary adenomas.23 Although ploidy analyses have shown many pituitary adenomas to have either diploid or aneuploid complements, neither have been reliably correlated with the invasiveness or potential for recurrence of an adenoma. Similarly, S-phase fractions of pituitary adenomas, as determined by flow cytometry, do not appear to be prognostically informative. The S-phase fraction of cycling cells also has been determined in situ after the in vivo administration of bromodeoxyuridine (BrDU), a thymidine analog. The incorporated BrDU in cycling cells is revealed by immunohistochemistry analysis using an anti-BrDU antibody. Because pituitary adenomas have relatively low proliferation rates overall, the S-phase fractions usually are small, generally <0.5%. The consistent finding that tumors associated with Nelson syndrome have the highest S-phase fractions validates the aggressiveness observed clinically in this group of tumors. The narrow range of observed S-phase fractions in pituitary adenomas overall, however, limits the sensitivity of this technique in distinguishing subtle differences in S-phase fractions between aggressive and nonaggressive pituitary adenoma variants. Carcinoma of the Pituitary Gland. Despite their epithelial nature and the regularity with which many pituitary adenomas exhibit aggressive local behavior, it remains a mystery why so few are capable of metastatic dissemination. Metastasizing pituitary tumors of adenohypophysial origin (i.e., pituitary carcinomas) are rare. A recent review identified 52 cases of pituitary carcinoma to which 12 new cases were added.72 Pituitary carcinoma is a precisely defined entity that includes only those tumors of adenohypophysial type with demonstrated craniospinal or systemic metastases.1,10 In contrast to most human carcinomas, the usual histologic markers of malignancy (nuclear atypia and pleomorphism, mitotic activity, necrosis, hemorrhage, invasiveness) are insufficient to permit an unqualified diagnosis of pituitary carcinoma; these features are commonly seen in ordinary pituitary adenomas. Instead, the diagnosis is predicated on tumor behavior, being relatively independent of histologic features. Among reported cases of pituitary carcinoma, metastatic dissemination most often involved the cerebrospinal fluid axis. Numerous extraneural metastatic sites have also been reported, including bone, liver, lymph nodes, lung, kidney, and heart. Pituitary carcinomas primarily affect adults, although their clinical presentation is variable. In some patients, the initial course is indistinguishable from that of a benign pituitary adenoma. Local invasion is variably present, and the histologic characteristics of the tumor may be entirely benign. A protracted course, often punctuated by multiple local recurrences, is then followed by metastatic dissemination. In many such cases, a clear escalation in histologic abnormalities is observed when primary tumors are compared with metastatic deposits. In this setting, the process appears to be one of malignant transformation in a preexisting benign adenoma. Alternatively, the behavior of other pituitary carcinomas is indicative of de novo malignancy. Such tumors are biologically malignant from the onset, beginning as rapidly growing, relentlessly invasive, cytologically atypical tumors that promptly give way to metastatic dissemination. Although both hormonally active and nonfunctioning pituitary carcinomas have been described, the former appear to pre-dominate. Pituitary carcinomas composed of ACTH cells (particularly in the context of Nelson syndrome), prolactin cells, and GH cells are most frequently represented. As a rule, metastatic deposits should retain the immunotype of the primary tumor. Because the pituitary is often the recipient of metastatic carcinomas originating systemically, the far more common occurrence of metastatic deposits to the pituitary must be excluded by careful clinical and pathologic examination before arriving at a diagnosis of primary pituitary carcinoma. The factors that underlie the capacity for metastatic dissemination in pituitary carcinomas remain obscure. Their mode of spread, however, is more fully understood. Craniospinal involvement appears to begin with invasion of the subarachnoid space and eventual dissemination by cerebrospinal fluid flow. Intracranial deposits involving brain parenchyma likely develop as the result of tumor permeation into perivascular (Virchow-Robin) spaces or by venous sinus invasion. Extracranial spread of pituitary carcinomas involves both hematogenous and lymphatic routes. Invasion of the cavernous sinus provides the necessary venous access for transport of tumor cells to the internal jugular vein through the petrosal system. Whereas the pituitary lacks lymphatic drainage, invasion of the tumor into the skull base provides access to the rich lymphatic network, permitting systemic dissemination. Death from pituitary carcinoma usually is the consequence of extensive intracranial disease and not the result of extraneural deposits per se. Because neither the presence nor the treatment of systemic metastases appears to alter the prognosis of pituitary carcinomas, the frequency of metastases, particularly nonfunctioning ones, is likely underreported. CLASSIFICATION OF PITUITARY ADENOMAS BY HORMONE PRODUCTION1,2,10 Pituitary adenomas can be classified according to hormone production, on the basis of clinical presentation and blood hormone levels as assessed by radioimmunoassay. Thus, functioning adenomas, which produce GH, prolactin, ACTH, TSH, FSH, LH, or the a subunit, can be distinguished clinically from non-functioning adenomas, which are not associated with hormone excess. It is evident that all known adenohypophysial hormones can be secreted by pituitary adenoma cells. A unifying pituitary adenoma classification encompasses histologic, immunocytochemical, and ultrastructural features of tumor cells and emphasizes the importance of hormone production, cellular composition, and cytogenesis. Such a classification stresses structurefunction relationships and correlates the morphologic appearance of adenoma cells with their secretory activity. All known adenohypophysial cell types can give rise to adenoma. The prevalence of each type is as follows: prolactin cell adenoma, 26%; null cell adenoma, 17%; GH cell adenoma, 14%; corticotrope cell adenoma, 15%; plurihormonal adenoma, 13%; oncocytoma, 6%; gonadotrope cell adenoma, 8%; and thyrotrope cell adenoma, 1%. The morphologic features of pituitary adenomas can reveal their cellular composition, hormone production, and, in some cases, cytogenesis. Growth Hormone Cell Adenomas. Growth hormone cell adenomas are associated with elevated blood GH concentrations and acromegaly or gigantism (see Chap. 12). On light. microscopic examination they appear either acidophilic or chromophobic. On electron microscopic examination, acidophilic adenomas are densely granulated, and chromophobic adenomas are sparsely granulated. Separating these two subtypes has practical importance. Densely granulated tumors have a better prognosis; they have a slower growth rate, tend to be easier to remove surgically, and recur less frequently after surgery. The densely granulated adenoma cells resemble GH cells seen in the nontumorous pituitary and are characterized by well-developed, rough-surfaced endoplasmic reticulum, a conspicuous Golgi apparatus, and numerous spherical, evenly electron-dense secretory granules measuring 250 to 650 nm (Fig. 11-7). The cells constituting sparsely granulated adenomas differ from GH cells of the nontumorous adenohypophysis. They possess irregular indented nuclei, rough-surfaced and smooth-surfaced endoplasmic reticulum, a conspicuous Golgi apparatus, several centrioles and cilia, fibrous bodies composed of microfilaments and smooth-walled tubules, and a few spherical, evenly electron-dense secretory granules measuring 100 to 300 nm (Fig. 11-8). No correlation has been noted between serum GH levels and granularity of the cell cytoplasm; both densely and sparsely granulated GH cell adenomas can be associated with markedly or slightly elevated GH levels as well as with severe and rapidly progressing or mild and slowly advancing acromegaly.

FIGURE 11-7. Ultrastructural features of densely granulated growth hormone cell adenoma. Note well-developed, rough-surfaced endoplasmic reticulum (arrows) and numerous large secretory granules. 5600

FIGURE 11-8. Sparsely granulated growth hormoneproducing adenoma with fibrous bodies (arrows). The secretory granules are sparse and small. 5600

Prolactin Cell Adenomas. Prolactin cell adenomas are associated with hyperprolactinemia, amenorrhea, galactorrhea, and infertility (see Chap. 13). In men, symptoms include decreased libido and impotence. In some cases, endocrine manifestations are subtle or absent, and only local symptoms call attention to a pituitary tumor. Markedly elevated serum prolactin levels (i.e., >150200 ng/mL) confirm the clinical diagnosis of a prolactin-producing pituitary adenoma. Prolactin-producing adenomas can occur at any age but most often are diagnosed in young women. Based on surgical material, women are more frequently affected than men, but in unselected adult autopsies, no sex difference can be demonstrated. Since the introduction of bromocriptine therapy, the number of prolactin-producing tumors treated surgically has declined. Patients with prolactin cell adenomas can be effectively treated with various dopaminergic agonists. Depending on the sex of the patient, the treatment abolishes the amenorrhea, stops the galactorrhea, increases the libido, and cures the infertility (see Chap. 13); serum prolactin levels show a marked decline. The tumor regresses, as assessed by various imaging techniques. Local symptoms, such as visual disturbances, often disappear. The effects of treatment are reversible; if drug administration is discontinued, the tumor regrows, serum prolactin levels rise, and clinical symptoms reappear. As a result of bromocriptine medication, the tumor cells themselves decrease in size, indicating that tumor regression is mainly a result of a reduction in cell volume. In addition to an overall reduction in cytoplasmic volume, there also is a significant reduction in the volume density of the hormone's synthetic and secretory apparatus (rough endoplasmic reticulum and Golgi complex). Overall, this gives a hypercellular appearance to the tumor. The tumoristatic effects of bromocriptine are especially well visualized by electron microscopic examination, in which cells become irregularly shaped, contain heterochromatic nuclei, and maintain a scant cytoplasm with markedly involuted Golgi complexes. The size and volume density of secretory granules remain unaffected. All these ultrastructural changes are fully reversible on discontinuation of the drug. Prolonged exposure to bromocriptine also may result in fibrotic change within the tumor, an alteration that may adversely complicate surgical extirpation. The presence of cell necrosis after bromocriptine administration also has been reported. Whether bromocriptine is cytotoxic is difficult to assess because foci of necrosis, hemorrhage, and fibrosis may occur in prolactin-producing adenomas without dopaminergic agonist treatment. Histologically, prolactin cell adenomas most often are chromophobic or slightly acidophilic, exhibiting distinct prolactin immunostaining in the Golgi complex and secretory granules. On electron microscopic examination, prolactin cell adenomas can be separated into densely granulated and sparsely granulated variants. Densely granulated prolactin cell adenomas are rare. The adenoma cells resemble nontumorous resting prolactin cells and are characterized by prominent rough-surfaced endoplasmic reticulum, conspicuous Golgi complexes, and numerous spherical, oval, or irregularly shaped secretory granules measuring up to 700 nm. Sparsely granulated prolactin cell adenomas are the most frequent tumor type in the human pituitary. The adenoma cells possess a prominent rough-surfaced endoplasmic reticulum, a conspicuous Golgi apparatus, and sparse, spherical, and oval, or irregularly shaped, evenly electron-dense secretory granules measuring 150 to 300 nm (Fig. 11-9). The rough-surfaced endoplasmic reticulum membranes form concentric cytoplasmic whorls, called nebenkerns. Another characteristic ultrastructural feature of prolactin cell adenomas is the presence of misplaced exocytosisextrusion of secretory granules on the lateral side of the cell, distant from capillaries and intercellular extensions of the basement membrane. Granule extrusion also occurs on the capillary side of prolactin cells.
HISTOPATHOLOGY.

FIGURE 11-9. Sparsely granulated prolactin cell adenoma with prominent rough-surfaced endoplasmic reticulum membranes (arrows) and misplaced exocytosis (extrusion of secretory granules into the extracellular space, distant from the basement membrane; arrowheads). 7800

Prolactin cell adenomas may produce an amyloid-like substance and exhibit various degrees of calcification, sometimes so extensive as to be visible with imaging techniques. Amyloid deposition and calcification are characteristic but not pathognomonic; they most frequently occur in prolactin cell adenomas but occasionally may be present in other adenoma types. Corticotrope Adenomas. Corticotrope adenomas produce ACTH and other peptides of the pro-opiomelanocortin molecule, such as b-LPH and endorphins. They may secrete hormones excessively and may be associated with Cushing disease or Nelson syndrome (see Chap. 75). In Cushing disease, the excess ACTH stimulates the adrenal cortex and causes various degrees of hypercortisolism. Historically, bilateral adrenalectomy was undertaken because the ACTH-producing pituitary adenoma either was unrecognized or was unsuccessfully treated. With the development of increasingly precise imaging technology and superior microsurgical techniques, bilateral adrenalectomy is now a procedure of last resort that is rarely required in the management of Cushing disease. Accordingly, new cases of Nelson syndrome are becoming increasingly uncommon. When bilateral adrenalectomy is necessary to control refractory Cushing disease, the clinical picture is characteristic. As hypercortisolism regresses, the tumor frequently behaves in a disinhibited fashion, tending to grow more rapidly and to produce large amounts of ACTH and other pro-opiomelanocortinderived peptides. Patients become hyperpigmented, and the tumor tends to be far more aggressive and invasive than are those corticotropic adenomas that have an intact pituitary-adrenal axis. Furthermore, only a few patients with Nelson syndrome are cured by surgery and radiotherapy; up to 20% of these

patients succumb to uncontrolled local tumor growth. Such aggressive behavior has been ascribed to the loss of negative glucocorticoid feedback resulting from the adrenalectomy; this suggests that corticotrope adenomas, as a whole, are not entirely autonomous, but are subject to feedback and modulation of their secretory activity and growth rate by glucocorticoid hormones. Importantly, more than 80% of the adenomas responsible for Cushing disease are microadenomas. Because many are only a few millimeters in diameter, they frequently evade detection, even with the most sophisticated imaging techniques. Given their small size, these tumors present technical difficulties, both to surgeons and to pathologists. From a surgical standpoint, corticotrope microadenomas rarely are situated on the surface of the gland and are exposed only after a thorough dissection of a seemingly normal gland. Most arise in the midline of the pituitary, within the so-called mucoid wedge, an area in which corticotropes are most numerous. The surgical specimen provided to the pathologist frequently is small and fragmented; as a result, considerable patience and skill are required to differentiate adenomatous elements from the normal glandular tissue. Whereas 10% to 15% of microadenomas demonstrate local invasion, fully 60% of macroadenomas are grossly invasive. Histologically, corticotrope adenomas usually are basophilic and exhibit various degrees of PAS and lead-hematoxylin positivity. Immunoperoxidase staining reveals ACTH and other fragments of the pro-opiomelanocortin molecule in the cytoplasm of adenoma cells. On electron microscopic examination, adenomatous corticotropes often resemble nonad-enomatous corticotropes; they possess well-developed, rough-surfaced endoplasmic reticulum, prominent Golgi apparatuses, and numerous spherical secretory granules that vary in electron density, often line up along the cell membrane, and measure 250 to 700 nm (Fig. 11-10). In corticotrope adenomas removed from patients with Cushing disease, bundles of microfilaments usually are present, whereas in adenomas removed from patients with Nelson syndrome, microfilaments are inconspicuous or absent. Otherwise, no ultrastructural differences exist between the corticotrope adenomas of patients with Cushing disease and Nelson syndrome.
HISTOPATHOLOGY.

FIGURE 11-10. Electron microscopic appearance of densely granulated corticotrope cell adenoma. Note the irregular or dented shape of secretory granules and bundles of type I filaments (arrows). 9800.

A few actively secreting corticotrope adenomas are chromophobic and possess only sparse, fine, PAS-positive cytoplasmic granules. These tumors immunostain for ACTH and related peptides, indicating that they arise in and consist of corticotropes. On electron microscopic examination, chromophobic corticotrope adenoma cells are sparsely granulated and appear less differentiated than their basophilic, densely granulated counterparts. Chromophobic tumors usually are larger, grow faster, tend to be invasive, and recur more frequently than basophilic adenomas, which generally are small, measuring only a few millimeters in diameter. Silent Corticotrope Adenomas.1,2,10,22 Silent corticotrope adenomas immunostain for ACTH and related peptides but are not associated with clinical or biochemical evidence of ACTH excess. On light microscopic examination, these tumors are basophilic or chromophobic, show various degrees of PAS positivity, and immunostain for ACTH, b-LPH, and endorphins. On electron microscopic examination, silent corticotrope cell adenomas are a heterogeneous group. In some cases, the ultrastructural features of tumor cells are indistinguishable from those of actively secreting tumors, such as are associated with Cushing disease or Nelson syndrome. In other cases, an increase in the size and number of lysosomes, crinophagy (uptake of secretory granules by lysosomes), and marked underdevelopment or involution of the Golgi apparatus suggest a defect in various steps involving hormone synthesis, packaging, or discharge. Two distinct silent corticotrope adenomas, designated silent corticotrope adenomas subtypes I and II, have been identified as specific clinicopathologic entities. A third, no longer considered corticotropic in nature, is designated silent subtype III. For the most part, silent subtypes I through III present clinically as large and invasive nonfunctioning sellar masses.1,17 In some instances, hyperprolactinemia is present, sometimes at a level higher than that attributable to the stalk section effect. Accordingly, it has been suggested that some silent subtypes may facilitate prolactin release from nontumorous lactotropes or are themselves capable of prolactin secretion. These findings are particularly applicable to silent subtype III, a tumor frequently seen in women, and therefore often is diagnosed before operation as a prolactinoma. The reason that some silent subtype III tumors have been responsible for acromegaly remains a total mystery. A peculiarity of silent subtype tumors relates to their propensity to undergo apoplectic hemorrhage. In the experience of the authors, more than 40% of silent subtype adenomas presented in this fashion.22 In all, the silent subtype tumors are enigmatic entities, whose cytogenesis and overall biology warrants further study. Thyrotrope Adenomas. With fewer than 100 cases reported, TSH-secreting adenomas are the least common pituitary tumor phenotype, accounting for only 1% of all pituitary adenomas. Of reported cases, most thyrotrope adenomas have been large, aggressive macroadenomas, both compressive and invasive of surrounding structures.73 Usually, the accompanying clinical history is remarkable for some form of thyroid dysfunction. It once was believed that most arose in the context of long-standing primary hypothyroidism, presumably by way of feedback inhibitory loss, induction of thyrotrope hyperplasia, and, later, adenoma formation. Although such a perspective was compatible with earlier experimental studies in which thyroidectomy induced pituitary thyrotrope adenomas in rodents, careful clinicopathologic correlations of human thyrotrope adenomas indicate an alternate sequence of likely events.10,23 In many patients, the initial manifestations appear to be those of hyperthyroidism and goiter, events wholly compatible with TSH hypersecretion by the tumor. Because secondary (i.e., pituitary-dependent) hyperthyroidism previously was not a well-recognized condition, many such patients were incorrectly thought to have primary hyperthyroidism and were subjected to some form of thyroid ablation. This served to ameliorate symptoms, but sometimes was followed later by accelerated tumor growth, optic nerve compression, or recurrence of the hyperthyroid state. Only then was the pituitary correctly identified as the site of pathologic involvement. The invasive nature of these tumors appears to be related to two factors, the first of which is the typical diagnostic delay. A more cogent factor, however, is the loss of feedback inhibition. In the same way that endorgan ablation contributes to tumor aggressiveness in the context of Nelson syndrome, similar disinhibiting influences may be operative in the progression of TSH adenomas in the setting of prior thyroidectomy. The routine availability of sensitive TSH assays coupled with general awareness of the thyrotrope adenoma as a potential, though rare, cause of hyperthyroidism should permit more expeditious diagnosis of this tumor type, perhaps while it is still in the microadenoma stage.67 Furthermore, thyrotrope adenomas commonly cosecrete in excess a free a subunit that, if present, may be a helpful diagnostic clue suggesting a pituitary source over a primary thyroid cause of hyperthyroidism (see Chap. 15 and Chap. 42). On light microscopic examination, thyrotrope adenomas are chromophobic, containing a few small cytoplasmic granules mainly at the cell periphery that stain for PAS, aldehyde fuchsin, and aldehyde thionin. TSH can be demonstrated immunocytologically in the cytoplasm of the adenoma cells. Occasionally, immunostaining shows only slight or no TSH immunopositivity, suggesting that little hormone is stored in the cytoplasm or that abnormal TSH is produced that is not immunoreactive but may have bioactivity. On electron microscopic examination, thyrotrope adenomas consist of elongated, angular, or irregular cells with long cytoplasmic processes, scanty rough-surfaced endoplasmic reticulum, an inconspicuous Golgi apparatus, and numerous microtubules. Secretory granules are sparse and spherical, vary slightly in electron density, line up along the cell membrane, and measure 50 to 200 nm. Thyrotrope adenomas often differ in their ultra-structural appearance from nontumorous thyrotropes. In some cases, they consist of highly differentiated thyrotropes with an abundant, slightly dilated, rough-surfaced endoplasmic reticulum, a conspicuous Golgi apparatus, and varying numbers of secretory granules in the range of 50 to 250 nm.
HISTOPATHOLOGY.

Gonadotrope Adenomas. Gonadotrope adenomas produce FSH, FSH and LH, or, rarely, LH alone. They are identified more frequently in older patients. The clinical manifestations of gonadotrope cell adenomas are not clearly defined; patients may show varying degrees of hypogonadism, decreased libido, impotence, and elevated serum FSH/LH concentrations. On light microscopic examination, gonadotrope adenomas appear chromophobic and may contain sparse PAS-positive cytoplasmic granules. Immunostaining reveals FSH, LH, or both in the cytoplasm of adenoma cells. On electron microscopic examination, sexual dichotomy is evident. In men, adenomatous gonadotropes usually are less differentiated and have a few rough-surfaced endoplasmic reticulum membranes, a moderately developed Golgi apparatus, several microtubules, and sparse, spherical secretory granules that vary slightly in electron density, line up along the cell membrane, and measure 100 to 300 nm. In women, adenomatous gonadotropes are more differentiated and resemble their non-tumorous counterparts. The prominent honeycomb-like Golgi complex consists of several dilated sacculi and vesicles containing a few immature secretory granules. Secretory granules are sparse and randomly distributed, vary slightly in electron density, and measure 50 to 150 nm.
HISTOPATHOLOGY.

Null Cell Adenomas. Null cell adenomas have no histologic, immunocytologic, or electron microscopic markers and are not associated clinically and biochemically with any known hormone excess. Although these tumors are endocrinologically inactive and contain no known hormones, they possess the organelles necessary for hormone secretion and have secretory granules.1 It may be that these tumors produce biologically inactive hormone fragments, precursor molecules, or hormones unidentified at present. Together, null cell adenomas and their oncocytic variants (discussed later) account for almost one-fourth of all pituitary adenomas. Both are nonfunctioning sellar masses and are typically seen during middle or old age. They tend to be slow-growing lesions, with some likely growing for years subclinically before manifesting themselves clinically. Despite the regularity with which these tumors are encountered in clinical practice, and the fact that their existence has been known for almost two decades, fundamental questions concerning their causation, cytogenesis, and biology remain unanswered.2,10,74 That these tumors frequently share morphologic similarities with undifferentiated gonadotrope adenomas has fueled speculation that null cell adenomas may be neoplastic offshoots of gonadotropic lineage. Compelling support for such a notion was provided by the finding that 80% of null cell adenomas and oncocytomas express glycoprotein hormone genes.10 Furthermore, both gonadotropin release and gonadotropin-releasing hormone responsiveness have been demonstrated in null cell adenomas maintained in tissue culture. Alternatively, there is preliminary evidence favoring the existence of nonneoplastic null cells scattered throughout the normal pituitary.75 It is speculated that such normal null cells may be transitional, undifferentiated, or precursor cells that are capable of shifting from hormonally inactive to hormonally active states. Should null cells be validated as cellular constituents of the normal pituitary, null cell adenomas could be envisioned as their neoplastic derivatives. On light microscopic examination, null cell adenomas are chromophobic; immunocytologically, they contain no adenohypophysial hormones. In many tumors, however, small groups of adenoma cells or randomly scattered individual cells immunostain for one or more pituitary hormones, most frequently FSH or the a subunit, less frequently TSH, LH, and prolactin, and occasionally GH or ACTH. Consistent with these findings, in vitro studies reveal that most null cell adenomas produce FSH, TSH, LH, or the a subunit; these hormones can be demonstrated by radioimmunoassay in tumor culture. It is not clear whether this means multidirectional differentiation from an uncommitted precursor, tumor cell heterogeneity, or gradual dedifferentiation. On electron microscopic examination, null cell adenomas are characterized by closely apposed, polyhedral or irregularly shaped cells with pleomorphic or indented nuclei and poorly developed cytoplasm possessing a few rough endoplasmic reticulum profiles, a moderately developed Golgi apparatus, and microtubules, which may be abundant in some cells. The secretory granules are sparse and spherical, vary slightly in electron density, line up along the cell membrane, and measure 100 to 200 nm (Fig. 11-11). Oncocytic transformation (i.e., the increase of cytoplasmic volume occupied by mitochondria) is common in null cell adenomas.
HISTOPATHOLOGIC CHARACTERISTICS.

FIGURE 11-11. Null cell adenoma. The small adenoma cells contain poorly developed cytoplasmic organelles and sparse, small secretory granules. 9800

Because of the lack of endocrine symptoms, some null cell adenomas are recognized only when they enlarge and spread outside the sella, causing local symptoms such as visual disturbances, headache, or injury of cranial nerves. Oncocytomas. Oncocytomas represent the oncocytic variant of null cell adenomas. The term oncocytosis is used to describe tumor cells that exhibit intracellular mitochondrial accumulation. Thus, oncocytomas differ from null cell adenomas only insofar as the cells of the former contain massive numbers of large mitochondria. Both tumors share a similar clinical profile that is dominated by the neurologic and endocrinologic sequelae of an expansile, nonfunctioning sellar mass. Like null cell adenomas, oncocytomas are unaccompanied by clinical or biochemical evidence of oversecretion of adenohypophysial hormones, have no morphologic markers that would reveal their cytogenesis, occur most frequently in older men and women, and are rarely diagnosed in patients younger than 40 years. On light microscopic examination, oncocytomas are chromophobic or acidophilic. The acidophilia is not the result of staining of secretory granules but of the uptake of acid dyes by accumulating mitochondria. Immunostaining of oncocytoma cells fails to demonstrate pituitary hormones. In many cases, however, small groups of adenoma cells or randomly scattered individual cells immunostain for FSH, the alpha subunit, LH, or TSH, indicating that oncocytoma cells do not lose their potential to produce hormones or that they can differentiate to a hormone-producing cell line. Electron microscopic examination reveals abundant cytoplasmic mitochondria (Fig. 11-12), which may be extensive, filling as much as 50% of the cytoplasm (compared with ~8% normally). In some tumors, a transition can be seen between null cell adenoma and oncocytoma. The authors use the term oncocytoma only when abundant mitochondria are evident in practically every adenoma cell.
HISTOPATHOLOGY.

FIGURE 11-12. Pituitary oncocytoma showing abundance of mitochondria. 7000

Plurihormonal Adenomas. Plurihormonal adenomas produce more than one hormone and can be divided into monomorphous and plurimorphous types. Monomorphous plurihormonal adenomas consist of one morphologically distinct cell type that produces two or more hormones; the cell may differ morphologically from known adenohypophysial cells. Plurimorphous plurihormonal adenomas are composed of two or more morphologically distinct cell types, each producing different hormones; they are similar in ultrastructural appearance to their nontumorous counterparts. Immunocytologic techniques are required to establish the diagnosis. Electron microscopic examination may fail to reveal the cellular origin of the adenoma because ultrastructural features may not be distinct or the tumor may consist of cells not recognized in nontumorous adenohypophyses. In the experience of the authors, 12% of surgically removed pituitary adenomas are plurihormonal. The most frequent hormonal combination produced is GH and prolactin. Three morphologically distinct adenoma types produce GH and prolactin simultaneously: acidophil stem cell adenoma, mammosomatotrope adenoma, and mixed GH cellprolactin adenoma. Acidophil stem cell adenomas grow rapidly, often spreading into neighboring tissues. They are associated with various degrees of hyperprolactinemia. In some patients, clinical features of acromegaly may be apparent despite normal serum GH levels. Acidophil stem cell adenomas are monomorphous, bihormonal tumors that consist of one cell type, which is assumed to represent the common progenitor of GH cells and prolactin cells. Immunocytologic techniques demonstrate both prolactin and GH in the cytoplasm of the same adenoma cells. Immunostaining for GH often is weak or absent. On electron microscopic examination, acidophil stem cell adenomas are composed of closely apposed elongated cells with irregular nuclei and well-developed cytoplasm containing dispersed, short profiles of rough-surfaced endoplasmic reticulum, an inconspicuous Golgi apparatus, fibrous bodies containing microfilaments and smooth-walled tubules, multiple centrioles and cilia, and sparse, irregular secretory granules measuring 100 to 300 nm. Some exocytosis may be evident. Oncocytic
ACIDOPHIL STEM CELL ADENOMAS.

change and mitochondrial gigantism occur in most tumors. The correlation between tumor size and blood prolactin concentrations, which is apparent in patients with sparsely granulated prolactin cell adenomas, is often absent in patients with acidophil stem cell adenomas; relatively large tumors may be accompanied by only slight or moderate hyperprolactinemia. Mammosomatotrope cell adenomas are slowly growing tumors accompanied by elevated serum GH concentrations, acromegaly, and, in some cases, mild hyperprolactinemia. These monomorphous, bihormonal tumors consist of acidophilic cells. Immunocytologic methods demonstrate GH and prolactin in the cytoplasm of the same adenoma cells. On electron microscopic examination, the adenoma cells appear to be well differentiated and resemble densely granulated GH cells. The secretory granules are often irregular; they may be evenly electron dense or have a mottled appearance, and they measure 200 to 2000 nm. Exocytosis and large extracellular deposits of secretory material are characteristic features.
MAMMOSOMATOTROPE CELL ADENOMAS.

Mixed GH cellprolactin cell adenomas are associated with elevated serum GH levels, acromegaly, hyperprolactinemia, and, occasionally, galactorrhea, amenorrhea, decreased libido, and impotence. These bimorphous, bihormonal tumors are composed of two morphologically distinct cell types: densely or sparsely granulated GH cells and prolactin cells. The two cell types form small groups; in several areas, individual cells are interspersed. Immunostaining demonstrates GH and prolactin in the two different cell populations. Electron microscopic examination shows two morphologically distinct cell types. Every combination may occur; most frequently, densely granulated GH cells and sparsely granulated prolactin cells are identified.
MIXED GROWTH HORMONE CELLPROLACTIN CELL ADENOMAS.

Occasional, unusual plurihormonal pituitary adenomas produce bizarre combinations of two or more hormones, such as GH and TSH; prolactin and TSH; GH, prolactin, and TSH; and, less frequently, GH, prolactin, and ACTH, or GH, prolactin, FSH/LH, and the a subunit. Such tumors may be monomorphous or plurimorphous. The cell type or types constituting the tumors often cannot be identified, even with detailed electron microscopic investigation. In a few cases, however, two or more ultrastructurally distinct cell types resembling their nontumorous counterparts can be recognized. The hormone content and ultrastructural features of adenomas cannot always be correlated.
UNUSUAL PLURIHORMONAL ADENOMAS.

On light microscopic examination, the unusual plurihormonal adenomas consist of chromophobic, acidophilic, or basophilic cells, or a mixture of cells that stain differently with different histologic techniques. Clinically and biochemically, the secretion of several hormones may be apparent; acromegaly may be accompanied by hyperthyroidism, hypercorticism, or hyperprolactinemia. Some components may be silent. Immunostaining demonstrates hormones in the cell cytoplasm, but hormone production is not always reflected in hypersecretory symptoms clinically, or in increased serum hormone levels. Plurihormonal adenomas, which are difficult to classify, clearly show that the one cellone hormone theory, which has dominated pituitary cytophysiology and cytopathology for many years, is oversimplified and requires revision. Beyond the conceptual importance underlying the phenomenon of plurihormonality is a potentially important clinical issue. Although unproven, there is some suggestion that some plurihormonal tumors behave more aggressively than do their monohormonal counterparts.10,75,76 In the case of other endocrine tumors (e.g., pancreatic tumors, medullary carcinoma of the thyroid), plurihormonal tumors are thought to portend a more malignant course than that of monohormonal tumors. Evidence in support of a similar occurrence in the pituitary remains inconclusive. It is known, however, that most plurihormonal pituitary adenomas are macroadenomas at presentation, even in the presence of a hypersecretory syndrome. Further-more, ~50% of all plurihormonal pituitary adenomas are grossly invasive at the time of diagnosis.76 MALIGNANT PITUITARY LESIONS PRIMARY MALIGNANT NEOPLASMS Primary malignant neoplasms of the hypophysis include carcinomas and sarcomas; they are extremely rare. Primary Adenohypophysial Carcinomas. Primary adenohypophysial carcinomas, which are derived from anterior pituitary cells, may secrete GH, prolactin, or ACTH or may not be associated with hormone production. They are rare and were discussed earlier. Electron microscopic and immunocytologic studies fail to distinguish between benign and malignant tumors. Sarcomas. Sarcomas of the adenohypophysis include fibrosarcoma, osteosarcoma, and undifferentiated sarcoma. With few exceptions, virtually all have occurred after radiotherapy for either pituitary adenoma, craniopharyngioma, or retinoblastoma. Fibrosarcomas, in particular, are most commonly the consequence of radiotherapy to a pituitary adenoma. Histologically, these tumors exhibit marked cellular and nuclear pleomorphism, replete with mitotic figures, areas of necrosis, and hemorrhage. In many instances, the sarcomatous component can be seen to be intimately admixed within the substance of a persistent pituitary adenoma. Thus, it has been suggested that radiotherapy induces fibrosarcoma formation by transforming fibroblastic elements within the original adenoma. The latency period for sarcomatous transformation is variable; an average period of 11 years has been reported.10,19 Postirradiation sarcomas are virtually always high-grade malignancies typified by rapid growth, relentless local invasion, and a survival period rarely exceeding a few months. SECONDARY NEOPLASMS Secondary neoplasms of the pituitary most often are found incidentally at autopsy and are not associated with clinical symptoms or biochemical abnormalities. They occur in 1% to 5% of patients with cancer. The most commonly observed endocrine abnormality is diabetes insipidus (see Chap. 26), which occurs in patients with metastatic carcinoma of the posterior lobe, hypophysial stalk, or hypothalamus. Compression or destruction of the production site of hypothalamic releasing and suppressing hormones or interference with adenohypophysial blood flow (either by blocking transport of hypothalamic hormones to the adenohypophysis or by inducing ischemia) also may account for the development of endocrine symptoms. Local symptoms may be apparent; however, anterior hypopituitarism is rare because a substantial part of the adenohypophysis must be destroyed before a decrease in adenohypophysial hormone secretion becomes manifest clinically and biochemically. Hypophysial metastases usually occur at an advanced stage of neoplastic disease, when the malignant process involves several organs. Affected patients rarely live long enough to develop anterior hypopituitarism. Only rarely are symptoms of pituitary metastases the first manifestation of a systemic malignancy.77 Hypophysial metastases may come from different primary sources, such as carcinomas of the bronchus, colon, prostate, larynx, or kidney; malignant melanoma; sarcomas; and hematologic malignancies. In the last of these categories, plasmacytoma is notorious for its periodic presentation in the sellar region. Many plasmacytomas of the sellar region have occurred in the absence of known systemic disease, presenting as seemingly ordinary pituitary adenomas. Regardless of therapy, most eventually evolve into full-blown multiple myeloma. For women, carcinoma of the breast is the most common primary lesion that metastasizes to the pituitary. In men, carcinoma of the lung is the most culpable primary tumor. Metastases to the posterior lobe are more frequent than to the anterior lobe, presumably because the posterior lobe has a rich direct arterial blood supply. Metastatic tumor deposits usually occur first in the pituitary stalk or posterior lobe and permeate the anterior lobe, either by direct extension from the hypophysial stalk or posterior lobe, or through the portal circulation, by long or short portal vessels. However, isolated metastases may occur in the anterior lobe, indicating that the secondary tumor in the adenohypophysis is not invariably the result of prior involvement of the posterior lobe and hypophysial stalk; in the genesis of adenohypophysial metastases, routes other than the portal circulation must be considered. Tumor cells may also spread from perihypophysial tissues to the pituitary. In patients with disseminated carcinomatosis, infiltration of long portal vessels by carcinoma cells leads to vascular occlusion and subsequent ischemia, causing adenohypophysial infarcts. The necrotic foci are not large enough to cause hypopituitarism. CRANIOPHARYNGIOMAS Craniopharyngiomas are histologically benign tumors that are thought to arise from remnants of the Rathke pouch. Such embryonic squamous cell nests extend from the tuber cinereum to the pituitary gland, presumably along the track of an incompletely involuted hypophysial-pharyngeal duct. By virtue of their location, craniopharyngiomas simultaneously compromise the function of several intracranial structures and produce numerous clinical effects, including vision loss, anterior and posterior pituitary dysfunction, and increased intracranial pressure. Despite their benign nature, craniopharyngiomas can offer considerable resistance to successful treatment. Craniopharyngiomas account for ~3% of intracranial tumors. They primarily affect children, for whom they represent the most common nonglial brain tumor and 10% of all intracranial neoplasms. The age-related incidence of craniopharyngiomas is bimodal, with a major, early peak between 5 and 10 years of age, and a second, smaller peak between 50 and 60 years of age. A slight male preponderance has generally been noted. Topologically, most craniopharyngiomas are suprasellar in location. Those so situated can compress the optic apparatus, cause hydrocephalus by indenting the third ventricle, encroach on hypothalamic structures, distort the infundibulum and pituitary, and penetrate cerebrospinal fluid spaces to gain access to the middle and

posterior cranial fossae. Twenty percent of craniopharyngiomas originate within the sella, producing sellar enlargement in a fashion similar to that produced by pituitary adenomas. Almost half of all craniopharyngiomas are cystic, 15% are solid, and the remainder are made up of both solid and cystic elements. Although generally well circumscribed, craniopharyngiomas are not encapsulated lesions and therefore are often tenaciously adherent to basal neural and vascular structures. It is this feature of craniopharyngiomas that frequently undermines successful attempts at a safe and curative resection. There has been an increasing tendency to view craniopharyngiomas as being either classically adamantinomatous or papillary in nature, which is a distinction that some believe to be of clinical, pathologic, and prognostic significance.78,79 Adamantinomatous Craniopharyngioma. Adamantinomatous craniopharyngioma represents the classic cystic craniopharyngioma of childhood. It frequently is a bulky, partially calcified tumor that tenaciously insinuates itself around basal brain structures. On gross sectioning, it oozes a viscid admixture of shimmering cholesterol crystals and calcific desquamated debris that, in appearance and consistency, often is described as machinery oil. On light microscopic examination, this variant exhibits an intricate pattern of epithelial growth, including intermixed islands of solid and cystic epithelium within a matrix of variably cellular connective tissue. Nests or cords of columnar and squamous epithelial cells can be demonstrated. Lymphocytes, macrophages, clusters of foamy cells, cholesterol crystals, keratin deposits, necrotic debris, and polymorphonuclear leukocytes often are present. Calcification, ranging from areas visible only with a microscope to palpable concretions, is common. Rarely, frank bone formation may be present. Cyst formation is presumably the result of degeneration of squamous cells, accumulation of keratinous debris, and perivascular stromal degeneration. The results of immunohistochemical studies are conclusively negative for adenohypophysial hormones, indicating that craniopharyngiomas do not originate in adenohypophysial cells and are not capable of hormone production. The results of immunostaining for keratin are positive. Electron microscopic examination shows bundles of tonofilaments, prominent desmosomal attachments, and an absence of secretory granules (Fig. 11-13).

FIGURE 11-13. Epithelial component of craniopharyngioma showing tonofilaments (arrows) and absence of secretory granules. 8000

Papillary Craniopharyngioma. The clinical, radiologic, and pathologic profile of papillary craniopharyngioma deviates considerably from that of the conventional adamantinomatous variant.78,79 Accounting for ~10% of all craniopharyngiomas, the papillary variant occurs almost exclusively in adults. It usually is suprasellar in location and frequently involves, or arises within, the third ventricle. Most papillary variants are solid or have only a relatively minor cystic component. Papillary craniopharyngiomas are more discretely circumscribed than classic craniopharyngiomas and lack the calcification and machinery oil content so typical of adamantinomatous tumors. Liquid contents, although rarely present, generally are clear. Lacking tenacious adhesions to basal brain structures, the papillary craniopharyngioma is reputed to be more readily separable from surrounding structures. Histologically, it consists of a well-differentiated, albeit less complex, epithelial pattern than that of the adamantinomatous variant. Prominent, stratified, squamous-lined papillae, devoid of columnar palisading, microcystic degeneration, calcification, keratinous nodules, and cholesterol clefts, are characteristic. Whereas the papillary variant is regarded by some as being more amenable to complete and curative resection, not all are in agreement with this view.78,79,80 and 81 Both forms of craniopharyngioma are histologically benign. Mitotic figures and other features of histologic aggressiveness generally are not seen. Despite their histologic benignity, craniopharyngiomas, particularly the adamantinomatous type, are notorious for their high rate of postoperative recurrence. Even among tumors resected radically, a procedure sometimes accompanied by considerable functional deficit, recurrence rates as high as 25% have been reported. For lesser degrees of resection, recurrence of symptoms is virtually guaranteed, often within 3 years of surgery. Radiotherapy for incompletely excised lesions has proved effective in forestalling recurrence (see Chap. 22). Malignant transformation is exquisitely rare; only a single case of malignant craniopharyngioma has been described. Craniopharyngiomas, both adamantinomatous and papillary, have been shown to express estrogen receptor mRNA and protein. The significance of this finding remains to be determined.82 Rathke Cleft Cysts. Rathke cleft cysts are epithelial cysts apparently derived from remnants of the Rathke pouch. At autopsy, roughly one-fifth of pituitaries contain macroscopic remnants of the Rathke pouch in the form of discontinuous cystic remnants or microscopic clefts at the interface of the anterior and posterior lobes. Occasionally, these Rathke cleft remnants, as the result of progressive accumulation of colloidal secretions, become sufficiently large and compress surrounding structures. Most cases involving symptoms present as expansile intrasellar masses, occasionally having a suprasellar component. Only exceptionally are they entirely suprasellar in location. Local compressive effects are the basis for presentation in most cases involving symptoms, with headache, hypopituitarism, hyperprolactinemia, visual disturbance, and, rarely, diabetes insipidus being the principal clinical features.83 Rathke cleft cysts are thin-walled, uniloculate, and filled with fluid, the composition of which ranges from watery to mucinous. The cyst wall frequently is composed of a single layer of cuboidal or columnar, ciliated, or mucin-producing epithelium. Small numbers of adenohypophysial cells also may be evident. Calcification is rare, as is amyloid deposition. Although the epithelial pattern is considerably less complex than that of craniopharyngioma, the distinction between these two entities occasionally can be troublesome, emphasizing the importance of generous tissue sampling. Rarely, pituitary adenomas can be found to be admixed with elements of a Rathke cleft cyst. Such biopsy results usually represent the simultaneous sampling of two distinct lesions. Even rarer are more complex lesions, in which Rathke cyst components are intimately associated with adenoma and even squamous metaplasia; such lesions have been termed transitional cell tumors of the pituitary. Whether these are distinct clinicopathologic entities or simply the collision of two distinct lesions remains to be determined. Rathke cleft cysts are definitively treated by drainage and marsupialization of the cyst wall, a procedure that usually results in cure. Recurrences are unusual.

PATHOLOGY OF THE NEUROHYPOPHYSIS AND HYPOTHALAMUS 2 ,4,10,19

Clinically significant diseases of the posterior lobe are rare. Endocrinologically, they can be divided into conditions associated with increased or decreased vasopressin secretion. Several abnormalities are unaccompanied by endocrine alterations. INAPPROPRIATE SECRETION OF VASOPRESSIN Inappropriate secretion of vasopressin, or the Schwartz-Bartter syndrome, is the result of vasopressin hypersecretion, either from the posterior pituitary or from extrahypophysial neoplasms (see Chap. 27 and Chap. 219). Renal sodium loss and hyponatremia are characteristic features. Several diseases, including meningitis, myxedema, and cerebral lesions, may be associated with increased vasopressin discharge. Paraneoplastic (ectopic) vasopressin secretion may occur in various neoplasms, but is seen mainly in carcinoma of the bronchus. Vasopressin can be extracted from the extrapituitary tumors of patients with vasopressin excess, providing evidence for paraneoplastic (ectopic) production of the hormone. DIABETES INSIPIDUS Diabetes insipidus is characterized clinically by polyuria and polydipsia (see Chap. 26). In most cases it is caused by vasopressin deficiency resulting from the destruction of supraoptic and paraventricular nuclei, which is where vasopressin is synthesized, or by organic damage to the hypophysial stalk or posterior lobe, which is the site of vasopressin discharge. Morphologically, various lesions can be seen in the hypothalamus, especially in the nucleus supraopticus or along the supraopticohypophysial tract. Selective destruction of the posterior lobe results in only moderate and temporary polyuria and polydipsia. Causes include lesions resulting from head trauma, transection of the hypophysial stalk, meningoencephalitis, sarcoidosis, granulomas, Langerhans histiocytosis, primary tumors, metastatic carcinomas, lymphomas, and leukemias. Of note, pituitary adenomas, including large and invasive ones, are virtually never accompanied by diabetes insipidus as an initial feature of the tumor. The preoperative presence of diabetes insipidus in association with a sellar region mass argues strongly against a diagnosis of pituitary

adenoma, however suggestive the imaging studies may be. In idiopathic diabetes insipidus, no destructive lesions can be recognized grossly in the hypothalamus, hypophysial stalk, or posterior pituitary. Histologically, nerve cells of the supraoptic and paraventricular nuclei may show a marked reduction in number and size and loss of stainable neurosecretory material. Immunostaining demonstrates an absence of vasopressin in the hypothalamus, hypophysial stalk, and posterior lobe. The renal form of diabetes insipidus (nephrogenic diabetes insipidus) is caused by endorgan failure. Renal tubular cells fail to respond to the antidiuretic effect of vasopressin, resulting in polyuria and polydipsia. No lesions are evident in the hypothalamus, hypophysial stalk, or posterior lobe. Vasopressin synthesis and release are not impaired. BASOPHILIC CELL INVASION Basophilic cell invasion of the pituitary is a frequent autopsy finding in older men. It causes no clinical symptoms and cannot be detected by gross examination of the pituitary. Histologically, single or small groups of basophilic cells are seen to creep into the posterior lobe. In some cases, large groups of basophilic cells deeply invade the posterior lobe. The cytoplasm of basophilic cells is PAS-positive and contains ACTH and other fragments of the pro-opiomelanocortin molecule, indicating that these cells are related to corticotropes. However, they differ from the corticotropes located in the anterior lobe: they are smaller and denser, and, except for occasional cases, do not show Crooke hyalinization as a result of cortisol excess. Baso-philic cell invasion is not apparent before puberty and cannot be correlated with any clinical endocrine abnormality. From a practical standpoint, the phenomenon of basophil invasion is important only insofar as its presence in a surgical specimen should not be mistaken for a corticotrope adenoma invading the neural lobe of the gland. MISCELLANEOUS FINDINGS Squamous cell nests, which are glandular structures resembling salivary glands, and focal mononuclear cell infiltration are common incidental findings at autopsy in the posterior lobe and distal end of the hypophysial stalk. Hemorrhages, necroses, and granulomas were reviewed in the discussion on diseases of the anterior lobe. INTERRUPTION OF THE HYPOPHYSIAL STALK Interruption of the hypophysial stalk causes distinct changes along the entire supraopticohypophysial tract. Surgical transection or disruption of the hypophysial stalk secondary to head trauma or organic diseases leads to atrophy of the supraoptic and, to a lesser extent, the paraventricular nuclei, as well as the posterior lobe. Various diseases, such as infections, granulomas, sarcoidosis, and neoplasms, can destroy the supraoptic and paraventricular nuclei and disrupt the hypophysial stalk, thereby impairing the innervation of the posterior lobe. Because the normal functional activity of the posterior lobe depends on the integrity of its nerve supply, diabetes insipidus develops in the absence of innervation. Morphologically, the posterior lobe undergoes marked involution; loss of stainable neurosecretory material and hormone content can be demonstrated. On radio-immunoassay, vasopressin concentrations become undetectable; histologically, neurohypophysial tissue is replaced by a fibrous scar. Atrophy of the posterior lobe is noticeable in some cases of anterior hypopituitarism. In postpartum hypopituitarism, atrophy of the supraoptic and paraventricular nuclei and of the hypophysial stalk and posterior lobe may occur. NEOPLASMS Neoplasms in the posterior lobe, hypophysial stalk, and hypothalamus are uncommon. Secondary carcinomas were discussed earlier in relation to the anterior lobe. GRANULAR CELL TUMORS Granular cell tumors (choristomas, or tumorlets) are the most common primary tumors in the posterior lobe and distal part of the hypophysial stalk. They are found in 1% to 8% of unselected adult autopsies, mainly in the elderly. Usually small (12 mm or less), they can be detected histologically. Granular cell tumors are slow-growing, histologically benign, sharply demarcated but unencapsulated nodules that usually are not associated with clinical symptoms or biochemical abnormalities and are recognized incidentally at autopsy. Rare granular cell tumors grow rapidly and become large, causing headaches, vision disturbances, diabetes insipidus, and anterior hypopituitarism.10 Because of increased intracranial pressure or cranial nerve compression, surgical removal of the tumor may be necessary. Histologically, granular cell tumors consist of loosely apposed, large, spherical, oval, or polygonal cells with eccentric nuclei and abundant, coarsely granular cytoplasm. Numerous large granules that stain strongly with PAS, luxol fast blue, and alcian blue almost fill the entire cytoplasm. On electron microscopic examination, the granules correspond to large, membrane-bound, unevenly electron-dense lysosomes. Immunostains reveal an absence of adenohypophysial or neurohypophysial hormones, but may show S-100 protein in the cytoplasm. Granular cell tumors appear to originate in pituicytes, the special glial cells of the posterior lobe. GLIOMAS Other neoplasms are rare in the posterior lobe. Gliomas may originate in pituicytes of the posterior lobe; their histologic features are identical to those of glial tumors deriving from the central nervous system. Most gliomas involving the posterior lobe originate in the hypothalamus or median eminence and spread downward to the posterior lobe. Diabetes insipidus may develop as a result of massive destruction of the hypothalamus, the hypophysial stalk, or the posterior lobe. Most gliomas that involve the hypothalamus, stalk, and posterior lobe are low-grade astrocytomas of pilocytic type. Malignant gliomas arising in these structures are extremely rare, most being a long-term complication of radiotherapy for a sellar region tumor. Postirradiation gliomas occurring in the region generally have been either anaplastic astrocytomas or glioblastomas. The mean latency period is ~10 years.19 They invariably are aggressive lesions, with most patients dying within months of the diagnosis. GANGLIOCYTOMAS (HAMARTOMAS) OF THE SELLAR REGION Lesions composed of neurons can occasionally form symptomatic masses in the sellar region. In fact, constituting this group are a collection of entities, ranging from simple hypothalamic hamartomatous growths on the one hand, to intriguing intra-sellar neoplasms composed of both neuronal and adenohypophysial elements. As a group, however, they remain unified by their content of fully differentiated ganglion cells that appear mature and are accompanied by neuropil. The nomenclature surrounding these lesions has been as diverse as it has been confusing, wherein designations such as ganglioneuroma, neuronal hamartoma, gangliocytoma, choristoma, gangliocytoma-pituitary adenoma, pituitary adenomaadenohypophysial choristoma, and a medley of neologisms of one form or another have been inconsistently, interchangeably, and, at times, arbitrarily applied. Understandably, some of this variation in terminology reflects differing views on the presumed histiogenetic origins of these lesions, an issue that currently is far from settled. It is important to recognize that two topologically distinct variants exist: one arising in the hypothalamus and the second within the sella. Of the former entities that arise in or remain physically attached to the hypothalamus, the term hypothalamic neuronal hamartoma is applied. Of the intrasellar variants, the majority of which are intimately admixed within the substance of a pituitary adenoma, the terms pituitary adenoma-adenohypophysial neuronal choristoma (PANCH)84 or mixed pituitary adenoma-gangliocytoma are applied.85 Both hypothalamic and intrasellar variants are discussed separately. Hypothalamic Neuronal Hamartoma. Not uncommonly, a well-defined mass composed of mature central ganglionic tissue projecting in the leptomeninges from the base of the brain will be encountered. In fact, when carefully sought, minute, macroscopic and nodular foci of ectopic hypothalamic tissue may be found incidentally in ~20% of random autopsies. This is usually attached to the ventral hypothalamus, the adjacent pia, or on the surface of the proximal posterior cerebral arteries. Although such hamartomatous nodules are clinically insignificant, rarely they may grow to several centimeters in size and compress surrounding structures. Some retain a thick pedicular attachment to the hypothalamus, tuber cinereum, or mammillary bodies, whereas others form a sessile solitary mass. Most symptomatic examples occur in young males, in whom precocious puberty is the best known manifestation. In some instances, GnRH can be detected immunohistochemically within the neurons of such hamartomas, providing a neuroen-docrinologic basis for accelerated sexual maturation. This is not, however, a universal nor a necessary finding. Precocious puberty in immunonegative cases is presumably the result of hypothalamic compression. Rarely, hypothalamic neuronal hamartomas may produce GHRH and clinical acromegaly on a hypothalamic basis, a condition dubbed hypothalamic acromegaly. In such cases, the pituitary may show either hyperplasia or adenomas of GH-producing cells. In addition to other features of hypothalamic dysfunction (somnolence, hyperphagia, autonomic disturbances, and diabetes insipidus), these tumors are associated with a peculiar form of epilepsy, one characterized by laughing (gelastic seizures). Grossly, the mass in most symptomatic examples will be only 1 to 2 cm in size and often lies behind the pituitary stalk. It is pale, of firm consistency, and has homogeneous cut surfaces. Microscopically, the composition of the tissue resembles that of cerebral gray matter, both qualitatively and quantitatively. The neurons can vary in size, shape, and number, and both unipolar and multipolar forms are represented. Overall, these elements fully resemble mature hypothalamic neurons, although they are often disposed in clusters. Both myelinated and unmyelinated fibers course among them, forming compact bundles in some areas. Axonal processes may also appear to form ill-defined tracts, particularly among examples having pedicular attachment to the hypothalamus. In all examples, the neuronal elements are

supported by a normal complement of glial cells of different kinds, although a variable amount of fibrillary gliosis may be found. Immunohistochemical stains reveal a variety of hypothalamic-releasing peptides that normally reside within the cytoplasm of hypothalamic neurons: immunopositivity for GnRH, somatostatin, GHRH, and CRH may be detectable. In most cases, the immunohistochemical presence of these factors is regarded as a physiologic finding, and should not necessarily imply that these lesions are engaged in pathologic hormone hypersecretion. However, in the appropriate clinical context, such as with precocious puberty or the rare instance of acromegaly, the trophic effects of these hormones and their pathologic contribution to the endocrinopathy cannot be dismissed. If they are symptomatic and surgically accessible, therapy for hypothalamic neuronal hamartomas is directed primarily at relief of the mass effect. When technically feasible, surgical resection has been successful not only in ameliorating mass effects, but also in regressing secondary sexual characteristics in those experiencing precocious puberty and in improving seizure control. For lesions less amenable to total resection by virtue of their deep intrahypothalamic location or other factors limiting surgical accessibility, partial resection may also be of some symptomatic benefit; residual hamartomatous tissue should grow slowly, if at all. Hypothalamic Hamartoblastoma. A special variant of hypothalamic hamartoma is the hypothalamic hamartoblastoma, which occurs in infants and neonates and is associated with other multiple congenital abnormalities. The most frequent of the co-existing anomalies have included pituitary agenesis; dwarfism; dysmorphic facies; short, broad, or absent olfactory bulbs; hypoplastic thyroid and adrenals; cryptorchidism; various renal and cardiac malformations; anorectal atresia; syndactyly; and short metacarpals. This lethal syndrome is sometimes referred to as the Pallister-Hall syndrome. The hamartomatous lesion in this condition differs in several subtle respects from the more common hypothalamic hamartoma described earlier. As might be expected from the young age of affected patients, the lesion tends to be more cellular and less differentiated than the typical hypothalamic hamartoma found in older patients, hence the designation hamartoblastoma. A process morphologically intermediate between neoplasm and malformation, the hamartoblastoma is composed of ill-defined clusters of uniform, primitive-appearing, immature neurons unassociated with atypia or mitoses; neuronal differentiation appears incomplete. Intrasellar Adenohypophysial Neuronal Choristoma. The intriguing neoplasm of the sellar region known as an intrasellar adenohypophysial neuronal choristoma, which, for lack of better designations, has previously been known under the terms intrasellar adenohypophysial neuronal choristoma or the intrasellar gangliocytoma. It is a composite neoplasm composed of adenohypophysial cells and ganglion-like cells to which the appellation pituitary adenoma-neuronal choristoma (PANCH) has been applied in an attempt to highlight the duality of its composition.84 In a comprehensive review reiterating the composite neuronal and adenohypophysial nature of this neoplasm, the authors indicated a preference for the appellation mixed adenoma-gangliocytoma.85 This lesion, although sharing some histologic similarities with the hypothalamic hamartoma, differs in several topologic and biologic respects. First and foremost, adenohypophysial neuronal choristomas are intrasellar lesions that have no physical attachment to the hypothalamus. Second, they are generally associated withor more preciselyare intimately admixed with an endocrinologically functioning pituitary adenoma. Accordingly, they are virtually always symptomatic lesions of which the presence is heralded by a hypersecretory state. Finally, the neuronal component of this lesion, like the adenohy-pophysial component, are both genuinely neoplastic in nature. The basic lesion consists of islands of ganglion-like cells and accompanying neuropil interspersed within the substance of a pituitary adenoma (Fig. 11-14). The ratio of adenohypophysial cells to neuronal elements can vary considerably, but as a rule, both will be strongly represented. The adenohypophysial cells in virtually all instances are chromophobic and sparsely granulated in nature. They may be disposed in nests, sheets, or otherwise scattered among the neuronal elements. With respect to the latter, the ganglion-like cells are of varying size and number, contain an abundant cytoplasm, complete with peripheral Nissl substance and neuronal processes containing Herring bodies. The ganglion cells fully resemble normal hypothalamic neurons.

FIGURE 11-14. Adenohypophysial neuronal choristoma. The specimen is from a patient with acromegaly and a large intrasellar and suprasellar mass that was presumed to be an ordinary growth hormonesecreting adenoma. After transsphenoidal resection, the surgical specimen revealed an adenohypophysial neuronal choristoma intermixed within the substance of a growth hormonesecreting pituitary adenoma. Adenoma cells (arrowhead) intermixed with fully differentiated neurons (arrow) are evident in the specimen. These neurons were shown to contain growth hormonereleasing hormone. Original magnification 400

Of reported cases, most have occurred in the setting of acromegaly, wherein patients have had the clinical, radiologic, and endocrinologic features of a typical GH-producing pituitary adenoma. On examination of the surgical specimen, a somatotrope adenoma is encountered, invariably of the sparsely granulated type, in which are interspersed varying numbers of neuronal elements, the morphology of which resembles well-differentiated ganglionic cells. Some of these ganglion cells have been found to be immunoreactive for GHRH, whereas others are immunoreactive for somatostatin. Less frequently, the lesion occurs in the setting of Cushing disease, wherein the adenoma is of the corticotropic type accompanied by ganglion cells immunoreactive for CRH. Only rarely is the adenohypophysial tumor a prolactinoma or a clinically nonfunctioning-pituitary adenoma. Previously, hypotheses concerning the origin of these lesions centered on the seemingly displaced hypothalamic neurons in the pituitary fossa that, as a result of their secretion of trophic factors, might have a paracrine inductive effect on adjacent adenohypophysial cells, stimulating their growth and eventual neoplastic transformation. According to this view, pituitary adenoma formation is regarded as a secondary, downstream event. Subsequently, however, an alternative hypothesis has been proposed that features adenohypophysial cells as the instigators of the process, rather than seemingly passive targets of paracrine effect.84 Supported by compelling morphologic evidence, this hypothesis suggests that pituitary adenoma formation is the primary event, and the ganglionic component of the lesion arises as the consequence of neuronal metaplasia of neoplastic adenohypophysial cells. When subjected to careful ultrastructural study, morphologic changes that are indicative of a metaplastic process beginning in adenohypophysial cells that gradually gives way to cells having intermediate and, eventually, fully neuronal features can be seen. Whereas metaplastic change is a well-known phenomenon in both normal and neoplastic adenohypophysial cells, as it is in other neuroendocrine neoplasms, actual induction of such metaplastic change has yet to be demonstrated experimentally. A third hypothesis is that both adenohypophysial and neuronal components of this tumor arise from a common, but as yet hypothetical, progenitor cell sequestrated in the sella as an embryonal tissue nest. Because this entity is relatively new, with only small numbers of cases having been collected, and virtually no long-term follow-up data available, it is unclear whether the behavior of these mixed lesions differs significantly from that of the corresponding pituitary adenoma. The isolated intrasellar gangliocytoma is a lesion unassociated with a pituitary adenoma and anatomically distinct from the hypothalamus.85 Although an adenoma was not found in any of seven cases, three examples were associated either with Cushing disease or with acromegaly, a situation not readily explained by a purely gangliocytic tumor. The authors have not encountered such a pure gangliocytic lesion in the sella, particularly in the setting of a hypersecretory statewhen an accompanying pituitary adenoma or hyperplasia has always been present. CHAPTER REFERENCES
1. Hague K, Post KD, Morgello S. Absence of peritumoral Crooke's change is associated with recurrence in surgically treated Cushing's disease. Surg Neurol 2000; 53:77. 1a. Kovacs K, Horvath E. Tumors of the pituitary gland. In: Hartmann WH, Sobin LH, eds. Atlas of tumor pathology, fascicle 21, series 2. Washington, DC: Armed Forces Institute of Pathology, 1986:1. 2. Horvath E, Scheithauer B, Kovacs K, et al. Regional neuropathology: hypothalamus and pituitary. In: Graham D, Lantos P, eds. Greenfield's Neuropathology. London: Arnold, 1997:1007. 3. Lloyd RV, Chander WF, Kovacs K, Ryan N. Ectopic pituitary adenomas with normal anterior pituitary glands. Am J Surg Pathol 1986; 10:546. 4. Thapar K, Kovacs K, Scheithauer B, et al. Classification and pathology of sellar and parasellar tumors. In: Tindall G, Cooper P, Barrow D, eds. The Practice of Neurosurgery. Baltimore: Williams

5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84.

& Wilkins, 1996:1021. Bergland RM, Ray BS, Torack RM. Anatomical variations in the pituitary gland and adjacent structures in 225 human autopsy cases. J Neurosurg 1968; 28:93. Kaufman B, Chamberlin WB Jr. The ubiquitous empty sellar turcica. Acta Radiol Diagn (Stockh) 1972; 13:413. Weisberg LA, Housepian EM, Saur DP. Empty sella syndrome as a complication of benign intracranial hypertension. J Neurosurg 1975; 43:177. Brismar K, Efendic S. Pituitary function in the empty sella syndrome. Neuroendocrinology 1981; 32:70. McFadzean R, Teasdale G. Pituitary apoplexy. In: Landolt A, Vance M, Reilly P, eds. Pituitary Adenomas. Edinburgh: Churchill Livingstone, 1996:485. Thapar K, Kovacs K. Tumors of the sellar region. In: Bigner DD, McLendon RE, Bruner JM, eds. Russel and Rubinstein's pathology of tumors of the nervous system, 6th ed. Baltimore: Williams & Wilkins, 1998:561. Cardosa ER, Peterson EW. Pituitary apoplexy: a review. Neurosurgery 1984; 14:363. Berger SA, Edberg SC, David G. Infectious disease in the sella turcica. Rev Infect Dis 1986; 8:747. Domingue JN, Wilson CB. Pituitary abscesses: report of seven cases and review of the literature. J Neurosurg 1977; 46:601. Sano T, Kovacs K, Scheithauer BW, et al. Pituitary pathology in acquired immunodeficiency syndrome. Arch Pathol Lab Med 1989; 113:1066. Beressi N, Beressi JP, Cohen R, Modigliani E. Lymphocytic hypophysitis: a review of 145 cases. Ann Med Interne (Paris) 1999; 150:327. Feigenbaum SL, Martin MC, Wilson CB, Jaffe RB. Lymphocytic adenohy-pophysitis: a pituitary mass lesion occurring in pregnancy. Proposal for medical treatment. Am J Obstet Gynecol 1991; 164:1549. Lee JH, Laws ER Jr, Guthrie BL, et al. Lymphocytic hypophysitis: occurrence in two men. Neurosurgery 1994; 34:159. Capellan JIL, Olmedo C, Martin JM, et al. Intrasellar mass with hypopituitarism as a manifestation of sarcoidosis. J Neurosurg 1990; 73:283. Scheithauer BW. The neurohypophysis. In: Kovacs K, Asa SL, eds. Functional endocrine pathology. Boston: Blackwell, 1991:170. Nishio S, Mizuno J, Barrow DL, et al. Isolated histiocytosis X of the pituitary gland: case report. Neurosurgery 1987; 21:718. Tampanaru-Sarmesiu A, Stefaneanu L, Thapar K, et al. Transferrin and transferrin receptor in human hypophysis and pituitary adenomas. Am J Pathol 1998; 152:413. Horvath E, Kovacs K. The adenohypophysis. In: Kovacs K, Asa SL, eds. Functional endocrine pathology. Boston: Blackwell, 1991:245. Thapar K, Kovacs K, Laws ER Jr. The pathology and molecular biology of pituitary adenomas. Adv Tech Stand Neurosurg 1995; 22:4. Sano T, Asa SL, Kovacs K. Growth hormone-releasing hormoneproducing tumors: clinical, biochemical, and morphological manifestations. Endocr Rev 1988; 9:357. Scheithauer BW, Kovacs K, Randall RV. The pituitary in untreated Addison's disease: a histologic and immunocytologic study of 18 adenohypophyses. Arch Pathol Lab Med 1983; 107:484. Scheithauer BW, Kovacs K, Randall RV, Ryan N. Pituitary gland in hypothyroidism. Histologic and immunocytologic study. Arch Pathol Lab Med 1985; 109:499. Herman V, Fagan J, Gonsky R, et al. Clonal origin of pituitary adenomas. J Clin Endocrinol Metab 1990; 71:1427. Schulte HM, Oldfield EH, Allolio B, et al. Clonal composition of pituitary adenomas in patients with Cushing's disease: determination by X chromosome inactivation analysis. J Clin Endocrinol Metab 1991; 73:1302. Faglia G, Spada A. The role of the hypothalamus in pituitary neoplasia. Baillieres Clin Endocrinol Metab 1995; 9:225. Asa S, Kovacs K, Stefaneanu L, et al. Pituitary adenomas in mice transgenic for growth hormone-releasing hormone. Endocrinology 1992; 131:2083. Melmed S. Acromegaly. In: Melmed S, ed. The Pituitary. Cambridge: Black-well Science, 1995:413. Levy A, Lightman S. Growth hormone-releasing hormone transcripts in human pituitary adenomas. J Clin Endocrinol Metab 1992; 74:1474. Rauch C, Li J, Croissandeau G, et al. Characterization and localization of an immunoreactive growth hormone-releasing hormone precursor form in normal and tumoral human anterior pituitaries. Endocrinology 1995; 136:2594. Castro M, Brooke J, Bullman A, et al. Synthesis of corticotropin-releasing hormone (CRH) in mouse corticotropic tumour cells expressing the human proCRH gene, intracellular storage and regulated secretion. J Mol Endocrinol 1991; 7:97. Pagesy P, Li JY, Rentier-Delrue F, et al. Growth hormone and somatostatin gene expression in pituitary adenomas with active acromegaly and minimal plasma growth hormone elevation. Acta Endocrinol 1990; 122:745. Miller G, Alexander J, Klibanski A. Gonadotropin-releasing hormone messenger RNA expression in gonadotroph tumors and normal human pituitary. J Clin Endocrinol Metab l996; 81:80. Thapar K, Kovacs K, Stefaneanu L, et al. Overexpression of the growth hormone-releasing hormone gene in acromegaly associated pituitary tumor: an event associated with neoplastic progression and aggressive behavior. Am J Pathol 1997; 151:769. Thapar K, Kovacs K, Muller P. Clinical-pathologic correlations of pituitary tumors. Baillieres Clin Endocrinol Metab l995; 9:243. Molitch ME. Pathologic hyperprolactinemia. Endocrinol Metab Clin North Am 1992; 21:877. Scheithauer BW, Kovacs K, Randall RV, Ryan N. Effects of estrogen on the human pituitary: a clinicopathologic study. Mayo Clin Proc 1989; 64:1077. Stefaneanu L, Kovacs K, Horvath E, et al. In situ hybridization of estrogen receptor mRNA in human adenohypophysial cells and pituitary adenomas. J Clin Endocrinol Metab 1994; 78:83. Kovacs K, Stefaneanu L, Ezzat S, et al. Prolactin producing pituitary adenoma in a male to female transsexual patient following protracted estrogen administration: a morphologic study. Arch Pathol Lab Med 1994; 118:562. Vallar L, Spada A, Giannattasio G. Altered Gs and adenylate cyclase activity in human GH secreting pituitary tumors. Nature 1987; 330:566. Landis CA, Harsh G, Lyons J, et al. Clinical characteristics of acromegalic patients whose pituitary tumors contain mutant Gs protein. J Clin Endocrinol Metab 1990; 71:1416. Yoshimoto K, Iwahana H, Fukuda A, et al. Rare mutations of the Gs alpha subunit in human endocrine tumors. Mutation detection by polymerase chain reaction-primer introduced restriction analysis. Cancer 1993; 72:1286. Tjordman K, Stern N, Ouaknine G, et al. Activating mutations of the Gs alpha gene in nonfunctioning pituitary tumors. J Clin Endocrinol Metab 1993; 77:765. Williamson E, Daniels M, Foster S, et al. Gs alpha and gi2 alpha mutations in clinically non-functioning pituitary tumors. Clin Endocrinol 1994; 41:815. Williamson E, Ince P, Harrison D, et al. G-protein mutations in human pituitary adrenocorticotrophic hormone-secreting adenomas. Eur J Clin Invest 1995; 25:128. Karga HJ, Alexander JM, Hedley-White ET, et al. Ras mutations in human pituitary tumors. J Clin Endocrinol Metab 1992; 74:914. Herman V, Drazin NZ, Gonsky R, et al. Molecular screening of pituitary adenomas for gene mutations and rearrangements. J Clin Endocrinol Metab 1993; 77:50. Pei L, Melmed S, Scheithauer BW, et al. H-ras mutations in human pituitary carcinoma metastases. J Clin Endocrinol Metab l994; 78:842. Bystrom C, Larsson C, Blomberg C, et al. Localization of the MEN1 gene to a small region within chromosome 11q13 by deletion mapping tumors. Proc Natl Acad Sci U S A 1990; 87:1968. Chandrasekharappa S, Guru C, Manickam P, et al. Positional cloning of the gene for multiple endocrine neoplasiatype 1. Science 1997; 276:404. Thakkar R, Pook M, Wooding C, et al. Association of somatotrophinomas with loss of alleles on chromosome 11 and with Gsp mutations. J Clin Invest 1993; 91:2815. Boggild M, Jenkinson S, Pistorello M, et al. Molecular genetic studies of sporadic pituitary tumors. J Clin Endocrinol Metab l994; 78:387. Harris C, Hollstein M. Clinical implications of the p53 tumor-suppressor gene. N Engl J Med 1993; 329:1318. Thapar K, Scheithauer BW, Kovacs K, et al. p53 expression in pituitary adenomas and carcinomas: correlation with invasiveness and tumor growth fractions. Neurosurgery 1996; 38:765. Levy A, Hall L, Yeudall WA, et al. p53 gene mutations in pituitary adenomas: rare events. Clin Endocrinol 1994; 41:809. Buckley N, Bates AS, Broome J, et al. p53 protein accumulates in Cushings adenomas and invasive non-functional adenomas. J Clin Endocrinol Metab 1994; 79:1513. Jacks T, Fazeli A, Schmitt EM, et al. Effects of an Rb mutation in the mouse. Nature 1992; 359:295. Cryns VL, Alexander JP, Klibanski A, Arnold A. The retinoblastoma gene in human pituitary tumors. J Clin Endocrinol Metab 1993; 77:644. Zhu J, Leon S, Beggs A, et al. Human pituitary adenomas show no loss of heterozygosity at the retinoblastoma gene locus. J Clin Endocrinol Metab 1994; 78:922. Pei L, Melmed S, Scheithauer B, et al. Frequent loss of heterozygosity at the retinoblastoma susceptibility gene (RB) locus in aggressive pituitary tumors: evidence for a chromosome 13 tumor suppressor gene other than RB. Cancer Res 1995; 55:1613. Capra E, Rindi G, Pompei-Spina M, et al. Chromosomal abnormalities in a case of pituitary adenoma. Cancer Genet Cytogenet 1993; 681:40. Rock J, Babu V, Drumheller T, et al. Cytogenetic findings in pituitary adenomas: results of a pilot study. Surg Neurol 1993; 40:224. Alvaro V, Levy L, Deburay C, et al. Invasive human pituitary tumors express a point mutated alpha-protein kinase-C. J Clin Endocrinol Metab 1993; 77:1125. Klibanski A, Zervas N. Diagnosis and management of hormone-secreting pituitary adenomas. N Engl J Med 1991; 324:822. Scheithauer BW, Kovacs K, Laws ER Jr, Randall RV. Pathology of invasive pituitary adenomas with special reference to functional classification. J Neurosurg 1986; 65:733. Knosp E, Kitz K, Perneczky A. Proliferation activity in pituitary adenomas. Measurement by monoclonal antibody Ki. Neurosurgery 1989; 25:927. Thapar K, Kovacs K, Scheithauer B, et al. Proliferative activity and invasiveness among pituitary adenomas and carcinomas: an analysis using the MIB-1 antibody. Neurosurgery 1996; 38:99. Hsu DW, Hakim F, Biller BMK, et al. Significance of proliferating cell nuclear antigen index in predicting pituitary adenoma recurrence. J Neurosurg 1993; 78:753. Pemicone PJ, Scheithauer BW, Sebo TJ, et al. Pituitary carcinoma. A clinicopathologic study of 15 cases. Cancer 1997; 79:804. Gesundheit N, Petrick PA, Nissim M, et al. Thyrotropin-secreting pituitary adenomas. Clinical and biochemical heterogeneity. Case reports and follow-up of nine patients. Ann Intern Med 1989; 111:827. Kovacs K, Asa SL, Horvath E, et al. Null cell adenomas of the pituitary: attempts to resolve their cytogenesis. In: Leschago J, Kameya T, eds. Endocrine pathology update. New York: Field & Wood, 1990:17. Kovacs K, Horvath E, Asa SL, et al. Pituitary cells producing more than one hormone. Human pituitary adenomas. Trends Endocr Metab 1989; 1:104. Thapar K, Stefaneanu L, Kovacs K, et al. Plurihormonal pituitary tumors: beyond the one cell-one hormone theory. Endocr Pathol 1993; 4:1. Branch CL, Laws ER Jr. Metastatic tumors of the sella turcica masquerading as primary pituitary tumors. J Clin Endocrinol Metab 1987; 65:649. Burger P, Scheithauer B. Tumors of the central nervous system. Third series, fascicle 10. Washington, DC: Armed Forces Institute of Pathology, 1994. Crotty R, Scheithauer B, Young W, et al. Papillary craniopharyngioma: a clinicopathological study of 48 cases. J Neurosurg 1995; 83:206. Crotty T, Scheithauer BW, Young WF, Davis D. Papillary craniopharyngiomas: a morphological and clinical study of 46 cases. Endocr Pathol 1992; 3(Suppl):S6(abst). Weiner H, Wishoff J, Rosenberg M, et al. Craniopharyngiomas: a clinicopathologic analysis of factors predictive of recurrence and functional outcome. Neurosurgery 1994; 35:1001. Thapar K, Stefaneanu L, Kovacs K, et al. Estrogen receptor gene expression in craniopharyngioma: an in situ hybridization study. Neurosurgery 1994; 35:1012. Brassier G, Morandi X, Tayiar E, et al. Rathke's cleft cysts: surgical-MRI correlation in 16 symptomatic cases. J Neuroradiol 1999; 26:162. Horvath E, Kovacs K, Scheithauer BW, et al. Pituitary adenoma with neuronal choristoma (PANCH): composite lesion or lineage infidelity? Ultrastruct Pathol 1994; 18:565.

CHAPTER 12 GROWTH HORMONE AND ITS DISORDERS Principles and Practice of Endocrinology and Metabolism

SECTION A ADENOHYPOPHYSIS CHAPTER 12 GROWTH HORMONE AND ITS DISORDERS

GERHARD BAUMANN Growth Hormone Genes Growth Hormone Structure Growth Hormone Receptor Growth HormoneBinding Protein Somatotrope Development and Growth Hormone Secretion Neural Control Growth HormoneReleasing Hormone and the Growth HormoneReleasing Hormone Receptor Somatostatin and Somatostatin Receptors Growth HormoneReleasing Peptides and the Growth HormoneReleasing Peptide Receptor Feedback Control Other Control Mechanisms Regulation of Placental Growth Hormone Secretion Regulation of the Growth Hormone Receptor and Binding Protein Actions of Growth Hormone Insulin-Like Growth Factors and Their Binding Proteins Insulin-Like Growth Factor Receptor Measurement of Plasma Growth Hormone Dynamic Tests of Growth Hormone Secretion Clinical Use of Insulin-Like Growth Factor-I and Insulin-Like Growth FactorBinding Protein Measurements Abnormal Secretion of Growth Hormone Deficiency of Growth Hormone Growth Hormone Insensitivity Hypersecretion of Growth Hormone (Acromegaly) Hypersensitivity to Growth Hormone Chapter References

Growth hormone (GH) is a polypeptide hormone produced by the somatotrope cells in the pituitary gland. It is the master anabolic hormone and possesses numerous bioactivities related to somatic growth, body composition, and intermediary metabolism. Many of the biologic actions of GH are mediated through insulin-like growth factor-I (IGF-I), but GH also has direct effects independent of IGF-I. Unlike most other hormones, GH is species specific, not only in its structure but also partially in its function. Its one-way species specificity refers to the fact that primate GHs are active in lower (evolutionarily earlier) species, but GHs of lower species are inactive in primates, including humans. GH regulation and, in part, GH action also differ among species. This chapter focuses primarily on human GH and its biology.

GROWTH HORMONE GENES

In humans, GH is encoded by two genes on the long arm of chromosome 17.1 They are part of a gene-duplication cluster that also includes the genes for chorionic somatomammotropin (placental lactogen), a protein highly homologous with GH (Fig. 12-1). The GH genes are named GH-N (or GH-1) and GH-V (or GH-2); the former is expressed in the pituitary, the latter in the placenta. GH-V is also called placental GH. The two GH genes are similar in structure; both are composed of five exons and four introns; each spans ~1.6 kb.

FIGURE 12-1. Human growth hormone locus on chromosome 17q22-24. The top panel shows the organization of the locus with the two GH genes (GH-N and GH-V) and the three chorionic somatomammotropin genes (CS-L, CS-A, CS-B). CS-L is probably a pseudogene. The bottom panel shows the GH-N gene in more detail, with five exons and four introns. GH-N is expressed in the pituitary, the other genes in the locus exclusively in the placenta. (From Parks JS. Molecular biology of growth hormone. Acta Paediatr Scand 1989; 349[Suppl]:127.)

GROWTH HORMONE STRUCTURE

Human GH is heterogeneous and consists of several molecular variants (Fig. 12-2).2 The principal form is a 191-amino-acid, single-chain protein with a molecular weight of ~22,000. It is the most abundant form (~90%95% of pituitary GH), generally referred to as GH or 22K GH. Two disulfide loops are present, and the three-dimensional structure is a twisted bundle of four a-helices (Fig. 12-3).3 Two independent receptor-binding sites are located on opposite surfaces of GH, allowing for ligand-induced receptor dimerization (see later). GH-V (placental GH) has a similar primary structure; it differs from GH-N (22K GH) in 13 of the 191 amino-acid positions (see Fig. 12-2). One important difference is the glycosylation consensus site at asparagine 140; GH-V exists as a glycosylated as well as nonglycosylated protein. The second most abundant GH form after 22K GH is an mRNA splice variant that lacks an internal sequence of 15 amino acids. Its molecular weight is ~20,000, and hence it is known as 20K GH. It accounts for 5% to 7% of pituitary GH. Other minor variants include an Na-acylated and two deamidated variants (see Fig. 12-2). Little is known about the bioactivities or significance of these minor GH forms. In addition to the monomeric forms of GH just described, GH also exists as an oligomeric series of up to at least pentameric GH.4 Both noncovalent and disulfide-linked oligomers occur, and homo- as well as heterooligomers composed of the various monomeric forms exist in the pituitary and plasma. The biologic significance of GH oligomers is unclear, but they likely act as modulators of overall GH activity because of their different affinities for the GH receptor. The existence of so many molecular forms of GH is one reason for the difficulty with GH measurements and the discrepant results obtained by different assays (see later).

FIGURE 12-2. Primary structure of human growth hormone and its variants. The polypeptide shown is GH-N (22K). Amino-acid substitutions in GH-V are indicated next to the involved residues. The sequence connected by the heavy line (residues 3246) is deleted in 20K GH. The tree structure at Asn-140 denotes glycosylation in GH-V. The asterisks indicate sites of deamidation, the dot at the amino terminus acylation. (From Baumann G. Growth hormone heterogeneity: genes, iso-hormones, variants and binding proteins. Endocr Rev 1991; 12:424.)

FIGURE 12-3. Three-dimensional structure of human growth hormone (22K), depicted as a ribbon diagram. The four main a-helices are shown together with three minihelices within the connecting loops. Some residues mutated for technical purposes are indicated; they are not relevant in this context. (N, amino terminus; C, carboxy terminus.) (From Ultsch MH, Somers W, Kossiakoff AA, de Vos AM. The crystal structure of affinity-matured human growth hormone at 2 resolution. J Mol Biol 1994; 236:286.)

GROWTH HORMONE RECEPTOR

The GH receptor (GHR) is a 620-amino-acid, ~130 kDa, single-chain glycoprotein with a single transmembrane domain, a large extracellular domain containing the GH-binding site, and an intracellular domain involved in GH signaling (Fig. 12-4).5 The extracellular domain also occurs separately as a soluble GH-binding protein (GHBP, see later). The GHR is a member of the cytokine receptor family that also includes the receptors for prolactin, erythropoietin and other hematopoietic growth factors, many of the interleukins, and others.6 The GHR is encoded by a single gene located on the short arm of chromosome 5. The gene spans at least 87 kb and is divided into 10 exons and 9 introns.7 Exons 27 encode the extracellular domain, exon 8 the trans-membrane domain, and exons 9 and 10 the intracellular domain. The GHR is expressed ubiquitously, with the liver being the organ most enriched in GHRs. In addition to the full-length GHR, two variants of the GHR are found in humans. A version lacking the 22 amino acids encoded by exon 3 is differentially expressed among tissues8 and/or in different individuals.9 This internal deletion near the amino terminus has no known functional consequence. The second variant is a GHR truncated at nine amino acids beyond the transmembrane domain, so that it lacks most of the intracellular domain.10 This variant is also expressed ubiquitously. The absence of an intracellular domain renders this variant incapable of signaling and favors prolonged persistence on the cell membrane. The latter may be the reason why this receptor variant contributes substantially to GHBP generation (see later). This truncated GHR variant modulates GH action by forming heterodimers with full-length GHRs, thereby sequestering some of the GHRs in a nonfunctional state.

FIGURE 12-4. Schematic representation of the growth hormone receptor (GHR) complementary DNA (top) and protein (bottom). The GHR is encoded by 10 exons; exons 27 encode the extracellular domain, exon 8 the transmembrane domain, and exons 9 and 10 the cytoplasmic domain. The numbers in the upper panel denote the exons; those in the lower panel the amino acids. The transmembrane domain is shaded in black. (Adapted from Kelly PA, Djiane J, Postel-Vinay M-C, Edery M. The prolactin/growth hormone receptor family. Endocr Rev 1991; 12:235.)

GH initiates its action first by binding to the GHR through site 1 on one of its surfaces; this is then followed by binding of a second GHR to site 2 on the other surface of GH.11 This results in a complex containing two GHRs in association with GH (Fig. 12-5). This GH-induced dimerization of the GHR is critical for GHR signaling and GH action. The functional domains of the GHR are depicted in Figure 12-6. Intracellular signaling is initiated by binding of JAK2 (Janus kinase 2) to a proline-rich region (Box 1) in the proximal intracellular part of the GHR, followed by a JAK2-mediated tyrosine phosphorylation cascade involving JAK2 itself, the GHR, signal transducers and activators of transcription (Stats) 1, 3, and 5, insulin-receptor substrates (IRS) 1 and 2, components of the mitogen-activated protein kinase (MAPK), the protein kinase C, and phosphatidyl inositol-3 kinase pathways, and several other intracellular signaling and adapter proteins, not all of which are known (Fig. 12-7).12 Gene transcription is then activated through these pathways. Interestingly, as of this writing, the precise pathway responsible for activation of IGF-I gene transcription is not known.

FIGURE 12-5. Schematic representation of growth hormone (GH) induced dimerization of two growth hormone receptors (GHRs). Binding occurs first through site 1 on GH, followed by binding of a second GHR to site 2 on GH. The dimerized GHR (2:1) complex (lower right) is active in transducing the GH signal. At very high (pharmacologic) GH concentrations (lower left), the ratio of GH to GHR is high enough to saturate GHRs through site 1 binding, with the GHR trapped in an inactive 1:1 complex. This occurs at GH levels that are not seen in vivo. However, recognition of the potential existence of an inactive 1:1 complex was important in developing a GH antagonist (see text on treatment for acromegaly). (hGH, human growth hormone.) (From Fuh G, Cunningham BC, Fukunaga R, et al. Rational design of potent antagonists to the growth hormone receptor. Science 1992; 256:1677.)

FIGURE 12-6. Functional domains of the growth hormone receptor (extracellular domain is on top). The proline-rich Box 1 is crucial for JAK2 (Janus kinase 2) binding and initiation of most of the signaling events. (C, extracellular cysteines, only one of which is free [not disulfide linked]; N, potential N-linked glycosylation sites; Y, intracellular tyrosines, important for phosphorylation and docking of signaling molecules; WSXWS, tryptophan-serine motif; Stat, signal transducer and activator of transcription; SHC, Src homology containing protein; MAPK, mitogen-activated protein kinase; Spi 2.1, serine protease inhibitor 2.1; IRS, insulin receptor substrate.) (From Argetsinger LS, Carter-Su C. Mechanisms of signaling by growth hormone receptor. Physiol Rev 1996; 76:1089.)

FIGURE 12-7. Intracellular signaling pathways of the growth hormone receptor (GHR) (partial rendition). Principal pathways are the signal transducer and activator of transcription (Stat), protein kinase C (PKC), mitogen-activated protein kinase (MAPK), and insulin receptor substrate (IRS) pathways. The inner ellipse represents the nucleus. Transcriptional elements, their cognate transcription factors, and transactivated genes are shown. (GH, growth hormone; P, phosphorylated sites; JAK2, Janus kinase 2; PC, phosphatidyl choline; PLC, phospholipase C; DAG, diacylglycerol; PI3K, phosphatidyl-inositol-3 kinase.) (From Argetsinger LS, Carter-Su C. Mechanisms of signaling by growth hormone receptor. Physiol Rev 1996; 76:1089.)

The GHR binds GH variants with different affinities. The 20K variant as well as the oligomeric GH forms have lower affinity than monomeric 22K GH, but GH-V is equipotent to 22K GH. Little is known about the binding of the other GH variants.

GROWTH HORMONEBINDING PROTEIN

The GHBP is the soluble, extracellular domain of the GHR. In humans and many other species, the GHBP is generated from the GHR by proteolysis; in rodents it is derived from the GHR gene via alternative splicing.13 The GHBP circulates in plasma in nanomolar concentrations, sufficient to complex a substantial part (~50%) of plasma GH. The serum GHBP level appears to reflect the GHR abundance of the organism, especially in the liver. The biologic significance or importance of the GHBP is not known; it is evolutionarily conserved throughout the vertebrates and is generated by different mechanisms in different species, suggesting an important role. The GHBP modulates GH action through a variety of mechanisms. It inhibits GH action by competing with the GHR for ligand and by generating unproductive heterodimers with the GHR at the cell surface, as opposed to the GHR homodimers necessary for signaling. The GHBP also prolongs the half-life of GH in the circulation by complexing GH, thereby delaying its elimination. The net effect of these modulatory activities in the intact organism is not well understood. In the circulation, GH also binds (with low affinity) to one or more proteins related to a214-macroglobulin. This complex accounts for no more than 5% of total GH in serum.

SOMATOTROPE DEVELOPMENT AND GROWTH HORMONE

Several pituitary transcription factors are involved in pituitary somatotrope development. Their consideration is important here because mutations in their respective genes lead to abnormalities in the GH axis. The reader is referred to Chapter 8 and Chapter 11 for a full discussion of pituitary ontogeny. Developmental genes relevant to GH include the Prop-1, Pit-1 (also known as Pou1f1), and GH-releasing hormone receptor (GHRH-R) genes. These genes are expressed sequentially during anterior pituitary development; each is more restrictive in its impact on different cell types; and each is dependent on the activity of the preceding one.15 Prop-1 and Pit-1 are POU-homeodomain transcription factors, the GHRH-R is a seven transmembrane domain receptor signaling through the cyclic adenosine mono-phosphate (cAMP) pathway. Prop-1 is important for development of the gonadotrope, somatotrope, lactotrope, and thyroptrope lineages. Pit-1, under the direction of Prop-1, is involved in differentiation of somatotropes, lactotropes, and thyrotropes. The GHRH-R, which is exclusively expressed in somatotropes under the direction of Pit-1, is critical for the normal expansion of the somatotrope population. The GHRH-R is also necessary for GH synthesis and secretion (see later).

SECRETION
NEURAL CONTROL GH secretion is under neural control from the hypothalamus through at least two and possibly three hypophysiotropic factors: GH-releasing hormone (GHRH), somatostatin, and probably ghrelin (see following section). The GHRH neurons are located primarily in the arcuate and ventromedial nuclei, and somatostatin neurons are located primarily in the anterior periventricular area of the hypothalamus. GHRH and somatostatin release are controlled by a complex and incompletely understood neural network, involving a-adrenergic, dopaminergic, serotoninergic, cholinergic, and histaminergic inputs. In general, a-adrenergic, dopaminergic, serotoninergic, and cholinergic signals stimulate GH secretion. The limbic system plays an important role in GH secretion. A full discussion of the neural pathways regulating GH secretion is beyond the scope of this chapter. From a practical standpoint, it is important to know the physiologic stimuli leading to GH secretion, the principal pharmacologic agents used to test GH secretory capacity, and the peripheral feedback control of GH secretion. GROWTH HORMONERELEASING HORMONE AND THE GROWTH HORMONERELEASING HORMONE RECEPTOR GHRH is a 40- to 44-amino-acid peptide isolated first from pancreatic tumors that produced it ectopically,16,17 and subsequently from the hypothalamus (Fig. 12-8). Its gene, located on chromosome 20q, encodes a 108-amino-acid precursor from which GHRH is derived by proteolytic cleavage. It is expressed in highest concentration in the hypothalamus, but also in other parts of the brain, in the gut, and in other tissues. The extrahy-pothalamic role of GHRH is largely unknown; it may act as a sleep inducer. GHRH is released from the median eminence into the pituitary portal system and is the principal stimulatory hypophysiotropic factor promoting GH secretion. (As mentioned earlier, GHRH is also important for the development of somatotrope cells15 and for GH synthesis.18) GHRH, on reaching the somatotropes, interacts with the GHRH-R, which is a seven transmembrane, Gsa-coupled receptor that signals through the cAMP and Ca2+-channel pathways (Fig. 12-9). Activation of these pathways effects GH release from secretory granules as well as GH gene transcription. The GHRH-R is expressed in a variety of tissues, but its biologic role in extrapituitary sites is unknown. GHRH is rapidly inactivated in plasma by an amino peptidase that cleaves the N-terminal dipeptide. Ectopic production of GHRH can occur in carcinoid and pancreatic islet tumors.

FIGURE 12-8. Amino-acid sequence of growth hormonereleasing hormone (GHRH) 1-40 (top) and 1-44 (bottom). GHRH 1-44 has four additional amino acids at the C-terminal end of the molecule and is amidated.

FIGURE 12-9. Primary structure of the growth hormonereleasing hormone (GHRH) receptor, showing the seven transmembrane helices. The location of a nonsense mutation responsible for familial GHRH-resistant dwarfism is also shown (see text on congenital growth hormone deficiency). (From Maheshwari HG, Silverman BL, Dupuis J, Baumann G. Phenotype and genetic analysis of a syndrome caused by an inactivating mutation in the growth hormone releasing hormone receptor: dwarfism of Sindh. J Clin Endocrinol Metab 1998; 83:4065.)

SOMATOSTATIN AND SOMATOSTATIN RECEPTORS Somatostatin is a cyclical peptide that exists in two forms: somatostatin-14 and somatostatin-28, the latter being extended at the amino terminus (Fig. 12-10 see Chap. 169). In humans, both somatostatins are encoded by a single gene on the long arm of chromosome 3, and a 92-amino-acid precursor is differentially processed to the two somatostatin forms, in part in a tissue-specific manner. In the hypothalamus, somatostatin-14 is the predominant form. Somatostatin is widely expressed throughout the central nervous system, the gut, and the pancreas. In extrahypothalamic sites, somatostatin has inhibitory effects on insulin secretion, gut hormone secretion, gut motility, and pancreatic and gastrointestinal exocrine secretions. In the hypothalamic-pituitary system, somatostatin inhibits GH and thyroid-stimulating hormone (TSH) secretion. Five somatostatin receptor subtypes are known; normal human pituitary expresses subtypes 1, 2, and 5.19 These receptors are members of the seven transmembrane domain, G proteincoupled class. Interaction of somatostatin with its receptors induces coupling to Gi and Go proteins, which in turn inhibits cAMP production and Ca2+-channel fluxes, thereby blocking release of GH (and other hormones).19

FIGURE 12-10. Amino-acid sequences of somatostatin-14 (top) and somatostatin-28 (bottom). The cyclical nature is indicated by the Cys-Cys bond.

GROWTH HORMONERELEASING PEPTIDES AND THE GROWTH HORMONERELEASING PEPTIDE RECEPTOR GHRPs are a class of short peptides (56 amino acids) that are extremely potent as pharmacologic GH secretagogues. The first prototypes (GHRP-5 and GHRP-6) were described in the early 1980s,20 and many peptide and nonpeptide analogs have since been synthesized. GHRPs are not entirely specific for GH; they also act to release adrenocorticotropic hormone (ACTH) and prolactin, although the effect on these hormones is relatively modest. The cloning of a specific GHRP receptor in 1996 moved this field from the pharmacologic to the physiologic realm,21 and a natural ligand, ghrelin, has been identified.21a The GHRP receptor is also a seven transmembrane domain, G proteincoupled receptor that interacts with Ga11 It is expressed in the hypothalamus and to a lesser degree in the pituitary. Ghrelin is expressed in the stomach and hypothalamus; it must be considered as a potential candidate for the regulation of GH secretion. Of interest is that its action on GH secretion is dependent on a functional GHRH system, and that GHRH and GHRP have synergistic actions in vivo. In contrast, the effect of GHRP on ACTH and prolactin release is independent of GHRH. The principal site of action of GHRP on the release of GH, ACTH, and prolactin is the hypothalamus, although for GH a direct, minor effect is also present at the pituitary level. The precise role of the GHRP system in the regulation of GH secretion remains to be determined. FEEDBACK CONTROL Negative feedback on GH secretion is exerted by IGF-I (a long feedback loop) and by GH itself (a short feedback loop). IGF-I inhibits GH secretion at both the hypothalamic and pituitary levels by influencing GHRH and somatostatin production (hypothalamus) and by interfering with GHRH action (pituitary). GH inhibits its own secretion by modulating GHRH and somatostatin secretion in the hypothalamus. These feedback effects are superimposed on the neural control mentioned earlier. OTHER CONTROL MECHANISMS Other important control mechanisms for GH secretion are estrogen- and age-dependent changes. Estrogen has a generally stimulating effect on GH secretion, which results in a distinct sexual dimorphism of GH secretion during the reproductive years. Estrogen can also be used to prime a patient to maximize response to pharmacologic stimuli. The estrogen effect may in part be mediated by induction of a peripheral GH-resistant state, with lowering of serum IGF-I levels. Pronounced developmental changes occur in GH secretion over the life span. In late fetal and neonatal life, GH secretion is very high and partly unregulated, perhaps in part because of the immaturity of the GHR system and low IGF-I levels. After birth, GH secretion rapidly falls to childhood levels, to be up-regulated again during puberty in response to sex steroids. Thereafter, GH secretion declines progressively by ~15% per decade, reaching very low levels in old age. This process has been termed somatopause and is in part responsible for the body compositional changes associated with aging. Both genders are affected by this age-dependent decline.

The principal physiologic short-term regulators of GH secretion are (a) neural endogenous rhythm, (b) sleep, (c) stress, including exercise, and (d) nutritional and metabolic signals. The integrated result of the multiple inputs into the control of GH secretion is a diurnal rhythm of pulsatile secretion that is fairly constant in periodicity but varies widely in amplitude. The highest peaks in serum GH are seen during phase IV (slow wave) sleep, typically 1 to 2 hours after falling asleep. Pulses of smaller amplitude occur throughout the day, on average approximately every 2 hours.22 Many of these pulses are too small to be measured in conventional assays, and perhaps too small to have much biologic activity. Women of reproductive age generally have higher amplitudes as well as higher inter-peak GH levelsan effect that has been attributed to estrogen (Fig. 12-11). The extent to which metabolic changes due to intermittent meals influence GH secretion is unclear; available data suggest that under physiologic circumstances, such effects are minor at best. However, fasting, malnutrition, and obesity have profound effects on GH production (see later). The variability of serum GH levels makes it clear that sampling at single, random time points cannot be used for diagnostic purposes, and that dynamic testing under standardized conditions or diurnal sampling is necessary to arrive at a diagnosis of GH under- or overproduction.

FIGURE 12-11. Characteristic diurnal profiles of serum growth hormone (GH) levels in a young man (top) and a young woman (bottom). Women of reproductive age have a higher baseline and a greater average pulse amplitude than men. The ordinate is logarithmic to emphasize the presence of small secretory peaks throughout the day; note that the units for GH are pg/mL. The black bar indicates the period of sleep; the shaded bar shows the 1 ng/mL level that represents the detection limit in many conventional assays; the triangles on top indicate the secretory events as defined by a pulse-detection program. (Adapted from Winer LM, Shaw MA, Baumann G. Basal plasma growth hormone levels in man: new evidence for rhythmicity of growth hormone secretion. J Clin Endocrinol Metab 1990; 70:1678.)

There is no known differential secretion or specific stimulus for any of the GH variants. Indeed, they appear to be cosecreted in response to a variety of physiologic or pharmacologic stimuli. However, they have different plasma half-lives, and hence their relative proportions in plasma may differ from that in the pituitary. The average plasma half-life of GH (representing mostly monomeric 22K GH) is ~17 minutes.23 The 20K variant and oligomeric forms have longer half-lives. After secretion, GH binds to GHBP in the circulation. This occurs very rapidly, with maximal binding achieved within a few minutes. The amount of GH bound to GHBP varies, depending on the GHBP level in a given person, the GH concentration (which may partially saturate the GHBP), and the time after a secretory pulse. On the average, 40% to 50% of plasma GH is bound to the GHBP.24 The bound fraction has delayed clearance, dampens the oscillations of serum GH, and serves as a circulating GH reservoir. The GHBP level in serum is not influenced by GH pulses; it exhibits no or minimal diurnal variation. REGULATION OF PLACENTAL GROWTH HORMONE SECRETION GH-V or placental GH is secreted during pregnancy into the maternal (but not fetal) circulation. This process is not regulated by the factors just described for pituitary GH regulation. The principles regulating GH-V secretion are unknown; it may simply be released constitutively as a function of syncytiotrophoblast mass. Plasma GH-V levels increase progressively during the second trimester to reach a plateau in the third trimester (Fig. 12-12). Concomitantly, pituitary GH-N levels are suppressed, presumably via negative feedback by GH-V and IGF-I. GH-V binds to GHR with the same affinity as GH-N; its high plasma levels may be responsible for some of the fluid retention and changes in physical features seen in late pregnancy.

FIGURE 12-12. Plasma levels of the growth hormones GH-N and GH-V during pregnancy. Pituitary GH-N is gradually supplanted by placental GH-V. (hGH, human growth hormone.) (From Baumann G. Growth hormone heterogeneity: genes, isohormones, variants and binding proteins. Endocr Rev 1991; 12:424; as adapted from Frankenne F, Closset J, Gomez F, et al. The physiology of growth hormones [GHs] in pregnant women and partial characterization of the placental GH variant. J Clin Endocrinol Metab 1988; 66:1171.)

REGULATION OF THE GROWTH HORMONE RECEPTOR AND BINDING PROTEIN

For reasons of accessibility, little is known about the regulation of GHRs in human tissues. Therefore, much of the information about GHR regulation is based on (a) animal studies and (b) GHBP measurements in humans, using the GHBP as a surrogate for the GHR. Based on the comparison between direct GHR data in animals and GHBP data in humans, the GHBP level in serum appears to be a reasonable index of GHR abundance in tissues, primarily the liver. The main regulator of GHR/GHBP abundance is ontogeny. In fetal and neonatal life, GHR expression is very low, and serum GHBP levels are correspondingly low. This is a physiologic GH-resistant condition, with high GH and low IGF-I levels. Postnatally a rapid up-regulation of the GHR and GHBP occurs, coincident with the emergence of GH responsivity.25 This process continues throughout childhood, until in the late teens GHBP levels (and presumably GHR levels) reach adult levels. GHBP levels remain constant through adult life until approximately age 60, when a progressive decline ensues that continues until the tenth decade.26 This decline is accompanied by a decline in IGF-I levels and constitutes part of the somatopause. Thus, in old age the changes caused by the decreasing GH secretion are further amplified by the development of GH resistance. Similar changes in GHR expression have been shown in aging animals. Women of reproductive age have slightly higher GHBP levels than men. Estrogens, particularly if given orally, up-regulate serum GHBP, whereas androgens tend to lower GHBP. Interestingly, no discernible change is seen in GHBP level during puberty in either sex. The effect of GH itself on the GHBP in humans is controversial. The majority of studies find no significant change in GHBP levels in GH deficiency or in response to GH treatment. However, some reports show an up-regulation, and others a down-regulation of GHBP by GH. On balance, the effect of GH on serum GHBP (and probably, hence, on GHR) in humans is neutral or at least inconsistent. This differs from the case in rodents, in which the GHR (and the GHBP) expression is up-regulated by GH. Another important regulator of GHR/GHBP expression is nutritional status, probably, at least in part, mediated by insulin. A strong positive correlation exists between body mass index and serum GHBP, and IGF-I levels and GH responsivity vary in parallel as a function of adiposity and nutritional status.13

ACTIONS OF GROWTH HORMONE

Growth hormone has numerous biologic actions, many occurring in concert to enhance protein anabolism and tissue accretion. GH can act directly as well as indirectly through IGF-I, also known as somatomedin C. The mitogenic and proliferative actions of GH are mediated through IGF-I, whereas some of the metabolic actions are direct GH effects. GH has no specific target organ; it acts on most if not all tissues through the ubiquitously expressed GHR. INSULIN-LIKE GROWTH FACTORS AND THEIR BINDING PROTEINS The existence of GH-dependent factors that might mediate the action of GH was first suggested by the report that SO4 incorporation into growth cartilage chondroitin sulfate was reduced by hypophysectomy, but that exposure of cartilage to GH in vitro could not correct the abnormality. In contrast, serum from animals treated with GH was highly effective in restoring SO4 uptake, implicating a GH-dependent sulfation factor.27 Sulfation factor was later renamed somatomedin, and the somatomedin hypothesis was born. It soon became clear that somatomedin was identical to nonsuppressible insulin-like activity, which led to the characterization of IGF-I and IGF-II. The IGFs are proinsulin-like molecules that are produced in many tissues in response to GH and other regulators. IGF-I production is highly GH-dependent, whereas IGF-II production is less dependent on GH. IGFs act both locally in a paracrine/autocrine fashion and distantly in a hormone-like mode. They have mitogenic and metabolic activities and act through the type I IGF receptor, which is structurally similar to the insulin receptor. Six binding proteins for IGF (IGFBP) are present in serum and interstitial fluid; they may either enhance or decrease IGF activity.28 In addition, three IGFBP-related proteins bind IGFs with low affinity,29 for a total of nine proteins in the IGFBP family. IGFBP-1 is insulin dependent and has primarily inhibitory activity in vivo. IGFBP-2 is inversely GH dependent; its biologic role is incompletely understood. IGFBP-3 is the major IGFBP in serum; it is highly GH dependent, and it serves primarily to retain IGFs in the circulation by forming a 150-kDa ternary complex involving IGF, IGFBP-3, and another GH-dependent protein called acid-labile subunit (ALS). This complex is responsible for the high IGF concentrations in serum and serves as a circulating IGF reservoir. Most of the circulating IGF is bound in this complex. IGFBP-4, IGFBP-5, and IGFBP-6 are in part associated with the extracellular matrix and modulate IGF action through restricting or enhancing IGF access to the IGF receptor. IGFBP structure and activity is modulated by proteases that cleave the IGFBPs, thereby decreasing IGF-binding affinity. Protease activity is regulated by a variety of physiologic and pathologic conditions. IGFBPs and their fragments may have activities of their own (such as antianabolic activity) that are independent of their IGF-binding properties. A full description of the complex IGF-IGFBP system is beyond the scope of this chapter; from a clinical standpoint, an understanding of IGFBP-3 and IGFBP-2 is most important because they can be used diagnostically. INSULIN-LIKE GROWTH FACTOR RECEPTOR IGF-I and IGF-II bind to and signal through the IGF receptor (also known as type I IGF receptor). IGF-II also binds to the mannose-6-phosphate/type II IGF receptor, but to date no convincing evidence exists that the lysosomal pathway connected to this receptor is relevant to GH action. The type I IGF receptor is structurally homologous to the insulin receptor; it has a tetrameric structure with two extracellular b-subunits covalently connected to two b-subunits through disulfide bonds. The b-subunits have intrinsic tyrosine kinase activity and signal through a phosphorylation cascade involving IRS-1 and IRS-2, PI3-kinase, MAPK, and other pathways. The IGF receptor differs functionally from the insulin receptor in that it promotes primarily mitogenic/proliferative rather than metabolic activities. Because both receptors share intracellular pathways, it is of liver, where it is expressed at very low levels. The IGF system is described in detail in Chapter 173. Because of the widespread expression of the GHR, the IGFs, and the IGF receptor, discerning which of the ultimate biologic actions of GH are direct and which are IGF-mediated has been difficult. Table 12-1 lists the principal GH actions and attempts to assign them to direct and IGF-mediated pathways. In many cases, this assignment is tentative, and both direct and indirect mediation may occur.

TABLE 12-1. Bioactivities of Growth Hormone (GH)

The principal bioactivities of the GH/IGF system relevant to the intact organism and to clinical medicine are nitrogen retention, protein anabolism, linear growth, lipolysis, insulin antagonism (diabetogenesis), Na+ retention, and negative feedback on GH secretion (short- and long-loop feedback). GH is best viewed as the master postnatal anabolic hormone orchestrating a cascade of activities leading to lean body mass accretion. The exception is adipose tissue, in which GH is largely catabolic. Human GH has lactogenic properties because it can bind to prolactin receptors. This is not a property of animal GHs, and its biologic importance in humans is uncertain. Also unclear is the importance of GH for the immune system; no overt abnormality in immune function is seen in cases of GH deficiency or in GH resistance.

MEASUREMENT OF PLASMA GROWTH HORMONE

The measurement of GH in blood is problematic because of the heterogeneous nature of GH. This heterogeneity is one reason for the observation that different assay designs can yield different results for the same blood sample.30 Plasma GH is measured either by conventional polyclonal radioimmunoassay or by a variety of monoclonal immunoradiometric or immunoenzymatic assays. Monoclonal assays frequently yield lower readings than polyclonal assays. This is in part due to the fact that some of the GH variants are not fully reactive in monoclonal assays, but other, poorly understood matrix effects are also involved. The problem of nonreproducibility of results among assays and laboratories can present a diagnostic dilemma in the classification of a patient as GH deficient or normal. Discrepancies are particularly notable among monoclonal assays. The need exists for a universal standard and assay design that permits comparison of GH determinations among different laboratories. DYNAMIC TESTS OF GROWTH HORMONE SECRETION Because of the pulsatile nature of GH secretion, plasma GH levels vary widely in normal subjects. It follows that a single GH measurement is not diagnostic for an abnormality in the GH axis. Therefore, dynamic testing of the response of plasma GH to standardized provocative or inhibitory tests is mandatory for the evaluation of GH disorders. Table 12-2 lists the dynamic tests in clinical use. This topic has been comprehensively reviewed.31 Among the provocative tests, the insulin-tolerance test is considered the gold standard, as it is a potent and reliable stimulus for GH release. Its disadvantage is that induction of hypoglycemia can be risky, and, therefore, the test must be closely supervised. For the test to be valid, a drop in blood glucose by at least 50% from the starting level must be achieved. GHRH testing has not proven to be as useful as anticipated because the GH response is highly variable, presumably because of differences in prevailing somatostatin tone. Clonidine is used successfully in children, but in adults it is a relatively weak stimulus for GH release. In normal subjects, the highest serum GH levels are seen after combined GHRH-GHRP stimulation. However, the clinical utility of this potent test in the diagnosis of GH deficiency is not yet known.

TABLE 12-2. Dynamic Tests for Growth Hormone Secretion

For research purposes, circadian sampling of blood for GH measurements is probably the best procedure to assess spontaneous GH secretion. The frequency of sampling should be at least every 20 minutes for accurate estimation of secretion rate. This can be done over a 24-hour period or overnight, when most of the GH secretion occurs. A variant of circadian sampling is continuous blood withdrawal by a pump. This yields a pooled average GH level but does not permit detection of secretory pulses. These maneuvers are cumbersome and impractical for general diagnostic purposes; they are largely reserved for investigational use. An additional limitation is that the GH secretion rate is quite variable among normal subjects, and the boundary between normal and deficient GH secretion rates is ill defined. Measurement of urinary GH excretion has been advocated as a possible estimate of GH secretion rate. This method is technically feasible, but only a small fraction (0.01%) of the daily GH production is excreted in the urine. This, combined with substantial day-to-day variability and dependence on renal function, has made it difficult to establish urine GH determination as a reliable index of GH production rate. Measurement of urine GH as a clinical test has therefore been largely abandoned. The GHRH-stimulation test cannot be used to distinguish reliably between hypothalamic and pituitary GH deficiency. This is similar to the failure of testing with other hypothalamic releasing hormones (gonadotropin-releasing hormone [GnRH], thyrotropin-releasing hormone [TRH], corticotropin-releasing hormone [CRH]) to clearly differentiate hypothalamic from pituitary disease. Pituitary function tests must be interpreted in the context of other clinical assessments. They can yield misleading results in some conditions not associated with hypothalamic-pituitary disease. Examples are obesity, hypothyroidism, or hypercorticism, in which the GH response to all stimuli tends to be blunted. Conversely, malnutrition, catabolic conditions, or stress may be associated with elevated GH levels that are not normally suppressible. CLINICAL USE OF INSULIN-LIKE GROWTH FACTOR-I AND INSULIN-LIKE GROWTH FACTORBINDING PROTEIN MEASUREMENTS Plasma IGF-I can be used as a surrogate measurement for GH. Unlike GH, its plasma level fluctuates very little throughout the day. Because IGF-I is highly GH dependent, its serum level reflects GH secretory status and can be used as an index for integrated GH secretion. The same is true for IGFBP-3, another GH-dependent factor. However, such measurements must be interpreted in the context of other clinical information. IGF-I is also highly dependent on nutritional status and is influenced by age, catabolic disease, etc. IGF-I and IGFBP-3 levels are best used in conjunction with dynamic GH measurements or for longitudinal follow-up in conditions that are already diagnosed as primary GH disorders. IGFBP-2 is inversely GH dependent and may be used as an ancillary measurement. IGF-II, although GH dependent to some degree, has not proven to be of clinical use in the diagnosis of GH disorders. In the measurement of IGFs, separation of IGF from the IGFBPs before measurements is imperative to avoid spurious results due to interference of IGFBPs in the assay. Several strategies can be used to accomplish this, such as acid ethanol extraction, C18 column separation, and acid gel filtration.

ABNORMAL SECRETION OF GROWTH HORMONE

DEFICIENCY OF GROWTH HORMONE GH deficiency results from various causes, and the clinical manifestations depend on the age of the patient when the disease first occurs. The condition may be hereditary or acquired, and GH deficiency may be isolated or combined with other pituitary hormone deficiencies. Subdividing GH deficiency into childhood onset and adult onset categories is useful, because some of the clinical manifestations are different. ETIOLOGY Congenital Growth Hormone Deficiency. Table 12-3 lists the causes of GH deficiency. Among the congenital causes, between 5% and 30% have a familial pattern, which suggests a genetic basis. Known genetic causes of combined pituitary hormone deficiency involve the transcription factors Prop-1 and Pit-1, both members of the POU-homeodomain family of proteins critical for tissue differentiation. These genes are sequentially expressed during pituitary ontogeny, and Pit-1 is dependent on the expression of Prop-1 (of which the full name is Prophet of Pit-1). Inactivating mutations in the PROP1 gene lead to defective pituitary development, with TSH, luteinizing hormone, follicle-stimulating hormone, GH, and prolactin deficiency.32 A two-base deletion in a hot spot in the PROP1 gene, with resulting frameshift, is the most common cause.33 The clinical phenotype is one of familial dwarfism, with hypothyroidism and hypogonadism. Of interest, pituitary enlargement frequently occurs in affected patientsa finding whose pathogenesis is not clear. Mutations in the PIT1 gene are another cause of combined pituitary hormone deficiency.34 At least eight different mutations have been described; most are recessively inherited, but some are dominant negative (i.e., the abnormal gene product interferes with that derived from the normal allele). Pit-1 is important for development of somatotropes, lactotropes, and thyrotropes; affected patients suffer from GH, prolactin, and TSH deficiency. They do not develop pituitary enlargement as is seen in patients with PROP1 mutations. Known genetic causes of isolated GH deficiency include inactivating mutations in the GHRHR gene and in the GHN gene. GHRH action is required for somatotrope proliferation (late in pituitary development) and for GH synthesis and secretion. A defective GHRHR impacts all these mechanisms, and affected patients have pituitary hypoplasia and severe isolated GH deficiency.35,36 The clinical phenotype is one of proportionate dwarfism with relative microcephaly, normal fertility, and normal lactation.36 Mutations in the GHN gene affect GH production directly.37 The GH locus is prone to deletions because of gene duplication, which includes homologous sequences in the regions flanking the duplicated genes. Deletions of 6.7, 7.0, and 7.6 kb are typical; affected homozygous patients have severe GH deficiency (termed type IA). A similar phenotype occurs with other severely inactivating mutations, such as nonsense and frameshift mutations. A milder form (termed type IB) is seen with less disabling mutations, such as missense mutations in which some, albeit abnormal, GH is produced. In type IA, secondary resistance to exogenous GH administration is seen frequently, but not always, because of the formation of high-titer anti-GH antibodies to the foreign GH protein. Patients with type IB, on the other hand, respond well to exogenous GH because they are immune tolerant to GH. A special case in this category is a bioinactive GH.38 Dominantly inherited GHN gene mutations (termed type II) are caused by splice-site mutations in one allele that result in skipping of exon 3, with the resultant abnormal GH protein exerting a deleterious (dominant negative) influence on the normal GH produced by the intact allele. Improper disulfide pairing/aggregation or deranged transport to secretory granules are postulated mechanisms. In one case, a mutant GH was shown to act as an antagonist at the level of the GHR.39 Yet another type of isolated GH deficiency is inherited in an X-linked manner (termed type III). The gene involved is unknown but resides close to the gene for Bruton tyrosine kinase, an enzyme that is crucial for B-lymphocyte function. Hence, this type of GH deficiency is usually associated with hypo- or agammaglobu-linemia. Interestingly, defects in the GHRH gene have not been identified to date, despite the fact that many cases of idiopathic GH deficiency are thought to be due to GHRH deficiency.

TABLE 12-3. Causes of Growth Hormone (GH) Deficiency

The majority of cases of congenital GH deficiency (isolated or combined) are sporadic; they are thought to be caused by birth trauma, cerebral insults, and congenital malformations or tumors affecting hypothalamic-pituitary development or function. Approximately 10% of such patients have abnormal magnetic resonance imaging (MRI) scans. In one family affected with two cases of septo-optic dysplasia, an inactivating mutation has been found in the gene encoding the early developmental transcription factor Hesx1, also known as Rpx (Rathke Pouch Homeobox).40 The incidence of GH deficiency is estimated as 1 per 4000 to 10,000 births. Acquired Growth Hormone Deficiency. GH deficiency can be acquired at any time during the life span, depending on when a hypothalamic-pituitary lesion or insult occurs. As already indicated, normal aging is associated with declining GH secretion during adulthood, and senescence resembles progressive GH insufficiency. During the growing years, the hallmark of GH deficiency is growth retardation; in adult life, the signs are more subtle changes in body composition and metabolism. Because of the lack of obvious manifestations, GH deficiency in adults is often not recognized, being diagnosed only after a long delay, or not considered clinically important. Moreover, only recently has GH deficiency in adults become amenable to therapy. With pituitary lesions, the pattern of loss of pituitary hormones tends to follow a certain order: GH, gonadotropins, and prolactin tend to be lost early, whereas TSH and ACTH are affected later. This may be related to the abundance and/or spatial distribution of the respective cell types, with somato-tropes being the most prevalent (accounting for ~50% of the anterior pituitary mass). CLINICAL PRESENTATION AND DIAGNOSIS In childhood, GH deficiency soon declares itself because of growth retardation. The GH dependence of somatic growth begins at or shortly after birth. If coexistent TSH deficiency is present, the resulting hypothyroidism contributes to and aggravates growth retardation. Organic causes for GH deficiency should be sought, but frequently no lesion is found, leading to the diagnosis of idiopathic GH deficiency. This entity is attributed to partial GHRH deficiency, although direct proof is difficult to obtain. Patients with idiopathic GH deficiency usually have blunted, but not absent, GH responses to secretagogues. As already mentioned, the boundaries between normal, insufficient, and deficient GH secretion are ill defined because GH secretion rates are highly variable in normal people. This may render the diagnosis of idiopathic GH deficiency difficult. Growth retardation due to GH deficiency results in proportionate dwarfism and a progressive deviation from the normal growth curve. Untreated severe isolated GH deficiency results in adult heights of 120140 cm in males and 110130 cm in females. Other manifestations of GH deficiency during childhood are hypoglycemia, micropenis, and craniofacial abnormalities with saddle nose and facial hypoplasia. Patients tend to be chubby because of increased adipose tissue, particularly in the truncal area. Bone maturation is delayed by several years, and puberty is delayed by 2 to 3 years. Postpubertal men have a characteristic, high-pitched voice. The impact of GH deficiency on somatic growth and development is described in detail in Chapter 18, Chapter 92 and Chapter 198 Figure 12-13 shows the phenotype of adult patients with isolated GH deficiency. GH deficiency must be differentiated from other forms of short stature (constitutional, idiopathic, familial, etc.) that are much more common.

FIGURE 12-13. A, Two siblings with familial isolated growth hormone (GH) deficiency. Note the proportionate short stature and normal sexual development. B, The child-like, finely wrinkled face of a 41-year-old man with isolated GH deficiency.

In adults, GH deficiency should be suspected when a pituitary or hypothalamic lesion is found, or when a history of childhood GH deficiency is present. Patients who had idiopathic GH deficiency during childhood may or may not be GH deficient as adults, so that retesting of GH secretion in adulthood is required. Approximately two-thirds of such patients are found to have normal GH secretion as adults.41 One explanation for this phenomenon is that GH secretion rates during childhood and particularly during puberty are higher than in adult life, and that a partial limitation in GH production may yield levels that are inadequate during the growth years but sufficient in adulthood. Patients with pituitary lesions should be tested for GH deficiency even in the absence of other pituitary hormone deficiencies because somatotrope dysfunction may be an early manifestation of pituitary failure. This has therapeutic consequences, because such patients may benefit from GH treatment. The clinical manifestations of the adult GH deficiency syndrome are listed in Table 12-4.

TABLE 12-4. Clinical Manifestations of Growth Hormone Deficiency in Adults

Diagnosis relies on the demonstration of an underlying cause (a pituitary or hypothalamic lesion, a genetic defect, or a history of radiation), on provocative tests of GH secretion, and on IGF-I measurement. A low serum IGFBP-3 and a high IGFBP-2 level can be used as corroborative indicators. The typical, but arbitrary, cutoff for a normal peak GH level in response to provocative tests is 5 to 7 ng/mL in a polyclonal assay, and 2.5 to 3.5 ng/mL in a monoclonal assay. The Growth Hormone Research Society Consensus Guidelines for adult GH deficiency define a cutoff of 3 ng/mL in response to an insulin-tolerance test as severe GH deficiency, with a gray zone between 3 and 5 ng/mL, using a polyclonal assay.42 These cutoffs cannot be absolute because of the problems inherent in the assays mentioned above, and because of interindividual variations in GH response. Furthermore, tests that are less potent than the insulin-tolerance test may yield a lower GH response. Therefore, provocative tests must be interpreted in the clinical context. A finding of a low serum IGF-I level is helpful in the absence of other causes for decreased IGF production, such as nutritional deprivation. However, difficulty arises because of the wide normal range for IGF-I levels. IGF-I levels are highly age dependent and must be judged against age-appropriate normative values. In very young children, in whom IGF-I levels are naturally low, differentiation of GH deficiency from normal production may be difficult based on the IGF-I level. In adults, particularly older adults, considerable overlap also exists between the normal IGF-I range and that seen in GH deficiency (Fig. 12-14).

FIGURE 12-14. Serum insulin-like growth factor-I (IGF-I) levels in adult hypopituitary patients, aged 17 to 77 years (closed circles), compared to those in age-matched normal controls (open circles). All patients had growth hormone deficiency, as assessed by the insulin-tolerance test, and deficiency of other pituitary hormones. Note the overlap between normal and hypopituitary IGF-I levels. (S, assay sensitivity for IGF-I [25 ng/mL]; n, number of subjects tested.) (Adapted from Hoffman DM, O'Sullivan AJ, Baxter RC, Ho KKY. Diagnosis of growth-hormone deficiency in adults. Lancet 1994; 343:1064.)

TREATMENT GH deficiency can be treated successfully with recombinant human GH, given subcutaneously once daily at bedtime. The bedtime dosing is designed to at least partially mimic the normal GH secretory pulse occurring after sleep onset. The dose for children is 25 to 50 g/kg per day; for adults it is 3 to 12 g/kg per day. Studies have suggested that in adults administration of an absolute daily dose (e.g., 200400 g per day) may be more appropriate than dosing based on kilograms of body weight. This regimen is more physiologic because normal GH secretion is inversely related to adiposity, and weight-based dosing does not consider this. Women of reproductive age or those on estrogen treatment need higher GH doses to maintain normal IGF-I levels, as would be expected from the higher GH-secretion rates in normal women. GH therapy is highly effective in reversing the manifestations of GH deficiency. The effect on children is growth acceleration to normal or even catch-up growth velocity, and normal adult height can be achieved provided that treatment starts early and is maintained until epiphyseal fusion occurs. Pediatric use of GH is described in detail in Chapter 18, Chapter 92 and Chapter 198. In adults, the principal effects are normalization of body composition (i.e., a decrease in adipose tissue, particularly visceral fat; Fig. 12-15), an increase in lean body mass and body fluid, an increase in bone mineral density, and a decrease in low-density lipoprotein cholesterol.43 Other effects, such as improved muscular strength, endurance, energy, and psychosocial well-being, have also been reported in many, but not all, studies. Several years are probably required to realize the full benefit of GH therapy in adults. GH therapy for adults has begun only in the past decade, and its ultimate outcome is still not fully known. One important unanswered question is whether the shortened life span of hypopituitary patients44 can be prolonged with GH therapy.44a

FIGURE 12-15. Transverse computed tomographic scan at the L3L4 level before (A) and after 6 months of growth hormone therapy (B). Note the reduction of both visceral and subcutaneous adipose tissue (dark areas). (R, right; L, left.) (From Bengtsson B-A, Edn S, Lnn L, et al. Treatment of adults with growth hormone [GH] deficiency with recombinant human GH. J Clin Endocrinol Metab 1993; 76:309.)

GH dosing is monitored in children by observation of growth velocity and serum IGF-I, and in adults by serum IGF-I levels, which should be maintained in the age-appropriate normal range. Side effects are rare in children; adults, particularly the elderly, are more prone to side effects, such as transient edema, carpal tunnel syndrome, arthralgia, and myalgia, but these symptoms can be minimized or largely avoided by proper dosing. GH secretion naturally declines with age, and dosing accordingly should be adjusted based on age. Carbohydrate intolerance has not proved to be a major problem in GH-replacement therapy. Typically, insulin levels increase initially, but glucose levels remain normal. This is physiologically appropriate and inherent in the correction of the GH-deficient state. Nevertheless, glycemia should be monitored after institution of GH therapy, which may unmask a latent diabetic propensity. As adiposity diminishes over time, insulin levels tend to fall back to baseline values. GROWTH HORMONE INSENSITIVITY A condition that clinically resembles GH deficiency is GH insensitivity, which can range from absolute GH resistance to varying degrees of mild GH insensitivity. It may be congenital/genetic or acquired. GENETIC CAUSES Genetic GH resistance, also known as Laron syndrome, is due to inactivating mutations in the GHR gene.45 Inheritance is autosomal recessive. Over 250 patients have been reported worldwide; significant consanguineous clusters have been described in Israel, Ecuador, and the Bahamas. Over 35 different mutations (deletions and nonsense, splice-site, and missense mutations) have been identified to date. Most are found in the extracellular domain of the receptor and inactivate the GH binding site. Other mutations include one that interferes with receptor dimerization, some that result in skipping of exon 8 (which encodes the trans-membrane domain), and some that severely truncate the intra-cellular domain. Of interest is a mutation that has a dominant negative effect because the abnormal, truncated receptor interferes with normal receptor signaling owing to heterodimer formation between normal and mutant receptor.46 The type of mutation determines whether the GHBP activity in plasma is absent or low (the majority of cases), normal, or even high (e.g., when the mutant receptor lacks a transmembrane domain).47 The phenotype is similar to that of severe GH deficiency. Children are near-normal at birth but show all the signs and symptoms associated with GH deficiency described earlier. Facial dysplasia may be even more pronounced than in GH deficiency. Heterozygotes are physically normal. IGF-I and IGFBP-3 levels are very low, but in contrast to GH deficiency, GH levels are elevated, both in the basal state and after stimulation. This is due to lack of negative feedback on GH production by IGF-I and by GH. Treatment with GH is ineffective, but therapy with IGF-I has resulted in growth rates that approach, but do not quite reach, those seen with GH therapy in GH-deficient children. Numerous problems exist with IGF-I therapy, such as abnormal pharmacokinetics because of low IGFBP-3 levels, limited availability of IGF-I, and absence of direct GH effects, that render therapy for this rare condition difficult. Milder heterozygous mutations or polymorphisms in the GH receptor have been suggested as a possible explanation of idiopathic short stature.48 However, in view of the fact that heterozygous relatives of Laron syndrome patients are of normal height, it is difficult to explain how milder heterozygous mutations would cause short stature. Pygmies have a primary, possibly genetic, form of partial GH resistance. This has been attributed to low GHR expression, based on low serum GHBP levels.49 Pygmies may also have partial IGF-I insensitivity, as postulated from the lack of an IGF-mediated proliferative response of lymphocytes derived from pygmies.50 The exact reasons for pygmies' short stature remain to be elucidated.

ACQUIRED CAUSES Many medical conditions are associated with GH insensitivity of varying degrees (Table 12-5). These conditions are characterized by low IGF-I levels, elevated GH levels, and low GHBP levels. Decreased hepatic GHR expression has been demonstrated in corresponding animal models. GH insensitivity contributes to and aggravates catabolism in those disorders that are already prone to catabolism because of oversecretion of stress hormones such as glucocorticoids, catecholamines, and glucagon. To date, influencing the GH-resistant state has not been possible, except by treating the underlying cause. Attempts to overcome the resistance with large doses of exogenous GH have had mixed results, depending on the type of catabolic condition being treated. A study of GH treatment in critically ill patients resulted in a higher mortality. Therefore such therapy cannot be recommended at present except in the context of carefully controlled studies.

TABLE 12-5. Conditions Associated with Growth Hormone (GH) Resistance

HYPERSECRETION OF GROWTH HORMONE (ACROMEGALY) Overproduction of GH, usually by a pituitary tumor, causes acromegaly in adults and pituitary gigantism in prepubertal children. Acromegaly is a disease of insidious onset that usually is not recognized until the progressive overgrowth of connective tissue and bone has caused striking changes of appearance. The prevalence of acromegaly is estimated to be ~1 in 20,000.51 Its peak incidence is between the fourth and sixth decades. ETIOLOGY AND PATHOGENESIS The great majority (>99%) of cases of acromegaly are caused by a pituitary adenoma, either pure somatotrope or occasionally somatomammotrope in origin. The adenoma is usually sporadic but may occur as part of the multiple endocrine neoplasia type 1 (MEN1) syndrome. Rarely, acromegaly may result from somatotrope hyperplasia, induced by excess production of GHRH from an ectopic source (carcinoids, islet cell tumors, small cell carcinoma) or by hamartomas or gangliocytomas in the hypothalamic or pituitary area. GHRH was first isolated from such an ectopic source in the pancreas,16,17 but only 50 cases of ectopic GHRH-induced acromegaly have been described.52 Only one clearly documented case exists of acromegaly due to ectopic production of GH (by an islet cell tumor).53 McCune-Albright syndrome is another disorder that can be associated with acromegaly and gigantism. Although the role of a primary pituitary or hypothalamic disorder as the initiating cause of acromegaly was disputed for years, it is now clear that many GH-secreting tumors are monoclonal and harbor an activating mutation of the Gsa protein subunit. As indicated above, the GHRH-R signals through theGs-cAMP pathway, and constitutive activation of Gs essentially bypasses GHRH-regulated somatotrope growth, GH synthesis, and GH secretion. Thus, in a substantial proportion of tumors, a somatic mutation seems to be responsible for the disease.54 McCune-Albright syndrome, a disorder with widespread constitutive activation of Gsa, probably causes GH overproduction through the same mechanism.55 In MEN1, tumorigenesis is a function of the loss of both alleles of a tumor-suppressor gene, the menin gene.56 Thus, except in rare cases of ectopic (or eutopic) GHRH overproduction, acromegaly appears to be a primary pituitary disorder. CLINICAL AND LABORATORY MANIFESTATIONS The clinical manifestations of acromegaly relate to the hormonal effects of GH and IGF-I on the one hand, and to local tumor-mass effects on the other. Table 12-6 lists the principal clinical and laboratory findings recorded in a large series of patients with acromegaly; Figure 12-16, Figure 12-17 and Figure 12-18 illustrate typical features. The symptoms and signs of acromegaly can be divided into three categories: (a) physical changes related to excessive amounts of GH and IGF-I, (b) metabolic effects of excessive amounts of GH, and (c) local effects of the pituitary tumor.

TABLE 12-6. Clinical and Laboratory Findings in 57 Patients with Acromegaly

FIGURE 12-16. A 64-year-old man with acromegaly. Note the prominent jaw, the large zygomatic arches and supraorbital ridges. The bony overgrowth results in a comparative hollowing of the temporal region. The nose and ears are enlarged. The skin folds are exaggerated, the skin is tough and oily, and the sebaceous glands and pores are enlarged.

FIGURE 12-17. AC, Progressive acromegalic changes in a 58-year-old man. Old photographs are useful to evaluate whether a diagnosis of acro-megaly should be considered or to document progression of the disease.

FIGURE 12-18. A 64-year-old man with acromegaly. Note the prominent jaw, the large zygomatic arches and supraorbital ridges. The bony overgrowth results in a comparative hollowing of the temporal region. The nose and ears are enlarged. The skin folds are exaggerated, the skin is tough and oily, and the sebaceous glands and pores are enlarged.

Identifying acromegaly on the basis of appearance alone can be difficult. Because these changes occur slowly, often only the examination of old photographs can confirm the suspicion and help determine the time of onset (see Fig. 12-17). A delay in diagnosis is unfortunate because a surgical cure is much more likely when the tumor is small, and the more prominent bony changes are only partially reversible. In experienced surgical hands, transsphenoidal adenomectomy is curative for the majority of microadenomas (<1 cm in diameter), but cures are more difficult to achieve for macroadenomas (see later). Somatotrope tumors have a propensity for local invasion, including of bone, which makes eradication of all cells extraordinarily difficult once the tumor has progressed beyond the boundaries of the pituitary gland. The physical changes (see Fig. 12-16, Fig. 12-17 and Fig. 12-18 include a general coarsening of features. Acral growth of bone and soft tissues is seen (a finding that gave the disease its name). Patients typically have large, spadelike hands, with a characteristic swelling of subcutaneous tissues that is more doughy than edematous (see Fig. 12-18). This results in a change in ring, glove, and shoe sizekey anamnestic elements that should be actively sought. Bony overgrowth in the skull leads to frontal bossing, mandibular growth, prognathism, and dental malocclusion. Bony overgrowth leads to bone spurs in the spine and around the large joints, sometimes giving rise to spinal stenosis or other nerve compression syndromes. Carpal tunnel syndrome is a classic early sign of acromegaly; it is also seen with exogenous GH treatment for GH deficiency (see earlier). A barrel deformity of the chest may occur as a result of rib elongation and kyphosis. Laryngeal overgrowth and thickened vocal cords, together with cranial hyperpneumatization, causes a characteristic deep, sonorous voice. The tongue is large and hypertrophic, sometimes with dental impressions at its borders. Sleep apnea is a common finding, at least in part because of pharyngeal obstruction. The skin is coarse and oily, and skin tags are common. Patients complain of excessive sweating and body odor. Galactorrhea may be present in women. The large joints may show deformities, and the calvarium is thickened. Arthralgias and osteoarthritis-like arthropathies are common. General visceromegaly is present. Cardiac enlargement commonly occurs, and abnormalities invariably are seen on tests of dynamic cardiac function.57 This probably results from a combination of direct effects of GH and IGF-I on myocardial growth and the sequelae of hypertension, hyperlipidemia, and diabetes, all of which are associated with acromegaly. Before effective means of treating acromegaly were available, premature cardiovascular death was responsible for a significantly shortened life span. Acromegaly predisposes to colonic polyp formation,58 and the incidence of colon cancer is probably also increased.59 The metabolic effects of excess GH and IGF-I relate to sodium retention, insulin antagonism, phosphate retention, abnormal calcium metabolism, and heightened bone turnover. One-third of acromegalic patients have hypertension.60 Carbohydrate intolerance is common, and frank diabetes mellitus is seen in 10% to 20% of patients. Hyperlipidemias of various types occur in the presence and even absence of diabetes. Serum phosphorus levels are elevated, as are levels of 1,25-hydroxyvitamin D. There is increased calcium absorption and excretion, as well as hydroxyproline excretion as a manifestation of increased bone turnover. Local tumor effects include headaches, visual field abnormalities due to optic chiasm compression, oculomotor paresis due to cavernous sinus invasion, galactorrhea due to stalk compression, and panhypopituitarism (uncommon). Headaches are characteristic for acromegaly, may be severe, and often exceed those seen with other pituitary tumors of similar or larger size. A cephalgic property may be inherent in excess GH or some other substance cosecreted from somatotrope tumors, as is also suggested by the sudden and dramatic relief of headaches by administration of somatostatin or its analogs (see later). DIAGNOSTIC TESTING In acromegaly, GH secretion remains pulsatile, with increased amplitude and failure to completely cease between pulses. Pulse frequency may also be increased. GH levels vary widely, both within and outside the normal range. Therefore, a random GH level is not diagnostic of acromegaly, just as it cannot be diagnostic of hypopituitarism. The mixture of GH molecular variants secreted in acromegaly does not differ from the mixture secreted normally. Dynamic testing of GH suppressibility is required for diagnosis. The diagnosis of acromegaly has been facilitated by IGF-I measurement. An elevated serum IGF-I level is almost pathognomonic for acromegaly, as few other conditions cause high IGF-I levels. Furthermore, IGF-I is highly sensitive to even mild elevation of GH secretion. When acromegaly is suspected, serum IGF-I measurement should be performed as a screening test. This should then be followed by a standard oral glucose-tolerance test (75 or 100 g glucose). Glucose normally suppresses GH secretion, with plasma levels of <1 ng/mL within an hour or two of glucose administration. In acromegaly, GH levels are not normally suppressible and may even increase in response to glucose. Lack of suppression by glucose to <1 ng/mL is the definitive diagnostic test in the proper clinical setting suggesting acromegaly. The severity of clinical manifestations is notoriously poorly correlated with serum GH levels; correlation with IGF-I levels may be better. IGF-I may be elevated, and clinical disease present, even at a mean GH level of 5 ng/mL.61 The hallmark of acromegaly, even in such mild cases, is the absence of cessation of GH secretion, which results in higher than normal interpulse nadirs and a lack of glucose-induced suppressibility of GH to <1 ng/mL. Frequently, patients with acromegaly have a paradoxical GH response to TRH or GnRHfindings that support the diagnosis but are not entirely specific. Measurements of IGFBPs or GHBP are not useful in the diagnosis of acromegaly. In the ectopic GHRH syndrome, peripheral serum GHRH levels are substantially elevated.52 However, this syndrome is so rare that routine screening is not recommended. Pituitary imaging (MRI or computed tomographic scanning) frequently reveals a macroadenoma that may extend in any direction. Conventional radiography may show a ballooned sella turcica, the bony skull deformities described earlier, and enlarged sinuses.

TREATMENT The goal of treatment is two-fold: complete eradication of the tumor and normalization of GH and IGF-I secretion. The criteria for cure have changed over the years, with progressive lowering of the GH threshold from 10 to 5 to 2 ng/mL. Cure rates quoted in the literature are correspondingly variable. Criteria for a true cure should be very stringent, with normal suppressibility of GH and serum IGF-I in the normal range. Many patients, although substantially improved with respect to clinical disease, do not achieve this criterion, but rather have low-grade, residual acromegaly. The primary form of therapy is transsphenoidal adenomectomy. For microadenomas (<1 cm), the surgical cure rate in experienced hands ranges from 60% to 90%; for larger tumors it is from 35% to 71%.62,63 and 64 These numbers partly depend on the stringency of the criteria used to determine a cure. Early diagnosis is the key to a successful eradication of a small tumor. Selective removal of the adenoma should leave the rest of the pituitary intact, and postoperative hypopituitarism is uncommon. It cannot be emphasized enough that surgical outcomes depend on the experience and skill of the neurosurgeon; this procedure should be performed only in specialized centers. Radiation therapy is used for patients who are not cured by surgery alone. It can be delivered as conventional x-ray therapy, as proton beam therapy, or in the form of the gamma knife (multi-port, collimated cobalt-60 therapy). Radioactive pituitary implants have largely been abandoned. All three radiation modalities are attended by development of hypopituitarism. Moreover, they are slow in their onset of action, with conventional x-ray therapy requiring 5 to 10 years or more to lower GH levels to an acceptable range. Proton beam therapy acts faster but has been complicated by optic and oculomotor nerve damage and seizure disorders; its availability is limited to institutions that possess a cyclotron. Experience with gamma knife therapy is still limited; it is reserved for small tumors and, thus, is not well suited for those patients with macroadenomas who are most in need of radiation therapy. Medical therapy is available in the form of dopamine agonists (bromocriptine, cabergoline) and somatostatin analogs (octreotide). These agents are largely reserved for patients with residual disease after surgery; they have generally not been used as primary therapy, although one study suggested that octreotide monotherapy may be considered for patients for whom a surgical cure is not anticipated.65 In contrast to their great efficacy in the treatment of prolactinomas, dopamine agonists are not very effective in acromegaly, either in lowering GH levels or in shrinking tumor size. High doses are needed, and patient acceptance is poor because of side effects. Octreotide, on the other hand, is quite effective; its main drawback is that it has to be given by injection three times a day. However, long-acting analogs (lanreotide) or formulations (octreotide LAR) have been developed and are effective when injected in 20- to 40-mg doses at biweekly or monthly intervals. Octreotide can have a marked, immediate effect in relieving headachesa phenomenon that is poorly understood. Future medical therapy may include use of a GH antagonist that is presently in clinical trials. Antagonism at the GHR level is based on the disabling of binding site 2 in the GH molecule, thereby preventing productive GHR dimerization. Preliminary results show that this antagonist is very effective in blocking GH action in patients with residual acromegaly. The hope is that the combination of agents that decrease GH secretion with one that blocks GH action will finally meet the difficult challenge of treating patients who suffer from residual acromegaly despite the best efforts at optimizing surgical and radiologic therapy. HYPERSENSITIVITY TO GROWTH HORMONE GH hypersensitivity states have not been widely appreciated, but at least one such condition is very common: obesity. In obesity, IGF-I levels tend to be high or normal despite decreased GH secretion. GHBP levels are high, likely indicating increased GHR expression.13 This constellation is the opposite of that in malnutrition, which is associated with GH insensitivity. Obese children grow faster than lean children66an old observation that was not understood until recently. Hyperinsulinemia was proposed as an explanation, but insulin does not promote linear growth. Other than enhanced growth, the pathophysiologic consequences of the GH hypersensitivity associated with obesity are not clear; further studies are required to determine whether it contributes to increased morbidity. Considerable advances have been made in our understanding of the GHRHGHIGF axis during the past few years. The genetic control of pituitary development and somatotrope function has been partially elucidated, with human disease linked to several control factors, such as Rpx, Prop-1, Pit-1, GHRH-R, and Gsa. GH action has become far better understood through detailed molecular modeling of the GH-GHR interaction, the study of GHR mutations and their functional consequences, the partial elucidation of the GHR signaling cascade, and the assessment of GHR regulation through GHBP measurements in various physiopathologic conditions. GH therapy for adults with hypopituitarism has come of age, with several new insights gained about the importance of GH in maintaining normal body composition and metabolism during adult life. CHAPTER REFERENCES
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. Chen EY, Liao YC, Smith DH, et al. The growth hormone locus: nucleotide sequence, biology and evolution. Genomics 1989; 4:479. Baumann G. Growth hormone heterogeneity: genes, isohormones, variants and binding proteins. Endocr Rev 1991; 12:424. Ultsch MH, Somers W, Kossiakoff AA, de Vos AM. The crystal structure of affinity-matured human growth hormone at 2 resolution. J Mol Biol 1994; 236:286. Stolar MW, Amburn K, Baumann G. Plasma big and big-big growth hormone (GH) in man: an oligomeric series composed of structurally diverse GH monomers. J Clin Endocrinol Metab 1984; 59:212. Leung DW, Spencer SA, Cachianes G, et al. Growth hormone receptor and serum binding protein: purification, cloning and expression. Nature 1987; 330:537. Kelly PA, Djiane J, Postel-Vinay M-C, Edery M. The prolactin/growth hormone receptor family. Endocr Rev 1991; 12:235. Godowski PJ, Leung DW, Meacham LR, et al. Characterization of the human growth hormone receptor gene and demonstration of a partial gene deletion in two patients with Laron-type dwarfism. Proc Natl Acad Sci U S A 1989; 86:8083. Mercado M, Davila N, McLeod JF, Baumann G. Distribution of growth hormone receptor messenger ribonucleic acid containing and lacking exon 3 in human tissues. J Clin Endocrinol Metab 1994; 78:731. Wickelgren RB, Landin KL, Ohlsson C, Carlsson LM. Expression of exon 3-retaining and exon 3-excluding isoforms of the human growth hormone-receptor is regulated in an interindividual, rather than a tissue-specific manner. J Clin Endocrinol Metab 1995; 80:2154. Dastot F, Sobrier M-L, Duquesnoy P, et al. Alternatively spliced forms in the cytoplasmic domain of the human growth hormone (GH) receptor regulate its ability to generate soluble GH-binding protein. Proc Natl Acad Sci U S A 1996; 93:10723. Cunningham BC, Ultsch M, De Vos AM, et al. Dimerization of the extracellular domain of the human growth hormone receptor by a single hormone molecule. Science 1991; 254:821. Carter-Su C, Smit LS. Signaling via JAK kinases: growth hormone receptor as a model system. Recent Prog Horm Res 1998; 53:61. Baumann G. Growth hormone binding proteins: state of the art. J Endo-crinol 1994; 141:1. Kratzsch J, Selisko T, Birkenmeier G. Identification of transformed alpha 2-macroglobulin as a growth hormone-binding protein in human blood. J Clin Endocrinol Metab 1995; 80:585. Sornson MW, Wu W, Dasen JS, et al. Pituitary lineage determination by the Prophet of Pit-1 homeodomain factor defective in Ames dwarfism. Nature 1996; 384: 327. Rivier J, Spiess J, Thorner M, Vale W. Characterization of a growth hormone-releasing factor from a human pancreatic islet tumour. Nature 1982; 300:276. Guillemin R, Brazeau P, Bohlen P, et al. Growth hormonereleasing factor from a human pancreatic tumor that caused acromegaly. Science 1982; 218:585. Barinaga M, Yamamoto G, Rivier C, et al. Transcriptional regulation of growth hormone gene expression by growth hormone releasing factor. Nature 1983; 306:84. Reisine T, Bell GI. Molecular biology of somatostatin receptors. Endocr Rev 1995; 16:427. Bowers CY, Momany FA, Reynolds GA, Hong A. On the in vitro and in vivo activity of a new synthetic hexapeptide that acts on the pituitary to specifically release growth hormone. Endocrinology 1984; 114:1537. Smith RG, Van der Ploeg LH, Howard AD, et al. Peptidomimetic regulation of growth hormone secretion. Endocr Rev 1997; 18:621.

21a. Kojima M, Hosada H, Date Y, et al. Ghrelin is a growth-hormone-releasing acylated peptide from stomach. Nature 1999; 402:656. 22. Winer LM, Shaw MA, Baumann G. Basal plasma growth hormone levels in man: new evidence for rhythmicity of growth hormone secretion. J Clin Endocrinol Metab 1990; 70:1678. 23. Faria ACS, Veldhuis JD, Thorner MO, Vance ML. Half-time of endogenous growth hormone (GH) disappearance in normal man after stimulation of GH secretion by GH-releasing hormone and suppression with somatosta-tin. J Clin Endocrinol Metab 1989; 68:535. 24. Baumann G, Amburn K, Shaw M. The circulating growth hormonebinding protein complex: a major constituent of plasma growth hormone in man. Endocrinology 1988; 122:976. 25. Daughaday WH, Trivedi B, Andrews BA. The ontogeny of serum GH binding protein in man: a possible indicator of hepatic GH receptor development. J Clin Endocrinol Metab 1987; 65:1072. 26. Maheshwari H, Sharma L, Baumann G. Decline of plasma growth hormone binding protein in old age. J Clin Endocrinol Metab 1996; 81:995. 27. Salmon WD Jr, Daughaday WH. A hormonally controlled serum factor which stimulates sulfate incorporation by cartilage in vitro. J Lab Clin Med 1957; 49:825. 28. Clemmons DR. Insulin-like growth factor binding proteins and their role in controlling IGF actions. Cytokine Growth Factor Rev 1997; 8:45. 29. Burren CP, Wilson EM, Hwa V, et al. Binding properties and distribution of insulin-like growth factor binding proteinrelated protein 3 (IGFBP-rP3/NovH), an additional member of the IGFBP superfamily. J Clin Endocrinol Metab 1999; 84:1096. 30. Chtelain P, Bouillat B, Cohen R, et al. Assay of growth hormone levels in human plasma using commercial kits: analysis of some factors influencing the results. Acta Paediatr Scand 1990; 370(Suppl):56. 31. Shalet SM, Toogood A, Rahim A, Brennan BMD. The diagnosis of growth hormone deficiency in children and adults. Endocr Rev 1998; 19:203. 32. Wu W, Cogan JD, Pfffle RW, et al. Mutations in PROP1 cause familial combined pituitary hormone deficiency. Nature Genet 1998; 18:147. 33. Cogan JD, Wu W, Phillips JA 3rd, et al. The PROP1 2-base pair deletion is a common cause of combined pituitary hormone deficiency. J Clin Endo-crinol Metab 1998; 83:3346. 34. Tatsumi K, Miyai K, Notomi T, et al. Cretinism with combined hormone deficiency caused by a mutation in the PIT1 gene. Nat Genet 1992; 1:56. 35. Wajnrajch MP, Gertner JM, Harbison MD, et al. Nonsense mutation in the human growth hormone-releasing hormone receptor causes growth failure analogous to the little (lit) mouse. Nat Genet 1996; 12:88. 36. Maheshwari HG, Silverman BL, Dupuis J, Baumann G. Phenotype and genetic analysis of a syndrome caused by an inactivating mutation in the growth hormone releasing hormone receptor: dwarfism of Sindh. J Clin Endocrinol Metab 1998; 83:4065. 37. Phillips JA III, Cogan JD. Genetic basis of endocrine disease 6: molecular basis of familial human growth hormone deficiency. J Clin Endocrinol Metab 1994; 78:11. 38. Takahashi Y, Shirono H, Arisaka O, et al. Biologically inactive growth hormone caused by an amino acid substitution. J Clin Invest 1997; 100:1159. 39. Takahashi Y, Kaji H, Okimura Y, et al. Short stature caused by a mutant growth hormone. N Engl J Med 1996; 334:432. 40. Dattani MT, Martinez-Barbera JP, Thomas PQ, et al. Mutations in the homeobox gene HESX1/Hesx1 associated with septo-optic dysplasia in human and mouse. Nat Genet 1998; 19:125. 41. Tauber M, Moulin P, Pienkowski C, et al. Growth hormone (GH) retesting and auxological data in 131 GH-deficient patients after completion of treatment. J Clin Endocrinol Metab 1997; 82:352. 42. Consensus guidelines for the diagnosis and treatment of adults with growth hormone deficiency: summary statement of the Growth Hormone Research Society Workshop on Adult Growth Hormone Deficiency. J Clin Endocrinol Metab 1998; 83:379. 43. Carroll PV, Christ ER, Bengtsson BA, et al. Growth hormone deficiency in adulthood and the effects of growth hormone replacement: a review. J Clin Endocrinol Metab 1998; 83:382.

44. Rosn T, Bengtsson BA. Premature mortality due to cardiovascular disease in hypopituitarism. Lancet 1990; 336:285. 44a. Bulow B, Hagmar L, Eskilsson J, Erfurth EM. Hypopituitary females have a high incidence of cardiovascular morbidity and an increased prevalence of cardiovascular risk factors. J Clin Endocrinal Metab 2000; 85:574. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. Rosenfeld RG, Rosenbloom AL, Guevara-Aguirre J. Growth hormone (GH) insensitivity due to primary GH receptor deficiency. Endocr Rev 1994; 15:369. Ayling RM, Ross R, Towner P, et al. A dominant-negative mutation of the growth hormone receptor causes familial short stature. Nat Genet 1997; 16:13. Parks JS, Brown MR, Faase ME. The spectrum of growth-hormone insensitivity. J Pediatr 1997; 131:S45. Goddard AD, Dowd P, Chernausek S, et al. Partial growth-hormone insensitivity: the role of growth-hormone receptor mutations in idiopathic short stature. J Pediatr 1997; 131:S51. Baumann G, Shaw MA, Merimee TJ. Low levels of high-affinity growth hormonebinding protein in African pygmies. N Engl J Med 1989; 320:1705. Geffner ME, Bersch N, Bailey RC, Golde DW. Insulin-like growth factor I resistance in immortalized T cell lines from African Efe Pygmies. J Clin Endocrinol Metab 1995; 80:3732. Alexander L, Appleton D, Hall R, et al. Epidemiology of acromegaly in the Newcastle region. Clin Endocrinol (Oxf) 1980; 12:71. Faglia G, Arosio M, Bazzoni N. Ectopic acromegaly. Endocrinol Metab Clin North Am 1992; 21:575. Melmed S, Ezrin C, Kovacs K, et al. Acromegaly due to secretion of growth hormone by an ectopic pancreatic isletcell tumor. N Engl J Med 1985; 312:9. Landis CA, Harsh G, Lyons J, et al. Clinical characteristics of acromegalic patients whose pituitary tumors contain mutant G s protein. J Clin Endocrinol Metab 1990; 71:1416. Weinstein LS, Shenker A, Gejman PV, et al. Activating mutations of the stimulatory G protein in the McCune-Albright syndrome. N Engl J Med 1991; 325:1688. Chandrasekharappa SC, Guru SC, Manickam P, et al. Positional cloning of the gene for multiple endocrine neoplasia-type 1. Science 1997; 276:404. Martin JB, Kerber RE, Sherman MB, et al. Cardiac size and function in acromegaly. Circulation 1977; 56:863. Delhougne B, Deneux C, Abs R, et al. The prevalence of colonic polyps in acromegaly: a colonoscopic and pathological study in 103 patients. J Clin Endocrinol Metab 1995; 80:3223. Orme SM, McNally RJ, Cartwright RA, Belchetz PE. Mortality and cancer incidence in acromegaly: a retrospective cohort study. United Kingdom Acromegaly Study Group. J Clin Endocrinol Metab 1998; 83:2730. Minniti G, Moroni C, Jaffrain-Rea ML, et al. Prevalence of hypertension in acromegalic patients: clinical measurement versus 24-hour ambulatory blood pressure monitoring. Clin Endocrinol (Oxf) 1998; 48:149. Daughaday WH, Starkey RH, Saltman S, et al. Characterization of serum growth hormone (GH) and insulin-like growth factor I in active acromegaly with minimal elevation of serum GH. J Clin Endocrinol Metab 1987; 65:617. Davis DH, Laws ER, Ilstrup DM, et al. Results of surgical treatment for growth hormonesecreting pituitary adenomas. J Neurosurg 1993; 79:70. Abosch A, Tyrell JB, Lamborn KR, et al. Transsphenoidal microsurgery for growth hormone-secreting pituitary adenomas: initial outcome and long-term results. J Clin Endocrinol Metab 1998; 83:3411. Swearingen B, Barker FG II, Katznelson L, et al. Long-term mortality after transsphenoidal surgery and adjunctive therapy for acromegaly. J Clin Endocrinol Metab 1998; 83:3419. Newman CB, Melmed S, George A, et al. Octreotide as primary therapy for acromegaly. J Clin Endocrinol Metab 1998; 83:3034. Forbes GB. Nutrition and growth. J Pediatr 1977; 91:40.

CHAPTER 13 PROLACTIN AND ITS DISORDERS Principles and Practice of Endocrinology and Metabolism

CHAPTER 13 PROLACTIN AND ITS DISORDERS

LAURENCE KATZNELSON AND ANNE KLIBANSKI Normal Control of Prolactin Secretion Clinical Aspects of Prolactin Physiology Diurnal and Menstrual Cycle Variation Eating Stress Age Normal States Clinical Manifestations of Hyperprolactinemia Differential Diagnosis and Clinical Approach to Hyperprolactinemia Treatment for Prolactinomas Medical Therapy Surgery Radiation Therapy Management of Microprolactinomas Management of Macroprolactinomas Pregnancy and Prolactinomas Chapter References

NORMAL CONTROL OF PROLACTIN SECRETION

Prolactin secretion is controlled by dual inhibitory and stimula-tory factors (Fig. 13-1). This hormone is unique among anterior pituitary hormones, because it is primarily regulated through tonic inhibition. Two decades of investigation have demonstrated the presence of one or more prolactin-inhibiting factors (PIFs).1

FIGURE 13-1. Regulation of prolactin secretion. Prolactin release is under tonic inhibition by prolactin inhibiting factors, predominantly dopamine. Prolactin release is stimulated by a number of factors, including vasoactive intestinal peptide (VIP), thyroid-releasing hormone (TRH), and gonadotropin-releasing hormone (GnRH). Estrogens, pregnancy, and breast suckling stimulate prolactin release. Within the hypothalamus, serotinergic and dopaminergic pathways are stimulatory and inhibitory, respectively, to prolactin release. (GABA, g-aminobutyric acid.) (Modified from Molitch ME. Pathologic hyperprolactinemia. Endocrinol Metab Clin North Am 1992; 21:877.)

Dopamine is the most important PIF described. Multiple studies support the hypothesis that dopamine acts as an inhibiting factor. In vitro studies reveal that high-affinity dopamine receptors (D2) are present on lactotrope membranes, and after binding occurs, inhibition of adenylate cyclase is demonstrated.2,3 This results in a decrease in cyclic adenosine mono-phosphate (cAMP) production and the release of prolactin. Dopamine also directly inhibits prolactin biosynthesis at the level of RNA transcription. Dopamine is produced in higher nuclei in the brain and is secreted into the portal circulation to reach the pituitary. Infusion of dopamine in humans, resulting in serum dopamine concentrations similar to those found in portal blood, causes a reduction in prolactin secretion.4 Dopa-mine receptor blockade results in prolactin elevation.5 After dopamine is removed, as during pituitary stalk section, prolactin is rapidly released. These studies all point to a direct inhibitory effect of dopamine on pituitary lactotrope secretion. Most pharmacologic agents that cause prolactin release act either by blockade of dopamine receptors (e.g., haloperidol, phenothia-zines) or by dopamine depletion in the tuberoinfundibular neurons (e.g., reserpine, a-methyldopa). Another potential PIF includes a 56-amino-acid PIF identified within the precursor for gonadotropin-releasing hormone (GnRH). This GnRH-associated peptide (GAP) inhibits prolactin secretion and reduces prolactin secretion in rats, but the role of GAP in modulating prolactin secretion in humans remains unconfirmed.6 Data also support the role of g-aminobutyric acid (GABA) as a PIF. GABA is secreted into the portal circulation, with a resulting inhibitory effect on prolactin secretion, and GABA receptors have been detected on lactotropes.7 However, the physiologic importance of GABA in prolactin regulation is unclear. Stimulatory factors also regulate prolactin secretion. These substances may act directly on the pituitary or may act indirectly by means of dopaminergic blockade or depletion at the level of the hypothalamus. Estrogens are important physiologic stimulators of prolactin release.8 In vitro studies show that estradiol increases prolactin biosynthesis, consistent with a direct stimulatory effect of estrogen on lactotrope function.9 Chronic exposure to estrogens increases lactotrope number and size (i.e., pregnancy cells), and acute administration increases prolactin secretion within hours.10 Estrogens may also indirectly increase prolactin levels by altering dopaminergic tone and by increasing responsiveness to other neuromodulators. Thyrotropin-releasing hormone (TRH) stimulates the synthesis and release of prolactin in vivo and in vitro from normal and neoplastic lactotropes. Although pharmacologic doses of TRH result in a rapid release in prolactin after intravenous administration in humans, the physiologic role of TRH in modulating prolactin secretion is not established.11 For example, suckling leads to release of prolactin without an accompanying heightened release of TRH. Hypothyroidism results in an increase in TSH and the prolactin response to TRH, and elevations in basal prolactin levels may be seen in primary hypothyroidism. The suggestion has been made that decreased hypothalamic dopamine may play a role in the hyperprolactinemia associated with hypothyroidism. Vasoactive intestinal peptide (VIP) may selectively stimulate or potentiate the TRH effect on prolactin release.12 Evidence exists for VIP receptors on lactotropes and VIP may stimulate prolactin release in vitro. Immunoneutralization of VIP effects through administration of anti-VIP antisera diminishes the prolactin response to stimuli, including suckling, a finding that again suggests a role of VIP as a stimulatory factor.13 Data suggest that VIP may be produced in the pituitary and may stimulate prolactin release through a paracrine or autocrine mechanism.14 The clinical significance of VIP in prolactin regulation in humans is unknown. GnRH may also have stimulatory properties. The administration of GnRH induces the acute release of prolactin in normally cycling women and hypogonadal patients.15 Moreover, incubation of human lactotropes with GnRH in vitro results in prolactin secretion.16 These investigations suggest that GnRH, directly or through a paracrine mechanism involving gonadotropes, may be important in evoking prolactin release. Investigations have suggested a role for galanin as a potent stimulator of prolactin release. Galanin is a 29-amino-acid peptide widely distributed in the central and peripheral nervous system. In the rat, intracerebroventricular injections of galanin may increase prolactin levels.17 However, intravenous administration of galanin in humans does not raise serum prolactin levels.18 The physiologic role of galanin in human prolactin regulation remains controversial. Serotonin is another factor that may stimulate prolactin secretion.19 Administration of serotonin antagonists decreases prolactin levels. Serotonin agonists appear to enhance prolactin secretion through specific serotonin receptors, perhaps explaining why only specific serotonin antagonists are capable of lowering prolactin secretion.20 Other factors that may have stimulatory roles include bombesin, angiotensin II, histamine (H2) antagonists, and opiates.

Human prolactin structure has partial homology with growth hormone, perhaps accounting for the lactotropic activity of growth hormone. Heterogeneous forms of prolactin are found in the circulation. Eighty-five percent of prolactin (23 kDa) detected in the pituitary and secreted into serum is non-glycosylated, but glycosylated forms have been detected.21 Approximately 8% of prolactin extractable from the pituitary is dimeric, and an additional 1% to 5% is polymeric, linked by disulfide bonds.22 These forms include big, big-big, and little or native prolactins. The significance of these forms is unknown. The larger forms of prolactin may have decreased rates of binding to the prolactin receptor and possess diminished bioactivity relative to monomeric, nonglycosylated prolactin. These forms may represent nonspecific hormonal aggregates or binding of prolactin to serum proteins. Some patients have normal reproductive function but elevated serum prolactin values; the prolactin in these patients is composed of a relatively increased component of polymeric prolactin.23 In such patients, the elevated prolactin levels may reflect increased levels of polymeric prolactin with decreased bioactivity. The suggestion has been made that certain isotypes of prolactin, specifically iso-B prolactin, may be elevated in the sera of patients with infertility and pregnancy wastage.24 This isotype may be more resistant to bromocriptine therapy than native prolactin. However, these reports need to be confirmed. The significance of the remaining fraction of glycosylated prolactin is unknown.

CLINICAL ASPECTS OF PROLACTIN PHYSIOLOGY

The physiologic causes of hyperprolactinemia are summarized in Table 13-1. The following sections describe clinical aspects of prolactin physiology.

TABLE 13-1. Physiologic Causes of Hyperprolactinemia

DIURNAL AND MENSTRUAL CYCLE VARIATION Prolactin is secreted in a pulsatile fashion with 4 to 14 pulses per day (60% occur during sleep).25 Prolactin secretory pulses begin 60 to 90 minutes after the onset of sleep.26 The amplitude of pulses varies greatly among individuals, with peak levels occurring during the late hours of sleep. Such rises are not clearly associated with any specific stage of sleep. Although some studies have suggested that prolactin varies during the menstrual cycle, the precise nature of this relationship remains unclear. Several investigators have shown that prolactin levels are significantly higher during the ovulatory and luteal phases, particularly at midcycle.27 This midcycle rise may be the result of increased circulating periovulatory estradiol levels. However, other studies have not confirmed this finding. Prolactin is probably not necessary for ovulation, because ovulatory periods may occur in women taking bromocriptine, a medication that suppresses prolactin. EATING Abrupt rises in serum prolactin levels occur within an hour of eating in normal and pregnant hyperprolactinemic individuals, but they do not rise in those with prolactinomas. Amino acids metabolized from protein components of meals appear to be the main stimulants to prolactin secretion.28 STRESS Prolactin rises during stress, including physical exertion, surgery, sexual intercourse, insulin hypoglycemia, and seizures. The nature and teleologic significance of these changes is unknown. In women, nipple stimulation, chest wall trauma or surgery, and herpes zoster infection of the breast may result in increases in prolactin levels, in part through afferent neural pathways.29 In contrast, nipple stimulation in men does not cause increased prolactin levels. AGE Mean levels of prolactin are slightly higher in premenopausal women than in men, probably because of a direct effect of estrogen on pituitary prolactin secretion or estrogen-induced alterations in dopaminergic tone.30 Some studies suggest that prolactin levels progressively decline in women with age, particularly after the menopause.31 The responsiveness of prolactin to various pharmacologic agents (e.g., TRH) declines with age in women, probably because of postmenopausal estrogen deficiency. NORMAL STATES The only established role of prolactin is to initiate and maintain lactation. Prolactin levels increase progressively with pregnancy. Estrogens play a major role in stimulation of prolactin levels, which peak at term (at 100300 ng/mL).32 Lactation begins when estradiol levels fall at parturition. During the first 4 to 6 weeks after delivery, prolactin levels increase in the circulation to 60 times higher than baseline levels within 20 to 30 minutes of nursing.33 This elevation is associated with enhanced prolactin pulse amplitude without alteration of pulse frequency.34 The nursing stimulus effectively promotes acute prolactin release through afferent spinal neural pathways. With continued nursing, the nipple stimulation itself elicits progressively less prolactin release, and in the weeks after initiation of lactation, basal and nursing-induced prolactin pulses decrease, although lactation continues.33 Within 4 to 6 months after delivery, basal pro-lactin levels are normal, without a nursing-induced rise. The full explanation for the attenuation in basal and nursing-induced prolactin levels with continued nursing is unknown.

CLINICAL MANIFESTATIONS OF HYPERPROLACTINEMIA

The amenorrhea-galactorrhea syndrome is the classic description of the clinical manifestation of hyperprolactinemia. However, a spectrum of reproductive disorders may be seen. Prolactin elevations are found in approximately 20% of patients with secondary amenorrhea.35 Women with hyperprolactinemia may have more subtle abnormalities in gonadal function, including oligomenorrhea or alterations in luteal phase function. A subset of infertile women has been described with mild hyperprolactinemia in whom fertility was restored with bromocriptine therapy. Galactorrhea affects only ~30% of female patients with hyperprolactinemia, but the presence of galactor-rhea in a woman with an ovulatory disorder greatly increases the chance that hyperprolactinemia is the underlying cause of the amenorrhea.36 Patients with primary amenorrhea and delayed puberty may have hyperprolactinemia.37 Galactorrhea occurs in as many as 25% of women with normal serum prolactin levels. However, patients with idiopathic galactorrhea may demonstrate intermittent hyperprolactinemia. In a study of nine normoprolactinemic women with galactorrhea, eight patients had elevated levels of prolactin during sleep.38 Several studies have shown that infertile, normopro-lactinemic women with luteal phase defects may show improved luteal function or fertility after administration of bromocriptine therapy.38 Unrecognized hyperprolactinemia may occur in a subset of patients with presumed normoprolactinemic galactorrhea and luteal phase defects. Hypogonadism is frequently found in patients with hyperprolactinemia. In women, hypogonadism includes abnormal menstrual function, dry vaginal mucosa, and dyspareunia, and in men and women, the features include fatigue and diminished libido. Multiple potential mechanisms have been hypothesized for the induction of hypogonadism by prolactin, and the antigo-nadotropic actions of prolactin may occur at multiple levels. Frequently, the hypogonadism is associated with decreased or inappropriately normal levels of luteinizing hormone (LH) and follicle-stimulating hormone (FSH) relative to the state of estrogen deficiency. Multiple investigations suggest that prolactin may suppress spontaneous LH release through decreases in endogenous GnRH levels. In castrated rats, administration of graded doses of prolactin suppress LH levels, and prolactin appears to exert a negative feedback effect on its own secretion by means of a short-loop negative feedback at the level of the hypothalamus.39,40 This feedback may be mediated through an increase in dopamine inhibitory tone. This increased hypothalamic dopamine tone, along with opiates and other factors, may suppress GnRH with a resultant decrease in LH pulses. The restoration of ovulatory menstrual periods in hyperprolactinemic women by pulsatile exogenous GnRH administration confirms the importance of endogenous GnRH abnormalities as the key mechanism of hypogonadism in these women.41 Prolactin may modulate androgen secretion at the level of the adrenal gland and ovary, resulting in increased secretion of dehydroepiandrosterone sulfate and

testosterone.42 Altered ratios of estrogens and androgens may result in further abnormal gonadal function, with evidence of clinical hyperandrogenism. If the underlying cause of the increased prolactin is a pituitary macroadenoma, the adenoma could cause compression of the normal, adjacent pituitary gland with a resultant decrease in gonadotrope function. Men with hyperprolactinemia may have clinical manifestations of hypogonadism, such as decreased libido, impotence, infertility due to oligospermia, and gynecomastia. Galactorrhea is rare in hyperprolactinemic men because of a lack of estrogen priming of the breast.

DIFFERENTIAL DIAGNOSIS AND CLINICAL APPROACH TO HYPERPROLACTINEMIA

As shown in Table 13-1 and Table 13-2, hyperprolactinemia has multiple causes. The measurement of prolactin level should be repeated with the patient in a nonstimulated state, and, if possible, after an overnight fast in a nonstressed state. Because prolactin may be secreted to a modest degree after a breast examination, a subsequent mild increase in prolactin levels would warrant a repeat determination. Although Table 13-2 demonstrates the existence of several pathologic causes of prolactin elevation, pituitary tumors are clinically the most important. Prolactin-secreting pituitary adenomas are the most common type of pituitary tumors and may account for as many as 40% to 50% of all pituitary tumors.43 Hyperprolactinemia may be detected in as many as 40% of patients with acromegaly and has been reported in patients with Cushing disease. Hyper-prolactinemic patients with suggestive clinical manifestations should be evaluated for acromegaly and Cushing disease.

TABLE 13-2. Pathologic and Pharmacologic Causes of Hyperprolactinemia

Substantial elevation in prolactin (>150 ng/mL) in a nonpuerperal state usually indicates a pituitary tumor. Good correlation exists between radiographic estimates of tumor size and prolactin levels, and very high levels of prolactin are associated with larger tumors. Prolactinomas are classified as microade-nomas (<10 mm) and macroadenomas (>10 mm). The finding of a substantial elevation in serum prolactin in association with a pituitary lesion larger than 10 mm by radiographic analysis supports the diagnosis of a macroprolactinoma. Modest levels of prolactin elevation (25100 ng/mL) may be associated with several diagnoses. All other causes of hyper-prolactinemia should be excluded before a tumor is considered. Primary hypothyroidism and pregnancy should be excluded. Chronic renal disease is associated with elevations in prolactin, probably because of altered metabolism or clearance of prolac-tin or decreases in dopaminergic tone.44 Hemodialysis does not usually reverse the hyperprolactinemia. Multiple pharmacologic causes of hyperprolactinemia are found. Ingestion of phenothiazines and other neuroleptics is a common cause for elevations in serum prolactin.44a One diagnostic problem is the evaluation of patients with psychiatric disease who are receiving phenothiazines and are found to have an elevated prolactin level. A magnetic resonance image (MRI) should be obtained for patients whose prolactin levels are above 100 ng/mL. Levels lower than 100 ng/mL are consistent with neuroleptic administration, and a scan is unnecessary unless other symptoms suggest a pituitary tumor. This strategy is based on the finding that most patients receiving neuroleptics with modest prolactin elevations have no evidence of a pituitary abnormality on MRI. Other pharmacologic agents associated with hyperprolactinemia include reserpine, a-methyldopa, cimetidine, and opiates. Estrogen may increase prolactin levels, as is seen in pregnancy. However, the estrogen concentrations in typical oral contraceptives (e.g., 35 g of ethinyl estradiol) are not associated with hyperprolactinemia, and no evidence exists that post-menopausal replacement estrogen causes elevations in serum prolactin. Any intrasuprasellar mass may lead to modest prolactin elevations through stalk compression, and the evaluation should include an MRI. These masses include primary pituitary tumors or meningiomas and craniopharyngiomas. Hypothalamic disorders, including destructive lesions such as tumors and granulomatous diseases, may lead to hyperprolactinemia by interfering with normal dopaminergic tone. If an elevated serum prolactin level is not associated with primary hypothyroidism, pregnancy, or pharmacologic agents, a pituitary radiographic scan should be performed to rule out the presence of a prolactin-secreting pituitary tumor or other lesions. Microprolactinomas should be differentiated from macroprolac-tinomas. An MRI is the most sensitive tool for evaluating the sellar and suprasellar areas. If the scan shows normal sellar and extrasellar contents and no clear secondary cause of the elevated prolactin is present, the diagnosis of idiopathic hyperprolactine-mia is made. This syndrome may be the result of a small tumor that is beyond the sensitivity of the scanning technique. The evaluation should include assessment of gonadal status, such as the presence of oligomenorrhea or amenorrhea in women and of sexual dysfunction in men. This impacts therapy, as is described later. No stimulatory or suppressive endocrine tests are available that aid in the evaluation of elevated prolactin levels. For example, a TRH test cannot be used to diagnose a pituitary tumor; although tumors typically have blunted responses after TRH stimulation, this response can be seen with other disorders.

TREATMENT FOR PROLACTINOMAS

Treatment depends on whether the patient has hyperpro-lactinemia due to an underlying cause such as drugs or hypothyroidism, or due to a prolactinoma. If the evaluation suggests the presence of a microprolactinoma, three treatment options are available: medical therapy with a dopamine agonist, careful follow-up without treatment, and, rarely, surgery. All patients with macroadenomas should be treated. MEDICAL THERAPY Almost all patients with hyperprolactinemia due to pituitary disease can be effectively treated medically with the dopamine agonist bromocriptine (see Chap. 21). Bromocriptine lowers serum prolactin in patients with pituitary tumors and all other causes of hyperprolactinemia. A review of early studies of bromocriptine therapy for more than 400 hyperprolactinemic patients showed that normoprolactinemia or return of ovulatory menses occurred in 80% to 90% of patients.45 Bromocriptine effectively decreases prolactin levels, normalizes reproductive function, and reverses galactorrhea. In this series, return of menstrual function was accompanied in some patients by prolactin levels that were significantly reduced but not normal. This suggests that the reduction of prolactin levels in some patients to slightly elevated levels may be sufficient for return of gonadal function. Bromocriptine is also useful in treating galactorrhea in patients with normoprolactinemic galactorrhea. The onset of the effects of bromocriptine is rapid, usually occurring within 1 to 2 hours. The greatest decrease in prolactin levels occurs at the initiation of therapy; however, normalization may take weeks. The biologic half-life of bromocriptine is similar to its plasma half-life. Discontinuation of the drug is typically followed by a return of prolactin to elevated values. Bromocriptine decreases prolactin production and secretion, with a resultant reduction in lactotrope size and a subsequent decrease in tumor size.46 Therapy should be initiated slowly because side effects, including nausea, headache, dizziness, nasal congestion, and constipation, may occur. Gastrointestinal side effects may be minimized by starting with a very low dose (e.g., 1.25 mg, or one-half tablet) taken at night with a snack, and increasing the dose by 1.25 mg over 4- to 5-day intervals, as tolerated. This progression is continued until a dosage that normalizes prolac-tin levels is reached. The rate of dosage escalation is dictated by the clinical situation, such as the presence of mass effects. Side effects can usually be improved by continuing the medication at the same dosage or by temporarily reducing the dosage. If patients stop taking the medication for a few days, therapy should be reinstituted at a lower dosage, because side effects may return. Rarely, long-term therapy may result in side effects, including painless cold-induced digital vasospasm, alcohol intolerance, dyskinesia, and psychiatric reactions, including

fatigue, depression, and anxiety. To reduce the gastrointestinal side effects, bromocriptine has been administered intravaginally. Reductions in prolactin similar to that attained by oral bromocriptine have been achieved with the intravaginal route.47 Gastrointestinal side effects are less common, and therapy may be more effective with vaginally administered bromocriptine.48 Cabergoline is an ergoline derivative with selective, potent, and long-lasting dopaminergic properties and is highly effective in the management of hyperprolactinemia. Because of the ease of administration of cabergoline (i.e., once or twice weekly) and its improved side-effect profile relative to bromocriptine, patients show a high rate of compliance. Administration of cabergoline at doses as high as 1.0 mg twice weekly to 113 patients with micro-prolactinomas resulted in normalization of prolactin levels in 95%.49 In another study, administration of cabergoline resulted in normalization of prolactin levels in 25 of 26 patients with micro-prolactinomas.50 Cabergoline appears to be better tolerated than bromocriptine and may play an important role in the management of patients intolerant of or resistant to bromocriptine. In a multicenter, randomized, 24-week trial involving 459 women, cabergoline was more effective and better tolerated than bromocriptine.51 In a study of 27 patients with prolactinomas resistant to bromocriptine or the investigational dopamine agonist CV 205-502, including 19 subjects with macroadenomas, caber-goline administration resulted in normalization of prolactin values in 47% of patients with macroadenomas and in all patients with microadenomas.52 Therefore, use of cabergoline may be considered in all subjects with hyperprolactinemia, including those who are poorly tolerant of or resistant to bromocriptine.52a Other dopamine agonists are available, but not in the United States; these include CV 205-502 and a long-acting preparation of bromocriptine mesylate, Parlodel LAR. CV 205-502 is a non-ergot, long-acting dopamine agonist that appears to have D2-receptor binding compared with bromocriptine. Either CV 205-502 or bromocriptine was administered in a randomized double-blind fashion to 22 patients with microprolac-tinomas.53 In this study, 91% of the patients who received CV 205-502 and 56% of those who received bromocriptine had normalization of prolactin levels. Side effects were less common in the group receiving CV 205-502, and the drug may be useful in those patients intolerant to bromocriptine.54 CV 205-502 also is effective in the management of macroprolactinomas.55 Pergolide is a dopamine agonist approved by the U.S. Food and Drug Administration for the treatment of Parkinson disease. Although not approved for use in the management of hyperprolactinemia, studies have shown that pergolide has a comparable side-effect profile to bromocriptine and may reduce prolactin levels in patients unresponsive to bromocriptine.56,56b With the availability of cabergoline, pergolide is rarely used as an alternative to bromocriptine. SURGERY Although surgery is not a primary mode of management for patients with prolactinomas, it may be indicated in several settings. These include patients with large tumors causing visual field deficits unresponsive to bromocriptine, those unable to tolerate dopamine agonist therapy because of its side effects, those with cystic tumors that do not respond to medical therapy, and those with tumor apoplexy. A transsphenoidal approach is used almost exclusively. When the procedure is performed by experienced surgeons, the morbidity rate is negligible. The mortality rate is less than 0.27%, and the major morbidity rate is 3%.57 A theoretic advantage of curative surgery is avoidance of long-term medication. However, clinical evidence is lacking. Among 28 patients with microprolactinomas, 24 were cured with transsphenoidal surgery based on normalization of serum prolactin levels. After approximately 4 years, 50% of these initially cured patients had recurrence of hyperpro-lactinemia, although none had radiographic evidence of tumor growth.58 Another study found a recurrence rate of 39% after approximately 5 years.59 In one study of patients with normal prolactin values after surgery, the overall recurrence rate was 26% at a mean follow-up of 9.2 years.60 An immediately postoperative serum prolactin value of <5 ng/mL was associated with a recurrence rate of approximately 20%. These data suggest that, although surgery may result in normalization of prolactin levels initially in patients with microprolactinomas, risk of recurrence is relatively high. In addition, achievement of a low-normal serum prolactin level is an important predictive factor for long-term cure. Surgical cure rates for macroprolactinomas are approximately 32%, with cure defined as normal prolactin levels after surgery.45 Surgical cure is inversely proportional to serum pro-lactin levels and tumor size. Unfortunately, the recurrence rate in macroprolactinomas has been reported to be as high as 80% after curative surgery,58 although another report indicates that recurrence rates as low as 26% can be achieved.60 RADIATION THERAPY Conventional radiotherapy (45005000 rad) or, rarely, proton beam therapy may be indicated in patients who are not able to tolerate medical therapy.

MANAGEMENT OF MICROPROLACTINOMAS
The decision to institute medical therapy in patients with microprolactinomas is based on the metabolic consequences of hyperprolactinemia and tumor size. Patients with hyperpro-lactinemia are usually hypogonadotropic and have accompanying menstrual irregularities. Dopamine agonist therapy can restore menstrual function in most patients with amenorrhea. Luteal phase defects associated with hyperprolactinemia can also be reversed with bromocriptine therapy. Ovulation rates greater than 90% have been reported, with induction of pregnancy in more than 80% of patients.61 Galactorrhea is not an absolute indication for dopamine agonist therapy, unless the degree of galactorrhea is significantly bothersome to the patient. The presence of amenorrhea is an indication for medical therapy because of the risk of osteoporosis associated with hyperprolactinemic amenorrhea. Some women with micropro-lactinomas may choose not to have therapy if they show no evidence of hypogonadism or amenorrhea and if they do not desire fertility. Reduction of prolactin levels frequently restores libido and increases sperm counts in hyperprolactinemic men. Patients with microprolactinomas and those without radiographic evidence of pituitary tumors can sometimes be followed without therapy. Studies investigating the natural history of such tumors have shown that prolactin elevations usually remain stable and, in some cases, spontaneously normalize.62 In a study of 41 patients with idiopathic hyperpro-lactinemia, based on normal computed tomographic scans for 5.5 years, 67% of patients whose initial prolactin values were less than 57 ng/mL had normalization of prolactin levels.63 None of the patients with initial prolactin values above 60 ng/mL showed normalization. These and other data suggest that the degree of prolactin elevation is a prognostic factor for spontaneous resolution. When 38 untreated patients with microprolactinomas were followed for an average of 50.5 months, 36.6% had an increase, 55.3% had a spontaneous decrease, and 13.1% had no change in prolactin levels.64 A prospective study of untreated hyperpro-lactinemic women showed that basal menstrual function is an important variable in predicting progression of the prolactin level.65 In this study, patients with normal initial menstrual function were more likely to have normalization of prolactin levels, and patients with oligomenorrhea or amenorrhea were more likely to have no change or increases in prolactin levels. Most microprolactinomas do not exhibit evidence of further growth, and prolactin levels may spontaneously normalize. An important aspect of the natural history of microprolacti-nomas is that most tumors do not significantly increase in size. Although many of these studies used insensitive radiographic techniques, such as skull films and tomograms, they demonstrated that in patients with microprolactinomas and no radiographic evidence of a tumor, tumor size increased in 0% to 22% of patients.61,63,64,65 and 66 In a study of 43 patients with presumed microadenomas with a mean follow-up period of 5.4 years, only 2 patients showed evidence of tumor progression.66 In a prospective study, 27 women were followed for an average of 5.2 years.65 Of 14 women with normal baseline radiographic studies, 4 developed evidence of an adenoma, although none developed a macroadenoma. Of the 13 women with evidence of a tumor at baseline, only 2 showed worsening of radiographic findings. This study suggests that, although tumor growth may occur in as many as 22% of cases, it is rarely accompanied by clinical symptoms from mass effects. Follow-up of untreated patients should include serial measurement of prolactin levels and periodic MRI scans, because tumor progression may not be accompanied by an increase in prolactin levels. The presence of osteoporosis is a key factor in the decision to institute therapy. Hyperprolactinemia is associated with trabecular and cortical osteopenia. Fourteen young hyperprolactinemic women with prolactin levels ranging from 22 to 99 ng/mL and amenorrhea for 1 to 18 years had significantly decreased cortical bone density compared with normal women.67 Additional studies have shown that hyperprolactinemic women may have trabecular osteopenia with spinal bone density 10% to 25% below normal.68,69 and 70 Spinal bone density in hyperpro-lactinemic women correlates with serum androgen levels and relative percentage of ideal body weight.68,71 and 72 Decreased bone density is thought to be due to hypogonadism and not a direct effect of prolactin, because hyperprolactinemic women with normal menstrual function do not have associated bone loss.68 Figure 13-2 shows that hyperprolactinemic patients with hypogonadism have lower bone density than eugonadal hyper-prolactinemic women. The decreased bone density in the hyperprolactinemic women with amenorrhea also correlates with the duration of amenorrhea.70

FIGURE 13-2. Spinal bone density in 13 women with hyperprolactine-mic amenorrhea (top), 12 eumenorrheic hyperprolactinemic women (middle), and 11 women with hypothalamic amenorrhea (bottom). The mean 1 standard deviation for 19 normal women is shown by the solid and dashed horizontal lines, respectively. (Reprinted from Klibanski A, Biller BM, Rosenthal DI, et al. Effects of prolactin and estrogen deficiency in amenorrheic bone loss. J Clin Endocrinol Metab 1988; 67:124.)

An important question is whether treatment of hyperprolac-tinemic amenorrhea may result in improvement in bone density. A prospective series in which 32 women with hyperprolactine-mic amenorrhea were randomized to medical therapy or no therapy showed that resumption of menses with medical therapy was associated with a significant increase in cortical bone density, mostly during the first 12 months of therapy.70 However, the cortical bone density achieved in this series remained below that of normal women. Of 38 women with prolactinomas examined 2 to 5 years after surgery, cortical bone mass was below normal in both cured and uncured patients, a finding which suggests that remission of hyperprolactinemia may not lead to normalization of bone density.73 Therapy for hyperprolactinemic amenorrhea with resumption of menses may lead to improvement in but not normalization of bone density. Another important question is whether hyperprolactinemic amenorrhea is associated with progressive, accelerated bone loss. In a study of 52 hyperprolactinemic women and 41 controls over a mean period of 1.8 years, trabecular osteoporosis was marked in women with hyperprolactinemic amenorrhea, and bone loss progressed in untreated women by an average of 3.8% per year.71 Conversely, in another study of 56 hyperpro-lactinemic women, spinal bone density in women with amenorrhea did not decrease significantly over an average of 4.7 years.72 These studies suggest that a subgroup of women with hyperprolactinemic amenorrhea may be at risk for progressive osteoporosis, including those women with decreased percentages of ideal body weight and lower serum androgen levels. Because osteoporosis is common in women with hyperpro-lactinemic amenorrhea, the authors recommend treatment of amenorrheic women to improve bone density or at least prevent further deterioration.

MANAGEMENT OF MACROPROLACTINOMAS
Unlike patients with microprolactinomas, those with macropro-lactinomas always require therapy. Patients with macroade-nomas may have evidence of local mass effects due to the tumor mass, with resultant visual field abnormalities, and hypopitu-itarism due to compression of the normal pituitary gland. Aggressive management is required to prevent or reverse these complications. Bromocriptine therapy results in significant tumor shrinkage in as many as 75% of patients. Tumor size reduction may occur in weeks or over many months, and this reduction is frequently accompanied by an improvement in visual field abnormalities and pituitary function. Visual field deficits have been reported to improve within hours of institution of medical therapy. Of 27 patients treated with bromocriptine, 64% experienced a reduction in tumor size of at least 50%. 74 Tumor shrinkage often occurred within 6 weeks. Sixty-six percent of patients had normalization of prolactin levels, but the fall in prolactin levels did not always correlate with reductions in tumor size. Administration of cabergoline may lead to reduction in pro-lactin levels, tumor shrinkage, and restoration of gonadal function in patients with macroprolactinomas. When cabergoline was administered to 14 patients with macroprolactinomas, tumor shrinkage was observed in 13 (93%), and complete disappearance was documented in 2 patients.75 In one such study, cabergoline was administered to 15 patients with macroprolacti-nomas in a multicenter, 48-week trial (see Fig. 13-3).76 Normalization of prolactin levels was achieved in 73% of subjects, and mean reduction in tumor size was 31%. In this study, tumor shrinkage often occurred during the first 8 weeks; however, in some subjects, reductions in tumor volume were not seen until 24 weeks. Therefore, cabergoline may have an important role in the management of patients with macroprolactinomas.77 Dosages of 0.5 to 3.0 mg cabergoline per week are often sufficient for maximal effects. The drug may be useful as first line therapy in patients with macroprolactinomas or in those who are intolerant of or resistant to bromocriptine.52

FIGURE 13-3. Individual serum prolactin levels at baseline and 48 weeks according to final dose in macroprolactinoma patients treated with cabergoline. The dotted horizontal line indicates normal serum pro-lactin (<20ng/mL). The dotted vertical line refers to a patient who normalized PRL at week 6 but did not complete 48 weeks because of a complication (pituitary hemorrhage). (From Biller BM, Molitch ME, Vance ML, et al. Treatment of prolactin-secreting macroadenomas with the once-weekly dopamine agonist cabergoline. J Clin Endocrinol Metab 1996; 81:2338, with permission.)

The heterogeneity of prolactinoma responses to bromocriptine may be the result of variable density of the D2 receptor on tumor membranes. Good correlation exists between the density of dopaminergic binding sites and maximal inhibition of adenylate cyclase activation in bromocrip-tine-responsive and bromocriptine-resistant prolactinomas.78 This suggests that bromocriptine response depends on the density of the D2 receptor. A study of 46 prolactinomas in patients undergoing surgery because of failure to respond to bromocriptine or local compression showed no evidence of mutations in the D2 coding sequence.79 Alterations in density or function probably explain the lack of dopamine agonist effect in these patients. Medical therapy represents the initial option for patients with macroprolactinomas with no or stable visual field deficits because of the low cure rates associated with surgery in such patients. Although transsphenoidal surgery is not used as a primary therapy for patients with macroprolactinomas, no data are available directly comparing the use of dopamine agonists with surgery as primary therapy for macroadenomas. Specific circumstances, such as the presence of a cystic prolactinoma (which often does not respond fully to dopamine agonist therapy), may predict a poor initial response to bromocriptine therapy in these patients. Cabergoline or bromocriptine is a useful adjunctive therapy in patients with large tumors, for which complete resection has not been possible. Most men with diagnosed prolactinomas have macroade-nomas. In women, most tumors are microadenomas. In part, this difference may reflect the fact that women present earlier than men for evaluation because of complaints of menstrual disturbances. In men with hyperprolactinemia-induced hypog-onadism and normal residual pituitary function, 3 to 6 months may be required for testosterone levels to increase, normal sexual function to be restored, and sperm counts and motility to increase after prolactin levels normalize.80

PREGNANCY AND PROLACTINOMAS

Many women with hyperprolactinemia present with infertility; bromocriptine is typically used to normalize prolactin levels and allow normal ovulation to occur. After pregnancy is established, bromocriptine should be discontinued if no evidence of local tumor compression is seen. Pregnancy in normal women leads to increased pituitary size through estrogen-stimulated lactotrope hyperplasia. For patients with prolactinomas, the concern is that the high estrogen levels associated with pregnancy may lead to lactotrope stimulation and tumor growth, with resulting local complications, including visual field deficits, headaches, and diabetes insipidus. In a

review of the data on pregnancy outcomes in hyperprolactinemic patients, clinically significant tumor enlargement (e.g., causing headaches, visual deficits) were found in as many as 5.5% of patients with microprolactinomas.81 Patients with microprolactinomas should be followed carefully during pregnancy, with visual field monitoring at monthly intervals. In contrast to patients with microprolactinomas, 15.5% to 35.7% of patients with macroadenomas are at risk for clinically significant tumor enlargement during any trimester of pregnancy. Although some centers recommend transsphenoidal resection of the macroadenomas before conception, surgical resection does not prevent symptomatic enlargement during pregnancy. Monitoring of serum prolactin levels throughout pregnancy is not clinically useful, because prolactin levels increase markedly during pregnancy. The decision to reinstitute therapy depends on the development of clinical symptoms, not the serum prolactin level. If a complication due to tumor growth does occur, it is rapidly reversible with the reinstitution of bromocriptine therapy, which is then continued through term. The outcome of bromocriptine-induced pregnancies is comparable to that of normal pregnancies. A large, international experience with bromocriptine and pregnancy suggests that bromocriptine therapy does not result in complications for the fetus82 (see Chap. 110). Experience with cabergoline and pregnancy is much more limited. In the largest such study (226 pregnancies induced by cabergoline) no evidence was found of increased pregnancy- associated complications or birth defects in these women compared to women who did not receive cabergoline.83 However, until the experience with cabergoline is more widespread, bromocriptine is recommended as a first-choice therapy for hyper-prolactinemic women seeking pregnancy. After delivery, breast-feeding appears to be safe in patients not receiving bromocriptine. Patients with macroadenomas should continue to be followed closely, and the decision to institute therapy should depend on tumor size and clinical symptoms. Patients with hyperprolactinemia should undergo a complete evaluation to determine the underlying cause of the elevated prolactin level. If idiopathic hyperprolactinemia or a prolactinoma is the cause, dopamine agonist therapy is highly efficacious in lowering prolactin levels, restoring gonadal function, and reversing local tumor complications. CHAPTER REFERENCES
1. Blackwell RE. Hyperprolactinemia. Evaluation and management. Endocrinol Metab Clin North Am 1992; 21:105. 2. Foord SM, Peters JR, Dieguez C, et al. Dopamine receptors on intact anterior pituitary cells in culture: functional association with the inhibition of prolactin and thyrotropin. Endocrinology 1983; 112:1567. 3. Schettini G, Cronin MJ, MacLeod RM. Adenosine 3',5'-monophosphate (cAMP) and calcium-calmodulin interrelation in the control of prolactin secretion: evidence for dopamine inhibition of cAMP accumulation and prolactin release after calcium mobilization. Endocrinology 1983; 112:1801. 4. Leblanc H, Lachelin GC, Abu-Fadil S, Yen SS. Effects of dopamine infusion on pituitary hormone secretion in humans. J Clin Endocrinol Metab 1976; 43:668. 5. Rivera JL, Lal S, Ettigi P, et al. Effect of acute and chronic neuroleptic therapy on serum prolactin levels in men and women of different age groups. Clin Endocrinol 1976; 5:273. 6. Nikolics K, Mason AJ, Szonyi E, et al. A prolactin-inhibiting factor within the precursor for human gonadotropin-releasing hormone. Nature 1985; 316:511. 7. Grossman A, Delitala G, Yeo T, Besser GM. GABA and muscimol inhibit the release of prolactin from dispersed rat anterior pituitary cells. Neuroen-docrinology 1981; 32:145. 8. Raymond V, Beaulieu M, Labrie F, Boissier J. Potent antidopaminergic activity of estradiol at the pituitary level on prolactin release. Science 1978; 200:1173. 9. Maurer RA. Estradiol regulates the transcription of the prolactin gene. J Biol Chem 1982; 257:2133. 10. Scheithauer BW, Sano T, Kovacs KT, et al. The pituitary gland in pregnancy: a clinicopathologic and immunohistochemical study of 69 cases. Mayo Clin Proc 1990; 65:461. 11. Bowers CY, Friesen HG, Hwang P, et al. Prolactin and thyrotropin release in man by synthetic pyroglutamyl-histidyl-prolinamide. Biochem Biophys Res Commun 1971; 45:1033. 12. Kato Y, Iwasaki Y, Iwasaki J, et al. Prolactin release by vasoactive intestinal polypeptide in rats. Endocrinology 1978; 103:554. 13. Abe H, Engler D, Molitch ME, et al. Vasoactive intestinal peptide is a physiological mediator of prolactin release in the rat. Endocrinology 1985; 116:1383. 14. Hagen TC, Arnaout MA, Scherzer WJ, et al. Antisera to vasoactive intestinal polypeptide inhibit basal prolactin release from dispersed anterior pituitary cells. Neuroendocrinology 1986; 43:641. 15. Casper RF, Yen SS. Simultaneous pulsatile release of prolactin and luteiniz-ing hormone induced by luteinizing hormonereleasing factor agonist. J Clin Endocrinol Metab 1981; 52:934. 16. Denef C, Andries M. Evidence for paracrine interaction between gonado-trophs and lactotrophs in pituitary cell aggregates. Endocrinology 1983; 112:813. 17. Koshiyama H, Kato Y, Inoue T, et al. Central galanin stimulates pituitary prolactin secretion in rats: possible involvement of hypothalamic vasoactive intestinal polypeptide. Neurosci Lett 1987; 75:49. 18. Giustina A, Licini M, Schettino M, et al. Physiological role of galanin in the regulation of anterior pituitary function in humans. Am J Physiol 1994; 266:E57. 19. Wehrenberg WB, McNicol D, Frantz AG, Ferin M. The effects of serotonin on prolactin and growth hormone concentrations in normal and pituitary stalk-sectioned monkeys. Endocrinology 1980; 107:1747. 20. Falaschi P, Rosa M, Rocco A, et al. Effect of ritanserin, specific 5HT-2 antagonist, on PRL secretion in normal subjects and in different hyperprolactinaemic conditions. Clin Endocrinol 1991; 34:449. 21. Lewis UJ, Singh RN, Sinha YN, Vanderlaan WP. Glycosylated human prolactin. Endocrinology 1985; 116:359. 22. Whitaker MD, Klee GG, Kao PC, et al. Demonstration of biological activity of prolactin molecular weight variants in human sera. J Clin Endocrinol Metab 1984; 58:826. 23. Corenblum B. Asymptomatic hyperprolactinemia resulting from macro-prolactinemia. Fertil Steril 1990; 53:165. 24. Ben-David M, Schenker JG. Transient hyperprolactinemia: a correctable cause of idiopathic female infertility. J Clin Endocrinol Metab 1983; 57:442. 25. Veldhuis JD, Johnson ML. Operating characteristics of the hypothalamo-pituitary-gonadal axis in men: circadian, ultradian, and pulsatile release of prolactin and its temporal coupling with luteinizing hormone. J Clin Endocrinol Metab 1988; 67:116. 26. Sassin JF, Frantz AG, Weitzman ED, Kapen S. Human prolactin: 24-hour pattern with increased release during sleep. Science 1972; 177:1205. 27. Franchimont P, Dourcy C, Legros JJ, et al. Prolactin levels during the menstrual cycle. Clin Endocrinol 1976; 5:643. 28. Carlson HE. Prolactin stimulation by protein is mediated by amino acids in humans. J Clin Endocrinol Metab 1989; 69:7. 29. Boyd AED, Spare S, Bower B, Reichlin S. Neurogenic galactorrhea-amenorrhea. J Clin Endocrinol Metab 1978; 47:1374. 30. Katznelson L, Riskind PN, Saxe VC, Klibanski A. Prolactin pulsatile characteristics in postmenopausal women. J Clin Endocrinol Metab 1998; 83:761. 31. Vekemans M, Robyn C. Influence of age on serum prolactin levels in women and men. BMJ 1975; 4:738. 32. Rigg LA, Lein A, Yen SS. Pattern of increase in circulating prolactin levels during human gestation. Am J Obstet Gynecol 1977; 129:454. 33. Noel GL, Suh HK, Frantz AG. Prolactin release during nursing and breast stimulation in postpartum and nonpostpartum subjects. J Clin Endocrinol Metab 1974; 38:413. 34. Nunley WC, Urban RJ, Kitchin JD, et al. Dynamics of pulsatile prolactin release during the postpartum lactational period. J Clin Endocrinol Metab 1991; 72:287. 35. Franks S, Murray MA, Jequier AM, et al. Incidence and significance of hyperprolactinaemia in women with amenorrheoa. Clin Endocrinol 1975; 4:597. 36. Jacobs HS, Franks S, Murray MA, et al. Clinical and endocrine features of hyperprolactinaemic amenorrhoea. Clin Endocrinol 1976; 5:439. 37. Patton ML, Woolf PD. Hyperprolactinemia and delayed puberty: a report of three cases and their response to therapy. Pediatrics 1983; 71:572. 38. Asukai K, Uemura T, Minaguchi H. Occult hyperprolactinemia in infertile women. Fertil Steril 1993; 60:423. 39. Bergland RM, Page RB. Can the pituitary secrete directly to the brain? (Affirmative anatomical evidence). Endocrinology 1978; 102:1325. 40. Park SK, Selmanoff M. Hyperprolactinemia suppresses the luteinizing hormone responses to N-methyl-D-aspartate, epinephrine, and neuropeptide-Y in male rats. Endocrinology 1993; 133:2091. 41. Polson DW, Sagle M, Mason HD, et al. Ovulation and normal luteal function during GnRH treatment of women with hyperprolactinaemic amenorrhoea. Clin Endocrinol 1986; 24:531. 42. Lobo RA, Kletzky OA, Kaptein EM, Goebelsmann U. Prolactin modulation of dehydroepiandrosterone sulfate secretion. Am J Obstet Gynecol 1980; 138:632. 43. Klibanski A, Zervas NT. Diagnosis and management of hormone-secreting pituitary adenomas. N Engl J Med 1991; 324:822. 44. Cowden EA, Ratcliffe WA, Ratcliffe JG, et al. Hyperprolactinaemia in renal disease. Clin Endocrinol 1978; 9:241. 44a.Kleinberg DL, Davis JM, de Coster R, et al. Prolactin levels and adverse events in patients treated with risperidone. J Clin Psychopharm 1999; 19:57. 45. 46. 47. 48. 49. Molitch ME. Pathologic hyperprolactinemia. Endocrinol Metab Clin North Am 1992; 21:877. Bassetti M, Spada A, Pezzo G, Giannattasio G. Bromocriptine treatment reduces the cell size in human macroprolactinomas: a morphometric study. J Clin Endocrinol Metab 1984; 58:268. Vermesh M, Fossum GT, Kletzky OA. Vaginal bromocriptine: pharmacology and effect on serum prolactin in normal women. Obstet Gynecol 1988; 72:693. Kletzky OA, Vermesh M. Effectiveness of vaginal bromocriptine in treating women with hyperprolactinemia. Fertil Steril 1989; 51:269. Webster J, Piscitelli G, Polli A, et al. Dose-dependent suppression of serum prolactin by cabergoline in hyperprolactinaemia: a placebo controlled, double blind, multicentre study. European Multicentre Cabergoline Dose-Finding Study Group. Clin Endocrinol 1992; 37:534. 50. Muratori M, Arosio M, Gambino G, et al. Use of cabergoline in the long-term treatment of hyperprolactinemic and acromegalic patients. J Endo-crinol Invest 1997; 20:537. 51. Webster J, Piscitelli G, Polli A, et al. A comparison of cabergoline and bromocriptine in the treatment of hyperprolactinemic amenorrhea. N Engl J Med 1994; 331:904. 52. Colao A, Di Sarno A, Sarnacchiaro F, et al. Prolactinomas resistant to standard dopamine agonists respond to chronic cabergoline treatment [see comments]. J Clin Endocrinol Metab 1997; 82:876. 52a.Cabergoline in the treatment of hyperprolactinemia: a study in 455 patients. J Clin Endocrinol Metab 1999; 84:2518. 53. Homburg R, West C, Brownell J, Jacobs HS. A double-blind study comparing a new non-ergot, long-acting dopamine agonist, CV 205-502, with bromocriptine in women with hyperprolactinaemia. Clin Endocrinol 1990; 32:565. 54. van-der-Lely AJ, Brownell J, Lamberts SW. The efficacy and tolerability of CV 205-502 (a nonergot dopaminergic drug) in macroprolactinoma patients and in prolactinoma patients intolerant to bromocriptine. J Clin Endocrinol Metab 1991; 72:1136. 55. Vance ML, Lipper M, Klibanski A, et al. Treatment of prolactin-secreting pituitary macroadenomas with the long-acting non-ergot dopamine agonist CV 205-502. Ann Intern Med 1990; 112:668. 56. Kletzky OA, Borenstein R, Mileikowsky GN. Pergolide and bromocriptine for the treatment of patients with hyperprolactinemia. Am J Obstet Gynecol 1986; 154:431. 56aFreda PU, Andreadis CI, Khandji AG, et al. Long-term treatment of prolac-tin-secreting macroadenomas with pergolide. J Clin Endocrinol Metab 2000; 85:8. 57. Zervas NT. Surgical results in pituitary adenomas: results of an international study. In: Black PM, Zervas NT, Ridgway EC, Martin JB, eds. Secretory tumors of the pituitary gland. New York: Raven Press, 1984:377. 58. Serri O, Rasio E, Beauregard H, et al. Recurrence of hyperprolactinemia after selective transsphenoidal adenomectomy in women with prolacti-noma. N Engl J Med 1983; 309:280. 59. Schlechte JA, Sherman BM, Chapler FK, VanGilder J. Long term follow-up of women with surgically treated prolactin-secreting pituitary tumors. J Clin Endocrinol Metab 1986; 62:1296. 60. Feigenbaum SL, Downey DE, Wilson CB, Jaffe RB. Transsphenoidal pituitary resection for preoperative diagnosis of prolactin-secreting pituitary adenoma in women: long term follow-up. J Clin Endocrinol Metab 1996; 81:1711. 61. Bergh T, Nillius SJ, Wide L. Bromocriptine treatment of 42 hyperprolacti-naemic women with secondary amenorrhoea. Acta Endocrinol 1978; 88:435. 62. Weiss MH, Teal J, Gott P, et al. Natural history of microprolactinomas: six-year follow-up. Neurosurgery 1983; 12:180.

63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83.

Martin TL, Kim M, Malarkey WB. The natural history of idiopathic hyper-prolactinemia. J Clin Endocrinol Metab 1985; 60:855. Sisam DA, Sheehan JP, Sheeler LR. The natural history of untreated micro-prolactinomas. Fertil Steril 1987; 48:67. Schlechte J, Dolan K, Sherman B, et al. The natural history of untreated hyperprolactinemia: a prospective analysis. J Clin Endocrinol Metab 1989; 68:412. March CM, Kletzky OA, Davajan V, et al. Longitudinal evaluation of patients with untreated prolactin-secreting pituitary adenomas. Am J Obstet Gynecol 1981; 139:835. Klibanski A, Neer RM, Beitins IZ, et al. Decreased bone density in hyper-prolactinemic women. N Engl J Med 1980; 303:1511. Klibanski A, Biller BM, Rosenthal DI, et al. Effects of prolactin and estrogen deficiency in amenorrheic bone loss. J Clin Endocrinol Metab 1988; 67:124. Koppelman MC, Kurtz DW, Morrish KA, et al. Vertebral body bone mineral content in hyperprolactinemic women. J Clin Endocrinol Metab 1984; 59:1050. Klibanski A, Greenspan SL. Increase in bone mass after treatment of hyper-prolactinemic amenorrhea. N Engl J Med 1986; 315:542. Biller BM, Baum HB, Rosenthal DI, et al. Progressive trabecular osteopenia in women with hyperprolactinemic amenorrhea [see comments]. J Clin Endocrinol Metab 1992; 75:692. Schlechte J, Walkner L, Kathol M. A longitudinal analysis of premeno-pausal bone loss in healthy women and women with hyperprolactinemia. J Clin Endocrinal Metab 1992; 75:698. Schlechte JA, Sherman B, Martin R. Bone density in amenorrheic women with and without hyperprolactinemia. J Clin Endocrinol Metab 1983; 56:1120. Molitch ME, Elton RL, Blackwell RE, et al. Bromocriptine as primary therapy for prolactin-secreting macroadenomas: results of a prospective multi-center study. J Clin Endocrinol Metab 1985; 60:698. Ferrari C, Paracchi A, Mattei AM, et al. Cabergoline in the long-term therapy of hyperprolactinemic disorders. Acta Endocrinol 1992; 126:489. Biller BM, Molitch ME, Vance ML, et al. Treatment of prolactin-secreting macroadenomas with the once-weekly dopamine agonist cabergoline. J Clin Endocrinol Metab 1996; 81:2338. Ferrari CI, Abs R, Bevan JS, et al. Treatment of macroprolactinoma with cabergoline: a study of 85 patients. Clin Endocrinol 1997; 46:409. Pellegrini I, Rasolonjanahary R, Gunz G, et al. Resistance to bromocriptine in prolactinomas. J Clin Endocrinol Metab 1989; 69:500. Friedman E, Adams EF, Hoog A, et al. Normal structural dopamine type 2 receptor gene in prolactin-secreting and other pituitary tumors. J Clin Endocrinol Metab 1994; 78:568. De Rosa M, Colao A, Di Sarno A, et al. Cabergoline treatment rapidly improves gonadal function in hyperprolactinemic males: a comparison with bromocriptine. Eur J Endocrinol 1998; 138:286. Molitch ME. Pregnancy and the hyperprolactinemic woman. N Engl J Med 1985; 312:1364. Turkalj I, Braun P, Krupp P. Surveillance of bromocriptine in pregnancy. JAMA 1982; 247:1589. Robert E, Musatti L, Piscitelli G, Ferrari CI. Pregnancy outcome after treatment with the ergot derivative, cabergoline. Reprod Toxicol 1996; 10:333.

CHAPTER 14 ADRENOCORTICOTROPIN: PHYSIOLOGY AND CLINICAL ASPECTS Principles and Practice of Endocrinology and Metabolism

CHAPTER 14 ADRENOCORTICOTROPIN: PHYSIOLOGY AND CLINICAL ASPECTS

DAVID J. TORPY AND RICHARD V. JACKSON Synthesis and Processing of Proopiomelanocortin Functions of Adrenocorticotropin Regulation of Adrenocorticotropin Secretion Mechanism of Control of ACTH Secretion Measurement of Adrenocorticotropin Basal Plasma ACTH Concentrations: Clinical Applications Dynamic Testing of Adrenocorticotropin Secretory Function Clinical Manifestations of Disorders of Adrenocorticotropin Excess Clinical Manifestations of Adrenocorticotropin Deficiency Chapter References

The existence of an adrenocorticotropic factor was predicted by classic pituitary ablation/pituitary extract replacement experiments.1,2 Ovine and human adrenocorticotropin hormone (ACTH) were later isolated and sequenced.3,4 ACTH, a 39 amino acid peptide secreted by the corticotropes (located centrally in the anterior pituitary gland), stimulates adrenocortical steroid synthesis. Measurement of this peptide, formerly difficult, is now reliable and practical, although ACTH measurements are most useful when autonomous secretion or pathologic suppression of ACTH release is suspected. ACTH release is stimulated by the hypothalamic secretion of corticotropin-releasing hormone (CRH) and arginine vasopressin (AVP)in response to stress and the circadian rhythmand inhibited by negative feedback from circulating cortisol. A number of acquired or inherited disorders of ACTH secretion may be manifested by hypercortisolism, adrenal insufficiency, or hyperpigmentation.

SYNTHESIS AND PROCESSING OF PROOPIOMELANOCORTIN

When ACTH is synthesized in the anterior pituitary gland by the corticotropes, it is derived from a much larger precursor, a 241 amino acid peptide, proopiomelanocortin (POMC)5 (Fig. 14-1). The human POMC 8-Kb gene is located on chromosome 2p23.3, consisting of a 400 to 700 bp promoter, three exons, and two introns.6,7 The main regulators of POMC transcription are CRH and glucocorticoids. CRH increases POMC transcription through a cyclic adenosine monophosphatemediated mechanism, whereas glucocorticoids inhibit transcription. In the pituitary, pro-hormone convertase 1 (PC1) cleaves POMC into ACTH and two other large polypeptides, N-terminal peptide and b-lipotropin.8 The roles of these large peptides, which are cosecreted in equimolar amounts to ACTH, are unknown. b-lipotropin was so named because of a mild lipolytic activity9 that has doubtful relevance in humans. Formerly, b-lipotropin was measured by some investigators as a substitute for ACTH. In the brain, ACTH is cleaved by prohormone convertase 2 (PC2), yielding a-melanocyte-stimulating hormone (a-MSH) and corticotropin-like intermediate lobe peptide (CLIP), which are also produced in the human fetal pituitary intermediate lobe. These peptides are not produced in the adult, where the intermediate lobe is no longer present. Although MSH sequences (a-MSH, b-MSH, g-MSH) are contained within POMC fragments, ACTH itself has melanotropic action, increasing melanin synthesis, and may itself lead to pigmentation in hypersecretory states. No MSH peptides are released from the anterior pituitary in humans, nor are these peptides found in human blood. MSH activity in human blood results from MSH sequences contained within ACTH and possibly within b-lipotropin and the N-terminal fragment of POMC.

FIGURE 14-1. Processing of human proopiomel-anocortin (POMC). The processing proceeds in stages, yielding a variety of forms of secreted peptides; N-terminal peptide, adrenocorticotropic hormone (ACTH), and b-lipotropin are the principal circulating forms. Approximately 40% of ACTH has a posttranslational addition of a phosphate moiety. (g-MSH, gamma-melanocyte-stimulating-hormone; LPH, lipotropin; CHO, carbohydrate; PO4, phosphate.)

Animal studies have revealed a role for a-MSH in the regulation of food intake, acting through the brain melanocortin-4 receptor. Studies in Mexican-Americans have shown an association between inheritance of the POMC region of chromosome 2, serum leptin concentration, and obesity.10 POMC gene mutations can result in a monogenic disorder of early-onset obesity, adrenal insufficiency, and red hair.11

FUNCTIONS OF ADRENOCORTICOTROPIN
The primary action of ACTH is to promote steroidogenesisthat is, to enhance the synthesis and secretion of glucocorticoids, mineralocorticoids, and weak androgenic steroids of the adrenal cortex. However, the main physiologic controller of aldosterone release is the renin-angiotensin system (see Chap. 17). The N-terminal 18 amino acids are capable of cortisol release but are subject to rapid degradation; hence, the N-terminal 24 amino acids are used clinically to stimulate ACTH release (cosyntropin). ACTH acts on a specific 297-amino acid cell surface receptor that belongs to the Gs-proteincoupled 7-transmembrane superfamily of receptors. The ACTH receptor (also known as the melanocortin-2 receptor) gene is located at chromosome 18p11.1.12,13 The ACTH receptor acts via a cyclic adenosine monophosphatedependent second messenger pathway to increase adrenal lipoprotein uptake from plasma and increase the transcription rates of genes for enzymes involved in steroidogenesis. Increased low-density lipoprotein uptake is facilitated by an increase in cell-surface low-density lipoprotein receptors. In contrast, ACTH has a smaller effect on zona fasciculata hydroxymethylglutaryl-coenzyme A reductase levels and hence cholesterol synthesis.14,15 Acutely, ACTH increases the activity of the rate-limiting enzyme of adrenal steroidogenesis, the P450SCC enzyme that catalyzes the conversion of cholesterol to D5-pregnenolone. Chronically, ACTH increases the activity of other adrenal enzymes involved in steroidogenesis. ACTH also stimulates protein synthesis, resulting in adrenal hypertrophy and hyperplasia. The activity of phenylethanolamine-N-methyltransferase, the enzyme that catalyzes epinephrine production (see Chap. 85), is dependent on cortisol, and hence ACTH levels. ACTH has a trophic effect on tyrosine hydroxylase activity, the enzyme responsible for catalyzing the rate-limiting step in catecholamine biosynthesis, and increases the rate of melanin synthesis in melanocytes, which leads to skin pigmentation. REGULATION OF ADRENOCORTICOTROPIN SECRETION The anterior pituitary contains ~600 g of ACTH.16 ACTH, which is released on a pulsatile basis with peaks at ~30-minute intervals,17 is subject to intravascular enzymatic degradation with a plasma disappearance half-life of 7 to 12 minutes.18 Cortisol pulses follow those of ACTH by ~30 minutes, although not all cortisol peaks follow those of ACTH.19 The relationship between increasing levels of ACTH and consequent cortisol secretion is defined by a sigmoidal curve.20 Very high levels of ACTH do not further increase plasma cortisol concentrations, although the duration of a cortisol secretory burst continues to increase.21 Under some circumstances, the relation between ACTH and cortisol secretion is disturbed. In chronic stress, such as critical illness, a greater cortisol release occurs for an additional given ACTH stimulus because of adrenal hypertrophy and altered adrenal enzyme activities, which favor production of cortisol over that of adrenal androgens.22 Elevated ACTH levels after appropriate stimuli, with blunted or normal cortisol responses, are seen in myotonic dystrophy,23 a multisystem genetic disorder, and fibro-myalgia,24 an idiopathic pain syndrome. In these cases, the primary defect is thought to lie in ACTH regulation rather than an altered ACTH bioactivity or adrenal hyporesponsiveness. Regulation of ACTH release is subject to three themes: stress, the circadian rhythm, and glucocorticoid negative feedback. Stress, defined as a threat to homeostasis,

derives from such factors as sepsis, trauma, or emotion. In response to stress, ACTH release is greatly increased; consequently, the secretion of cortisol, the principal glucocorticoid in humans, can increase five-fold. The principal functions of cortisol during stress are to restrain the immune system, by reducing production of potentially damaging cytokines25; augment the effects of catecholamines on the vascular system; mobilize glucose and fatty acids for metabolic use; and sharpen cognition to allow appropriate behavioral responses. During starvation, cortisol acts to increase hepatic glucose production for use as energy. ACTH secretion follows a light-entrained circadian rhythm with peak cortisol blood levels attained at ~6:00 a.m. to 8:00 a.m. and the lowest concentrations at approximately midnight.26,27 and 27a In animals, the circadian rhythm is based on serotonergic pathways arising from the suprachiasmatic nucleus,28 although the mechanismor indeed the functionof circadian ACTH release in humans is unknown. In humans, severe disturbances of circadian ACTH release following transmeridian travel require several days to be reentrained29 (see Chap. 6). Normally, glucocorticoid negative feedback controls ACTH release; however, stress and circadian factors make ACTH release less susceptible to feedback inhibition. At physiologic levels of cortisol, the brain, rather than the pituitary, is the main site of feedback inhibition.30 Cortisol acts at two types of receptors in the hypothalamus and hippocampus to inhibit CRH release; these are type-1 (mineralocorticoid) and type-2 (classic glucocorticoid) receptors. The high affinity type-1 receptors are thought to mediate suppression of CRH release in response to basal conditions, whereas type-2 receptors mediate stress-level cortisol inhibition.31,32 Both rapid and delayed feedback of glucocorticoids on ACTH secretion are observed. Rapid feedback is responsive to the rate of change in glucocorticoid concentrations; delayed feedback responds to absolute circulating glucocorticoid levels.33 The progressive rise in plasma ACTH, which occurs during pregnancy, is associated with increased total and free cortisol (two- to three-fold). This may be due to stimulation of the maternal pituitary corticotropes by placental CRH.34 Transiently reduced central CRH secretion and relative hypocortisolism in the postpartum period may contribute to the mood and autoimmune phenomena observed at this time. Leptin is a peptide hormone produced by adipocytes. It was isolated after studies of the leptin-deficient ob/ob (obese) mouse.35 Leptin acts on specific neuropeptide/neurotransmitter pathways of the central nervous system and through these modulatory effects inhibits appetite. Starvation is associated with low leptin levels and hypercortisolism. Moreover, plasma leptin and cortisol have an inversely related circadian rhythm.36 These findings suggest a close relationship between the hypothalamicpituitaryadrenal (HPA) axis and leptin. Leptin inhibits ACTH secretion by inhibiting CRH synthesis.37 Leptin also inhibits adrenocortical enzyme activity.38 Hence, low leptin levels may account for the physiologic hypercortisolism of starvation, allowing stimulation of fuel catabolism and liberating amino acids and glucose39 (see Chap. 186). MECHANISM OF CONTROL OF ACTH SECRETION As previously stated, pituitary ACTH secretion is subject to regulation by the hypothalamic hormones, principally CRH and AVP (Fig. 14-2).40 These hypophysiotropic factors are secreted into the hypothalamic-pituitary portal circulation at the median eminence from neurons arising in the parvicellular portion of the hypothalamic paraventricular nucleus. CRH, a 41 amino acid peptide discovered in 1981,41 is regarded as the main proximate regulator of ACTH secretion, acting via the CRH-R1 receptor.42,43 The effects of CRH and AVP are synergistic with respect to ACTH release.44 Both CRH and AVP have been used to stimulate ACTH release in clinical studies.

FIGURE 14-2. Neuroendocrine regulation of the hypothalamicpituitaryadrenal (HPA) axis in humans. The circadian rhythm, stress, and feedback inhibition from circulating cortisol are the major regulators of HPA axis function. The two major secretagogues controlling adrenocorticotropin hormone (ACTH) release are corticotropin-releasing hormone (CRH) and arginine vasopressin (AVP). Noradrenergic input from the brainstem comprises a major stimulatory path to CRH release and links the HPA axis and the sympathetic nervous system.

The hypophysiotropic peptides are themselves regulated by a host of neurotransmitter pathways. Research into the role and importance of these pathways in humans has relied on indirect studies of pharmacologic agents through measurement of ACTH and cortisol release. In vitro studies of hypothalamic organ systems and animal preparations have allowed further insights. Importantly, CRH secretion is stimulated by noradrenergic neurons arising in the brainstem.45 These are themselves innervated by CRH neurons, which provide a reverberating feedback loop to link the two great effectors of the stress response (the HPA axis and the sympathetic nervous system)46 and are under tonic inhibition by central opioid pathways.47 Other neurotransmitter systems that stimulate the CRH neuron include acetylcholine, serotonin, and neuropeptide Y. CRH secretion and central noradrenergic nuclei are inhibited by the gamma amino butyric acid/benzodiazepine system.48 Other neurotransmitter systems with inhibitory effects on the CRH neuron include substance P and arcuate nucleus-derived POMC. A major regulator of ACTH release is the immune system. Cytokinesproduced in inflammatory sites with a time course favoring consecutive release of TNF-a, interleukin-1, and interleukin-6 (IL-6)are potent releasers of ACTH and cortisol. Acutely, IL-6 may act principally at the brain and hypothalamus to indirectly cause release of ACTH, although the pituitary and adrenal can also be stimulated directly.49 In this regard, IL-6 in particular may be considered a hormone that is the major link between the immune system and the HPA axis. The result is a classic feedback loop between IL-6 and the HPA axis, in which cytokines stimulate the axis and cortisol suppresses the immune system and release of cytokines.50 MEASUREMENT OF ADRENOCORTICOTROPIN Traditional radioimmunoassays rely on a single-site antibody that measures both intact ACTH and abnormal ACTH, or POMC fragments, such as may be secreted from ectopic ACTH-secreting tumors. The highest correlation between the radioimmunoassay and bioassay of ACTH is found with midportion assays (i.e., assays using antisera that cross-react with the mid-portion of the ACTH molecule). These antisera bridge the sites of proteolytic cleavage of ACTH in the 15-amino acid to 18-amino acid portion, thus binding to the portion of the molecule responsible for the steroidogenic activity.51 Greater sensitivity and specificity have been reported for the two-site immunoradiometric assay. Two different antibodies are used, each directed to a different portion of the ACTH molecule. For ACTH to be detected in this assay, each site must be bound to its respective antibody. The detection limit is as low as 2 to 3 pg/mL of plasma, making it possible to measure ACTH without extraction, even in samples with suppressed ACTH levels. Nonetheless, these immunoradiometric assays may fail to detect aberrant large molecular mass forms of ACTH, resulting in misleadingly low levels.52 Detection of large molecular mass forms of ACTH is usually indicative of ectopic (extrapituitary) ACTH production. Samples for the measurement of ACTH must be collected and immediately placed on ice for plasma separation within a short period. Plasma must be stored frozen until assay. This is necessary to prevent catabolism of ACTH by circulating peptidases. Mishandling of samples for clinical ACTH measurement leads to falsely low ACTH levels. BASAL PLASMA ACTH CONCENTRATIONS: CLINICAL APPLICATIONS In many circumstances, because the levels of plasma or urinary cortisol are directly related to those of ACTH and can be measured more economically, cortisol levels can be used as an indirect index of ACTH activity. However, there are several circumstances in which the measurement of ACTH is invaluable. These include the differential diagnosis of ACTH-dependent or independent Cushing syndrome (see Chap. 75); measurement of ACTH in the inferior petrosal sinuses for definitive diagnosis of pituitary Cushing syndrome; the differential diagnosis of adrenal insufficiency; and the monitoring of Nelson syndrome. Typical plasma ACTH concentrations found in pituitary-adrenal disorders are shown in Figure 14-3. Substantial overlap exists between the ACTH values in ectopic Cushing and pituitary-dependent Cushing syndrome; hence, ACTH can not be used to differentiate these disorders, although ACTH levels >2000 pg/mL, in the setting of

hypercortisolism, are virtually pathognomic of ectopic Cushing syndrome. Very high ACTH values are also seen in Nelson syndrome. Even in severe, untreated primary adrenal insufficiency, plasma ACTH rarely exceeds 2000 pg/mL. Low or undetectable plasma ACTH values occur in patients with hypopituitarism during or after exogenous corticosteroid administration and in patients with cortisol-secreting tumors (e.g., adrenal adenoma).

FIGURE 14-3. Plasma adrenocorticotropic hormone (ACTH) concentrations in pituitary-adrenal disorders, as determined with an assay using an N-terminal ACTH antiserum. Patients with treated Cushing disease had previous bilateral adrenalectomy, no history of pituitary irradiation, and were taking replacement doses of corticosteroids. Patients with Nelson syndrome had the classic features, except visual field defects were not present in all cases.

Measurement of plasma ACTH in samples obtained simultaneously from the inferior petrosal sinuses and peripheral plasma forms the basis of the currently most definitive test for differentiation of pituitary and ectopic sources of ACTH in ACTH-dependent Cushing syndrome. Ratios of >2:1 before CRH injection and >3:1 after CRH injection indicate pituitary Cushing syndrome.53 The test is generally only necessary in selected cases of ACTH-dependent Cushing syndrome that have eluded diagnosis through less invasive testing. In adrenal insufficiency, ACTH concentrations can be used to distinguish pituitary from adrenal causes. Inappropriately low or low-normal ACTH levels in the setting of adrenocortical deficiency are indicative of ACTH deficiency that is due to pituitary or hypothalamic disease. High ACTH levels in this setting are seen in primary adrenal insufficiency.

DYNAMIC TESTING OF ADRENOCORTICOTROPIN SECRETORY FUNCTION

In states of possible autonomous ACTH hypersecretion, dexamethasone is used to inhibit ACTH secretion, and the plasma cortisol concentration or urinary cortisol excretion can be used to estimate ACTH secretory function (see Chap. 74). This technique is used in the diagnosis and differential diagnosis of Cushing syndrome and has been used to demonstrate hypercortisolism in melancholic depression. Many stimuli have been used to stimulate ACTH secretion for clinical diagnostic purposes. Indications include Cushing syndrome and hypoadrenalism. The most frequently used test of HPA reserve is the cosyntropin stimulation test. This test is based on the rationale that if adrenal cortisol reserves are normal, the hypothalamic CRH and pituitary ACTH reserves must also be normal. Typically 250-g ACTH (124) is administered intravenously and cortisol levels measured at baseline, 30 minutes, and 60 minutes. If plasma cortisol levels rise above 19 g/dL at 30 (or 60) minutes, this is interpreted as a normal adrenal response. Peak cortisol levels of 16 g/dL at 60 minutes occur after intramuscular cosyntropin.54 In general, this is used to demonstrate both pituitary ACTH-secretory integrity and normal adrenal-cortisol secretory function, because adrenal atrophy develops rapidly in states of ACTH deficiency. An abnormal cosyntropin test is highly specific for adrenal insufficiency. In cases of pituitary or hypothalamic damage, in which sufficient time may not have passed to allow adrenal atrophy from ACTH deprivation, the cosyntropin test is normal. However, glucocorticoid deficiency crises have occurred in individuals with a normal cosyntropin test, especially in patients with central hypoadrenalism (pituitary or hypothalamic causes) who may be mistakenly diagnosed as normal in as many as 40% of cases.55 Low sensitivity of the 250-g cosyntropin test has led to studies using a lower, more physiologic, 1-g dose, which produces cortisol responses in normal subjects comparable to those obtained with 250-g cosyntropin and appears more reliable in the diagnosis of central adrenal insufficiency.56 A safe, reliable single test for assessing the functional reserve of the HPA axis57 is still not available. Therefore, good clinical judgment is needed to select the correct test, or combination of tests, to determine whether there is an abnormality of HPA axis function. Insulin-induced hypoglycemia involves the intravenous injection of 0.15 IU/kg insulin into a fasting subject. Hypoglycemia induces a profound and sustained hypercortisolism (cortisol >19 g/dL). Careful medical supervision is essential, and the test is contraindicated in patients older than age 60 or in those who have had myocardial ischemia or epilepsy. This test, because of vast accumulated experience, is often regarded as the gold standard against which other tests of HPA axis reserve are compared. The CRH test (see Chap. 74, Chap. 75 and Chap. 76) involves the intravenous injection of CRH (1 g/kg) with measurement of ACTH and cortisol at 1, 15, 30, 45, and 60 minutes. The test directly measures corticotrope function and offers the potential to separate pituitary from hypothalamic causes of ACTH deficiency, because pituitary lesions produce an attenuated ACTH/cortisol response to CRH.58 Peak cortisol levels are similar to those observed in the ACTH stimulation test. Metyrapone inhibits the 11-hydroxylase enzyme, thereby inhibiting cortisol synthesis and leading to high levels of its immediate precursor, 11-deoxycortisol, which is stimulated by a lack of cortisol feedback and consequent ACTH hypersecretion (see Chap. 74). This test is used in the differential diagnosis of ACTH-dependent Cushing syndrome; patients with ectopic Cushing syndrome are less sensitive to glucocorticoid feedback, so the 11-deoxycortisol response is lower than in pituitary Cushing syndrome59 (see Chap. 74, Chap. 75 and Chap. 219).

CLINICAL MANIFESTATIONS OF DISORDERS OF ADRENOCORTICOTROPIN EXCESS

Hyperpigmentation is common when ACTH levels exceed 300 pg/mL (Fig. 14-4). The highest ACTH levels are seen in Addison disease, Nelson syndrome, and ectopic ACTH secretion. Clinically, the hyperpigmentation is found particularly in areas of increased pressure (elbows, knuckles, knees) and is accentuated in areas of normal pigmentation (areolae, genitalia, palmar creases). Surfaces (such as the mucosal) that are not normally pigmented may exhibit hyperpigmentation. Scars acquired after the onset of ACTH excess also may exhibit hyperpigmentation (see Chap. 76).

FIGURE 14-4. Pigmented lunulae in a patient with the paraneoplastic adrenocorticotropic hormone (ACTH) syndrome who developed severe hyperpigmentation of rapid onset 3 months previously. More commonly, pigmentation of the nails related to excess ACTH is longitudinal rather than transverse, as in this patient. (For other examples of hyperpigmentation related to excess ACTH, see Chap. 76 and Chap. 219.)

Most cases of primary adrenal insufficiency are due to autoimmunity; infiltration or infection of the adrenal gland; or, rarely, X-linked adrenoleukodystrophy. Each of these may be associated with specific clinical features such as other immune disorders (type II autoimmune polyglandular syndrome), infectious manifestations, or neurologic disease, respectively.

Congenital resistance to ACTH may occur in isolated familial glucocorticoid deficiency, an autosomal recessive disorder that is due to a mutation of the ACTH receptor.60 Affected children have hypoglycemic episodes, hyperpigmentation, and failure to thrive. Mineralocorticoid production is normal, and there is no cortisol response to exogenous ACTH. Patients with the Allgrove syndrome (ACTH resistance, achalasia, and alacrima) have ACTH resistance but lack mutations in the gene for the ACTH receptor.61 Nelson syndrome entails the development of an invasive ACTH-secreting macroadenoma after bilateral adrenalectomy to alleviate the hypercortisolism of pituitary ACTH-dependent Cushing syndrome.62 Severe pigmentation occurs, and ACTH levels may be used as a tumor marker, particularly after pituitary surgery has made the identification of a pituitary tumor on magnetic resonance imaging unreliable. The development of Nelson syndrome could be prevented in children by presurgical pituitary irradiation.63 Nelson syndrome is not a universal concomitant of bilateral adrenalectomy; an association between the development of Nelson syndrome and the presence of a glucocorticoid receptor mutation has been found.64 Because bilateral adrenalectomy is now rarely performed in pituitary corticotropinoma cases, Nelson syndrome is correspondingly rare. Autonomous excess secretion of ACTH produces hypercortisolism and Cushing syndrome.65,66 The source of ACTH is generally a pituitary adenoma, although 10% to 20% of cases are due to an ectopic neuroendocrine tumor, most commonly a bronchial carcinoid. Clinical features of Cushing syndrome include centripetal obesity, muscle weakness, osteoporosis, skin thinning with bruising and striae, glucose intolerance, hirsutism, and hypertension. The presence of hyperpigmentation in the context of Cushing syndrome tends to suggest ectopic paraneoplastic ACTH production, in which the ACTH levels are often very high (see Chap. 219).

CLINICAL MANIFESTATIONS OF ADRENOCORTICOTROPIN DEFICIENCY

A number of pituitary pathologies can lead to ACTH deficiency, including pituitary tumor, autoimmune lymphocytic hypophysitis (often postpartum), and granulomatous infiltration. Pituitary hormone deficiencies occasionally develop without apparent cause, although the most common cause in adults is a pituitary tumor, or treatment with surgery or irradiation. ACTH deficiency is generally a late manifestation of pituitary disease, following growth hormone deficiency (poor growth in children), gonadotropin deficiency (loss of menses in women, loss of libido in men), and hypothyroidism (see Chap. 45) that is generally mild in pituitary disease but may be symptomatic with fatigue and cold intolerance. ACTH deficiency produces hypoadrenalism. Loss of pigmentation may be apparent, particularly in children. Hypoadrenalism commonly causes fatigue and hypotension, which is often postural. Hyponatremia may be evident because glucocorticoids contribute to the maintenance of renal free water excretion; symptomatic hypoglycemia can occur. Mineralocorticoid production is regulated chiefly by angiotensin II and the potassium ion, thus mineralocorticoid deficiency does not develop in states of ACTH deficiency. Isolated ACTH deficiency is rare, because pituitary pathologies such as tumor, radiation, or autoimmune disease typically spare the corticotropes until deficiencies of growth hormone, gonadotropins, and, finally, thyroid-stimulating hormone emerge. However, most cases of isolated ACTH deficiencies are likely to be autoimmune because of an association with other autoimmune disorders67 or antipituitary antibodies.68 Congenital ACTH deficiency that is due to a defect in POMC cleavage has been described.69 Undetectable ACTH levels, normal cortisol levels, and increased cortisol secretory sensitivity to ACTH have been reported in a single patient. Two point mutations of the ACTH receptor gene were detected, suggesting the existence of an ACTH hypersensitivity syndrome.70 Acquired isolated CRH deficiency has also been suggested, based on a lack of ACTH response to insulin hypoglycemia but a normal response to exogenous CRH.71 The CRH stimulation test will probably not allow sufficiently accurate separation of pituitary or hypothalamic causes of ACTH deficiency58 to be used as a gold standard test, but it may be useful along with other tests to assist clinical decision making. A transient functional deficiency of ACTH release occurs after prolonged supraphysiologic glucocorticoid exposure, such as occurs in response to treatment with antiinflammatory doses of glucocorticoids or to correction of Cushing syndrome. Rapid restoration of ACTH release with CRH infusion suggests that the defect may be due predominantly to reduced CRH secretion.72 This functional deficiency may last for as many as 2 years before normal HPA axis dynamics are restored, although a 6- to 9-month recovery course is often observed.73 CHAPTER REFERENCES
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. Smith PE. Hypophysectomy and a replacement therapy in the rat. Am J Anat 1930; 45:205. Collip JB, Anderson EM, Thomson DL. The adrenocorticotropic hormone of the anterior lobe. Lancet 1933; 2:347. Li CH, Beschwind II, Cole RD, et al. Amino acid sequence of a corticotropin. Nature 1955; 176:687. Lee TH, Lerner AB, Buettner-Janusch U. Isolation and structure of human corticotropin (ACTH). J Am Chem Soc 1959; 81:6084. Eipper BA, Mains RE. Structure and biosynthesis of pro-adrenocorticotropin/endorphin and related peptides. Endocr Rev 1980; 1:1. Satoh H, Mori S. Subregional assignment of the proopiomelanocortin gene (POMC) to human chromosome band 2p23.3 by fluorescence in situ hybridization. Cytogenet Cell Genet 1997; 76:221. Chang AC, Cochet M, Cohen SN. Structural organization of human genomic DNA encoding the proopiomelanocortin peptide. Proc Natl Acad Sci U S A 1980; 77:4890. Marcinkiewicz M, Day R, Seidah NG, Chretien M. Ontogeny of the prohormone convertases PC1 and PC2 in the mouse hypophysis and their colocalization with corticotropin and a-melanotropin. Proc Natl Acad Sci U S A 1993; 90:4922. Li CH, Barafi L, Chretien M, Chung D. Isolation and structure of b-LPH from sheep pituitary glands. Nature 1965; 208:1093. Comuzzie AG, Hixson JE, Almasy L, et al. A major quantitative trait locus determining serum leptin levels and fat mass is located on human chromosome 2. Nat Genet 1997; 15:273. Krude H, Biebermann H, Luck W, et al. Severe early-onset obesity, adrenal insufficiency and red hair pigmentation caused by POMC mutations in humans. Nat Genet 1998; 19:155. Mountjoy KG, Robbins LS, Mortrud MT, Cone RD. The cloning of a family of genes that encode the melanocortin receptors. Science 1992; 257:1248. Magenis RE, Smith, L, Nadeau JH, et al. Mapping of the ACTH, MSH, and neural (MC3 and MC4) melanocortin receptors in the mouse and human. Mammalian Genome 1994; 5:503. Shima S, Mitsunaga M, Nakao T, et al. Effect of ACTH on cholesterol dynamics in rat adrenal tissue. Endocrinology 1972; 98:808. Brody RI, Black VH. Differential ACTH response of immunodetectable HMG CoA reductase and cytochromes P450 (17 alpha) and P450 (21) in guinea pig adrenal outer zone cell types, zona glomerulosa and zona fasciculata. Endocr Res 1991; 17:195. Frohman LA. Diseases of the anterior pituitary. In: Felig P, Baxter JD, Broadu AE, Frohman LA, eds. Endocrinology and Metabolism, 2nd ed. New York: McGraw-Hill 1987:247. Gallagher TF, Yoshida K, Roffwarg HD, et al. ACTH and cortisol secretory patterns in man. J Clin Endocrinol Metab 1973; 36:1058. Krieger DT, Allen W. Relationship of bioassayable and immunoassayable plasma ACTH and cortisol concentration in animal subjects and in patients with Cushing' disease. J Clin Endocrinol Metab 1975; 10:675. Horrocks PM, Jones AF, Ratcliffe WA, et al. Patterns of ACTH and cortisol pulsatility over twenty-four hours in normal males and females. Clin Endocrinol (Oxf) 1990; 32:127. Schurmeyer TH. On the relationship between ACTH and cortisol secretion. Horm Metab Res Suppl 1987; 16:6. Keller-Wood ME, Dallman MF. Corticosteroid inhibition of ACTH secretion. Endocr Rev 1984; 5:1. Reincke M, Lehmann R, Karl M, et al. Severe illness. Neuroendocrinology. Ann NY Acad Sci 1995; 771:556. Grice JE, Jackson RV, Hockings GI, et al. Adrenocorticotropin hyperresponse to the corticotropin-releasing hormone-mediated stimulus of naloxone in patients with myotonic dystrophy. J Clin Endocrinol Metab 1995; 80:179. Crofford LJ, Engleberg NC, Demitrack MA. Neurohormonal perturbations in fibromyalgia. Baillieres Clin Rheumatol 1996; 10:365. Munck A, Guyre PM, Holbrook, NJ. Physiological functions of glucocorticoids in stress and their relation to pharmacological actions. Endocr Rev 1984; 5:25. Sack RL, Lewy AJ, Blood ML, et al. Circadian rhythm abnormalities in totally blind people: incidence and clinical significance. J Clin Endocrinol Metab 1992; 75:127. Krieger DT. Rhythms of ACTH and corticosteroid secretion in health and disease and their experimental modification. J Steroid Biochem 1975; 6:785.

27a.Luboshitzky R. Endocrine activity during sleep. J Pediatr Endocrinol Metab 2000; 13:13. 28. Banky Z, Molnar J, Csernus V, Halasz B. Further studies on circadian rhythms after local pharmacological destruction of the serotoninergic innervation of the rat suprachiasmatic region before the onset of the corticosterone rhythm. Brain Res 1988; 445:222. 29. Desir D, Van Cauter E, Fang VS, et al. Effects of jetlag on hormonal patterns. I. Procedures, variations in total plasma proteins, and disruption of adrenocorticotropin cortisol periodicity. J Clin Endocrinol Metab 1981; 52:628. 30. Levin N, Shinsako J, Dallman M. Corticosterone acts on the brain to inhibit adrenalectomy-induced adrenocorticotropin secretion. Endocrinology 1988; 122:694. 31. Reul JHM, de Kloet ER. Two receptor systems for corticosterone in rat brain: microdistribution and differential occupation. Endocrinology 1985; 117:2505. 32. Ratka A, Sutanta W, Bloemers M, de Kloet ER. On the role of brain mineral-ocorticoid (Type I) and glucocorticoid (Type II) receptors in neuroendocrine regulation. Neuroendocrinology 1989; 50:117. 33. Dallman MF, Akana SF, Levin N, et al. Corticosteroids and the control of function in the hypothalamopituitaryadrenal (HPA) axis. Ann NY Acad Sci 1994; 746:22. 34. Magiakou MA, Mastorakos G, Rabin D, et al. Hypothalamic corticotropin-releasing hormone suppression during the postpartum period: implications for the increase in psychiatric manifestations at this time. J Clin Endocrinol Metab 1996; 81:1912. 35. Zhang Y, Proneca R, Maffei M, et al. Positional cloning of the mouse obese gene and its human homologue. Nature 1994; 372:425. 36. Licinio J, Mantzoros C, Negrao AB, et al. Human leptin levels are pulsatile and inversely related to pituitary-adrenal function. Nat Med 1997; 3:575. 37. Heiman ML, Ahima RS, Craft LS, et al. Leptin inhibition of the hypothalamic pituitaryadrenal axis in response to stress. Endocrinology 1997; 138:3859. 38. Bornstein SR, Uhlmann K, Haidan A, et al. Evidence for a novel peripheral action of leptin as a metabolic signal to the adrenal gland: leptin inhibits cortisol release directly. Diabetes 1997; 46:1235. 39. Flier JS. Clinical review 94: What' in a name? In search of leptin' physiologic role. J Clin Endocrinol Metab 1998; 83:1407. 40. Orth DN, Jackson RV, DeCherney GS, et al. Effect of synthetic ovine corticotropin releasing factor: dose response of plasma ACTH and cortisol. J Clin Invest 1983; 71:587. 41. Vale W, Spiess J, Rivier C, Rivier J. Characterization of a 41-residue ovine hypothalamic peptide that stimulates secretion of corticotropin and b-endorphin. Science 1981; 213:1394. 42. Castro MG, Morrison E, Perone MJ, et al. Corticotrophin-releasing hormone receptor type 1: generation and characterization of polyclonal antipeptide antibodies and their localization in pituitary cells and cortical neurones in vitro. J Neuroendocrinol 1996; 8:521.

43. Webster EL, Lewis DB, Torpy DJ, et al. In vivo and in vitro characterization of antalarmin, a nonpeptide corticotropin-releasing hormone (CRH) receptor antagonist: suppression of pituitary ACTH release and peripheral inflammation. Endocrinology 1996; 137:5747. 44. DeBold CR, Sheldon WR, DeCherney GS, et al. Arginine vasopressin potentiates adrenocorticotropin release induced by ovine corticotropin-releasing factor. J Clin Invest 1984; 73:533. 45. Al-Damluji S, Perry L, Tolin S, et al. Alpha-adrenergic stimulation of corticotrophin secretion by a specific central mechanism in man. Neuroendocrinology 1987; 45:68. 46. Chrousos GP, Gold PW. The concepts of stress and stress system disorders. Overview of physical and behavioral homeostasis. JAMA 1992; 267:1244. 47. Jackson RV, Grice JE, Hockings GI, Torpy DJ. Naloxone-induced ACTH release: mechanism of action in humans. Clin Endocrinol (Oxf) 1995; 43:423. 48. Torpy DJ, Grice JE, Hockings GI, et al. Alprazolam blocks the naloxone-stimulated hypothalamopituitaryadrenal axis in man. J Clin Endocrinol Metab 1993; 76:388. 49. Papanicolaou DA, Wilder RL, Manolagas SC, Chrousos GP. The pathophysiologic roles of interleukin-6 in human disease. Ann Intern Med 1998; 128:127. 50. Chrousos GP. The hypothalamicpituitaryadrenal axis and immune-mediated inflammation. N Engl J Med 1995; 332:1351. 51. Nicholson WE, Davis DR, Sherrell BJ, Orth DN. Rapid radioimmunoassay for corticotropin in unextracted human plasma. Clin Chem 1984; 30:259. 52. Findling JW, Engeland WC, Raff H. The use of immunoradiometric assay for the measurement of ACTH in human plasma. Trends Endocrinol Metab 1990; 1:283. 53. Oldfield EH, Doppman JL, Nieman LK, et al. Petrosal sinus sampling with and without corticotropin releasing hormone for the differential diagnosis of Cushing' syndrome. N Engl J Med 1991; 325:897. 54. Longui CA, Vottero A, Harris AG, Chrousos GP. Plasma cortisol responses after intramuscular corticotropin 1-24 in healthy men. Metabolism 1998; 47:1419. 55. Hockings GI, Strakosch CR, Jackson RV. Secondary adrenocortical deficiency: avoiding potentially fatal pitfalls in diagnosis and treatment. Med J Aust 1997; 166:400. 56. Thaler LM, Blevins L Jr. The low dose (1 g) adrenocorticotropin stimulation test in the evaluation of patients with suspected central adrenal insufficiency. J Clin Endocrinol Metab 1998; 83:2726. 57. Streeten DH, Anderson GH Jr, Bonaventura MM. The potential for serious consequences from misinterpreting normal responses to the rapid adrenocorticotropin test. J Clin Endocrinol Metab 1996; 81:285. 58. Schulte HM, Chrousos GP, Avgerinos P, et al. The corticotropin releasing hormone test: a possible aid in the evaluation of patients with adrenal insufficiency. J Clin Endocrinol Metab 1984; 58:1064. 59. Avgerinos PC, Nieman LK, Oldfield EH, Cutler GB Jr. A comparison of the overnight and the standard metyrapone test for the differential diagnosis of adrenocorticotrophin-dependent Cushing' syndrome. Clin Endocrinol (Oxf) 1996; 45:483. 60. Weber A, Topperi J, Harvey RD, et al. Adrenocorticotropin receptor gene mutations in familial glucocorticoid deficiency: relationships with clinical features in four families. J Clin Endocrinol Metab 1995; 80:65. 61. Heinrichs C, Tsigos C, Deschepper J, et al. Familial adrenocorticotropin unresponsiveness associated with alacrima and achalasia: biochemical and molecular studies in two siblings with clinical heterogeneity. Eur J Pediatr 1995; 154:191. 62. Negesser SK, van Seters AP, Kievit J, et al. Long-term results of total adrenalectomy for Cushing' disease. World J Surg 2000; 24:108. 63. Jennings AS, Liddle GW, Orth DN. Results of treating childhood Cushing' disease with pituitary irradiation. N Engl J Med 1977; 297:957. 64. Karl M, von Wichert G, Kempter E, et al. Nelson' syndrome associated with a somatic frame mutation in the glucocorticoid receptor gene. J Clin Endocrinol Metab 1996; 81:124. 65. Newell-Price J, Trainer P, Besser M, Grossman A. The diagnosis and differential diagnosis of Cushing' syndrome and pseudo-Cushing' states. Endocr Rev 1998; 19:647. 66. Orth DN. Cushing' syndrome. N Engl J Med 1995; 332:791. 67. Shigemasa C, Kouchi T, Veta Y, et al. Evaluation of thyroid function in patients with isolated ACTH deficiency. Am J Med Sci 1992; 304:279. 68. Sugiura M, Hashimoto A, Shizawa M, et al. Heterogeneity of anterior pituitary cell antibodies detected in insulin dependent diabetes mellitus and ACTH deficiency. Diabetes Res 1980; 3:11. 69. Nussey SS, Soo SC, Gibson S, et al. Isolated congenital ACTH deficiency. A cleavage enzyme defect? Clin Endocrinol (Oxf) 1993; 39:381. 70. Hiroi N, Yakushiji F, Shimojo M, et al. Human ACTH hypersensitivity syndrome associated with abnormalities of the ACTH receptor gene. Clin Endocrinol (Oxf) 1998; 48:129. 71. Nishihara E, Kimura H, Ishimaru T, et al. A case of adrenal insufficiency due to acquired hypothalamic CRH deficiency. Endocr J 1997; 44:121. 72. Gomez MT, Magiakou MA, Mastorakos G, Chrousos GP. The pituitary corticotroph is not the rate limiting step in the postoperative recovery of the hypothalamicpituitaryadrenal axis in patients with Cushing syndrome. J Clin Endocrinol Metab 1993; 77:173. 73. Graber RL, Ney RL, Nicholson WE, et al. Natural history of pituitary-adrenal recovery following long-term suppression with corticosteroids. J Clin Endocrinol Metab 1965; 25:11.

CHAPTER 15 THYROID-STIMULATING HORMONE AND ITS DISORDERS Principles and Practice of Endocrinology and Metabolism

CHAPTER 15 THYROID-STIMULATING HORMONE AND ITS DISORDERS

JOSHUA L. COHEN Thyroid-Stimulating Hormone Structure, Biosynthesis, and Function Glycoprotein Hormone Structure Biosynthesis of Thyroid-Stimulating Hormone Thyroid-Stimulating Hormone Receptor Biologic Actions of Thyroid-Stimulating Hormone Control of Thyroid-Stimulating Hormone Secretion: Neural and Humoral Modulators Neural Control of Thyroid-Stimulating Hormone Secretion Thyrotropin-Releasing Hormone Thyroid Hormones Dopamine and Somatostatin Steroid Hormones Integrated Control of Thyroid-Stimulating Hormone Secretion Control of Thyroid-Stimulating Hormone Secretion: Physiologic Modulation Effects of Age and Sex Daily Rhythm Effects of Stress Effects of Fasting and Severe Illness Clinical Applications of Thyroid-Stimulating Hormone Measurement Thyroid-Stimulating Hormone Assay Thyroid-Stimulating Hormone in Primary Hypothyroidism Thyroid-Stimulating Hormone in Hyperthyroidism Evaluation of Therapy for Thyroid Diseases Disorders of Thyroid-Stimulating Hormone Control Disorders Causing Hypothyroidism Disorders Causing Hyperthyroidism Defects in Thyroid-Stimulating Hormone Receptor Function Clinical Use of Recombinant Human Thyroid-Stimulating Hormone Chapter References

THYROID-STIMULATING HORMONE STRUCTURE, BIOSYNTHESIS, AND FUNCTION

GLYCOPROTEIN HORMONE STRUCTURE Thyroid-stimulating hormone (TSH, thyrotropin) is one of a family of glycoprotein hormones that also includes follicle-stimulating hormone (FSH), luteinizing hormone (LH), and human chorionic gonadotropin (hCG, Table 15-1). These hormones each consist of two dissimilar subunits, a and b, held together by strong noncovalent bonds.1 Each subunit consists of a polypeptide core that is stabilized by internal disulfide bonds and glycosylated at specific residues. The a subunits of all the glycoprotein hormones are identical and are highly conserved among different species. The b subunits of different hormones have extensive sequence homologies.

TABLE 15-1. Characteristics of Human Glycoprotein Hormones

The glycoprotein hormones bind to membrane receptors in their target tissues. Hormonal specificities are determined by the respective b subunits. However, hCG and LH have weak intrinsic TSH bioactivity. The isolated a and b subunits are not bioactive. However, both a -TSH and b-TSH subunits contain receptor-binding domains. Modifications of amino-acid residues or of the carbohydrate side chains can result in TSH analogs with enhanced bioactivity or altered metabolic clearance.2 TSH has a molecular mass of ~28 kDa. It is glycosylated at two sites on the a subunit and at one site on the b subunit. Glycosylation is required for subunit association, intracellular processing of the precursors to the secretory form of the hormone, and metabolic clearance of secreted hormone. Chemically deglycosylated TSH binds to its receptor but does not elicit a biologic response, suggesting that the glycosyl residues may have a role in receptor activation.3 The x-ray crystallographic structure of hCG is known,4 and a homologous structure for TSH has been proposed.5 The aand b subunits of the glycoprotein hormones share a common structural motif consisting of a central cystine knot formed by disulfide bonds, with two hairpin loops on one side of the knot and a single loop on the other. This cystine knot structure is found in growth factors and hormones, including platelet-derived growth factor, nerve growth factor, inhibins, and others.5 Another structural element presumed common to hCG, TSH, and the other glycoprotein hormones is a seat-belt region near the carboxyl terminus of the b subunit that wraps around the a subunit to maintain the heterodimeric structure. BIOSYNTHESIS OF THYROID-STIMULATING HORMONE Separate genes encode the a-TSH and b-TSH subunits.6 The human a gene is present in a single copy on chromosome 6 and is transcribed in all pituitary and placental cells synthesizing glycoprotein hormones. The human b-TSH gene is on chromosome 1. Hence, regulation of TSH gene transcription requires control of separate DNA regulatory elements for each subunit gene. Although all of the b subunits probably evolved from a common ancestor, b subunit genes are on separate chromosomes and do not form a single linkage group (see Table 15-1). The a-TSH and b-TSH apoprotein cores are transcribed as prehormones, starting with leader sequences of hydrophobic amino acids that direct the nascent chains through the rough endoplasmic reticulum membrane (see Chap. 3). Pituitary glands contain excess a subunit relative to their b subunit content, and free subunits as well as intact hormone are secreted by the thyrotrope. Free a subunit generally can be detected in serum from euthyroid persons. It is increased in persons with primary hypothyroidism; free b-TSH also can be detected. Free a subunit (secreted by gonadotropes) also is increased in postmenopausal women. Free a subunit contains an additional oligosaccharide group that prevents it from combining with b-TSH.5 Thyrotropin-releasing hormone (TRH) stimulates a-TSH and b-TSH subunit gene transcription by inducing or activating specific transcriptional regulatory factors.7,8 Triiodothyronine (T3) causes a rapid fall in transcription of the a-TSH and b-TSH genes.9 The synthesis of b-TSH appears to be the rate-limiting step and a major regulatory point in the control of TSH. Regulatory regions involved in T3 suppression of subunit gene transcription (negative T3 response elements, TREs) have been identified for both the a-TSH and b-TSH genes.6 These negative TREs contain nucleotide sequences for interaction with the b receptor.10 Glycosylation of the subunits begins cotranslationally, with the transfer of preassembled oligosaccharides to specific asparagine residues.3 Initial core glycosylation is required for the polypeptide chains to assume their tertiary structures and associate into heterodimers of a-TSH and b-TSH. Glycosylation of TSH also is under

hormonal control by TRH and thyroid hormone.3 TSH from patients with severe primary hypothyroidism and TSH released acutely by TRH stimulation differ in glycosylation compared to basal TSH.11 TSH extracted from sera of patients with nonthyroid illnesses also has alterations in glycosylation. Because alterations in glycosylation affect its biologic properties, control of TSH glycosylation may be of physiologic significance. THYROID-STIMULATING HORMONE RECEPTOR The TSH receptor is a member of the superfamily of guanine nucleotide regulatory protein (G protein)coupled receptors.12,13 Binding of TSH stimulates receptor interaction with the a subunit of the Gs protein. This leads to the release of guanosine diphosphate (GDP) from the a subunit and its replacement with guanosine triphosphate (GTP) as well as dissociation of the G protein into a and bg subunits that stimulate adenylate cyclase, increasing intracellular cyclic adenosine monophosphate (cAMP) and activating protein kinase A. At higher TSH concentrations, the receptor may interact with other G proteins, leading to modulation of other intracellular signaling pathways such as the Ca2+, phosphatidylinositol phosphate, protein kinase C cascade.14 The TSH receptor gene was cloned by probing thyroid complementary DNA (cDNA) libraries with oligonucleotide probes complementary to cDNA sequences coding for segments of the LH/hCG and FSH hormone receptors thought likely to be conserved in the TSH receptor.15 The TSH receptor gene, located on chromosome 14, contains 10 exons coding for a 764-amino-acid polypeptide (including a 21-amino-acid leader sequence). The apoprotein core of the TSH receptor has a molecular mass of 84.5 kDa.16 Functional recombinant receptor has been expressed in stably trans-fected Chinese hamster ovary cells and has been used for bioassay of TSH and of thyroid-stimulating immunoglobulin (TSI) activity. The TSH receptor and other G proteincoupled receptors contain three major structural and functional domains (Fig. 15-1) The extracellular region is the site of receptor-ligand interaction and also is involved in signal transduction. The membrane-spanning region contains seven hydrophobic transmembrane segments joined by short extracellular and cytoplasmic connecting loops. The transmembrane segments and connecting loops interact with the G-protein a subunit. The cytoplasmic tail contains phosphorylation sites that may be targets for protein kinases to regulate receptor activity. Binding of TSH to the extracellular domain results in a conformational change that alters the interaction of the membrane-spanning domain with the Gsasubunit, leading to activation of the GDP/GTP cycle.

FIGURE 15-1. Schematic illustration of the human thyrotropin receptor with the sites of mutations resulting in constitutive activation or loss of function.82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99 and 100 The initial 21 amino acids form the leader sequence and are cleaved during intracellular processing of the receptor. Germline activating mutations have been found in families with hereditary thyrotoxicosis and in neonates with nonautoimmune thyrotoxicosis. Somatic activating mutations have been found in solitary toxic nodules and toxic multinodular glands. (A database of thyroid-stimulating hormone receptor mutations is accessible through the internet at http://www.uni-leipzig.de/innere.)

The extracellular region of the TSH receptor contains 398 amino acids, including six potential glycosylation sites. A major structural feature of the extracellular domain of the glycoprotein receptors is the presence of leucine-rich repeats (LRRs). Proteins containing LRRs generally have functions involving interaction with polypeptide ligands.17 Based on the crystallo-graphic structure of ribonuclease inhibitor, an LRR-containing protein, a model of the extracellular domain of the TSH receptor has been proposed.18 The LRRs form a concave surface in the extracellular domain that can provide an extensive area for interaction with the hormone. This model is compatible with studies that have identified multiple regions in the TSH a and b subunits involved in hormone binding. BIOLOGIC ACTIONS OF THYROID-STIMULATING HORMONE Within the thyroid, TSH stimulates virtually all metabolic and cellular processes involved in the synthesis and secretion of thyroid hormones, including iodine uptake and organification, thyroglobulin synthesis, iodotyrosine coupling, colloid droplet formation, and iodothyronine secretion.19 TSH also stimulates intermediary metabolism as well as protein and nucleic acid synthesis and thyroid growth. Clinically observable effects of TSH on the normal thyroid gland include thyroid gland enlargement, increased radioactive iodine uptake, and increased secretion of thyroxine (T4) and T3.

CONTROL OF THYROID-STIMULATING HORMONE SECRETION: NEURAL AND HUMORAL MODULATORS

NEURAL CONTROL OF THYROID-STIMULATING HORMONE SECRETION The thyrotrope secretes TSH in response to humoral signals (Fig. 15-2). The hypothalamic neurohormones TRH, somatostatin, and dopamine are released under control of the central nervous system. Thyroid hormones feed back to suppress TSH release. Other hormones, including corticosteroids and cytokines, also can modulate TSH secretion.

FIGURE 15-2. Interactions of the major humoral mediators of thyroid-stimulating hormone (TSH) secretion. Solid arrows denote stimulatory effect; shaded arrows denote inhibitory effect. (DA, dopamine; TRH, thyrotropin-releasing hormone; T4, thyroxine; T3, triiodothyronine.)

Physiologic alterations in TSH secretion in response to factors such as circadian rhythm, cold exposure (in neonates and lower animals), and stress are controlled by the central nervous system through pathways that project to hypothalamic nuclei to modulate hypothalamic releasing hormone secretion. Numerous neurotransmitters, including biogenic amines, amino acids, endogenous opioids, and neuropeptides, have been found to modulate TSH secretion in animals.20,21 For example, in rats, the TSH response to cold exposure appears to be mediated by a-adrenergic pathways. THYROTROPIN-RELEASING HORMONE TRH, a tripeptide (pyroglutamyl-histidyl-proline amide), was one of the first hypothalamic releasing hormones to be isolated and characterized (Fig. 15-3). The mammalian TRH prohormone contains multiple copies of the sequence Gln-His-Pro-Gly, suggesting that several TRH molecules could be derived from each precursor.22 The highest density of TRH-containing cells in the central nervous system is found in the hypothalamic paraventricular nucleus.23 TRH also is distributed

extensively in the extrahypothalamic brain and in nonneuronal tissue, including the heart and testis.24 High-affinity TRH receptors are widely distributed; many behavioral, pharmacologic, and neurophysiologic effects have been reported after TRH administration. Hence, TRH may function as a neurotransmitter.21 However, probably only TRH released into the hypothalamic-pituitary portal circulation is involved in control of TSH secretion. Transcription of TRH is also under negative feedback regulation by T3. TREs have been identified in the regulatory region of the TRH gene.24

FIGURE 15-3. Chemical structure of thyrotropin-releasing hormone (TRH). TRH is the smallest identified hypophysiotropic hormone. Cyclization of the glutamate and amidation of the proline are required for full bioactivity.

The binding of TRH to its G-proteincoupled receptor on the thyrotrope stimulates hydrolysis of the membrane lipid phosphatidylinositol 4,5-bisphosphate to yield inositol 1,4,5-triphosphate and diacylglycerol25; each of these may function as an intracellular second messenger (see Chap. 4). Inositol triphosphate stimulates the mobilization of calcium from intracellular stores. Free calcium is bound rapidly by the calcium-binding protein calmodulin. The resulting complex activates kinases that phosphorylate proteins involved in the exocytosis of secretory granules. Diacylglycerol activates protein kinase C, leading to phosphorylation of other proteins that also may be needed for TSH secretion. TRH also stimulates adenylate cyclase in the pituitary. Experimental evidence for the role of TRH in the control of TSH includes the observation that hypothalamic lesions that reduce TRH content cause a decrease in TSH secretion and hypothyroidism. In addition, the administration of anti-TRH antiserum to neutralize endogenous TRH causes a decrease in basal and cold-stimulated TSH levels in normal and hypothyroid rats.21 TRH administration also stimulates prolactin release. The physiologic role of TRH in the normal control of prolactin secretion is unclear, however (see Chap. 13). Prolactin, but not TSH, is elevated in nursing women. The administration of anti-TRH antibody does not block the physiologic prolactin rise during pregnancy or suckling. Hyperprolactinemia and galactorrhea have been observed in primary hypothyroidism. Normally, TRH does not stimulate the secretion of other pituitary hormones. However, growth hormone release is stimulated by TRH administration in many patients with acromegaly and in some patients with renal failure, liver disease, anorexia nervosa, and depression. TRH stimulates the release of adrenocorticotropin in some patients with Cushing disease or Nelson syndrome, and the release of intact gonadotropin, a subunit, or free b subunit in patients with pituitary adenoma of gonadotrope origin.26 THYROID HORMONES Serum TSH levels are extremely sensitive to changes in circulating thyroid hormone concentrations. The administration of small doses of T4 or T3 to euthyroid persons suppresses basal TSH levels and blunts the TSH response to TRH. Conversely, the administration of iodide, which slightly decreases thyroid secretion of T4 and T3, causes increased basal TSH concentrations and enhanced TSH responses to TRH.27 An inverse loglinear relationship is seen between changes in TSH and alterations in free T4 or free T3 levels.28 Thyroid hormone suppression of TSH secretion is initiated by T3 binding to a nuclear receptor. T4-induced suppression of TSH is mediated by its intrapituitary deiodination to T3.29 Monodeiodinase activity is high in the pituitary, and 50% to 60% of intrapituitary T3 comes from local conversion of T4 to T3. Hence, the serum TSH concentration may be more dependent on serum T4 than on serum T3. In addition to inhibiting TSH biosynthesis, T3 administration also reduces the number of TRH receptors on thyrotropes and might thereby reduce their sensitivity to TRH.30 DOPAMINE AND SOMATOSTATIN The administration of dopamine, L -dopa, or bromocriptine to normal or hypothyroid individuals decreases their basal TSH concentrations and maximum TSH responses to TRH.21 In addition, the dopamine-receptor blocking agents, metoclopramide and domperidone, increase TSH concentrations in both euthyroid and hypothyroid persons, a finding that supports a physiologic role for endogenous dopamine secretion in suppressing TSH release. Because neither dopamine nor domperidone crosses the bloodbrain barrier, the effects of dopamine most likely are mediated directly at the thyrotrope. Somatostatin may be a physiologic inhibitor of TSH secretion. The infusion of somatostatin lowers TSH levels in hypothyroid patients and suppresses the normal nocturnal rise in TSH.21 In addition, the release of somatostatin from the hypothalamus is stimulated by T3.31 Finally, the administration of anti-somatostatin antiserum to rats causes an increase in basal serum TSH levels, and in cold- or TRH-stimulated TSH release. STEROID HORMONES Steroid hormones modulate TSH secretion. Corticosteroids inhibit basal and TRH-stimulated TSH levels and block the normal nocturnal surge in TSH. The effects of steroids on pituitary responsiveness to TRH may be mediated by modulation of TRH receptors on thyrotropes. INTEGRATED CONTROL OF THYROID-STIMULATING HORMONE SECRETION The secretion of TSH is controlled by the integrated thyrotrope response to humoral signals. The inhibitory effect of thyroid hormones on TSH secretion results in a closed-loop negative-feedback control system for the tight regulation of the concentration of thyroid hormone, which is sensed by the thyrotrope. A deviation of the thyroid hormone concentration from the setpoint of the control system alters TSH release, thereby causing a change in the secretion of thyroid hormones that returns their concentrations toward the setpoint. The setpoint of this thyrostat is established by stimulatory and inhibitory hypothalamic releasing hormones (TRH, somatostatin, dopamine) and other mediators that determine the sensitivity of the thyrotrope to inhibition by T3. Thus, differences in thyroid hormone levels among individuals, or differences in responsiveness to environmental or physiologic conditions, reflect differences in the thyrotrope set-points. Support for this model is provided by the observation that hypothalamic lesions to thyroidectomized rats blunt the expected increase in TSH and enhance the suppressive effects of low doses of thyroid hormones. Chronic intrathecal infusion of TRH in patients with amyotrophic lateral sclerosis produces a sustained rise in serum thyroid hormone and TSH levels, suggesting that the infusion raises the pituitary setpoint.32 In children with hypothyroidism resulting from idiopathic TRH deficiency, TSH release is more sensitive to inhibition by T4, suggesting that thyrotrope sensitivity to thyroid hormone suppression is increased in the absence of TRH.33

CONTROL OF THYROID-STIMULATING HORMONE SECRETION: PHYSIOLOGIC MODULATION

EFFECTS OF AGE AND SEX

TSH is first detectable in fetal serum at ~13 weeks' gestational age, approximately the same time as the onset of fetal thyroid iodine uptake. The TSH level remains low until the 18th to 20th week of gestation, when it rises abruptly. At birth, the serum TSH concentration is ~10 U/mL. It then rises rapidly, reaching levels of 75 to 150 U/mL by 30 minutes after birth.34 Levels then decline and are within the usual childhood range by 2 to 3 days after birth. The serum TSH concentration declines slightly during childhood and adolescence.35 The mean TSH concentration in euthyroid adults is 1.4 to 2.0 U/mL. Levels do not differ significantly between men and women. Serum TSH levels do not change between adolescence and the age of 60 years. A study of the 24-hour profile of TSH secretion using a sensitive TSH assay found that mean TSH levels were decreased in healthy older men compared to younger control subjects.36 The normal diurnal rhythm was preserved (see Chap. 199). DAILY RHYTHM TSH levels rise to a peak between midnight and the early morning hours; a nadir in TSH concentration occurs in late afternoon.21 TSH secretion is pulsatile, most likely in response to pulsatile release of TRH.37 Six to 10 major pulses per 24 hours are noted. The nocturnal rise in TSH levels is associated with increased TSH pulse amplitude. Alterations in dopaminergic and somatostatinergic tone may modulate the TSH pulse size. EFFECTS OF STRESS The serum TSH concentration is transiently depressed after stressful medical procedures such as treadmill exercise or gastroscopy, and for 1 to 2 days after elective surgery. These transient decreases may result from the inhibitory effects of increased serum cortisol. Although exposure to cold is an important stimulus for TSH release in some animals, temperature effects in humans are more limited. Exposure of neonates 3 hours after birth to a decreased ambient temperature causes an increase in serum TSH levels compared to those in age-matched control subjects.34 Elevated serum TSH levels also are noted during hypothermic cardiac surgery in infants.38 In adults, small and inconsistent effects on serum TSH values have been reported as a result of acute or chronic cold exposure.21 EFFECTS OF FASTING AND SEVERE ILLNESS Fasting is associated with decreased 5'-monodeiodinase activity, which causes a fall in extrathyroidal conversion of T4 to T3 and an increase in reverse T3 (see Chap. 36). Because similar changes have been observed in patients with numerous acute or chronic illnesses, fasting has been used as a model for the effects of illness on thyroid function. Despite the low T3 concentration, basal TSH levels are either unchanged or decreased, and the TSH response to TRH generally is impaired. Thyrotrope responsiveness to thyroid hormones is intact because the administration of T3 during a fast results in further suppression of TSH, whereas the administration of iodide (to reduce circulating thyroid hormone concentrations) restores normal TSH responsiveness to TRH.39 Basal TSH levels generally have been reported to be normal in euthyroid persons during mild or moderate illness. However, the pituitary-hypothalamic regulation of TSH may be impaired in some severely ill patients. In a study of patients who became severely ill while undergoing bone marrow transplantation, serum TSH decreased to subnormal or undetectable levels in most patients in whom serum T4 values declined.40 The decrease in TSH generally preceded the decline in T4. With recovery, serum T4 and TSH levels returned to normal. These findings suggest that the setpoint of circulating thyroid hormone is decreased during fasting and severe illness. This could be the result of altered secretion of hypothalamic-releasing hormones, such as TRH or somatostatin. The post-mortem hypothalamic content of TRH mRNA, measured by in situ hybridization, has been reported to be reduced in severely ill individuals.23 Infusion of the cytokines interleukin-1 (IL-1) or tumor necrosis factor-a in rats alters TSH levels and thyroid cell function.41 Also, administration of IL-1b to rats causes a decrease in the hypothalamic content of TRH mRNA.42 These immunomodulators may play a role in the hypothalamic and pituitary responses to stress and illness. During recovery from severe illness, TSH levels may be transiently elevated.43

CLINICAL APPLICATIONS OF THYROID-STIMULATING HORMONE MEASUREMENT

THYROID-STIMULATING HORMONE ASSAY The radioimmunoassays initially used for routine clinical measurement of TSH have been replaced by more sensitive two-site noncompetitive immunometric assays (IMAs). These assays use two antibodies with specificity toward separate epitopes on TSH. One antibody is attached to a solid-phase support and the other antibody is labeled. Enzymes, luminescent or fluorescent compounds, and iodine-125 have been used as antibody labels.28,44 Luminescent labels now are used most commonly. Binding of the two antibodies to TSH results in the formation of a labeled sandwich that is separated from the noncomplexed reagents and measured. IMAs are specific for the measurement of TSH and generally free from interference. Rarely, falsely elevated serum TSH levels have been reported because of the presence in serum of heterophilic antibodies that neutralize the reagent anti-TSH antibody.45 With the introduction of the IMAs, a confusing nomenclature has developed based on the claimed sensitivity (e.g., highly sensitive, ultrasensitive, supersensitive). The American Thyroid Association has proposed that assays should be characterized by a functional criterion and that a TSH assay should be designated as sensitive only if sera from thyrotoxic persons yield results >3 log standard deviations below the mean of normal euthyroid persons.46 More than 95% of sera from thyrotoxic persons would be expected to have TSH levels below the lower limit of normal in an assay meeting this criterion. Most commercial IMAs appear capable of meeting that standard.47 Clinical chemists have traditionally reported as the analytic sensitivity or detection limit of an assay the lowest TSH level statistically distinguishable from zero concentration by measurement of replicate samples in the same assay run. Such a definition of sensitivity may be clinically misleading, because a single measurement of a specimen containing TSH at the analytic threshold concentration would yield a result of zero 50% of the time. Furthermore, analytic sensitivity is a function of within-assay variance and does not assess the reliability of between-assay comparisons, which are more likely to be clinically useful in the diagnosis and treatment of an individual patient. As an alternative to analytic sensitivity, the proposal has been made that assay sensitivity be characterized by a criterion based on interassay variability. Specifically, the lower limit of interassay quantitative measurement,48 or functional sensitivity, of an assay is the TSH concentration for which the interassay coefficient of variation is less than some preestablished threshold (generally 20%) to permit reliable quantitative comparisons between specimens measured in different assay runs. A generational classification of TSH assays has been proposed49 (Table 15-2). Each generation is approximately an order of magnitude more sensitive than the previous one. Both second- and third-generation assays distinguish suppressed TSH levels in hyperthyroidism from normal values. However, the third-generation assay further differentiates partial suppression of basal TSH concentrations in some patients with subclinical hyperthyroidism, nonthyroid illness, glucocorticoid therapy, and other clinical states (Table 15-3) from the more complete suppression of basal TSH concentrations in overt hyperthyroidism.50

TABLE 15-2. Properties of Thyroid-Stimulating Hormone (TSH) Assays

TABLE 15-3. Clinical Influence on Basal Thyroid-Stimulating Hormone (TSH) and TSH Response to Thyrotropin-Releasing Hormone

Measurement of TSH is frequently used as the initial, and sometimes sole, thyroid function test.51 This approach is generally sensitive and specific in the ambulatory population, in which the finding of a normal TSH level is strong evidence that a patient is euthyroid, and an abnormal TSH has a high likelihood of being due to thyroid dysfunction. However, abnormally high or low TSH values (compared to those of an ambulatory euthyroid control population) are frequently noted in hospitalized patients as a result of the effects of nonthyroid illness, acute psychiatric illness, or glucocorticoid therapy.52,53 Therefore, diagnoses of hypothyroidism or hyperthyroidism in hospitalized patients should be based on clinical evaluation, measurement of free thyroid hormone levels, and other indices of thyroid function, rather than on TSH measurement alone. The TRH stimulation test has been used in the assessment of mild thyroid dysfunction and in the functional evaluation of the hypothalamicpituitarythyroid axis.54 The test entails measuring serum TSH levels at baseline and after the bolus intravenous administration of TRH. A dose-response relation between administered TRH and peak TSH levels is observed for TRH doses of 6.25 to 400 g. In clinical practice, a TRH dose sufficient to produce a maximal TSH response is used. The peak TSH response occurs 20 to 40 minutes after TRH administration. If a primary thyroid disorder is suspected, measurement of TSH levels at baseline and at 20 or 30 minutes after TRH administration is sufficient (Fig. 15-4). If pituitary or hypothalamic dysfunction is suspected, TSH measurements should be continued for 2 to 3 hours at 30- to 60-minute intervals. During the first 5 minutes after TRH administration, side effects may include mild nausea, headache, a transient rise in blood pressure, light-headedness, a peculiar taste sensation, a flushed feeling, and urinary urgency.54

FIGURE 15-4. Typical thyroid-stimulating hormone (TSH) responses to the administration of thyrotropin-releasing hormone (TRH) under different conditions. Basal TSH is suppressed in overt thyrotoxicosis and does not respond to TRH. The blunted TSH response in patients with nonthyroid illness may be similar to the response in patients with subclinical hyperthyroidism. Patients with subclinical hyperthyroidism or nonthyroid illness may have a basal TSH below the detection threshold for second-generation TSH assays.

TRH testing may be viewed as a means of amplifying and detecting small differences in TSH secretion due, most importantly, to alterations in serum T4 or T3 concentrations. A slight excess of T4 or T3 blunts or completely blocks the TSH response to TRH, whereas small decrements in thyroid hormone levels enhance the response. The peak TSH response to TRH is proportional to the basal serum TSH level.55 Expressed as a multiple of the basal TSH, the peak TSH is a mean 8 to 9.5 times higher. However, considerable variability is seen in individual responses (range: 3- to 23-fold increment in euthyroid persons). In addition to thyroid hormone concentrations, other factors can alter the TSH response to TRH (see Table 15-3). In patients with severe illnesses, the TSH response to TRH is likely to be diminished. Cortisol and other neurohumoral factors secreted in response to stress, malnutrition, and the administration of glucocorticoids or dopamine all may contribute to the blunted TSH response. In patients with a subnormal basal TSH level, however, the magnitude of the TSH response to TRH does not distinguish those in whom TSH is suppressed as a result of intercurrent illness from those in whom it is suppressed as a result of partial suppression of the pituitary by slight excess of free thyroid hormone (i.e., patients with autonomous thyroid nodules or exogenous thyroid hormone suppression).55 In general, if the basal TSH level exceeds the functional sensitivity threshold of the assay system and, therefore, can be accurately measured, then measurement of the TRH-stimulated TSH level does not provide additional information regarding the cause of the suppressed TSH. With the improvement in sensitivity of TSH assays, the TRH-stimulation test is not generally required in the evaluation of primary hypothyroidism or hyperthyroidism with suppressed TSH. However, the TRH-stimulation test may be useful in the evaluation of central hypothyroidism, in the rare patient with TSH-dependent hyperthyroidism, and in some patients with functioning pituitary tumors that respond to TRH stimulation (e.g., acromegaly). THYROID-STIMULATING HORMONE IN PRIMARY HYPOTHYROIDISM The basal serum TSH concentration is increased in patients with intrinsic failure of the thyroid gland (primary hypothyroidism) of all causes. The magnitude of the increase is roughly proportional to the severity of disease.56 In general, basal TSH levels show a better inverse correlation with serum T4 levels than with serum T3 levels; this is because of the importance of the uptake of serum T4 and its intracellular deiodination as a source of T3 in the thyrotrope.29 In some persons, elevated TSH levels may be found with normal serum T3 concentrations but decreased serum T4 values. Such findings are common in patients with early thyroid gland failure, patients with mild iodine deficiency, and some patients with Graves disease who have been given long-term antithyroid drug treatment. The isolated elevation of serum TSH levels with normal serum T4 and T3 concentrations in the absence of clinical signs or symptoms of hypothyroidism has been termed subclinical hypothyroidism. This condition has an overall prevalence of 2% to 7% and is particularly common in older women. Overt hypothyroidism develops at a rate of 5% to 10% per year in persons with elevated TSH levels and positive antithyroid antibodies.57 Measurement of the serum TSH level remains a sensitive test for the diagnosis of primary hypothyroidism in severely ill patients because basal TSH levels, although sometimes partially attenuated, remain higher than normal during intercurrent illness in patients with moderate or severe hypothyroidism. The diagnosis of hypothyroidism should be confirmed by measurement of the free T4 value, however. Patients with mildly elevated TSH and normal free T4 concentrations generally should undergo repeated thyroid function testing after discharge from the hospital to confirm the diagnosis of hypothyroidism. The TSH response to TRH is exaggerated in patients with primary hypothyroidism. However, TRH testing should not be needed in the evaluation of suspected hypothyroidism if the basal serum TSH level is elevated. THYROID-STIMULATING HORMONE IN HYPERTHYROIDISM Circulating TSH is suppressed in hyperthyroidism of all causes except in the rare patients with TSH-dependent thyrotoxicosis. In clinically hyperthyroid patients, the basal TSH level measured with a third-generation assay (functional sensitivity of <0.01 U/mL) is <0.01 U/mL; in most patients, TSH is undetectable.28,58 Serum TSH

generally remains undetectable after TRH stimulation in these patients. Slight overproduction of T4 and T3 in patients with autonomous nodular goiter or mild Graves disease may suppress TSH values without raising free T4 or free T3 levels above the normal range (subclinical hyperthyroidism). Suppression of the hypothalamic-pituitary axis may be less complete in patients with subclinical hyperthyroidism. In some, the basal TSH level is undetectable, whereas in others, it is detectable but suppressed below the normal range for ambulatory persons. If the basal TSH value is detectable, a TSH increment in response to TRH also is measurable. If the basal TSH level is undetectable, a slight increment after TRH stimulation sometimes may be noted.28,58 A second-generation TSH assay distinguishes suppressed TSH in hyperthyroidism from euthyroid levels. However, patients with partial TSH suppression also may have undetectably low TSH levels with this assay (see Fig. 15-4). If more precise measurement of thyrotrope suppression is required (i.e., to distinguish partial TSH suppression due to nonthyroid illness from thyrotoxicosis in hospitalized patients), measurement with a more sensitive assay or TRH-stimulation testing may be helpful. EVALUATION OF THERAPY FOR THYROID DISEASES The adequacy of thyroid hormone replacement therapy in primary hypothyroidism can be evaluated by measuring the serum TSH concentration. Normalization of serum TSH levels should be sought as a therapeutic end point in adults.59 The threshold for suppression of TSH has been reported to be elevated in infants with congenital hypothyroidism, who may show a persistently elevated TSH level in spite of the administration of adequate doses of replacement T4.60 After the return of free T4 from elevated to normal levels, the thyrotrope may remain suppressed for 4 to 6 weeks.61 Therefore, TSH is not an accurate indicator of thyroid status in thyrotoxic patients whose thyroid hormone levels are changing. Free T4 (and free T3) should be measured in the short-term follow-up of hyperthyroid patients being treated with antithyroid drugs or during the immediate period after treatment with radioactive iodine. TSH should be measured by a third-generation assay to confirm complete suppression by exogenous thyroid hormone in patients with thyroid cancer. If a second-generation assay is used, then TSH should be measured after TRH stimulation to determine the degree of suppression.28,58

DISORDERS OF THYROID-STIMULATING HORMONE CONTROL

DISORDERS CAUSING HYPOTHYROIDISM CENTRAL HYPOTHYROIDISM Central hypothyroidism results from failure of the pituitary to secrete biologically active TSH. Central hypothyroidism should be considered in patients who have clinical features of hypothyroidism or hypothalamic-pituitary dysfunction, low serum T4 concentrations, and serum TSH levels that are not elevated. In patients with no clinical manifestations of myxedema and no other evidence of hypopituitarism, other causes of low serum T4 without elevated serum TSH also should be considered, such as decreased thyroid hormonebinding proteins or severe illness. Patients with central hypothyroidism constitute several groups (Table 15-4), as described below.

TABLE 15-4. Causes of Central Hypothyroidism

Combined Pituitary Hormone Deficiency (Idiopathic Hypopituitarism). Patients with combined pituitary hormone deficiency (CPHD) have no antecedent histories of disease or injury that could cause hypopituitarism. Clinical features vary. Cases may be sporadic or familial in occurrence. Hypothyroidism alone is rare; patients generally present in childhood with growth hormone deficiency. Patients then go on to develop varying degrees of TSH, prolactin, and gonadotropin deficiencies. However, the patients do not have diabetes insipidus or neurologic deficits. In some families with CPHD, mutations in pituitary-specific transcriptional factors (Pit-1, Prop-1) have been identified (Table 15-5).62,63 and 64

TABLE 15-5. Identified Molecular Defects in Thyroid-Stimulating Hormone (TSH) and TSH Receptor Function

Isolated Thyroid-Stimulating Hormone Deficiency. In patients with isolated thyroid-stimulating hormone deficiency, basal serum TSH values may be low, normal, or slightly elevated; TRH stimulation generally causes a rise in serum TSH levels. The chronic administration of TRH can restore thyroid hormones to normal levels. Hence, the hypothyroidism is thought to result from impaired release of TRH from the hypothalamus. Rarely, familial cases with mutations in the b-TSH gene (see Table 15-5) have been described. Mutations have included a truncated b-TSH transcript as a result of a premature stop codon; a single amino-acid substitution resulted in a b-TSH that was unable to heterodimerize, and a frame-shift mutation resulted in reduced amounts of TSH with decreased bioactivity.65,66 In another patient, loss-of-function mutations of the TRH receptor were reported (see Table 15-5).67 Hypothalamic Lesions. Destructive lesions of the hypothalamus may result in central hypothyroidism. Generally, multiple pituitary hormones are involved, and patients often have diabetes insipidus. Serum prolactin levels may be mildly elevated because of interruption of the tonic dopaminergic lactotrope inhibitory pathway. In addition to hormone deficiencies, patients may have neurologic abnormalities and other manifestations of hypothalamic disease, such as disturbances of autonomic function, temperature regulation, food and water intake, and sleep cycle, as well as emotional lability. TSH with reduced bioactivity has been found in the serum of patients with hypothalamic hypothyroidism; long-term treatment with TRH increases the bioactivity of their TSH.68 These findings might explain why some patients with hypothalamic disorders have hypothyroidism despite showing normal or slightly increased levels of immunoreactive TSH. Pituitary Lesions. Pituitary lesions also cause hypothyroidism. Hypothalamic disturbances, visual field cuts, and parasellar abnormalities may be observed if an

extrasellar extension of the lesion is present. Pituitary adenomas rarely cause hypothyroidism if they have not grown large enough to cause distortion and enlargement of the margins of the sella turcica. 69 THYROID-STIMULATING HORMONE AND THYROTROPIN-RELEASING HORMONE TESTING IN HYPOTHALAMIC-PITUITARY DISEASE Basal TSH levels may be normal, low, or slightly elevated in patients with central hypothyroidism. Basal TSH values do not correlate with free T4 levels, and TRH-stimulated TSH values are not proportional to basal TSH values.55 Therefore, clinical evaluation and measurement of free T4 should be used in confirming the diagnosis of suspected central hypothyroidism and in titrating thyroid hormone replacement therapy in patients with central hypothyroidism. Initially, the proposal was made that TRH testing should differentiate pituitary from hypothalamic causes of central hypothyroidism, with pituitary dysfunction causing a blunted or absent TSH response and hypothalamic disease resulting in a normal or exaggerated TSH rise that is delayed or prolonged. Although such classic patterns are found in many patients, the responses often overlap. Normal or hypothalamic patterns have been observed in patients with pituitary lesions, and the serum TSH response has been flat in some patients with suprasellar disease.70 Thus, the anatomic site of the central lesion may not always correlate with the expected functional consequences on TRH responsiveness. Abnormal serum TSH responses to TRH (i.e., blunted or delayed peaks) also are noted frequently in patients with pituitary or hypothalamic diseases even when the patients appear clinically euthyroid and their serum thyroid hormone levels are normal.71 DISORDERS CAUSING HYPERTHYROIDISM THYROID-STIMULATING HORMONEINDUCED HYPERTHYROIDISM Hyperthyroidism induced by TSH has been described in a small group of thyrotoxic patients in whom serum TSH levels are normal or increased (see Chap. 42). If signs and symptoms of thyrotoxicosis are absent in a patient with elevated thyroid hormone concentrations and normal serum TSH levels, other causes of hyperthyroxinemia with nonsuppressed TSH also should be considered (Table 15-6). Patients with TSH-induced hyperthyroidism have clinical evidence of a stimulated thyroid gland, including diffuse goiter and elevated radioactive iodine uptake. Extrathyroidal manifestations of Graves disease are lacking, and the assay of TSI is negative.72 Two groups of abnormalities have been described.

TABLE 15-6. Causes of Hyperthyroxinemia with Nonsuppressed Thyroid-Stimulating Hormone (TSH)

Thyroid-Stimulating HormoneProducing Pituitary Tumors. In patients with TSH-producing pituitary tumors, basal intact TSH levels are elevated but are not stimulated by TRH or suppressed by exogenous thyroid hormone. Free b-TSH is undetectable; however, free a subunit levels are elevated and the molar ratio of free a subunit to intact TSH is >1.73,74 Computed tomography or magnetic resonance imaging generally reveals a pituitary lesion. However, these patients are best distinguished from those with nonneoplastic hypersecretion of TSH by measurement of the ratio of free a subunit to intact TSH and the response of TSH to TRH. TSH-producing pituitary tumors may cosecrete growth hormone or prolactin. The initial treatment of TSH-producing pituitary tumors is surgical resection, sometimes followed by postoperative irradiation. If hyperthyroidism persists, the long-acting somatostatin analog octreotide acetate has been shown to decrease serum TSH levels and cause reduction in tumor size.73,75 Finally, thyroid ablation can be used for persistent thyrotoxicosis. Nonneoplastic Hypersecretion of Thyroid-Stimulating Hormone. TSH-dependent hyperthyroidism has been described in the absence of a TSH-secreting pituitary tumor. These patients have selective pituitary resistance to thyroid hormone (PRTH), a variant of the syndrome of generalized resistance to thyroid hormone (GRTH).76,77 The thyroid hormone resistance syndromes are inherited in an autosomal dominant fashion. Considerable clinical overlap is found between GRTH and PRTH. The free thyroid hormone levels are elevated in both. Clinical manifestations of hyperthyroidism in PRTH are variable. Laboratory findings in PRTH and GRTH are similar and are distinguished from the findings in subjects with TSH-secreting pituitary tumors. Basal serum TSH levels are normal or slightly elevated and are stimulated by TRH. The administration of T3 suppresses basal TSH levels and blunts the TSH response to TRH. Serum free a subunit levels are normal, and the molar ratio of free a subunit to intact TSH is <1. Thus, TSH secretory dynamics are qualitatively normal. The thyroid hormone resistance syndromes are the result of mutations in the T3 nuclear receptor (TR). Similar mutations in the TRb gene have been described in both GRTH and PRTH.74 Hence GRTH and PRTH are alternate phenotypic expressions of a common underlying genotype. The mechanism responsible for this phenotypic variation has not yet been established. Because both the normal and mutant receptor genes may be expressed, the dominant receptor isoform (normal or mutant) may vary in different organs and in different individuals. The treatment of nonneoplastic TSH-dependent hyperthyroidism is difficult. Antithyroid drug therapy may increase TSH levels further, causing enlargement of the goiter and possibly enhancing abnormal thyrotrope growth. Thyroxine administration may suppress the TSH but worsens symptoms of thyrotoxicosis. The pharmacologic suppression of TSH secretion has been attempted. Some patients have responded to bromocriptine or octreotide. Other patients have been treated successfully with 3,5,3'-triiodothyroacetic acid (TRIAC) or with D-thyroxine.74 PARANEOPLASTIC THYROID-STIMULATING HORMONE PRODUCTION A convincing case of ectopic production of TSH by a nonpituitary tumor has not been reported. A patient with hepatocellular carcinoma, thyrotoxicosis, and nonsuppressed TSH has been described.78 Neither TSH nor TRH immunoreactivity could be detected in the tumor, and the hypothesis was made that the tumor was producing a factor with TRH bioactivity. HYPERTHYROIDISM IN TROPHOBLASTIC DISEASE Hyperthyroidism has been observed in patients with trophoblastic neoplasms such as hydatidiform mole and choriocarcinoma.79 Research has found that the thyrotropic activity isolated from hydatidiform moles copurifies with hCG and that highly purified hCG has weak intrinsic thyrotropic bioactivity75 (see Chap. 42, Chap. 112 and Chap. 219). Because hCG concentration may be high in patients with trophoblastic tumors, even weak intrinsic thyrotropic activity could stimulate the thyroid enough to cause thyrotoxicosis. The TSH bioactivity of hCG is enhanced by the removal of amino acids from its carboxyl terminus and by the removal of sialic acid residues from its carbohydrate side chains.80 Hence, the proposal has been made that the thyroid stimulator in hyperthyroidism resulting from trophoblastic disease may be a structurally variant hCG, either transcribed from an abnormal hCG gene or altered by post-translational modifications. DEFECTS IN THYROID-STIMULATING HORMONE RECEPTOR FUNCTION Since the cloning of the human TSH receptor, a number of mutations and polymorphisms have been reported (see Table 15-5).81,82,83 and 84 Several groups of patients have been described. AUTONOMOUSLY FUNCTIONING THYROID NODULES Mutations to the TSH receptor or to the Gs protein that result in constitutive activation of the cAMP cascade occur in some autonomously functioning thyroid nodules

(AFTNs).85,86,87,88,89,90 and 91 TSH receptoractivating mutations have been reported in 8% to 82% of solitary AFTNs as well as in autonomous nodules within multinodular thyroid glands. These mutations are found only in the autonomous nodule and not in adjacent normal thyroid tissue or in peripheral cells; hence, they are nongermline somatic mutations. Activating somatic mutations to Gsa also have been reported in 0% to 38% of AFTNs. In addition, hyperthyroidism is frequently noted in patients with McCune-Albright syndrome who are mosaics for somatic Gsa-activating mutations acquired during embryogenesis. Not all AFTNs are the result of mutations to these components of the cAMP cascade, however, because autonomous nodules without mutations to either protein are also encountered frequently. Transfected cells expressing the mutant receptor or Gs protein have elevated basal levels of cAMP compared to cells transfected with the wild-type receptor or Gs. Stimulation by TSH generally causes a further increment in cAMP levels. The receptor-activating mutations are predominantly localized to the membrane-spanning region of the receptor within the trans-membrane segments and connecting loops that interact with Gsa (see Fig. 15-1). Presumably, as a result of the conformational change in the receptor caused by the mutation, it is more effective in constitutively activating Gsa. FAMILIAL NONAUTOIMMUNE HYPERTHYROIDISM Hereditary thyrotoxicosis with diffuse hyperplasia of the thyroid has been described in several families.92,93,94,95,96 and 97 Inheritance of the syndrome is autosomal dominant. It is distinct from Graves disease because no markers of autoimmunity such as TSI, lymphocytic infiltrates, or ophthalmopathy are found. Clinical characteristics vary. Individuals generally present during infancy or early childhood, but some have been young adults at the time of diagnosis; hyperthyroidism ranges from mild to severe. Goiter may not be present initially but tends to grow over time. Hyperthyroidism generally recurs if the patient is treated with subtotal thyroidectomy or with less than complete radioiodine ablation. TSH receptoractivating mutations have been identified and characterized in these patients. The mutations are located in the same regions of the TSH receptor as the activating mutations in AFTNs (see Fig. 15-1); in some cases they are identical. However, the constitutive activity of the TSH receptor in this familial syndrome is generally less than the constitutive activity of the mutated receptors in AFTNs. Severe congenital activating mutations may have a high likelihood of being lethal in utero. In addition to the familial cases, several sporadic cases of congenital hyperthyroidism with activating mutations to the TSH receptor have been reported. These apparently are de novo germline mutations, because neither parent carried the mutated gene. Constitutive activating mutations of G proteincoupled receptors also have been described in other clinical syndromes, including familial male precocious puberty (LH receptor), continued spermatogenesis after hypophysectomy (FSH receptor), autosomal dominant hypoparathyroidism (calcium receptor), and in a rare form of dwarfism associated with PTH-independent hypercalcemia (PTH receptor).81 RESISTANCE TO THYROID-STIMULATING HORMONE Several cases of TSH resistance as a result of TSH receptor loss-of-function mutations have been reported (see Table 15-5).98,99 and 100 All patients have elevated TSH levels; in most, the radioactive iodine uptake and the free T4 are normal or only slightly low, consistent with partial TSH unresponsiveness compensated by the elevated TSH. Cases of severe congenital hypothyroidism with hypoplasia of the thyroid also occur, however. Inheritance is autosomal recessive; affected individuals are compound heterozygous or homozygous for TSH receptor mutations. Heterozygotes are unaffected. TSH-stimulated cAMP production is reduced or absent in cells transfected with the mutant receptor. Several mechanisms may cause the loss-of-function mutation. Some mutant TSH receptors cannot be detected in the plasma membrane. In some of these mutants, a premature stop codon results in production of a truncated polypeptide; in others an amino acid substitution or deletion may lead to abnormal intracellular processing of the mutant TSH receptor and its consequent failure to appear on the cell surface. A second group of loss-of-function mutations exhibits decreased TSH binding as a result of mutations in the extracellular TSH-binding domain. Finally, in a third group of mutations, TSH binding is normal or near normal, but the cAMP response is impaired. These mutations, which have been localized within the extracellular and membrane-spanning domains (see Fig. 15-1), presumably result in defective coupling to Gsa. TSH resistance occurs in patients with pseudohypoparathyroidism type 1a, who have Albright hereditary osteodystrophy (brachydactyly, subcutaneous ossifications, short stature, round face; see Chap. 60). These patients are heterozygous for loss-of-function mutations in the Gsa gene that result in partial loss of Gsa activity. They have variable patterns of hormone dysfunction involving end-organ responses to PTH, TSH, and gonadotropins in addition to the skeletal deformities.101

CLINICAL USE OF RECOMBINANT HUMAN THYROID-STIMULATING HORMONE

Recombinant human TSH has been produced by cotransfecting Chinese hamster ovary cells with cDNA coding for a and b-TSH.102 Glycosylation of recombinant human TSH is not identical to that of native human TSH. It is bioactive, however, and has undergone initial clinical trials in patients with thyroid cancer.103 Results of whole body radioiodine scans in patients who had undergone thyroidectomy were evaluated after patients were treated with recombinant human TSH to stimulate radio-iodine uptake, and were compared to scans obtained in the same patients after withdrawal of T3 suppression and stimulation by endogenous TSH. In most patients, visualization of residual thyroid activity and metastatic uptake was comparable after both stimulation procedures.104 CHAPTER REFERENCES
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. Pierce JG, Parsons TF. Glycoprotein hormones: structure and function. Annu Rev Biochem 1981; 50:465. Szkudlinski MW, The NG, Grossmann M, et al. Engineering human glycoprotein hormone superactive analogues. Nature Biotech 1996; 14:1257. Magner JA. Thyroid-stimulating hormone: biosynthesis, cell biology and bioactivity. Endocr Rev 1990; 11:354. Lapthom AJ, Harris DC, Littlejohn A, et al. Crystal structure of human chorionic gonadotropin. Nature 1994; 369:455. Grossmann M, Weintraub BD, Szkudlinski MW. Novel insights into the molecular mechanisms of human thyrotropin action: structural, physiological, and therapeutic implications for the glycoprotein hormone family. Endocr Rev 1997; 18:476. Chin WW, Can FE, Burnside J, Darling DS. Thyroid hormone regulation of thyrotropin gene expression. Recent Prog Horm Res 1993; 48:393. Kim MK, McClaskey JH, Bodenner DL, Weintraub BD. An AP-1-like factor and the pituitary-specific factor Pit-1 are both necessary to mediate hormonal induction of human thyrotropin b gene expression. J Biol Chem 1993; 268:23366. Steinfelder HJ, Wondisford FE. Thyrotropin (TSH) b-subunit gene expressionan example for the complex regulation of pituitary hormone genes. Exp Clin Endocrinol Diabetes 1997; 105:196. Shupnik MA, Ridgway EC. Triiodothyronine rapidly decreases transcription of the thyrotropin subunit genes in thyrotropic tumor explants. Endocrinology 1985; 117:1940. Can FE, Wong NCW. Characteristics of a negative thyroid hormone response element. J Biol Chem 1994; 269:4175. Magner JA, Kane J, Chou ET. Intravenous thyrotropin (TSH)-releasing hormone releases human TSH that is structurally different from basal TSH. J Clin Endocrinol Metab 1992; 74:1306. Boume HR, Sanders DA, McCormick F. The GTPase superfamily: a conserved switch for diverse cell functions. Nature 1990; 348:125. Ji TH, Grossmann M, Ji I. G protein-coupled receptors. I. Diversity of receptor-ligand interactions. J Biol Chem 1998; 273:17299. Laugwitz KL, Allgeier A, Offermanns S, et al. The human thyrotropin receptor: a heptahelical receptor capable of stimulating members of all four G protein families. Proc Natl Acad Sci U S A 1996; 93:116. Parmentier M, Libert F, Maenhaut C, et al. Molecular cloning of the thyrotropin receptor. Science 1989; 246:1620. Vassart G, Dumont JE. The thyrotropin receptor and the regulation of thyrocyte function and growth. Endocr Rev 1992; 13:596. Kobe B, Deisenhofer J. A structural basis of the interactions between leucine-rich repeats and protein ligands. Nature 1995; 374:183. Jiang X, Dreano M, Buckler DR, et al. Structural predictions for the ligand-binding region of glycoprotein hormone receptors and the nature of hormone-receptor interactions. Structure 1995; 3:1341. Tong W. Actions of thyroid-stimulating hormone. In: Greep RO, Astwood EB, eds. Handbook of physiology, section 7, Endocrinology; vol III, Thyroid. Baltimore: Williams & Wilkins, 1974:255. Toni R, Lechan RM. Neuroendocrine regulation of thyrotropin-releasing hormone (TRH) in the tuberoinfundibular system. J Endocrinol Invest 1993; 16:715. Morley JE. Neuroendocrine control of thyrotropin secretion. Endocr Rev 1981; 2:396. Lechan RM, Wu P, Jackson IMD, et al. Thyrotropin-releasing hormone precursor: characterization in rat brain. Science 1986; 231:159. Fliers E, Wiersinga WM, Swaab DF. Physiological and pathophysiological aspects of thyrotropin-releasing hormone gene expression in the human hypothalamus. Thyroid 1998; 8:921. Wilber JF, Xu AH. The thyrotropin-releasing hormone gene cloning, characterization, and transcriptional regulation in the central nervous system, heart and testis. Thyroid 1998; 8:897. Gershengorn MC. Mechanism of thyrotropin releasing hormone stimulation of pituitary hormone secretion. Annu Rev Physiol 1986; 48:515. Daneshdoost L, Gennarelli TA, Bashey HM, et al. Recognition of gonadotroph adenomas in women. N Engl J Med 1991; 324:589. Vagenakis AG, Rapoport B, Azizi F, et al. Hyperresponse to thyrotropin-releasing hormone accompanying small decreases in serum thyroid hormone concentrations. J Clin Invest 1974; 54:913. Spencer CA, LoPresti JS, Patel A, et al. Applications of a new chemiluminometric thyrotropin assay to subnormal measurement. J Clin Endocrinol Metab 1990; 70:453. Larsen PR. Thyroid-pituitary interaction. N Engl J Med 1982; 306:23. Gershengorn MC. Thyrotropin-releasing hormone receptor: cloning and regulation of its expression. Recent Prog Horm Res 1993; 48:341 Berelowitz M, Maeda K, Harris S, Frohman LA. The effect of alterations in the pituitary-thyroid axis on hypothalamic content and in vitro release of somatostatin-like immunoreactivity. Endocrinology 1980; 107:24. Kaplan MM, Taft JA, Reichlin S. Munsat TL. Sustained rises in serum thyrotropin, thyroxine, and triiodothyronine during long term, continuous thyrotropin-releasing hormone treatment in patients with amyotrophic lateral sclerosis. J Clin Endocrinol Metab 1986; 63:808. Sato T, Ishiguro K, Suzuki Y, et al. Low setting of feedback regulation of TSH secretion by thyroxine in pituitary dwarfism with TSH-releasing hormone deficiency. J Clin Endocrinol Metab 1976; 42:385. Fisher DA, Odell WD. Acute release of thyrotropin in the newborn. J Clin Invest 1969; 48:1670. Penny R, Spencer CA, Frasier SD, Nicoloff JT. Thyroid-stimulating hormone and thyroglobulin levels decrease with chronological age in children and adolescents. J Clin Endocrinol Metab 1983; 56:177. Van Coevorden A, Laurent E, Decoster C, et al. Decreased basal and stimulated thyrotropin secretion in healthy elderly men. J Clin Endocrinol Metab 1989; 69:177. Brabant G, Prank K, Ranft U, et al. Physiological regulation of circadian and pulsatile thyrotropin secretion in normal men and women. J Clin Endocrinol Metab 1990; 70:403. Wilber JF, Baum D. Elevation of plasma TSH during surgical hypothermia. J Clin Endocrinol Metab 1970; 31:372. Gardner DF, Kaplan MM, Stanley CA, Utiger RD. Effect of tri-iodothyronine replacement on the metabolic and pituitary response to starvation. N Engl J Med 1979; 300:579.

40. Wehmann RE, Gregerman RI, Burns WH, et al. Suppression of thyrotropin in the low-thyroxine state of severe nonthyroidal illness. N Engl J Med 1985; 312:546. 41. Pang X-P, Hershman JM, Mirrel CJ, Pekary AE. Impairment of hypothalamic-pituitary-thyroid function in rats treated with human recombinant tumor necrosis factor-a (cachectin). Endocrinology 1989; 125:76. 42. Kakucska I, Romero LI, Clark BD, et al. Suppression of thyrotropin releasing hormone gene expression by interleukin 1-beta in the rat: implications for nonthyroid illness. Neuroendocrinology 1994; 59:129. 43. Bacci V, Schussler GC, Kaplan TB. The relationship between serum tri-iodothyronine and thyrotropin during systemic illness. J Clin Endocrinol Metab 1982; 54:1229. 44. van Heyningen V, Abbott SR, Daniel SG, et al. Development and utility of a monoclonal-antibody-based, highly sensitive immunoradiometric assay of thyrotropin. Clin Chem 1987; 33:1387. 45. Brennan MD, Klee GO, Preissner CM, Hay ID. Heterophilic serum antibodies: a cause for falsely elevated serum thyrotropin levels. Mayo Clin Proc 1987; 62:894. 46. Hay ID, Bayer MF, Kaplan MM, et al. American Thyroid Association assessment of current free thyroid hormone and thyrotropin measurements and guidelines for future clinical assays. Clin Chem 1991; 37:2002. 47. Spencer CA, Takeuchi M, Kazarosyan M. Current status and performance goals for serum thyrotropin (TSH) assays. Clin Chem 1996; 42:140. 48. Bayer MF. Performance criteria for appropriate characterization of (highly) sensitive thyrotropin assays. Clin Chem 1987; 33:630. 49. Nicoloff JT, Spencer CA. The use and misuse of the sensitive thyrotropin assays. J Clin Endocrinol Metab 1990; 71:553. 50. Saller B, Brod'a N, Heydarian R, et al. Utility of third generation thyrotropin assays in thyroid function testing. Exp Clin Endocrinol Diabetes 1998; 106:S29. 51. Ross DS, Daniels GH, Gouveia D. The use and limitations of a chemiluminescent thyrotropin assay as a single thyroid function test in an outpatient endocrine clinic. J Clin Endocrinol Metab 1990; 71:764. 52. Attia J, Margetts P, Guyatt G. Diagnosis of thyroid disease in hospitalized patients: a systematic review. Arch Intern Med 1999; 159:658. 53. Spencer C, Eigen A, Shen D, et al. Specificity of sensitive assays of thyrotropin (TSH) used as a screen for thyroid disease in hospitalized patients. Clin Chem 1987; 33:1391. 54. Hershman JM. Clinical application of thyrotropin-releasing hormone. N Engl J Med 1974; 290:886. 55. Spencer CA, Schwarzbein D, Guttler RB, et al. Thyrotropin (TSH)-releasing hormone stimulation test responses employing third and fourth generation TSH assays. J Clin Endocrinol Metab 1993; 76:494. 56. Bigos ST, Ridgway EC, Kourides IA, Maloof F. Spectrum of pituitary alterations with mild and severe thyroid impairment. J Clin Endocrinol Metab 1978; 46:317. 57. Tunbridge WMG, Brewis M, French JM, et al. Natural history of autoimmune thyroiditis. BMJ 1981; 282:258. 58. Ross DS, Ardisson U, Meskell MJ. Measurement of thyrotropin in clinical and subclinical hyperthyroidism using a new chemiluminescent assay. J Clin Endocrinol Metab 1989; 69:684. 59. Helfand M, Crapo LM. Monitoring therapy in patients taking levothyroxine. Ann Intern Med 1990; 113:450. 60. Schultz RM, Glassman MS, MacGillivray MH. Elevated threshold for thyrotropin suppression in congenital hypothyroidism. Am J Dis Child 1980; 134:19. 61. Vagenakis AG, Braverman LE, Azizi F, et al. Recovery of pituitary thyrotropic function after withdrawal of prolonged thyroid-suppression therapy. N Engl J Med 1975; 293:681. 62. Cohen LE, Wondisford FE, Radovick S. Role of Pit-1 in the gene expression of growth hormone, prolactin, and thyrotropin. Endocrinol Metab Clin North Am 1996; 25:523. 63. Pellegrini-Bouiller I, Blicar P, Barlier A, et al. A new mutation of the gene encoding transcription factor Pit-1 is responsible for combined pituitary hormone deficiency. J Clin Endocrinol Metab 1996; 81:2790. 64. Deladoy J, Flck C, Bykgebiz A, et al. Hot spot in the PROP1 gene responsible for combined pituitary hormone deficiency. J Clin Endocrinol Metab 1999; 84:1645. 65. Doeker BM, Pfffle RW, Pohlenz J, Andler W. Congenital central hypothyroidism due to a homozygous mutation in the thyrotropin bsubunit gene follows an autosomal recessive inheritance. J Clin Endocrinol Metab 1998; 83: 1762. 66. Hayashizaki Y, Hiraoka Y, Endo Y, Matsubara K. Thyroid stimulating hormone (TSH) deficiency caused by a single base substitution in the CAGYC region of the b-subunit. EMBO J 1989; 8:2291. 67. Collu R, Tang J, Castagn J, et al. A novel mechanism for isolated central hypothyroidism: inactivating mutations in the thyrotropin-releasing hormone receptor gene. J Clin Endocrinol Metab 1997; 82:1561. 68. Beck-Peccoz P, Amr S, Menzezes-Ferreira MM, et al. Decreased receptor binding of biologically inactive thyrotropin in central hypothyroidism. N Engl J Med 1985; 312:1085. 69. Klijn JGM, Lamberts SWJ, De Jong FH, et al. The importance of pituitary tumour size in patients with hyperprolactinaemia in relation to hormonal variables and extrasellar extension of tumour. Clin Endocrinol (Oxf) 1980; 12:341. 70. Patel YC, Burger HG. Serum thyrotropin (TSH) in pituitary and/or hypothalamic hypothyroidism: normal or elevated basal levels and paradoxical responses to thyrotropin-releasing hormone. J Clin Endocrinol Metab 1973; 37:190. 71. Snyder PJ, Jacobs LS, Rabello MM, et al. Diagnostic value of thyrotrophin-releasing hormone in pituitary and hypothalamic diseases. Ann Intern Med 1974; 81:751. 72. Weintraub BD, Gershengorn MC, Kourides IA, Fein H. Inappropriate secretion of thyroid stimulating hormone. Ann Intern Med 1981; 95:339. 73. Brucker-Davis F, Oldfield EH, Skarulis MC, et al. Thyrotropin-secreting pituitary tumors: diagnostic criteria, thyroid hormone sensitivity, and treatment outcome in 25 patients followed at the National Institutes of Health. J Clin Endocrinol Metab 1999; 84:476. 74. McDermott MT, Ridgway EC. Central hyperthyroidism. Endocrinol Metab Clin North Am 1998; 27:187 75. Gesundheit N, Petrick PA, Nissim M, et al. Thyrotropin-secreting pituitary adenomas: clinical and biochemical heterogeneity. Ann Intern Med 1989; 111:827. 76. Refetoff S, Weiss RE, Usala SJ. The syndromes of resistance to thyroid hormone. Endocr Rev 1993; 14:348. 77. Beck-Pecoz P, Forloni F, Cortelazz D, et al. Pituitary resistance to thyroid hormones. Horm Res 1992; 38:66. 78. Helzberg JH, McPhee MS, Zarling EJ, Lukert BP. Hepatocellular carcinoma: an unusual course with hyperthyroidism and inappropriate thyroid-stimulating hormone production. Gastroenterology 1985; 88:181. 79. Higgins HP, Hershman JM, Kenimer JG, et al. The thyrotoxicosis of hydatidiform mole. Ann Intern Med 1975; 83:307. 80. Hoermann R, Broecker M, Grossmann M, et al. Interaction of human chorionic gonadotropin (hCG) and asialo-hCG with recombinant human thyrotropin receptor. J Clin Endocrinol Metab 1994; 77: 1009. 81. Spiegel AM. Mutations in G proteins and G proteincoupled receptors in endocrine disease. J Clin Endocrinol Metab 1996; 81:2434. 82. Van Sande J, Parma J, Tonacchera M, et al. Somatic and germline mutations of the TSH receptor gene in thyroid diseases. J Clin Endocrinol Metab 1995; 80:2577. 83. Bodenner DL, Lash RW. Thyroid disease mediated by molecular defects in cell surface and nuclear receptors. Am J Med 1998; 105:524. 84. Morris JC. The clinical expression of thyrotropin receptor mutations. Endocrinologist 1998; 8:195. 85. Parma J, Duprez L, Van Sande J, et al. Somatic mutations in the thyrotropin receptor gene cause hyperfunctioning thyroid adenomas. Nature 1993; 365:649. 86. Holzapfel HP, Fhrer D, Wonerow P, et al. Identification of constitutively activating somatic thyrotropin receptor mutations in a subset of toxic multinodular goiters. J Clin Endocrinol Metab 1997; 82:4229. 87. Parma J, Duprez L, Van Sande J, et al. Diversity and prevalence of somatic mutations in the thyrotropin receptor and G s a genes as a cause of toxic thyroid adenomas. J Clin Endocrinol Metab 1997; 82:2695. 88. Russo D, Arturi F, Wicker R, et al. Genetic alterations in thyroid hyperfunctioning adenomas. J Clin Endocrinol Metab 1995; 80:1347. 89. Duprez L, Parma J, Costagliola S, et al. Constitutive activation of the TSH receptor by spontaneous mutations affecting the N-terminal extracellular domain. FEBS Lett 1997; 409:469. 90. Tonacchera M, Chiovato L, Pinchera A, et al. Hyperfunctioning thyroid nodules in toxic multinodular goiter share activating thyrotropin receptor mutations with solitary toxic adenoma. J Clin Endocrinol Metab 1998; 83:492. 91. Fhrer D, Holzapfel HP, Wonerow P, et al. Somatic mutations in the thyrotropin receptor gene and not in the Gs a protein gene in 31 toxic thyroid nodules. J Clin Endocrinol Metab 1997; 82:3885. 92. Leclre J, Bn MC, Aubert V. Clinical consequences of activating germline mutations of TSH receptor, the concept of toxic hyperplasia. Horm Res 1997; 47:158. 93. Duprez L, Parma J, Van Sande J, et al. Germline mutations in the thyrotropin receptor gene cause non-autoimmune autosomal dominant hyperthyroidism. Nat Genet 1994; 7:396. 94. Fhrer D, Wonerow P, Willgerodt H, Paschke R. Identification of a new thyrotropin receptor germline mutation (Leu629 Phe) in a family with neonatal onset of autosomal dominant nonautoimmune hyperthyroidism. J Clin Endocrinol Metab 1997; 82:4234. 95. Schwab KO, Gerlich M, Broecker M, et al. Constitutively active germline mutation of the thyrotropin receptor gene as a cause of congenital hyperthyroidism. J Pediatr 1997; 131:899. 96. Grters A, Schneberg T, Biebermann H, et al. Severe congenital hyperthyroidism caused by a germ-line neo mutation in the extracellular portion of the thyrotropin receptor. J Clin Endocrinol Metab 1998; 83:1431. 97. Khoo DHC, Parma J, Rajasoorya C, et al. A germline mutation of the thyrotropin receptor gene associated with thyrotoxicosis and mitral valve prolapse in a Chinese family. J Clin Endocrinol Metab 1999; 84:1459. 98. Sunthoruthepvarakui T, Gottschalk ME, Hayashi Y, Refetoff S. Resistance to thyrotropin caused by mutations in the thyrotropin-receptor gene. N Engl J Med 1995; 332:155. 99. de Roux N, Misrahi M, Brauner R, et al. Four families with loss of function mutations of the thyrotropin receptor. J Clin Endocrinol Metab 1996; 81:4229. 100. Gagn N, Parma J, Deal C, et al. Apparent congenital athyreosis contrasting with normal plasma thyroglobulin levels and associated with inactivating mutations in the thyrotropin receptor gene: are athyreosis and ectopic thyroid distinct entities? J Clin Endocrinol Metab 1998; 83:1771. 101. Levine MA, Downs RW Jr, Moses AM, et al. Resistance to multiple hormones in patients with pseudohypoparathyroidism. Am J Med 1983; 74:545. 102. Thotakura NR, Desai BK, Bates LG, et al. Biological activity and metabolic clearance of a recombinant human thyrotropin produced in Chinese hamster ovary cells. Endocrinology 1991; 128:341. 103. Meier CA, Braverman LE, Ebner SA, et al. Diagnostic use of recombinant human thyrotropin in patients with thyroid carcinoma (phase I/II study). J Clin Endocrinol Metab 1994; 78:188. 104. Haugen BR, Pacini F, Reiners C, et al. A comparison of recombinant human thyrotropin and thyroid hormone withdrawal for the detection of thyroid remnant or cancer. J Clin Endocrinol Metab 1999; 84:3877. 105. Spiegel AM, Weinstein US, Shenker A. Abnormalities in G protein coupled signal transduction pathways in human disease. J Clin Invest 1993; 92:1119.

CHAPTER 16 PITUITARY GONADOTROPINS AND THEIR DISORDERS Principles and Practice of Endocrinology and Metabolism

CHAPTER 16 PITUITARY GONADOTROPINS AND THEIR DISORDERS

WILLIAM J. BREMNER, ILPO HUHTANIEMI, AND JOHN K. AMORY Structure, Synthesis, and Storage of Gonadotropins Subunit Composition Carbohydrate Content Control of Secretion of Gonadotropins Pulsatile Activity Controlling Factors Effects of Luteinizing HormoneReleasing Hormone on Gonadotropes Therapeutic Usefulness of Luteinizing HormoneReleasing Hormone Pulsatile Versus Continuous Administration Luteinizing HormoneReleasing Hormone Agonists Luteinizing HormoneReleasing Hormone Antagonists Gonadotropin-Releasing HormoneAssociated Peptide Effects of Gonadal Hormones on Gonadotropins Gonadal Steroids Inhibin Luteinizing Hormone and Follicle-Stimulating Hormone in Peripheral Blood Clearance Measurement Gonadal Effects of Gonadotropins Abnormalities of Gonadotropin Secretion and Action Gonadotrope Adenomas of the Pituitary Chapter References

The gonadotropins, luteinizing hormone (LH) and follicle-stimulating hormone (FSH), are large glycoproteins secreted from the anterior pituitary gland in response to hormonal signals from the brain and the gonads (Fig. 16-1). LH and FSH serve as intermediary messengers in the neuroendocrine system, which transmits environmental and central nervous system (CNS) information to the reproductive system. By transmitting the effects of exercise, diet, stress, and, in some species, of photoperiod and olfactory impulses, LH and FSH can stimulate ovarian and testicular function and therefore alter the physiology, fertility, and behavior of the host organism. In addition, pituitary gonadotropin secretion responds to hormonal signals returning from the gonads to allow integration of the function of the reproductive system, for example, during the menstrual cycle.

FIGURE 16-1. Schema for the control of gonadotropin secretion. (*Estrogen can be inhibitory or stimulatory to luteinizing hormone releasing hormone [GnRH], depending on the level and duration of exposure, and perhaps other factors.) (, stimulatory; , inhibitory; stim., stimulates; LH, luteinizing hormone; FSH, follicle-stimulating hormone.)

STRUCTURE, SYNTHESIS, AND STORAGE OF GONADOTROPINS

SUBUNIT COMPOSITION The a and b subunit structure of LH and FSH1,2 is similar to that of thyroid-stimulating hormone and human chorionic gonadotropin (hCG; see Chap. 15 and Chap. 112). Within each species, the a subunit is essentially identical in structure among the four hormones, whereas the b subunit differs among the four hormones and determines the biologic function of the dimeric molecule. The two subunits must be bound together through noncovalent interactions to be bioactive; neither has significant activity alone. CARBOHYDRATE CONTENT The approximate molecular mass of LH is 28 kDa and of FSH is 33 kDa; the a subunit common to both has a molecular mass of 14 kDa. These weights are approximate because of heterogeneity in the oligosaccharide (carbohydrate) moieties. The common a subunit has two N-linked carbohydrate side chains, LH-b has one, and FSH-b has two. Human chorionic gonadotropin-b has, in addition to two N-linked carbohydrate moieties, four O-linked moieties in its C-terminal 32-amino-acid extension. The exact structure of these carbohydrate side chains is unknown and variable, but they contribute ~16% of the weight of the LH molecule. The composition of the carbohydrate side chains of a given hormone shows a certain degree of microheterogeneity.3 The composition of these gonadotropin isoforms apparently varies according to the physiologic state and may, thus, determine the intrinsic bioactivity of the circulating gonadotropin at a given moment. However, the physiologic and pathophysiologic significance of this variability still remains unclear. The sialic acid component of the carbohydrate side chain varies markedly in amount among the gonadotropins: 20 residues per molecule in hCG, 5 in FSH, and 1 or 2 in LH. The sialic acid content is directly related to the half-life of the hormone circulating in blood; hCG is cleared slowly, LH relatively rapidly, and FSH at an intermediate rate. In addition, the high degree of terminal sulfation of the carbohydrate termini in LH, as well as the rapid elimination of this type of glycoprotein through action of a specific liver receptor, contribute to the faster elimination rate of LH in comparison to FSH.4 Progressive desialylation of the gonadotropins shortens their half-lives and, therefore, decreases their bioactivity in vivo, whereas their in vitro bioactivity may be retained.5,6 Some of the deglycosylated hormones bind to receptors in vitro but do not stimulate adenylate cyclase, thereby functioning as competitive antagonists for the native hormone.7 The a and b subunits are products of separate genes for which the structures have been determined.8,9 A single gene encodes the a subunit for the four glycoprotein hormones. The amino acid sequence contains a leader or signal peptide that is removed before addition of the carbohydrate moieties (see Chap. 3). An important advance has been the expression in cultured mammalian cells of complementary DNA (cDNA) for the gonadotropins.2 In mammalian cells, processing and glycosylation of the subunits, as well as combination and secretion, occurred in large quantities. In previously used bacterial cell systems, this had not been possible. The hormones produced by this recombinant technology are much purer than those obtained from human pituitaries or urine and can be produced in large quantities with constant quality. Significantly, they lack the postulated capability to transmit Creutzfeldt-Jakob disease, something which has halted the use of hormones prepared from human pituitaries10 (see Chap. 12 and Chap. 18). With this technology, one can also prepare gonadotropin analogs of different conformation and amino-acid or carbohydrate content, and test the structure and function relationships of these molecules, including possible antagonist compounds. Recombinant gonadotropins are now available for human LH, human FSH, and hCG.11 The availability of these agents facilitates greatly the care of patients with gonadotropin deficiencies and aids ovarian priming procedures involved in artificial reproductive techniques such as in vitro fertilization (IVF) and intracytoplasmic sperm injection (ICSI). The cells that synthesize and store gonadotropins in the anterior pituitary, the gonadotropes, are distributed singly in acini made up largely of other cell types. Most

gonadotropes stain with antibodies to both LH and FSH (or their b subunits), a finding that implies that the same cell produces both gonadotropins.12 A few cells stain for either LH or FSH, but not for both (see Chap. 11). Because of its better supply and similarity in action to LH, hCG has generally been used in diagnostic testing and in therapy (see Chap. 93 and Chap. 97). Human menopausal gonadotropin (or menotropin), which is obtained from the urine of postmeno-pausal women, has an FSH-like and LH-like action; one preparation (Pergonal) has been used extensively to induce ovulation or spermatogenesis (see Chap. 97). Purified FSH and LH from urinary sources, as well as the recombinant form of FSH, are now available for clinical use. CONTROL OF SECRETION OF GONADOTROPINS The synthesis and secretion of LH and FSH are under the control of luteinizing hormonereleasing hormone (LHRH, also called gonadotropin-releasing hormone, GnRH), and gonadal steroids and peptides, including inhibin and activin (see Fig. 16-1). GnRH is a decapeptide that is a product of hypothalamic neurosecretion and is carried by the hypothalamico-hypophysial portal system to the gonadotropic cells of the anterior pituitary (Fig. 16-2). Binding of GnRH to receptors on the pituitary gonadotropes causes the release of both FSH and LH.

FIGURE 16-2. Amino-acid composition of luteinizing hormonereleasing hormone (GnRH) and some GnRH agonist analogs.

During embryonic life, the cell bodies of GnRH-secreting neurons are detected in the olfactory placode, from where they gradually migrate to their final locations in the hypothalamus. This migration of GnRH neurons is disturbed in the most common X-linked form of Kallmann syndrome (hypogonadotropic hypogonadism and anosmia), due to mutation in an extracellular matrix protein, anosmin, and disturbance in development of the olfactory bulbs and tracts.13 Other forms of hypogonadotropic hypogonadism have been found to be due to inactivating mutations in the GnRH-receptor gene.14 After the embryonic migration period, cell bodies of GnRH neurons are found predominantly in two regions of the primate hypothalamus: (a) anteriorly, especially in the medial preoptic area and the interstitial nucleus of the stria terminalis, and (b) in the tuberal regions, particularly the arcuate nucleus and adjacent paraventricular nucleus (see Chap. 8).15,16 Physiologically important projections from these neurons go to the median eminence, where they terminate near the capillary bed of the portal system. Projections also are found to other CNS areas, such as the amygdala, hippocampus, and periaqueductal gray area, and to the posterior pituitary. The function of these projections is unknown, but the CNS projections could be important in the behavioral effects reported for GnRH. Stimulatory effects of GnRH on sexual behavior due to a direct CNS site of action have been reported in rodents. PULSATILE ACTIVITY GnRH is secreted in pulses that vary in frequency from approximately one per hour to as few as one or two per 24 hours. In sheep, direct measurement of GnRH in portal blood simultaneously with the measurement of LH in peripheral blood demonstrated that each LH pulse is preceded by a GnRH pulse.17 To generate pulsatile secretory activity, many GnRH neurons must be synchronized to secrete almost simultaneously. This synchronized neural activity can be recorded by electrodes placed in the hypothalamus and correlates with pulsatile increases of LH into the circulation.18 The neural mechanisms underlying this synchronous activity are not well understood (see Chap. 6). CONTROLLING FACTORS GnRH secretory activity is affected by many CNS neurotrans-mitters and by gonadal steroids as well as other hormones. Opiates suppress gonadotropin secretion.19 This effect is almost certainly mediated through an effect on GnRH secretion, because little evidence exists that opiates exert a direct effect on pituitary responsiveness to GnRH in vitro. Opiate antagonists, such as naloxone, can stimulate LH secretion in humans, suggesting a significant tonic inhibitory effect of endogenous opiates on GnRH secretion.20 An important interaction occurs between the opiate effect and the prevailing steroid milieu, because naloxone administration increases LH levels in women in the late follicular and luteal phases but not in those in the early follicular phase or in postmenopausal women.20,21 In monkeys, little b-endorphin enters the portal blood during the early follicular phase or in ovariectomized animals.22 Levels of b-endorphin are markedly increased by estrogen and progesterone administration, a result that implies that these steroids may exert their inhibitory feedback effects partly through endogenous opiates.23 In men, similar mechanisms may occur, because testosterone markedly slows the frequency of pulsatile LH secretions, and opiate blockade inhibits this effect.24,25 Catecholamines that arise in central neurons probably are important in the control of GnRH, although relatively few data supporting this concept are available for humans. Available evidence suggests that, in general, norepinephrine is stimulatory to GnRH secretion, and dopamine may be inhibitory, but further work is necessary, particularly studies on the effects of different steroidal milieux.26 The roles of other neurotransmitters, including serotonin, g-aminobutyric acid, and epinephrine, in the physiology of GnRH secretion are even less clear. Prolactin and glucocorticoids may exert inhibitory effects. Gonadal steroids exert profound effects on gonadotropin secretion; in many situations, whether these effects are exerted at a hypothalamic or a pituitary level is difficult to ascertain. However, a change in the frequency of pulsatile LH secretion is generally assumed to reflect an effect on the hypothalamus, because the frequency of LH pulses closely mimics that of GnRH pulses. In studies based on this assumption, progesterone has been demonstrated to have an inhibitory effect on GnRH, converting the normally rapid LH pulses of the follicular phase to the much slower frequency characteristic of the luteal phase.23 Similarly, testosterone administration leads to markedly slower LH pulse frequency in men, presumably through a hypothalamic effect.24 Estradiol exerts both inhibitory and stimulatory effects, depending on the level and duration of exposure. EFFECTS OF LUTEINIZING HORMONERELEASING HORMONE ON GONADOTROPES GnRH plays a central role in the biology of reproduction. GnRH interacts with high-affinity receptors on gonadotrope cells of the anterior pituitary. The GnRH receptor belongs to the family of seven transmembrane domain receptors, interacting on ligand binding with Gq.27 This leads sequentially to activation of different phospholipases and release of free Ca2+ and lipid-derived molecules as second messengers.28 Activation of plasma membrane Ca2+ channels, mobilization of Ca2+ from intracellular stores, and activation of calmodulin-stimulated cell responses play an important role in GnRH action.29 In addition, the activation of phospholipase C (PL-C) is an early response, followed by those of phospholipase A-2 (PLA-2) and phospholipase D (PLD). Generation of second messengers inositol-1,4,5-triphosphate and diacyl-glycerol (DAG) mobilize intracellular pools of Ca2+ and activate protein kinase C (PKC). Both Ca2+-dependent conventional and CA2+ independent novel PK isoforms are activated during this process. Arachidonic acid (AA), liberated by activated PLA-2, also participates in PKC activation. Crosstalk between Ca2+, AA, and selected lipoxygenase products (e.g., leukotriene C4), and the different PKC isoforms might generate compartmentalized signal transduction cascades on GnRH stimulation of the gonadotrope-producing cells. These include the mitogen-activated protein kinase (MAPK) cascade. Activation of c-jun and c-fos by GnRH stimulation can participate in transcriptional regulation through formation of the transcription factor AP-1. At least partly dissimilar signal transduction systems mediate the GnRH-stimulated acute secretion of gonadotropins and the more prolonged stimulation of new gonadotropin synthesis. The events previously described lead to changes in intracellular protein kinase levels that are thought to be important in stimulating gonadotropin secretion and synthesis, although the detailed mechanisms of these stimulatory effects are unknown (see Chap. 4). Gonadotropin synthesis occurs through the classic process of ribosomal formation of peptide chains, followed by posttranslational modifications in the endoplasmic reticulum and Golgi apparatus (see Chap. 3). These modifications include cleavage of segments off the amino terminus of both the a and b subunits and the subsequent addition of carbohydrate moieties to form the mature

gonadotropin molecules in the secretory granules. After synthesis and storage in granules, gonadotropins are available for release almost immediately after GnRH stimulation. After exocytosis, the gonadotropins diffuse rapidly into the nearby capillaries and appear in the venous effluent of the pituitary.

THERAPEUTIC USEFULNESS OF LUTEINIZING HORMONERELEASING HORMONE

GnRH and its agonist analogs are important in the treatment of GnRH deficiency states, including menstrual and fertility disorders in women (see Chap. 96 and Chap. 97) and hypothalamic causes of hypogonadism in men, such as Kallmann syndrome (see Chap. 115). However, GnRH infusion tests rarely give diagnostic information beyond that obtainable by measuring basal gonadotropin and gonadal steroid levels. PULSATILE VERSUS CONTINUOUS ADMINISTRATION30 Gonadotropin secretion does not follow the GnRH input signal exactly, particularly when GnRH is administered in a nonphysiologic pattern. Soon after GnRH became available for experimental use, researchers noted that continuous administration of this hormone (e.g., by constant intravenous infusion) did not lead to constant LH output but rather to a biphasic pattern of increase, implying that two pools of releasable hormone existed.31 After the maximal response, a decline was seen in LH responsiveness that could not be explained by gonadal inhibitory feedback.32 Pulsatile administration of GnRH was observed to cause persistent pulsatile secretion of LH, whereas continuous GnRH infusion caused a desensitization of pituitary responsiveness.33 Continuous high-dose administration of GnRH over several days leads to almost complete failure of the ability of the pituitary to respond to GnRH secreted endogenously, yielding an experimentally induced hypogonadotropic hypogonadal state.34 LUTEINIZING HORMONERELEASING HORMONE AGONISTS The ability of continuous GnRH administration to inhibit pituitary gonadotropin production has been expanded greatly with the use of GnRH agonists (see Fig. 16-2).35 Once-daily administration of the agonist analogs results in several days of increased gonadotropin secretion, followed by desensitization and a hypogonadotropic state with very low levels of gonadotropin and gonadal steroid production, or medical castration. Such agonists have proved useful36 in the treatment of many conditions, such as prostate cancer (Chap. 225), endometriosis (Chap. 98), precocious puberty (Chap. 92), uterine leiomyomas,37 breast cancer, polycystic ovarian disease, cyclic porphyria, and acne, and for ovulation suppression and in vitro fertilization. Potentially useful applications include male contraception and the treatment of benign prostatic hypertrophy. LUTEINIZING HORMONERELEASING HORMONE ANTAGONISTS Antagonist analogs of GnRH are undergoing active development.35 These substances bind to GnRH receptors in the pituitary without stimulating gonadotropin secretion and block the effect of endogenous GnRH. Unlike GnRH agonists, they are immediately inhibitory, a potential advantage in the treatment of hormonally dependent conditions, such as prostate cancer. Antagonists may also prove superior to agonists in situations in which total gonadotropin inhibition is required, such as male contraceptive development, because even low levels of LH and FSH can partially support spermatogenesis.38,39 In nonhuman primate studies, the administration of an GnRH antagonist, together with physiologic replacement dosages of testosterone (because the animals would otherwise be androgen deficient), reliably eliminates spermatogenesis.40 Because the suppressive effect of the antagonist appears fully reversible, this combination shows promise for male contraceptive development (see Chap. 123). GONADOTROPIN-RELEASING HORMONEASSOCIATED PEPTIDE GnRH (i.e., gonadotropin-releasing hormone) is synthesized in the hypothalamus as a part of a larger peptide, which is cleaved to yield GnRH and a larger fragment called gonadotropin-releasing hormoneassociated peptide.41 Some reports have demonstrated that this peptide stimulates secretion of gonadotropin, particularly FSH, from pituitary cells in vitro and inhibits prolactin secretion. The physiologic role of this peptide remains to be explored, however.

EFFECTS OF GONADAL HORMONES ON GONADOTROPINS

GONADAL STEROIDS Gonadal steroids and proteins exert direct pituitary effects on gonadotropin secretion along with their effects on GnRH secretion. Estradiol, for example, exerts both negative and positive feedback effects directly on the pituitary. Testosterone administration inhibits LH and FSH secretion by a direct pituitary effect in men.42 This effect is mediated in part by aromatization of testosterone to estradiol in vivo. Progesterone and testosterone also exert direct pituitary effects when studied in vitro, although these steroids have not been studied as definitively in vivo as has estradiol. INHIBIN The major protein hormone product of the gonads is inhibin, a substance that has been studied for over 50 years,43 and now has been characterized structurally.44,45 Two 32-kDa forms of the glycoprotein, termed inhibin A and inhibin B and containing a and b subunits, have been described. Inhibin inhibits pituitary FSH secretion selectively, although at high levels it may also inhibit LH. Inhibin is produced in the Sertoli cells of the testis and in granulosa cells of the ovary, mainly under the stimulatory influence of FSH46,54 The structure of inhibin bears an interesting homology to that of other substances such as transforming growth factor and antimllerian hormone.44,47 Antimllerian hormone, also a product of Sertoli cells, leads to regression of the mllerian duct in the male embryo. Interestingly, alternative combinations of the subunits of inhibin (b-b instead of a-b) exert stimulatory effects on FSH secretion in vitro rather than the inhibitory effect of the inhibin molecule.48,49 These b-b combinations are called activins. Whether or not activins are secreted from the gonads, and what their effects might be in vivo, are areas of current research activity. Surprisingly, one of the activins stimulates hemoglobin production in human bone marrow in vitro.50 Many other effects of this family of substances have been described, including stimulation of early embryonic development and immunologic alterations. Follistatin, another protein produced by the gonads, also inhibits FSH secretion in primates.51 The precise roles of inhibins, activins, and follistatin in the control of gonadotropin secretion are active areas of investigation. In women, inhibin A is elevated in the luteal phase of the menstrual cycle, whereas inhibin B is elevated in the follicular phase.52 Levels of both decline with age, implying a decrease in the quality of the remaining ovarian follicles. Inhibin B provides feedback inhibition of FSH, and a fall in inhibin B pre-dates the oligomenorrhea and increases in FSH that signal menopause. In men, inhibin B is the physiologically important form of inhibin.53 It exhibits a reciprocal relationship with FSH and a diurnal rhythm that appears to be independent of FSH or testosterone feedback.54

LUTEINIZING HORMONE AND FOLLICLE-STIMULATING HORMONE IN PERIPHERAL BLOOD

LH and FSH circulate in blood predominantly in the monomeric form found in the pituitary gland; little evidence exists for prehormones or smaller active fragments. However, changes in properties, such as the ratio of bioactivity to immunoactivity, have been reported in various states of aging, steroid environments, and GnRH analog administration. Some of these changes are apparently due to inaccuracies in determination of gonadotropin immunoreactivity.55 In monkeys, ovariectomy leads to a slight increase in molecular size and a decrease in the clearance rate of LH and FSH, which can be reversed by estrogen treatment. In aging men, the ratios of bioactivity to immunoactivity for both LH and FSH decrease.56,57 During GnRH antagonist administration, FSH bioactivity decreases much more quickly than does immunoactivity. This phenomenon has been explained by the demonstration of FSH isoforms in human serum that block the bioeffect of FSH on the gonads.58 Apparently, the secretion of these isoforms is stimulated by the GnRH antagonist. In all these situations, the conventional radioimmunoassays yield inaccurate assessments of the level of bioactive gonadotropins in blood. This error can be usually eliminated by use of two-site immunoassays. CLEARANCE The gonadotropins are cleared from the blood by both the kidney and the liver. Small amounts are also bound to the gonads, but this accounts for little of the hormone clearance. Some 10% to 20% of the hormone appears in a bioactive form in urine; most is metabolized in the hepatic and renal parenchyma. The half-life of FSH (4 to 5 hours) is much longer than that of LH (30 to 60 minutes); this is due partly to the higher sialic acid content of FSH, which impairs clearance, particularly by the liver, and partly to the high proportion of sulfated carbohydrate termini in LH, which accelerates its hepatic clearance.4

MEASUREMENT After the classical in vivo bioassays, the preferred method of gonadotropin measurements used to be radioimmunoassay using polyclonal antiserum. These assays now appear suboptimal due to their low sensitivity and poor specificity; in particular, they are unable to distinguish between low-normal and low levels. The second generation of immunoassays, using the non-competitive immunometric principle and nonradioactive signaling systems (enzyme, fluorescence, chemiluminescence) are superior in terms of sensitivity and specificity. Gonadotropin concentrations are expressed in terms of partially purified preparations of pituitary or urinary hormones. The use of standards of different purity, and various antisera, leads to different reference ranges reported by individual laboratories and makes their comparisons difficult. This problem will be partly overcome when recombinant gonadotropins are adopted as standards. However, the gonadotropin patterns during physiologic changes, such as aging (see Chap. 199), puberty (see Chap. 91), pregnancy (see Chap. 112), and the menstrual cycle (see Chap. 95), as well as in pathologic conditions, are similar in reports from various laboratories. Most of the antibodies used in LH assays cross-react with hCG, so that LH measurements are artifactually elevated when hCG levels in serum are high, as in pregnancy and choriocarcinoma (see Chap. 111 and Chap. 112). The possible presence of a common genetic variant of LH59 should be kept in mind, because some commonly used immunoassay methods do not detect this structurally aberrant form of LH. Sensitive, specific bioassays are now available for the measurement of both LH and FSH in human serum. More time-consuming and difficult to perform than immunoassays, these bioassays are not in common clinical use. In general, bioassays have confirmed the conclusions obtained with immunoassays. Some discrepancies are found, however (see above), and bioassay analysis is indicated when the immunoassay result does not fit the clinical picture.

GONADAL EFFECTS OF GONADOTROPINS

The only known bioeffects of LH and FSH are in the gonads. LH and FSH stimulate cell growth and maintenance in both the testis and ovary (see Chap. 94 and Chap. 113). As classically defined, LH stimulates steroidogenesis in both sexes, particularly testosterone synthesis, from Leydig cells in the male and from theca cells in the female. FSH stimulates spermatogenesis in the male and follicular development and estradiol secretion in the female. LH also induces ovulation from the mature follicle in the female and exerts a partial stimulatory effect on spermatogenesis in the male, probably mediated through increases in intratesticular levels of testosterone. LH and FSH were named for their initially described roles in females. Both gonadotropins act through classic protein hormonereceptor mechanisms, involving a G proteinassociated seven-transmembrane domain receptor.60,61,62 After ligand binding, adenylate cyclase is activated, leading to increases in intracellular cyclic adenosine monophosphate (cAMP), which is the main second messenger involved in gonadotropin action. The cAMP activates protein kinase, and the resulting protein phosphorylation is thought to be important in the cellular effects of the gonadotropins. In the testis, LH directly stimulates the synthesis of a steroidogenic acute regulatory (StAR) protein, which plays a key role in the transfer of cholesterol from the outer to the inner mitochondrial membrane. This is the site of the first step in steroid hormone biosynthesis from cholesterol to pregnenolone. Thereafter, the metabolic steps in the steroidogenic pathway leading to testosterone take place in the smooth endoplasmic reticulum. Testosterone exerts stimulatory effects on Sertoli cells and on spermatogenesis39 (see Chap. 113). In men, this effect is probably mediated through the stimulatory effect of LH on intratesticular testosterone levels, because LH has no known direct effect on the seminiferous epithelium. FSH binds to Sertoli cell and spermatogonial membranes in the testis. FSH is the major stimulator of seminiferous tubule growth during development. Because the tubules account for ~80% of the volume of the testis, FSH is of major importance in determining testicular size. FSH is important in the initial maturation of spermatogenesis during puberty; however, adult men can maintain sperm production despite very low blood levels of FSH if LH levels are normal.39 The total numbers of sperm produced in the absence of FSH are low. Normalization of FSH levels leads to quantitatively normal sperm production.63 These findings imply that the major physiologic role of FSH in men is to stimulate quantitatively normal levels of spermatogenesis. Additional descriptions of ovarian and testicular effects of gonadotropins are presented in Chapter 90, Chapter 91, Chapter 94, Chapter 95 and Chapter 113. ABNORMALITIES OF GONADOTROPIN SECRETION AND ACTION An outline of the causes of gonadotropin abnormalities is presented in Table 16-1. Other aspects of the pathophysiology of gonadotropins, as well as detailed discussions of the etiologies, diagnosis, and treatment of gonadotropin abnormalities, are included in Chapter 17, Chapter 92, Chapter 96, Chapter 97, Chapter 103, Chapter 114, Chapter 115. Pituitary tumors that produce gonadotropins are discussed below.

TABLE 16-1. Abnormalities of Gonadotropins

Genetic studies have detected several mutations in the gonadotropin and gonadotropin-receptor genes. The ligand mutations are exclusively of the loss-of-function type, whereas both loss- and gain-of-function mutations have been discovered in the receptor genes. In the latter, the signal transduction system of the receptor is partially activated in the absence of ligand hormone, resulting in constitutive activation of the hormonal effects. These mutations have been very educational in terms of unraveling certain details of the physiology of gonadotropin action as well as the pathogenesis of certain disorders of reproductive function. Mutations in the FSH and LH b-subunit genes are extremely rare, probably due to the key role of these hormones in regulation of reproduction; no a-subunit mutations have so far been described. An inactivating LH-b mutation was found to cause absence of Leydig cells, lack of spontaneous puberty, and infertility in a male with normal early sexual differentiation.64 Two females with homozygous mutations of the FSH-b gene have been described in the literature.65,66 They both had primary amenorrhea with poor development of secondary sexual characteristics. One azoospermic male with similar mutations in the FSH-b gene has been described.67 However, he may have suffered from an additional unrelated disturbance of Leydig cell function, because his testosterone level was low, and his LH level was high. Inactivating LH-receptor mutations67,68 and 69 in genetic males cause pseudohermaphroditism with severe Leydig cell hypoplasia, a finding that indicates a crucial role for the LH receptor in the stimulation of fetal testicular testosterone production and male sexual differentiation. In females, the phenotype of these mutations is anovulatory infertility and hypoestrogenism.69,70 An inactivating mutation in the FSH-receptor gene, an Ala Val point mutation at position 198, has been detected in the extracellular domain of the receptor.71 The homozygous females have hypergonadotropic primary or early-onset secondary amenorrhea with arrest of follicular development, a diagnosis often termed resistant ovary syndrome. The male phenotype was surprising, because only variable oligoasthenozoospermia was found in these normally masculinized men, but no azoospermia or complete infertility.72 This indicates that FSH action is not an absolute requirement for the pubertal initiation and maintenance of spermatogenesis or fertility. However, qualitatively and quantitatively normal spermatogenesis appears to be dependent on FSH action. Partially inactivating mutations of the FSH-receptor gene have been described; the phenotype of these individuals includes secondary amenorrhea, high circulating gonadotropin levels, and arrest of follicular growth at the early antral stage,73 similar to that of patients with completely inactivating mutations. The role of FSH in ovarian and testicular function as indicated by the inactivating FSH ligand and receptor mutations is corroborated by very similar phenotypes observed in FSH-b74 and FSH-receptor75 knock-out mice. Also, constitutively activating mutations are known for the LH and FSH-receptor genes. The former results in the familial male-limited, gonadotropin-independent, precocious puberty, testotoxicosis.76,77 The constitutive activity of the LH-receptor results in onset of testicular testosterone synthesis without LH action. No phenotype has been described in women affected with activation of LH-receptor mutations, apparently because premature LH activation has no effects on ovarian function without

previous FSH priming, or because the prepubertal ovary does not express the LH receptor. One case of a possible activating FSH-receptor mutation has been described in a male.78 The subject was a male who had been hypophysectomized due to pituitary adenoma. Despite unmeasurable gonadotropin levels, he had persistently normal spermatogenesis. A mutation was found in his FSH-receptor gene, resulting in marginal constitutive activation of the FSH-receptor and probably in maintenance of spermatogenesis in the absence of FSH.

GONADOTROPE ADENOMAS OF THE PITUITARY

Many pituitary adenomas, previously classified as nonfunctional, produce gonadotropins or their subunits.79 Indeed, as many as 40% to 50% of all macroadenomas may originate in gonadotropes.80 These tumors (Table 16-2) are generally recognized because of mass effects, such as visual impairment, headache, or the findings of sellar enlargement and pituitary mass on radiologic examination. The gonadotropins and their subunits that are produced only uncommonly lead to a clinical syndrome, partly explaining why these tumors were previously thought to be non-functional. Gonadotropin subunits have no known bioeffects, and the intact hormones are rarely produced in sufficient amount to cause a clinical syndrome. In fact, paradoxically, deficiencies of pituitary hormone secretion, especially of LH, are more commonly found, because the tumor mass compresses the normal pituitary, thereby impairing normal hormone production.

TABLE 16-2. Gonadotrope Adenomas of the Pituitary

The hormones produced by gonadotrope adenomas are, in decreasing order of frequency, FSH, LH, a subunit, and LH-b subunit. The elevations may be small, however, and could easily be mistaken for the elevations seen in mild primary hypogonadism. The administration of thyrotropin-releasing hormone can sometimes be useful because, in normal subjects, this hormone only rarely increases the secretion of gonadotropins or their subunits, whereas it may stimulate the secretion of these substances in patients with gonadotrope adenomas.81 Magnetic resonance imaging evaluation is important in assessing the existence and extent of a gonadotrope adenoma. In addition, magnetic resonance imaging is helpful in determining the position of the optic chiasm, the possibility of hemorrhage in the pituitary, and the differentiation of an adenoma from an aneurysm. Management also should include a thorough evaluation of pituitary function, including thyroid and glucocorticoid axes, as well as the exclusion of a prolactinoma, because these tumors may be treated medically. Therapy for gonadotrope adenomas has been reviewed.82 Gonadotrope adenomas are principally treated with resection via the transsphenoidal approach. Indications for surgery include visual field and other neurologic deficits. Improvements are seen in as many as 90% of cases.83 Irradiation can be used in patients with residual tumor after initial surgery to prevent recurrence, in patients who are poor surgical candidates, or in those whose tumors are surgically inaccessible; however, the incidence of pituitary dysfunction afterward is significant.84 Medical therapies, such as administration of GnRH antagonists and somatostatin analogs, have been reported to decrease hormone levels; however, they do not cause regression of tumor size. Their role is probably limited to treating patients with aggressive tumors for whom other therapies have failed. CHAPTER REFERENCES
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. Bousfield GR, Perry WM, Ward DN. Gonadotropins: chemistry and biosynthesis. In: Knobil E, Neill JD, eds. The physiology of reproduction. New York: Raven Press, 1994:1749. Gharib SD, Wierman ME, Shupnik MW, et al. Molecular biology of the pituitary gonadotropins. Endocr Rev 1990; 11:177. Ulloa-Aguirre A, Midgley AR Jr, Beitins IZ, Padmanabhan V. Follicle-stimulating isohormones: characterization and physiological relevance. Endocr Rev 1995; 16:765. Fiete D, Srivastava V, Hindsgaul O, Baenziger JU. A hepatic reticuloendothelial cell receptor specific for SO4-4GalNAc beta 1, 4GLcNAc beta 1,2Man alpha that mediates rapid clearance of lutropin. Cell 1991; 67:1103. Van Hall EW, Vaitukaitis GT, Ross GT, et al. Effects of progressive desialylation on the rate of disappearance of immunoreactive hCG from plasma in rats. Endocrinology 1971; 89:11. Dufau ML, Catt KJ, Tsuruhara T. Retention of in vitro biological activities by desialyated human luteinizing hormone and chorionic gonadotropin. Biochem Biophys Res Commun 1971; 44:1022. Manjunath P, Sairam MR. Biochemical, biological and immunological properties of chemically deglycosylated human choriogonadotropin. J Biol Chem 1982; 257:7109. Fiddes JC, Goodman HM. The gene encoding the common alpha subunit of the four human glycoprotein hormones. J Mol Appl Genet 1981; 1:3. Talmadge K, Vamvakopoulos NC, Fiddes JC. Evolution of the genes for the beta subunits of human chorionic gonadotropin and luteinizing hormone. Nature 1984; 307:37. Powell-Jackson J, Weller RO, Kennedy P, et al. Creutzfeldt-Jakob disease after administration of human growth hormone. Lancet 1985; 2:244. LeCotonnec J-Y, Porchet HC, Beltrami V, et al. Clinical pharmacology of recombinant human follicle-stimulating hormone (FSH). 1. Comparative pharmacokinetics with urinary FSH. Fertil Steril 1994; 61:669. Pelletier G, Robert F, Hardy J. Identification of human anterior pituitary cells by immunoelectron microscopy. J Clin Endocrinol Metab 1978; 46:534. Seminara SB, Hayes FJ, Crowley WF Jr. Gonadotropin-releasing hormone deficiency in the human (idiopathic hypogonadotropic hypogonadism and Kallmann's syndrome): pathophysiological and genetic considerations. Endocr Rev 1998; 19:521. De Roux N, Young J, Misrahi M, et al. A family with hypogonadotropic hypogonadism and mutations in the gonadotropin-releasing hormone receptor. N Engl J Med 1997; 337:1597. Barry J. Immunofluorescence study for LRF neurons in man. Cell Tissue Res 1977; 181:1. King JC, Anthony ELP. GnRH neurons and their projections in humans and other mammals. Peptides 1984; 5(Suppl 1):195. Clarke IJ, Cummins JT. The temporal relationship between gonadotropin releasing hormone (GnRH) and luteinizing hormone (LH) secretion in ovariectomized ewes. Endocrinology 1982; 111:1737. Wilson RC, Kesner JS, Kaufman J-M, et al. Central electrophysiologic correlates of pulsatile luteinizing hormone secretion in the rhesus monkey. Neuroendocrinology 1984; 39:256. Reid RL, Hoff JD, Yen SSC, Li CH. Effects on pituitary hormone secretion and disappearance rates of exogenous b-endorphin in normal human subjects. J Clin Endocrinol Metab 1981; 51:1179. Ropert JR, Quigley ME, Yen SSC. Endogenous opiates modulate pulsatile LH release in humans. J Clin Endocrinol Metab 1981; 52:583. Reid RL, Quigley ME, Yen SSC. The disappearance of opioidergic regulation of gonadotropin secretion in postmenopausal women. J Clin Endocrinol Metab 1983; 57:1107. Wehrenberg WB, Wardlaw SL, Frantz AG, Ferin M. b-Endorphin in hypophyseal portal blood: variations throughout the menstrual cycle. Endocrinology 1982; 111:879. Soules MR, Steiner RA, Clifton DK, et al. Progesterone modulation of pulsatile luteinizing hormone secretion in normal women. J Clin Endocrinol Metab 1984; 58:378. Matsumoto AM, Bremner WJ. Modulation of pulsatile gonadotropin secretion by testosterone in men. J Clin Endocrinol Metab 1984; 58:378. Veldhuis JD, Rogol AD, Samojlik E, Ertel NH. Role of endogenous opiates in the expression of negative feedback actions of androgens and estrogens on pulsatile properties of luteinizing hormone secretion in man. J Clin Invest 1984; 74:47. Barraclough CA, Wise PM. The role of catecholamines in the regulation of pituitary luteinizing hormone and follicle-stimulating hormone secretion. Endocr Rev 1982; 3:91. Sealfon SC, Weinstein H, Millar RP. Molecular mechanisms of ligand interaction with the gonadotropin-releasing hormone receptor. Endocr Rev 1997; 18:180. Conn PM. Gonadotropin-releasing hormone action. In: Adashi EY, Rock JA, Rosenwaks Z, eds. Reproductive endocrinology, surgery and technology. Philadelphia: LippincottRaven, 1996:163. Naor Z, Harris D, Shacham S. Mechanism of GnRH receptor signaling: combinatorial cross-talk of Ca 2+ and protein kinase C. Front Neuroendocrinol 1998; 19:1. Urban RJ, Evans WS, Rogol AO, et al. Contemporary aspects of discrete peak-detection algorithms. 1. The paradigm of the luteinizing hormone pulse signal in men. Endocr Rev 1988; 9:3. Bremner WJ, Paulsen CA. Two pools of luteinizing hormone in the human pituitary: evidence from constant administration of luteinizing hormone-releasing hormone. J Clin Endocrinol Metab 1974; 39:811. Bremner WJ, Findlay JK, Lee VWK, et al. Feedback effects of the testis on pituitary responsiveness to GnRH infusions in the ram. Endocrinology 1980; 106:329. Belchetz PE, Plant TM, Nakai Y, et al. Hypophysial responses to continuous and intermittent delivery of hypothalamic gonadotropin-releasing hormone. Science 1978; 202:631. Veldhuis JD. Pathophysiologic features of episodic gonadotropin secretion in man. Clin Res 1988; 35:11. Pechstein B, Nagaraja NV, Hermann R, et al. Pharmacokinetic-pharmaco-dynamic modeling of testosterone and luteinizing hormone suppression by cetrorelix in healthy volunteers. J Clin Pharmacol 2000; 40:266. Vickery BH. Comparison of the potential for therapeutic utilities with gonadotropin-releasing hormone agonists and antagonists. Endocr Rev 1986; 7:115. Friedman AF, Harrison-Atlas D, Barbieri RL. A randomized, placebo-controlled, double-blind study evaluating the efficacy of leuprolide acetate depot in the treatment of uterine leiomyomata. Fertil Steril 1989; 51:251. Matsumoto AM, Karpas AE, Paulsen CA, et al. Reinitiation of sperm production in gonadotropin-suppressed normal men by administration of follicle stimulating hormone. J Clin Invest 1983; 72:1005. Matsumoto AM, Paulsen CA, Bremner WJ. Stimulation of sperm production in gonadotropin-suppressed normal men by physiological dosages of human luteinizing hormone. J Clin Endocrinol Metab 1984; 59:882. Bremner WJ, Bagatell CJ, Steiner RA. Gonadotropin-releasing hormone antagonist plus testosterone: a potential male contraceptive. J Clin Endocrinol Metab 1991; 73:465. Nikolics K, Mason AJ, Szonyi E, et al. A prolactin-inhibiting factor within the precursor for human gonadotropin-releasing hormone. Nature 1985; 316:512. Sheckter CB, Matsumoto AM, Bremner WJ. Testosterone administration inhibits gonadotropin secretion by an effect on the human pituitary. J Clin Endocrinol Metab 1989; 68:397. Baker HWG, Bremner WJ, Burger HG, et al. Testicular control of FSH secretion. Recent Prog Horm Res 1977; 32:429. Mason AJ, Hayflick JS, Ling N, et al. Complementary DNA sequences of ovarian follicular fluid inhibin show precursor structure and homology with transforming growth factor B. Nature 1985; 318:659.

45. Forage RG, Ring JM, Brown RW, et al. Cloning and sequence analysis of cDNA species encoding for the two subunits of inhibin from bovine follicular fluid. Proc Natl Acad Sci U S A 1986; 83:3091. 46. McLachlan RI, Matsumoto AM, Burger HG, et al. The relative roles of follicle-stimulating hormone and luteinizing hormone in the control of inhibin secretion in normal men. J Clin Invest 1988; 82:880. 47. Cate RL, Mattaliano RJ, Hession C, et al. Isolation of the bovine and human genes for mllerian inhibiting substance and expression of the human gene in animal cells. Cell 1986; 45:685. 48. Vale W, Rivier J, Vaughan J, et al. Purification and characterization of an FSH releasing protein from ovarian follicular fluid. Nature 1986; 321:776. 49. Ling N, Ying S-Y, Ueno N, et al. Pituitary FSH is released by a heterodimer of the b-subunits from the two forms of inhibin. Nature 1986; 321:779. 50. Yu J, Shao L, Lemas V, et al. Importance of FSH-releasing protein and inhibin in erythrodifferentiation. Nature 1987; 330:765. 51. Meriggiola MC, Dahl KD, Mather JP, Bremner WJ. Follistatin decreases activin-stimulated FSH secretion with no effect on GnRH-stimulated FSH secretion in prepubertal male monkeys. Endocrinology 1994; 134:1967. 52. Klein N, Illingworth P, Groome NP, et al. Decreased inhibin-B secretion is associated with the monotropic rise in older ovulatory women: a study of serum and follicular fluid inhibin-A and B levels in spontaneous menstrual cycles. J Clin Endocrinol Metab 1996; 81:2742. 53. Anawalt BD, Bebb RA, Matsumoto AM, et al. Serum inhibin B levels reflect Sertoli cell function in normal men with testicular dysfunction. J Clin Endocrinol Metab 1996; 81:3341. 54. Carlsen E, Olsson C, Petersen JH, et al. Diurnal rhythm in serum levels of inhibin B in normal men: relation to testicular steroids and gonadotropins. J Clin Endocrinol Metab 1999; 84:1664. 55. Jaakkola T, Ding Y-Q, Kellokumpu-Lehtinen P, et al. The ratios of serum bioactive/immunoreactive LH and FSH in various conditions with increased and decreased gonadotropin secretion: re-evaluation by ultrasensitive immunometric assay. J Clin Endocrinol Metab 1990:1496. 56. Warner BA, Dufau M, Santen RJ. Effects of aging and illness on the pituitary testicular axis in men: qualitative as well as quantitative changes in luteinizing hormone. J Clin Endocrinol Metab 1985; 60:263. 57. Tenover JS, Dahl KD, Hsueh AJW, et al. Serum bioactive and immunoreactive follicle-stimulating hormone levels and the response to clomiphene in healthy young and elderly men. J Clin Endocrinol Metab 1987; 64:1103. 58. Dahl KD, Bicsak TA, Hsueh AJW. Naturally occurring antihormones: secretion of FSH antagonists by women treated with a GnRH analog. Science 1988; 239:72. 59. Huhtaniemi I, Pettersson K. Mutations and polymorphisms in gonadotropin subunit genes: clinical relevance. Clin Endocr (Oxf) 1998; 48:675. 60. Catt KJ, Harwood JP, Clayton RN, et al. Regulation of peptide hormone receptors and gonadal steroidogenesis. Recent Prog Horm Res 1980; 36:557. 61. Segaloff DL, Ascoli M. The lutropin/choriogonadotropin receptor 4 years later. Endocr Rev 1993; 14:1496. 62. Simoni M, Gromoll J, Nieschlag E. The follicle-stimulating hormone receptor: biochemistry, molecular biology, physiology and pathophysiology. Endocr Rev 1997; 18:739. 63. Matsumoto AM, Karpas AE, Bremner WJ. Chronic human chorionic gonadotropin administration in normal men: evidence that follicle-stimulating hormone is necessary for the maintenance of quantitatively normal spermatogenesis in man. J Clin Endocrinol Metab 1986; 62:1184. 64. Weiss J, Axelrod L, Whitcomb RW, et al. Hypogonadism caused by a single amino acid substitution in the b subunit of luteinizing hormone. N Engl J Med 1992; 326:179. 65. Matthews CH, Borgato S, Beck-Peccoz P, et al. Primary amenorrhea and infertility due to a mutation in the b-subunit of follicle stimulating hormone. Nat Genet 1993; 5:83. 66. Layman LC, Lee E-J, Peak DB, et al. Delayed puberty and hypogonadism caused by mutations in the follicle-stimulating hormone b-subunit gene. N Engl J Med 1997; 337:607. 67. Phillip M, Arbelle JE, Segev Y, Parvari R. Male hypogonadism due to a mutation in the gene for the b-subunit of follicle-stimulating hormone. N Engl J Med 1998; 338:1729. 68. Kremer H, Kraaij R, Toledo SPA, et al. Male pseudohermaphroditism due to a homozygous missense mutation of the luteinizing hormone receptor gene. Nat Genet 1995; 9;10. 69. Latronico AC, Anasti J, Amhold I, et al. Testicular and ovarian resistance to luteinizing hormone caused by inactivating mutations of the luteinizing hormone-receptor gene. N Engl J Med 1996; 334:507. 70. Toledo SP, Brunner HG, Kraaij R, et al. An inactivating mutation of the luteinizing hormone receptor causes amenorrhea in a 46,XX female. J Clin Endocrinol Metab 1996; 81:3850. 71. Aittomki K, Dieguez-Lucena JL, Pakarinen P, et al. Mutation in the follicle-stimulating hormone receptor gene causes hereditary hypergonadotropic ovarian failure. Cell 1995; 82:959. 72. Tapanainen JS, Aittomki K, Jiang M, et al. Men homozygous for an inactivating mutation of the follicle-stimulating hormone (FSH) receptor gene present variable suppression of spermatogenesis and fertility. Nat Genet 1997; 15:205. 73. Beau I, Touraine P, Meduri G, et al. A novel phenotype related to partial loss of function mutations of the FSH receptor. J Clin Invest 1998; 102:1352. 74. Kumar TR, Wang L, Lu N, Matzuk MM. Follicle-stimulating hormone is required for ovarian follicle maturation but not male fertility. Nat Genet 1997; 15:201. 75. Dierich A, Sairam MR, Monaco L, et al. Impairing follicle-stimulating hormone (FSH) signaling in vivo: targeted disruption of the FSH receptor leads to aberrant gametogenesis and hormonal imbalance. Proc Natl Acad Sci U S A 1998; 95:13612. 76. Kremer H, Mariman E, Otten BJ, et al. Co-segregation of missense mutations of the luteinizing hormone receptor gene with familial male-limited precocious puberty. Hum Mol Genet 1993; 2:1779. 77. Shenker A, Laue L, Kosusgi S, et al. A constitutively activating mutation of the luteinizing hormone receptor in familial male precocious puberty. Nature 1993; 365:652. 78. Gromoll J, Simoni M, Neischlag E. An activating mutation of the follicle-stimulating hormone receptor autonomously sustains spermatogenesis in a hypophysectomized man. J Clin Endocrinol Metab 1996; 81:1367. 79. Snyder PJ. Gonadotroph cell adenomas of the pituitary. Endocr Rev 1985; 6:552. 80. Daneshdoost L, Gennarelli TA, Bashey HM, et al. Identification of gonadotroph adenomas in men with clinically nonfunctioning adenomas by the luteinizing hormone beta-subunit response to thyrotropin-releasing hormone. J Clin Endocrinol Metab 1993; 77:1352. 81. Daneshdoost L, Gennarelli TA, Bashey HM, et al. Recognition of gonadotroph adenomas in women. N Engl J Med 1991; 324:589. 82. Shomali ME, Katznelson L. Medical therapy for gonadotroph and thyrotroph tumors. Endocrinol Metab Clin North Am 1999; 28:223. 83. Black PM, Zervas NT, Candia G. Management of large pituitary adenomas by transsphenoidal surgery. Surg Neurol 1998; 29:443. 84. Snyder PJ, Fowble BF, Schatz NJ, et al. Hypopituitarism following radiation therapy of pituitary adenomas. Am J Med 1986; 81:457.

CHAPTER 17 HYPOPITUITARISM Principles and Practice of Endocrinology and Metabolism

CHAPTER 17 HYPOPITUITARISM
JOSEPH J. PINZONE Pathogenesis of Hypopituitarism Tumors Other Mechanical or Compressive Lesions Infarction Radiation Autoimmune Infiltrations and Infections Miscellaneous Diagnosis General Principles Clinical Presentation and Diagnostic Tests Adrenocorticotropin Deficiency Symptoms Signs Laboratory Assessment Thyroid-Stimulating Hormone Deficiency Symptoms Signs Laboratory Assessment Gonadotropin Deficiency in Men Symptoms Signs Laboratory Assessment Gonadotropin Deficiency in Women Symptoms Signs Laboratory Assessment Growth Hormone Deficiency Symptoms Signs Laboratory Assessment Prolactin Dysregulation Symptoms Signs Laboratory Assessment Anterior Pituitary Hormone Hypersecretion Vasopressin Dysregulation Oxytocin Dysregulation Conditions that Mimic Hypopituitarism Endocrine Replacement Therapy General Principles Glucocorticoid Replacement Thyroid Hormone Replacement Gonadal Steroid Replacement Testosterone Replacement Estrogen Replacement Growth Hormone Replacement Vasopressin Replacement Prolactin and Oxytocin Chapter References

The pituitary gland comprises anterior and posterior divisions. The anterior pituitary gland secretes adrenocorticotropic hormone (ACTH), thyroid-stimulating hormone (TSH), follicle-stimulating hormone (FSH), luteinizing hormone (LH), growth hormone (GH), and prolactin. The posterior pituitary gland secretes antidiuretic hormone (ADH) and oxytocin. Signs and symptoms of hypopituitarism result from reduced secretion of one or more of these hormones. Although some patients have total failure of pituitary function (i.e., panhypopituitarism), more commonly, only one or several hormones are deficient (i.e., partial hypopituitarism). Deficiency of only one pituitary hormone is called selective or isolated hypopituitarism. This chapter will deal primarily with deficiencies of anterior pituitary hormones of adults (see Chap. 25 and Chap. 26 for a more complete discussion of deficiencies of ADH and oxytocin and Chap. 18 and Chap. 198 for pediatric neuroendocrine and GH dysregulation). The most common selective deficiencies are those of GH or the gonadotropins. In progressive loss of pituitary function, as with a slowly growing pituitary adenoma, TSH, ACTH, and ADH are often the last to be diminished, not uncommonly in that order. Because the causes and manifestations of hypopituitarism are protean, the prevalence of hypopituitarism is difficult to estimate accurately. However, the major cause of hypopituitarism is pituitary tumors. Prevalence of pituitary tumors in the United States was reported to be 22,517, although this is probably an underestimate.1 Anterior pituitary hormone deficiencies lead to end-organ failure (e.g., diminished ACTH leads to decreased adrenal cortisol production). Syndromes associated with pituitary hormone insufficiency are termed secondary or central (e.g., decreased adrenal cortisol production that is due to diminished ACTH secretion is termed secondary or central hypoadrenalism). Technically, hormone deficiency syndromes resulting from hypothalamic failure should be called tertiary, but hypothalamic hormones are not measured in clinical practice. Consequently, conditions resulting from inadequate levels of either hypothalamic or pituitary hormones are referred to as secondary hormone deficiencies. In general, pituitary hormone levels vary widely throughout the day and are dependent on feedback inhibition by target organ hormones. An exception to this is the maintenance of normal prolactin levels via tonic inhibition of prolactin secretion by dopamine.2,3 An essential concept in the evaluation of pituitary dysfunction is that except for prolactin, measurement of pituitary hormone levels alone has no role in the diagnosis of pituitary hormone insufficiency. For example, a normal level of TSH does not rule out secondary hypothyroidism. It is important to remember that low thyroid hormone levels should lead to increased levels of TSH. Therefore, a normal or frankly low level of TSH in the presence of low thyroid hormone levels may be evidence of secondary hypothyroidism. A variety of symptoms and signsresulting from mass lesions (e.g., tumors) affecting the pituitary gland and from pituitary hormone deficienciesmay suggest hypopituitarism. Moreover, manifestations of hypopituitarism are often nonspecific and subtle; recognition is frequently delayed and misdiagnosis is common. The signs and symptoms of hormonal deficiency are inadequately relieved by symptomatic measures but respond well to specific hormone replacement therapy.

PATHOGENESIS OF HYPOPITUITARISM
Table 17-1 lists many of the causes of hypopituitarism. There are no reliable statistics for the percentage of patients in each subgroup, but pituitary tumors represent the largest single category, and a pituitary adenoma (see Chap. 11) is the most common single cause of hypopituitarism in adults.

TABLE 17-1. Pathogenesis of Hypopituitarism

TUMORS Pituitary adenomas originate within the sella from one of the cell types found in the anterior pituitary. They may be secretory, with a resultant increase in the serum concentration of one of the pituitary hormones and evidence of, for example, Cushing syndrome or acromegaly. Alternatively, the adenoma may be clinically nonfunctioning, without evidence of hormonal hypersecretion; however, some clinically nonfunctioning tumors secrete hormonal precursors with diminished or absent biologic activity.4 Even when an adenoma is hypersecretory, concomitant evidence of hypopituitarism is common because of the tumor effects on adjacent nontumorous pituitary tissue. Patients with microadenomas (<10 mm)may or may not have hormone deficiencies; they most commonly have diminished gonadotropin secretion.5,6 At least 30% of patients with macroadenomas (>10 mm) have a deficiency of one or more pituitary hormones, most commonly GH.1 Craniopharyngiomas(see Chap. 11) may be sellaror more often, suprasellarand frequently contain calcifications. These tumors are congenital and may cause growth retardation, diabetes insipidus (DI), or mass-related symptoms. They commonly cause hypopituitarism in children, but symptoms may not appear until adulthood.7 Other tumors can occur in the pituitary or hypothalamic area and cause compression and hypopituitarism. Among the many possibilities are meningiomas, gliomas, and metastases, especially from breast and lung cancer. Pinealomas (see Chap. 10) are worthy of special mention because they often respond well to radiotherapy. If the computed tomography (CT) or magnetic resonance imaging (MRI) scan shows a tumor in the pineal area, the serum and cerebrospinal fluid should be tested for a-fetoprotein and b-human chorionic gonadotropin, and the cerebrospinal fluid cytology should be checked. In the past, when biopsy of the pineal area was associated with a much greater risk, treatment was customarily initiated based on the previously described tests alone if the results were compatible with the diagnosis of pinealoma. However, these results were sometimes misleading. Biopsies in the pineal area can be done more safely now, and some physicians recommend that the diagnosis be confirmed by biopsy before treatment is initiated.8 The mechanisms by which pituitary tumors cause hypopituitarism include mechanical compression of normal pituitary tissue and interference with delivery of hypothalamic hormones to the pituitary via the hypothalamic-hypophysial portal system. In addition, supraphysiologic prolactin levels may diminish hypothalamic gonadotropin-releasing hormone (GnRH) secretion, resulting in decreased gonadotropin secretion and hypogonadism. A patient with a pituitary tumor of any size may have a deficiency of one or any combination of hormones. Therefore, upon diagnosis of a pituitary tumor, before therapy, the hypothalamicpituitaryadrenal (HPA) axis and hypothalamicpituitarythyroid (HPT) axis should be evaluated, and prolactin should be measured. Evaluation of other axes may be undertaken, depending on individual circumstances. Hypopituitarism caused by a pituitary tumor may be reversible; surgical removal of the tumor or shrinkage via medical therapy is often accompanied by return of normal pituitary function. However, treatment of pituitary tumors with surgery or radiotherapy may worsen or cause hypopituitarism. OTHER MECHANICAL OR COMPRESSIVE LESIONS The discovery of an enlarged sella turcica on a skull radiographic film suggests a pituitary tumor. However, the empty sella syndrome (see Chap. 11 and Chap. 20) is another common cause of an enlarged sella.9 The sella may be enlarged and the pituitary gland compressed by pressure from meningeal tissue that has herniated into the sella. Although the pituitary is atrophic and the sella cavity appears empty on radiologic studies, because there is sufficient residual functioning pituitary tissue, the patient is typically endocrinologically normal; nevertheless, hormonal deficiencies occasionally occur. A CT or MRI scan of the head can usually differentiate a pituitary tumor from the empty sella syndrome. Usually no pituitary tumor is present, but there occasionally may be a microadenoma in the residual pituitary tissue. A neurosurgical procedure in the pituitary-hypothalamic region may lead to pituitary tissue destruction and hypopituitarism.10 Severe head trauma, particularly if complicated by a basal skull fracture, may result in interruption of the pituitary stalk and consequent hypopituitarism.11 Rarely, an intrasellar carotid artery aneurysm can cause compression and destruction of pituitary tissue and hypopituitarism.12 INFARCTION The pituitary gland enlarges during pregnancy, making it more vulnerable to a reduction in blood flow. Sheehan syndrome results from ischemic pituitary necrosis, which typically follows an episode of severe hypotension during pregnancy, usually a result of major blood loss at the time of delivery or shortly thereafter.13 Symptoms are often obvious soon after delivery, but they may be delayed. The patient typically is unable to lactate postpartum and experiences amenorrhea or hypomenorrhea. Hypothyroidism usually occurs, and there may be concomitant hypoadrenalism. Sheehan syndrome has become much less common as obstetric techniques have improved. Rarely, the physician may see similar pituitary necrosis in a male or nonpregnant female patient who presumably has severely compromised pituitary blood flow from, for example, diabetes mellitus or sickle cell disease. The acute and often dramatic symptoms of pituitary apoplexy (see Chap. 11) result from hemorrhagic infarction of a pituitary adenoma; hypopituitarism may result.14 A pituitary tumor may have been diagnosed previously, or apoplexy may be its first recognized clinical manifestation. The patient may report severe headache and then become comatose; there may be an accompanying loss of vision and ophthalmoplegia from pressure on adjacent neural structures. A head CT scan shows hemorrhage in the pituitary area, and decompressive surgery may be needed on an emergency basis. Occasionally, a pituitary tumor may be totally destroyed by pituitary apoplexy, resulting in remission of hormonal and mechanical effects and cure of hormonal hypersecretion.14a An imaging study of a patient with a pituitary tumor may show evidence of old hemorrhage into the tumor, but the patient typically cannot recall any symptoms that might have been associated or may remember a severe headache without other complications. Such limited hemorrhages are much more common than the larger hemorrhages that cause the clinical syndrome of pituitary apoplexy. Pituitary infarction from any cause may be life threatening; hormone replacement therapy with glucocorticoids and thyroid hormone should not be delayed and should be initiated before diagnosis. Pituitary imaging and neurosurgical consultation should be obtained immediately. With expedient, proper management, most patients do well. RADIATION Radiotherapy (see Chap. 22), specifically to the pituitary area or as part of broader treatment of a head or neck tumor, may cause hypopituitarism.15,16 Radiotherapy-induced hypopituitarism is dose dependent and often delayed. Most patients who receive radiotherapy will go on to develop partial or complete hypopituitarism. However, the length of time from the administration of radiotherapy to the onset of pituitary hormone deficiencies is highly variable. Therefore, patients who have received any form of pituitary or head and neck radiotherapy should undergo evaluation of pituitary function at regular intervals for life. Patients should be educated about symptoms of pituitary hormone deficiency and should be instructed to contact a physician if any of these symptoms develop. Some practitioners recommend evaluation of pituitary hormone axes every 3 months for the first year after radiotherapy and then yearly thereafter.1 The regular interim visits should include evaluation of the HPA and HPT axes and measurement of prolactin. Axes that are failing but not yet in the abnormal range should be evaluated more frequently, or hormone replacement therapy should be initiated, depending on the clinical situation. In general, practitioners should have a low threshold for initiating hormone replacement therapy. AUTOIMMUNE Lymphocytic hypophysitis refers to an autoimmune process that involves the anterior pituitary gland; it is most commonly seen in women in late pregnancy or

postpartum, although it may occur in nonpregnant women or in men. There may be clinical signs of a pituitary mass or hypopituitarism, and there may be accompanying evidence of autoimmune disease elsewhere in the endocrine system.17 INFILTRATIONS AND INFECTIONS Many infiltrative or infectious processes can occasionally cause anterior or posterior pituitary dysregulation (see Table 17-1 and Chap. 11 and Chap. 213). These diseases usually present with their respective classic signs and symptoms. Occasionally, hypopituitarism is the first, only, or most severe manifestation. MISCELLANEOUS There are several miscellaneous causes of hypopituitarism, including the inevitable idiopathic classification.18 One unfortunately commonalthough potentially preventablecause is the discontinuation of hormone therapy in someone who has been receiving replacement for a known endocrine deficiency or who has been receiving pharmacologic doses for a nonendocrine disease. Patients taking glucocorticoid replacement therapy for primary or secondary hypoadrenalism have limited or no endogenous glucocorticoid production and should receive larger doses when significant physiologic stress occurs. In particular, pharmacologic (supraphysiologic) doses of glucocorticoids given as therapy for nonendocrine diseases, such as rheumatoid arthritis, may lead to adrenal atrophy and impair one's ability to produce endogenous glucocorticoids. Importantly, patients who have received pharmacologic doses of glucocorticoids for 2 weeks or more may be at risk for hypoadrenalism, even after the glucocorticoids have been discontinued. After discontinuation, these patients may have the capacity to produce sufficient amounts of endogenous glucocorticoids during normal day-to-day functioning, but have a diminished capacity to produce sufficient glucocorticoids during times of significant physiologic stress for as long as 1 year afterward (see also Chap. 78). Another category worthy of comment is Kallmann syndrome (see Chap. 16 and Chap. 115), a form of isolated gonadotropin deficiency that sometimes can be diagnosed by history alone.19 Usually, the patient consults a physician because he failed to go through puberty. If a very poor or absent sense of smell has been present since birth, it is likely that the patient has Kallmann syndrome, a congenital and sometimes familial form of hypogonadotropic hypogonadism, which is caused by a developmental failure of neuronal migration of the olfactory neurons and of the neurons responsible for GnRH secretion. Aplasia of the olfactory gyri and absence of the olfactory bulbs and tracts may be detected by MRI scan. This syndrome is not accompanied by a pituitary tumor or by clinical failure of the other pituitary functions. The physician must specifically inquire about the sense of smell, because the patient usually does not volunteer this seemingly insignificant detail. (Because the anosmia or hyposmia is congenital in Kallmann syndrome, a later onset might indicate a tumor, trauma, or sinusitis.) Selective deficiency of GH may occur as a familial or sporadic phenomenon (see Chap. 12 and Chap. 198).

DIAGNOSIS
GENERAL PRINCIPLES The history and physical examination (Fig. 17-1; see Table 17-1 and Table 17-2) will usually help in the selection of subsequent laboratory studies. Some hormonal deficiencies or even a mass lesion may be inapparent from the history and physical examination alone and can be detected only by an appropriate laboratory and radiologic survey. The laboratory evaluation for hypopituitarism can be divided into initial screening tests and supplemental tests that should be considered if the initial results are abnormal (Table 17-3).

FIGURE 17-1. Hypopituitarism. A 55-year-old man with recent-onset hypopituitarism caused by a clinically nonfunctioning pituitary tumor. The skin is pale, and there is an almost total lack of body hair. Test results revealed a deficiency of gonadotropins and thyroid-stimulating hormone.

TABLE 17-2. Symptoms and Signs of Hypopituitarism

TABLE 17-3. Laboratory Evaluation of Hypopituitarism

The most effective initial approach to detecting adrenal or gonadal hypofunction is to assay the target gland hormones, even if the underlying defect is thought to be at the pituitary level. The serum concentrations of the target gland hormones are less variable than pituitary hormone levels and are generally more useful for differentiating subnormal from normal target gland function. If any of these target hormone values are diminished, the serum levels of the corresponding pituitary hormones should be measured. Because primary hypothyroidism is commonly seen in clinical practice, and because serum TSH is a very sensitive test used to diagnose primary hypothyroidism, many clinicians are in the habit of measuring TSH to diagnose hypothyroidism. However, if secondary hypothyroidism is suspected, the target gland hormone (free thyroxine [FT4]) should be measured. Evaluation of GH deficiency in adults is typically not undertaken unless definite pituitary pathology

is documented. Serum prolactin is usually measured as part of any evaluation for hypogonadism or hypopituitarism. If a target gland hormone is definitely deficient, and the corresponding pituitary hormone concentration is clearly elevated, the diagnosis is primary target-gland failure. If, however, the pituitary hormone value is subnormal, this is consistent with deficiency at the level of the pituitary or hypothalamus. In the presence of frank target gland failure, a normal pituitary hormone concentration in the serum may represent biologically hypoactive hormone.20 The tests recommended in Table 17-3 usually provide good discrimination between normal and abnormal endocrine states, but false-positive and false-negative results are seen, often in conjunction with certain medications and diseases. Table 17-4 lists the factors that can lower hormone values and cause false-positive diagnoses, and those that raise the hormone values and could obscure a true-positive result. Four general mechanisms account for most of this interference. First, the target gland hormones of the thyroid, adrenal gland, and gonads circulate partially bound to specific carrier proteins; any medication or disease that changes the hepatic synthesis of a carrier protein also changes the concentration of total hormone, even if the hormonally active free fraction remains constant throughout. Second, hypersecretion or hyposecretion of one pituitary hormone may cause a change in the concentration of another pituitary hormone. The latter abnormality disappears when the former is corrected by appropriate therapy. The suppression of gonadotropins by hyperprolactinemia is one example. Another is the increase of serum TSH concentration that occasionally can be seen in untreated primary or secondary hypoadrenalism.21 Third, diseases that affect urinary excretion (e.g., uremia, nephrotic syndrome) or alter hepatic degradation (e.g., various liver diseases, hypothyroidism)as well as severe, stressful illnessmay alter hormone concentrations. The euthyroid sick syndrome, common in intensive care units or among other similarly stressed patients, mimics the laboratory findings of secondary hypothyroidism.22 Fourth, some medications are a potential source of error. Phenytoin and high doses of salicylates can bind to the hormonal binding sites on thyroxine (T4) binding proteins; this results in a decrease in total T4 concentration, but the FT4 determination remains unaltered. The information summarized in Table 17-4 can alert the physician to most cases of potential misinterpretation. However, if the clinical picture and the test results do not initially seem to agree, the prudent course is to review both carefully before rejecting either.

TABLE 17-4. Some Causes for Decreased or Increased Hormone Values Other Than Intrinsic Disease of the Respective Endocrine Gland

All patients with hypopituitarism should have an MRI or CT scan (see Chap. 20). In many cases, an MRI study is preferable, because it provides better anatomic detail, may detect a microadenoma that is not visible on a CT scan, is better able to demonstrate whether a mass lesion is impinging on the chiasm, and can better differentiate an aneurysm from other mass lesions. However, an MRI is more expensive than a CT scan, and in some clinical situations a CT scan is sufficient (e.g., if the only purpose is to rule out a large tumor that would be easily seen with either technique). With current methods, some small pituitary tumors cannot be detected by either imaging method. Absence of a visible mass does not definitively rule out a tumor, and periodic follow-up examinations may be indicated. When MRI or CT is performed, it should be a dedicated study with thin sections through the sellar and suprasellar area, performed with and without contrast.1,23 If a pituitary mass is found, neuroophthalmologic examination should be conducted by an ophthalmologist, including the determination of visual fields by computerized threshold or Goldmann perimetry (see Chap. 19). Periodic follow-up examinations also may be indicated, as after surgical removal of a pituitary tumor to determine whether there has been a change from the preoperative examination or as one of several ways to monitor a pituitary mass lesion for possible progression. Additional tests may be indicated when certain specific diagnoses are being considered. In a patient with DI and a negative MRI result with no obvious underlying cause, sarcoidosis should be considered, and a cerebrospinal fluid examination for protein; a chest radiograph; and hepatic, bone marrow, or lymph node biopsy should be performed. Rarely, hemochromatosis (see Chap. 131) or other infiltrative or infectious disease may manifest as hypopituitarism (see Table 17-1). The physician should consider such a possibility and carry out appropriate screening tests if there are clues suggesting one of these conditions. In a patient with a known pituitary or hypothalamic tumor, some preoperative endocrine testing is indicated, although the neurosurgery may result in further changes in endocrine status (Table 17-5). Hypothyroidism and hypoadrenalism should be sought. Untreated hypothyroidism increases the risk of an operative procedure; it is prudent to consider delaying pituitary exploration until the hypothyroidism is safely corrected.24

TABLE 17-5. Management of Hypopituitarism: Intercurrent Illnesses, Long-Term Follow-Up, and Perioperative Care at the Time of Pituitary Surgery

If hypoadrenalism and hypothyroidism coexist, the hypoadrenalism should be treated before, or along with, any hypothyroidism; restoration to a euthyroid state without correction of concomitant hypoadrenalism sometimes precipitates acute hypoadrenalism with adrenal crisis. If a pituitary tumor is clinically nonfunctioning, the physician may wish to obtain additional tests preoperatively to determine if the tumor is secreting one or more hormonal products that may be clinically silent. Prolactin, FSH, LH, and the a-Subunit are all substances that under certain circumstances may fail to produce associated signs or symptoms. The a-subunit is a polypeptide chain that is part of the structures of FSH, LH, and TSH; the serum concentration of free a-subunit is increased above normal if a pituitary tumor secretes this peptide. If hypersecretion of any of these products is found, another measurement later in the course can serve as an additional criterion of the completeness of surgical resection or tumor recurrence. It is especially important to measure prolactin because the first-line therapy of a prolactin-secreting adenoma is usually dopamine agonist therapy and not surgery (see Chap. 13). After pituitary surgery, close endocrinologic and neurosurgical evaluation is required (see Chap. 23 and Table 17-5). The immediate postoperative period should include measurement of serum sodium and urine-specific gravity at regular intervals, at first several times per day and then less frequently, for as long as 2 weeks after surgery. This regimen should diagnose immediate or delayed onset of DI or the syndrome of inappropriate ADH (SIADH); these conditions may occur with any manipulation of the pituitary or hypothalamus (see Chap. 25 and Chap. 208). Stress doses of glucocorticoid supplementation should have been initiated preoperatively and may be tapered as the patient improves postoperatively. However, maintenance doses of glucocorticoids should continue until adequate functioning of the HPA axis is confirmed. Obviously, replacement of any target gland deficiencies should continue postoperatively. The first follow-up appointment should occur as early as needed, but definitely within the first month after discharge from the hospital. Intervals between subsequent visits during the first year after surgery can be individualized. Each postoperative visit should include a history and physical examination designed to determine the presence of hypopituitarism. Evaluation of HPA and HPT axis function and serum sodium should occur multiple times during the first postoperative year. Evaluation of HPA axis function and management of glucocorticoid therapy after surgery for Cushing disease or Cushing syndrome may be quite complex (see Chap. 14, Chap. 23 and Chap. 75). Patients cured of Cushing disease or Cushing syndrome may require a protracted course of glucocorticoid supplementation. Medical therapy of Cushing syndrome may render the

patient hypoadrenal, and replacement glucocorticoids and/or mineralocorticoids may be necessary. The initial evaluation of hypopituitarism (including stimulation and suppression tests, 24-hour urine collections, visual field testing, and head CT or MRI scan) often can be conducted on an outpatient basis, unless the patient is acutely ill at the time of presentation. It is best to start with simple tests of baseline target-organ secretion and then to be selective in applying the many available additional tests. If a pituitary tumor is discovered and surgery is planned, the initial hormonal survey can be limited, and more extensive evaluation can be deferred until after surgery, at which time the need for long-term replacement therapy is determined. Patients with hypopituitarism should have laboratory tests repeated annually or more often, depending on the clinical situation. In progressive forms of hypopituitarism, testing enables the detection of new deficits before they become symptomatic, and testing also can be used to monitor the adequacy of established hormonal regimens. The evaluation of pituitary and end-organ function is more completely discussed elsewhere (see Chap. 33, Chap. 74 and Chap. 114). For long-term surveillance, if hypopituitarism has been caused by a definite or suspected mass lesion, annual visual field testing is usually indicated. The frequency of head CT scanning or MRI is individualized, because both tests are expensive and the former involves radiation and iodide-contrast exposure. During the year immediately after removal of a mass lesion, the neurosurgeon or endocrinologist may wish to obtain more than one scan. If the lesion remains stable, the frequency of MRI or CT scans may revert to annual examinations for 1 or 2 years and then less fequently, unless there is some clinical suggestion of a recurrence. If a secretory tumor has been resected, a later rise in the serum concentration of the relevant hormone is one reason to suspect a recurrence and to repeat the CT or MRI scan. It is important to recognize when symptoms are not caused by hypopituitarism. Patients with proven hypopituitarism may also develop other, more common conditions with symptoms that may be confused with the nonspecific symptoms of endocrine disease. For instance, anxiety and depression are common, and patients with hypopituitarism are not immune to such problems. In the presence of a new symptom, it should not be automatically assumed that the hormonal regimen should be changed. If endogenous serum hormone concentrations are normal, if the patient is already at the conventional upper limit of hormonal drug dosage, or if the patient has previously been stable on that dosage for many months, another diagnosis should be sought. A careful history and physical examination should be obtained, and hormone measurements should be repeated and other pertinent tests performed before making any medication changes. CLINICAL PRESENTATION AND DIAGNOSTIC TESTS Patients with hypopituitarism first seek medical attention for relief of symptoms and signs related to (a) a space-occupying lesion in the sella turcica or parasellar region, and/or (b) pituitary hormone dysregulation (see Table 17-2). SPACE-OCCUPYING LESION OF THE SELLA TURCICA OR PARASELLAR REGION In general, the larger the lesion in the sella turcica or parasellar region, the more likely is the lesion to compress surrounding structures, leading to pressure-related phenomena. A patient with a functioning pituitary tumor may present to a clinician with symptoms or signs of hormone excess (see Chap. 12, Chap. 13, Chap. 14, Chap. 15 and Chap. 16), pressure symptoms, and/or symptoms of diminished pituitary hormone secretion. Most commonly, a patient with a nonfunctioning pituitary tumor or other space-occupying lesion first seeks medical attention for pressure symptoms, even when hormonal deficiencies antedate these phenomena. Headache, caused by traction on the meningeal membranes, is a common manifestation of a space-occupying sellar lesion. Often, such headaches are nonpulsatile, dull and poorly localized, and difficult or impossible to differentiate from many other forms of headache. Routine radiologic studies cannot be recommended to look for the small percentage of headache patients with pituitary tumors, unless there are additional clues such as new onset, progressive worsening, or incapacitating headaches; or an accompanying hormonal, ophthalmologic, or neurologic abnormality. Space-occupying lesions may compress the optic chiasm and lead to reports of diminished visual acuity or decreased peripheral vision. Further evaluation during the physical examination or by more formal methods of visual field testing may reveal deficits suggesting chiasmal compression (e.g., bitemporal hemianopia, superior lateral quadrantanopia; see Chap. 19).25 Extraocular muscle paresis caused by pressure on the cranial nerves or nuclei lateral to the sella is seen occasionally. A large lesion can cause hypothalamic symptoms (see Chap. 9) or result in anosmia, seizures, or even symptoms of frontal or temporal lobe dysfunction. Stroke and subarachnoid bleeding are rare but more acute manifestations associated with very large sellar lesions and pituitary apoplexy, both of which may interrupt the sellar or parasellar vasculature. PITUITARY HORMONE DYSREGULATION When a patient is suspected of having definite hypoadrenalism, hypothyroidism, or hypogonadism, the corresponding pituitary hormone should be assayed to determine whether the deficiency is primary (i.e., disease of the target gland itself) or secondary (i.e., insufficient stimulation of the target gland by the corresponding pituitary hormone). If the relevant pituitary hormone concentration is not clearly elevated, as it should be in the absence of feedback control, but instead is in the normal or subnormal range, a pituitary or hypothalamic cause must be suspected. To more accurately diagnose hypopituitarism, the pituitary hormone value should be checked before hormonal replacement is begun, because the exogenous target gland hormone would suppress the corresponding pituitary hormone. However, treatment, if urgent, need not always wait for the result. Each anterior pituitary hormone is part of a tightly regulated axis composed of hypothalamic factors and target organ factors. Each component of the axis is involved in feedback regulation of the other components (see Chap. 12, Chap. 13, Chap. 14, Chap. 15 and Chap. 16). Pituitary hormones can be diminished alone or in combination; therefore, consideration of hypothalamic-pituitary target organ axes individually will provide a logical framework for diagnosing the complex and overlapping manifestations of hypopituitarism.

ADRENOCORTICOTROPIN DEFICIENCY
The HPA axis is the most crucial and may be the most challenging pituitary axis to evaluate, because symptoms and signs of ACTH and cortisol deficiency are often nonspecific. Selective ACTH deficiency is rare; usually, other pituitary hormones are deficient as well.26 SYMPTOMS Physiologic effects of cortisol are protean. Consequently, cortisol deficiency often leads to subtle, nonspecific manifestations, but may be life threatening. Symptoms include weakness, fatigue, weight loss, and diminished sense of well-being. If a patient has neuroglycopenic symptoms and documented fasting hypoglycemia, evaluation for ACTH deficiency is indicated, along with a search for other causes (see Chap. 158 and Chap. 161). Abdominal distress, including nausea and vomiting, is often noted. A history of recently discontinued glucocorticoids should be sought. SIGNS Protracted secondary ACTH deficiency may lead to pale skinincluding nipples and areolaewith decreased ability to tan. Diminished axillary and pubic hair may occur, especially in women. Postural hypotension is often seen; the pulse rate may be inappropriately normal or slow if concomitant hypothyroidism is present. With more acute hypoadrenalism, or during time of significant physiologic stress, patients may present with shock, which can be resistant to therapy. LABORATORY ASSESSMENT The HPA axis comprises a tightly regulated feedback loop. The hypothalamus secretes corticotropin-releasing hormone (CRH) into the portal circulation, which is transported to the pituitary. There it acts to stimulate ACTH secretion into the peripheral circulation. ACTH stimulates cortisol secretion from the adrenal glands. Cortisol feeds back to inhibit CRH and ACTH secretion. Most cortisol is bound to cortisol-binding globulin (CBG), although the much smaller free or unbound component is the bioactive fraction. Factors that alter CBG production may affect total serum cortisol levels, which may not accurately reflect the free cortisol fraction. Estrogen stimulates hepatic production of CBG, whereas cirrhosis, nephrotic syndrome, and hyperthyroidism may lower CBG. However, in most clinical contexts, the total serum cortisol level is used to determine whether the adrenal glands are functioning properly. Many glucocorticoid preparations cross-react with the cortisol assay and should be avoided within 24 hours of testing. In secondary hypoadrenalism, ACTH deficiency leads to decreased cortisol production by the adrenals; however, mineralocorticoid production typically remains adequate. Importantly, with most causes of primary hypoadrenalism, mineralocorticoid insufficiency is also present. Serum cortisol levels in normal persons and patients with hypoadrenalism often overlap considerably. Normally, serum cortisol levels demonstrate a diurnal variation. Peak cortisol levels usually occur between 6:00 and 8:00 a.m. and nadir at around 10:00 p.m. to midnight. With secondary hypoadrenalism, this diurnal variation may be maintained, but cortisol levels will be lower at any given time throughout the day. If CRH or ACTH secretion is abruptly interrupted, serum cortisol will drop to very low levels, often within hours. Therefore, pituitary apoplexy, Sheehan syndrome, and postoperative ACTH deficiency should be treated as emergent, potentially life-threatening situations. If hypoadrenalism is suspected and the patient is hypotensive, therapy with hydrocortisone, 100 mg intravenously every 8 hours, should not be delayed for diagnostic purposes. However, in this situation, measurement of the serum cortisol and ACTH level just before administration of glucocorticoids will be helpful diagnostically. Although the literature is

often conflicting on this subject, one review provided the following guidelines: While a patient is hypotensive, a serum cortisol level 18 g/dLis evidence of adequate HPA axis function. A level of 13 to 18 g/dL is indeterminate, and hydrocortisone administration should be continued until further testing can be carried out. A serum cortisol level of 5 to 13 g/dL is presumptive evidence of hypoadrenalism, and a level <5 g/dL is regarded as definite evidence of hypoadrenalism.27 If hypoadrenalism is diagnosed, a high ACTH level suggests primary hypoadrenalism, whereas a low or normal ACTH level suggests secondary hypoadrenalism. With patients who are more stable, a variety of diagnostic tests can be utilized to determine whether hypoadrenalism exists. With each of these tests, an ACTH level may be drawn depending on the clinical context. Here, too, a frankly high ACTH level in a hypoadrenal patient would argue for primary hypoadrenalism, whereas a normal or low level would argue for secondary hypoadrenalism. The simplest test of the HPA axis is a cortisol level drawn between 6:00 and 8:00 a.m. A value of 19g/dL indicates an intact HPA axis, whereas a level <3g/dLis definite evidence of hypoadrenalism.27 A value of 3 to 19g/dLis indeterminate and requires further testing. Because most patients with or without hypoadrenalism will have values in the indeterminate range, this test is of limited value. The cosyntropin stimulation test is a convenient, safe, and generally reliable dynamic test of the HPA axis. Cosyntropin, the synthetic bioactive portion of ACTH, is injected intravenously at a dose of 250 g, and serum cortisol is measured at 0, 30, and 60 minutes. A cortisol level 18g/dLat any point indicates that the adrenal glands can be stimulated adequately.27 In most situations, this means that the HPA axis is intact. A lower value might indicate secondary hypoadrenalism. (See Chap. 76 and Chap. 241 for the use of the ACTH test in primary hypoadrenalism [Addison disease].) It should be noted that if secondary hypoadrenalism has occurred within the recent past (usually taken to mean the preceding few weeks), atrophy of the adrenal gland, normally seen with ACTH deficiency, may not have had sufficient time to occur. Consequently, endogenous adrenal production of cortisol may be insufficient because of lack of endogenous ACTH, whereas pharmacologically augmented cortisol production using intravenously administered cosyntropin may still be possible. Cosyntropin, 250 g intravenously, is a supraphysiologic stimulus of cortisol production. This has led to interest in the use of ACTH 1 g intravenously in place of a 250-g dose as being possibly a more discriminating test of HPA axis integrity.28 The most reliable dynamic test of HPA axis function is the insulin tolerance test (ITT). The rationale is that severe hypoglycemia produces maximal physiologic stress and should stimulate the entire HPA axis leading to augmented cortisol production. The test is carried out by administering regular insulin 0.1 to 0.15 U/kg intravenously and measuring serum cortisol and blood glucose levels at 0, 30, and 60 minutes. Blood glucose at 30 and/or 60 minutes must be <40 mg/dL for the test to be valid. A cortisol level 18g/dL indicates an intact HPA axis, whereas a value <18 g/dL may indicate secondary hypoadrenalism. This test is contraindicated in elderly patients and with patients who have seizures or those with cardiovascular or psychiatric disease. A physician must be present and the intravenous catheter must remain in place with a 50% dextrose solution available in case of hypoglycemia. The metyrapone test, like the ITT, is a test of the entire HPA axis. Metyrapone blocks 11b-hydroxylase activity in the adrenal gland; 11b-hydroxylase catalyzes the conversion of 11-desoxy-cortisol (11-S) to cortisol in the final step of the cortisol biosynthetic pathway. Therefore, if metyrapone is given to a normal individual, cortisol production decreases; this stimulates the pituitary to secrete ACTH. ACTH then stimulates the cortisol biosynthetic pathway in the adrenal gland. However, because metyrapone is blocking the final step, 11-S accumulates. Proper functioning of the HPA axis is confirmed by an 11-S value >7.0 g/dL. In a patient with secondary hypoadrenalism, diminished cortisol does not lead to a rise in ACTH, and the 11-S level remains 7g/dL. The metyrapone test requires an inpatient stay and may precipitate symptomatic adrenal insufficiency. A less reliable test of hypoadrenalism, the urine free cortisol (UFC), measures free cortisol in a 24-hour urine collection. This test measures only the amount of cortisol that exceeds the binding capacity of CBG and is excreted in the urine; UFC levels are within the reference range in 20% of patients with hypoadrenalism.29 In contrast, UFC is a good screening test for cortisol excess (see Chap. 14).

THYROID-STIMULATING HORMONE DEFICIENCY

Symptoms and signs of secondary hypothyroidism are similar to those seen with primary hypothyroidism, although they often are milder. The essential diagnostic differences relate to interpretation of the results of the laboratory assessment. SYMPTOMS Cold intolerance, constipation, fatigue, and lethargy are common symptoms. Weight gain despite diminished appetite may occur. Patients or family members may note a slowing of cognitive and motor function as well as a deepening and hoarseness of the voice. Muscle cramping, including symptoms of carpal tunnel syndrome, is sometimes present. Women of childbearing years may report menorrhagia. SIGNS On physical examination, a myxedematous appearance with pale, cool skin that appears dry and doughy, and a large protruding tongue may be noted. Pulse rate is often slow and pulse pressure narrowed. Hair may be dry and sparse. Relaxation phase of the reflexes is prolonged. With more advanced hypothyroidism, hypothermia, hypoglycemia, stupor, and respiratory depression may, if left untreated, lead to death. In addition, children may demonstrate other age-dependent manifestations. Congenital hypothyroidism, although rarely of pituitary origin, may lead to cretinism (see Chap. 47). Linear growth may be inhibited in children and adolescents, leading to short stature. LABORATORY ASSESSMENT The hypothalamus releases thyrotropin-releasing hormone (TRH) into the portal circulation. TRH is transported to the pituitary gland, where it causes release of TSH into the peripheral circulation. TSH binds to the thyroid, leading to release of T4 and, to a lesser extent, the bioactive thyroid hormone, triiodothyronine (T3). T4 is converted to T3 in peripheral tissues, including the pituitary. Both T4 and T3 exist primarily bound to thyroid-binding globulin and other blood proteins. The free fractions of T4 and T3 are the clinically important components. Many laboratories now utilize direct measurement of FT4, which is the preferable way to measure serum T4. However, sometimes the total T4 and T uptake are used to calculate a FT4 index (FT4I), which estimates the FT4 by taking into account variations in binding protein levels. In secondary hypothyroidism, TSH deficiency leads to low levels of FT4. Therefore, an FT4 (or FT4I), as well as a sensitive TSH assay, must be obtained. The FT4 will be low, whereas TSH will be normal or subnormal. In contrast, with primary hypothyroidism, FT4 is diminished but TSH is elevated as a compensatory response. The T3 level can be normal or near normal in early hypothyroidism and is not usually measured as part of the initial workup. The FT4 can differentiate hypothyroidism from most nonthyroidal laboratory abnormalities.20 However, if the clinical setting suggests the possibility of the euthyroid sick syndrome (see Chap. 36), the FT4 may be low, with a normal or subnormal TSH. This is thought to be a compensatory response that down-regulates metabolism during physiologic stress, thereby preventing excessive catabolism. The underlying illness that is causing euthyroid sick syndrome should be treated; therapy with thyroid hormone is not recommended and may be harmful. In this situation, additional tests may be helpful; nevertheless, the clinician should be cautious in diagnosing hypothyroidism secondary to pituitary disease when a patient is experiencing a physically stressful illness or a period of caloric deprivation. Rarely, a patient with long-standing primary hypothyroidism may develop significant pituitary enlargement and even chiasmal symptoms. Presumably, this occurs from chronic stimulation of the pituitary gland by the negative-feedback mechanism.30 Furthermore, a patient rarely may have a pituitary tumor that secretes bioactive TSH or may have resistance to thyroid hormone. Both of these conditions can result in elevated or normal serum concentrations of TSH with concomitantly elevated thyroid hormone levels (see Chap. 21 and Chap. 32). A stimulation test with TRH is sometimes proposed to help differentiate secondary from tertiary (hypothalamic) hypothyroidism, but normal and abnormal results overlap considerably. Hence, this test result is often equivocal and should not be relied on as the sole criterion (see Chap. 33).

GONADOTROPIN DEFICIENCY IN MEN

Most manifestations of decreased gonadotropin secretion in men are attributable to the resultant low testosterone levels. However, testosterone is converted to estrogens, and some manifestations of low LH and FSH may be due to low estrogen levels. Symptoms and signs of hypogonadism in men may be present for years before being appreciated and often are mistakenly attributed to normal aging. The patient's sexual partner may be the first to notice and, when appropriate, should also be queried about symptoms and signs.

SYMPTOMS In men with secondary hypogonadism, decreased libido, impotence, and infertility are common. The patient may note that he shaves less frequently and that there is a decline in the rate of progression of male pattern hair loss. With long-standing hypogo-nadism, fatigue and a loss of muscularity may be noted. A history of hyposmia or anosmia may be elicited from patients with Kallmann syndrome. If symptoms begin before or during puberty, the patient may demonstrate lack of deepening of the voice, no need for deodorant, and concerns about social acceptance. In some cases, it may be difficult to differentiate hypogonadotropic hypogonadism from delayed puberty (see Chap. 7, Chap. 91 and Chap. 92).31 SIGNS Evidence on physical examination includes pale skin with decreased axillary, facial, and pubic hair. Small, soft testes and a small prostate gland may be noted. In cases where onset was before or during puberty, a small penis and eunuchoid body habitus may be appreciated. Osteopenia and osteoporosis may be associated with hypogonadism from any cause (see Chap. 64). LABORATORY ASSESSMENT GnRH is secreted in a pulsatile fashion into the pituitary portal circulation and is the major regulator of LH secretion. For every pulse of GnRH secreted into the portal circulation, a resultant pulse of LH is secreted into the peripheral circulation. Peripheral LH stimulates testosterone production by the testes. Along with other factors, GnRH also regulates FSH secretion. Spermatogenesis is dependent on testosterone and FSH. The initial test of this axis should be a serum total testosterone or a serum free testosterone. Testosterone, like T4, is partially bound to serum proteins, and any alteration in the binding proteins also changes the total testosterone level. If a total serum testosterone is obtained initially and is found to be abnormal, this finding should be confirmed with a serum free testosterone measurement (see Chap. 114 for a discussion of free testosterone measurements).32 Moreover, a low serum testosterone measurement should be followed up with a serum LH and FSH determination. In secondary hypogonadism, LH and FSH levels will be in the normal or subnormal range despite a low total or free testosterone level. In primary hypogonadism, total testosterone and free testosterone levels are low, but LH and FSH levels are elevated as a compensatory response. When viewed by adult norms, normal children have hypogonadotropic values of LH and FSH, and these assays are usually omitted prepubertally. (Hyperprolactinemia is a possible cause of hypogonadotropic hypogonadism, and a determination of prolactin should be part of any laboratory assessment of hypogonadism.) Because a stimulation test with GnRH rarely provides clinically useful information for the individual patient, the author does not currently recommend it as part of routine management (see Chap. 16). If desired, in adults, testicular function can be evaluated with a semen analysis. Moreover, routine blood work in hypogonadal men may demonstrate a normocytic anemia.

GONADOTROPIN DEFICIENCY IN WOMEN

Disrupted menstrual function, a very early manifestation of gonadotropin deficiency, often prompts premenopausal women to seek medical attention earlier in the course of hypogonadism than do men. SYMPTOMS In women with secondary hypogonadism, menstrual dysfunction that manifests as dysfunctional uterine bleeding, oligomenorrhea, amenorrhea, or infertility is common. Diminished libido may occur. With persistence of hypogonadism, atrophy of breast tissue and of the vagina and labia may be noted. The latter may lead to dyspareunia. If prolactin is increased, galactorrhea may result. With Sheehan syndrome, failure to lactate postpartum may be noted. Hypogonadism with onset before or during puberty may result in delayed puberty; reports of lack of development of secondary sexual characteristics, including delayed menarche, and concerns about social acceptance may be present. SIGNS Physical examination may reveal decreased axillary and pubic hair, and breast and genital atrophy (as well as expressible galactorrhea, if hyperprolactinemia is present). Osteopenia and osteoporosis are associated with hypogonadism from any cause (see Chap. 64). Signs of hypogonadism in adolescent girls include persistence of prepubertal body habitus with delayed breast, genital, and pubic hair development. LABORATORY ASSESSMENT Peripheral LH stimulates estradiol and to a lesser extent testosterone production by the ovary. Along with other factors, GnRH also regulates FSH secretion. Ovulation is dependent on estradiol and FSH. In general, no laboratory assessment is necessary if the patient reliably describes entirely normal menstrual cycles; the hormone assays usually are normal if the menses are normal. One exception to this is that hyperprolactinemia may be present even if menses are normal. If the menses are not perfectly normal, a serum estradiol value should be obtained, and if it is low, the physician also should obtain serum FSH, LH, and prolactin concentrations. Postmenopausal women suspected of having hypogonadism should have serum FSH, LH, and prolactin measured. Postmenopausal women typically have increased FSH and LH levels; thus, normal or subnormal values are consistent with secondary hypogonadism. Increased prolactin is associated with many causes of hypopituitarism.

GROWTH HORMONE DEFICIENCY

Extensive research has led to improved characterization of the consequences of GH deficiency in adults and to a better understanding of the risks and benefits of treatment. Patient subgroups include those who have childhood-onset GH deficiency that persists into adulthood and those who have adult-onset GH deficiency. Although the presentation and response to therapy overlap considerably, variation between the subgroups has been noted.33 The adult-onset subgroup frequently has other concomitant pituitary hormone deficiencies. Both subgroups may exhibit a varied constellation of symptoms and signs, and often have difficult diagnostic and management issues. GH replacement is expensive, requires many resources, and provides variable although often profoundbenefit to adult patients. Therefore, it is important to identify patients at risk for GH deficiency and prudently utilize confirmatory diagnostic tests on the patients who might be appropriate candidates for GH replacement. Because of the complexity of the diagnostic and management issues surrounding GH deficiency in adults, it is advisable to have an endocrinologist involved throughout the care of these patients. SYMPTOMS Adult patients with GH deficiency may report increased abdominal adiposity and reduced muscle mass that leads to decreased strength, reduced vitality and energy, and a diminished exercise capacity. Impaired psychological well-being is a major component of the GH deficiency syndrome in adults. A depressed and sometimes labile mood, anxiety, and social isolation are common symptoms.34 Children with GH deficiency may report short stature compared with peers. A history of childhood GH deficiency should be sought in all patients with suspected hypopituitarism. SIGNS Signs are often very subtle and nonspecific. Affect may be depressed or labile. Patients may be obese with centrally distributed fat. Skin may be thin and dry, and extremities may be cool because of poor venous circulation.34 Parental heights may be used to calculate the predicted height of the patient; this may afford an objective criterion on which to base a diagnosis of short stature in cases of childhood-onset GH deficiency (see Chap. 7). LABORATORY ASSESSMENT The hypothalamus secretes GH-releasing hormone into the pituitary portal circulation. GH-releasing hormone is transported to the pituitary, where it stimulates pulsatile GH secretion into the peripheral circulation. GH acts at liver and other tissues to induce insulin-like growth factor-I (IGF-I) synthesis and secretion. IGF-I is thought to mediate most of the metabolic and growth-enhancing effects of GH (see Chap. 12). GH levels vary significantly throughout the day, and considerable disagreement exists about what constitutes laboratory confirmation of GH deficiency in adults. One

definition of GH deficiency, in a patient with a compatible history, is a negative response to a standard GH stimulation test. A negative response may be taken to mean a peak serum GH of <5 ng/mL when measured by radioimmunoassay using a polyclonal antibody or <2.5 ng/mL when measured by immunoradiometric assay using monoclonal antibodies.35 Because hypoglycemia stimulates GH secretion, the ITT is a reliable way to detect GH deficiency in adults. Serum samples for GH and glucose are collected every 15 to 30 minutes for 90 minutes after administration of insulin. The same prior-mentioned caveats and patient contraindications apply to the ITT when it is utilized to evaluate for GH deficiency. Other provocative testing agents have been used, including clonidine and arginine. For an outline of GH stimulation tests and a discussion of diagnostic criteria for GH deficiency in children, see Chapter 12, Chapter 18 and Chapter 198. A serum IGF-I level is not a sensitive and specific test on which, alone, to base the diagnosis of GH deficiency. There are many factorsincluding a complex set of binding proteins, age, and genderthat determine serum IGF-I levels. A low serum IGF-I value is helpful but not diagnostic, and a normal value does not rule out GH deficiency. Osteopenia and osteoporosis are more prevalent in adults with GH deficiency; bone densitometry may be helpful as corroborative diagnostic information and to guide therapy. Measurement of plasma lipid and lipoprotein levels often reveals a profile consistent with an increased risk of coronary artery disease.34 Formal muscle strength, exercise testing, and body composition testing is individualized.

PROLACTIN DYSREGULATION
Prolactin elevation may occur because of oversecretion by a pituitary adenoma that may or may not be detectable on pituitary imaging. Moreover, several medications and certain physiologic processes may increase prolactin (see Table 17-4); a slightly elevated value should be repeated (see Chap. 13).36 Rarely, serum prolactin may be elevated because of macroprolactinemia.37 In this condition, the serum contains a large-molecular-mass form of prolactin that is immunoreactive but without significant bioactivity. If the serum prolactin level is >200 ng/mL but no pituitary tumor is visible by appropriate MRI, this possibility should be considered. SYMPTOMS Hyperprolactinemia may be asymptomatic; an elevated serum value may be a clue to the presence of a pituitary tumor or pituitary stalk impingement.36 However, prolactin elevation from any cause may result in signs and symptoms of hypogonadism in men or women.37,38 A history of osteopenia or osteoporosis may be present. SIGNS If hyperprolactinemia results in hypogonadism, the usual signs in men and women may occur (see respective sections of this chapter). Gynecomastia may occur in men. Galactorrhea occurs more commonly in women because estrogen priming of breast tissue facilitates this process. LABORATORY ASSESSMENT Serum prolactin should be measured as part of any general screening for pituitary dysfunction (see Chap. 13). A persistently elevated prolactin value, without a pharmacologic, physiologic, or nonpituitary pathologic explanation (see Table 17-4) should be considered abnormal and likely due to a prolactin-secreting pituitary adenoma, even if pituitary imaging studies fail to demonstrate an abnormality. A moderately elevated serum prolactin may occur in disease or injury involving the hypothalamus. For serum prolactin, the upper limit of the normal range depends on which assay is used, but is typically 15 ng/mL for men and 20 ng/mL for women. Pituitary microadenomas that secrete prolactin (microprolactinomas) and macroadenomas that do not secrete prolactin but impinge on the pituitary stalk are typically associated with serum prolactin levels <250 ng/mL.36 In contrast, macroadenomas that secrete prolactin (macroprolactinomas) are typically associated with prolactin levels >250 ng/mL.36 Low prolactin levels are rare but may lead to failure of postpartum lactation.

ANTERIOR PITUITARY HORMONE HYPERSECRETION

If the serum concentration of a pituitary hormone is shown to be inappropriately elevated (not just appropriately increased in response to a target gland deficiency), there may be a pituitary tumor present and concomitant hyposecretion of other pituitary hormones. Patients with symptoms of the amenorrhea-galactorrhea syndrome (see Chap. 13), acromegaly (see Chap. 12), or Cushing disease (see Chap. 14 and Chap. 75) must be strongly suspected of harboring a hypersecretory pituitary tumor. The mechanism of any associated hyposecretion could be the compression of normal surrounding pituitary tissue by the tumor. However, in the case of a prolactin-secreting adenoma, concomitant hypogonadism may be caused by the hormonal effect of a high serum level of prolactin, suppressing gonadotropin production and decreasing the gonadal response to the gonadotropins.39 Almost always, first-line treatment of a prolactinoma is dopamine agonist therapy; hence, measurement of serum prolactin should be part of the preoperative evaluation.

VASOPRESSIN DYSREGULATION
Vasopressin (i.e., ADH) is synthesized in the magnocellular neurons of the anterior hypothalamus and transported and stored in the posterior pituitary gland. Vasopressin is secreted into the peripheral circulation and acts on the renal tubular cells to prevent free water loss. Serum osmolarityand, to a lesser extent, volume statusregulate posterior pituitary vasopressin secretion. Deficiency of vasopressin may lead to polyuria, polydipsia, and free water loss with resultant hypernatremia. This condition is known as central diabetes insipidus (central DI) and can be temporary or permanent. It is common to lose several liters of water daily. Central DI usually indicates damage to the hypothalamus, either via suprasellar extension of a sellar or brain mass, trauma, or after extensive surgical resection. Appropriate secretion of vasopressin can often be seen with posterior pituitary damage if the hypothalamus remains intact. A patient who is alert, able to sense thirst, and can get to unlimited free water usually can drink enough to keep pace with free water losses. Of course, these patients need exogenous vasopressin treatment. However, a patient with DI and an altered sensorium or who is not able to get to free water is in a life-threatening situation; free water replacement and exogenous vasopressin treatment are urgently required. DI has other nonpituitary causes (see Chap. 25 and Chap. 206). Vasopressin excess causes SIADH and may result in free water retention and resultant hyponatremia. Among other causes, hypothalamic and/or pituitary pathology may lead to SIADH. This condition, too, may be temporary or permanent. SIADH may be asymptomatic initially, and when more advanced, severe hyponatremia can result in stupor, seizure, coma, or death. Treatment is primarily with free water restriction. SIADH also has nonpituitary causes (see Chap. 25 and Chap. 206). Diagnosis of central DI or SIADH is typically made by a history of polyuria or polydipsia, or by detecting serum sodium abnormalities in patients with known hypothalamic/pituitary pathology. Serum sodium and urinespecific gravity should be measured on all patients suspected of hypopituitarism. Patients who have undergone pituitary manipulation should have serum sodium measured routinely in the immediate postoperative period and at regular intervals for 2 weeks after surgery, because onset of SIADH or DI can be delayed (see Table 17-5). More extensive testing may be necessary (see Chap. 25 and Chap. 206).

OXYTOCIN DYSREGULATION
Oxytocin deficiency is rare and may result in problems during parturition (see Chap. 25). Oxytocin is not routinely measured.

CONDITIONS THAT MIMIC HYPOPITUITARISM

The syndrome of hypopituitarism frequently includes nonspe-cific symptoms, such as weakness and fatigue. Because such symptoms are frequently psychogenic or, if organic, may be secondary to a variety of diseases, it is not cost-effective to screen specifically for hypopituitarism unless there are additional clues. Two specific syndromes may superficially mimic hypopituitarism. Anorexia nervosa (see Chap. 128) is a form of deliberate chronic starvation, most commonly seen in young women and likely to have a psychiatric cause.40 The presence of amenorrhea with low gonadotropin levels and borderline low thyroid function test results suggests pituitary disease, but these are compensatory hormonal changes seen in any form of starvation. A patient with anorexia nervosa appears obviously emaciated, an unexpected component of hypopituitarism. In the syndrome of autoimmune polyglandular hypofunction(see Chap. 197), hormonal deficiencies of two or more endocrine glands are common, and if these are target glands of the pituitary, a pituitary cause naturally is considered.41 However, measurement of the relevant pituitary hormones in this syndrome indicates that these are primary rather than secondary deficiencies. Furthermore, there may be other endocrine deficiencies that cannot be explained by hypopituitarism, such as diabetes mellitus or hypoparathyroidism. Circulating serum antibodies directed against endocrine tissues are often present. Lymphocytic hypophysitis also may be a component of this syndrome, but this represents a true cause of hypopituitarism.

ENDOCRINE REPLACEMENT THERAPY

GENERAL PRINCIPLES Hypopituitarism can be an exceptionally satisfying condition to treat; most of its endocrine signs and symptoms can be completely relieved by suitable hormone replacement therapy. There are two general principles for prescribing a replacement regimen. First, the physician should perform suitable testing and treat only patients who are demonstrably borderline or deficient. Second, when there is failure of an endocrine target gland (e.g., thyroid, adrenal, gonad), regardless of whether the deficiency is primary or secondary, the physician should replace the target-gland hormone (e.g., cortisol) rather than the corresponding pituitary hormone (e.g., ACTH). Table 17-6 lists some commonly used endocrine replacement medications and typical adult doses.

TABLE 17-6. Maintenance Medications for Hypopituitarism

When testing demonstrates borderline-low secretion of a hormone, particularly cortisol, in an asymptomatic patient, it is unclear whether the patient should be treated. Many patients with lower-than-normal glucocorticoid secretion can tolerate even a major stress adequately, but, if possible, it is prudent not to take this risk. If the baseline serum cortisol concentration is borderline deficient, some physicians believe that extended treatment is better because the patient acquires the habit of taking glucocorticoid and may be more likely to think of increasing the dose appropriately at the time of an intercurrent illness. Others prefer to have the patient keep a supply of cortisol at home and take it only for intercurrent illness. The choice of treatment plan should be individualized. More informed choices can be made if the physician becomes familiar with local costs of the various relevant diagnostic tests and of medications. In general, the costs of maintenance endocrine medications are low and are not of significant economic concern, with the exceptions of GH therapy and gonadotropin treatment to promote fertility. For long-term follow-up, the most cost-effective strategy is to minimize hospitalizations. A central part of this strategy is to schedule follow-up office visits every 6 to 12 months after the patient is clinically stable. At these visits, the physician can answer questions and continue to educate the patient about his or her disease, as well as maintain a good rapport with the patient. This will increase the likelihood that the patient will phone for advice at a time when prompt treatment of an intercurrent illness can help prevent complications and hospitalization. GLUCOCORTICOID REPLACEMENT Cortisol (i.e., hydrocortisone) or cortisone can be used for replacement therapy. Cortisone is converted to cortisol after ingestion. Cortisol is the predominant glucocorticoid secreted by the human adrenal gland, and an adult produces ~10 mg daily in the unstressed state.42 There is a diurnal rhythm, with greater secretion in the morning and less in the evening. The serum half-life is ~1 hour. When cortisol is used for adrenal replacement in the adult, it is commonly given in dosages of 30 mg daily; for example, 20 mg each morning and 10 mg each afternoon or evening. For smaller individuals, 20 mg daily is sufficient (e.g., 10 mg twice daily). This cortisol dosage exceeds normal daily adrenal secretion, but there are losses in absorption; also, undoubtedly, one to three doses daily is a less efficient mode of administration than the more continuous natural secretion. In the occasional patient who cannot reliably remember to take more than one dose daily, it may be best to give the entire dose in the morning (rather than risk undertreatment), or to use a glucocorticoid with a longer pharmacologic half-life. Some authors suggest every other day therapy. Cortisone is used interchangeably with cortisol, with 25 mg of cortisone considered equivalent to 20 mg of cortisol. Usually, there is little to favor one compound over the other, except habit or price. However, in severe liver disease, cortisol is preferable because the conversion of cortisone to cortisol occurs in that organ. Prednisone also can be substituted for cortisol, using the assumption that 5 mg of prednisone is equivalent to 20 mg of cortisol. Replacement doses of dexamethasone are not well defined, and there is considerable individual variation; the plasma disappearance half-life is longer than 4 hours, and once-daily dosing is often adequate. A typical dose would be 0.5 mg given at bedtime, with a range of 0.25 to 0.75 mg. Although cortisol can be assayed in blood and urine, such determinations are not particularly useful in determining a replacement dose of cortisol for an individual patient. The serum half-life is short, so that the serum concentration can be above or below the normal range much of the time, even when a patient is responding well to the dose. The free cortisol in a 24-hour urine sample is often elevated on conventional cortisol dosage regimens. Most patients do well on any of the aforementioned dosage regimens, manifesting no overt evidence of hypoadrenalism or Cushing syndrome. (However, one report suggests that in men with Addison disease, commonly used replacement doses of glucocorticoids may be associated with a low bone mineral density.43) Even allowing for the lesser efficiency of intermittent oral administration, it seems likely that the customarily used replacement doses are appropriate for most individuals. If, while using one of these regimens, a patient still has chronic symptoms that suggest hypoadrenalism, the physician should check for medication error or noncompliance. If neither is found, the symptoms are likely to be due to a cause other than hypoadrenalism. Some individuals experience a pleasant high with supraphysiologic doses of glucocorticoid, leading them to complain when they are returned to true replacement doses.44 A patient's desire for euphoria is not a reason to continue higher-than-replacement doses indefinitely. A patient with untreated or undertreated hypopituitarism potentially may be more vulnerable at the time of an intercurrent illness because of a lack of hormonal homeostatic mechanisms. Before therapy, or if therapy is insufficient, such patients are very sensitive to infections, surgical procedures, or drugs such as sedatives or narcotics. Glucocorticoid doses conventionally are increased for significant physical illness or for operative procedures in an attempt to mimic the normal physiologic response to such situations. Interestingly, studies of patients undergoing operative procedures suggest that conventionally prescribed stress doses of glucocorticoid are larger and are maintained for a longer period than usually would be needed to mimic normal cortisol production.45,46 A renal transplantation group reports that they no longer increase baseline immunosuppressive doses of glucocorticoid (5 to 10 mg of prednisone daily), even when patients are stressed by sepsis or surgery, and they believe such doses have been sufficient.47 It is likely that the same would be true for intercurrent illnesses. However, until further data are available, the recommendation is to give conventional stress doses during times of intercurrent illness and perioperatively when there is no known contraindication to that approach. For some situations, howeversuch as surgery of a diabetic patient whose disease becomes difficult to control on higher glucocorticoid doses, or for a patient who becomes psychotic with high glucocorticoid dosesthere are reasons to minimize the dosage. In such a situation, the author gives only the equivalent of two or three times the usual replacement doses on the day of surgery and taper to replacement doses within 3 days in the absence of postoperative complications. The patient should be monitored carefully for possible signs of hypoadrenalism and given supplemental glucocorticoid if indicated. Hypoadrenal symptoms are readily reversed if recognized and treated promptly. Table 17-5 summarizes current recommendations concerning glucocorticoid administration in the presence of intercurrent illness or perioperatively. Patients whose hypoadrenalism is secondary to ACTH deficiency usually secrete normal or near-normal quantities of aldosterone in response to the still intact renin-angiotensin system, although their responses to salt restriction are not completely normal.48 Hyponatremia in a patient with untreated hypopituitarism is most likely due to hypersecretion of ADH, rather than to mineralocorticoid deficiency; there usually is no evidence of salt or volume depletion, and the hyponatremia is rapidly corrected by glucocorticoid.49 In contrast to patients with Addison disease, patients with secondary hypoadrenalism rarely need replacement therapy with the mineralocorticoid fludrocortisone. Moreover, cortisol has some intrinsic mineralocorticoid activity; occasionally it may be necessary to provide a small dose of fludrocortisone for a patient taking a synthetic glucocorticoid, such as dexamethasone, that has less mineralocorticoid activity. Perioperative management of patients undergoing surgery that will affect the pituitary gland presents special challenges (see Chap. 23 and Table 17-5). Evaluation of HPA axis function and management of glucocorticoid therapy after surgery for Cushing disease or Cushing syndrome may be quite complex (see Chap. 14 and Chap. 75). Patients cured of Cushing disease may require a protracted course of glucocorticoid supplementation. Medical therapy of Cushing syndrome may render the patient hypoadrenal, and replacement glucocorticoids and/or mineralocorticoids may be necessary.

THYROID HORMONE REPLACEMENT

L -thyroxine

(L-T4) is the preferred replacement medication for patients with hypothyroidism (see Chap. 45). A typical replacement dosage of L -T4 is 75 to 150 g daily, with some patients requiring as little as 50 g and others as much as 200 g daily; the total replacement dose is usually ~1.6 g/kg per day. Elderly patients may need ~25 to 50 g less than younger individuals.50,51 and 52 One dose daily is sufficient, because the half-life of L -T4 is ~1 week. The serum level of FT4 is used to judge the adequacy of the L -T4 dose. The aim for most patients is to keep the FT4 in the mid-normal range. If this goal is achieved, it is assumed that the body will convert T4 to T3 at a physiologically appropriate rate, and monitoring of the serum T3 level is not required. The serum TSH, which is most helpful in choosing the correct replacement dosage in primary hypothyroidism, cannot serve this purpose in secondary hypothyroidism, a condition caused by a deficiency of TSH. If an individual is elderly, has cardiovascular disease, or is severely hypothyroid, it is prudent to start with 25 g per day and gradually increase the dose by 25 g per month until the FT4 or FT4I normalizes. If the patient has none of these risk factors, the initial dose of 50 g per day may be increased more rapidly in 25- to 50-g increments. The correction of hypothyroidism in a patient with unrecognized hypoadrenalism can precipitate overt hypoadrenal crisis; the physician should begin glucocorticoid therapy before or along with the L -T4 if there is concomitant hypoadrenalism. Severe hypothyroidism, resulting in myxedema coma, may initially require large doses of L -T4. GONADAL STEROID REPLACEMENT A hypogonadal state is not life threatening, and emergency therapy is not needed. Adults with a deficiency are treated with estrogens or androgens to improve sexual functioning and for a general feeling of enhanced vigor and well-being (Fig. 17-2). In the premenopausal woman, estrogen helps to protect against osteopenia and osteoporosis. However, if hyperprolactinemia is present, therapy with bromocriptine should be tried first, because the resultant suppression of prolactin may return the gonadotropins and the target gland hormones to normal. There also is concern that if a prolactinoma is present, its growth may be stimulated by estrogens. The same phenomenon may be seen with the administration of testosterone, presumably because of partial conversion of testosterone to estradiol.53 Moreover, replacement of the gonadal hormones is relatively ineffective in the presence of hyperprolactinemia.38 Testosterone therapy in hypogonadal men may avert or improve osteoporosis, restores skin and body hair to normal, and is beneficial to muscle function. Gonadal hormone replacement may be contraindicated in patients with a history of certain medical conditions (e.g., breast cancer, phlebitis, and pulmonary embolism in women, and cancer of the prostate or breast in men). Consequently, patients receiving long-term gonadal replacement therapy should be monitored for the possible development of breast or prostate cancer to allow prompt discontinuation of the hormone. TESTOSTERONE REPLACEMENT Men may be treated with parenteral, transdermal, or oral androgens (see Fig. 17-2; see Chap. 119). The parenterally and transdermally administered androgen preparations are potent and have fewer important side effects than oral androgens. A testosterone transdermal system consists of a patch that is applied directly to the skin surface at 10:00 p.m. and delivers testosterone continuously; this regimen results in a serum testosterone concentration profile that mimics the normal circadian variation observed in healthy young men. A typical starting dose is one 5-mg patch. The dose can be increased or decreased to keep the morning testosterone level in the normal range. The most common adverse effect is irritation at the site of application of the patch. This can be treated with over-the-counter topical hydrocortisone cream applied after patch removal.

FIGURE 17-2. Clinically nonfunctioning pituitary tumor in a 53-year-old man. A, Skull radiograph film shows an enlarged, ballooned sella turcica (arrow) with eroded anterior and posterior clinoid processes. B, The patient has pale, pasty facies with no beard and fine wrinkling of the skin. Testing revealed gonadotropin deficiency. C, One year after having received testosterone injections every 3 weeks, the patient has responded to therapy.

A traditional, reliable method of androgen replacement in men uses long-acting testosterone esters given intramuscularly, with effects lasting as long as 3 weeks. Typical dosages are 200 mg of testosterone cypionate or testosterone enanthate every 2 to 3 weeks or 300 mg every 3 weeks. Because the active androgen is testosterone itself, allergic or idiosyncratic reactions are not expected, unless there is a reaction to one of the other ingredients (e.g., cottonseed oil, sesame oil). Dose and schedule of administration should be adjusted to keep maximum and minimum levels in the normal range. Oral androgens are often ineffective and may be associated with side effects such as cholestatic hepatitis, hepatic tumors, and peliosis hepatis. Their use is rapidly falling out of favor. Elderly men with benign prostatic hypertrophy may develop urinary retention with the initiation of testosterone therapy; therefore, smaller doses may be appropriate. Moreover, the initiation of androgen therapy in a man who has not previously attained puberty results in physical changes and changes in libido, degree of aggression, and general outlook. Some male patients who have been hypogonadal for long periods may abandon androgens because these mental and behavioral changes are psychologically threatening. Therefore, starting doses should be smaller and should be increased very gradually (see Chap. 92 and Chap. 119). ESTROGEN REPLACEMENT Premenopausal women are usually given cyclic oral estrogens, along with a progestogen for the final portion of the cycle (see Chap. 100). A typical regimen is 1 to 2 mg of micronized estradiol or 0.9 to 1.25 mg of conjugated estrogens daily for 25 consecutive days each month, taken orally (or estradiol, 0.05 to 0.1 mg per day, transdermally), with the addition of a progestogen, such as 5 to 10 mg of medroxyprogesterone acetate or 0.35 mg of norethindrone daily for the last 12 to 14 days of estrogen administration. Monthly withdrawal bleeding is expected. Standard oral contraceptive preparations containing a cyclic estrogen and progestogen are another option (see Chap. 104). Postmenopausal estrogen replacement regimens are discussed in Chapter 100. In secondary hypogonadism, the gonads are usually intrinsically normal; they are lacking gonadotropin stimulation. Female patients with hypopituitarism are potentially capable of having fertility temporarily restored.54 The use of gonadotropins and other drugs to restore fertility in such patients is discussed in Chapter 97 and Chapter 103. GROWTH HORMONE REPLACEMENT Adults with GH deficiency may start therapy with 0.006 mg/kg per day (0.018 IU/kg per day) given as a daily subcutaneous injection. The dose may be increased based on relief of symptoms and signs, and should be limited by adverse effects. The aim is to keep the IGF-I level in the age- and sex-matched normal range. The maximum dose is 0.0125 mg/kg per day (0.0375 IU/kg per day).35 Adverse effects include peripheral edema, arthralgias and myalgias, headache, paresthesias, and carpal tunnel syndrome. A child with short stature resulting from GH deficiency should be evaluated for possible treatment if the epiphyses remain open (see Chap. 12, Chap. 18 and Chap. 198).55 VASOPRESSIN REPLACEMENT ADH deficiency may lead to partial or total DI and has several forms of treatment (see Chap. 26). DI may be masked in the presence of hypoadrenalism and may come

to clinical attention only after glucocorticoid therapy is initiated. PROLACTIN AND OXYTOCIN Prolactin and oxytocin deficiency are not treated in clinical practice. CHAPTER REFERENCES
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. Vance ML. Hypopituitarism. N Engl J Med 1994; 330:1651. LeBlanc H, Lachelin CL, Abu-Fadil S, et al. Effects of dopamine infusion on pituitary hormone secretion in humans. J Clin Endocrinol Metab 1976;43:668. DeRivera JL, Lal S, Ettigi P, et al. Effects of acute and chronic neuroleptic therapy on serum prolactin levels in men and women of different age groups. Clin Endocrinal (Oxf) 1976; 5:273. Snyder P. Clinically nonfunctioning pituitary adenomas. Endocrinol Metab Clin North Am 1993; 22:163. Stevenaert A, Beckers A, Vandalem JL, Hennen G. Early normalization of luteinizing hormone pulsatility after successful transsphenoidal surgery in women with microprolactinomas. J Clin Endocrinol Metab1986; 62:1044. Sauder SE, Frager M, Case GD, et al. Abnormal patterns of pulsatile luteinizing hormone secretion in women with hyperprolactinemia and amenorrhea: response to bromocriptine. J Clin Endocrinol Metab 1984; 59:941. Styne DM. The therapy for hypothalamic-pituitary tumors. Endocrinol Metab Clin North Am 1993; 22:631. Dempsey PK, Kondziolka D, Lunsford LD. Stereotactic diagnosis and treatment of pineal region tumours and vascular malformations. Acta Neu-rochir (Wien) 1992; 116:14. Buchfelder M, Brockmeister S, Pichl J, et al. Results of dynamic endocrine testing of hypothalamic pituitary function in patients with a primary empty sella syndrome. Horm Metabol Res 1989; 21:57. Nelson AT Jr, Tucker HSG Jr, Becker DP. Residual anterior pituitary function following transsphenoidal resection of pituitary macroadenomas. J Neurosurg 1984; 61:577. Edwards OM, Clark JDA. Post-traumatic hypopituitarism. Medicine (Baltimore) 1986; 65:281. Ooi TC, Russell NA. Hypopituitarism resulting from an intrasellar carotid aneurysm. Can J Neurol Sci 1986; 13:70. Barkan AL. Case report: pituitary atrophy in patients with Sheehan's syndrome. Am J Med Sci 1989; 298:38. Rolih CA, Ober KP. Pituitary apoplexy. Endocrinol Metab Clin North Am 1993; 22:291.

14a.Imaki T, Yamada S, Horada S, et al. Amelioration of acromegaly after pituitary infarction due to gastrointestinal hemorrhage from gastric ulcer. Endocr J 1999; 46:147.
15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. Littley MD, Shalet SM, Beardwell CG, et al. Radiation-induced hypopituitarism is dose-dependent. Clin Endocrinol (Oxf) 1989; 31:363. Constine LS, Woolf PD, Cann D, et al. Hypothalamic-pituitary dysfunction after radiation for brain tumors. N Engl J Med 1993; 328:87. Cosman F, Post KD, Holub DA, Wardlaw SL. Lymphocytic hypophysitis: report of 3 new cases and review of the literature. Medicine (Baltimore) 1989; 68:240. Van der Werff ten Bosch JJ, Bot A. Growth of males with idiopathic hypopituitarism without growth hormone treatment. Clin Endocrinol (Oxf) 1990; 32:707. Rugarli EI, Ballabio A. Kallmann syndrome: from genetics to neurobiology. JAMA 1993; 270:2713. Beck-Peccoz P, Amr S, Menezes-Ferreira MM, et al. Decreased receptor binding of biologically inactive thyrotropin in central hypothyroidism. N Engl J Med 1985; 312:1085. Topliss DJ, White EL, Stockigt JR. Significance of thyrotropin excess in untreated primary adrenal insufficiency. J Clin Endocrinol Metab 1980; 50:52. Surks MI, Chopra IJ, Mariash CN, et al. American Thyroid Association guidelines for use of laboratory tests in thyroid disorders. JAMA 1990; 263:1529. Elster AD. Modern imaging of the pituitary. Radiology 1993; 187:1. Ladenson PW, Levin AA, Ridgway EC, Daniels GH. Complications of surgery in hypothyroid patients. Am J Med 1984; 77:261. Melen O. Neuro-ophthalmologic features of pituitary tumors. Endocrinol Metab Clin North Am 1987; 16:585. Orme SM, Belchetz PE. Isolated ACTH deficiency. Clin Endocrinol (Oxf) 1991; 35:213. Grinspoon SK, Biller BMK. Laboratory assessment of adrenal insufficiency. J Clin Endocrinol Metab 1994; 79:923. Talwar V, Lodha S, Dash RJ. Assessing the hypothalamo-pituitary-adrenocortical axis using physiological doses of adrenocorticotropic hormone. QJM 1998; 91:285. Snow K, Jiang NS, Kao PC, Scheithauer BW. Biochemical evaluation of adrenal dysfunction: the laboratory perspective. Mayo Clin Proc 1992; 67:1055. Yamamoto K, Saito K, Takai T, et al. Visual field defects and pituitary enlargement in primary hypothyroidism. J Clin Endocrinol Metab 1983; 57:283. Whitcomb RW, Crowley WF Jr. Male hypogonadotropic hypogonadism. Endocrinol Metab Clin North Am 1993; 22:125. Rosenfield RL. Plasma testosterone binding globulin and indexes of the concentration of unbound androgens in normal and hirsute subjects. J Clin Endocrinol Metab 1971; 32:717. Attanasio AF, Lamberts SWJ, Matranga AMC, et al. Adult growth hormone (GH)-deficient patients demonstrate heterogeneity between childhood onset and adult onset before and during human GH treatment. J Clin Endocrinol Metab 1997; 82:82. Carroll PV, Christ ER, Bengtsson BA, et al. Growth hormone deficiency in adulthood and the effects of growth hormone replacement: a review. J Clin Endocrinol Metab 1998; 83:382. Physicians' Desk Reference. 53rd ed. Montvale, NJ: Medical Economics Company, 1999. Molitch ME. Pathologic hyperprolactinemia. Endocrinol Metab Clin North Am 1992; 21:877. Jackson RD, Wortsman J, Malarkey WB. Macroprolactinemia presenting like a pituitary tumor. Am J Med 1985; 78:346. Carter JN, Tyson JE, Tolis G, et al. Prolactin-secreting tumors and hypogonadism in 22 men. N Engl J Med 1978; 299:847. Murray FT, Cameron DF, Ketchum C. Return of gonadal function in men with prolactin-secreting pituitary tumors. J Clin Endocrinol Metab 1984; 59:79. Wade TD, Bulik CM, Neale M, Kendler KS, Anorexia nervosa and major depression: shared genetic and environmental risk factors. Am J Psychiatry 2000; 157:469. Neufeld M, Maclaren NK, Blizzard RM. Two types of autoimmune Addison's disease associated with different polyglandular autoimmune syndromes. Medicine (Baltimore) 1981; 60:355. Esteban NV, Loughlin T, Yergey AL, et al. Daily cortisol production rate in man determined by stable isotope dilution/mass spectrometry. J Clin Endocrinol Metab 1991; 72:39. Zelissen PMJ, Croughs RJM, van Rijk PP, et al. Effect of glucocorticoid replacement therapy on bone mineral density in patients with Addison disease. Ann Intern Med 1994; l20:207. Flavin DK, Fredrickson PA, Richardson JW, Merritt TC. Corticosteroid abuse: an unusual manifestation of drug dependence. Mayo Clin Proc 1983; 58:764. Udelsman R, Norton JA, Jelenich SE, et al. Responses of the hypothalamic-pituitary- adrenal and renin-angiotensin axes and the sympathetic system during controlled surgical and anesthetic stress. J Clin Endocrinol Metab 1987; 64:986. Chernow B, Alexander HR, Smallridge RC, et al. Hormonal responses to graded surgical stress. Arch Intern Med 1987; 147:1273. Bromberg JS, Alfrey EJ, Barker CF, et al. Adrenal suppression and steroid supplementation in renal transplant recipients. Transplantation 1991; 51:385. Williams GH, Rose LI, Dluhy RG, et al. Aldosterone response to sodium restriction and ACTH stimulation in panhypopituitarism. J Clin Endocrinol Metab 1971; 32:27. Oelkers W. Hyponatremia and inappropriate secretion of vasopressin (antidiuretic hormone) in patients with hypopituitarism. N Engl J Med 1989; 321:492. Davis FB, LaMantia RS, Spaulding SW, et al. Estimation of a physiologic replacement dose of levothyroxine in elderly patients with hypothyroidism. Arch Intern Med 1984; 144:1752. Helfand M, Crapo LM. Monitoring therapy in patients taking levothyroxine. Ann Intern Med 1990; l13:450. Toft AD. Thyroxine therapy. N Engl J Med 1994; 331:174. Prior JC, Cox TA, Fairholm D, et al. Testosterone-related exacerbation of a prolactin-producing macroadenoma: possible role for estrogen. J Clin Endocrinol Metab 1987; 64:391. Yen SSC. Female hypogonadotropic hypogonadism. Endocrinol Metab Clin North Am 1993; 22:29. Saggese G, Federico G, Barsant S. Management of puberty in growth hormone deficient children. J Pediatr Endocrinol Metab 1999; 12(Suppl 1):329.

CHAPTER 18 HYPOTHALAMIC AND PITUITARY DISORDERS IN INFANCY AND CHILDHOOD Principles and Practice of Endocrinology and Metabolism

CHAPTER 18 HYPOTHALAMIC AND PITUITARY DISORDERS IN INFANCY AND CHILDHOOD

ALAN D. ROGOL Growth Pattern Concepts Growth Velocity Body Segment Ratio Skeletal and Dental Maturation Other Growth Concepts Growth Failure Hypopituitarism Deprivational Dwarfism Cushing Disease Diagnosis and Treatment of Growth Failure Tall Stature Endocrine Dysfunction Chapter References

Disorders of the hypothalamus and pituitary gland in infants and children often manifest as abnormalities of growth. Pediatric manifestations may involve secondarily the thyroid gland, sexual development, or glucose homeostasis (see Chap. 47 and Chap. 161).

GROWTH PATTERN CONCEPTS

A child's growth pattern arises from a complex mixture of genetic potential, nutrition, psychological factors, and the secretion and interaction of many hormones (see Chap. 7, Chap. 91 and Chap. 198). GROWTH VELOCITY The patterns of growth and adolescent development can provide useful data that signal specific problems or simplify differential diagnoses. The rate of stature change, the growth velocity, may be derived from the growth chart (see Chap. 7). The growth chart also has a graph that depicts percentiles of weight for attained linear stature. This information may prove useful in evaluating problems such as malabsorption syndromes or chronic illness, in which most children are relatively underweight for height, or hypopituitarism, in which children are relatively overweight for height. BODY SEGMENT RATIO The body segment ratio (upper segment/lower segment ratio) is determined by measuring the lower segment (pubis to soles), subtracting that value from the height to calculate the upper segment (crown to pubis), and dividing the upper by the lower. These proportions change throughout development. At birth, the trunk is relatively long compared with the extremities, but by the end of puberty, the extremities are relatively longer than the trunk (Fig. 18-1). The body segment ratio at birth is ~1.7:1; it becomes ~1:1 by 8 to 10 years of age, reflecting the growth of long bones. Blacks are relatively long-limbed compared with whites and have upper/lower segment ratios of ~0.90 after puberty (Fig. 18-2). Tables of normal body segment ratios and arm span measurements have been published.1

FIGURE 18-1. Changes in body proportions with growth. The horizontal lines divide the figures into quarters. Note the following normal phenomena: progressive decrease in the relative size of the head, increase in the relative size of the lower extremities, and the progressive anatomic descent of the midpoint of the body. (From Sinclair D. Human growth after birth. Oxford: Oxford University Press, 1985:128.)

FIGURE 18-2. Normal standards for upper segment/lower segment (US:LS) body ratio during growth. (Measurements were obtained from 2100 Baltimore school children.) There is a significantly lower US:LS ratio for blacks (B) than for whites (A) at all ages because of a shorter upper segment and longer lower segment in blacks. There were no sex differences within either racial group up until approximately age 15 years. (Modified from McKusick VA. Heritable disorders of connective tissue, 3rd ed. St. Louis: Mosby, 1966:50.)

The body segment ratio is affected in certain growth disorders. For example, many patients with chondrodystrophy have relatively short limbs compared with their upper segment; hypogonadal patients have increased limb length compared with their upper segment (eunuchoid proportions); and children with hypothyroidism may have a more infantile upper/lower segment ratio. Others have used the span (fingertip-to-fingertip) to delineate altered body proportions. The height and the span normally are within 5 cm of each other. SKELETAL AND DENTAL MATURATION Skeletal and dental maturation may be used to assess a child's developmental (physiologic) status. A number of methods for obtaining the bone age have been developed, but the method of Greulich and Pyle2 has proved to be the most practical screening procedure. A single radiograph of the left hand and wrist is obtained, and the epiphyseal development is compared with that of children of normal stature using an atlas. Other systems evaluate ossification centers in the hand, knee, and foot (Fig. 18-3). Together, these criteria allow a child's growth and developmental pattern to be determined (Fig. 18-3; Table 18-1). These data should help to determine whether a child's growth is within the range of normal (see Chap. 7 and Chap. 217) or should be further evaluated.

TABLE 18-1. Age-at-Appearance Percentiles for Major Postnatal Ossification Centers

FIGURE 18-3. The 20 ossification centers of maximum predictive value in boys (A) and girls (B). The postnatal centers that have the greatest utility in skeletal assessment are in the hand, the knee, and the foot. The numbers indicate the relative predictive value of these postnatal ossification centers, number 1 being the most predictive. (From Garn SM, Rohman CG, Silverman FN, et al. Med Radiogr Photogr 1967; 43:45. Reprinted courtesy of Eastman Kodak Company.)

OTHER GROWTH CONCEPTS Height age is the age for which a person's present height would be at the 50th percentile on the growth chart. Short stature usually is given a statistical definition, that is, a specific number of standard deviations (SD) below the expected mean for children of a certain chronologic age. A common definition is 23 SD below the mean for age; 2.5 SD, equivalent to the third percentile, often is used. Tall stature may be defined as 23 SD above the mean height for age; 2.5 SD, equivalent to the 97th percentile, often is used. Growth velocity, or rate of incremental change, also has a normal range. Growth failure is often defined below the 10th to 25th percentile of the mean height velocity for age. The lowest values for both boys and girls between the ages of 2 and 12 years approximate 1.75 inches (4.5 cm) per year. The average height increment during these years is ~2.5 inches (6.3 cm). These principles and their use in the diagnosis, differential diagnosis, and therapy for growth-retarded children are described in Chapter 92 and Chapter 198. Dental maturation is discussed in Chapter 217.

GROWTH FAILURE
Growth failure may be classified as follows: hypopituitarism; emotional deprivation; chromosomal defects; systemic illness (moderate to severe); metabolic diseases; and chondrodystrophies. This classification is not all-inclusive but rather serves as a framework for discussing the approach to children with a failure to grow adequately. The major emphasis of this chapter is on hypopituitarism and emotional deprivation. HYPOPITUITARISM Hypothalamic and pituitary dysfunction (partial or complete) may be classified using a number of systems: congenital versus acquired; isolated, partial, or panhypopituitarism; transient versus permanent; idiopathic versus organic; and primary (pituitary) versus secondary (affecting releasing factors). The outline shown in Table 18-2 is a convenient list with which to organize a differential diagnosis of hypopituitarism in infants and children. It is not all-inclusive but highlights major categories of pituitary insufficiency. The criteria for the diagnosis of the growth hormone (GH) deficiency of hypopituitarism are as follows: short stature; growth failure; no significant underlying illness; no evidence for emotional deprivation; delayed bone age; normal body proportions; low circulating levels of insulin-like growth factor-I (IGF-I); and subnormal responses to at least two stimuli for GH release.

TABLE 18-2. Classification of Hypothalamic and Pituitary Insufficiency

CONGENITAL HYPOPITUITARISM Developmental Pituitary Disorders. Developmental pituitary disorders may be so severe that they are incompatible with life. In some children, there is hypoplasia or absence of the pituitary gland, but in others, developmental anomalies of the hypothalamus appear to be responsible for the lack of pituitary development.3,4 Anencephaly and holoprosencephaly (including arrhinencephaly, holotelencephaly, cyclopia, and cebocephaly) usually are incompatible with life. This group of anomalies compose a spectrum of developmental anomalies associated with failure of complete midline cleavage of the embryonic fore-brain. The children either have no pituitary or a hypoplastic gland and have maldevelopment of the target organs. Atrophy of the adrenal glands always is present, and death usually occurs from adrenal insufficiency. A partial form of holoprosencephaly exists (Kallmann syndrome) in which patients have anosmia caused by the agenesis of the olfactory lobes, and hypogonadotropic hypogonadism secondary to failure of hypothalamic gonadotropin-releasing hormone secretion5 (see Chap. 16 and Chap. 115). Absence or hypoplasia of the anterior pituitary is also an uncommon cause of hypopituitarism. The former usually is incompatible with life, but the latter has a spectrum ranging from severe to mild deficiency. The clinical presentation depends on the amount of functioning hypothalamic and pituitary tissue. The main features are severe

hypoglycemia and shock because of adrenal failure. The anterior pituitary lobe develops from an upward diverticulum of the primitive buccal cavity (Rathke pouch) in the nasopharynx and must migrate to its usual location within the sella turcica. There are a number of ectopic sites in which the gland may lodge, from the submucosa of the nasopharynx (pharyngeal pituitary) to the base of the brain. The pituitary stalk is the main neural connection between the hypothalamus (median eminence) and the posterior lobe of the pituitary. In addition to these neurons, there are the capillary loops of the hypothalamic-pituitary portal system providing the principal blood supply to the various portions of the anterior lobe. Pituitary stalk interruption syndrome (hypoplasia of the anterior lobe, ectopic position of the posterior lobe, and interruption of the stalk) is strongly correlated with multiple anterior pituitary hormone deficiencies.6 Injury may be developmental or secondary to transection of the stalk. A number of other anomalies of the craniofacial area may coexist with hypopituitarism. The syndrome of basal encephalocele and hypothalamicpituitary dysfunction should be mentioned because these patients all have a nasoepipharyngeal mass or unexplained nasal polyp.7 The diagnosis should be considered when evaluating a nasal mass, especially in conjunction with the associated findings of hypertelorism, broad nasal root, midline facial defects, optic atrophy, or optic coloboma. Inherited Pituitary Disorders GENETIC DISORDERS OF GROWTH HORMONE DEFICIENCY. Five to 30 percent of children with growth hormone deficiency have an affected first-degree relative, which is consistent with a genetic cause. These genetic anomalies may result in isolated or combined pituitary hormone deficiencies. The more common are POU1F1, homolog of the mouse Pit-1, and PROP-1 deficiencies, which lack not only GH, prolactin (Prl), and thyroid-stimulating hormone (TSH), but also the gonadotropins. The specific genes encode members of the homeodomain family of transcription factors, which play an important role in the development of the human pituitary gland.8 Other deficiencies are caused by growth hormone-releasing hormone receptor mutations. ISOLATED GROWTH HORMONE DEFICIENCY AND PITUITARY DWARFISM. The classification of familial GH deficiency primarily is descriptive and is based on the inheritance of the appropriate phenotype and lack of response to pharmacologic stimuli for GH secretion.9 Six distinct groups based on the mode of inheritance and other hormone deficiencies have been defined (see Table 18-2). In only one, isolated GH deficiency type IA, have the pathophysiologic characteristics been defined: absence of the structural gene for GH (hGH-N).10 The others apparently are a result of the lack of synthesis or secretion of the hypothalamic-releasing hormone, growth hormonereleasing hormone (GHRH), or to an excessive secretion of the inhibitory hormone, somatostatin. However, heterogeneity of structure and function exists within and among families with isolated GH deficiency and within and among families with pituitary deficiency, making the determination of the precise location of the defect difficulthypothalamus or pituitary.11 Boys with severe GH deficiency and other anterior pituitary deficits may present with microphallus and hypoglycemia. The physical examination of the genitalia should alert the physician to determine the level of plasma glucose frequently and to use GH and cortisol to treat refractory hypoglycemia.12 Congenital Tumors CRANIOPHARYNGIOMA. The most common tumor in the area of the pituitary of children is the craniopharyngioma13 (see Chap. 11). It arises from the embryonic remains of the craniopharyngeal duct (Rathke pouch) and is of epithelial origin. This neoplasm commonly is cystic and often contains dark, thick, viscous fluid (machinery oil). Although congenital, it is so slow-growing that signs and symptoms often are not manifested until late in the first decade or in the second decade of life or even into adulthood. In addition to the signs of hypopituitarism, children and adolescents may have neurologic symptoms, including prolonged frontal headache, vomiting, or vision deficitsdecreased acuity, diplopia, or photophobia. On examination, papilledema, optic nerve atrophy, and impaired visual fields are found in most; they indicate raised intracranial pressure. Older children and adolescents may have obesity or amenorrhea. In the asymptomatic patient, the diagnosis may be made by noting flecks of calcium in the suprasellar region or changes within the suprasellar area as evaluated on computed tomographic (CT) or magnetic resonance imaging (MRI; Fig. 18-4) studies. In a symptomatic patient, CT or MRI scanning is extremely helpful in confirming the diagnosis and determining the extent of the neoplasm.

FIGURE 18-4. Computed tomographic (CT) scans of the brain of a 3-year-old girl with a craniopharyngioma. A, Sagittal view showing large suprasellar mass. B, Coronal view showing compression of pituitary gland and large suprasellar mass. (Courtesy of Dr. W. Cail, University of Virginia.)

Although craniopharyngioma is most commonly found in children, optic and third ventricle gliomas and arteriovenous malformations must be considered. The treatment is surgical and may be possible through the transsphenoidal approach. Aspiration of the cyst can decompress the mass and relieve symptoms but rarely is curative. Radiation therapy including intracystic application of radioactive pellets, although controversial, may be effective as adjuvant therapy to surgery. Tumor recurrence, even after total removal, is common. Hypopituitarism, especially GH deficiency, may be found preoperatively. Postoperatively, it is present in most patients, often in association with other defects in anterior pituitary function and central diabetes insipidus. Meticulous follow-up of growth velocity and pituitary target-organ function, and early replacement therapy greatly decrease the morbidity and mortality after surgery. DIENCEPHALIC SYNDROME. The diencephalic syndrome usually comprises a tumor in the diencephalon and the clinical picture of severe emaciation with relative conservation of growth rate, alert appearance, and relatively few neurologic signs. It is nearly always, but not exclusively, found in infancy.14,15 There is a striking lack of subcutaneous fat. Often, the child seems inappropriately happy (Fig. 18-5). Some clinicians believe that it is related to the age of onset of compression of the hypothalamus because it is not seen in patients with craniopharyngiomas or other tumors within the same anatomic area that present later in childhood, although such tumors displace the third ventricle in a manner similar to that of opticochiasmatic gliomas. Although endocrinologic deficits may be present, they are inconstant and do not help in the diagnostic process. Most tumors are gliomas located in the anterior hypothalamus or in the optico-chiasmatic system. The syndrome in infants with tumors placed more posteriorly is characterized by the early onset of vomiting and the absence or late onset of nystagmus, tremor, pallor, polyuria, papilledema, or optic atrophy. These patients are more likely to have malignant cells and raised protein concentrations in the cerebrospinal fluid than are children with more anteriorly placed tumors. These latter patients more often have nystagmus and optic atrophy. Vomiting appears later.16

FIGURE 18-5. A 10-month-old child with diencephalic syndrome (left next to a normal child of the same age (right). Despite striking emacia tion, the infant appeared happy and alert and was extremely hyperac tive. There was nystagmus. The feet and hands appeared large in comparison with the body. Basal growth hormone levels were high and were not suppressed adequately with glucose administration. At cra niotomy, there was an optic glioma involving the optic nerves and chi asma that extended into the anterior hypothalamus and the posterior portion of the third ventricle. (From Hger A, Thorell JI. Studies on growth hormone secretion in a patient

with the diencephalic syndrome of emaciation. Acta Paediatr Scand 1973; 62:231.)

Radiographic studies in patients with anteriorly placed tumors may show enlarged optic foramina and, less commonly, sellar alterations, evidence of increased intracranial pressure, and, rarely, calcification. Computed tomographic and MRI scans may be the best tools for defining the exact nature and extent of disease because the optic nerves, chiasm, and sellar and suprasellar regions can be visualized precisely. The prognosis is poor, and the treatment remains controversial. Surgery is rarely curative, but it is important to obtain a pathologic diagnosis. Radiation therapy can dramatically reduce the mass of the tumor. However, the natural history is uncertain and variable. Thus, treatment for these often slow-growing tumors varies from no therapy, to surgical excision, to radiation therapy; however, one must be concerned with ultimate brain growth if a child younger than 2 to 3 years old undergoes radiation therapy. Because most series have been small and their data were collected over several decades when the approach to diagnosis, therapy, intensive postoperative care, and endocrinologic replacement therapy was in great flux, it is not surprising that no single therapeutic protocol has proved to be clearly superior. Defects in the Structure, Metabolism, or Secretion of Growth Hormone or Insulin-Like Growth Factor I BIOLOGICALLY INACTIVE GROWTH HORMONE. Children with growth failure who have normal or elevated basal GH concentrations or a normal or increased response to pharmacologic stimuli, but diminished IGF-I concentrations that cannot be attributed to malnutrition, chronic illness, or other causes, may secrete a bioinactive (subactive) GH molecule. Bone and dental development are significantly delayed, and the body proportions are more appropriate for chronologic than height age. Although some investigators have proposed this hypothesis for short stature in the few children who fit into the diagnostic category, there is sparse evidence. Clearly, it is not the public health problem it was originally considered.17 Findings in a single patient indicate that aggregation of serum GH may be responsible.18 However, a mutation in the GH gene itself can produce a mutated form of GH (found to be expressed in Escherichia coli), which is clinically associated with short stature in the affected child.19 The mutated GH binds avidly to the GHG receptor in the IM-9 cell line, but does not stimulate intracellular signaling pathways. Thus, it inhibits the bioeffects of native GH. The growth hormone insensitive syndrome (GHIS) represents another defect in the mechanism of GH action3,20 (Fig. 18-6). These patients with familial dwarfism and clinical features of GH deficiency tend to be of normal birth weight, but growth velocity is retarded soon after birth. Motor development, bone maturation, and dental eruption are slow, and the anterior fontanelle may close later than average. The facial bones grow more slowly than the cranial vault. When the child has a small mid-face and mandible and a bulging forehead, macrocephaly should be considered. The children tend to be obese and have a high-pitched voice. Although many of the original patients were of Semitic origin, subsequent patients have come from many ethnic backgrounds.

FIGURE 18-6. Three Ecuadorian patients with growth hormone receptor deficiency resulting from a point mutation at codon 180 of exon 6 of the growth hormone receptor: a boy aged 26 months (height, 8.2 SD), a girl aged 4.8 years (height 7.4 SD), and a woman aged 19 years (height, 6.5 SD). The vertical dimension of the face is decreased, the nasal bridge is hypoplastic, and the forehead is of normal dimension, giving the impression of craniomegaly, especially in the children. (Courtesy of J. Guevara-Aguirre, Institute for Endocrinology, Metabolism, and Reproduction, Quito, Ecuador; and A.L. Rosenbloom, University of Florida College of Medicine.)

GH values are high basally and are elevated in response to pharmacologic stimuli, but IGF-I levels are low. The circulating GH molecules are biologically active. The pathogenesis of the defect is failure of the liver to respond to GH by generating IGF-I. Because the extracellular domain of the GH receptor is a growth hormonebinding protein (GHBP), this protein might be expected to be absent or defective in GHIS.21 The initial direct evidence for a GH receptor defect in GHIS was that hepatocytes obtained at biopsy from a patient with this disorder did not bind tracer quantities of radiolabeled GH, although samples from control subjects undergoing abdominal surgery did.22 Subsequently, the similarity of the circulating GHBP to the GH receptor (the former is the extracellular domain of the latter) led to the finding that GHBP was absent in patients with GHIS. The clinical and biochemical characteristics of GHIS have been reviewed.23 Prominent biochemical abnormalities include low circulating levels of IGF-I, IGF-II, insulin-like growth factorbinding protein-3 (IGFBP-3), and GHBP; and levels of IGF-I, IGF-II, and IGFBP-3 that are higher in adults than in children without a change in the level of GHBP.22 A number of treatment trials with recombinant human IGF-I are under way. In all the young patients, the growth rate is accelerated. Many patients had growth rates during therapy in the range of 8 to 10 cm per year. Because of the low IGFBP-3 levels at the onset of treatment and the lack of a buffer for the injected IGF-I, some patients experienced hypoglycemia. GROWTH HORMONE NEUROSECRETORY DYSFUNCTION. Children with dysregulation of GH secretion meet the following criteria: short stature; growth failure; bone age 2 or more years behind chronologic age; IGF-I levels low for age; and results for provocative tests for GH within normal limits.24 The children with GH deficiency meet these same criteria except that they have subnormal peak responses to provocative stimuli. Results of functional tests for other pituitary and target-gland hormones are normal. Malnutrition, systemic disease, and psychosocial dwarfism must be considered. The normal child usually has more than six secretory episodes of GH per day, but children within this category usually have fewer than four. The number of secretory episodes, the amount of GH secreted, and the peak GH concentrations all are intermediate between values found for GH-deficient and normal children. As greater numbers of slowly growing children are tested for GH neurosecretory dysfunction, the boundaries between normal and subnormal become more indistinct. It is not possible to determine with certitude from the 24-hour GH secretory pattern which children will respond favorably to therapy with GH. Given the expense and difficulty in determining a 24-hour secretory pattern, it may be prudent in selected children with growth failure to administer a 6-month therapeutic trial with GH. Marked acceleration in growth rate is the criterion for continued therapy. However, a number of such children respond to gonadal steroid hormone therapy with marked increases in endogenous GH secretion and growth rate.25,26,27 and 28 In fact, for boys, if the child is within the pubertal age range, treatment with low doses of testosterone (50100 mg intramuscularly each month) would be more appropriate than GH. The question of whether low-dose estradiol will augment GH secretion and ultimate growth in girls has been insufficiently explored. ACQUIRED HYPOPITUITARISM During Birth. Perinatal complications, including breech or face presentation, can lead to pituitary insufficiency because of vascular compromise to the anterior hypophysis or to the hypothalamic area. Review of the birth records of patients with non-familial hypopituitarism indicates that more than half of all hypopituitary patients have an abnormal perinatal historyan incidence much higher than in the normal population. Magnetic resonance imaging studies may show an ectopic posterior pituitary gland (bright spot) or interruption of the pituitary stalk. After Birth. Trauma continues to be an important cause for hypopituitarism beyond the neonatal period. A blow to the head or face can cause bleeding in the hypothalamic area or may shear the pituitary stalk. Varying degrees of hypopituitarism result, but when the pituitary remains relatively intact, all the anterior pituitary hormones circulate in low concentrations except for high levels of prolactin. The infiltrative diseases, whether infectious or granulomatous, may invade the hypothalamic areas that synthesize the releasing hormones or carry these factors to the median eminence. Pituitary insufficiency occurs whether the underlying disease is Langer-hans cell histiocytosis (relatively commonly associated with hypopituitarism), meningitis or encephalitis caused by a number of organisms (e.g., tuberculosis), or is a result of autoimmune phenomena from primary or metastatic deposits in the hypothalamic area (see Chap. 11 and Chap. 17). These latter causes are uncommon in adults and rare in children and adolescents. All forms of hypothalamicpituitary dysfunction may occur after intracranial surgery, irradiation therapy, or chemotherapy for neoplastic diseases. Data indicate growth

failure and an abnormal GH secretory pattern among children who have survived acute lymphoblastic leukemia but who had received irradiation to the craniospinal axis or intrathecal therapy with antineoplastic agents.29 The growth rate of such children should be followed carefully to determine which of the children might benefit from GH therapy (see Chap. 198). Aneurysms of the internal carotid area and arteriovenous malformations in the hypothalamic area are uncommon. However, they represent emergent but potentially treatable causes of hypopituitarism. Their diagnosis has been made much simpler with the confirmation provided by CT or MRI scanning and vascular contrast studies. Their treatment is mainly surgical. Lymphocytic hypophysitis is rare, especially in children. It may be part of a multiple endocrine organ autoimmune syndrome or appear as an isolated condition (see Chap. 11). There are no specific clues to the diagnosis except when it is found in association with other endocrine gland dysfunction. However, this condition may be responsible for varying degrees of hypopituitarism postpartum, especially in women without an obvious hypotensive episode. DEPRIVATIONAL DWARFISM The deprivational dwarfism syndrome (psychosocial dwarfism) is characterized by behavioral aberrations, emotional disturbances in association with growth failure, and abnormalities of pituitary function30 (Fig. 18-7). These are transient and are reversed by changes in the child's home environment. As originally described, truly bizarre behavior (e.g., eating from garbage pails or dog dishes, drinking from toilet bowls, and exhibiting severe sleep disturbances) was common. Growth failure was invariant and, early after hospitalization, test results for GH and corticotropin (ACTH) reserve were usually abnormal. The transient nature of the hypopituitary state became obvious when these children grew and gained weight at phenomenal speed during hospitalizationa rate of 1 inch per month was not unusual. The results of pituitary function tests for GH and ACTH reserve often reverted to normal within days to weeks. Return to the same home environment can cause the syndrome to recur.

FIGURE 18-7. Left, A 5-year, 4-month-old child with deprivational dwarfism related to parental neglect, abuse, and malnutrition. The child is 84.5 cm (33.5 in.) in height, corresponding to a height age of 22 months; there is moderate mental retardation. There was patchy alopecia of the scalp, moderate hirsutism of the body, and a paucity of sweating. Note the sparse abdominal fat and the prominent abdomen; hepatomegaly was present. Right, the child was hospitalized, attended to, and provided unlimited food intake. Four months later, he had gained 5.2 kg (11.5 lb) and had grown 3.8 cm (1.5 in.). He became more active and playful, the body hair diminished, and he perspired normally. (From Copps SC, Gerritsen T, Smith DW, Waisman HA. Urinary excretion of 3,4-dihydroxyphenylalanine [DOPA] in two children of short stature and malnutrition. J Pediatr 1963;62:208.)

Dysfunction of central neurotransmitter function has been postulated to explain the hypopituitarism and the sleep disturbance. The role of malnutrition is controversial but probably is unimportant in those children older than 2 years of age who can fend for themselves, although in a disturbed manner. However, for younger infants, inadequate caloric intake may play some role. The clinical presentations are inconstant but include variable age of onset, short stature, growth failure, bizarre eating and drinking (polydipsia) behavior, and sleep disturbance with nighttime roaming and foraging. The family social situation is best described as disorganized and disrupted. Often, only one member of a larger sibship is affected. The diagnosis is clinical, based on the detailed history and observations of the growth failure in the child and the family's behavior. The results of endocrine tests for GH and ACTH (e.g., arginine and insulin tolerance tests) are likely to be abnormal during the first few days of hospitalization but may quickly revert to normal. There are no data concerning the use of direct tests of pituitary reserve by GHRH or corticotropin-releasing hormone. Removal of such children from their environment is the only method of confirming the diagnosis and treating them. Truly remarkable acceleration of linear growth (catch-up growth) and weight gain are not unusual during the first few months in a more nurturing environment. The prognosis for weight gain, growth, and pubertal development are excellent, provided these children remain outside the deleterious environment or that environment is sufficiently changed to become more nurturing. Long-term psychological therapy for both the parents and the child is often necessary. CUSHING DISEASE Cushing disease is mentioned here not because the pituitary adenomas associated with this condition cause growth failure, but because the excessive circulating levels of cortisol in response to the elevated ACTH concentrations are potent inhibitors of growth. In fact, excess glucocorticoids from any source can adversely affect the growth rate by inhibiting both GH secretion and GH action. Diseases of glucocorticoid excess are considered in Chapter 75 and Chapter 83. DIAGNOSIS AND TREATMENT OF GROWTH FAILURE The differential diagnosis of growth failure is broad; even when hypopituitarism is documented by proper testing in the appropriate clinical situation, the precise diagnosis may be elusive. The key features of the major disease entities and their criteria for diagnosis have been outlined in other chapters. The treatment of hypopituitarism is dependent on the hormonal deficits and may include anterior and posterior pituitary hormones and some target-organ hormones or analogs. The replacement therapy strategy for diminished gonadotropin, thyroid-stimulating hormone, and ACTH concentrations is to use the target-gland hormones (Table 18-3). For diabetes insipidus, deamino-D-arginine vasopressin is preferred.

TABLE 18-3. Replacement Therapy for Pituitary Hormonal Deficiencies

Until recently, the only therapeutic option for GH deficiency was to administer human cadaveric pituitary-derived material, usually at a dose of 1 to 2 U intramuscularly three times a week. With the more purified material produced within the last decade, the incidence of circulating antibody formation was low, and only rarely was the titer high enough to attenuate growth. However, because of the limited supply, not all children with hypopituitarism could be treated, and detailed pharmacologic studies

to determine the optimal dose, route, and interdose interval could not be performed. With the recent concern over the transmission of Creutzfeldt-Jakob disease by materials derived from human brain, the recombinant DNA-produced GHs are the only agents presently available. The last several years have been exciting in terms of the development of agents for the treatment of GH insufficiency. First, the human GH structural gene has been cloned and expressed in bacteria that can produce the hormone indefinitely. The purified product is available and as potent as the native hormone. Second, GHRH has been extracted and purified, its structure determined, and large quantities of it and other stimulatory analogs produced. Early trials with subcutaneous administration in an intermittent, pulsatile fashion have proved efficacious.31,32 Studies to determine the optimal form, dose, route, and frequency of administration are being vigorously pursued. More recent experience shows that GH-deficient children accelerate their growth when the GH dosage of 0.3 mg/kg per week is administered subcutaneously (typically subdivided into daily doses, e.g., 0.045 mg/kg per day) or when GHRH therapy is given subcutaneously (~8 g/kg per dose twice daily) or using a portable minipump (2 g/kg per dose every 3 hours).33,34 Because most children with hypopituitarism, even those with structural lesions, have suprasellar defects, this form of therapy holds promise for large numbers of GH-deficient children.35 A hexapeptide, GH-releasing peptide (GHRP), has been developed synthetically but is not homologous to the GHRH-related peptides of 40 or 44 amino acids. Its activity in humans is not diminished by somatostatin, and its ability to cause release of GH when submaximal quantities of GHRH are present is at least additive and possibly synergistic. The role of this peptide or its analogs, or the native homolog if one exists, has not yet been defined, but preliminary data from a clinical trial with one analog are available.36,37 IGF-I is also available for therapeutic protocols in the treatment of GH-deficient patients. Its major use is in patients with GHIS and in patients with GH gene deletion treated with GH and producing antihuman GH antibodies. Preliminary data are consistent with growth rates in the range of 8 to 12 cm per year, similar to those among GH-deficient children treated with recombinant GH. The other conditions (chromosomal defects, moderate to severe systemic illness, metabolic diseases, and chondrodystrophies)38 are beyond the scope of this chapter. An extensive differential diagnosis of syndromes associated with short stature has been published.39 TALL STATURE NONENDOCRINE DYSFUNCTION Usually, children more than 2 to 3 SD above the mean height for age are considered tall. Most, of course, have a normal growth rate and are considered to be constitutionally tall without a pathologic cause. Many are the children of tall parents and are merely fulfilling their genetic potential. In addition to constitutional tall stature, other nonendocrine conditions associated with tall stature or excessive growth rate are cerebral gigantism, Marfan syndrome, homocystinuria, and Beckwith-Wiedemann syndrome. Endocrine dysfunctional disorders include gigantism and aberrant sexual maturation (sexual precocity, hypogonadism). From a cultural viewpoint, excessively tall girls are more likely than boys to seek therapy to diminish their predicted adult height. Although, in girls, estrogen therapy in large doses has been used in the past, with some diminution from predicted adult height, the actual and potential side effects are severe enough that this form of therapy cannot be recommended except in particularly unusual circumstances.40 Open discussion of the actions and potential deleterious effects of these potent sex steroids is often sufficient to dissuade their use. CEREBRAL GIGANTISM Cerebral gigantism (Sotos syndrome) is a disorder in which a rapid growth rate and tall stature are noted in infancy and early childhood.41 These children tend to be large at birth and grow excessively the first 3 or 4 years of life (Fig. 18-8). There are a number of dysmorphic features in addition to the tall stature: prominent forehead, high-arched palate, hypertelorism, and a long head are found in more than 80% of the children, who also are likely to be developmentally retarded. The excessive body and skeletal growth and the increased incidence of malignancies suggest that this condition is primarily a disorder of growth factors, but there are no data relevant to the etiology and pathogenesis.

FIGURE 18-8. A, Patient with cerebral gigantism at age 11 years. B, Growth curve (mean and 2 SD) of the patient shown in (A). (From Hook EB, Reynolds JW. Cerebral gigantism: endocrinological and clinical observations of six patients including a congenital giant, concordant monozygotic twins, and a child who achieved adult gigantic size. J Pediatr 1967; 70:900.)

Diagnosis is based on the constellation of signs noted and cerebral ventricular widening on CT or MRI scans. The endocrine function is normal. No specific therapy exists, and although they become tall adults, few are excessively tall. The developmental delay and mental retardation do not progress. MARFAN SYNDROME Marfan syndrome is a heritable disorder of connective tissue in which the patient tends to be tall and to have long slim limbs, arachnodactyly, lax joints, dislocation (subluxation) of the lens, and dilation with or without dissecting aneurysm of the ascending aorta42 (see Chap. 189). During growth and development, there is a marked tendency to scoliosis (Fig. 18-9). The prognosis depends on the vascular complications. The diagnosis is clinical, with therapy directed toward the vascular lesions. The specific abnormality has now been identified.43

FIGURE 18-9. A 14-year-old girl with Marfan syndrome. There is excessive height (183 cm; 72 in.), a thin body habitus, and normal pubertal development. The face is long and thin, and the limbs, hands, and feet are elongated and spidery. Scoliosis necessitated the wearing of a body brace. There is muscle hypotonia and diminution of subcutaneous fat. There also was subluxation of the lenses. The patient was treated unsuccessfully with estrogen in an attempt to hasten skeletal maturation. Transmission of this condition is autosomal dominant. (From Rallison ML. Growth disorders in infants, children and adolescents. New York: Churchill Livingstone, 1986:330.)

HOMOCYSTINURIA Homocystinuria, also a heritable disorder of connective tissue, is characterized by tall stature, fair complexion, mild mental retardation (inconstant), slim skeletal build, and arachnodactyly resembling Marfan syndrome but with severe osteoporosis and subluxation of the lens. The latter usually is not present at birth but has been described in patients 2 years old42 (see Chap. 191). The pathogenesis is cystathionine synthase deficiency, which is characterized by elevated plasma concentrations of methionine and homocystine and by the excretion of homocystine in the urine. The diagnosis is based on the clinical findings and the presence of homocystine in the urine. The prognosis depends on the marked tendency to form life-threatening arterial and venous thromboses. Many patients die of coronary or carotid occlusion in the second or third decade of life.42 The goals of dietary therapy are to eliminate potentially toxic substances accumulating proximal to the enzymatic deficiency and to supply those substances that are deficient distal to the defect.44 BECKWITH-WIEDEMANN SYNDROME Beckwith-Wiedemann syndrome is characterized by exomphalos (omphalocele), macroglossia, neonatal hypoglycemia, and postnatal gigantism42 (Fig. 18-10). The birth weight usually is increased, but polyhydramnios and prematurity are common. Although the early growth rate is often slow, these infants subsequently show accelerated growth with macrosomia, a large muscle mass, and thick subcutaneous tissue. Most children grow near the 90th percentile and have an advanced bone age. A characteristic face with macroglossia and a protruding tongue are prominent. Visceromegaly is invariant, and there is a susceptibility to abdominal malignancies because of the loss of heterozygosity of the chromosomal locus 11p15.5, an area that contains the genes for IGF-II and a tumor-suppressor gene. The most serious problem is neonatal hypoglycemia, usually with elevated circulating concentrations of insulin; however, as these infants mature, the episodes of hypoglycemia wane. The prognosis depends on the early control of the hypoglycemic episodes. This may require systemic glucocorticoid therapy for the first few months of life.

FIGURE 18-10. A 2.5-month old infant with Beckwith-Wiedemann syndrome. There was macroglossia at birth, and an omphalocele (congenital protrusion of omentum and intestine through an abdominal opening) necessitated surgical repair in the first 24 hours of life. It had been noted that the baby was big, had large facial features, and had mild weakness of the limbs. At the time this photograph was made, the infant was in the 97th percentile for weight and the 90th percentile for length. There were prominent inner canthal folds, marked enlargement of the tongue, and a high-arched palate. The nose was broad, and there was a long upper lip philtrum. There was a flat, red flame nevus in the center of the forehead and other nevi on the cheeks as well as on the eyelids. (The midline nevus as well as the bilateral double linear indentations of the ear lobes are typical.) Other findings in this infant included a short neck, large muscle mass, hepatosplenomegaly, and a residual ventral hernia. Occasionally girls with this syndrome may have clitoromegaly at birth, and some boys have cryptorchidism. The final adult height attained is relatively tall. Mild mental retardation is common, but not uniform. The relative tongue size may regress or may necessitate surgery. The condition usually appears sporadically, but autosomal-recessive transmission has been suspected. (From Filippi G, McKusick VA. The Beckwith-Wiedemann syndrome. Medicine [Baltimore] 1970; 49:279.)

ENDOCRINE DYSFUNCTION PITUITARY GIGANTISM Pituitary gigantism (GH excess) is an extremely uncommon cause of accelerated growth and tall stature. The usual cause is a pituitary adenoma composed of somatotropes. However, excessive GHRH from the hypothalamus or from an ectopic source (e.g., a pancreatic islet cell tumor) may lead to a syndrome indistinguishable from that caused by a pituitary adenoma.45,46 Theoretically, lack of somatostatin secretion or excessive secretory episodes of GHRH could cause an increased release of GH. Clinically, the patients (often adolescents) have an augmented growth ratecrossing previously defined growth percentiles in an upward direction (Fig. 18-11). They may have evidence of the acral enlargement characteristic of an adult with acromegaly (see Chap. 12). Headache, visual field abnormalities (classically, bitemporal hemianopsia), decreased visual acuity, and other signs or symptoms of increased intracranial pressure may be prominent.

FIGURE 18-11. Pituitary gigantism in two children. Growth hormone excess before the fusion of the epiphyses is characterized predominantly by rapid growth rate and excessive size and weight for age. After epiphyseal closure, growth hormone excess induces acromegalic features, including overgrowth of soft tissue, thickening of skin, widening of bones, and prominence of acral parts of the body (face, hands, feet).Many pituitary giants exhibit both increased size and acral changes, particularly if the growth hormone excess persists into adulthood. Gigantism diagnosed during childhood also may show varying degrees of acromegaly. A, A 31-month-old girl with pituitary gigantism who began to grow rapidly in her second year of life. The rapid growth and large size had been noted by the parents, who also remarked on the excessive sweating and strong body odor. The height was 104 cm (41 in.), which was 5.5 cm (2.5 in.) greater than the estimated 97th percentile. The child resembled a 4-year-old. The facial features were normal, and the hands and feet were only slightly prominent. B and C, A girl who is 5 years, 8 months old with pituitary gigantism. Her height of 118 cm (50.5 in.) corresponds approximately to a height age of 8 years, 6 months. However, this child had considerably enlarged hands and feet and marked enlargement of the jaw, nose, and ears. (A from Espiner EA, Carter TA, Abbott GD, Wrightson P. Pituitary gigantism in a 31-month-old girl: endocrine studies and successful response to hypophysectomy. J Endocrinol Invest 1981; 4:445; B and C from Hurxthal LM. Pituitary gigantism in a child five years of age: effect of x-radiation, estrogen therapy and self-imposed starvation diet during an eleven year period. J Clin Endocrinol Metab 1961; 21:343.)

Laboratory confirmation of this basically clinical diagnosis includes high circulating levels of GH and especially IGF-I and failure of the elevated GH concentrations to be suppressed after an oral glucose load. Standard CT imaging techniques with and without contrast medium, and especially MRI, are useful in defining the anatomic limits of the pituitary tumor (see Chap. 20). Transsphenoidal microsurgery often is curative, although some groups advocate radiation therapy, either as primary treatment or in addition to intracranial surgery.

With large tumors, craniotomy with direct visualization and tumor removal is preferred (see Chap. 23). ABERRANT SEXUAL MATURATION Aberrant sexual maturation, whether precocious in children or hypogonadal in adolescents, is associated with tall stature and an excessive growth rate47,48 (see Chap. 77 and Chap. 90, Chap. 91 and Chap. 92). The more commonly encountered conditions are listed in Table 18-4.

TABLE 18-4. Aberrant Sexual Maturation

Precocious Sexual Development. Sexual precocity occurs more commonly in girls than in boys and may be secondary to early maturation of the hypothalamicpituitarygonadal axis (precocious puberty) or a result of sex steroid production from the adrenal or gonad (sexual precocity, pseudoprecocious puberty). The signs of sexual maturation usually precede those of accelerated growth velocity, just as they do during normal pubertal development. Precocious puberty may be caused by organic lesions (Fig. 18-12) or may be idiopathic (see Chap. 92).

FIGURE 18-12. Nine-year-old boy with rapid growth and sexual precocity resulting from a hamartoma of the hypothalamus.

The introduction of a long-acting gonadotropin-releasing hormone (GnRH) agonist, whose activity is based on the pharmacologic precept of down-regulation (desensitization) of the GnRH receptor on the gonadotropes, has made the goals for therapy of this condition realizable.47,48 and 49 These goals are arrest or reversal of secondary sexual maturation; decrease of the linear growth rate to a normal prepubertal velocity; slowing of skeletal maturation to increase adult stature; preservation of fertility as an adult; and prevention or amelioration of any emotional disturbance. The therapy for organic lesions causing sexual precocity, including precocious puberty, is mainly surgery or radiation with, if needed, postoperative replacement of target-gland hormones. The most common cause of heterosexual precocity in girls is congenital virilizing adrenal hyperplasia (adrenogenital syndrome; see Chap. 77). The excessive growth is caused by the failure of adequate ACTH suppression by cortisol and the subsequent overproduction of biologically active adrenal androgens. During the first few years, the goal of administering adequate but not excessive amounts of glucocorticoids is not always easily obtainable. Although children with inadequate adrenal suppression grow excessively, the bone age matures at an even greater rate, causing the patient to be tall as a child but to enter puberty early and become short as an adult. Hypogonadal Syndromes. A number of hypogonadal syndromeseither hypogonadotropic or hypergonadotropicexist. The pathophysiologic disturbance that leads to excessive stature and eunuchoid body proportions is failure of epiphyseal maturation (bone age) beyond ~13 years of age and, thus, continued, albeit often slow, growth into the third decade of life (see Chap. 115).50 The therapy is mainly with gonadal steroids to produce secondary sexual characteristics and increased libido (see Table 18-3). Treatment of hypogonadotropic patients with GnRH or the gonadotropins themselves is appropriate when fertility is desired (see Chap. 16, Chap. 97 and Chap. 115). CHAPTER REFERENCES
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. Wilkins L. The diagnosis and treatment of endocrine disorders in childhood and adolescence. Springfield, IL: Charles C Thomas Publisher, 1965. Greulich WW, Pyle SI. Radiographic atlas of skeletal development of the hand-wrist, 2nd ed. Palo Alto, CA: Stanford University Press, 1959. Laron Z. The hypothalamus and the pituitary gland. In: Hubble D, ed. Pediatric endocrinology. Oxford: Blackwell Scientific, 1969:35. Rimoin DL, Schimke RN. Genetic disorders of the endocrine glands. St. Louis: Mosby, 1971:11. Lieblich JM, Rogol AD, White BJ, Rosen SW. Syndrome of anosmia with hypogonadotropic hypogonadism (Kallmann syndrome). Am J Med 1982; 73:506. Triulzi F, Scotti G, diNatale B, et al. Evidence of a congenital midline brain anomaly in pituitary dwarfs: a magnetic resonance imaging study in 101 patients. Pediatrics 1994; 93:409. Lieblich JM, Rosen SW, Guyda H, et al. The syndrome of basal encephalocele and hypothalamic-pituitary dysfunction. Ann Intern Med 1978; 89:910. Procter AM, Phillips JA III, Cooper DN. The molecular genetics of growth hormone deficiency. Hum Genet 1998; 103:255. Rimoin DL. Hereditary forms of growth hormone deficiency and resistance. Birth Defects 1976; 12:15. Phillips JA III. The growth hormone (hGH) gene and human disease. In: Banberry Report 14: Recombinant DNA applications to human disease. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory, 1983; 305. Rogol AD, Blizzard RM, Foley TP Jr, et al. Growth hormone releasing hormone and growth hormone: genetic studies in familial growth hormone deficiency. Pediatr Res 1985; 19:489. Lovinger RD, Kaplan SL, Grumbach MM. Congenital hypopituitarism associated with neonatal hypoglycemia and microphallus: four cases secondary to hypothalamic hormone deficiencies. J Pediatr 1975; 87:1171. Paja M, Lucas T, Garcia-Uria J, et al. Hypothalamic-pituitary dysfunction in patients with craniopharyngioma. Clin Endocrinol 1995; 42:467. Burr IM, Slonim AE, Danish RK, et al. Diencephalic syndrome revisited. J Pediatr 1976; 88:439. Tanabe M, Watanabe T, Hori T: Von Recklinghausen's disease with diencephalic syndrome in an adult: case report. J Neurosurg 1994; 80:556. Rogol AD. Pituitary and parapituitary tumors of childhood and adolescence. In: Givens JR, ed. Hormone-secreting pituitary tumors. Chicago: Year Book Medical Publishers, 1982:349. Kowarski AA, Schneider J, Ben-Galim E, et al. Growth failure with normal serum RIA-GH and low somatomedin activity: somatomedin restoration and growth acceleration after exogenous GH. J Clin Endocrinol Metab 1978; 47:461. Valenta LJ, Sigel MB, Lesniak MA, et al. Pituitary dwarfism in a patient with circulating abnormal growth hormone polymers. N Engl J Med 1985; 312:214. Takahashi Y, Kaji H, Okimura Y, et al. Short stature caused by a mutant growth hormone. N Engl J Med 1996; 334:432. Duquesnoy P, Sobrier ML, Duriez B, et al. A single amino acid substitution in the extracytoplasmic domain of the human growth hormone (GH) receptor confirms familial GH resistance (Laron syndrome) with positive GH-binding activity by abolishing receptor homodimerization. EMBO J 1994 13:1386. Leung DW, Spencer SA, Cachianes G, et al. Growth hormone receptor and serum binding protein: purification, cloning and expression. Nature 1987; 330:537. Eshet R, Laron Z, Pertzelan A, et al. Defect of human growth hormone receptors in the liver of two patients with Laron-type dwarfism. Isr J Med Sci 1984; 20:8. Rosenfeld RG, Rosenbloom AL, Guevara-Aguirre J. Growth hormone (GH) insensitivity due to primary GH receptor deficiency. Endocr Rev 1994; 15:369. Constine LS, Woolf PF, Cann D, et al. Hypothalamic-pituitary dysfunction after radiation for brain tumors. N Engl J Med 1993; 328:87. Parker MW, Johanson AJ, Rogol AD, et al. Effect of testosterone on somatomedin C concentrations in pubertal boys. J Clin Endocrinol Metab 1984; 58:87. Link K, Blizzard RM, Evans WS, et al. The effect of androgens on the pulsatile release and the twenty-four hour mean concentration of growth hormone in peripubertal males. J Clin Endocrinol Metab 1986; 62:159. Mauras N, Blizzard RM, Link K, et al. Augmentation of growth hormone secretion during puberty: evidence for a pulse amplitude-modulated phenomenon. J Clin Endocrinol Metab 1987; 64:596. Mauras N, Blizzard RM, Rogol AD. Androgen-dependent somatotroph function in a hypogonadal adolescent male: evidence for control of exogenous androgens on growth hormone release. Metabolism 1989; 38:286. Blatt J, Bercu BB, Gillian C, et al. Reduced pulsatile growth hormone secretion in children after therapy for acute lymphoblastic leukemia. J Pediatr 1984; 104:182. Powell GF, Brasel JA, Blizzard RM. Emotional deprivation and growth retardation simulating idiopathic hypopituitarism. I. Clinical evaluation of the syndrome. N Engl J Med 1967; 276:1271.

31. Thorner MO, Reschke J, Chitwood J, et al. Acceleration of growth in two children treated with human growth hormone-releasing factor. N Engl J Med 1985; 312:4. 32. Rogol AD, Blizzard RM, Vance ML, et al. Growth hormone releasing hormone: studies in vitro and in vivo. In: Hintz RL, Rosenfeld RG, eds. Growth abnormalities: contemporary issues in endocrinology and metabolism. New York: Churchill Livingstone, 1987; 4:13. 33. Frasier SD. Dose effect relationships and chronobiologic considerations in growth hormone administration. In: Laron Z, Butenandt O, Raiti S, eds. Clinical use of growth hormone: present and future aspects, Basel: Karger, 1987: 37. (Laron Z, ed. Pediatric and adolescent endocrinology, vol 16.) 34. Thorner MO, Rogol AD, Blizzard RM, et al. Acceleration of growth rate in growth hormone-deficient children treated with human growth hormone releasing hormone. Pediatr Res 1988; 24:145. 35. Thorner MO, Rochiccioli P, Colle M, et al. Once daily subcutaneous growth hormone-releasing hormone therapy accelerates growth in growth hormone-deficient children during the first year of therapy. J Clin Endocrinol Metab 1996; 81:1189. 36. Pihoker C, Badger TM, Reynolds GA, Bowers CY. Treatment effects of intranasal growth hormone releasing peptide-2 (GHRP-2) in children with growth hormone insufficiency. J Endocrinol 1997; 155:79. 37. Meriqa V, Merriam GR, Bowers CY, et al. Effects of eight months treatment with graded doses of a (GH)-releasing peptide in GH deficient children. J Clin Endocrinol Metab 1998; 83:2355. 38. Aceto T Jr, Bunger PF, Krell E, et al. Differential diagnosis of short stature and/or slow growth. Minn Med 1980; 63:469. 39. Bailey JA II. Disproportionate short stature, diagnosis and treatment. Philadelphia: WB Saunders, 1973. 40. Wettenhall HNB, Cahill C, Roche AF. Tall girls: a survey of 15 years of management and treatment. J Pediatr 1975; 86:602. 41. Sotos JF, Cutler EA, Dodge P. Cerebral gigantism. Am J Dis Child 1977; 131:625. 42. Maher ER, Reik W. Beckwith-Wiedemann syndrome: imprinting in clusters revisited. J Clin Invest 2000; 105:247. 43. Hayward C, Grook DJ. Fibrillin-1 mutations in Marfan's syndrome and other type-1 fibrillinopathies. Journal Hum Mutations 1997; 10:415. 44. Scriver CR, Rosenberg LE. Amino acid metabolism and its disorders. Philadelphia: WB Saunders, 1973. 45. Haigler ED, Hershman JM, Meador CK. Pituitary gigantism: a case report and review. Arch Intern Med 1973; 132:588. 46. Thorner MO, Perryman RL, Cronin MJ, et al. Somatotroph hyperplasia: successful treatment of acromegaly by removal of a pancreatic tumor secreting a growth hormone releasing factor. J Clin Invest 1982; 70:965. 47. Kaplan SL, Grumbach MM. Clinical review 14: pathophysiology and treatment of sexual precocity. J Clin Endocrinol Metab 1990; 71:785. 48. Kulin HE, Bourguignon JP. Central precocious puberty. In: Bardin CW, ed. Current therapy in endocrinology and metabolism. Toronto: Marcel Dekker, 1994. 49. Mansfield MJ, Beardsworth DE, Loughlen JS, et al. Long-term control of precocious puberty with a long-acting analogue of luteinizing hormone-releasing hormone. N Engl J Med 1983; 309:1286. 50. Van Dop C, Burstein S, Conte F, Grumbach MM. Isolated gonadotropin deficiency in boys: clinical characteristics and growth. J Pediatr 1987; 111:684.

CHAPTER 19 THE OPTIC CHIASM IN ENDOCRINOLOGIC DISORDERS Principles and Practice of Endocrinology and Metabolism

CHAPTER 19 THE OPTIC CHIASM IN ENDOCRINOLOGIC DISORDERS

R. MICHAEL SIATKOWSKI AND JOEL S. GLASER Anatomy and Embryology of the Optic Chiasm Clinical Manifestations of Chiasmal Syndromes Vision Changes Headache Sensory Phenomena Eye Movement Disorders Pallor of the Optic Discs Evaluation of Field Defects Developmental Chiasmal Dysplasias De Morsier Syndrome Forebrain Basal Encephaloceles Hypogonadism Syndromes Pituitary Dwarfism Parachiasmal Lesions Diencephalic Syndrome Radiologic Investigation of Gliomas Craniopharyngiomas Rathke Cleft Cysts Arachnoid Cysts Suprasellar Dysgerminoma Pituitary Adenomas Symptomatology Vision Changes Effect of Pregnancy Diagnosis Therapy Acromegaly Pituitary Apoplexy Vision Aspects of Therapy Follow-Up of Treated Pituitary Adenomas Empty Sella Syndrome Miscellaneous Lesions of the Optic Chiasm Traumatic Chiasmal Syndrome Metastatic Lesions Inflammatory Lesions Vascular Abnormalities Chapter References

Situated in the floor of the third ventricle, in close proximity to the hypothalamus, hypophysial stalk, and pituitary gland, the optic chiasm is commonly involved by intracranial disease processes, many of which have endocrinologic manifestations. Such disorders include, but are not limited to, pituitary adenomas, optic gliomas, and congenital forebrain anomalies.

ANATOMY AND EMBRYOLOGY OF THE OPTIC CHIASM

The optic chiasm may be considered the Grand Central Station of the vision sensory system, containing some 2.4 million afferent axons. Many of the disease processes that involve the intracranial optic nerves likewise involve the chiasm. Because of the relationship of the nerves and chiasm to the basal structures of the anterior and middle cranial fossae, pituitary adenomas and parasellar meningiomas frequently encroach on the anterior vision pathways. Failure of early diagnosis of chiasmal disorders may endanger the life of the patient and lessen the likelihood of reversal of visual deficits. The chiasm is situated in the anteroinferior recess of the third ventricle. The inferior aspect of the chiasm usually is 8 to 13 mm above the nasotuberculum line (i.e., the plane of the diaphragma sellae or clinoid processes). The intracranial portion of the optic nerves is inclined as much as 45 degrees from the horizontal and measures 17 2.5 mm in length (Fig. 19-1). The chiasm measures ~8 mm from anterior to posterior notch, 12 mm across, and 4 mm in height. The inferior surface of the chiasm projects more or less directly above the bony dorsum sellae (79%). When the chiasm lies more anteriorly over the diaphragm sellae, it is said to be prefixed (17%), and behind the dorsum, postfixed (4%).1, 2 and 3 The lateral aspect of the chiasm is embraced by the supraclinoid portion of the internal carotid artery. The anterior cerebral arteries of the circle of Willis (Fig. 19-2) pass over the dorsal surface of the optic nerves as they converge. The optic nerves are fixed at the intracranial entrance of the optic canals, the dorsal aspect of which is formed by an unyielding falciform fold of dura.

FIGURE 19-1. A, The relationships of the optic nerves (ON) and chiasm (X) to sellar structures and the third ventricle (3). (C, anterior clinoid; D, dorsum sellae; P, pituitary gland in sella.) B, Base of the brain, showing some of the pertinent anatomic structures. The cranial nerves are numbered. (A, From Glaser JS. Topical diagnosis: the optic chiasm. In: Glaser JS. Neuro-ophthalmology, 3rd ed. Philadelphia: Lippincott, 1999:81; B, From Akesson EJ, Loeb JA, Wilson-Pauwels L. Thompson's core textbook of anatomy, 2nd ed. Philadelphia: JB Lippincott, 1990:81.)

FIGURE 19-2. The circle of Willis and its environs. A, Relationship to intracranial contents, view from below. B, Anatomy of major intracranial arteries (caudal, top and

dorsal, bottom). (From Akesson EJ, Loeb JA, Wilson-Pauwels L. Thompson's core textbook of anatomy, 2nd ed. Philadelphia: JB Lippincott, 1990:81.)

The pituitary gland lies within the sella turcica and is covered by the diaphragma sellae, through which the hypophysial stalk passes. The chiasmatic (suprasellar) cistern is the subarachnoid space between the chiasm and the pituitary gland into which pituitary macroadenomas grow, often becoming large before affecting the optic nerves or chiasm (see Chap. 11 and Chap. 20). Basal mass lesions, even of moderate size, do not necessarily encroach on the chiasm. For example, pituitary adenomas must extend well above the confines of the sella turcica to contact the chiasm. Conversely, in the presence of chiasmal visual field defects, advanced suprasellar extension of an adenoma may be predicted. Smaller tumors may be detected clinically only when signs of unilateral optic nerve compression evolve or when secretory manifestations accrue. Although the anatomic variations in the position of the chiasm, as well as its arterial supply, have been studied,3, 4 it is difficult to draw conclusions regarding the preferential vulnerability of a particular portion of the chiasm based on blood supply. Moreover, it is unclear whether field defects are due to direct compression of vision axons or to interference with vasculature (or to both). At craniotomy, major stretching, distortion, and thinning of the nerves and chiasm are commonly encountered, shedding little light on the mechanisms of impairment of function. The intrinsic nerve fiber anatomy of the optic nerves and chiasm provides the functional substrate for the clinical evolution of field defects. The retinal topographic pattern is preserved, even at the junction of the optic nerves with the chiasm (Fig. 19-3). Superior retinal quadrants are represented in the superior portion of the nerve, inferior retina below, with nasal and temporal retinal fibers maintaining their relative positions in the optic nerve. Most axially located fibers within the optic nerves and chiasm are macular in origin. Some 2.4 million nerve fibers, ~1.2 million per nerve, enter the optic chiasm, with a crossed/uncrossed fibers ratio of 53:47.5

FIGURE 19-3. Retinotopic organization of visual fibers in the anterior visual pathways (after Hoyt). Diagram of homonymous retinal quadrants and their fiber projections, anterior aspect. (SN, superior nasal; ST, superior temporal; it, inferior temporal; in, inferior nasal.) Note: The superior fibers retain a superior course; the inferior fibers retain an inferior position; the anterior notch (1) is occupied by inferonasal (superior temporal field) fibers; inferior homonymous fibers, contralateral eye (2), and ipsilateral eye (3) converge in the chiasm, but superior homonymous fibers converge in the chiasm in the optic tract (4); the posterior notch (5) is occupied by the superior nasal (inferior temporal field) fibers as well as by macular fibers. (From Glaser JS. Topical diagnosis: the optic chiasm. In: Glaser JS. Neuro-ophthalmology, 3rd ed. Philadelphia: Lippincott, 1999:90.)

Macular axons constitute the largest portion of these fibers, which cross primarily in the central and posterior portions of the chiasm. The concept of Wilbrand knee (i.e., that contralateral, inferior nasal fibers cross in the anterior notch of the chiasm and loop forward into the terminal portion of the opposite optic nerve) is no longer universally accepted (see next paragraph). Superior nasal fibers cross more posteriorly in the chiasm. The lateral portions of the chiasm are composed of uncrossed superior and inferior temporal retinal fibers. Elegant anatomic autoradiography studies have been made of the chiasm in normal monkeys and in both monkey and human cadavers that had undergone monocular enucleation.6 In normal monkeys, the optic nerve fibers cross the chiasm without entering the contralateral optic nerve, but after short-term monocular enucleation, fibers from the normal optic nerve begin to approach the entry zone of the degenerating optic nerve. Only after long-term enucleation (in both humans and monkeys) was Wilbrand knee identified. Thus, because Wilbrand knee does not occur in normal humans, the phenomenon of a superior temporal hemianopia in the so-called junctional scotoma must be due to herniation of optic nerve fibers subserving the superotemporal visual field into the chiasm after prolonged compression and subsequent neural atrophy. The development of the embryonic optic chiasm begins between the fourth and sixth weeks of gestation (8- to 10-mm stage), when the optic nerves start to converge,7 and a true chiasm is clearly evident by the end of the second month of gestation. The structures of the diencephalon, including the posterior lobe of the hypophysis, the tuber cinereum, and mammillary bodies, are well defined by the third month of gestation, but the infundibular pouch contacts the stomodeal hypophysial pouch even during the fourth week of development. Congenital absence of the chiasm is a rare, but well-documented, phenomenon. It may occur in association with marked microphthalmia and aplasia of the optic nerves and tracts,8 or, conversely, in children with normal eyeballs and only minimal optic nerve anomalies.9 The latter cases are associated with strabismus and nystagmus and are likely to be secondary to developmental visuotopic misrouting.

CLINICAL MANIFESTATIONS OF CHIASMAL SYNDROMES

VISION CHANGES Most chiasmal syndromes are caused by extrinsic tumors: classically, pituitary adenomas or suprasellar meningiomas and craniopharyngiomas. With few exceptions, these slow-growing tumors produce insidiously progressive visual deficits in the form of variations on a bitemporal theme (Fig. 19-4). Sophisticated neuroradiologic and psychophysical testing has attempted to correlate the degree of vision loss with the degree of anatomic displacement of the chiasm10; these studies have demonstrated that vision loss occurs well before it is detectable by conventional outpatient methods of assessing visual field and visual acuity.11 Asymmetry of field loss is the rule, such that one eye may show advanced deficits, including reduced acuity, whereas only relative temporal field defects are present in the other eye. Unless acuity is diminished, patients report rather vague vision symptomsfor example, trouble seeing to the side, a history of fender accidents to their automobiles, or that, when passed by another automobile, the overtaking vehicle suddenly appears in their lane. The first clue to the presence of an hemianopic defect may be the manner in which acuity charts are read (i.e., with the right eye, only the left letters are seen, and with the left eye, only the right letters are seen). Progressive painless loss of peripheral field or central acuity (symmetric or asymmetric) may go unnoticed by many children as well as by some adults. Unilateral visual defects, especially in children, frequently are found during routine school vision tests. Rarely, vision may decrease precipitously when parachiasmal masses enlarge abruptly, as with infarction of an adenoma (pituitary apoplexy). Chiasmal visual field loss may differ depending on the anatomic area affected. Anterior lesions may cause an ipsilateral optic neuropathy with a contralateral superotemporal field deficit. Lesions in the body of the chiasm usually produce more symmetric bitemporal field loss, with normal visual acuity. Posterior lesions (with prefixed chiasm) may cause incongruous homonymous field loss secondary to involvement of the optic tract. Usually, many vision symptoms are vague and nonspecific until central acuity fails in one or both eyes. Unfortunately, it is all too common that visual field loss is already marked before initial perimetry is accomplished.

FIGURE 19-4. Chiasmal field defects. A, Junctional scotoma combines typical optic nerve defect (central scotoma) in left eye with temporal hemianopia in right (see also C). B, Classic bitemporal hemianopia. Riddoch phenomenon (movement perception) demonstrable in shaded area of the left field. C, Automated perimetry (Humphrey) of a junction-type field defect demonstrates diffuse field loss in the left eye (left) and temporal hemianopia in the right (right). D, Automated perimetry from patient with pituitary tumor compressing chiasm from below demonstrates asymmetric bitemporal superior quadrantic defects. (LE, left eye; RE, right eye.)

Vision loss of a chiasmal pattern during pregnancy should suggest the possibility of an enlargement of a preexisting pituitary adenoma or of an estrogen-dependent meningioma. There is little evidence to support the notion of a physiologic enlargement of a normal pituitary gland during pregnancy that is sufficient to encroach on the anterior visual pathways, nor is lactation optic neuritis a valid concept. Even now, vision loss is the first palpable clinical manifestation of pituitary macroadenomas, but this symptom cannot be construed as an early sign. From the Mayo Clinic experience,12>40% of adenomas present as vision symptoms, and 70% show field defects. HEADACHE Chronic headaches, mild or severe, are noted by most patients with pituitary adenomas.13 Headache symptoms are variable but may be most marked where advancing pituitary adenomas are restrained by a taut diaphragma sellae. In acromegaly, chronic headaches may indicate paranasal sinus enlargement, with or without active sinusitis. In children, headaches usually are not thoroughly evaluated until nausea, vomiting, and behavioral changes occur, at which point an intracranial lesion should be suspected. Additionally, obesity, precocious or delayed sexual development, somnolence, and diabetes insipidus should alert the clinician to hypothalamic dysfunction (see Chap. 9, Chap. 18 and Chap. 26). SENSORY PHENOMENA Peculiar sensory phenomena may be noted by patients with bitemporal field defects, resulting in a nonparetic form of strabismus or diplopia and in difficulty with visual tasks requiring depth perception (e.g., use of a screwdriver, threading a needle, and the like). Loss of portions of normally superimposed binocular fields results in the absence of corresponding points in visual space (and on the retina) and subsequently diminished fusional capacity. In essence, the patient has two free-floating nasal hemi-fields with no interhemispheral linkage to keep them aligned. Vertical and horizontal slippage produces doubling of images, gaps in otherwise continuous visual panorama, and steps in horizontal lines. A series of 260 patients with pituitary adenoma included some degree of double vision preoperatively in 98 patients, but a demonstrable ocular palsy was present in only 14.13 Additionally, without temporal fields, objects beyond the point of binocular fixation fall on nonseeing nasal hemiretina, so that a blind area exists with extinction of objects beyond the fixation point. EYE MOVEMENT DISORDERS The association of extraocular muscle palsies with chiasmal field defects implies involvement of the structures in the cavernous sinus, usually a sign of rapid expansion of a pituitary adenoma. Only rarely is tumor diagnosis delayed sufficiently for obstruction of the ventricular system to occur, with elevation of intracranial pressure and unilateral or bilateral sixth nerve palsies (Fig. 19-5 and Fig. 19-6).

FIGURE 19-5. Diagram illustrating nerves and structures in and near the cavernous sinus. (a, artery; ant., anterior; n., nerve; post., posterior; sup., superior.) (From Glaser JS. Topical diagnosis: the optic chiasm. In: Glaser JS. Neuro-ophthalmology, 3rd ed. Philadelphia: Lippincott, 1999:408.)

FIGURE 19-6. A, Acromegalic patient with involvement of left oculomotor nerve. B, Patient had reported a gradual onset of left-sided ptosis and diplopia with a mild headache.

Patients with large parasellar masses may display seesaw nystagmus, with alternate depression and extorsion of one eye and elevation and intorsion of the other. This results from expansion of tumors within the third ventricle, with secondary midbrain compression, rather than from chiasmal involvement per se.14 PALLOR OF THE OPTIC DISCS Pallor of the optic discs, although an anticipated physical sign of chiasmal interference, is not a prerequisite for diagnosis. In a series of 156 cases of pituitary tumors, optic atrophy was found in only 155 of 312 eyes (50%)15, disc pallor was present in 34% of the adenomas studied at the Mayo Clinic.12 Optic atrophy may not be present even when vision symptoms have lasted as long as 2 years.16 Furthermore, extensive field loss in chiasmal syndromes may be associated with normal or

minimally pale discs. Therefore, it is unwise to rely on the presence of optic atrophy as an indication of chiasmal interference; such findings are corroborative evidence at best, after the visual fields have been carefully evaluated. It is somewhat risky to predict on the basis of disc appearance the ultimate level of vision anticipated after chiasmal decompression. As a rule, the more atrophic the disc, the less likely is the return of function in defective areas of field. However, vision recovery to surprisingly good levels may occur despite relatively advanced disc pallor. Papilledema is a less common finding and results from large tumors compressing the third ventricle, with resultant hydrocephalus. EVALUATION OF FIELD DEFECTS The importance of establishing that the vertical meridian forms the central border of the defect (see Fig. 19-4) is paramount in distinguishing chiasmal interference from deficits that mimic temporal hemianopia. Such mimicking conditions include tilted discs (congenital inferior scleral crescents); nasal sector retinitis pigmentosa; bilateral cecocentral scotomas; papilledema with greatly enlarged blind spots; nutritional optic neuropathy; retinal inflammatory disease; and redundant overhanging upper lid tissue.17 The endocrinologist should be aware of the general pattern of evolution of chiasmal field defects and of the useful screening procedures that are appropriate in the context of the general physical examination. Visual field defects caused by chiasmal interference may be characterized by the following17: depressions initially occur in the central 20-degree field (therefore, exploration of the periphery is time-consuming and insensitive) (see Fig. 19-4); the central edge of the defect is aligned along the vertical meridian that passes through the point of visual fixation, in one or both eyes; the defect is more readily apparent with red targets than, for example, white against black (Fig. 19-7); and the loss of monocular acuity (reading vision) that is not explained by uncorrected refractive error or ocular disease (cataract, macular degeneration, etc.) is evidence of prechiasmal optic nerve compression until proved otherwise. Three additional caveats are that normal visual fields do not preclude the presence of a parachiasmal lesion (e.g., a microadenoma does not cause field defects); the screening technique described does not replace other formal perimetric techniques, such as Goldmann or automated perimetry; and the assessment of visual fields in no way obviates the need for anatomic studies, that is, computed tomography (CT) or magnetic resonance imaging (MRI). Only limited information is available on plain films; subtle changes in the bony structures of the sella may be easily overlooked by the inexperienced. Radiologic measurements of the sella, whether linear or volumetric, are not intrinsically important. In marginal cases, such measurements are unreliable; in obvious cases, they are superfluous; and in neither case is the problem of suprasellar extension solved. Thus, plain films have little, if any, use today in the diagnostic evaluation of patients with parasellar disease.

FIGURE 19-7. A, Finger-counting fields in adults. Four quadrants of each eye should be tested. Patient may name or hold up same number of fingers. Simultaneous finger counting may bring out subtle hemian-opic defect. B, Hand comparison in hemianopic depression appears darker, in shadow, or blurred. (From Glaser JS. Topical diagnosis: the optic chiasm. In: Glaser JS. Neuro-ophthalmology, 3rd ed. Philadelphia: Lippincott, 1999:3133.)

The advent of thin-section CT and of gadolinium-enhanced MRI has greatly simplified the diagnosis of intracranial mass lesions. Pneumoencephalography no longer is indicated, and cerebral angiography is reserved for those patients in whom the precise configuration of neighboring vessels may be relevant to transsphenoidal or transcranial surgical approaches to intrasellar or suprasellar tumors. The question of full or empty sella is clearly settled by coronal-section MRI (Fig. 19-8).

FIGURE 19-8. Empty sella syndrome. An MRI shows flattened remnant of pituitary (arrowheads) on sellar floor; chiasm above (arrow).

DEVELOPMENTAL CHIASMAL DYSPLASIAS The chiasm is infrequently the site of developmental anomalies; at times, it is related to malformation of other diencephalic mid-line structures, including the third ventricle. Embryonic dysgenesis may result in the abnormal development of the primitive optic vesicles with resulting unilateral or bilateral anophthalmos (absent globe) or useless microphthalmic cysts. Such gross ocular abnormalities may occur in isolation or may be associated with a spectrum of neural defects, including major malformations that preclude survival. Perhaps of greater clinical import are the more subtle ocular dysplasias that accompany those anterior forebrain malformations compatible with long life. Certain specific congenital anomalies of the optic disc may indicate the presence of otherwise occult forebrain malformations. DE MORSIER SYNDROME The clinical constellation of pituitary dysfunction and optic nerve hypoplasia (Fig. 19-9) is recognized as the de Morsier syndrome.18,19 Variable findings may include neonatal hypotonia; seizures; prolonged bilirubinemia; deficiencies of growth hormone, adrenocorticotropic hormone, or vasopressin; panhypo-pituitarism; mental retardation; or a combination of these.20 The association of this entity with midline cerebral anomalies (i.e., absence of septum pellucidum, agenesis of the corpus callosum) has been well documented.21 Additional neuropathologic findings have included extensive atrophy of the optic nerves and chiasm, aplasia of olfactory bulbs and tracts, and hypoplasia of the hypophysial stalk and infundibulum.22 Although the exact etiology of this condition has not been determined, it is well known to occur in association with maternal gestational diabetes and maternal use of phenytoin, quinine, alcohol, and lysergic acid.19 Genetic defectsincluding partial deletion of chromosome 1423 and mutations in the PAX324 and HESX125 geneshave been reported. Others have hypothesized that this group of defects represents a vascular disruption sequence.26

FIGURE 19-9. De Morsier syndrome. A, Hypoplastic optic discs. Note depigmented ring around discs. B, MRI, coronal view, demonstrates single ventricle with absence of septum pellucidum.

The wide clinical spectrum of patients with bilateral optic nerve hypoplasia is now well recognized. In one study, only 30 of 40 such children had coexistent central nervous system anomalies. Concomitant endocrinologic defects were associated with posterior pituitary ectopia rather than agenesis of the corpus callosum or other central nervous system abnormalities.27 Some investigators feel that endocrine function may be predicted on the basis of pituitary appearance on MRI,28 but others have documented normal-appearing glands in patients with central diabetes insipidus.29 In 35 patients with bilateral optic nerve hypoplasia, endocrine dysfunction was seen in only 27% (growth hormone deficiency and hypothyroidism being most common), and abnormal neuroimaging in 46%.30 Eighty percent of these patients were legally blind (34% had no light perception in either eye), but 10% of eyes had vision 20/60 or better. The full-blown syndrome (i.e., septo-optic dysplasia with panhypopituitarism) occurred in only 11.5% of these patients. An important clinical point is that children with such disorders are at risk for sudden death during febrile illnesses secondary to hypocortisolism, thermoregulatory disturbances, and dehydration.31 The occurrence of segmental superior optic nerve hypoplasia has been observed in infants born to mothers with gestational diabetes32; this, perhaps, represents a forme fruste of the de Morsier syndrome. FOREBRAIN BASAL ENCEPHALOCELES A variety of optic nerve abnormalities (hypoplasia, colobomata, peripapillary staphylomata, and the morning glory disc) may be associated with other developmental defects; principally, they take the form of midline craniofacial anomalies,33,34 including hypertelorism; defects of the lip, palate, and maxilla; anencephaly35 and prosencephaly36; skeletal malformations37; or combinations B thereof. Of special neuro-ophthalmologic interest is the association of optic disc malformation with forebrain basal encephaloceles. Herniated brain tissue may present as pulsating exophthalmos (spheno-orbital encephalocele usually associated with neurofibro-matosis), hypertelorism with a pulsatile nasopharyngeal mass (transsphenoidal encephalocele), or a frontonasal mass, with or without hypertelorism (frontoethmoidal encephalocele). Congenital disc anomalies, such as hypoplasia or the coloboma dysplasia variety38 (Fig. 19-10), associated with hypertelorism or other mid-facial malformation, are evidence of basal encephalocele until proved otherwise. There are also reports of isolated retinal colobomas, sparing the optic nerve, in association with sphenoethmoidal encephaloceles. The physical findings of transsphenoidal or transethmoidal basal encephalocele39 are listed as follows:

FIGURE 19-10. Transsphenoidal encephalocele. A, Large dysplastic optic disc (coloboma of nerve). B, Polytomogram demonstrates transsellar herniation of forebrain (arrows).

Midline facial anomalies: Broad nasal root, hypertelorism, mid-line lip defect, wide bitemporal skull diameter, cleft palate Nasopharyngeal mass: Midline epipharyngeal space in location pulsatile; often symptoms of nasal obstruction may be confused with a nasal polyp; rare in infancy Hypopituitarism/dwarfism, ocular: Congenital disc anomalies such as colobomatous dysplasias, chiasmal field defects, poor vision, exotropia In the evaluation of basal encephalocele, neuroradiologic imaging of the cranial base is indicated. Biopsy or attempted resection of posterior nasopharyngeal masses should be avoided because these masses invariably are encephalomeningoceles, and surgical manipulation may result in meningitis with tragic outcome. Because both optic disc hypoplasia and colobomatous dysplasia may reflect developmental defects of the chiasm, it is essential to recognize that such ophthalmologic anomalies may be a clue to underlying brain and endocrine defects. HYPOGONADISM SYNDROMES Hypogonadotropic hypogonadism (Kallmann syndrome) is an autosomal dominant disorder characterized by low levels of serum follicle-stimulating hormone and luteinizing hormone and by eunuchoid features40 (see Chap. 115). Other pituitary hormone levels are usually normal, but midline facial anomalies and unilateral renal agenesis may be present. This disorder can be associated with anosmia, hypoplasia of the nose and eyes, hypogeusia (taste deficiency),41 retinitis pigmentosa,42 and Mbius syndrome (congenital facial diplegia and horizontal gaze palsies with esotropia) with peripheral neuropathy.43,44 Hypoplastic development of the olfactory bulbs and the hypothalamus has been documented. PITUITARY DWARFISM A large series45 of 101 pituitary dwarfs revealed MRI evidence of posterior pituitary ectopia in 59. Pituitary volume did not change with hormonal therapy. There was a 3:1 male predominance, and a higher incidence of breech delivery (32% versus 7%) and congenital brain anomalies (12% versus 7%) in children with posterior pituitary ectopia. Defective induction of mediobasal brain structures in early embryonic life is the presumed cause, and mutation in the PIT-1 gene may be implicated.

PARACHIASMAL LESIONS
Most chiasmal syndromes are caused by extrinsic masses: classically, pituitary adenomas, suprasellar meningiomas, craniopharyngiomas, and internal carotid artery aneurysms. Although certain patterns of vision failure may suggest the location and type of lesion, such clinical impressions often prove fallible in the face of neuroradiologic procedures (and at craniotomy). At any rate, the diagnostic evaluation of all nontraumatic chiasmal syndromes is stereotyped: to rule in or out the presence of a potentially treatable mass lesion. Inflammatory or infectious causes are extremely rare, and chiasmal involvement by head trauma or radionecrosis is uncommon. The distinction is made by age-related incidence; accompanying signs and symptoms; and typical, if not diagnostic, radiologic appearance.

OPTIC AND HYPOTHALAMIC GLIOMAS Primary astrocytic tumors of the anterior visual pathways assume two major clinical forms: the relatively benign glioma (juvenile pilocytic astrocytoma) of childhood and the rare malignant glioblastoma of adulthood. With the exception of vision loss and anatomic location, these two groups have little in common, and the assumption that the progressive malignant form stems from the static childhood form is untenable. For the indolent glioma of childhood, clinical presentation is predicated on the location and extent of the tumor. Gliomas may be separated into roughly three topographic groups: unilateral optic nerve (orbital, intracranial, or both, but not involving the chiasm); principally chiasmal mass; or simultaneous infiltration of the hypothalamus. Strictly intraorbital gliomas present as insidious proptosis of variable degree, and although vision is usually diminished, remarkably good vision function is not uncommon. Strabismus, disc pallor, or disc swelling may be observed. Progressive proptosis, even if abrupt, or increased visual deficit does not imply an aggressive activity of the tumor, a hemorrhage, or necrosis. These tumors may rarely extend intracranially and have a low (~10%) mortality rate.46 Chiasmal gliomas are more common than the isolated orbital type. These tumors present with unilateral or bilateral vision loss, strabismus, optic atrophy, and/or infantile nystagmus. The nystagmus may mimic spasmus nutans (usually unilateral or asymmetric nystagmus), complete with head nodding and torticollis, or may show a coarse, conjugate mixed horizontalrotary pattern, especially when vision is severely defective. Seesaw nystagmus may also rarely be seen in these patients. Children with extensive basal tumors also show hydrocephalus and signs and symptoms of increased intracranial pressure. Hypothalamic signs include precocious puberty, obesity, dwarfism, hypersomnolence, and diabetes insipidus. Usually, the non-vision complications of extensive gliomas occur in infancy or early childhood, and onset of obstructive signs or hypothalamic involvement much beyond the age of 5 years is uncommon. Gliomas of the optic nerves and chiasm may be associated with neurofibromatosis in 20% to 40% of cases; the patients either show other characteristic stigmata or have affected relatives.47, 48 and 49 Indeed, there is good evidence that children with neurofibromatosis type 1 and chiasmal gliomas have a better long-term prognosis than those with chiasmal gliomas alone.46 Absence of neurofibromatosis, electrolyte abnormalities, and intracranial hypertension are all indicators of a poor prognosis.50,51 Treatment of benign optic gliomas in childhood is somewhat controversial, with most pediatric and neuro-ophthalmologists favoring a conservative approach. When there is documented progressive vision loss, significant tumor growth, or obstructive signs, intervention is obviously indicated. Debulking surgery, radiation therapy, and chemotherapy have all been advocated. Newer radiation techniques may allow a higher percentage of treated patients to grow normally,52 and, when chemotherapy is indicated, reports suggest success with a combined carboplatin/vincristine regimen,53 as well as oral etoposide (VP-16).54 DIENCEPHALIC SYNDROME When the hypothalamus is the site of childhood glioma, a dien-cephalic syndrome evolves.55 Findings consist of emaciation, despite adequate food intake, that develops after a period of normal growth; hyperactivity and euphoria; skin pallor (without anemia); hypotension; and hypoglycemia. Other notable signs include nystagmus and disc pallor, to which may be added sexual precocity and laughing seizures.56 Twelve cases of histologically proven opticochiasmatic glioma with diencephalic syndrome were culled from a 22-year review.57 There were 6 men and 6 women, all with failure to thrive but with normal linear growth; none had stigmata of neurofibromatosis. Two patients died in the immediate postoperative period, and 10 patients received radiotherapy with reversal of their diencephalic syndrome (weight gain, deposition of subcutaneous fat, normal development). Six of 10 are alive, 3 being considered normal, and 3 are blind, retarded, or both. Clinical evidence of bilateral optic nerve involvement was seen in 10 of these 12 cases, but it was not possible during surgery to determine the origin of these tumors. The diencephalic syndrome appears to be related to the age at which the hypothalamus becomes compressed; none of the 12 patients (and only 4% of all published cases) had onset of symptoms after 2 years of age. Craniotomy with biopsy and radiotherapy is often indicated, as tumors involving the hypothalamus appear to be larger and more aggressive than other astrocytomas arising in this region.58 RADIOLOGIC INVESTIGATION OF GLIOMAS The radiologic investigation of suspected gliomas is now sophisticated to the extent that neuroradiologic biopsy may obviate tissue diagnosis. The typical, but variable, findings include CT and MRI evidence of enlarged orbital or intracranial optic nerves; enlarged chiasm; a homogeneous hypothalamic mass; enlarged optic canals; and J-shaped or gourd-shaped sellae. The demonstration of such typical dysplastic changes of the sella turcica and optic canals, coupled with CT or MRI evidence of intrinsic chiasmal mass, so strongly suggests the diagnosis of glioma that histopathologic affirmation probably is superfluous and hazardous to vision. Analysis of visual fields in patients with chiasmatic gliomas has shown no consistent relationship between the pattern of field defects and the location, size, or extent of tumor; in 12 of 20 patients, the putative bitemporal pattern of chiasmal involvement was absent.59 Central scotomas or measurable depression of the central field occurred in 70% of the eyes; therefore, the absence of bitemporal hemianopia, or one of its variants, cannot be interpreted as a sign that the glioma does not involve the chiasm. CRANIOPHARYNGIOMAS Craniopharyngiomas (see Chap. 11) are developmental tumors that arise from vestigial epidermoid remnants of Rathke pouch, scattered as cell nests in the infundibulohypophysial region. These tumors are usually admixtures of solid cellular components and variable-sized cysts containing oily mixtures of degenerated blood and desquamated epithelium or of necrotic tissue with cholesterol crystals. Calcification of such debris may be radiologically detectable, a helpful diagnostic sign. In rare instances, these tumors may present in the neonate, attesting to their congenital origin.60 Craniopharyngiomas constitute 3% of all intracranial tumors (~15% in children), with two distinctive modes of presentationin childhood and in adulthoodwith a peak in the 40- to 70-year-old age range (Fig. 19-11).

FIGURE 19-11. Age distribution of craniopharyngioma. (Data from Svolos D. Craniopharyngiomas: a study based on 108 verified cases. Acta Chir Scand Suppl 1969; 403:1; Matson DD, Crigler JF. Management of craniopharyngioma in childhood. J Neurosurg 1969; 30:377; Bartlett JR. Craniopharyngiomas: an analysis of some aspects of symptomatology, radiology and histology. Brain 1971; 94:725. Note: Matson and Crigler series limited to children younger than 16 years of age.)

The symptomatology of childhood craniopharyngioma is variable, depending on the position and mass of the tumor.61,62 Frequently, progressive vision loss goes unnoticed until a level of severe impairment is reached, or unless headache, vomiting, or behavioral changes occur because of hydrocephalus. Obesity, delayed sexual development, somnolence, and diabetes insipidus attest to hypothalamic dysfunction, and other endocrinopathies may be present. Increased intracranial pressure is not uncommon, and papilledema may be observed. Some optic atrophy is usually present, but its absence does not conflict with the presence of chronic compression of the anterior visual pathways, even with severe vision loss. Suprasellar or intrasellar calcification is a rather constant radiologic finding in childhood craniopharyngiomas, occurring in 80% to >90% of affected children. Cystic areas frequently occur in craniopharyngioma but rarely in opticochiasmatic gliomas. CT scanning retains special sensitivity in diagnosis, being superior to MRI in detecting calcifications and cyst formations. However, involvement of adjacent structures is more clearly defined by MRI. In adults, defects in visual field or acuity are the initial symptoms, although increased intracranial pressure or endocrine dysfunction less frequently occur. Visual field defects often take the form of asymmetric bitemporal hemianopia or homonymous patterns, indicating optic tract involvement. Intracranial calcification is seen much

less regularly in adults than in children. The surgical therapy for craniopharyngiomas ranges from total (or at least radical) excision63 or postoperative radiotherapy after partial removal of tumor, to radiation therapy administered after simple biopsy or cyst decompression.64 Transsphenoidal decompression may be indicated for large tumors filling the sella. When possible, total removal of the tumor is ideal, but radical manipulations should not be attempted when adhesions to the optic nerves, chiasm, carotid arteries, or hypothalamus are present. The more conservative approach of simple decompression of the anterior visual pathways and relief of third-ventricle obstruction appears judicious, and postoperative radiation therapy has established efficacy. Endocrine replacement therapy is anticipated in the vast majority of cases, often for life. As with pituitary adenoma and meningioma, craniopharyngiomas may enlarge abruptly during pregnancy.65 RATHKE CLEFT CYSTS Although previously regarded as a rare lesion in the sellar area, these cysts derive from Rathke cleft, an embryonic vestige of Rathke pouch. In a series of 18 patients with this lesion,66 7 presented with visual disturbance or bitemporal hemianopia, and 7 presented with a variety of endocrine dysfunctions. Unlike craniopharyngiomas, partial removal or decompression of these cysts with one procedure is usually sufficient, and regrowth is less common. ARACHNOID CYSTS Enlarging loculations of cerebrospinal fluid (CSF) contained in arachnoidal cysts infrequently present as a chiasmal syndrome. These may arise, for example, in the floor of the third ventricle, causing chiasmal compression, a J-shaped sella, and occasional precocious puberty.67 Women with benign intrasellar cysts have been reported,68 showing bitemporal hemianopia, headache, optic atrophy, and panhypopituitarism. Another patient presented with obesity and amenorrhea but without visual defects.69 SUPRASELLAR DYSGERMINOMA Primary suprasellar dysgerminomas (atypical teratoma, ectopic pinealoma) are rare causes of chiasmal interference, but they constitute a more or less distinguishable clinical syndrome. These tumors likely arise from cell rests in the anterior portion of the third ventricle and are not directly related to the pineal itself, although histologically, they resemble atypical pineal teratomas. A review of 64 cases67 revealed that the classic triad consists of early diabetes insipidus; visual field loss, not necessarily of a clearly chiasmal pattern (owing to infiltration of the anterior visual pathways); and hypopituitarism. Symptoms commence at the end of the first or during the second decade of life. Girls are affected more frequently, with a peak incidence at 10 to 20 years of age. Usually, plain film radiology of the sella is normal, but MRI readily reveals the lesion. Frequently, there is growth retardation. The diagnosis is confirmed by CSF cytology, measurement of human chorionic gonadotropin, or both, but often biopsy is necessary.70 The radical excision of tumor invading the optic nerves and chiasm, infundibulum, and floor of third ventricle is not possible, but radiotherapy offers excellent palliation, if not a cure. Because subarachnoid seeding of the neuraxis is a distinct possibility, more extensive radiation may be indicated. Long-range endocrine replacement is critical. PITUITARY ADENOMAS Asymptomatic pituitary adenomas occur in >20% of pituitary glands, and some degree of adenomatous hyperplasia can be found in almost every pituitary gland.71 A postmortem study72 of pituitaries removed from 120 patients without clinical evidence of pituitary tumors revealed a 27% incidence of microadenomas, of which 41% stained for prolactin, without gender difference. To generalize, >1 in 10 people in the general population dies harboring a prolactinoma. The incessant parade of this clinical syndrome is, therefore, not surprising. Tumor of the pituitary gland is the single most common intracranial neoplasm that produces neuro-ophthalmologic symptomatology, and chiasmal interference is overwhelmingly the most frequent presentation (see Chap. 11). Strictly speaking, a microadenoma refers to a tumor that is 10 mm or less in diameter and confined to the sella. Symptomatic adenomas occur infrequently before 20 years of age but are common from the fourth through seventh decades of life. When these tumors do occur in childhood, most are asymptomatic. When symptoms are present, headache, visual field loss, and endocrinopathies are the most common. Dissimilar to adults, in children there is a definite male predominance, and many tumors are hemorrhagic.73 Histologic staining characteristics alone do not correlate well with patterns of growth or clinical symptomatology. A functional classification of pituitary adenomas, as elaborated by electron microscopy and immunohistochemistry, has replaced the previous simplistic classification of eosinophilic, basophilic, and nonfunctioning (see Chap. 11). SYMPTOMATOLOGY Nonocular symptoms include chronic headaches (severe or mild) in more than two-thirds of patients, fatigue, impotence or amenorrhea, sexual hair change, or other signs of gonadal, thyroid, or adrenal insufficiency (see Chap. 17). Prediagnostic signs and symptoms, affecting vision or otherwise, may exist for months to years before diagnosis is established. VISION CHANGES With pituitary tumors, vision failure may take the form of a rather limited number of field patterns. As suprasellar extension evolves, a single optic nerve may be compromised, with resultant progressive monocular vision loss in the form of a central scotoma. More frequently, as the tumor splays apart the anterior chiasmal notch, superotemporal hemianopic defects occur (Wilbrand knee, as discussed previously). However, this well-touted superior bitemporal hemianopia is almost always accompanied by minor or major hemianopic scotomas approaching the fixational area along the vertical meridian (see Fig. 19-4). Asymmetry of field defects is common, the eye with the greater field deficit also being likely to show diminished central vision. Marked asymmetry is not uncommon, such that one eye may be blind and the other may show a temporal hemianopic defect, the so-called junctional scotoma; this combination is as exquisitely localizing to the chiasm as is the classic bitemporal hemianopia. Adenomas extending posteriorly produce incongruous homonymous hemianopias by optic tract involvement; central vision usually is diminished, at least in the ipsilateral eye. In late stages, the only suggestion of the chiasmal character of field defects may be minimal preservation of the nasal field of one eye. The absence of field defects in patients undergoing evaluation for amenorrhea or a sella enlargement that is incidentally discovered does not imply the absence of an adenoma. For example, many patients with acromegaly do not show field defects, and microadenomas by definition do not escape the confines of the sella. EFFECT OF PREGNANCY The effect of pregnancy on pituitary adenomas is of interest diagnostically and therapeutically. Enlargement of preexisting pituitary tumors during the third trimester of pregnancy may occur,74 with reduction in size postpartum. That an otherwise normal pituitary gland may enlarge owing to the changes of pregnancy alone, causing symptoms affecting vision, is controversial.75 Nevertheless, a 30-week pregnant woman with an enlarged pituitary and bitemporal hemianopia that regressed spontaneously postpartum was reported76; a retrospective diagnosis of lymphocytic hypophysitis was made. DIAGNOSIS Many pituitary tumors deform the sella turcica sufficiently to be detected by plain film techniques, but, normal or otherwise, such procedures must be considered preliminary or superfluous. CT with contrast or gadolinium-enhanced MRI is mandatory when chiasmal lesions are suspected (see Chap. 20). THERAPY The rational approach to treatment of pituitary adenomas has evolved radically over the past 2 decades with the advent of thin-section CT; MRI; transsphenoidal microsurgery; hormonal assays; and dopamine agonists (e.g., bromocriptine), potent inhibitors of pituitary synthesis and release of prolactin. The choice of treatment is open to discussion, with enthusiastic advocates in each camp, but the prime consideration is the ultimate well-being of the patient. Patients with high surgical risk, especially the elderly, should not be subjected to frontal craniotomy. After uncomplicated transsphenoidal surgery alone, vision recovery approaches 90%.77 Radiation therapy, used either primarily or postoperatively, has great efficacy,78 and stereotactic radiosurgery has been shown to be effective for select patient groups.79

The administration of bromocriptine may rapidly improve vision function when prolactinomas compress the chiasm. In a study80 of 10 men with field defects caused by prolactinomas (initial prolactin level range 153514,200 ng/mL) who were treated with 7.5 to 30 mg per day bromocriptine, an increase in vision usually began within days of commencing therapy, and CT evidence of a decrease in tumor volume was documented somewhat later. Pregnancy apparently is not a contraindication for bromocriptine therapy.81 An extraordinary, rare complication of chiasmal herniation from shrinkage of a pituitary tumor treated with bromocriptine has been reported; recovery of vision ensued after a decrease in the dosage.82 With the advent of pergolide, another ergot-derived dopamine agonist, comes another viable treatment alternative, with apparently fewer frequent side effects of hypotension, nausea, and headache. Also, cabergoline has a very long duration of action, as well as fewer adverse effects. Quinagolide, a non-ergot long-acting prolactin inhibitor, a pure D2 agonist, is also useful.83 Finally, the long-acting somatostatin analog octreotide may be effective in the treatment of somatotropic, thyrotropic, gonadotropic, and nonfunctioning adenomas.84 In many cases, hormonal therapy of prolactinomas results in rapid improvement in vision function, often independently of decrease in tumor size. ACROMEGALY Acromegaly is the relatively rare clinical condition related to adenomatous secretion of growth hormone, with resultant hypertrophy of bones, soft tissues, and viscera (see Chap. 12). Sellar changes, when present, are indistinguishable from those caused by other adenomas. Of 1000 pituitary adenomas, 144 of 228 acromegalic patients had visual field defects.96 Possibly, this relatively high incidence of visual defects reflects delay in diagnosis in a series commenced 5 decades ago. Diabetes mellitus in acromegaly may be associated with typical retinopathy.85 Increase in corneal thickness and elevated ocular tension (glaucoma) has been reported,86 and CT scan has revealed thickened extraocular muscles.87 An unusual developmental condition with a dominant inheritance pattern, the so-called ACL (acromegaly, cutis verticis, leukoma) syndrome, has been described.88 This syndrome consists of acromegaloid features combined with severe ridging of the skin of the scalp (cutis verticis gyrata) and corneal whitening (leukoma). Pathologic examination of corneal leukoma has demonstrated a propensity for the nasal limbus, with whorl-like accumulations of disorganized collagen material and mucinous deposits.89 Signs tend to increase with age, with variable family penetrance. PITUITARY APOPLEXY Pituitary apoplexyan acute change in adenoma volume resulting from hemorrhage, edematous swelling, or necrosisis not rare, although the appropriate diagnosis may be elusive (see Chap. 17). Perhaps some 10% of pituitary adenomas undergo such acute or subacute changes,90 with clinical signs and symptoms including change in headache pattern (often severe frontal cephalgia), rapid drop in visual function, unilateral or bilateral ophthalmoplegia, epistaxis or CSF rhinorrhea, and other complications of blood or necrotic debris in the CSF. In a review of 320 verified pituitary adenomas,91 evidence of hemorrhage was found in 98 cases (18.1%). There was a high incidence of giant or large recurrent adenomas (41%). The mean age was 50 years (range, 1771 years). The clinical course included acute apoplexy (7 cases); subacute apoplexy (11 cases); recent silent hemorrhages (13 cases); and old silent hemorrhages (27 cases). Sella enlargement was present in all patients. These patients need not be stuporous, but rapid deterioration and obtundation are highly suggestive. There appears to be a tendency for such events to take place in intrasellar secretory adenomas confined by a competent diaphragma sellae. Ischemic necrosis causes sudden expansion of the tumor with acute compression of neighboring structures, including the optic nerves and chiasm and the ocular motor nerves in the cavernous sinus. Although this syndrome should now be well known, delay in diagnosis is frequent. Common misdiagnoses usually include meningitis, ruptured intracerebral aneurysm, or sphenoidal mucocele. Almost all cases show abnormal sellae on plain skull series. The CT and MRI scans are typical, if not diagnostic.92 MRI and CT scans distinguish between many tissue densities, and MRI can detect the presence of blood; the finding of acute or subacute bleeding within a tumor based in an enlarged sella is highly suggestive of pituitary apoplexy. Although in a few cases (limited suprasellar extension and intact or improving vision) corticosteroid replacement and other expectant medical management may suffice, as a rule, rapid transsphenoidal decompression of the often hemorrhagic tumor should be accomplished without delay to minimize devastating visual consequences; final endocrine status is less likely to be affected. VISION ASPECTS OF THERAPY The medical, surgical, and radiation therapies of pituitary adenomas are covered elsewhere (see Chap. 21, Chap. 22, Chap 23 and Chap. 24). The present role of irradiation of pituitary adenomas is problematic, considering the palpable failure rate and question of untoward side effects. Radiation therapy does indeed appear to reduce the rate of recurrence of pituitary adenomas.93 However, optic nerve and chiasm damage have occurred secondary to radiation necrosis anywhere from 2 months to 6 years after treatment.94 Radiation retinopathy, empty sella syndrome, cranial neuropathies, and further pituitaryhypothalamic disturbances may result from radiation therapy of pituitary lesions. There are also anecdotal reports of sarcomas, gliomas, and meningiomas occurring after radiation treatment of pituitary adenomas.95 The addition of bromocriptine before, during, or after radiotherapy may be helpful in controlling tumor secretion and size until the radiation treatment reaches its maximal effect. After uncomplicated surgical decompression, visual acuity and fields may return rapidly within 24 to 48 hours or improve weekly (Fig. 19-12). Such restoration is dependent on the duration of visual morbidity and the degree of pallor of the optic discs. After surgery, if careful ophthalmoscopy reveals attrition of the retinal nerve fiber layer, corresponding field defects are permanent. For the most part, what vision returns does so by 3 to 4 months, although continued improvement to 1 year postoperatively is possible. Although fortunately the exception rather than the rule, vision loss is a well-known complication of both transsphenoidal surgery and craniotomy. Failure of vision recovery within the 24-48 hour postoperative interval is highly suggestive of occult hemorrhage in the tumor bed or from related vessels. MRI is essential and decompression may be necessary.

FIGURE 19-12. A, Preoperative automated visual fields from a 68-year-old man with a nonfunctioning pituitary adenoma. Note dense bitemporal defect. B, Same patient, 10 weeks after transsphenoidal decompression, enjoys dramatic recovery and near-normalization of visual fields. (Left, left eye; right, right eye.)

FOLLOW-UP OF TREATED PITUITARY ADENOMAS From the standpoint of detecting recurrence, the follow-up of treated adenomas has been problematic. Even as adenomas must be large initially to cause visual defects, so must recurrences be substantial before defects again evolve. Although progressive vision failure may be the incontestable impetus for reoperation, consecutive perimetry may not be counted on to reveal early tumor recurrence. One should obtain an anatomic assessment, as provided by CT scanning or MRI. Recurrence of vision failure may be caused by regrowth of tumor, arachnoidal adhesions associated with progressive empty sella syndrome, or delayed radionecrosis. Tumor recurrence is, by far, the most common mechanism of vision deterioration, but field examination alone may not make this distinction.

EMPTY SELLA SYNDROME

Extension of the subarachnoid space into the sella turcica through a deficient sellar diaphragm may manifest itself clinically and radiologically as a syndrome mimicking pituitary adenoma. The empty sella may be defined as nontumorous remodeling that results from a combination of incomplete diaphragma sellae and CSF fluid pressure.96 Diaphragmal openings are common; in one study, defects >5 mm were found in 39% of normal autopsy cases.97 The sella is characteristically enlarged, but an empty sella may be of normal size. Primary empty sella occurs spontaneously and may be associated with arachnoidal cysts or, possibly, infarction of the diaphragma and pituitary. Secondary empty sella follows pituitary surgery or radiotherapy (see Chap. 11 and Chap. 17) and may also be seen in cases of elevated intracranial pressure (e.g., pseudotumor cerebri or hydrocephalus). Neuroradiographic evidence of a reversible empty sella syndrome after therapy for idiopathic intracranial hypertension has been reported.98 Visual field defects, hypopituitarism, headaches, and spinal fluid rhinorrhea occasionally occur. A thorough review of the clinical and radiographic characteristics of primary empty sella99 has revealed the following features: obese women predominate, ranging in age from 27 to 72 years, with a mean age of 49 years; headache is a common symptom; there is no vision impairment because of chiasmal interference; usually, an enlarged sella turcica is found serendipitously on radiologic studies obtained for evaluation of headaches, syncope, or other symptoms; pseudo-tumor cerebri was present in 13% of patients; approximately two-thirds of the patients had normal pituitary function; and the remaining one-third demonstrated endocrine disturbances, including panhypopituitarism and growth hormone, gonadotropin, and thyrotropin deficiency. In another series of patients with primary empty sella,100 the following features are noteworthy: all 19 were female; 12 patients initially reported headache; in 7, vision disturbances were prominent subjective symptoms (blurred vision, diplopia, micropsia); 3 patients had bilateral papilledema, and pseudotumor cerebri was diagnosed; and 2 patients demonstrated minimal, relative hemianopias without obvious cause. Additionally, visual field defects typical of those seen in glaucoma are well documented in patients with empty sella syndrome; the normal intraocular pressures implicate the empty sella syndrome as a potential cause of so-called low-tension glaucoma.101 Secondary empty sella occurs after pituitary surgery or radiotherapy, wherein adhesions form between the tumor capsule (or sellar diaphragm) and the nerves and chiasm. Retraction of these adhesions into the empty sella draws the chiasm and nerves downward, with resulting visual defects. Packing the sellar cavity to elevate the diaphragma (chiasmapexy) has been suggested for prophylactic purposes102 or after the fact. Primary empty sella may rarely occur in children in association with multiple congenital anomalies, including the de Morsier syndrome.103

MISCELLANEOUS LESIONS OF THE OPTIC CHIASM

TRAUMATIC CHIASMAL SYNDROME Vision loss that follows closed-head trauma usually is attributed to contusion or laceration of the optic nerves occurring abruptly at the time of impact. Much less frequently, a chiasmal syndrome may be identified by the pattern of field loss and associated deficits, including diabetes insipidus, anosmia, CSF rhinorrhea, and fractures of the sphenoid bone. From a report of several such patients,104 it was clear that neither the degree of vision loss nor the extent of diabetes insipidus was necessarily related to the severity of craniocerebral trauma. Transient diabetes insipidus was present in approximately one-half of these patients. Rarely, panhypopituitarism may occur.105 The traumatic chiasmal syndrome may occur more commonly than recognized because of its frequent association with extensive basilar skull fractures and its concomitant altered level of consciousness and high mortality rate.106 Lesions of the hypophysial stalk and, more frequently, of the hypothalamus may follow blunt head trauma. Hypothalamic lesions have been noted in 42% of patients who died after head trauma.107 Ischemic lesions and microhemorrhages were attributed to shearing of small perforating vessels. METASTATIC LESIONS Pituitary metastases are uncommon manifestations of systemic cancer and, initially, may be difficult to distinguish from simple adenomas. To ascertain the incidence of pituitary tumors in cancer patients and to characterize the clinical presentations of metastases, the experience at Memorial Sloan-Kettering Cancer Center was reviewed.108 Also, a series of 500 consecutive autopsies was analyzed, with inclusion of examination of the pituitary gland. In the clinical series, histologic diagnosis was made in 60% of patients. Radiologic evaluation, including polytomography and CT, did not reliably distinguish metastasis from adenoma, but the clinical syndromes were distinctive. In the metastasis group, the review108 revealed an 82% incidence of diabetes insipidus but vision loss in only 11%. In the autopsy series, metastases were found in 36% of cases and adenomas in 1.8%. Two other reported cases109 of sellar metastases showed diplopia resulting from palsies of the third, fourth, and sixth cranial nerves and eventually diabetes insipidus in one patient. In another report, a man with known colon carcinoma developed panhypopituitarism, hyperprolactinemia, chiasmal field loss, and a right third nerve palsy but no diabetes insipidus.110 Several generalizations emerge from these reports of meta-static involvement of the pituitary gland: either the anterior or posterior lobe may be involved; diabetes insipidus is more common than with simple adenoma; cranial nerve palsies are more common than simple adenomas; hyperprolactinemia may be seen, but the serum prolactin level usually is <200 ng/mL. INFLAMMATORY LESIONS Chiasmal neuritis has been well documented and may occur in multiple sclerosis, systemic lupus erythematosus, and a variety of other vasculitic and autoimmune disorders. Such a case with positive serology for Lyme disease has been reported.111 Sarcoidosis has a well-known predilection for leptomeninges at the base of the brain. The hypothalamus, pituitary, and optic chiasm all are involved in a variety of cases.112 Chiasmal arachnoiditis, secondary to tuberculosis or syphilis, is now a rare cause of the chiasmal syndrome in the United States but may be seen with increasing frequency in various immunodeficient states. VASCULAR ABNORMALITIES Such lesions are indeed rare, but cavernous malformations appear to be the most common intrachiasmal vascular anomaly.113 Suprasellar hemangioma with progressive vision loss has also been reported to occur in von Hippel-Lindau disease.114 CHAPTER REFERENCES
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. Walker AE. The neurosurgical evaluation of the chiasmal syndromes. Am J Ophthalmol 1962; 54:563. Schaeffer JP. Some points in the regional anatomy of the optic pathway, with special reference to tumors of the hypophysis cerebri and resulting in ocular changes. Anat Rec 1924; 28:243. Bergland RM, Ray BS, Torack RM. Anatomical variations in the pituitary gland and adjacent structure in 225 human autopsy cases. J Neurosurg 1968; 28:93. Bergland RM, Ray BS. The arterial supply of the human optic chiasm. J Neurosurg 1969; 31:327. Kupfer C, Chumbley L, Downer J de C. Quantitative histology of optic nerve, optic tract and lateral geniculate nucleus of man. J Anat 1967; 101:393. Horton JC. Wilbrand knee of the primate optic chiasm is an artefact of monocular enucleation. Trans Am Ophthalmol Soc 1997; 95:579. Barber AN, Ronstrom GN, Muelling RJ Jr. Development of the visual pathway: optic chiasm. Arch Ophthalmol 1984; 52:447. Scott IU, Warman R, Altman N. Bilateral aplasia of the optic nerves, chiasm, and tracts in an otherwise healthy infant. Am J Ophthalmol 1997; 124:409. Apkarian P, Bour LJ, Barth PG, et al. Non-decussating retinal-fugal fibre syndrome. An inborn achiasmatic malformation associated with visuo-topic misrouting, visual evoked potential ipsilateral asymmetry and nystagmus. Brain 1995; 118:1195. Ikeda H, Yoshimoto T. Visual disturbances with pituitary adenoma. Acta Neurol Scand 1995; 92:157. Gutowski NJ, Heron JR, Scase MO. Early impairment of foveal magno- and parvocellular pathways in juxta chiasmal tumours. Vision Research 1997; 37:1401. Hollenhorst RW, Younge BR. Ocular manifestations produced by adenomas of the pituitary gland: analysis of 1000 cases. In: Kohler PO, Ross GT, eds. Diagnosis and treatment of pituitary tumors. Amsterdam: Excerpta Medica, 1973:53. Elkington SG. Pituitary adenoma: preoperative symptomatology in a series of 260 patients. Br J Ophthalmol 1968; 52:322. Schmidt D, Kommerell G. Schaukel-Nystagmus (see-saw nystagmus) mit bitemporaler Hemanopie als Folg von Schadelhirnstraumen. Graefes Arch Clin Exp Ophthalmol 1969; 178:349. Chamlin M, Davidoff LM, Feiring EH. Ophthalmologic changes produced by pituitary tumors. Am J Ophthalmol 1955; 40:353. Wilson P, Falconer MA. Patterns of visual failure with pituitary tumors: clinical and radiological correlations. Br J Ophthalmol 1968; 52:94. Glaser JS. Neuro-ophthalmology. Hagerstown, MD: Harper & Row, 1978:135. de Morsier G. tudes sur les dysraphias cranio-encephaliques. III. Agenesie du septum lucidum avec malformation du tractus optique: la dysplasie septo-optique. Schweiz Arch Neurol Psychiatr 1956; 77:267. Lambert SR, Hoyt CS, Narahara MH. Optic nerve hypoplasia. Surv Ophthalmol 1987; 32:1. Glaser JS. The optic chiasm. In: Lessell S, van Dalen JTW, eds. Neuro-ophthalmology 1982. Amsterdam: Excerpta Medica, 1982:61. de Morsier G. Median cranioencephalodysraphias and olfactogenital dysplasia. World Neurol 1962; 3:485. Cagianut B, Sigg P, Isler W, Friede R. Dysplasia opticoseptalis. Klin Monatsbl Augenheilkd 1980; 176:699.

23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. 103. 104. 105. 106. 107. 108. 109. 110. 111. 112. 113. 114.

Lemyre E, Lemieux N, Decarie JC, Lambert M. Del (14)(q22.1q23.2) in a patient with anophthalmia and pituitary hypoplasia. Am J Med Genet 1998; 77:162. Carey ML, Friedman TB, Asher JH Jr, Innis JW. Septo-optic dysplasia and WS1 in the proband of a WS1 family segregating for a novel mutation in PAX3 exon 7. J Med Genet 1998; 35:248. Dattani MT, Martinez-Barbera JP, Thomsa PQ, et al. Mutations in the homeobox gene HESX1/Hesx1 associated with septo-optic dysplasia in human and mouse. Nat Genet 1998; 19:125. Lubinsky MS. Hypothesis: septo-optic dysplasia is a vascular disruption sequence. Am J Med Genet 1997; 69:235. Brodsky MC, Glasier CM. Optic nerve hypoplasia. Clinical significance of associated central nervous system abnormalities on magnetic resonance imaging. Arch Ophthalmol 1993; 111(1):66. Sorkin JA, David PC, Meacham LR, et al. Optic nerve hypoplasia: absence of posterior pituitary bright signal on magnetic resonance imaging correlates with diabetes insipidus. Am J Ophthalmol 1996; 122:717. Abernethy LJ, Qunibi MA, Smith CS. Normal MR appearances of the posterior pituitary in central diabetes insipidus associated with septo-optic dysplasia. Pediatr Radiol 1997; 27:45. Siatkowski RM, Sanchez JC, Andrade R, Alvarez A. The clinical, neurora-diographic, and endocrinologic profile of patients with bilateral optic nerve hypoplasia. Ophthalmology 1997; 104:493. Brodsky MC, Conte FA, Taylor D, et al. Sudden death in septo-optic dysplasia. Report of 5 cases. Arch Ophthalmol 1997; 115:66. Landau K, Bajka JD, Kirchschlager BM. Topless optic disks in children of mothers with type I diabetes mellitus. Am J Ophthalmol 1998; 125:605. Edward WC, Layden WE. Optic nerve hypoplasia. Am J Ophthalmol 1970; 70:950. Walton DS, Robb RM. Optic nerve hypoplasia: a report of 20 cases. Arch Ophthalmol 1970; 84:572. Boniuk V, Ho PK. Ocular findings in anencephaly. Am J Ophthalmol 1979; 88:613. Greenfield PS, Wilcox LM Jr, Weiter JJ, Adelman L. Hypoplasia of the optic nerve in association with porencephaly. J Pediatr Ophthalmol Strabismus 1980; 17:75. Lloyd L, Buncic JR. Hypoplasia of the optic nerve and disc. In: Smith JS, ed. Neuro-ophthalmology focus 1980. New York: Masson, 1979:85. Apple DJ, Rabb MF, Walsh PM. Congenital anomalies of the optic disc. Surv Ophthalmol 1982; 27:3. Pollack JA, Newton TH, Hoyt WF. Transsphenoidal and transethmoidal encephaloceles: a review of clinical and roentgen features of 8 cases. Radiology 1968; 90:442. Males JS, Townsend JS, Schneider RA. Hypogonadotrophic hypogonadism with anosmia: Kallmann's syndrome. Arch Intern Med 1973; 131:501. Bosma JF, Henkin RI, Christiansen RL, Herdt JR. Hypoplasia of the nose and eyes, hyposmia, hypogeusia, and hypogonadotrophic hypogonadism in two males. J Craniofac Genet Dev Biol 1981; 1:153. Chang RJ, Davidson BJ, Carlson HE, et al. Hypogonadotrophic hypogonadism associated with retinitis pigmentosa in a female sibship: evidence for gonadotropin deficiency. J Clin Endocrinol Metab 1981; 53:1179. Rubinstein AE, Lovelace RE, Behrens MM, Weisberg LA. Moebius syndrome in association with peripheral neuropathy and Kallmann syndrome. Arch Neurol 1975; 32:480. Abid F, Hall R, Hudgson P, Weiser R. Moebius syndrome peripheral neuropathy and hypogonadotrophic hypogonadism. J Neurol Sci 1978; 35:309. Triulzi F, Scott G, diNatale B, et al. Evidence of a congenital midline brain anomaly in pituitary dwarfs: a magnetic resonance imaging study in 101 patients. Pediatrics 1994; 93:409. Alvord EC, Loftin S. Gliomas of optic nerve and chiasm: outcome by patients' age, tumor site, and treatment. J Neurosurg 1988; 68:85. Blatt J, Jaffe R, Deutsch M, et al. Neurofibromatosis and childhood tumors. Cancer 1986; 57:1225. Glaser JS, Topical diagnosis: the optic chiasm. In: Duane TD, Jaeger EA, eds. Clinical ophthalmology. Philadelphia: JB Lippincott, 1989:6. Imes RK, Hoyt WF. Childhood chiasmal gliomas: update on the fate of patients in the 1969 San Francisco study. Br J Ophthalmol 1986; 70:179. Shuper A, Horev G, Kornreich L, et al. Visual pathway glioma: an erratic tumour with therapeutic dilemmas. Arch Dis Child 1997; 76:259. Optic chiasm astrocytomas of childhood. 1. Long-term follow-up. Pediatr Neurosurg 1997; 27:121. Collet-Solberg PF, Sernyak H, Satin-Smith M, et al. Endocrine outcome in long-term survivors of low-grade hypothalamic/chiasmatic glioma. Clin Endocrinol (Oxf) 1997; 47:79. Gropman AL, Packer RJ, Nicholson HS, et al. Treatment of diencephalic syndrome with chemotherapy: growth, tumor response, and long term control. Cancer 1998; 83:166. Chamberlain MC, Grafe MR. Recurrent chiasmatic-hypothalamic glioma treated with oral etoposide. J Clin Oncol 1995; 13:2072. Russell A. A diencephalic syndrome of emaciation in infancy and childhood. Arch Dis Child 1951; 26:274. Pitlyk PJ, Miller RH, Johnson GM. Diencephalic syndrome of infancy presenting with anorexia and emaciation. Mayo Clin Proc 1965; 40:327. De Sousa AC, Kalsbeck JE, Mealey J, Fitzgerald J. Diencephalic syndrome and its relation to optico-chiasmatic glioma: review of twelve cases. Neuro-surgery 1979; 4:207. Poussaint TY, Barnes PD, Nichols K, et al. Diencephalic syndrome: clinical features and imaging findings. AJNR Am J Neuroradiol 1997; 18:1499. Glaser JS, Hoyt WF, Corbett J. Visual morbidity with chiasmal glioma: long-term studies of visual fields in untreated and irradiated cases. Arch Ophthalmol 1971; 85:3. Tabaddor K, Shulman K, Del Canto MC. Neonatal craniopharyngioma. Am J Dis Child 1984; 128:381. McLone DG, Raimondi AJ, Naidich TP. Craniopharyngiomas. Childs Brain 1982; 9:188. Carmel PW, Antunes JS, Chang CH. Craniopharyngiomas in children. Neurosurgery 1982; 11:382. Fischer EG, Welsh K, Belli JA, et al. Treatment of craniopharyngioma in children 19711980. J Neurosurg 1985; 62:496. Hoffman HJ, Buncic JR. Craniopharyngioma in children. In: Smith JS, ed. Neuro-ophthalmology update. New York: Masson, 1977:241. Sachs BP, Smith SK, Casser J, Van Iddekinge B. Rapid enlargement of craniopharyngioma in pregnancy. Br J Obstet Gynaecol 1978; 85:577 Sumid M, Uozimi T, Mukada K, et al. Rathke cleft cysts: correlation of enhanced MR and surgical findings. AJNR Am J Neuroradiol 1994; 15:525. Bowman CB, Farris BK. Primary chiasmal germinoma: a case report and review of the literature. J Clin Neuro Ophthalmol 1990; 10:9. Spanziante R, de Devitiis E, Stella L, et al. Benign intrasellar cysts. Surg Neurol 1981; 15:274. Verier Mine O, Salomez-Granier F, Buvat J, et al. propos d'un cas de kyste arachnoidien intrasellaire. Semaine Des Hpitaux 1983; 59:408. Izquierdo JM, Rougerie J, Lapras C, Sanz F. The so-called ectopic pinealomas: a cooperative study of 15 cases. Childs Brain 1979; 5:505. Kernohan JW, Sayre GP. Tumors of the pituitary gland and infundibulum. In: Atlas of tumor pathology, ser 1, fascicle 36. Washington, DC: Armed Forces Institute of Pathology, 1956:15. Burrow GN, Wortzman G, Rewcastle NB, et al. Microadenomas of the pituitary and abnormal sellar tomograms in unselected autopsy service. N Engl J Med 1981; 304:156. Poussaint TY, Barnes PD, Anthony DC, et al. Hemorrhagic pituitary adenomas of adolescence. AJNR Am J Neuroradiol 1996; 17:1907. Falconer MA, Stafford-Bell MA. Visual failure from pituitary and parasellar tumours occurring with favorable outcome in pregnant women. J Neurol Neurosurg Psychiatry 1975; 38:919. Miller NR. Walsh and Hoyt's clinical neuro-ophthalmology, 4th ed. Baltimore: Williams & Wilkins, 1982:123. Mikami T, Uozimi T, Yamamaka M. Lymphocytic adenohypophysitis: MRI findings of a suspected case. No Shinkei Geka 1989; 17:871. Powell M. Recovery of vision following transsphenoidal surgery for pituitary adenomas. Br J Neurosurg 1995; 9:367. McCord MW, Buatti JM, Fennell EM, et al. Radiotherapy for pituitary adenoma: long-term outcome and sequelae. Int J Radiat Oncol Biol Phys 1997; 39:437. Yoon SC, Suh TS, Jang HS, et al. Clinical results of 24 pituitary macroadenomas with linac-based stereotactic radiosurgery. Int J Radiat Oncol Biol Phys 1998; 41:849. Moster ML, Savino PJ, Schatz NJ, et al. Visual function in prolactinoma patients treated with bromocriptine. Ophthalmology 1985; 92:1332. Crosignani P, Ferrari C, Mattei AM. Visual field defects and reduced visual acuity during pregnancy in two patients with prolactinoma: rapid regression of symptoms under bromocriptine. Case reports. Br J Obstet Gynaecol 1984; 91:821. Taxel P, Waitzman DM, Harrington JF Jr, et al. Chiasmal herniation as a complication of bromocriptine therapy. J Neuro-ophthalmol 1996; 16:252. Grochowicki M, Khalfallah Y, Vighetto A, et al. Ophthalmic results in patients with macroprolactinomas treated with a new prolactin inhibitor CV 205502. Br J Ophthalmol 1993; 77:785. Warnet A, Harris AG, Renard E, et al. A prospective multicenter trial of octreotide in 24 patients with visual defects caused by nonfunctioning and gonadotropin-secreting pituitary adenomas. Neurosurgery 1997; 41:786. Amiya T, Toibana M, Hashimoto M, et al. Diabetic retinopathy in acromegaly. Ophthalmologica 1978; 176:74. Bramsen T, Klauber A, Bjerre P. Central corneal thickness and intraocular tension in patients with acromegaly. Acta Ophthalmol 1980; 58:971. Dal Pozzo G, Boschi MC. Extraocular muscle enlargement in acromegaly. J Comput Assist Tomogr 1982; 6:706. Rosenthal JW, Kloepfer HW. An acromegaloid, cutis verticis gyrata, corneal leukoma syndrome: a new medical entity. Arch Ophthalmol 1962; 68:722. Azar P, Rothschild H, Rosenthal JW, Ichinose H. Histopathology of corneal leukoma and fibroma in the ACL syndrome. Hum Pathol 1983; 14:188. David NJ, Gargano FP, Glaser JS. Pituitary apoplexy in clinical perspective. In: Glaser JS, Smith JL, eds. Neuro-ophthalmology, vol 8. Symposium of the University of Miami and the Bascom Palmer Eye Institute. St. Louis: CV Mosby, 1975:140. Symon L, Mohanty S. Haemorrhage in pituitary tumours. Acta Neurochir (Wien) 1982; 65:41. Davis PC, Hoffman JC, Spencer T, et al. MR imaging of pituitary adenomas: CT, clinical, and surgical correlation. AJNR Am J Neuroradiol 1987; 8:107. Grossman A, Cohen BL, Charlesworth M, et al. Treatment of prolactinomas with megavoltage radiotherapy. Br Med J 1984; 288:1105. Warman R, Glaser JS. Radionecrosis of optico-hypothalamic glioma. Neuro-ophthalmology 1989; 9:299. Capo H, Kupersmith M. Efficacy and complications of radiation therapy of anterior visual pathway tumors. Neurol Clin 1991; 9(1):179. Kaufman B. The empty sella turcica, a manifestation of the intrasellar subarachnoid space. Radiology 1968; 90:931. Bergland RM, Ray BS, Torack RM. Anatomical variations in the pituitary gland and adjacent structures in 225 human autopsy cases. J Neurosurg 1968; 28:93. Zagardo MT, Cail WS, Kelman SE, Rothman MI. Reversible empty sella in idiopathic intracranial hypertension: an indicator of successful therapy? AJNR Am J Neuroradiol 1996; 17:1953. Neelon FA, Goree JA, Lebovitz HE. The primary empty sella: clinical and radiographic characteristics and endocrine function. Medicine (Baltimore) 1973; 52:73. Berke JP, Buxton LF, Kokmen E. The empty sella. Neurology 1975; 25:1137. Beattie AM, Trope GE. Glaucomatous optic neuropathy and field loss in primary empty sella syndrome. Can J Ophthalmol 1991; 26(7):377. Olson DR, Guiot G, Derome P. The symptomatic empty sella: prevention and correction via the transsphenoidal approach. J Neurosurg 1972; 37:533. Wilkinson IA, Duck SC, Gager WE, Daniels DL. Empty sella syndrome: occurrences in childhood. Am J Dis Child 1982; 136:245. Savino PJ, Glaser JS, Schatz NJ. Traumatic chiasmal syndrome. Neurology 1980; 30:963. Gilad E, Dickerman Z, Laron Z, et al. Traumatic chiasmal syndrome with panhypopituitarism. Neuro Ophthalmol 1986; 6:79. Heinz GW, Nunery WR, Grossman CB. Traumatic chiasmal syndrome associated with midline skull fractures. Am J Ophthalmol 1994; 117:90. Crompton MR. Hypothalamic lesions followed closed head injury. Brain 1971; 94:165. Max MB, Deck F, Rottenberg DA. Pituitary metastases: incidence in cancer patients and clinical differentiation from pituitary adenoma. Neurology 1981; 31:998. Neeten SA, Mahler K, Bultinck J, Martin JJ. Sellar metastatic disease. Neuro Ophthalmol 1982; 2:255. Leramo OB, Booth JD, Zinman B, et al. Hyperprolactinemia, hypopituitarism, and chiasmal compression due to carcinoma metastatic to the pituitary. Neurosurgery 1981; 8:477. Scott IU, Silva-Lepe A, Siatkowski RM. Chiasmal optic neuritis in Lyme disease. Am J Ophthalmol 1997; 123:136. Sherman JL, Stern B. Sarcoidosis of the CNS: comparison of unenhanced and enhanced MRI images. AJR Am J Roentgenol 1990; 155:1293. Shibuya M, Baskaya MK, Saito K, et al. Cavernous malformations of the optic chiasm. Acta Neurochir (Wien) 1995; 136:29. Balcer LJ, Galetta SL, Curtis M, et al. von Hippel-Lindau disease manifesting as a chiasmal syndrome. Surv Ophthalmol 1995; 39:302.

CHAPTER 20 DIAGNOSTIC IMAGING OF THE SELLAR REGION Principles and Practice of Endocrinology and Metabolism

CHAPTER 20 DIAGNOSTIC IMAGING OF THE SELLAR REGION

ERIC BOUREKAS, MARY OEHLER, AND DONALD CHAKERES Advantages and Disadvantages of Imaging Techniques Normal Imaging Anatomy Skull Radiographs Computed Tomographic Imaging Magnetic Resonance Imaging Normal Variants Empty Sella Congenital Anomalies Neoplasms Craniopharyngioma Optic Chiasm and Hypothalamic Gliomas Pituitary Adenomas Infundibular Masses Hypothalamic Hamartomas Sellar and Parasellar Meningioma Schwannomas and Neurofibromas Metastasis Germ-Cell Tumors Miscellaneous Entities Pituitary Apoplexy Sphenoid Sinus Disease Aneurysm Sarcoidosis Primary Hypothyroidism and Nelson Syndrome Postoperative Changes Conclusion Chapter References

Radiographic imaging of the sella and hypothalamic region is important in the overall evaluation of many patients with endocrine and metabolic disorders. Often, imaging is crucial, because different pathology in this region may present with similar clinical findings. Imaging can help identify a number of diagnostic entities based on anatomicopathologic characteristics. For example, in patients with hypopituitarism, one can often pinpoint the etiology more precisely, determining, for example, whether it is secondary to congenital absence of the gland, septo-optic dysplasia, transection of the pituitary stalk, a large pituitary mass, a suprasellar arachnoid cyst, a hypothalamic glioma, or an eosinophilic granuloma. In addition, in those patients who have known endocrine disorders such as acromegaly, imaging characterizes other important anatomic aspects of the pathology, such as the size of the mass and possible involvement of the optic chiasm or cavernous sinuses. Imaging of the venous drainage of the pituitary fossa is also used to allow for venous sinus sampling of blood for hormone analysis. Finally, imaging helps direct interventional procedures (i.e., surgery, radiotherapy) in the sellar region and plays a role in postoperative evaluation. In this chapter, the various imaging modalities and their advantages and disadvantages are discussed in terms of their specific applications. A review is provided of the normal imaging anatomy and of the more commonly encountered pathologic entities, including normal variations, congenital anomalies, inflammations, neoplasms, trauma, and vascular disorders. The focus is primarily on disorders associated with endocrine malfunction.

ADVANTAGES AND DISADVANTAGES OF IMAGING TECHNIQUES

Traditionally, the first imaging examination in the evaluation of the pituitary-sellar region was a skull radiograph series.1 Plain radiographs offer the advantage of being noninvasive and inexpensive; they yield a certain amount of limited information about the adjacent bony structures (Fig. 20-1). However, potentially important secondary soft-tissue changes can only be inferred, although soft-tissue calcifications can be seen. Thus, in general, routine radiographs provide only minimal information related to the sella, and hence are generally not obtained.

FIGURE 20-1. Lateral radiograph of sella. Labeled structures include planum sphenoidale (curved solid white arrows), tuberculum sella (short black arrow), lamina terminalis (open white arrow), sphenoid sinus (black arrowheads), anterior clinoid (long thin black arrow), and posterior clinoid (open curved white arrow). This patient has a relatively well-pneumatized sphenoid sinus. The chiasmatic sulcus is not well marginated.

Computed tomography (CT) was the first imaging modality to directly visualize the pituitary gland, hypothalamus, and optic chiasm.2 The bony structures in this region can be well evaluated with CT (Fig. 20-2). It is more sensitive than either plain radiographs or magnetic resonance imaging (MRI) in the detection of calcifications within soft tissues. However, intravenous contrast agents frequently are necessary to improve the image contrast and to enhance the vasculature, and CT involves radiation exposure.

FIGURE 20-2. Normal sellar and suprasellar CT. A, Coronal CT scan obtained with contrast enhancement through the anterior third ventricle (short curved solid white arrow), supraclinoid carotid artery (thick black arrow), calcified cavernous carotid artery (long curved solid white arrow), pituitary gland (open white arrow), anterior clinoid (small black arrowhead), cavernous sinus (long white arrow), incidental fat in the cavernous sinus (large black arrowhead), and sphenoid sinus (white arrowhead). B, Axial CT scan made with contrast enhancement through the suprasellar cistern. The labeled structures include anterior clinoids (small black arrowhead), posterior clinoids (curved solid white arrow), optic canal (curved open white arrow), pneumatized planum sphenoidale (large black arrowhead), pituitary

stalk (long thin black arrow), and supraclinoid carotid artery (short black arrow).

A major disadvantage of CT is that soft-tissue characterization is less than ideal. Artifacts from beam hardening, related to the dense bone in the skull base, obscure soft-tissue detail. Additionally, on direct coronal imaging, there often are artifacts that are due to metallic dental fillings. Intravenous contrast is used to improve soft-tissue contrast and highlight the vasculature, but the contrast agent may cause life-threatening reactions. MRI is the most important single overall imaging modality for the sellar region because it is effective in simultaneously characterizing the soft tissues, cerebrospinal fluid (CSF) spaces, and blood vessels (Fig. 20-3).3 It is an extremely flexible imaging modality for which contrast can be extensively manipulated. The images may be acquired using a number of different techniques, each of which has specific advantages. For example, MRI (as magnetic resonance angiography [MRA]) can substitute for routine catheter angiography in most situations.

FIGURE 20-3. Normal sellar and suprasellar MRI. A, Sagittal contrast-enhanced T1-weighted 3D gradient echo image. Labeled structures include the pituitary gland (PG), pituitary stalk (PS), optic chiasm (OC), corpus callosum (CC), septal vein (SV), inferior sagittal sinus (ISS), massus intermedius (MI), anterior cerebral artery (AC), anterior third ventricle (AAT), brainstem (BS), cerebral aqueduct (CA), and mammillary body (MB). B, Coronal contrast-enhanced T1-weighted image with angiographic technique. Labeled structures include the cavernous carotid artery (small black arrowhead), supraclinoid carotid artery (short black arrow), pituitary stalk (short white arrow), optic chiasm (long black arrow), infundibular recess of the third ventricle (curved solid white arrow), sphenoid sinus (open white arrow), cavernous sinus (curved open white arrow), and pituitary gland (large black arrowheads). C, Axial 3D gradient echo contrast-enhanced T1-weighted pituitary fossa. Labeled structures include the cavernous carotid artery (CA), sphenoid sinus (SS), cavernous sinus (CS), basilar artery (BA), third cranial nerve (3CN), superior orbital fissure (SOF), brainstem (BS), and pituitary gland (PG). D, Axial 3D gradient echo contrast-enhanced T1-weighted suprasellar cistern. Labeled structures include supraclinoid carotid artery (CA), suprasellar cistern (SSC), pituitary stalk (PS), basilar artery (BA), optic nerve (ON), rectus gyrus (RG), and brainstem (BS). E, Axial 3D gradient echo contrast-enhanced T1-weighted hypothalamus optic complex. Labeled structures include anterior cerebral artery (ACA), mammillary bodies (MB), anterior third ventricle (ATV), interpeduncular cistern (IPC), optic tract (OT), hypothalamus (HT), cerebral aqueduct (CA), and midbrain (MB). F, Axial spin-echo noncontrast-enhanced T1-weighted anterior posterior pituitary. Labeled structures include anterior pituitary (white arrow) and posterior pituitary (black arrow).

Although not hampered to the same extent as CT by artifacts, some magnetic susceptibility artifacts may be related to bone and air interfaces and to dental fillings; these are not as significant a problem with MRI as with CT. Moreover, MRI allows multiplanar acquisitions more readily than CT. There is no ionizing radiation with MRI, and the contrast agents are associated with far fewer serious reactions. Because of these advantages, MRI has become the method of choice for the evaluation of the pituitary gland and adjacent soft tissues, unless there is specific interest in the bone or in soft-tissue calcifications. With precautions, the risks associated with MRI are low. Contraindications to MRI include cardiac pacemaker, a cerebral spring aneurysm clip, a metallic foreign body near a vital structure (e.g., within the eye, the spinal canal, or near a major blood vessel), neural stimulators, and some medical implants. For these patients, CT remains a possible alternative. Cerebral angiography is still an important imaging technique and remains the gold standard for details of vasculature.4 It is the examination of choice for the evaluation of aneurysms of the sellar/parasellar region and in the evaluation of profuse bleeding during or after transsphenoidal surgery, which may be secondary to injury of the sphenopalatine arteries or the internal carotids.5 Inferior petrosal venous sinus angiography and blood sampling can also be helpful.6 The main disadvantage of angiography is the risk of serious complications. In general, although the incidence of complications is low, the results can be catastrophic (infarcts and vascular occlusions). Even venous studies may have serious complications, such as brainstem infarction. The procedure is uncomfortable or painful for the patient and is associated with radiation exposure and potential contrast complications.

NORMAL IMAGING ANATOMY

SKULL RADIOGRAPHS An evaluation of subtle changes of the bony architecture of the sella is not of great clinical use because there are many normal variations in the size, shape, and cortical margins of the sella. In part, this is due to differences in the configuration of the sphenoid bone and sphenoid sinus. It is important to realize that variations in the shape of the sella do not always reflect the pathology of the adjacent structures. The sella is usually evaluated with lateral and posteroanterior plain radiographs (tomography has been largely replaced by CT). On the lateral projection, the sella turcica is a -shaped structure that is seen as a crisp, dense cortical bone margin (see Fig. 20-1). Those segments of the sella that are outlined by the underlying air spaces of the sinuses exhibit a thin cortical margin, which is referred to as the lamina dura. If the sinus is opacified or not pneumatized and filled with bone marrow, the sellar margin is thicker. The planum sphenoidale is a flat, thin bone that forms the roof of the anterosuperior sphenoid sinus and extends posteriorly to terminate in the anterior clinoid processes. Between the anterior clinoid processes, the planum sphenoidale ends at a small arc-shaped ridge of bone called the limbus; this forms the anterior margin of a small bony depression called the chiasmatic sulcus (which conforms to the optic chiasm and nerves). This concavity is where the optic chiasm divides into the optic nerves. The lateral margins of the chiasmatic depression extend into the optic canals that, in turn, extend anteriorly and laterally. Superior to the optic canals are the anterior clinoids, which are visible on lateral radiograph. The inferoposterior margin of the chiasmatic sulcus is the tuberculum sella, which forms the most anterior and superior margin of the sella turcica (hypophyseal fossa). The diaphragma sellae is the superior dural covering of the sella that extends from the tuberculum sella to the posterior clinoids. It cannot be seen on plain radiographs, but marks the true roof of the sella turcica. Usually, the floor of the sella is not perfectly hemicircular in configuration, but is more oval in shape. The normal range of dimensions of the bony sella from anterior to posterior on lateral radiograph is 5 to 16 mm, with a mean of 10.6 mm. The normal range of dimensions in depth is 4 to 12 mm, with a mean of 8.1 mm. 7 The anteroposterior projection of the sella usually is less informative than the lateral, but it may be useful in defining the lateral margins of the sphenoid sinus and the floor of the sella. The floor of the sella frequently bows inferiorly slightly. Septations of the sphenoid sinus are frequently seen and may account for variations in the shape of the sella and pituitary gland. COMPUTED TOMOGRAPHIC IMAGING On CT, bone and calcified structures have a high density (similar to that seen in plain radiographs), whereas the air spaces are of low density. The CSF spaces are of intermediate density, with the fat structures (such as periorbital fat) being lower in density than the CSF. The brain and pituitary gland usually have a slightly higher density than the CSF. With the use of intravenous contrast agents, the blood vessels, cavernous sinus, pituitary gland, and pituitary stalk all enhance intensely, thus increasing their apparent density (see Fig. 20-2). This helps define the pituitary gland and the cranial nerves in the cavernous sinus. Rarely, intrathecal contrast is used for the evaluation of a CSF fistula. In general, contrast-enhanced direct coronal CT images of the sella are the most informative, although axial sections are easier for patient positioning. If one is not able to perform direct coronal imaging, computer reconstructions of axial images into sagittal and coronal sections can be obtained, although their detail is limited. On

coronal CT, the bony sella is seen as a flat to minimally concave structure with the sphenoid sinus directly inferior (see Fig. 20-2A). The lateral walls of the sella are formed by the cavernous sinuses, which are visualized on enhanced CT as contrast-filled structures bounded laterally by dura. The individual cranial nerves (oculomotor, abducens, and trochlear) can be seen in the cavernous sinus. Meckel cave, which is a CSF space extending from the posterior fossa, contains the fifth cranial nerve. The dural margins of the cave enhance, while the CSF spaces are seen as low-density, vertical, oval cavities. The lateral margins of the pituitary gland, which presents as an oval-shaped soft-tissue structure, may be difficult to differentiate from those of the cavernous sinus. The pituitary gland usually has a flat superior margin; however, it is common for adolescents and menstruation-aged women to have a slightly convex upward superior margin. This finding can also be seen in small tumors and is less common in men. The infundibulum or pituitary stalk is seen as a thin, usually midline tubular structure, extending inferiorly from the median eminence of the arc-shaped hypothalamus to insert on the superior margin of the pituitary gland, usually in the anterior third. Both the pituitary gland and the infundibulum enhance with intravenous contrast; this is partially due to the portal venous plexus and partially due to the lack of a bloodbrain barrier. Directly superior to the infundibulum, in the anterior third of the sella turcica, is the optic chiasm, which is oriented horizontally across the top of the sella. The middle portion of the chiasm is rectangular in shape. Farther anteriorly, the chiasm is more figure-8 shaped as it divides into the individual optic nerves. Axial CT images of the sella demonstrate the air-filled sphenoid inferiorly. Asymmetry of the floor is common. The carotid arteries, which are frequently calcified in elderly patients (see Fig. 20-2A), are seen lateral to the gland in the cavernous sinus. The cavernous sinuses form the lateral flat margins adjacent to the temporal lobes. Medially, the margins between the cavernous sinus and normal pituitary may not be clear. Incidental rests of fat are frequently seen in the cavernous sinus and should not be confused with air. The pituitary gland has a circular configuration, and the suprasellar cistern has a five-pointed-star configuration. Centrally, the pituitary stalk should be visible. The anterior and posterior clinoids are also well demonstrated on axial images (see Fig. 20-2B). MAGNETIC RESONANCE IMAGING MRI is the examination of choice in the evaluation of the sella and parasellar regions. Unlike CT and plain radiographs, MRI does not use ionizing radiation to generate images; they are generated based on the intrinsic magnetic properties of hydrogen atoms. Respective parameters include proton density, relaxation times (T1 and T2), magnetic susceptibility, and motion. Importantly, only hydrogen atoms are detected with MRI. Regions of cortical bone or air have no signal (signal voids). The image contrast with MRI is plastic; thus, the tissues can have almost any type of signal. With routine, spin-echo, T1-weighted images, tissues with high triglyceride fat content (orbital fat, fatty bone marrow) have bright or high signal intensity. With T1-weighted images, the CSF has a low signal intensity, whereas the brain has an intermediate signal intensity. T2-weighted images demonstrate low signal intensity for fatty structures, intermediate signal intensity for the brain, and high signal intensity for tissues with high water content (e.g., CSF, cystic masses). Motion also has an imaging impact: The signal of moving atoms can have high, low, or misregistered signal intensities, depending on the type of motion and the pulse sequence. On MRI (see Fig. 20-3), the bony portions of the sella lack signal, except for areas of fatty marrow that have a high signal intensity on T1-weighted images. The normal pituitary gland varies in size with age, growing linearly in the first years of life and developing physiologic hypertrophy at adolescence.8 In fact, the gland tends to decrease in height and increase in width and anteroposterior dimension with age.9 In adult patients younger than 50 years of age, the average height of the pituitary gland is 5.7 1.4 mm, and the average length is 10.8 1.2 mm. 10 In newborns younger than 6 weeks of age, the superior margin of the pituitary is usually concave, with an overall globular configuration.9 In infants between the ages of 6 weeks and 2 years of age, the superior margin is convex in 46% and flat in 43%.9 The superior margin is generally flat in adults, but in 56% of menstruating women, it may be convex.8 In newborns younger than 6 weeks of age, both the anterior and posterior lobes of the pituitary are of high signal, whereas in infants 6 weeks to 2 years of age, the anterior lobe is of low signal in 52% and intermediate signal in 43%, and the posterior pituitary is of intermediate signal in 55% and of high signal in 37%.9 In adults, the normal gland is isointense with white matter, while the posterior pituitary or neurohypophysis is usually of higher signal intensity on T1-weighted images11 (see Fig. 20-3F). This is usually seen best on sagittal images as a bright spot posterior to the adenohypophysis of the gland, usually in the posterior one-third of the sella. There has been much speculation about the cause of the high signal intensity within the posterior pituitary gland. It is probably related to neurosecretory granules. When the gland is ectopic, this bright spot is also seen proximally along the pituitary stalk, suggesting an accumulation of these substances in the hypothalamus.11 The normal infundibulum is 3.5-mm thick near its origin from the hypothalamus and 2.8-mm thick at its midpoint.12 On sagittal images, the stalk can be seen oriented slightly obliquely and inserting on the anterior one-third of the pituitary gland (see Fig. 20-3A). On coronal images, the stalk is usually midline and is seen extending from the hypothalamus to the upper margin of the pituitary (see Fig. 20-3B). With gadolinium enhancement, the pituitary gland and infundibulum enhance brightly because there is no normal bloodbrain barrier in this region (see Fig. 20-3C and D). There is a characteristic enhancement pattern when one performs dynamic imaging with contrast flowing through the pituitary portal system. The enhancement starts in the infundibulum, proceeds to the central gland, and then spreads to involve the more lateral parts of the gland.13 Dynamic gradient echo MRI has been shown to be superior to conventional contrast-enhanced spin-echo imaging in delineating the margins and extent of tumor.14 The optic chiasm and the hypothalamus are intimately associated and cannot be easily separated on axial images (see Fig. 20-3E). On coronal images (see Fig. 20-3B), the pituitary gland, infundibulum, and optic chiasm have a configuration similar to that seen on CT. On most spin-echo MRIs, the cavernous sinus contains tubular areas of low signal intensity related to flowing blood in the cavernous internal carotid arteries (see Fig. 20-3F). The signal can be high with some sequences. The most common pulse sequence is the time-of-flight MRA technique. With gadolinium enhancement, the remainder of the cavernous sinus enhances (except for the cranial nerves).

NORMAL VARIANTS
EMPTY SELLA The sella may not be completely filled with tissue. When the sella is partially filled with CSF in the suprasellar cistern, it is referred to as an empty sella (Fig. 20-4). A partially empty sella is a common incidental finding and usually should be considered a normal variation in anatomy. The empty sella may be related to incomplete formation of the diaphragma sella, allowing CSF into the pituitary fossa (see Chap. 11). A partially empty sella can be seen in association with pseudotumor cerebri, hydrocephalus, previous pituitary surgery, or a previous pituitary mass that has shrunk. It can also be seen after irradiation, after trauma, or as sequela of apoplexy. Occasionally, an empty sella is associated with a CSF leak or fistula. The term empty sella syndrome has been used to describe a variable clinical constellation of findings, such as headache, endocrine dysfunction, and visual disturbances, which are seen in association with a partially empty sella.15

FIGURE 20-4. Empty sella. A, Sagittal T1-weighted noncontrast-enhanced image. Note that the pituitary stalk (small black arrowhead) extends all of the way to the floor of the sella (thick black arrow). The pituitary gland is very thin and not well marginated. In this case, the clivus (large black arrowhead) has a large amount of fat generating a high signal intensity region. The other normal structures include the optic chiasm (curved white arrow) and sphenoid sinus (open white arrowhead). B, Coronal contrast-enhanced T1-weighted image. The pituitary stalk (curved white arrow) is seen extending all of the way to the floor of the sella. The gland is seen as a thin, crescent-shaped structure (thick black arrow) at the base of the sella. Other labeled structures include the hypothalamus and optic chiasm (white straight arrows) and the anterior third ventricle (open white arrow).

With an empty sella, the sellar bony margins may be expanded, forming an oval or J shape on plain radiograph or CT. It is often of normal size. The CSF spaces extend into the sella; the pituitary gland may be normal, small, or large in size. The optic nerve, optic chiasm, and optic tracts may herniate into the empty sella. The main finding is visualization of the pituitary stalk extending into the fossa, which is the best way to distinguish an empty sella from a cystic mass.

CONGENITAL ANOMALIES HYPOPLASTIC PITUITARY The pituitary gland can be congenitally hypoplastic. Many of these patients present with a combination of endocrine deficiencies in childhood. Plain radiographs of the sella demonstrate a small sella turcica, which may measure only a few millimeters in dimension. The presence of a small bony cavity is a sign that the pituitary gland never developed to a normal size rather than an indication of some type of destructive process. Although CT and MRI confirm that the pituitary gland is small, the hypothalamus and other adjacent structures usually appear intact. ECTOPIC POSTERIOR PITUITARY GLAND An ectopic posterior pituitary gland is a relatively common etiology for hypopituitarism and growth hormone (GH) disturbances in young patients.16 On plain radiograph and CT, the pituitary fossa may be small or normal in size, with an ectopically placed pituitary gland. The pituitary gland itself may be hypoplastic or normal in appearance. Because CT and plain radiographs cannot differentiate the anterior and posterior lobes of the pituitary gland or demonstrate the nodule adjacent to the hypothalamus, there is really no role for these procedures in the evaluation of an ectopic posterior pituitary. With T1-weighted MRI, the normal round or ovoid bright spot in the posterior sella, representing the posterior pituitary gland, is not seen. The ectopic posterior pituitary tissue is usually located in the region of the median eminencethe insertion of the infundibulum to the hypothalamus. Nevertheless, the posterior pituitary gland maintains its characteristic signal properties, appearing as a small, high signal nodule protruding from the undersurface of the hypothalamus on T1-weighted images (Fig. 20-5). This type of abnormality of the hypothalamus has been described in other entities, including pituitary dwarfism and traumatic transection of the infundibulum, and as a normal variant in asymptomatic patients.16,17 and 18

FIGURE 20-5. Ectopic posterior pituitary gland. A, Sagittal unenhanced T1-weighted image through the midline. Note the high signal bright spot of the ectopic posterior pituitary (long straight white arrow) at the junction of the posterior margin of the optic chiasm and hypothalamus. The sella (curved white arrow) is essentially empty, and the anterior pituitary and infundibulum are not well seen. B, Coronal contrast-enhanced T1-weighted image through the optic chiasm, hypothalamus, and the ectopic posterior pituitary (long white arrow). The infundibulum is not well seen, and the other adjacent structures appear normal.

RATHKE CLEFT CYST Rathke cleft cysts are congenital variations derived from the embryologic precursors of the anterior lobe of the pituitary. The cysts have an epithelial lining and are rarely symptomatic. They may be seen incidentally and can mimic pituitary tumors. They usually occur in the midline of the anterior or superior sella. On plain radiographs of the skull, the sella is usually normal in appearance, although large Rathke cleft cysts can expand the sella. Rarely, calcifications may be seen. On CT, the sella may be normal or slightly expanded. The cyst can usually be seen as an area of lower attenuation that is similar in signal intensity to CSF. Smaller cysts may be difficult to differentiate from small cystic tumors, and larger cysts may be confused with craniopharyngiomas. These lesions infrequently affect the cavernous sinus or optic chiasm. On MRI, Rathke cleft cysts are usually seen as small cysts displacing the normal enhancing gland. The cysts are usually isointense with CSF on all pulse sequences. Occasionally, they may have a more unusual signal that is due to hemorrhaging. ENCEPHALOCELE Encephaloceles are defects in the skull and dura that may be associated with herniation of the brain or meninges. Encephaloceles can be congenital or acquired. When congenital, there is an anomaly in the midline structures, resulting in a bony defect in the anterior cranial fossa, the superior nose, and the sella; this is a rare anomaly that is more common in the Asian population. An encephalocele may be associated with facial deformities as well. On plain radiographs, the defect in the anterior cranial fossa and sella may be noted, but it often is small. On CT, the bony defect can usually be seen, and the brain can be observed to herniate into the sella. The findings on MRI are similar, but the detail of the brain is better seen, whereas the bony defect is less well seen. Brain herniation and anomalous vessels must be distinguished from other abnormalities, such as nasal polyps or masses, because inadvertent biopsy of an encephalocele that is mistaken for a nasal mass can be catastrophic. SUPRASELLAR ARACHNOID CYSTS Suprasellar arachnoid cysts are parenchymal in origin and are caused by sequestered remnants of the arachnoid membrane. They can cause endocrine dysfunction, most commonly precocious puberty. On CT, these cysts may be seen to distort the chiasm, the floor of the sella, or the clinoids. They may cause hydrocephalus by compressing the third ventricle superiorly (Fig. 20-6). On plain radiographs, in skeletally immature patients, the sutures of the skull may be widened if hydrocephalus is present. The sella may also be expanded, and there may be dysplasia of the clinoid processes with remodeling but no destruction. These bony findings are also seen on CT, but additionally, the cyst is visualized as a low-density structure with an attenuation similar to CSF. When present, hydrocephalus is noted by the presence of ventriculomegaly. It may be difficult to see the cyst as a separate structure when there is severe hydrocephalus (see Fig. 20-6A); thus, it is important to look for the membranous walls of the cyst, which may invert into the third ventricle.

FIGURE 20-6. Suprasellar arachnoid cyst. A, Axial non contrast-enhanced CT image of a young child with precocious puberty. The temporal horns are markedly dilated (thin black arrows). The suprasellar cistern is expanded (thick white arrows), but the cyst walls are not visible. B, Sagittal non contrast-enhanced T1-weighted magnetic resonance image through the midline. The sella is small, and the clinoids are somewhat deformed (short black arrows). There is a large cere-brospinal fluid space filling the suprasellar region (straight white arrows). The third ventricle is inverted, and the brainstem is displaced posteriorly. The cyst extends all of the way to the foramen of Monro (small black arrowhead) and columns of the fornix (large black arrowhead). C, Axial contrast-enhanced T1-weighted image through the region of

the foramen of Monro. The lateral ventricles are dilated (curved white arrows). The third ventricle is inverted and filled by the suprasellar cyst. The walls of the cyst are faintly seen as a thin septum (long white arrows).

On MRI, the findings are similar to CT, although the bony changes may be more difficult to see. MRI offers two advantages over CT in this setting. First, the margins of the cyst may be seen as separate from the ventricles; second, the additional planes of imaging may also help to delineate the cyst from the dilated ventricles (see Fig. 20-6B and C). Also, MRI is more sensitive than CT in the detection of transependymal migration of CSF as an indication of obstructive hydrocephalus. SEPTO-OPTIC DYSPLASIA Septo-optic dysplasia is a congenital malformation of the midline structures with absence of the septum pellucidum and a hypoplasia of the optic nerves, optic chiasm, and pituitary gland (Fig. 20-7). Septo-optic dysplasia can be considered the mildest form of holoprosencephaly. The patients have a variable degree of endocrine dysfunction, often including diabetes insipidus. Either vision or endocrine symptoms may predominate. Usually plain radiographs are normal, but the sella may be small. CT demonstrates the absence of the septum pellucidum. The optic chiasm is usually small, and the pituitary gland may be normal in size or hypoplastic. The findings on MRI are similar to those on CT, demonstrating an absence of the septum pellucidum and a small optic chiasm and optic nerves.

FIGURE 20-7. Septo-optic dysplasia. This adult presented with polydipsia, diabetes insipidus, and optic atrophy. A, Coronal contrast-enhanced T1-weighted image through the suprasellar region. The frontal horns are somewhat deformed, and the septum pellucidum is not visible in the midline (open white arrow). The pituitary gland (black arrow) is small. The infundibulum is also small (solid white arrow). The optic chiasm is not well seen. B, Axial proton-density image through the lateral ventricle. Again, the columns of the fornix and the septum pellucidum (open white arrow) are absent.

NEOPLASMS
CRANIOPHARYNGIOMA Craniopharyngiomas (Fig. 20-8, Fig. 20-9, Fig. 20-10), which are formed from ectodermal elements of Rathke pouch, are composed of squamous epithelial structures. They can occur anywhere from the floor of the third ventricle (hypothalamus) to the pharyngeal tonsil, with the majority being found in the suprasellar region. These tumors have a bimodal incidence, with peaks in the first and fifth decades of life; they comprise 9% of pediatric brain tumors.19 Clinicopathologically, two distinct subtypes are recognized: the adamantinous, which tend to occur in children, and the squamous-papillary variants, which tend to occur in adults.20 Craniopharyngiomas may present because of a mass effect on the chiasm, headaches, hydrocephalus, or pituitary and hypothalamic dysfunction.

FIGURE 20-8. Cystic craniopharyngioma. A, Contrast-enhanced sagittal T1-weighted image through the midline. A 2.5-cm inverted, pear-shaped cystic mass (straight white arrows) fills the sella and extends into the suprasellar cistern. The regions of the optic chiasm and pituitary stalk are not well demonstrated. There is no hydrocephalus. B, Axial T2-weighted image through the suprasellar cistern. The cystic craniopharyngioma demonstrates high signal intensity (black straight arrows) similar to that of cerebrospinal fluid. The suprasellar carotid arteries (solid white arrow) are seen as low signal intensity flow voids. They are displaced laterally by the mass.

FIGURE 20-9. Mixed cystic and solid craniopharyngioma. This patient had a family history of multiple endocrine neoplasia syndrome type 1, with hypopituitarism. He did not have a primary pituitary tumor, but rather a craniopharyngioma. A, Axial contrast-enhanced CT scan through the suprasellar cistern. There is an irregular mixed-density mass filling most of the suprasellar cistern (curved white arrows). There are small calcifications (thick black arrow) within the central posterior portion of the mass. There is no hydrocephalus. B, Coronal CT scan through the region of the sella. The floor of the sella is completely destroyed (small black arrowheads), and the mass extends to fill much of the sphenoid sinus. The clinoids are not well defined, and there is a mixed-density, partially calcified mass in the suprasellar region (small black arrows). C, Coronal contrast-enhanced T1-weighted magnetic resonance image through the sella in a location similar to that of B. The mass demonstrates intense contrast enhancement. There is envelopment of the carotid artery on the left (straight black arrow). The lesion extends into the suprasellar cistern and obliterates the region of the optic chiasm and pituitary stalk (solid white arrow). The lesion also grows into the temporal and frontal lobes above the clinoids on the left (black arrowheads). The cystic spaces within the tumor are lower in signal intensity, similar to that of cerebrospinal fluid.

FIGURE 20-10. Solid sellar and suprasellar craniopharyngioma. A, Sagittal contrast-enhanced T1-weighted magnetic resonance imaging (MRI) scan through the midline demonstrates a large mushroom-shaped mass extending out of the deformed sella into the anterior third ventricle (black arrows). The mass obscures the optic chiasm and other anterior third ventricular structures. The mass is associated with hydrocephalus and expansion of the lateral ventricles and bowing of the corpus callosum. B, Anterior view reconstruction of a contrast-enhanced 3D gradient echo acquisition shows the large enhancing sellar and suprasellar mass (white arrows). The MR angiogram (MRA) demonstrates elevation of the anterior cerebral arteries bilaterally (black arrows).

If the lesion is large enough, plain radiographs of the skull demonstrate remodeling of the sella turcica and clinoid processes. The amorphous or curvilinear calcifications, which are present in most pediatric tumors and half of adult tumors, are more readily detected by CT than on plain radiographs. On CT, craniopharyngiomas can be mixed cystic and solid and often exhibit enhancement of the more solid portions (see Fig. 20-9). Hemorrhage is not an uncommon finding, particularly within cystic portions of the tumors. Most commonly, these tumors are suprasellar in location. Craniopharyngiomas may grow to displace the optic chiasm superiorly, to displace the normal pituitary gland and stalk, to extend into the cavernous sinuses, and even to encase or occlude the carotid arteries. The imaging characteristics of craniopharyngiomas on MRI are variable, reflecting the wide range of components histologically composing these tumors. The tumors may be cystic (see Fig. 20-8), mixed cystic and solid (see Fig. 20-9), or primarily solid (see Fig. 20-10). High signal intensity on T1- and T2-weighted images is seen in cysts with high cholesterol content or with subacute hemorrhage. Craniopharyngiomas can also be of low signal intensity on T1-weighted images if the cyst contains a large amount of keratin.21 Fluid levels can be seen in cystic regions. Adamantinous craniopharyngiomas tend to be primarily cystic or mixed cystic-solid lesions that occur in children and adults, whereas squamous-papillary subtypes tend to be predominantly solid or mixed solid-cystic and occur in adults. Distinguishing between the two has a prognostic significance, because adamantinous tumors tend to recur. MRI can be helpful in distinguishing between the two: Encasement of vessels, a lobulated shape, and the presence of hyperintense cysts is suggestive of adamantinous tumors; and a round shape, presence of hypointense cysts, and a predominantly solid appearance is seen with squamous-papillary tumors.20 The overall sensitivity for detecting tumor is higher with MRI, and the potential for displaying anatomy in multiple planes provides excellent data for surgical planning (see Fig. 20-10). Some lesions that can be confused with craniopharyngiomas include arachnoid cysts, dermoid tumors, meningiomas, and aneurysms (if calcified). OPTIC CHIASM AND HYPOTHALAMIC GLIOMAS Gliomas involving the optic chiasm and hypothalamus present an imaging problem. It is often difficult or impossible to separate the origin of these tumors because of their intimate association.22 If there is extension of tumor along the optic tracts or optic nerves, it is much easier to determine the site of origin, because there is a characteristic growth pattern for optic pathway gliomas but not for hypothalamic gliomas. Histologically, these gliomas, which are more common in children, are slow-growing pilocytic astrocytomas. Optic gliomas are more common in patients with neurofibromatosis type 1 (Fig. 20-11). In adults, optic gliomas tend to be more aggressive and usually represent glioblastomas. Usually, plain radiographs are not helpful, except when the tumors extend along the optic nerves and cause expansion of the optic canals. CT and MRI demonstrate enlargement of the optic chiasm or hypothalamus, which is particularly well seen on coronal images (see Fig. 20-11A). These lesions usually enhance homogeneously with both CT and MRI contrast-enhanced images, and, in the case of optic nerve tumors, there may be extension of abnormal signal and enhancement along the optic tracts and radiations. The differential diagnosis includes craniopharyngiomas, sarcoid, metastases, or lymphomas.

FIGURE 20-11. Hypothalamic and optic pathway glioma. This young man presented with vision problems. He had a history of neurofibromatosis, type 1. A, Coronal contrast-enhanced T1-weighted magnetic resonance image. The optic chiasm and hypothalamus (solid white arrows) are thickened and globular in configuration. The anterior third ventricular recess is not well seen. The pituitary stalk and pituitary gland (open white arrow) are normal in configuration. B, Sagittal noncontrast-enhanced T1-weighted magnetic resonance image. An irregular mass fills the anterior third ventricle, optic chiasm, and hypothalamic area (long thin white arrows). A segment of the lesion may be cystic and appears as low signal regions. The pituitary gland and infundibulum are normal (open white arrow).

PITUITARY ADENOMAS Pituitary adenomas can be classified according to function and size. Lesions smaller than 1 cm are classified as microadenomas (Fig. 20-12), and those larger than 1 cm are classified as macroadenomas (Fig. 20-13, Fig. 20-14 and Fig. 20-15). The most common functioning adenomas are prolactinomas (see Chap. 13). Other functioning adenomas include adrenocorticotropic hormone-secreting tumors, thyroid-stimulating hormone-secreting tumors, and GH-secreting tumors (see Chap. 12, Chap. 14 and Chap. 15). Nonfunctioning adenomas account for ~40% of all pituitary adenomas. Adenocarcinomas of the pituitary gland are rare; in fact, metastasis to the gland is more common (see Chap. 11). Pituitary adenomas are usually seen in adults and are uncommon in children. When seen in childhood, they are usually seen in adolescent boys and are commonly macroadenomas, particularly prolactinomas that tend to be hemorrhagic.23

FIGURE 20-12. Pituitary microadenoma. This young woman presented with hyperprolactinemia. A, Sagittal T1-weighted noncontrast-enhanced magnetic resonance image. The anterior portion of the pituitary gland (solid white arrow) demonstrates slightly decreased signal. In general, the volume of the sella is not enlarged. The suprasellar structures are intact. B, Coronal contrast-enhanced T1-weighted image. The superior margin of the pituitary gland is slightly convex and asymmetrically

enlarged on the right. The low signal microadenoma (measuring 8 5 mm) is seen on the right (small black arrows). The lesion crosses the midline. This section is anterior to the infundibulum and also shows the anterior cerebral arteries (small black arrowheads), the optic chiasm (straight white arrow), the supraclinoid carotid artery (curved white arrow), and the cavernous carotid arteries (large black arrowhead). A few of the cranial nerves in the cavernous sinus are also visible (small white arrows).

FIGURE 20-13. Acromegaly and macroadenoma of the pituitary. This is a sagittal noncontrast-enhanced T1-weighted image of the midline skull. The scalp is thickened (solid white arrow). The skull is also markedly thickened (straight black arrows). The frontal sinuses are expanded (open white arrow). The pituitary gland is enlarged (small black arrowheads) and fills much of the clivus and sphenoid sinus. It does not extend into the suprasellar cistern or involve the brainstem.

FIGURE 20-14. Pituitary macroadenoma. A, Coronal contrast-enhanced CT scan through the sella. The pituitary gland is slightly enlarged. The superior surface is bowed cephalad into the suprasellar cistern (straight white arrows). The optic chiasm and pituitary stalk are not significantly deformed. The floor of the sella may be partially eroded because there is no cortical margin (open white arrow). Involvement of the cavernous sinus and carotid arteries is not clearly demonstrated. B, Coronal magnetic resonance image at a level similar to that of Figure 20-14A. The soft-tissue contrast is much better. Again, the gland is slightly enlarged and measures >1 cm, with the superior margin (solid white arrow) protruding into the suprasellar cistern without coming in contact with the optic chiasm (open white arrow). The carotid arteries (solid black arrows) are displaced slightly laterally in the cavernous sinus regions. There does not appear to be clear-cut involvement of the cavernous sinus. C, Sagittal T1-weighted noncontrast-enhanced magnetic resonance image. Again, the small macroadenoma is seen filling the sella and remodeling the sella (solid white arrows). The clivus is slightly remodeled, and the posterior clinoids are not well defined. There is an incidental infarct in the pons (open white arrow).

FIGURE 20-15. Invasive pituitary macroadenoma. This coronal contrast-enhanced 3D gradient echo magnetic resonance image through the sella demonstrates a large irregular aggressive skull-base mass (white arrow) with the pituitary gland enhancing diffusely. The pituitary stalk is displaced to the right. The low-signal-intensity left carotid artery (black arrow) is enveloped by the mass and is displaced inferiorly. The mass protrudes into the left suprasellar cistern. It is in contact with the left medial temporal lobe after breaking through the left cavernous sinus. The mass is extending through the left foramen ovale into the masticator space.

The imaging appearance of pituitary adenomas is nonspecific, and no inference to histology can be made from the sellar patterns. However, additional clues may be present, related to other secondary endocrine changes. For instance, with GH-secreting tumors, acromegaly occurs, and one may visualize thickening of the scalp or enlargement of the mandible on radiograph or physical examination (see Fig. 20-13); these tumors tend to be larger than 5 mm. Cushing adenomas usually are microadenomas, but compression vertebral fractures and a buffalo hump deformity may be clues. Prolactinomas are more variable in size; they usually are microadenomas, but may be macroadenomas. Nonfunctioning tumors tend to be large. Enlargement of the pituitary gland may result from many etiologies, not just neoplasia. End-organ failure is a cause of gland enlargement, such as is seen with primary hypothyroidism or surgical removal of the adrenals (Nelson syndrome; see later in this chapter and Chap. 75). If the functional status of a suspected adenoma or pituitary mass is in question, venous sampling of the petrosal sinuses can be performed by means of a catheter placed from the femoral vein into the internal jugular vein and then advanced into the greater petrosal veins.6 Analysis of blood samples can help determine the type of adenoma and the location of a lesion not detected by other imaging modalities. Sampling is also of value to demonstrate that the hormone originated from the gland rather than from an ectopic site. Although inferior petrosal sinus sampling is usually performed, it has been shown that bilateral, simultaneous cavernous sinus sampling, using corticotropin-releasing hormone, is as accurate as inferior petrosal sinus sampling in detecting Cushing disease and is perhaps more accurate in lateralizing the abnormality within the pituitary gland.24 The sella is usually normal in size with microadenomas and CT usually demonstrates no bony expansion, although there may be some asymmetry in the shape of the pituitary gland (see Fig. 20-12). CT, using thin-section coronal images and intravenous contrast, has been used successfully to detect microadenomas. The adenoma is identified as either a hypodense or hyperdense region in the gland after contrast enhancement. Cushing disease adenomas are more difficult to detect by CT, possibly because of their relative enhancement with respect to the normal gland.25 The recommended modality for examining a pituitary adenoma is MRI, with coronal and sagittal imaging. Detection is best with high-resolution techniques, such as three-dimensional imaging. The coronal plane is the most sensitive imaging plane, and T1-weighted spin-echo and three-dimensional imaging sequences are the best pulse sequences. The use of gadolinium enhancement is somewhat controversial,26,27 although the vast majority of radiologists believe that contrast is essential in the evaluation of the sella and parasellar regions. Usually, the tumors enhance less-than-normal tissue. Dynamic imaging can be of value in defining the abnormal segment of the gland.13 Of the pituitary macroadenomas, a higher percentage of these are nonfunctioning adenomas. Plain radiographs of the skull may demonstrate bony expansion or

erosion of the sella; at times, the masses can be huge, with wide destruction of the skull base (to the extent that the site of origin is not clear). Calcifications are rare. The sensitivity for detecting macroadenomas by CT is higher than for microadenomas; the CT examination should use thin-section coronal and axial imaging with intravenous contrast. Generally, the margins of the macroadenomas are more readily defined by MRI than by CT. Involvement of the optic chiasm, cavernous sinus, sphenoid sinus, orbit, temporal lobes, and carotid arteries can all be seen using MRI. In prolactinomas, MRI is used to evaluate the patient's response to bromocriptine therapy. A decrease in tumor size can be seen as early as 1 week after the start of therapy. Additionally, MRI can detect posttherapy hemorrhage into macroadenomas and mass effect or inferior herniation of the chiasm as a result of a decrease in the tumor size.28 In macroadenomas, subacute hemorrhage is readily detected by MRI because the breakdown products of hemoglobin have paramagnetic or diamagnetic effects, depending on their chemical composition. Moreover, MRI is good for evaluating invasion into the adjacent cavernous sinus and for documenting the patency of the carotid arteries (see Fig. 20-15). INFUNDIBULAR MASSES The thickness of the normal pituitary stalk averages 3.5 mm at the median eminence and 2.8 mm near its midpoint. The normal stalk enhances markedly on CT with contrast and on MRI with gadolinium. The most common clinical problem associated with disease of the pituitary stalk is diabetes insipidus. When this is present, there usually is absence of the normal hyperintensity of the posterior pituitary. On T1-weighted MRI, diabetes insipidus may be found to occur as a result of transection of the pituitary stalk. The differential diagnosis of a thickened stalk includes sarcoidosis, tuberculosis, histiocytosis X, and ectopic posterior pituitary as well as germinoma. A thickened stalk can also be due to an extension of a glioma within the hypothalamus. In patients with neurosarcoidosis and tuberculous infiltration of the stalk, the chest radiograph is generally abnormal and may be helpful in the differentiation from histiocytosis X. Clinically, patients with histiocytosis X may have skin lesions, otitis media, or bone lesions in addition to interstitial lung disease.12 HYPOTHALAMIC HAMARTOMAS A hamartoma of the tuber cinereum usually presents as precocious puberty in a young child.29 It is important to differentiate this lesion from a hypothalamic glioma because the prognosis for hamartoma is much more favorable. Imaging is best with MRI thin-section coronal and sagittal planes (Fig. 20-16). The findings are usually characteristic: The mass arises from the undersurface of the hypothalamus and is exophytic. The nodular mass (<1 cm) hangs into the suprasellar cistern adjacent to the mammillary bodies. On T1-weighted images, the signal is isointense with normal brain; on T2-weighted images, there is mild hyperintensity or isointensity. These lesions usually do not enhance with contrast administration.

FIGURE 20-16. Hypothalamic hamartoma. This young girl presented with precocious puberty. A, Noncontrast-enhanced sagittal T1-weighted image through the midline. A small nodule measuring ~5 mm in dimension (solid white arrow) is seen protruding from the under-surface of the hypothalamus into the suprasellar cistern just anterior to the mammillary bodies (small black arrowhead). The pituitary gland and infundibulum are normal (open white arrow). There does not appear to be any deformity of the anterior third ventricle or invasion of the brain. B, Axial proton-density magnetic resonance image. The small hamartoma (straight white arrow) is just anterior to the bifurcation of the basilar artery in the suprasellar cistern. The infundibulum is seen directly anterior to this (small black arrowhead). Other labeled structures include the supraclinoid carotid arteries (thin black arrows) and the posterior cerebral arteries (curved white arrows).

SELLAR AND PARASELLAR MENINGIOMA Meningiomas usually occur in the parasellar region rather than within the true sella. They are derived from meningeal cells and are intimately associated with the dura. They can arise from the planum sphenoidale, the diaphragma sella, the optic nerve sheaths, the clinoids, or within the cavernous sinuses. Meningiomas may demonstrate calcification on plain radiographs. On CT, they are usually hyperdense and may be mistaken for hemorrhagic lesions or aneurysms. Hyperostosis of the adjacent bones, including the clinoids, is characteristic, and there may be expansion of the sphenoid sinus by air. On MRI, meningiomas tend to be isointense with gray matter and may be difficult to see on T1-weighted images without gadolinium. With both iodinated CT contrast agents and gadolinium MRI agents, meningiomas usually demonstrate marked, homogeneous enhancement. SCHWANNOMAS AND NEUROFIBROMAS Schwannomas, which are tumors derived from the myelin sheath of peripheral nerves, can be found involving the cranial nerves within the cavernous sinus and parasellar regions. In general, pituitary function is not affected; however, often the cranial nerves III, IV, V, and VI are affected within the cavernous sinuses or in the suprasellar and prepontine cisterns. Schwannomas may remodel the foramina of the skull where the individual nerves exit. When multiple lesions are seen, neurofibromatosis should be considered. On CT, schwannomas usually are hyperdense lesions with homogeneous enhancement, and they may be hard to differentiate from meningiomas. On MRI, they may be isointense or hyperintense to gray matter on T1-weighted images, and they enhance homogeneously. METASTASIS Metastasis to the sellar, suprasellar, or parasellar regions may arise in the sphenoid bone or sinus, cavernous sinus, pituitary gland, hypothalamus, or surrounding soft tissues (Fig. 20-17 and Fig. 20-18). Endocrine symptoms are uncommon with pituitary metastasis but are often seen when the hypothalamus is involved. It may be difficult to distinguish a metastasis from a primary pituitary abnormality on the basis of imaging alone; however, the presence of bony destruction or the history of a known primary tumor may be helpful.

FIGURE 20-17. Metastatic suprasellar and pituitary ependymoma. A, Coronal contrast-enhanced magnetic resonance imaging scan through the sella. The pituitary gland, stalk, and hypothalamus are all infiltrated by an aggressive irregular mass (white arrows). This is a secondary CSF seeding metastasis from a primary ependymoma of the lower thoracic cord. This type of pattern can be seen in sarcoid, histiocytosis X, eosinophilic granuloma, lymphoma, leukemia, and carcinoma. It is

not uncommon that the sellar metastasis presents before the spinal or other primary tumor site. B, Sagittal T1-weighted midline thoracolumbar spin-echo image. The primary conus ependymoma (white arrows) is seen as an isointense expansile mass of the conus. Note that there is also thickening of some of the lower lumbar roots consistent with other drop metastases.

FIGURE 20-18. Hypothalamic and pituitary metastases. This patient presented with squamous cell carcinoma of the esophagus and diffuse metastatic bone disease. The patient became comatose. A, Sagittal non contrast-enhanced T1-weighted magnetic resonance image. There is a 2-cm mass (solid white arrows) seen filling the anterior third ventricle and extending toward the pituitary. There is also deformity of the superior portion of the pituitary gland (curved white arrow). The low signal changes in the marrow of the clivus (black arrowhead) are secondary to metastatic disease in this area as well. B, T2-weighted axial image through the suprasellar region. There is an irregular mass (open curved white arrows) deforming the suprasellar cistern and displacing the adjacent vessels.

GERM-CELL TUMORS Suprasellar germ-cell tumors include seminomas (germinomas), teratomas, embryonal tumors, choriocarcinomas, and tumors of mixed histology. Some of these tumors are associated with simultaneous lesions in the pineal gland. These tumors are much more common in boys and tend to involve the pituitary stalk. When germ-cell tumors involve only the pituitary stalk, they can be difficult to differentiate from other primary processes involving the infundibulum (Fig. 20-19). If there is an associated pineal mass, it is virtually diagnostic of a germ-cell tumor.

FIGURE 20-19. Invasive pituitary germinoma. A, This coronal contrast-enhanced magnetic resonance imaging (MRI) spin-echo scan through the sella shows an expansile pituitary mass (white arrows) filling the sella. It protrudes into the suprasellar cistern (black arrow) and is displacing the adjacent brain structures cephalad. It does not appear to be invading the cavernous sinus. There is some odd contrast enhancement in the anterior third ventricle separate from the mass. B, Another coronal contrast-enhanced MRI spin-echo scan slightly farther anteriorly demonstrates that the germinoma is quite aggressive and is directly invading the left basal ganglion (black arrows).

MISCELLANEOUS ENTITIES
PITUITARY APOPLEXY Pituitary apoplexy is the result of necrosis of the anterior lobe of the pituitary. When seen in a postpartum woman, it is referred to as Sheehan syndrome. It is usually the result of hemorrhage into the gland, but may also be nonhemorrhagic in nature.30 Pituitary apoplexy is usually associated with hemorrhage into a preexisting adenoma, although it can occur in a normal gland (Fig. 20-20).30 Although CT in general is sensitive in making the diagnosis of acute hemorrhage, the changes in the pituitary gland may be difficult to appreciate. The imaging modality of choice for evaluation of hemorrhage in the pituitary is MRI. Also, this allows for evaluation of the optic chiasm, which may be compressed.

FIGURE 20-20. Hemorrhagic pituitary apoplexy. A, Sagittal T1-weighted spin-echo midline magnetic resonance imaging (MRI) demonstrates a high-signal-intensity abnormality of the sella (white arrow). It demonstrates bulging of the superior contour of the pituitary gland without major deformity of the optic chiasm. The high signal intensity without contrast suggests subacute hemorrhage. B, The pituitary apoplexy hemorrhage (white arrow) demonstrates low signal intensity on the T2-weighted axial image. This suggests that the hemorrhage is in the acute phase. Note the high-signal-intensity cerebro spinal fluid (CSF) spaces.

SPHENOID SINUS DISEASE The sphenoid sinus can be involved with expansile (mucocele), inflammatory (Wegener granulomatosis), infectious (pyocele), or neoplastic (squamous cell carcinoma or lymphoma) processes. The inflammatory disease may cause irritation or may directly extend to involve the adjacent sella turcica. This is most severe with necrotizing infections (such as mucormycosis) in immunocompromised and diabetic patients. In this setting, it may be difficult to differentiate inflammatory from neoplastic disease, such as squamous cell or adenocarcinoma of the sphenoid sinus and chordoma of the clivus. Endocrine symptoms are rare.

ANEURYSM Aneurysms may involve any of the intracranial vessels and are usually found at branch points. Aneurysms of the cavernous supraclinoid carotid and anterior cerebral arteries may be located in the sella or suprasellar region (Fig. 20-21). Compressive symptoms of aneurysms depend on their location. Aneurysms that extend into the suprasellar region may compress the optic chiasm; thus, patients may present with bitemporal hemi-anopia. Aneurysms may also compress the pituitary or infundibulum and present with diabetes insipidus or other endocrine abnormalities. If there has been a subarachnoid hemorrhage, the patient may present with severe headache or with a neurologic deficit related to vasospasm.

FIGURE 20-21. Supraclinoid internal carotid artery aneurysm. This patient presented with Cushing syndrome and no history of subarachnoid hemorrhage or headache. A, Sagittal noncontrast-enhanced T1-weighted image. There is a circular signal void measuring ~2 cm (short white arrows) in the suprasellar region. This is interposed between the optic chiasm (long white arrows) and the septal region of the frontal lobes. The pituitary gland (large black arrowheads) is essentially normal. B, Contrast-enhanced coronal T1-weighted magnetic resonance angiogram (MRA). The supraclinoid carotid aneurysm (solid black arrows) is seen to have a high signal rim at the periphery. This is related to the abnormal blood flow in aneurysms at the periphery rather than at the center of the lesion. The anterior cerebral arteries (small black arrowheads) and the pituitary stalk (open curved white arrow) are displaced to the left. The pituitary gland is normal.

On plain radiograph, curvilinear or ring-shaped calcifications may be seen in the wall of the aneurysm. Large aneurysms can remodel the skull base and erode the clinoids. On CT, these lesions tend to be of high density on unenhanced examinations, with marked enhancement after intravenous contrast. On spin-echo MRI images, aneurysms possess the characteristics of very low signal intensity or signal void and may demonstrate phase encoding or flow artifacts. MRA may be helpful in these cases, because it provides noninvasive screening for differentiating an aneurysm from some other causes of lesions of low signal intensity seen on the MRI (such as calcification). Arteriovenous malformations may also occur; they reveal characteristic large feeding arteries and several serpentine flow voids, representing the draining veins. Contrast angiography is still the gold standard for the evaluation of small aneurysms and arteriovenous malformations, but MRI can provide a good noninvasive screening examination in most cases. (See also ref. 30a and ref. 30b.) SARCOIDOSIS Neurosarcoidosis has a propensity to involve the basilar cisterns and the suprasellar region, specifically the infundibulum. This is best evaluated with MRI, which can show thickening and enhancement of the meninges in the basal cisterns and surrounding the suprasellar cisterns as well as thickening of the normal pituitary stalk. In 3% to 16% of patients with sarcoidosis, the nervous system is involved, and sarcoidosis can mimic almost any other lesion.31,32 PRIMARY HYPOTHYROIDISM AND NELSON SYNDROME Both primary hypothyroidism and Nelson syndrome33 represent hypertrophic changes to the pituitary gland caused by end-organ failure. In the case of primary hypothyroidism, thyrotropin-releasing hormone is hypersecreted because of the lack of feedback inhibition by thyroxine on the hypothalamus. The pituitary gland may be enlarged. In Nelson syndrome, adrenocorticotropic hormone is hypersecreted by a pituitary tumor, commonly after surgical removal of the adrenal glands for the treatment of Cushing syndrome. This causes hyperpigmentation of the skin and tumorous enlargement of the pituitary gland that can appear to behave aggressively and even metastasize outside of the cranial cavity (see Chap. 11, Chap. 14 and Chap. 75). POSTOPERATIVE CHANGES Many pituitary resections are performed by means of a trans-sphenoidal approach. In the postoperative patient, the floor of the sella is generally found to be distorted. The sphenoid sinus may be filled with implanted gelatin foam or biologic materials, such as muscle or fat. The gland itself may have an unusual shape (including club-shaped), or it may be in a lateral position, and the infundibulum may be deviated with respect to the center of the fossa (Fig. 20-22). If a frontal approach has been used, skull defects may be noted, and there often are focal areas of increased water content in the frontal lobes, suggesting local injury to the brain.

FIGURE 20-22. Postoperative changes. This patient had resection of a large pituitary adenoma. This coronal contrast-enhanced T1-weighted magnetic resonance image demonstrates displacement of the pituitary stalk to the left (curved solid white arrow). The floor of the sella is asymmetric and slopes inferiorly to the right. There is a cyst seen filling the residual sella (open white curved arrow). There is no mass displacing the stalk. This displacement is more likely related to postsurgical changes.

The pituitary gland may be found to have regenerated to a normal shape and size. Large suprasellar masses can deform the optic chiasm, but the chiasm may return to its normal size and location after the mass is gone. The chiasm may have herniated into the sella after being stretched (Fig. 20-23). There may be residual or recurrent tumor, particularly if the preoperative evaluation reveals extension into or invasion of the cavernous sinus. The sella may be partially empty in the postoperative patient. Because of these dynamic postoperative changes, the recommended time for follow-up examination is ~4 to 6 months after surgery to allow for resolution of hematomas, fluid collections, and resorption of implant packing materials.34,35

FIGURE 20-23. Postoperative sellar changes. A, Contrast-enhanced coronal T1-weighted magnetic resonance image. This patient had resection of a pituitary tumor. The anterior third ventricle and optic chiasm structures are herniated (long white arrows) into the region of the sellar surgical defect. The anterior cerebral arteries (solid black arrows) are displaced somewhat inferiorly as well. There is no residual mass. The floor of the sella is seen sloping far to the right. B, Sagittal noncontrast-enhanced T1-weighted magnetic resonance image. The hypothalamus and optic chiasm (long white arrows) are seen drooping in a J-like configuration into the surgical defect of the slightly expanded sella.

CONCLUSION
Imaging of the sella and pituitary fossa can be an important part of evaluating patients with endocrine or metabolic dysfunction. A review of the findings in both normal and diseased states has been presented. An emphasis has been placed on MRI, because in most cases, this has become the modality of choice for pituitary imaging. CHAPTER REFERENCES
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. Underwood LE, Radcliffe WB, Guinto FC Jr. New standards for the assessment of sella turcica volume in children. Radiology 1976; 126:651. Reich E, Zelch JV, Alfidi RJ, et al. Computed tomography in the detection of juxtasellar lesions. Radiology 1976; 118:333. Elster AD. Modern imaging of the pituitary. Radiology 1993; 187:3. Kishore PRS, Kaufman AB, Melichar FA. Intrasellar carotid anastomosis simulating pituitary microadenoma. Radiology 1979; 132:381. Raymond J, Hardy J, Czepko R, Roy D. Arterial injuries in transsphenoidal surgery for pituitary adenoma: the role of angiography and endovascular treatment. AJNR Am J Neuroradiol 1997; 18:655. Miller DL, Doppman JL, Nieman LK, et al. Petrosal sinus sampling: discordant lateralization of ACTH-secreting pituitary microadenomas before and after stimulation with corticotropin-releasing hormone. Radiology 1990; 176:429. Lusted LB, Keats TE. Atlas of roentgenographic measurement, 4th ed. Chicago: Year Book Medical Publishers, 1978:59. Elster AD, Chen M, Williams DW III, Key LL. Pituitary gland: MR imaging of physiologic hypertrophy in adolescence. Radiology 1990; 174:682. Dietrich RB, Lis LE, Greensite FS, Pitt D. Normal appearance of the pituitary gland in the first 2 years of life. AJNR Am J Neuroradiol 1995; 16:1413. Doraiswamy PM, Potts JM, Axelson DA, et al. MR assessment of pituitary gland morphology in healthy volunteers: age- and gender-related differences. AJNR Am J Neuroradiol 1992; 13:1297. Mark LP, Haughton VM. The posterior sella bright spot: a perspective. AJNR Am J Neuroradiol 1990; 11:701. Tien RD, Newton TH, McDermott MW, et al. Thickened pituitary stalk on MR images in patients with diabetes insipidus and Langerhans cell histiocytosis. AJNR Am J Neuroradiol 1990; 11:707. Miki Y, Matsuo M, Nishizawa S, et al. Pituitary adenomas and normal pituitary tissue: enhancement patterns on Gadopentetate-enhanced MR imaging. Radiology 1990; 177:36. Escott EJ, Rao VM, Ko WD, Guitierrez JE. Comparison of dynamic contrast-enhanced gradient-echo and spin-echo sequences in MR of head and neck neoplasms. AJNR Am J Neuroradiol 1997; 18:1411. Chakeres DW, Curtin A, Ford G. Magnetic resonance imaging of pituitary and parasellar abnormalities. Radiol Clin North Am 1989; 27:267. Kelly WM, Kucharczyk W, Kucharczyk J, et al. Posterior pituitary ectopia: an MR feature of pituitary dwarfism. AJNR Am J Neuroradiol 1988; 9:454. Kurioiwa T, Okabe Y, Hasuo K, et al. MR imaging of pituitary dwarfism. AJNR Am J Neuroradiol 1991; 12:161. Benshoff ER, Katz BH. Ectopia of the posterior pituitary gland as a normal variant: assessment with MR imaging. AJNR Am J Neuroradiol 1990; 11:711. Young SC, Zimmerman REA, Nowell MA, et al. Giant cystic craniopharyngiomas. Neuroradiology 1987; 29:468. Sartoretti-Schefer S, Wichman W, Aguzzi A, Valavanis A. MR differentiation of adamantinous and squamous-papillary craniopharyngiomas. AJNR Am J Neuroradiol 1997; 18:77. Pusey E, Kortman KE, Flannigan BD, et al. MR of craniopharyngiomas: tumor delineation and characterization. AJNR Am J Neuroradiol 1987; 8:443. Albert A, Lee BCP, Saint-Louis L, et al. Magnetic resonance imaging of optic chiasm and optic pathway. AJNR Am J Neuroradiol 1986; 7:255. Poussaint TY, Barnes PD, Anthony DC, et al. Hemorrhagic pituitary adenomas in adolescence. AJNR Am J Neuroradiol 1996; 17:1907. Oliverio PJ, Monsein LH, Wand GS, Debrun GM. Bilateral simultaneous cavernous sinus sampling using corticotropin-releasing hormone in the evaluation of Cushing disease. AJNR Am J Neuroradiol 1996; 17:1669. Peck WW, Dillon WP, Norman D, et al. High-resolution MR imaging of pituitary microadenomas at 1.5 T: experience with Cushing disease. AJR Am J Roentgenol 1989; 9:149. Stadnik T, Stevenaert A, Beckers A, et al. Pituitary microadenomas: diagnosis with two- and three-dimensional MR imaging at 1.5 T before and after injection of gadolinium. Radiology 1990; 176:422. Chong BW, Kucharczk W, Singer W, et al. Pituitary gland MRI: a comparative study of healthy volunteers and patients with microadenomas. AJNR Am J Neuroradiol 1994; 15:675. Lundin P, Bergstrom K, Nyman R, et al. Macroprolactinomas: serial MR imaging in long-term bromocriptine therapy. AJNR Am J Neuroradiol 1992; 13:1287. Hahn FJ, Leinbrock LG, Huseman CA, Makos MM. The MR appearance of hypothalamic hamartoma. Neuroradiology 1988; 30:67. Lavallee G, Marcos R, Palardy J. MR of nonhemorrhagic postpartum pituitary apoplexy. AJNR Am J Neuroradiol 1995; 16:1939.

30a. Jager HR, Grieve JP. Advances in non-invasive imaging of intracranial vascular disease. Ann R Coll Surg Engl 2000; 82:1. 30b. Wardlaw JM, White PM. The detection and management of unruptured intracranial aneurysms. Brain 2000; 123:265. 31. Hayes WS, Sherman JL, Stern BJ, et al. Magnetic resonance and CT evaluation of intracranial sarcoidosis. AJR Am J Roentgenol 1987; 8:1043. 32. Lexa FJ, Grossman RI. MR of sarcoidosis in the head and spine: spectrum of manifestations and radiographic response to steroid therapy. AJNR Am J Neuroradiol 1994; 15:973. 33. Nagesser SK, van Seters AP, Kievit J, et al. Long-term results of total adrenalectomy for Cushing's disease. World J Surg 2000; 24:108. 34. Steiner E, Knosp E, Herold CJ, et al. Pituitary adenomas: findings of postoperative MR imaging. Radiology 1992; 185:522. 35. Kaufman B, Tomsak RL, Kaufman BA, et al. Herniation of the suprasellar visual system and third ventricle into empty sellae: morphologic and clinical considerations. AJNR Am J Neuroradiol 1989; 10:65.

CHAPTER 21 MEDICAL TREATMENT OF PITUITARY TUMORS AND HYPERSECRETORY STATES Principles and Practice of Endocrinology and Metabolism

CHAPTER 21 MEDICAL TREATMENT OF PITUITARY TUMORS AND HYPERSECRETORY STATES

DAVID H. SARNE Prolactin Hypersecretion Microprolactinomas Macroprolactinomas Idiopathic Hyperprolactinemia Prolactin Hypersecretion Associated with Other Lesions Details of Bromocriptine Therapy Therapy with Other Dopaminergic Agonists Adrenocorticotropic Hormone Hypersecretion Centrally Acting Drugs Peripherally Acting Drugs Growth Hormone Hypersecretion Bromocriptine Somatostatin Analogs Estrogen Therapy Gonadotropin Hypersecretion Gonadotropin-Secreting Adenomas True Precocious Puberty Hypersecretion of Thyroid-Stimulating Hormone Thyroid-Stimulating HormoneSecreting Adenoma Isolated Pituitary Resistance to Thyroid Hormone Nonsecretory Adenomas Chapter References

PROLACTIN HYPERSECRETION
MICROPROLACTINOMAS In women with microprolactinomas (tumor diameter of <10 mm) who require therapy for menstrual irregularities, infertility, or galactorrhea, bromocriptine is the treatment of choice.1,2 and 3 A dose of 2.5 to 10 mg per day relieves the symptoms and normalizes serum prolactin in nearly 90% of patients. Approximately 10% of these patients fail to respond to bromocriptine regardless of the dosage, but many treatment failures are related to an inability to tolerate a sufficient dosage. With treatment, prolactin levels may normalize within a few days to several months. A reduction in microadenoma size has been reported in up to 75% of treated patients. When bromocriptine is discontinued after 1 to 2 years, 20% of patients maintain normal prolactin levels.4 In most patients, hyperprolactinemia recurs, although 25% have prolactin levels less than half the pretreatment value. No test predicts which patients will have sustained normalization. Prolactin levels usually reach a stable plateau within 2 months, but serum prolactin can reach the pretreatment level in <2 weeks. Some authors recommend that bromocriptine therapy be discontinued in patients with prolactin-secreting microadenomas after 1 to 2 years, with reinstitution of therapy only in those patients who again develop elevated levels. The surgical treatment of microadenomas yields a high initial cure rate (85%), but the procedure is associated with complications. Moreover, with longer follow-up periods, half of these cured patients have a recurrence of hyperprolactinemia5 (see Chap. 13 and Chap. 23. Surgical treatment should be reserved for patients whose microadenomas fail to respond to bromocriptine or other dopaminergic agents or for those who are unable to tolerate the side effects of the drug regimen. Because studies of untreated patients have revealed that many microadenomas remain small for years, withholding treatment in patients with asymptomatic hyperprolactinemia may be reasonable.6,7 Nevertheless, although patients may not be disturbed by amenorrhea or galactorrhea, prolonged hypogonadism increases the loss of skeletal mass, so bromocriptine therapy is recommended for these patients, although for women not interested in fertility, estrogen therapy is also an option.3,8 MACROPROLACTINOMAS The treatment of macroprolactinomas (tumor diameter of 10 mm) remains controversial. If the goal is to achieve lasting normalization of prolactin after discontinuing therapy, then surgery or radiation therapy is most effective; however, bromocriptine controls prolactin levels in more patients than does either surgery or radiation.9 A treatment regimen of 5 to 15 mg per day of bromocriptine causes serum prolactin levels to fall to <10% of the pretreatment value in 90% of patients and normalizes prolactin levels in ~66% of patients.10 This decline in serum prolactin is usually accompanied by cessation of galactorrhea, return of normal menses in women, and improvement of libido in men. (The resolution of impotence in men occurs less frequently.) The maximal reduction of prolactin usually is seen after 2 to 3 months of therapy; however, some patients require up to 9 months of therapy for maximal suppression of prolactin. The gonadotropin abnormalities frequently found in these patients commonly improve with bromocriptine therapy.11 Abnormal secretion of growth hormone, adrenocorticotropic hormone (ACTH), or thyroid-stimulating hormone (TSH) occurs less frequently than gonadotropin abnormalities but is less likely to resolve. A reduction in tumor size is seen in 90% of treated patients; most exhibit a size reduction equal to or exceeding 50%.10,12 The extent of the reduction cannot be predicted by either basal or treatment prolactin levels. Prolactin levels usually decline before a reduction in tumor size is noted. In some patients, despite an almost total disappearance of the lesion, serum prolactin levels remain mildly elevated. A significant reduction in tumor size may occur within 6 weeks, but such a reduction may not be evident for at least 6 months, and the process may continue for up to 1 year. In the absence of progressive tumor growth, bromocriptine should be administered for at least 6 months to 1 year before one concludes that it is ineffective. Usually, prolactin levels remain elevated if treatment with bromocriptine does not reduce the tumor's size. Rarely, the adenoma increases in size despite the suppression of serum prolactin levels.13 When therapy is discontinued, hyperprolactinemia and an increase in tumor size usually follow.14 Reexpansion of the adenoma can be rapid and dramatic. Therefore, treatment withdrawal should be undertaken cautiously and cannot be recommended routinely. If discontinuation of therapy is not accompanied by an increase in serum prolactin levels or by tumor regrowth, necrosis of the adenoma has probably occurred. The role of bromocriptine in preparing patients for primary surgical treatment remains controversial.15 Bromocriptine therapy should not be discontinued before surgery because rapid tumor reexpansion may result. In patients who are to undergo irradiation, treatment with bromocriptine for 1 to 2 years until the therapy is fully effective is associated with a more rapid improvement in hyperprolactinemia, amelioration of symptoms, and reduction in tumor growth. Bromocriptine is the treatment of choice in patients in whom both surgery and radiation therapy have failed. IDIOPATHIC HYPERPROLACTINEMIA When treatment of idiopathic hyperprolactinemia is indicated, bromocriptine is the preferred mode of therapy.16 A treatment regimen of 2.5 to 7.5 mg per day of bromocriptine restores normal prolactin levels, corrects gonadotropin dysfunction, and relieves symptoms in >90% of patients. Surgery and radiation therapy are not indicated for this condition. PROLACTIN HYPERSECRETION ASSOCIATED WITH OTHER LESIONS Nonprolactin-secreting adenomas and other central nervous system lesions may increase prolactin levels by interfering with normal hypothalamic inhibition. Bromocriptine, 2.5 to 7.5 mg per day, usually normalizes serum prolactin levels in these patients. The normalization of prolactin is usually associated with the resolution of galactorrhea. Amenorrhea and infertility may also resolve with this mode of therapy; however, when the underlying lesion has disrupted the normal hypothalamic pituitary axis or has destroyed the gonadotropes, use of bromocriptine does not restore menses or fertility. DETAILS OF BROMOCRIPTINE THERAPY

Bromocriptine (2-Br-a-ergocryptine mesylate) is a semisynthetic ergot alkaloid. It specifically binds to and stimulates dopamine receptors. One-third of an oral dose is absorbed, and peak serum levels are reached 1 to 3 hours after oral administration. It is extensively metabolized by the liver, with the metabolites being excreted almost entirely by biliary secretion.17 Less than 5% of the drug is excreted in the urine. Maximal suppression of prolactin occurs 6 to 8 hours after a single dose, and suppression may be maintained for 12 to 14 hours. Treatment with bromocriptine should be initiated with a dose of one-half of a 2.5-mg tablet taken with food just before bedtime, followed by a regimen of 1.25 mg given with food every 8 to 12 hours. Less than 1% of treated patients experience a first-dose phenomenon, characterized by marked faintness or dizziness. This is observed most commonly in elderly patients and in those with a previous history of fainting, peripheral vascular disease, or use of vasodilators. Increases in dosage should be gradual, no more than 2.5 to 5 mg within a period of a few days to 1 week. The total daily dose is usually divided and administered every 8 to 12 hours. Side effects are usually dose related, with a rapid development of tolerance. Many side effects are potentiated by alcohol, the use of which should be avoided in sensitive patients. To tolerate bromocriptine therapy, some patients may need to begin with a dosage of 0.625 mg per day (one-fourth tablet), thereafter increasing the dosage at 1-week intervals. Nausea is the most common side effect and occurs in up to 25% of treated patients. The nausea is usually mild, may be minimized by administration of the drug with food and by the initial use of low doses, and generally improves with time.18 Constipation is also frequently reported, and some patients experience abdominal cramps. Seven patients receiving high doses of bromocriptine for the treatment of acromegaly were reported to have had major gastrointestinal hemorrhage associated with peptic ulcer disease (three of these episodes were fatal).18 However, bromocriptine has not been associated with an increased incidence of peptic ulcer disease. A slight decline in blood pressure is commonly observed in treated patients; however, patients usually remain asymptomatic. Mild orthostatic hypotension has also been noted.18 The decrease in blood pressure is probably related to both a relaxation of vascular smooth muscle and central inhibition of sympathetic tone. As with the gastrointestinal side effects, symptomatic hypotension usually improves with time. Vascular side effects, including digital vasospasm, livedo reticularis, and erythromelalgia, occur infrequently and are usually associated with bromocriptine doses that exceed those used in the treatment of hyperprolactinemia. Significant mental changes, including hallucinations, have been noted, most commonly in elderly patients receiving large doses of bromocriptine. In two patients, a dose of 5 to 7.5 mg of bromocriptine, administered for treatment of hyperprolactinemia, was reported to have caused psychotic delusions. However, one of these patients had a known history of schizophrenia in remission, and the other was under severe emotional stress. Other side effects of bromocriptine include nasal stuffiness, headache, and fatigue. Women taking bromocriptine should be advised to use mechanical contraception and, if pregnancy is desired or suspected, to discontinue bromocriptine whenever expected menses are >2 days late. Visual fields should be evaluated regularly during pregnancy. If evidence of tumor enlargement is found, a choice is made between continued observation, treatment with bromocriptine, or transsphenoidal surgery, depending on the status of the individual patient. In the United States, women are usually advised to discontinue bromocriptine therapy during pregnancy; in Europe, however, treatment is commonly continued. A review of 1410 pregnancies in 1335 women who received bromocriptine while pregnant revealed that the incidence of spontaneous abortions (11.1%) and congenital anomalies (3.5%) was no higher than that seen in the general population.19 In women not taking other fertility agents, a slightly increased incidence of twin pregnancies (1.8%) was seen. A retrospective study of 64 children born to 53 mothers who took bromocriptine while pregnant revealed no evidence of adverse effects on motor or psychological development.20 THERAPY WITH OTHER DOPAMINERGIC AGONISTS Several other dopamine agonists have been developed that may be useful in the treatment of hyperprolactinemia. A parenteral formulation of long-acting bromocriptine has been effective, with intramuscular injections given every 4 weeks. Pergolide is an ergoline derivative that can be given once daily in a dose of 50 to 100 g.21 Although it is similar to bromocriptine in its effectiveness and side effects, some patients who do not tolerate bromocriptine may tolerate pergolide.22 The nonergot dopamine agonist quinagolide (CV 205-502) can be administered in dosages of 0.1 to 0.5 mg per day, with fewer side effects than bromocriptine or pergolide. Quinagolide was effective in patients who were unable to tolerate bromocriptine and in some patients who failed to respond adequately to bromocriptine.23 Cabergoline is a long-acting ergoline derivative that can be effective when given weekly or biweekly in doses of 0.5 to 2.0 mg. Its efficacy and side effects profile are similar to or better than those of bromocriptine.24 In several studies, tumor shrinkage and normalization of prolactin levels have occurred in patients who could not tolerate bromocriptine or failed to respond adequately.25,26 and 27

ADRENOCORTICOTROPIC HORMONE HYPERSECRETION

When Cushing syndrome is caused by a pituitary tumor (Cushing disease), transsphenoidal surgery is the treatment of choice.3,28 Radiation therapy, by comparison, is less often successful and may take 1 to 2 years to be effective3 (see Chap. 22). Drug treatment is generally not used as a primary mode of therapy except in patients who refuse surgery or irradiation. However, drug treatment may be appropriate in severely ill patients with marked hypokalemia, psychiatric disturbances, infection, or poor wound healing or in patients awaiting transsphenoidal surgery. Medical therapy is also useful in reducing cortisol levels and ameliorating symptoms until pituitary irradiation is fully effective. Finally, drug therapy may be useful in patients in whom surgery and radiation therapy have failed. Patients with Cushing disease who are treated by adrenalectomy may develop large, ACTH-secreting, pituitary macroadenomas (Nelson syndrome). The response of such lesions to both surgery and irradiation has been disappointing. Agents used in the treatment of ACTH hypersecretion can be divided into two classesthose that act centrally to reduce ACTH release and those that act peripherally to reduce cortisol production or block its effect (Table 21-1; see Chap. 1). Centrally acting agents are preferred if a drug is to be used for primary therapy; moreover, they are the only agents appropriate for the treatment of Nelson syndrome. Peripherally acting drugs are the preferred agents for rapid preoperative treatment of severely ill patients awaiting surgery. When the treatment regimen involves the chronic use of peripherally acting drugs, the resultant reduction in cortisol and in negative feedback may be followed by an increase in ACTH hypersecretion, thereby necessitating increased dosages of the drug.

TABLE 21-1. Treatment of Adrenocorticotropic Hormone Hypersecretion

CENTRALLY ACTING DRUGS BROMOCRIPTINE Unlike the excellent results achieved with bromocriptine therapy in patients with hyperprolactinemia, long-term administration of the drug, even at dosages of 20 to 30 mg per day, effectively reduces ACTH hypersecretion in only a few patients.29 Although a single 2.5-mg dose of bromocriptine reduces ACTH levels in ~40% of patients, many of these short-term responders fail to improve significantly with long-term treatment. Conversely, some patients who fail to respond to a single dose of bromocriptine demonstrate marked improvement in symptoms and in ACTH hypersecretion with prolonged therapy.30 Neither the pretreatment ACTH and cortisol

levels nor the tumor size can be used to predict accurately the response to therapy. CYPROHEPTADINE The antiserotoninergic effect of cyproheptadine hydrochloride is thought to be the mechanism whereby ACTH secretion is reduced; however, this drug also has anticholinergic, antihistaminic, and antidopaminergic effects. Thirty percent to 50% of patients with Cushing disease achieve an initial clinical remission with this agent.31 Usually, when the drug is discontinued, elevated cortisol levels and symptomatic disease promptly return. No clinical features can predict which patients will respond to cyproheptadine. Importantly, many authors report poor efficacy and significant side effects with this drug. Occasionally, patients with Nelson syndrome have been reported to improve with administration of cyproheptadine. VALPROIC ACID The anticonvulsant agent valproic acid (and its derivatives) is a g-aminobutyric acid transaminase inhibitor that decreases ACTH hypersecretion in some patients with Cushing disease or Nelson syndrome. Reduction of tumor size with valproate sodium has been reported in a single instance.32 The drug is highly protein bound and has a serum half-life of 6 to 16 hours. Capsules should be swallowed whole and not chewed to avoid local irritation to the mouth and pharynx. Nausea and vomiting are commonly experienced at the time therapy is initiated. Tolerance to these side effects develops rapidly, and symptoms may be reduced by administering the drug with meals. Fatal hepatic failure has occurred in several patients receiving this drug as an anticonvulsant agent. Liver function tests should be performed before the initiation of therapy and at regular intervals during the first year. The drug should not be used in patients with a history of liver disease and should be discontinued if evidence of hepatic dysfunction is found. However, hepatic dysfunction has been known to progress even after discontinuation of the drug. An increased incidence of neural tube defects has been reported in children whose mothers received this agent during the first trimester of pregnancy. PERIPHERALLY ACTING DRUGS METYRAPONE Metyrapone reduces the production of cortisol by inhibiting 11-b-hydroxylation in the adrenal gland. The dosage is titrated to maintain normal serum cortisol levels (which should be evaluated at multiple intervals throughout the day) or titrated to keep the 24-hour urine free cortisol level within the physiologic range. The maintenance dosage varies from 250 mg three times a day to 1000 mg four times a day.30,33 The metabolism of metyrapone is accelerated by administration of phenytoin (Dilantin). The most common side effect is gastrointestinal irritation, which can be avoided by administering the drug with food. Despite improvement in serum cortisol levels, some women note worsening of hirsutism and acne during therapy.33 Cost and side effects may be reduced and efficacy enhanced by combining metyrapone with aminoglutethimide, with 1 g per day of each administered in divided doses. Although the manufacture of metyrapone tablets has been discontinued, capsules remain available from the manufacturer. MITOTANE Mitotane (1,1-dichloro-2-[o-chlorophenyl]-2-[p-chlorophenyl]-ethane or o,p'-DDD) suppresses the function of the zona fasciculata and zona reticularis of the adrenal cortex. The drug has been known to cause necrosis of the adrenal gland, producing acute adrenal insufficiency. Mitotane is inappropriate for rapid treatment because control of cortisol secretion requires 2 to 4 months of therapy.34 It may be useful in the treatment of patients awaiting the full effect of radiation therapy or in those in whom surgery and irradiation have failed.3,30 AMINOGLUTETHIMIDE Aminoglutethimide reduces cortisol production by inhibiting the conversion of cholesterol to D5-pregnenolone. During short-term therapy, serum cortisol levels usually are suppressed to less than one-half of pretreatment values. In some patients, glucocorticoid insufficiency occurs, necessitating concurrent glucocorticoid replacement therapy. When aminoglutethimide is used to treat patients with Cushing disease, a secondary increase in ACTH levels frequently leads to escape from acceptable control.30 Few patients have been treated for >3 months. Therapy is begun with administration of one 250-mg tablet every 6 hours. This dosage is then increased by 250 mg per day every 1 to 2 weeks until a total daily dose of 2 g is reached. Significant side effects occur in two-thirds of patients treated with this agent. The most frequent effects of the drug include drowsiness, which occurs in 33% of patients; skin rashes, which affect 16%; and nausea and vomiting, which occur in 13%. Other significant side effects include vertigo and depression. In general, side effects decrease with smaller doses and often improve or disappear after 1 to 2 weeks of continued therapy. Skin rashes may represent allergic or hypersensitivity reactions; if these are severe or persistent, the drug should be discontinued. Interference with thyroid hormone synthesis may produce hypothyroidism. Decreased estrogen synthesis may produce menstrual irregularities and increased hirsutism and acne in some women. Two cases of pseudohermaphroditism were reported in female infants of mothers who took this drug while pregnant. Because aminoglutethimide increases dexamethasone metabolism, hydrocortisone or cortisone acetate is preferred if glucocorticoid replacement therapy is needed. Inhibition of aldosterone synthesis may produce mineralocorticoid deficiency, presenting with orthostatic or persistent hypotension, which may require therapy with fludrocortisone acetate (Florinef). TRILOSTANE Trilostane is an inhibitor of the 3-b-hydroxysteroid dehydrogenase: D4,D5-isomerase enzyme system. It is generally less effective than the agents described earlier, and results are highly variable.35 Therapy is initiated with 30 mg of trilostane four times a day. This dosage is then increased as required to control serum cortisol and urinary cortisol levels, with an increase every 3 to 4 days until a total dose of 480 mg per day is reached. Significant side effects occur in half of treated patients. Gastrointestinal symptoms are the most common of these, with abdominal pain and discomfort being reported in 16% of patients, diarrhea in 17%, and nausea and vomiting in 5%. Trilostane has been reported to decrease progesterone levels, which has led to cervical dilation and termination of pregnancy in some women. KETOCONAZOLE Ketoconazole is an antimycotic agent that decreases serum cortisol by inhibiting cholesterol synthesis through blockade of the 14-demethylation of lanosterol. Ketoconazole may also inhibit 11-hydroxylation and may decrease the binding of glucocorticoid to its receptor. This drug has been reported to be effective in the treatment of patients with Cushing disease in whom surgery and other drug therapy have proved unsuccessful.3,30,36 After oral administration, the drug is rapidly absorbed. An acid pH is required for absorption; therefore, in patients who are also taking antacids or antihistaminic H2-inhibitors, the drug should be administered 2 hours after such therapy. Patients with achlorhydria may need to dissolve the tablets in aqueous hydrochloric acid. In serum, the drug is 99% protein bound. In patients with Cushing disease, therapy is initiated with 400 mg of ketoconazole administered every 12 hours for 1 month; this dosage is then decreased to 400 to 600 mg per day. Urinary cortisol levels were reported to decline significantly within 1 day after onset of therapy. In patients receiving conventional antifungal doses (200400 mg per day), the most common side effects are nausea and vomiting, occurring in 3%, and abdominal pain, occurring in 1.5%. Hepatotoxicity has been reported to occur in 1 in 10,000 treated patients; this condition usually resolves on discontinuation of the drug. However, one fatal case of hepatic necrosis that progressed despite discontinuation of the drug was reported. GLUCOCORTICOID RECEPTOR ANTAGONIST Mifepristone (RU 486) is a synthetic steroid agonist antagonist that blocks the binding of glucocorticoids to their receptor. It is under investigation as a potential therapeutic agent in the treatment of Cushing disease.37

GROWTH HORMONE HYPERSECRETION

Transsphenoidal surgery remains the treatment of choice for growth hormonesecreting adenomas (see Chap. 23). The overall rate of cure (defined as serum growth hormone levels of <5 ng/mL) is ~68% (88% of microadenomas and 59% of macroadenomas).38 Approximately 75% of patients who undergo conventional radiation therapy eventually achieve normalization of serum growth hormone levels over several years (see Chap. 22). The cure rate with proton beam irradiation of tumors without extrasellar extension approaches 95%. Medical therapy has generally been reserved for those patients who refuse surgery or radiation therapy or in whom these treatment modalities have been unsuccessful as well as for those patients awaiting the full effects of irradiation. Pretreatment with somatostatin analogs may improve surgical outcomes.39 In patients with acromegaly secondary to paraneoplastic (ectopic) growth hormonereleasing hormone secretion (see Chap. 12 and Chap. 219), removal of the tumor secreting growth hormonereleasing hormone is the treatment of choice. If this is impossible, medical treatment may be indicated. BROMOCRIPTINE Bromocriptine, at a dosage of 20 to 60 mg per day, is effective in many patients with acromegaly. At this dosage level, growth hormone levels are suppressed by more than half in ~70% of patients, but normalization is observed in <25%.40,41 Symptomatic improvement (decreased sweating, decreased softtissue swelling, improvement of sexual functioning, and decreased joint swelling) and an improvement in glucose tolerance are noted in 80% to 90% of patients.40,41 This discrepancy between growth hormone reduction and clinical improvement has been attributed to changes in the form of growth hormone secreted. Most authors recommend the use of insulin-like growth factor-I (IGF-I; also known as somatomedin C), rather than multiple determinations of growth hormone levels, to monitor therapy. In a few patients, a decrease in tumor size was observed. As with hyperprolactinemia, other dopamine agonists also appear to be effective in the treatment of acromegaly.42,43 Bromocriptine maintenance therapy in patients with acromegaly usually requires a minimum of 10 to 20 mg per day; and some patients require up to 80 mg per day. Bromocriptine dosage should be increased slowly to the maximum tolerated dose or until growth hormone levels normalize. Because the drug's half-life may vary, control may be assessed by measuring growth hormone levels three times the morning after a regular evening dose and then every 2 hours for 8 hours after the patient's usual morning dose. If serum growth hormone levels have not been suppressed significantly after several months of the maximum tolerated dosage, the drug should be discontinued. Most patients who respond to long-term bromocriptine therapy exhibit a marked reduction in growth hormone levels within 4 to 8 hours after a single 2.5-mg dose. Patients have been treated for up to 5 years, with maintenance of growth hormone suppression and clinical improvement. Discontinuation of the drug is usually followed by a rapid increase in growth hormone to pretreatment levels and a return of symptoms. Because patients who have undergone previous radiation therapy may have a normal serum growth hormone level years after treatment, bromocriptine use in these patients should be discontinued for 4 to 8 weeks every 1 to 2 years to reevaluate growth hormone secretion. SOMATOSTATIN ANALOGS Although somatostatin, a physiologic growth hormone inhibitor, lowers serum growth hormone levels in patients with acromegaly, it requires continuous infusion. In addition, the suppression of insulin secretion leads to hyperglycemia, and rebound growth hormone secretion is often observed after the infusion is discontinued. The longer-acting somatostatin analog octreotide acetate (SMS 201-995; see Chap. 169) is considerably more potent in the suppression of growth hormone. Growth hormone levels are suppressed for 8 to 12 hours after the subcutaneous administration of this analog. Octreotide has become the drug of choice for the medical therapy of acromegaly. It has been used successfully as primary therapy in patients with acromegaly44 as well as in patients in whom surgery and radiation therapy have failed and in whom control has not been achieved with dopaminergic drugs.3,45,46 and 47 Therapy is initiated with 50 g given subcutaneously every 12 hours, and the dose is increased based on the clinical and biochemical response. Many patients achieve a maximal response with a dosage of 100 g three times per day, but some may require that octreotide be administered at higher doses or every 6 hours. Some have been treated with up to 500 g three times per day or by continuous subcutaneous infusion. Symptomatic improvement is noted in 90% of patients within days to weeks. In patients who respond to the drug, growth hormone reductions are apparent after the first dose, and an IGF-I response is noted within 2 weeks. A reduction in growth hormone secretion occurs in 70% of patients, and mean growth hormone levels are usually normalized if the pretreatment value was <20.45,46 IGF-I levels are reduced in 90% of patients and normalized in 60%.45,46 Partial tumor shrinkage has been found in 20% to 50% of patients.45,46 Some patients who do not respond adequately to either octreotide or a potent dopamine agonist have a greater response when the two agents are administered together. Mild to moderate side effects are noted by one-third of patients, including pain at the injection site, abdominal pain, diarrhea and steatorrhea, vitamin B12 malabsorption, gastritis, and worsening of glucose tolerance (although this also improved in many patients with effective therapy of the acromegaly).45 Of greatest concern has been the development of gallstones, which may occur in up to 18% of patients. Some physicians administered a cholelitholytic agent with octreotide.48 Gallstones resolved after octreotide was discontinued. No rebound increase in growth hormone levels after discontinuation of the drug has been reported. Sandostatin LAR is a long-acting formulation of octreotide. Intramuscular injections of 20 to 30 mg every 4 weeks have been shown to be effective for up to 3 years in suppressing growth hormone and IGF-I levels and in inducing tumor shrinkage.49,50 As with octreotide, side effects include pain at the injection site, gastrointestinal symptoms, and, infrequently, development of gallstones or B12 malabsorption49,50 Lanreotide is a depot preparation of somatostatin. Intramuscular injections of 30 mg every 10 to 14 days have also been shown to be effective in suppressing growth hormone and IGF-I levels and in inducing tumor shrinkage.51,52 As with the other treatment modalities, symptomatic improvement may be greater than the changes in IGF-I or growth hormone. Side effects are similar to those of octreotide and Sandostatin LAR.52a Further studies are needed to determine which of the two long-acting preparations is more effective.52b ESTROGEN THERAPY A decline in IGF-I levels in acromegalic patients treated with estrogen has suggested that this might be a useful therapy. However, estrogen appears to directly suppress IGF-I generation and does not decrease growth hormone or appreciably influence the symptomatology.

GONADOTROPIN HYPERSECRETION
GONADOTROPIN-SECRETING ADENOMAS The standard therapy for gonadotropin-secreting adenomas is surgical resection, often followed by radiation therapy. However, this approach is rarely curative. Marked reductions (but not normalization) of follicle-stimulating hormone and a-subunit secretion have been seen in patients receiving bromocriptine after unsuccessful surgical therapy or as primary therapy.53 Octreotide therapy has reduced a-subunit secretion and produced tumor shrinkage in a few patients.54 Medical therapy should be considered only for those patients who refuse surgical and radiation therapy or in whom such treatment has failed. Partial suppression of follicle-stimulating hormone was reported in male patients receiving testosterone or human chorionic gonadotropin (which increases endogenous testosterone).53 Clomiphene citrate or estrogen failed to suppress gonadotropin secretion in these patients, and in one patient, estrogen therapy increased luteinizing hormone and a-subunit secretion. TRUE PRECOCIOUS PUBERTY Previously, treatment of true precocious puberty was limited to the use of medroxyprogesterone acetate or cyproterone acetate. These agents failed to arrest bone age advancement adequately and were associated with significant side effects, including adrenal suppression, hypertension, and excessive weight gain. Moreover, they led to reduced fertility and an increased incidence of reproductive tumors in some patients. Long-acting gonadotropin-releasing hormone agonists are effective in the management of central precocious puberty. These agents include leuprolide, nafarelin, and histrelin.55,56 Although they initially stimulate further hormonal release, their continued use produces a marked suppression of spontaneous and stimulated gonadotropin release. This suppression is accompanied by a reduction in the circulating level of sex steroids (see Chap. 16 and Chap. 225). Although children treated with these agents continue to grow, a marked reduction in height velocity and a retardation in the advancement of bone age are seen. In

some patients, an improvement in predicted attained height occurs. The gain in height is greatest with early initiation of therapy. In girls, a reduction in breast diameter, regression of pubic hair, and a cessation of menses (except for a menstrual period after the initial rapid rise and fall in estrogen level) may be seen.55,56 By ultrasonography, uterine and ovarian size decrease, and ovarian cysts decrease or resolve. In boys, testicular volume is reduced and the frequency of erections and aggressive behavior is markedly decreased.55,56 On discontinuation of therapy, serum gonadotropin and sex steroid levels rapidly return to pretreatment pubertal values, height velocity and bone age increase, and the development of secondary sexual characteristics progresses.55,57 Leuprolide, histrelin, and buserelin are administered as once-daily subcutaneous injections, but nafarelin is administered as a nasal spray; leuprolide is also available in a sustained-release preparation for depot intramuscular injections. Gonadotropin and sex steroid hormone levels are elevated during the first week of therapy but are then significantly reduced as long as therapy is maintained. If the plasma concentrations of sex steroids are not adequately suppressed in multiple samples obtained over several hours (estradiol level of <15 pg/mL in girls and testosterone level of <20 ng/dL in boys), the dosage is increased until adequate suppression is attained. The most common side effect is erythema at the injection site, seen in nearly one-third of treated patients. In two-thirds of these patients, the reaction does not progress. Two patients with erythema and wheal formation were desensitized and were treated with oral diphenhydramine hydrochloride (Benadryl). No patient developed immunoglobulin G antibodies against the analog. One-fourth of the girls treated with these agents developed hot flashes, which in one patient persisted for 2 years.

HYPERSECRETION OF THYROID-STIMULATING HORMONE

THYROID-STIMULATING HORMONESECRETING ADENOMA The rare patients who have TSH-secreting adenoma present with increased serum levels of TSH and symptomatic thyrotoxicosis. They are usually treated surgically58,59 (see Chap. 15 and Chap. 42). Some suppression of TSH has been observed with bromocriptine therapy, but this has been inadequate to normalize thyroid hormone levels; a slight reduction in tumor size was reported in one patient. Conventional antithyroid therapy treats the thyrotoxicosis but not the primary lesion. Trials with octreotide acetate, however, appear promising.47,60 ISOLATED PITUITARY RESISTANCE TO THYROID HORMONE Patients with isolated pituitary resistance to thyroid hormone also present with TSH hypersecretion and symptomatic thyrotoxicosis, but have no demonstrable tumor. Unlike patients with TSH-secreting adenomas, TSH in these patients is suppressed by thyroid hormone administration.58 Because patients are thyrotoxic, treatment with conventional thyroid hormone is not suitable for long-term management. Treatment with propylthiouracil or methimazole reduces thyroid hormone levels but markedly increases TSH secretion. Bromocriptine therapy has been found to reduce TSH levels in some of these patients; usually, however, the serum TSH level remained elevated, and the patients continued to exhibit signs of mild thyrotoxicosis.61 The combination of propylthiouracil and bromocriptine may produce euthyroidism with only a minimally elevated serum TSH. In one family, the use of 60 g of triiodothyronine once daily led to TSH suppression, normalization of serum thyroxine, and amelioration of symptoms, but this mode of treatment was unsuccessful in other patients.62 In one patient, the relatively inactive triiodothyronine metabolite triiodothyroacetic acid (TRIAC) was used to normalize TSH and thyroid hormone values.63 NONSECRETORY ADENOMAS The standard treatment modality for nonsecretory adenomas, if large, is transsphenoidal adenomectomy. However, in a series of 11 patients treated with bromocriptine (15 to 60 mg per day), 9 patients showed an average reduction in tumor size of 38%, whereas 2 patients exhibited a progressive increase in tumor size with treatment.64 Bromocriptine and other dopamine agonists may be useful as adjunctive therapy in patients with nonsecretory lesions or in those who cannot tolerate surgery, but their use cannot be recommended routinely as primary therapy. CHAPTER REFERENCES
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. Vance ML, Thorner MO. Prolactinomas. Endocrinol Metab Clin North Am 1987; 16:731. Colao A, Annunziato L. Treatment of prolactinomas. Ann Med 1998; 30:452. Shimon I, Melmed S. Management of pituitary tumors. Ann Intern Med 1998; 129:472. Moriondo P, Travaglini P, Nissim M, et al. Bromocriptine treatment of microprolactinomas: evidence of stable prolactin decrease after drug withdrawal. J Clin Endocrinol Metab 1985; 60:764. Serri O, Rasio E, Beauregard H, et al. Recurrence of hyperprolactinemia after selective transsphenoidal adenomectomy in women with prolactinoma. N Engl J Med 1983; 309:280. March CM, Kletzky OA, Davajan V, et al. Longitudinal evaluation of patients with untreated prolactin-secreting pituitary adenomas. Am J Obstet Gynecol 1981; 139:835. Schlechte J, Dolan K, Sherman B, et al. The natural history of untreated hyperprolactinemia: a prospective analysis. J Clin Endocrinol Metab 1989; 68:412. Greenspan SL, Neer RM, Ridgway EC, Klibanski A. Osteoporosis in men with hyperprolactinemic hypogonadism. Ann Intern Med 1986; 104:777. Wollesen F, Bendsen BB. Effect rates of different modalities for treatment of prolactin adenomas. Am J Med 1985; 78:114. Molitch ME, Elton RL, Blackwell RE, et al. Bromocriptine as primary therapy for prolactin-secreting macroadenomas: results of a prospective multicenter study. J Clin Endocrinol Metab 1985; 60:698. Warfield A, Finkel DM, Schatz NJ, et al. Bromocriptine treatment of prolac-tin-secreting pituitary adenomas may restore pituitary function. Ann Intern Med 1984; 101:783. Wass JAH, Williams J, Charlesworth M, et al. Bromocriptine in management of large pituitary tumours. BMJ 1982; 284:1908. Dallabonzana D, Spelta B, Oppizzi G, et al. Reenlargement of macroprolac-tinomas during bromocriptine treatment: report of two cases. J Endocrinol Invest 1983; 6:47. Thorner MO, Perryman RL, Rogol AD, et al. Rapid changes of prolacti-noma volume after withdrawal and reinstitution of bromocriptine. J Clin Endocrinol Metab 1981; 53:480. Landolt AM, Keller PJ, Froesch ER, Mueller J. Bromocriptine: does it jeopardize the result of later surgery for prolactinomas? Lancet 1982; 2:657. Martin TL, Kim M, Malarkey WB. The natural history of idiopathic hyper-prolactinemia. J Clin Endocrinol Metab 1985; 60:855. Vance ML, Evans WS, Thorner MO. Bromocriptine. Ann Intern Med 1984; 100:78. Spark RF, Dickstein G. Bromocriptine and endocrine disorders. Ann Intern Med 1979; 90:949. Turkalj I, Braun P, Krupp P. Surveillance of bromocriptine in pregnancy. JAMA 1982; 247:1589. Raymond JP, Goldstein E, Konopka P, et al. Follow-up of children born of bromocriptine-treated mothers. Horm Res 1985; 22:239. Ciccarelli E, Grottoli S, Miola C, et al. Double blind randomized study using oral or injectable bromocriptine in patients with hyperprolactinemia. Clin Endocrinol 1994; 40:193. Molitch ME. Pathologic hyperprolactinemia: neuroendocrine regulation of prolactin secretion. Endocrinol Metab Clin North Am 1992; 21:877. Brue T, Pellegrini I, Gunz G, et al. Effects of the dopamine agonist CV 205-502 in human prolactinomas resistant to bromocriptine. J Clin Endocrinol Metab 1992; 74:577. Webster J, Piscitelli G, Polli A. A comparison of cabergoline and bromocriptine in the treatment of hyperprolactinemic amenorrhea. N Engl J Med 1994; 331:904 Colao A, Di Sarno A, Sarnacchiaro F, et al. Prolactinomas resistant to standard dopamine agonists respond to chronic cabergoline treatment. J Clin Endocrinol Metab 1997; 82:876. Biller BM, Molitch ME, Vance ML, et al. Treatment of prolactin-secreting macroadenomas with the once-weekly dopamine agonist cabergoline. J Clin Endocrinol Metab 1996; 81:2338. DeRosa M, Colao A, Di Sarno A, et al. Cabergoline treatment rapidly improves gonadal function in hyperprolactinemic males: a comparison with bromocriptine. Eur J Endocrinol 1998; 138(3):286. Mampalam TJ, Tyrrell JB, Wilson CB. Transsphenoidal microsurgery for Cushing disease: a report of 216 cases. Ann Intern Med 1988; 109:487. Boscaro M, Benato M, Mantero F. Effect of bromocriptine in pituitary-dependent Cushing's syndrome. Clin Endocrinol 1983; 19:485. Miller JW, Crapo L. The medical treatment of Cushing's syndrome. Endocr Rev 1993; 14:443. Krieger DT. Physiopathology of Cushing's disease. Endocr Rev 1983; 4:22. Koppeschaar HPF, Croughs RJM, van't Verlatt JW, et al. Successful treatment with sodium valproate of a patient with Cushing's disease and gross enlargement of the pituitary. Acta Endocrinol 1984; 107:471. Jeffcoate WJ, Rees LH, Tomlin S, et al. Metyrapone in long-term management of Cushing's disease. BMJ 1977; 2:215. Luton JP, Mahoudeau JA, Bouchard P, et al. Treatment of Cushing's disease by o,p' DDD. N Engl J Med 1979; 300:459. Dewis P, Anderson DC, Builock DE, et al. Experience with trilostane in the treatment of Cushing's syndrome. Clin Endocrinol 1983; 18:533. Sonino N. The use of ketoconazole as an inhibitor of steroid production. N Engl J Med 1987; 317:812. Nieman LK, Chrousos GP, Kellner C, et al. Successful treatment of Cushing's syndrome with the glucocorticoid antagonist RU 486. J Clin Endo-crinol Metab 1985; 61:536. Serri O, Somma M, Comtois R, et al. Acromegaly: biochemical assessment of cure after long-term follow-up of transsphenoidal selective adenomec-tomy. J Clin Endocrinol Metab 1985; 61:1185. Stevenaert A, Beckers A. Presurgical octreotide: treatment in acromegaly. Metabolism 1996;45:72. Wass JAH, Thorner MO, Morris DV, et al. Long-term treatment of acromegaly with bromocriptine. BMJ 1977; 1:875. Jaffe CA, Barkan AL. Treatment of acromegaly with dopamine agonists. Endocrinol Metab Clin North Am 1992; 21:713. Cozzi R, Attanasio R, Barausse M, et al. Cabergoline in acromegaly: a renewed role for dopamine agonist treatment? Eur J Endocrinol 1998; 139(5):516. Abs R, Verhelst J, Maiter D, et al. Cabergoline in the treatment of acromegaly: a study in 64 patients. J Clin Endocrinol Metab 1998; 83:374. Newman CB, Melmed S, George A, et al. Octreotide as primary therapy for acromegaly. J Clin Endocrinol Metab 1998; 83:3034. Lamberts SWJ, Uitterlinden P, Verschoor L. Long-term treatment of acromegaly with the somatostatin analogue SMS 201-995. N Engl J Med 1985; 313:1576. Ezzat S, Snyder PJ, Young WF, et al. Octreotide treatment of acromegaly: a randomized, multicenter study. Ann Intern Med 1992; 117:711. Comi RJ. Pharmacology and use in pituitary tumors. In: Gorden P, moderator. Somatostatin and somatostatin analog (SMS 201-995) in treatment of hormone-secreting tumors of the pituitary and gastrointestinal tract and nonneoplastic diseases of the gut. Ann Intern Med 1989; 110:35. Montini M, Gianola D, Pagani MD. Cholelithiasis and acromegaly: therapeutic strategies. Clin Endocrinol 1994; 40:401. Flogstad AK, Halse J, Bakke S, et al. Sandostatin LAR in acromegalic patients: long-term treatment. J Clin Endocrinol Metab 1997; 82:23. Davies PH, Stewart SE, Lancranjan I, et al. Long-term therapy with long-acting octreotide (Sandostatin-LAR) for the management of acromegaly. Clin Endocrinol (Oxf) 1998; 48:311. Caron P, Morange-Ramos I, Cogne M, Jaquet P. Three year follow-up of acromegalic patients treated with intramuscular slow-release lanreotide. J Clin Endocrinol Metab 1997; 82(1):18. Giusti M, Gussoni G, Cuttica CM, Giordano G. Effectiveness and tolerability of slow release lanreotide treatment in active acromegaly: six-month report on an Italian multicenter study. J Clin Endocrinol Metab 1996; 81:2089.

52a. Turner HE, Lindsell DR, Vadivale A, et al. Differing effects on gallbladder motility of lanreotide SR and octreotide LAR for the treatment of acromegaly. Eur J Endocrinol 1999; 141:590.

52b. Cozzi R, Dellabonzana D, Attonasio R, et al. A comparison between octreotide-LAR and lanreotide-SR in the chronic treatment of acromegaly. Eur J Endocrinol 1999; 141:267. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. Snyder PJ. Clinically nonfunctioning pituitary adenomas. Endocrin Metab Clin North Am 1993; 22:163. Katznelson L, Alexander JM, Klibanski A. Clinically nonfunctioning pituitary adenomas. J Clin Endocrinol Metab 1993; 76:1089. Wheeler MD, Styne DM. The treatment of precocious puberty. Endocrinol Metab Clin North Am 1991; 20:183 Lee PA. Advances in the management of precocious puberty. Clin Pediatr 1994; 33:54. Manasco PK, Pescovitz OH, Feuillan PP, et al. Resumption of puberty after long-term luteinizing hormonereleasing hormone agonist treatment of central precocious puberty. J Clin Endocrinol Metab 1988; 67:638. Beck-Peccoz P, Persani L, Mantivani S, et al. Thyrotropin-secreting pituitary adenomas. Metabolism 1996;45:75. Brucker-Davis F, Oldfield EH, Skarulis MC, et al. Thyrotropin-secreting pituitary tumors: diagnostic criteria, thyroid hormone sensitivity, and treatment outcome in 25 patients followed at the National Institutes of Health. J Clin Endocrinol Metab 1999; 84:476. Chanson P, Weintraub BD, Harris AG. Octreotide therapy for thyroid-stimulating hormonesecreting pituitary adenomas: a follow-up of 52 patients. Ann Intern Med 1993; 119:236. Takamatsu J, Mozai T, Kuma K. Bromocriptine therapy for hyperthyroidism due to increased thyrotropin secretion. J Clin Endocrinol Metab 1984; 58:934. Rosler A, Litvin Y, Hage C, et al. Familial hyperthyroidism due to inappropriate thyrotropin secretion successfully treated with triiodothyronine. J Clin Endocrinol Metab 1982; 54:76. Beck-Peccoz P, Piscitelli G, Cattaneo MG, Faglia G. Successful treatment of hyperthyroidism due to nonneoplastic pituitary TSH hypersecretion with 3,5,3'-triiodothyroacetic acid (TRIAC). J Clin Endocrinol Invest 1983; 6:217. Wollesen F, Andersen T, Karle A. Size reduction of extrasellar tumors during bromocriptine treatment. Ann Intern Med 1982; 96:281.

CHAPTER 22 RADIOTHERAPY OF PITUITARYHYPOTHALAMIC TUMORS Principles and Practice of Endocrinology and Metabolism

CHAPTER 22 RADIOTHERAPY OF PITUITARYHYPOTHALAMIC TUMORS

MINESH P. MEHTA Pituitary Adenomas Prolactinomas Growth HormoneSecreting Tumors Adrenocorticotropic HormoneSecreting Tumors Non-HormoneSecreting Tumors Giant Pituitary Adenomas Craniopharyngiomas Overall Survival Local Recurrence Morbidity and Mortality Dose Considerations Radiotherapy for Recurrent Craniopharyngioma Hypothalamic Neoplasms Germinoma Parasellar and Skull Base Meningioma Glioma Complications of Brain Irradiation Endocrine Complications Vision System Complications Postirradiation Brain Necrosis Vascular Complications Radiation-Induced Carcinogenesis Radiation Techniques Goals of Radiation Therapy and Conventional External Beam Techniques Innovative Radiation Techniques Chapter References

Radiotherapeutic management of pituitary adenomas and tumors of the hypothalamus requires a thorough understanding of the various hypersecretory syndromes and their wide spread clinical manifestations, a working knowledge of the diagnostic principles, and an appreciation for the roles of surgery and medical management. Treatment with external beam radiotherapy typically results in little radiographic change in the tumor itself, and endocrine changes may require years to become detectable or stabilized. Meaningful evaluation of pituitary or hypothalamic tumors requires long-term follow-up to assess response and to observe complications. This chapter addresses the overall management of pituitary adenomas, craniopharyngiomas, hypothalamic tumors (including germ-cell neoplasms), parasellar meningiomas, and astrocytomas. This discussion emphasizes the role of and indications for radiotherapy. Standard external-beam techniques for radiotherapy are outlined, and less common techniquessuch as intracavitary radioisotope instillation, brachytherapy, particle beam radiotherapy, and stereotactic radiosurgeryare discussed. The complications of conventional radiotherapy are also addressed.

PITUITARY ADENOMAS
The role of radiotherapy in the management of pituitary adenomas remains poorly defined because of several factors, including the occasional need for emergent surgical decompression; the availability of competing therapeutic alternatives, such as transsphenoidal adenomectomy and medical treatment with dopamine agonists; the slow decline in hypersecretion after irradiation; the lack of large, well-controlled prospective randomized trials; concerns regarding possible long-term toxic effects of radiotherapy; and the false belief that benign tumors are not effectively treated by radiation therapy. Available data typically represent experience with a small number of cases accrued over a period of several years or decades and contain the inherent bias of referral patterns at tertiary care centers. Most patients have not been observed for sufficiently long periods to allow adequate interpretation of response and toxicity. Despite these limitations, it has become clear that the role of radiotherapy in the management of pituitary neoplasms falls into two distinct categories: the control of hypersecretion when other modalities have failed or are contraindicated and the control of mass effects. For most pituitary adenomas, external beam irradiation, delivered in conventional fraction sizes of 1.8 to 2 Gy to a total of 45 Gy or more, appears to suffice. Newer techniquessuch as radiosurgery, interstitial implantation, and particle beam therapyare under investigation. PROLACTINOMAS Prolactinomas, predominantly macroadenomas, were previously classified with the nonsecreting tumors as chromophobe adenomas; before the prolactin assay era, the major goal of radiotherapy was the control of the mass effect. With the availability of the prolactin assay, it became clear that some macroadenomas are functionally active. Prolactin-secreting microadenomas usually are effectively managed by bromocriptine or, rarely, by transsphenoidal adenomectomy. Long-term control of hyperprolactinemia after surgery alone is rare. Large tumors and tumors that persistently secrete prolactin despite resection can be treated medically with bromocriptine, a dopamine agonist that decreases prolactin levels and causes tumor shrinkage. The continued use of bromocriptine may be complicated by toxic effects, and its discontinuation often results in resumption of tumor growth and of prolactin hypersecretion. Despite prior therapy with bromocriptine, tumor growth may resume in women who become pregnant. Radiotherapy is indicated if hyperprolactinemia persists despite transsphenoidal tumor resection or the use of bromocriptine. The results of several single-institution studies are summarized in Table 22-1.1,2,3,4,5,6,7,8 and 9 Although irradiation causes a dramatic decrease in prolactin levels in some patients, the response is not always predictable. Serum assays obtained after radiotherapy have demonstrated a slow and variable decline in prolactin levels. Transient hyperprolactinemia, lasting as long as 2 years, has been described in patients irradiated for other pituitary-hypothalamic conditions.

TABLE 22-1. Control of Prolactinomas with Radiotherapy

After irradiation, prolactin levels decrease by an average of 60% by the end of the first year; after 3 or more years, the mean prolactin level decreases to one-tenth of the preirradiation value.5 In another study, 16 (44%) of 36 prolactin-secreting adenomas were controlled after radiotherapy.8 Another 36 female patients (12 with macroprolactinomas and 24 with microprolactinomas) were irradiated to a total dose of 45 Gy (1.8-Gy fractions) with a three-field technique.9 All patients underwent baseline and periodic reassessment of anterior and posterior pituitary function at intervals of 1 year or less while off bromocriptine for at least 2 months; they also had dynamic screening with thyrotropin-releasing hormone and luteinizing hormonereleasing hormone and hypoglycemic stimulation every 2 to 3 years. The preirradiation prolactin levels ranged from 1150 to 34,000 mU/L. With a mean follow-up of 8.5 years (range, 314 years), the postirradiation serum prolactin levels fell to normal (i.e.,

<360 mU/L) in 18 patients (50%). Another 10 patients (28%) had prolactin levels just above the normal range (378780 mU/L). Only 2 patients (6%) demonstrated an increase in prolactin levels; another patient had a radiographically confirmed recurrence. Neither the pretreatment prolactin level nor the size of the tumor influenced the outcome from radiotherapy. In a smaller study, long-term control of hyperprolactinemia was achieved in 7 (70%) of 10 patients after a total dose of 45 Gy (1.8-Gy fractions) of radiation.7 These studies elegantly demonstrated that radiotherapy and dopamine agonists are useful for long-term control of subtotally resected macroprolactinomas. Despite the paucity of long-term longitudinal studies of prolactin secretion after radiation therapy, studies indicate that over 2 to 13 years, prolactin levels return to normal or near-normal levels in >75% of irradiated patients.7,9 However, physiologic symptoms, such as amenorrhea from markedly elevated prolactin production, persist even when these levels fall to baseline. For example, in a series of 24 patients with prolactin-secreting macroadenomas treated with transsphenoidal surgery, dopamine-agonist therapy, and 45 Gy radiotherapy, tumor control and prolactin reductions were achieved in all, but amenorrhea persisted in the majority.10 GROWTH HORMONESECRETING TUMORS Initial reports of the effectiveness of radiotherapy for the treatment of growth hormone (GH)secreting tumors relied on clinical assessments of response, because direct measurements of GH levels were unavailable. Several investigators have reported clinical control rates of 80% to 90% after irradiation to doses of 40 Gy or more using conventional fraction sizes of 1.8 to 2 Gy per day.11,12,13 and 14 Clinical experience also suggests a dose-response relationship; control improves with doses as great as 40 Gy, and toxic effects occur at doses higher than 50 to 54 Gy. In 105 patients with GH-secreting pituitary adenoma treated to a total dose of 42 to 55 Gy, no improvement in local control was found at doses higher than 45 Gy. Moreover, the advent of the serum GH assay revealed that the functional response by these tumors to radiotherapy occurred slowly. The current definition of cure requires a GH level of <5 mU/L (2 ng/mL), with the understanding that long-term survival is dependent on such a reduction. Although radiation-induced GH reduction requires a lag period, as opposed to the immediate decline seen after surgery, the hypothalamic effects of radiation abrogate the endogenous somatostatin tone, thereby abolishing several responses that may enhance GH secretion. In a 20-patient study, arginine increased GH hypersecretion in those with a prior history of acromegaly whose GH levels had normalized after surgery. This phenomenon could not be demonstrated in patients postradiation.15 Patients treated after incomplete pituitary adenomectomy with high-voltage or with proton beam irradiation achieved an 80% decrease in GH levels after 4.5 years.14 In another report, control of GH secretion was achieved in 9 (60%) of 15 patients after irradiation.8 For 56 acromegalic patients in whom surgery had been unsuccessful, treatment with 50 Gy of radiation brought a 50% reduction in preirradiation GH levels in 51 patients (91%) at 26 months; for 40 of these patients, there was a further 50% decrease in GH levels at 42 months.16 At 2, 5, and 10 years after radiotherapy, endocrinologic control rates (defined by a drop in GH levels to <10 ng/mL) were 38%, 73%, and 81%, respectively.13 These data suggest that in patients who respond to radiotherapy, the GH depression may follow a first-order reaction, with a half-life of just longer than 2 years. 16 Because endocrinologic normalization can occur almost immediately after successful transsphenoidal resection, this approach has become the standard. Irradiation of GH-secreting pituitary microadenomas and most macroadenomas is limited to patients with persistent GH hypersecretion after resection and to those in whom surgery is otherwise contraindicated. Overall, ~75% of such patients achieve eventual control, defined as a GH level <10 ng/mL or a lack of progression of growth of the adenoma4,8,11,12,13 and 14,17,18,19,20,21,22 and 23 (Table 22-2). Additionally, some data suggest that macroadenomas may have a higher failure rate after radiotherapy, as exemplified in a 21-patient series in which 5 of 6 failed patients had macroadenomas.23

TABLE 22-2. Control of Growth HormoneSecreting Adenomas with Radiotherapy

ADRENOCORTICOTROPIC HORMONESECRETING TUMORS CUSHING DISEASE Cushing disease is typically associated with adrenocorticotropic hormone (ACTH)secreting pituitary microadenomas. The standard management for this syndrome remains transsphenoidal adenomectomy, which yields a 75% success rate in the control of hypercortisolism (see Chap. 23 and Chap. 75). Radiotherapy is restricted to inoperable patients or those in whom hypercortisolism persists after resection. Several reports of pituitary irradiation for Cushing disease are summarized in Table 22-3.8,21,22,23,24,25,26,27,28 and 29 The control rates range from 50% to more than 80%. Although a dose-response relationship has not been established in a prospective, randomized trial, retrospective data suggest that the control of hypercortisolism requires doses of 40 to 50 Gy. Cortisol levels normalized in 53% of patients treated with radiotherapy to the extent that no further treatment was required.26

TABLE 22-3. Control of Adrenocorticotropic HormoneSecreting Adenomas with Radiotherapy

In another study of 21 patients with Cushing disease for whom irradiation was the primary treatment, a total of 45 Gy in 25 fractions was delivered using a three-field technique, and patients were observed for 5.8 to 15.5 years (median of 9.5 years).29 Initially, all patients were treated with metyrapone to normalize cortisol levels; at the latest follow-up evaluation, 57% had normal mean cortisol levels throughout the day and were off all therapy. Five patients required additional treatment with bilateral adrenalectomy or transsphenoidal hypophysectomy, yielding a failure rate for treatment of ACTH-secreting adenomas with radiotherapy of ~25%. The time to normalization of cortisol levels ranged from 0.7 to 10.5 years (median, 4 years). In one study of 15 children with Cushing disease treated with radiotherapy, an 80% control rate was achieved.28 The time to remission of hypercortisolism was only 9 to 18 months, in contrast to the substantially longer remission times for GH- and prolactin-secreting adenomas. NELSON SYNDROME

Nelson syndrome consists of hypersecretion of ACTH and melanocytestimulating hormone, with dermal hyperpigmentation and aggressive growth of a pituitary adenoma, after bilateral adrenalectomy. Limited data suggest that the syndrome can successfully be treated with megavoltage pituitary irradiation. In a series of 15 patients observed for a median of 9.6 years (range, 1.517.3 years), clinical, radiographic, and endocrinologic improvement was observed in 14 cases (93%) after 45 Gy of megavoltage irradiation.29 NON-HORMONESECRETING TUMORS The primary goal of therapy for endocrinologically inactive pituitary adenomas is control of the mass effect, which typically manifests with impaired vision, headaches, and pressure-induced atrophic hypopituitarism. The mass effect primarily is a consequence of tumors that were previously referred to as chromophobe adenomas; occasionally, it is caused by GH-secreting macroadenomas. Modern pathologic assessment has demonstrated that some tumors that had been assumed to be nonfunctional chromophobe adenomas are instead prolactin-secreting tumors; these are discussed separately later. Other so-called nonfunctional tumors secrete gonadotropins or their subunits (see Chap. 16). The consequences of a mass effect from pituitary adenomas can be quite severe. For example, in a report of 140 patients with macroadenomas, visual field deficits were identified in 92%; in 10%, the tumors invaded the brain or the nasopharynx.30 In a subgroup of patients managed surgically without postoperative radiotherapy, the 5-, 10-, and 20-year recurrence-free survival rates were only 38%, 14%, and 0%, respectively. Postoperative radiotherapy increased the recurrence-free survival rates at 5, 10, and 20 years to 96%, 86%, and 73%, respectively. In that same study, 23 patients with relatively minor visual field deficits or with underlying medical conditions precluding surgery were selected for treatment with radiotherapy only; the 15-year recurrence-free survival rate was 93%.30 Radiotherapy normalized visual field deficits in more than two-thirds of patients in whom one quadrant or less was affected, an outcome that compared favorably with the results obtained with surgery. However, in patients with more extensive visual loss, visual restoration was more rapid and effective after surgery than after radiotherapy. The potential for blindness from such large tumors mandates immediate surgical debulking if feasible; however, the probability of residual disease and late recurrence is significant. Of 112 patients with nonfunctional pituitary adenomas, the actuarial progression-free survival rates after primary irradiation (25 patients) or postoperative irradiation (87 patients) was 97%, 89%, 87%, and 76% at 5, 10, 15, and 20 years, respectively.31 No demonstrable difference in local control rates was identified between patients who underwent postoperative or primary irradiation. Because those data represent the experience of a single institution over more than 2 decades, radiother-apeutic techniques and doses varied. A further analysis of the dose-response relationship revealed no significant advantage in local control at dose levels that ranged between 35.7 and 62.3 Gy. Some data, however, do suggest the existence of a dose-response relationship; a 78% local control rate was obtained after 40 to 50 Gy, compared with a 56% local control rate after 30 to 40 Gy of radiation.32 A summary of local control rates for nonfunctional adenomas after radiotherapy is presented in Table 22-4.7,8,11,12,21,22,31,33,34

TABLE 22-4. Control of Nonfunctioning Pituitary Adenomas with Radiotherapy

Additional data supporting a role for radiotherapy in the preservation and improvement of vision in selected patients come from a report of 25 patients with pituitary macroadenomas causing vision impairment. These patients were treated with radiation therapy alone. Twenty-three patients underwent neuro-ophthalmologic evaluation before and after radiation therapy. With a median follow-up of 3 years, 78% of patients whose pretreatment visual field deficits had less than dense hemianopia and who also did not have diffuse optic atrophy experienced visual field improvement. Only 1 patient experienced tumor progression.35 GIANT PITUITARY ADENOMAS The giant pituitary adenoma is a rare lesion, and its definition varies among case reports.36 A study of 31 patients with this tumor reserved the term for lesions with 40 mm or more of suprasellar extension.34 Eleven of the tumors were secretory, and 20 were nonfunctional. Four patients were treated with surgery, 2 with radiotherapy, and 25 with surgery plus radiotherapy. The patients who received combined-modality treatment were those with significant residual disease after surgery. With a mean follow-up of 8 years, the recurrence rate in either single-modality treatment group was 67% (4 of 6 patients). The use of combined-modality therapy yielded a local control rate of 84%. Based on limited experience, this rare subcategory of nonfunctional pituitary adenoma is best treated with surgery plus radiotherapy.

CRANIOPHARYNGIOMAS
Craniopharyngiomas are relatively rare neoplasms that arise from epithelial remnants of the Rathke pouch and are typically found in the suprasellar region in children or adolescents; they account for ~5% of all intracranial neoplasms in childhood.37 These tumors tend to grow slowly, and the patients often present with compression of adjacent neural structures, such as the optic chiasm. Hypopituitarism also may occur. Surgical decompression is the optimal treatment for symptom relief and immediate palliation. However, the location, proximity, and adhesiveness of the tumor to adjacent neural structures often preclude an effective total resection. Aggressive attempts at total resection carry high rates of morbidity and mortality.38 The management options for craniopharyngioma include total resection, which is applicable only to a very small proportion of patients; subtotal resection alone; or subtotal resection and postoperative radiotherapy. To understand the role of postoperative radiotherapy, it is necessary to compare the results for total or subtotal resection alone with subtotal resection plus postoperative radiotherapy. In the absence of a prospective randomized trial, the only data available for comparison are retrospective single-institution reports that span several decades. including an institutional bias regarding therapeutic preference and the impact of technologic advances on diagnosis, resection, and radiotherapy. Although flawed, this database represents the only opportunity to compare the various treatment options for the management of craniopharyngioma. OVERALL SURVIVAL A review of the English language literature from the mid-1960s through 1998 yields 34 reports from which management and outcome data can be summarized39 (Table 22-5). The data are broken down into three categories representing total resection, subtotal resection, and surgery plus postoperative radiotherapy. Actuarial 5- and 10-year survival rates after total resection were 81% and 69%. After subtotal resection alone, the survival rates were 53% and 37% at 5 and 10 years, and if radiotherapy was added, the survival rates were 89% and 77% at 5 and 10 years, respectively. There is a trend toward longer survival in later reports, which probably reflects improvements in neuroimaging, neurosurgery, radiotherapy, and overall medical management. Nonetheless, the outcome for patients undergoing subtotal resection is inferior to that of patients who also receive postoperative radiotherapy. Although the inferior outcome of patients who undergo subtotal resection could reflect a patient selection bias, the overall 5- and 10-year survival rates of 53% and 37% do not indicate a benign disease process.

TABLE 22-5. Survival Rates 5 and 10 Years after Three Therapeutic Approaches for Craniopharyngioma: Retrospective Data from 34 Reports

LOCAL RECURRENCE Another important factor in the prognosis and management of craniopharyngioma is the local recurrence rate. A review of 31 published studies suggests that local recurrence rates are 29% (90 of 308 patients) after total resection, 73% (163 of 224) after subtotal resection, and 17% (104 of 596) after surgery plus radiotherapy. Even after aggressive resection, recurrences are reported for as many as one-third of patients.40 The 20-year experience in childhood craniopharyngioma from the Joint Center for Radiation Therapy in Boston is a 10-year actuarial rate for freedom from progression of 31% after resection only, 100% after radiotherapy only, and 86% for patients treated with resection plus radiotherapy at the time of diagnosis. A point to consider is that surgical reports comment only on patients in whom resection was attempted. Patients in whom resection was not feasible were not reported as failure of intent to resect. Reporting data for an intent to resect category could significantly lower the overall local control rates for this treatment approach. MORBIDITY AND MORTALITY The morbidity and mortality rates that occur with total resection of craniopharyngiomas are significant. The operative mortality ranges from 2% to 43% (mean, 12%), and the morbidity ranges from 12% to 61% (mean, 30%). Surgical morbidity includes damage to the hypothalamus, which may result in diabetes insipidus or other endocrine anomalies (range, 30%57%; mean, 40%), visual impairment (range, 10%35%; mean, 19%), obesity, and memory impairment. In another report, 48% of patients developed obesity, and 57% developed memory impairment.41 In the largest surgical series reported, the operative mortality for total resection was 17%, and the incidence of significant morbidity was 16%.42 In contrast, subtotal resections carry a mortality of ~1%. The risk of late radiation damage to the hypothalamus, pituitary, and optic nerve is relatively low after conventionally fractionated irradiation (i.e., 1.8-Gy fractions to a total dose of 54 Gy). With these dose recommendations, visual impairment ranges between 1% and 1.5%.43 The addition of radiotherapy to subtotal resection does not increase the mortality and adds little to morbidity. The impairment of cognitive performance is difficult to assess, because few studies address the issue in detail and because the protocols for neuropsychologic evaluation are not standardized. In an analysis of 35 patients with craniopharyngioma who underwent neurologic and neuropsychologic assessment, a lower morbidity rate was observed after subtotal resection followed by radiotherapy than after attempted radical tumor resection alone.44 From this experience, the researchers concluded that primary irradiation caused less frontal lobe dysfunction than radical subfrontal excision. Visual perceptual tests, visual acuity, and ocular motility improved in one-third of conservatively treated patients but deteriorated in patients who underwent radical tumor resection. A common deficit, independent of the treatment modality, was mild impairment of manual dexterity, which became evident in sequential tapping or in slow, smooth pursuit tasks. This deficit may represent the underlying basis for the awkwardness, the slowness, and the stiffness of gross motor function described in patients with craniopharyngioma. An impairment of orbital-frontal lobe functions, which mainly manifests as persevering responses, was more pronounced in patients who had radical subfrontal excision than in those who had primary radiotherapy. This deficit persisted as long as 19 years after the resection. The intelligence quotients of the surgical and the radiotherapy groups remained within normal limits. Only 1 of the 18 patients in the primary irradiation group showed signs of frontal lobe disorder and had difficulty in school, but 2 patients required some tutoring for mild or moderate learning disability.44 Most of the surgically treated patients were unable to maintain regular employment, achieve expected educational goals, or enjoy a normal family and social life. A review suggests that although total excisions may be appropriate for relatively small craniopharyngiomas, heroic attempts at total resection typically result only in partial resection with enhanced morbidity and mortality rates.45 In light of the excellent results achieved after subtotal resection plus postoperative radiotherapy, this should be considered the standard management. DOSE CONSIDERATIONS No prospective randomized trials analyzing dose-response relationships are available for the treatment of craniopharyngioma with radiotherapy. However, most institutions have routinely irradiated these patients to a total dose of 50 to 55 Gy in 1.8- to 2-Gy fractions. A long-term analysis (20 years) of dose-response data revealed a local failure rate of 50% with <54 Gy but only 15% with 54 Gy or higher doses.46 In a retrospective analysis of patients treated with 51.3 to 70 Gy, a higher incidence of radiation-associated complications was identified in those who received >60 Gy (with an actuarial incidence of optic neuropathy of 30% and brain necrosis of 12.5%), without any concomitant improvement in tumor control.47 RADIOTHERAPY FOR RECURRENT CRANIOPHARYNGIOMA Limited data suggest that radiotherapy at recurrence, after prior resection, yields a 10-year progression-free survival rate of >70% for patients with craniopharyngiomas.48 These results are similar to those obtained with subtotal resection and preliminary irradiation and suggest that radiotherapy may be delayed in very young children, in whom its toxic effects may be more pronounced. However, routinely delaying irradiation may entail additional morbidity from continued tumor growth, the possible requirement for repeat surgery, and irradiation of larger volumes at relapse. Delaying radiotherapy cannot be recommended for older children. Recurrences appear to occur from 3 to 192 months (median, 12 months) after subtotal resection.43

HYPOTHALAMIC NEOPLASMS
GERMINOMA Germinomas typically arise in the floor of the third ventricle and have a propensity to invade and compress adjacent neural structures, such as the optic chiasm, or to spread in the periventricular space, which permits craniospinal seeding. Although germinomas are the most common embryonal cell tumors, choriocarcinoma, endodermal sinus tumors, and teratomas may also occur at this site. The initial intervention for tumors arising in the floor of the third ventricle is often surgical. Surgery is indicated for diagnosis and to decompress the ventricles or to relieve pressure on the chiasm or the pituitary stalk. The rare teratomas in this location are amenable to surgical cure. Current staging recommendations include contrast-enhanced magnetic resonance imaging (MRI) of the entire craniospinal axis. The incidence of spinal seeding varies and is often a function of the thoroughness of the investigation. The author recommends the use of triple-strength contrast during spinal imaging to assist in the detection of seeding of the leptomeninges or the cauda equina. Cerebrospinal fluid and serum evaluation for a-fetoprotein and the b-subunit of human chorionic gonadotropin (b-hCG) are recommended. Endodermal sinus tumors and, to a lesser extent, embryonal carcinomas cause elevations of a-fetoprotein; choriocarcinomas cause elevations of b-hCG. If either marker is elevated, repeat levels should be obtained postoperatively, after allowing time for the markers to decline. These markers allow monitoring of the treatment and follow-up phases. Historically, the standard management of germinomas has consisted of craniospinal irradiation. So dramatic is the response of these tumors to radiation therapy that some researchers recommended a low test dose of limitedfield irradiation (to ~20 Gy) as a diagnostic test in place of a biopsy.49 This practice is supported by a report that surgery is associated with a 41% incidence of spinal dissemination compared with a rate of only 2% in nonbiopsied patients.50 Although the issues of total dose and whether the spinal axis should be prophylactically radiated remain controversial, the overall disease-free survival rate for patients with germinoma is usually in excess of 80%.51 Of all patients with confirmed intracranial germ-cell tumors treated at the Hospital of Sick Children from 1952 to 1989, 25 patients with germinoma treated with radiotherapy had a 5-year survival rate of 85%, and 13 patients with nongerminoma germ-cell tumors treated with radiotherapy had a 5-year survival rate of 45.5%.52 The efficacy of chemotherapy for treatment of germ-cell neoplasms in other body sites has led to similar trials for intracranial germ-cell tumors. Preliminary results for

treatment with platinum-based combination chemotherapy for 10 patients with intracranial germ-cell tumors are encouraging. Sevenpatients received primary chemotherapy consisting of vincristine, etoposide, and carboplatin before craniospinal axis irradiation; 3 patients had complete responses, 3 had partial responses, and 1 patient had stable disease. All 7 patients were alive and disease free at a median of 12 months after treatment. The author's treatment for germinomas, attaining complete response after two courses of chemotherapy, is a lowered radiation dose prescription, but nongerminomatous germ-cell tumors receive standard total dose irradiation.53 Because of its considerable morbidity, the value of craniospinal irradiation in very young children with nonseminomatous tumors (e.g., yolk sac tumors) is controversial. PARASELLAR AND SKULL BASE MENINGIOMA Adjuvant postoperative external beam irradiation is effective treatment for parasellar meningiomas, because complete resections are usually impossible. Highly customized treatment fields, usually based on computed tomography (CT)- or MRI-guided treatment planning, are routinely used to minimize irradiation of normal tissues. Typically, the radiotherapy prescription is a dose of 54 Gy in 30 fractions of 1.8 Gy each.54 In a series of 186 patients with meningiomas treated with megavoltage photon irradiation between 1963 and 1983, the 10-year actuarial causespecific survival rate was 67%.55 Radiotherapy alone resulted in improvement of neurologic performance in 12 (38%) of 32 patients with inoperable tumors. Multivariate analysis revealed that histology, extent of resection, and performance status at the time of presentation for radiotherapy were independent prognostic variables. In a series of 115 patients with benign meningioma treated for primary or recurrent disease, 36 patients were treated by subtotal resection plus external beam irradiation, and 79 patients were treated by subtotal resection alone.56 The progression-free survival rate for 17 patients irradiated after initial subtotal resection was 88% at 8 years, compared with 48% for a similar group of patients treated with surgery alone. Sixteen patients whose tumors were incompletely resected at the time of first recurrence were irradiated; 78% were progression free at 8 years. Only 11% of the patients treated with surgery alone were progression free at 8 years (P = 0.001). Twenty-five patients were irradiated with photons alone at doses of 45 to 60 Gy and at a median follow-up time of 57 months; 6 (24%) had recurrences. Eleven patients were treated with combined 10-MV photons and 160-MV protons using three-dimensional treatment planning. After 53 months, none of these patients had a recurrence. These studies support a role for radiotherapy in the treatment of incompletely resected or inoperable meningioma of all histologic types.55,56 Patients who undergo complete resection of the typical benign meningioma do not require adjuvant irradiation. The roles for stereotactically implanted high-activity iodine-125 seeds (i.e., brachytherapy), radiosurgery, or endocrinologic manipulation in the management of inoperable skull base meningioma are evolving, and these modalities should be considered when the indications are appropriate.57,58,59 and 60 GLIOMA The full histologic spectrum of gliomas, from the benign juvenile pilocytic variant to glioblastoma multiforme, is encountered in the hypothalamus; chiasmal gliomas may also invade the hypothalamus. Occasionally, oligodendroglioma, mixed tumors, and ganglioglioma may be encountered. The management of juvenile pilocytic astrocytoma and ganglioglioma is surgical. When surgery fails, external beam radiotherapy provides highly effective adjuvant treatment. Other glial tumors usually are not amenable to total resection, and postoperative irradiation is frequently used as adjuvant treatment. In the management of 33 children with hypothalamic-chiasmatic gliomas, the median time to tumor progression was 70 months in patients who received irradiation and 30 months in those who did not (P <0.05).61 Nonirradiated patients who progressed were treated with irradiation. Clinical or radiographic improvement occurred in 11 (46%) of 24 irradiated patients. The 5- and 10-year survival rates for irradiated patients were 93% and 74%, respectively.61 To delay radiotherapy in very young children and to minimize the incidence of long-term complications, there is increasing interest in the initial treatment of histologically benign glial tumors with chemotherapy. In a study of 6 children with optic pathway gliomas who were treated with carboplatin at the time of progression, the median age at diagnosis was 2 years (range, 4 months to 7 years), and the interval between diagnosis and treatment with carboplatin ranged between 7 months and 6.5 years (median, 1.8 years).62 Disease stabilization was observed in all patients, suggesting that carboplatin can arrest growth of progressive optic pathway gliomas in young children and may permit a delay in the use of radiotherapy.62 Among 19 children between the ages of 15 weeks and 15 years (median, 3.2 years) with chiasmal or hypothalamic gliomas who were treated with nitrosourea-based chemotherapy, 12 patients (7 juvenile pilocytic astrocytomas, 2 astrocytomas, 2 highly anaplastic astrocytomas, and 1 subependymal giant cell astrocytoma) received their chemotherapy immediately after diagnosis because of progressive symptoms.63 Another 7 patients (all astrocytomas) received chemotherapy at the time of tumor progression. In 15 (83%) of 18 patients that could be evaluated, the tumor responded to or stabilized after chemotherapy. With a median follow-up of 79 weeks, median time to tumor progression had not been reached, and no tumor-related deaths had occurred. Improvement or stabilization of visual field function was observed in 16 patients. These results suggest that nitrosourea-based chemotherapy is useful for the initial treatment of children with chiasmal or hypothalamic gliomas and allows deferral of irradiation until such time as the tumor progresses.

COMPLICATIONS OF BRAIN IRRADIATION

Modern megavoltage radiotherapy produces minimal acute toxicity, including temporary alopecia, mild dermatitis, and a serous otitis media if the middle ear is included in the treatment field. These acute toxic effects typically are grade 2 or less. The focus of the treatment planning process is to minimize late toxic effects, which are uncommon but can be devastating. Late toxic effects occur predominantly in tissue with a slow turnover time and reflect a combination of direct cellular and indirect vascular injury. The factors that predict a higher rate of complications include very young age, large total dose, large fraction size, large irradiated volume of normal tissue, and underlying medical conditions. In an unselected series of 134 patients who had undergone pituitary-hypothalamic irradiation over an 18-year period, 97% of whom had been treated to 45 to 50 Gy, complications attributable to radiotherapy occurred in 7 patients (2 second malignancies, 2 auditory deteriorations, and 3 vision deteriorations), underscoring the 5% or less risk from a carefully prescribed course of radiotherapy.64 ENDOCRINE COMPLICATIONS Despite a low rate of cell proliferation in pituitary adenomas, the sequential assessment of hormone levels has demonstrated that the secretory functions of the pituitary gland are relatively susceptible to irradiation.65 The incidence of hypopituitarism was analyzed in a group of 165 patients who underwent cranial irradiation to total doses of 37.5 to 42.5 Gy in 2.25- to 2.65-Gy fractions.66 The analysis, which spanned a 10-year period, tested anterior pituitary function using insulin hypoglycemia or glucagon stimulation; thyrotropin-releasing hormone and luteinizing hormonereleasing hormone levels; and basal estimations of GH, prolactin, thyroid hormones, and testosterone or estradiol. Tests were repeated at 6- to 12-month intervals. Hyposecretion developed most rapidly for GH and least rapidly for thyroid-stimulating hormone; gonadotropins and ACTH declined at an intermediate rate. The time required for 50% of patients with normal pituitary function to develop a hypofunctional pituitary was 1.2 years for GH, 3 years for luteinizing hormone and follicle-stimulating hormone (Fig. 22-1), and 3.2 years for ACTH. Almost two-thirds of patients exhibited the GHluteinizing hormone/follicle-stimulating hormoneACTHthyroid-stimulating hormone sequence of hypopituitarism. In another report of 84 patients, radiotherapy resulted in local control of the majority of pituitary adenomas, but by 10 years of follow-up, the prevalence of hypopituitarism rose from 29% to 92%, suggesting that residual pituitary function is highly susceptible to long-term negation after radiotherapy. Patients should, therefore, be appropriately evaluated in follow-up and counseled regarding the almost universal need for hormone replacement after radiation.66

FIGURE 22-1. Actuarial probabilityof maintaining normal pituitary function. Growth hormone (GH) declines most rapidly, followed by luteinizing hormone

(LH)/follicle-stimulating hormone (FSH) and adrenocorticotropic hormone (ACTH); thyroid-stimulating hormone (TSH) is most resistant to a decline. (Adapted from Littley MD, Shalet SM, Beardwell CG, et al. Hypopituitarism following external radiotherapy for pituitary tumors in adults. Q J Med 1989; 70:145.)

Before attributing hypopituitarism to irradiation, the clinician must consider that a large number of patients present with a compromised endocrine status before commencing radiotherapy and that the preexisting disease and the extent of tumor resection contribute significantly to this process. Additionally, radiation-induced hyperprolactinemia developed in 44% of patients; the mean prolactin rose from 227 to 369 mU/L by 2 years, but the level returned to normal in the following 2 years.66 VISION SYSTEM COMPLICATIONS Vision loss as a sequela of irradiation usually occurs within 2 years. When the literature was reviewed for reports of blindness as a complication of pituitary irradiation, only two cases of total vision loss were found when the fraction size was maintained below 2 Gy.67 In a series of almost 500 patients with pituitary adenomas treated to a total dose of as much as 50 Gy in fraction sizes of 2 Gy or less, no optic nerve or chiasmal injury was reported.30 A logistic regression dose-response analysis of injury to cranial nerves after proton beam radiation therapy delivered in 1.8-Gy equivalent fractions revealed that the risk of cranial nerve damage was 1% at a total dose equivalent to 60 Gy, and it increased to 5% at a 70-Gy equivalent dose.68 POSTIRRADIATION BRAIN NECROSIS Brain necrosis is an uncommon complication of cranial irradiation with modern radiotherapeutic technology, fraction size, and total dose prescriptions. In a series of more than 650 patients, only 2 (0.3%) developed brain necrosis, and these cases were associated with the use of opposed lateral fields.30 In the author's experience, opposed lateral fields are safe if used with very-high-energy photons (10 MV), small field sizes, fraction size no greater than 1.8 Gy, and total doses no higher than 50.4 Gy. Radiation necrosis most often occurs 9 months to 2 years after radiotherapy. The hallmarks of radiation necrosis include an elevated cerebrospinal fluid protein while glucose remains normal, a focal delta that slows on electroencephalography, and T1-weighted magnetic resonance images that show diffuse gadolinium-enhancing lesions in the optic chiasm and hypothalamus. After parasellar irradiation, patients with necrosis may present clinically with progressive vision impairment and dementia. Neurocognitive effects side effects may occur.68a Current treatment planning and dose prescription algorithms are aimed at maintaining the rate of clinically significant necrosis at <1%. VASCULAR COMPLICATIONS Anecdotal reports have linked the occurrence of single or multiple intracranial aneurysms to irradiation for pituitary adenoma.69 It has been suggested that postirradiation cerebral pathologyfrom localized to multifocal radiation necrosis and from localized to diffuse vasculopathyis not rare in children, whose young nervous systems are particularly sensitive to ionizing radiation.70 Of 156 patients irradiated for pituitary adenomas, 7 experienced cerebral strokes at intervals of 3.2 to 14.6 years after irradiation.71 However, the observed incidence was not significantly greater than the expected value of 3.5 cerebral strokes (P = 0.078) for this population. No definitive data support the notion that irradiated patients are at a higher risk of developing strokes. RADIATION-INDUCED CARCINOGENESIS To quantify the risk of second brain tumors after childhood cranial irradiation, a retrospective analysis of 305 patients treated for pituitary adenomas was performed. Four tumors, all gliomas, occurred within the radiation field after a latency period of 8 to 15 years. This group of patients had a 16 times greater relative risk of developing a malignant brain tumor than an age- and gender-adjusted population. The cumulative actuarial risk of a secondary glioma after radiation therapy was 1.7% at 10 years and 2.7% at 15 years. Meningiomas and sarcomas also have been reported to occur within the radiation field years after irradiation, including the low doses used for treatment of tinea capitis.72

RADIATION TECHNIQUES
GOALS OF RADIATION THERAPY AND CONVENTIONAL EXTERNAL BEAM TECHNIQUES The primary goal of irradiation of tumors in the pituitary-hypo-thalamic region is delivery of the prescribed tumor dose with minimal exposure to critical normal surrounding structures, particularly the visual and auditory systems, brainstem, and temporal lobes, and to excessively large volumes of hypothalamus. The modern radiotherapeutic process is based on precise delineation of the target and critical structures, which is achieved using contrast-enhanced, thinslice CT scanning. Further improvements in tumor localization are possible by correlation of CT and MRI data.73 The newer localization techniques can correct for the geometric distortion of MRI data sets and provide three-dimensional CT and magnetic resonance images and target volume correlation. Most patients with neuroendocrine tumors undergo the treatment planning process and the actual radiation treatments with a head immobilization device. The author's practice is to use a thermoplastic material that is molded to the patient's face and immobilized to a baseplate. The reproducibility of this device is within 2 to 3 mm on a daily basis during a conventional 5- to 6-week course of radiotherapy. The treatment planning process itself is typically carried out at a workstation with three-dimensional capability. The objective of the treatment planning process is to evaluate the relative dose distribution to the tumor and surrounding normal tissues of a variety of field arrangements. The optimal plan is selected and is further refined by beam modulation with devices such as beamattenuating wedges, compensators, or customized blocks. The use of two opposed lateral fields with low-energy photons is not recommended because of a slightly higher risk of temporal lobe radionecrosis. Satisfactory dose distribution generally is achieved with a three-field technique that uses two opposed lateral beams and a vertex or coronal field directed downward from the scalp to the pituitary fossa. This technique is less suitable for small tumors with minimal suprasellar extension, because the vertex field traverses a considerable amount of normal brain. Other acceptable techniques include a four-field arrangement, which typically requires wedges in at least two fields, and the use of an arc technique. The bicoronal 110-degree arc technique with a reversing wedge filter moving in each beam is commonly used. With this setup, the linear accelerator rotates from the level of the ears superiorly to the midline, creating a 110-degree coronal arc. A similar arc is then created on the contralateral side. Although these arc fields can exhibit substantial nonuniformity of dose distribution because of the curvature of the skull, this can be corrected by using wedge filters in the field. Similarly, an axialoblique bilateral arc arrangement with wedges can be used if the patient's head can be flexed sufficiently to exclude the eyes from the treatment field. The author's standard practice is to generate multiple plans for each patient, evaluate the isodose distributions, and select the most appropriate treatment technique. With such refined treatment planning, the dose gradient within the tumor is typically <5%, and the dose falloff to surrounding critical structures is very sharp. An example of such an isodose distribution is presented in Figure 22-2, which illustrates an axialoblique arc pair with 15-degree wedges.

FIGURE 22-2. Isodose distribution using the arc technique (first setting) reveals a sharp dose falloff from the center of the target.

The advent of three-dimensional treatment planning has introduced a new level of sophistication into the simulation process. Using 3-D techniques, three standard 2-D techniques (opposed-lateral two-field setup, 110-degree bilateral arcs, and a three-field technique) were compared with a single 330-degree rotational arc method (avoiding the eyes) and a four-field noncoplanar arc technique in an attempt to identify the optimal technique for minimizing exposure of normal tissues surrounding the pituitary.74 These observations have significant practice implications. The two-field opposed-lateral technique using 6 MV photons resulted in the highest dose to the temporal lobes and, therefore, should not be used routinely. The dose to the temporal lobes could easily be reduced with this technique, simply by using a higher energy beam, such as 18 MV. The three-field technique was superior to the opposed-lateral technique in reducing temporal lobe dose. Both the bilateral arc and single 330-degree rotation technique further decreased temporal lobe dose, in comparison with the fixed two- and three-field techniques. The four-field noncoplanar arc technique yielded the lowest temporal lobe dose but also resulted in the highest lens dose. After the treatment planning process, the exact field setup is determined during a simulation session. The patient is repositioned in the same configuration as used for the treatment-planning CT or MRI studies, and the central intersection point of the radiation fields, or isocenter, is determined using a coordinate transform system from the planning CT. The isocenter is verified with orthogonal radiographs. External reference land marksspecific points that had been selected at the time of the treatment-planning CT and that were visualized with barium strips or palatesare revisualized on the orthogonal films using radiopaque markers. These serve as external fiducial coordinate reference points that allow the transformation of CT location to orthogonal films. Because the thermoplastic mask provides rigid and reproducible immobilization, there is no need for skin marks; all necessary marks are placed directly on the mask. Daily alignment of the patient at the treatment machine is carried out using reference marks on the mask and a three-point laser system. At the first treatment session, verification portal images are obtained; in routine cases, weekly portal images are obtained. More complex setups using noncoplanar fields incorporate a portal verification imaging system. The author uses a total dose of 45 Gy in 1.8-Gy fractions for patients with pituitary adenoma. This dose is delivered in five fractions per week over a 5-week period. Although a clear dose-response relationship has not been established, most reports suggest that 40 to 45 Gy in 1.8- to 2.25-Gy fraction sizes produces high local control rates for most pituitary microadenomas.75 For craniopharyngioma and meningioma, the total dose is increased to 54 Gy. For low-grade glioma, the total dose is a function of the exact histologic type, the extent of residual disease, and the proximity of critical neural structures to the treatment field. INNOVATIVE RADIATION TECHNIQUES PARTICLE BEAM IRRADIATION OF THE PITUITARY GLAND In the early 1950s, studies were performed with stereotactic-charged particles in patients with metastatic breast carcinoma to produce pituitary hormone suppression. Since then, more than 3500 patients have been treated worldwide to reduce primary pituitary tumors and to suppress pituitary function for the management of diabetic retinopathy, breast cancer, prostate cancer, and other conditions. Most of this experience was accrued at four institutions: the University of California at Berkeley, Massachusetts General Hospital, the Institute for Theoretical and Experimental Physics in Moscow, and the Institute of Nuclear Physics in St. Petersburg. The results in the management of more than 2000 patients with pituitary tumors have been excellent and are summarized in the following sections. Treatment of Acromegaly. More than 200 patients with acromegaly have been treated with helium ion radiosurgery to a total dose of 30 to 50 Gy in four fractions delivered over 5 days. GH levels dropped more than two-thirds within the first year; by 4 years, the mean GH level had fallen below the normal value of 5 ng/mL. Although patients with microadenoma appeared to have an earlier and more dramatic decline, there was no difference by 4 years in GH levels between patients with microadenomas or macroadenomas.76 Reported remission rates in acromegalic patients by 4 years have been 80% and 90%.77,78 Treatment of Cushing Disease. In a cohort of 44 patients with ACTH-secreting adenomas treated with stereotactic radiotherapy, the mean basal cortisol levels and dexamethasone suppression test results returned to normal within 1 year of treatment and remained normal during long-term follow-up. Although early failures were reported with a regimen of alternateday irradiation, a total dose of 60 to 150 Gy in three to four fractions was successful in 40 of 42 patients. The mean ACTH level decreased from 90 pg/mL before treatment to 58 pg/mL 1 year after treatment. Plasma cortisol suppression by dexamethasone and plasma 11-deoxycortisol response to metyrapone normalized 1 year after treatment. Approximately two-thirds of 175 patients with Cushing disease treated with particle beam irradiation exhibited complete remission with restoration of clinical and laboratory parameters to normal.77 Another 224 patients had similar responses.78 Treatment of Prolactinomas. In a cohort of 29 patients with prolactinoma treated with particle beam irradiation, 19 of 20 patients had marked decreases in their prolactin levels within 1 year. Partial or total remission of prolactin secretion was observed in 85% of patients after treatment with particle beam proton radiosurgery to doses ranging between 100 and 120 Gy.79 RADIOSURGERY: PITUITARY ADENOMA AND CRANIOPHARYNGIOMA Radiosurgery, a technique in which a single large fraction of radiation is deposited within a small intracranial target, has been used in several varieties of pituitary-hypothalamic tumors.79a,79b The appeal of this technique as compared with fractionated techniques includes a single fraction of radiation (as opposed to several weeks of fractionated radiotherapy), the ability to minimize exposure of surrounding normal brain tissue, and the possibility of a more rapid endocrine decline. A major limitation to the use of stereotactic radiotherapy for the treatment of pituitary adenomas has been the low tolerance of the optic chiasm (probably <8 Gy) to radiation. Nevertheless, preliminary results using stereotactic radiosurgery have been reported. In an analysis of 77 patients, 42 (82%) of 51 patients with Cushing disease were described as cured.80 Dose and tumor subtype may be critical factors in achieving an early response in secreting tumors. No endocrinologic response was observed with greater than 6 months follow-up in 14 patients having prolactinoma, Cushing syndrome, or Nelson syndrome after 8 to 20 Gy maximum dose, whereas significant hormonal declines occurred within 6 months in 14 acromegalic patients treated similarly.81 In contrast, when higher dosesin the range of 25 to 60 Gy maximumwere used, endocrinologic normalization occurred within 2 years in 36 of 37 patients, and major radiographic shrinkage occurred in all 13 patients observed for more than two years.82 Similarly, in 27 patients treated with maximum doses of 25 to 75 Gy, 13 of 16 patients with secreting tumors experienced endocrinologic normalization or a >50% decrease within 10 months.83 At lower maximum doses of 10 to 27 Gy, there was a significant endocrinologic response within 12 months in both prolactinomas and GH-secreting tumors, with 11 of 13 prolactinomas becoming normal within a year. Unlike the experience with fractionated radiation, a major radiographic response was observed in 15 of 21 patients with serial imaging studies.84 In perhaps the best comparison between single-fraction radiosurgery and fractionated radiotherapy,85 the Mantel log-rank test was used to obtain the Kaplan-Meier estimate of time to GH normalization in 16 patients treated with radiosurgery (25 Gy minimum, 50 Gy maximum) and 40 patients treated with standard radiotherapy (40 Gy). The mean time to normalization was 1.4 years in the radiosurgery group as compared with 7.1 years in the fractionated radiotherapy group (P <0.0001). Radiosurgical treatment has been reported for 61 patients with meningiomas (42 patients), craniopharyngiomas (4), pituitary adenomas (5), gliomas (3), and miscellaneous lesions in and around the cavernous sinus (7).86 Long-term tumor control rates have not been analyzed, but an evaluation of toxicity revealed a 19% incidence of cranial neuropathies. Others have reported smaller series; the data are difficult to interpret because of short follow-up and patient heterogeneity. Clearly, these small experiences remain intriguing but need to be tested more rigorously through clinical trials. RADIOSURGERY: MENINGIOMA Several reports indicate >90% local control rates for meningioma treated with subtotal resection and radiosurgery. These reports include parasellar meningioma as a subset of other intracranial meningiomas. One of the few reports specifically evaluating petroclival meningioma described the experience with 62 patients, of whom 39 (63%) had previously undergone surgical resection and 7 (11%) had received fractionated external beam radiation therapy.87 The radiosurgery marginal dose ranged from 11 to 20 Gy. With a median follow-up of 37 months, the tumor volume decreased in 14 (23%), remained stable in 42 (68%), and increased in 5 patients (8%), for an overall control rate of 92%. Complications from radiosurgery were rare. Five patients (8%) developed new cranial nerve deficits within 24 months, which resolved completely in two patients within another 6 months. In an Austrian report, 97 patients with skull base meningiomas were treated with radiosurgery; 53 had partial removal or recurrent growth and 44 underwent radiosurgery as primary treatment. The mean peripheral tumor dose was 13.8 Gy (range: 725 Gy). In 78 patients, follow-up scans were available. Follow-up imaging revealed decreased volume in 31 cases (40%), stabilization in 44 cases (56%), and progression in 3 cases (4%). In 8 cases, marked central tumor necrosis was seen.88 When radiosurgery was performed on 50 patients with skull base meningiomas5 sellar, 26 cavernous sinus, and 12 petroclival, with a mean peripheral dose of 18.0 Gythe 1- and 2-year tumor control rates were 97% and 100%, respectively.89 In the largest linear-acceleratorbased meningioma radiosurgery report to date, 127 patients with 155 meningiomas were observed for 31 months. The median marginal dose was 15 Gy, and 82 of the tumors were skull based. Freedom from progression was observed in 107 patients (84.3%) at a median time of 22.9 months. Actuarial tumor control for the patients with benign meningiomas was 100%, 92.9%, 89.3%, 89.3%, and 89.3% at 1, 2, 3, 4, and 5 years, respectively. Six patients

(4.7%) had permanent complications attributable to the radiosurgery.90 Whereas the preliminary retrospective reports provide encouraging early data, many questions remain unanswered about the precise role of radiosurgery in these patients, especially in terms of patient selection, appropriate dose definition, and toxicity. As stated earlier, cranial neuropathies remain a concern. In a retrospective review, the endurance of the visual pathways and cranial nerves was evaluated after radiosurgery in 50 patients who had undergone radiosurgery for skull base tumors. With a mean follow-up of 40 months, the actuarial incidence of optic neuropathy was zero for patients who received <10 Gy, 26.7% for patients receiving <15 Gy, and 77.8% for those who received >15 Gy (P <0.0001). No neuropathy was seen in patients whose cavernous sinus cranial nerves received 5 to 30 Gy, suggesting that the visual pathways exhibit a much higher sensitivity to single-fraction radiation than do other cranial nerves.91 Yet another infrequent complication to consider is the development of radiation-induced edema. To evaluate the causative factors after radiosurgery, a retrospective study was performed in 34 patients. The minimum dose was 12 Gy, and the follow-up was 1 to 3 years. Edema developed preferentially in nonbasal tumors, especially those around the midline and sagittal sinus. In all but one case in which radiation-induced edema was observed, the marginal tumor dose was 18 Gy or more.92 INTRACAVITARY INSTILLATION OF RADIOISOTOPES Intracavitary instillation of isotopes such as yttrium-90, gold-198, and phosphorus-32 has been attempted for the treatment of cystic craniopharyngioma.93 This therapeutic approach is attractive, because ~60% of craniopharyngiomas present as a single large cyst, and early refilling is frequent after drainage. Intermittent aspiration of the cyst by stereotactic puncture or drainage through an Ommaya reservoir is frequently necessary. Some reports have suggested that prognosis is worse for tumors with a large cystic component, reflecting the extensive attachment of the cyst wall to adjacent vital structures. The major advantage of intracystic radioisotope instillation is the reduced radiation dose to adjacent normal structures such as the optic chiasm, hypothalamus, and surrounding brain. Intra-cavitary radiotherapy can be used in patients who have previously been irradiated. The maximum range of particles from phosphorus-32 in soft tissue is ~8 mm, with more than one-half the dose absorbed within the first 1.5 mm, allowing ablation of secretory cells within the cyst wall without significant exposure of the surrounding brain to radiation. An acute inflammatory reaction has been reported after the procedure, leading some investigators to recommend the routine prophylactic use of corticosteroids. This treatment approach may be useful in the management of recurrent cystic craniopharyngioma. Stereotactic instillation of radioisotopes into the cystic cavity of craniopharyngioma has been combined with radiosurgical treatment of the solid component.94 With follow-up ranging from 10 to 23 years, 31 (74%) of 42 patients are alive. Moreover, the patients remained socially well adapted and maintained a high rate of full-time work and a low rate of intercurrent disease. The authors advocate a less aggressive surgical approach to craniopharyngioma. Intracavitary yttrium-90 was used as primary treatment for 31 patients with craniopharyngioma, with an 84% overall survival rate for follow-up ranging from 2 to 80 months (median, 44 months). Visual acuity improved or stabilized in 42%, and visual fields improved in 48%.95 Local control, defined as cyst stabilization (6 patients), reduction (12), or complete resolution (10), was observed in 28 (90%) of 31 patients. These results further support the use of intracavitary radioisotope instillation in the initial management of craniopharyngioma. INTERSTITIAL IMPLANTATION The interstitial implantation of radioisotopes for the management of intracranial neoplasms has a long history, but technical difficulties have limited its widespread application in tumors of the pituitary-hypothalamic region. Permanent interstitial implantation of pituitary tumors was used in England in the 1960s. In the 1980s, the technique of transnasal transsphenoidal implantation of iodine-125 was introduced.96 Iodine-125 implantation has the advantage of rapid dose attenuation outside the target volume, with the ability to minimize normal tissue complications. It also has several hypothetical radiobiologic advantages. Because implanting tumors in this location is complex, its role in the management of recurrent pituitary-hypotha-lamic tumors is based only on anecdotal reports. CHAPTER REFERENCES
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. Kleinberg DL, Noel GL, Frantz AG. Galactorrhea: a study of 235 cases, including 48 with pituitary tumors. N Engl J Med 1977; 296:589. Gomez F, Reyes FI, Faiman C. Nonpuerperal galactorrhea and hyperpro-lactinemia: clinical findings, endocrine features and therapeutic responses in 56 cases. Am J Med 1977; 62:648. Antunes JL, Housepian EM, Frantz AG, et al. Prolactin-secreting pituitary tumors. Ann Neurol 1977; 2:148. Sheline GE. The role of conventional radiation therapy in the treatment of functional pituitary tumors. In: Linfoot JA, ed. Recent advances in the diagnosis and treatment of pituitary tumors. New York: Raven Press, 1979:289. Frantz AG, Cogen EH, Chang H, et al. Long-term evaluation of the results of transsphenoidal surgery in radiotherapy in patients with prolactinoma. In: Chrosignani PG, Rueben BL, eds. Endocrinology of human infertility: new aspects. New York: Grune & Stratton, 1981:161. Mehta AE, Reyes FL, Faiman C. Primary radiotherapy of prolactinomas. Am J Med 1987; 83:49. Rush SC, Newal J. Pituitary adenoma: the efficacy of radiotherapy as the sole treatment. Int J Radiat Oncol Biol Phys 1989; 17:165. Tran LM, Blount L, Horton D, et al. Radiation therapy of pituitary tumors results in 95 cases. Am J Clin Oncol 1991; 14:25. Tsagarakis S, Grossman A, Plowman PN, et al. Megavoltage pituitary irradiation in the management of prolactinomas: long-term follow-up. Clin Endocrinol (Oxf) 1991; 34:399. Rush S, Cooper PR. Symptom resolution, tumor control, and side effects following postoperative radiotherapy for pituitary macroadenomas. Int J Radiat Oncol Biol Phys 1997; 37:1031. Sheline GE, Goldberg MB, Feldman R. Pituitary irradiation for acromegaly. Radiology 1961; 76:70. McCollough WM, Markus RB, Rhoton AL, et al. Long-term follow-up of radiotherapy for pituitary adenoma: the absence of late recurrence after greater than or equal to 4500 cGy. Int J Radiat Oncol Biol Phys 1991; 21:607. Eastman RC, Gorden P, Roth J. Conventional super-voltage in radiation is an effective treatment for acromegaly. J Clin Endocrinol Metab 1979; 48:931. Ludecke DK, Lutz BS, Niedworok G. The choice of treatment after incomplete adenenomectomy in acromegaly: proton versus high voltage radiation. Acta Neurochir (Wien) 1989; 96:32. Peacey SR, Toogood AA, Shalet SM. Hypothalamic dysfunction in cured acromegaly is treatment modality dependent. J Clin Endocrinol Metab 1998; 83:1682. Trampe EA, Lundell G, Lax I, Werner S. External irradiational growth hormone in producing pituitary adenomas: prolactin as a marker of hypothalamic and pituitary effects. Int J Radiat Oncol Biol Phys 1991; 20:655. Kramer S. Indications for, and results of, treatment of pituitary tumors by external radiation. In: Kohler PO, Ross GT, eds. Diagnosis and treatment of pituitary tumors. New York: Excerpta Medica, 1973:217. Williams RA, Jacobs HS, Kurtz AB, et al. The treatment of acromegaly with special reference to transsphenoidal hypophysectomy. Q J Med 1975; 44:79. Pistenma DA, Goffinet DR, Bagshaw MA, et al. Treatment of acromegaly with megavoltage radiation therapy. Int J Radiat Oncol Biol Phys 1976; 1:885. Werner S, Trampe E, Palacios P, et al. Growth hormone producing pituitary adenomas with concomitant hypersecretion of prolactin are particularly sensitive to photon irradiation. Int J Radiat Oncol Biol Phys 1985; 11:1713. Grigsby PW, Stokes S, Marks JE, Simpson JR. Prognostic factors and results of radiotherapy alone in the management of pituitary adenomas. Int J Radiat Oncol Biol Phys 1988; 15:1103. Hughes MN, Llamas KJ, Yelland ME, et al. Pituitary adenomas: long-term results for radiotherapy alone and post-operative radiotherapy. Int J Radiat Oncol Biol Phys 1993; 27:1035. Plataniotis GA, Kouvaris JR, Vlahos L, et al. Radiation therapy alone for growth hormone-producing pituitary adenomas. Acta Oncol 1998; 37:97. Dohan FC, Raventos A, Boucot N, et al. Roentgen therapy in Cushing's syndrome without adrenocortical tumor. J Clin Endocrinol Metab 1957; 17:8. Heuschele R, Lampe I. Pituitary irradiation for Cushing's syndrome. Radiol Clin Biol 1967; 36:27. Orth DN, Liddle GW. Results of treatment in 108 patients with Cushing's syndrome. N Engl J Med 1971; 285:243. Edmonds MW, Simpton WJK, Meakin JW. External irradiation of the hypophysis for Cushing's disease. C M A J 1972; 107:860. Jennings AS, Liddle GW, Orth DN. Results of treating childhood Cushing's disease with pituitary irradiation. N Engl J Med 1977; 297:257. Howlett TA, Plowman PN, Wass JAH, et al. Megavoltage pituitary irradiation in the management of Cushing's disease and Nelson's syndrome: long-term follow-up. Clin Endocrinol (Oxf) 1989; 31:309. Sheline GE, Tyrrell B. Pituitary adenomas. In: Phillips TL, Pistenma DA, eds. Radiation oncology annual, 1983. New York: Raven Press, 1984:1. Flickinger JC, Nelson PB, Martinez AJ, et al. Radiotherapy of non-functional adenomas with a pituitary gland: results with long-term follow-up. Cancer 1989; 63:2409. Chang CH, Pool JL. The radiotherapy of pituitary chromophobe adenomas. Radiology 1967; 89:1005. Chun M, Masko GB, Hetelekidis S. Radiotherapy in the treatment of pituitary adenomas. Int J Radiat Oncol Biol Phys 1988; 15:305. Fisher BJ, Gaspar LE, Noon EB. Giant pituitary adenomas: a role of radiotherapy. Int J Radiat Oncol Biol Phys 1993; 25:677. Rush SC, Kupersmith MJ, Lerch I, et al. Neuro-ophthalmological assessment of vision before and after radiation therapy alone for pituitary mac-roadenomas. J Neurosurg 1990; 72:594. Jefferson G. Extrasellar extension of pituitary adenomas. Proc R Soc Med 1940; 33:433. Laws ER. Craniopharyngiomas in children and young adults. Prog Exp Tumor Res 1987; 30:335. Amacher AL. Craniopharyngioma: the controversy regarding radiotherapy. Childs Brain 1980; 6:57. Brada M, Thomas DGT. Craniopharyngioma revisited. Int J Radiat Oncol Biol Phys 1993; 27:471. Hoffman HJ, Silva DE, Humphries RP, et al. Aggressive surgical management of craniopharyngiomas in children. J Neurosurg 1992; 76:47. Hetelekidis S, Barnes PD, Tao ML, et al. Twenty year experience in childhood craniopharyngioma. Int J Radiat Oncol Biol Phys 1993; 27:189. Yasargil M, Curle M, Kis M, et al. Total removal of craniopharyngiomas: approaches and long-term results in 144 patients. J Neurosurg 1990; 73:3. Weiss M, Sutton L, Marcialo V, et al. The role of radiation therapy in the management of childhood craniopharyngioma. Int J Radiat Oncol Biol Phys 1989; 17:1313. Cavazzutti V, Fischer EG, Welch K, et al. Neurological and psychophysiological sequelae following different treatments of craniopharyngioma in children. J Neurosurg 1983; 59:409. Rajan B, Ashley S, Gorman C, et al. Craniopharyngioma: long-term results following limited surgery and radiotherapy. Radiother Oncol 1993; 26:1. Regine WF, Kramer S. Pediatric craniopharyngiomas: long-term results of combined treatment with surgery and radiation. Int J Radiat Oncol Biol Phys 1992; 24:611. Flickinger JC, Lunsford LD, Singer J, et al. Megavoltage external beam irradiation of craniopharyngiomas: analysis of tumor control and morbidity. Int J Radiat Oncol Biol Phys 1990; 19:117. Jose CC, Rajan B, Ashley S, et al. Radiotherapy for the treatment of recurrent craniopharyngioma. Clin Oncol (R Coll Radiol) 1992; 4:287. Nakagawa K, Aoki Y, Akanuma A, et al. Radiation therapy of intracranial germ cell tumors with radiosensitivity assessment. Radiat Med 1992; 10:55. Wara WW, Jenkin DT, Evans A. Tumors of the pineal and suprasellar region: Children's Cancer Study Group treatment results, 19601975. Cancer 1979; 43:698. Linstadt D, Wara W, Edwards M, et al. Radiotherapy of primary intracranial germinomas: the case against routine craniospinal irradiation. Int J Radiat Oncol Biol Phys 1988; 15:291. Hoffman HJ, Otsubo H, Hendrick EB, et al. Intracranial germ-cell tumors in children. J Neurosurg 1991; 74:545. Sebag-Montefiore DJ, Douek E, Kingston JE, Plowman PN. Intracranial germ cell tumors: I. experience with platinum based chemotherapy and implications for curative chemoradiotherapy. Clin Oncol (R Coll Radiol) 1992; 4:345. Schad LR, Gademann G, Knopp M, et al. Radiotherapy treatment planning of basal meningiomas: improved tumor localization by correlation of CT and MR imaging data. Radiother Oncol 1992; 25:56.

55. Glaholm J, Bloom HJ, Crow JH. The role of radiotherapy in the management of intracranial meningiomas: the Royal Marsden Hospital experience with 186 patients. Int J Radiat Oncol Biol Phys 1990; 18:755. 56. Miralbell R, Linggood RM, de la Monte S, et al. The role of radiotherapy in the treatment of subtotally resected benign meningiomas. J Neurooncol 1992; 13:157. 57. Capo H, Kupersmith MJ. Efficacy and complications of radiotherapy of anterior visual pathway tumors. Neurol Clin 1991; 9:179. 58. Kumar PP, Patil AA, Leibrock LG, et al. Brachytherapy: a viable alternative in the management of basal meningiomas. Neurosurgery 1991; 29:676. 59. Kondziolka D, Lunsford LD, Coffey RJ, Flickinger JC. Stereotactic radiosurgery of meningiomas. J Neurosurg 1991; 74:552. 60. Goodwin JW, Crowley J, Eyre HJ, et al. A phase II evaluation of tamoxifen in unresectable or refractory meningiomas: a Southwest Oncology Group study. J Neurooncol 1993; 15:75. 61. Rodriguez LA, Edwards MS, Levin VA. Management of hypothalamic gliomas in children: an analysis of 33 cases. Neurosurgery 1990; 26:242. 62. Moghrabi A, Friedman HS, Burger PC, et al. Carboplatin treatment of progressive optic pathway gliomas to delay radiotherapy. J Neurosurg 1993; 79:223. 63. Petronio J, Edwards MS, Prados M, et al. Management of chiasmal and hypothalamic gliomas of infancy and childhood with chemotherapy. J Neurosurg 1991; 74:701. 64. Fisher BJ, Gaspar LE, Noone B. Radiation therapy of pituitary adenoma: delayed sequelae. Radiology 1993; 187:843. 65. Colao A, Cerbone G, Cappabianca P, et al. Effect of surgery and radiotherapy on visual and endocrine function in nonfunctioning pituitary adenomas. J Endocrinol Invest 1998; 21:284. 66. Littley MD, Shalet SM, Beardwell CG, et al. Hypopituitarism following external radiotherapy for pituitary tumors in adults. Q J Med 1989; 70:145. 67. Millar JL, Spry NA, Lamb DS, Delahunt J. Blindness in patients after external beam irradiation for pituitary adenomas: two cases occurring after small daily fractional doses. Clin Oncol (R Coll Radiol) 1991; 13:291. 68. Urie MM, Fullerton B, Tatsuzaki H, et al. A dose-response analysis of injury to cranial nerves and/or nuclei following proton beam therapy. Int J Radiat Oncol Biol Phys 1992; 23:27. 68a. Meyers CA, Geara F, Wong PF, Morrison WH. Neurocognitive effects of therapeutic irradiation for base of skull tumors. Int J Radiat Oncol Biol Phys 2000; 46:51. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. Matsukado Y, Hiraki T. Multiple intracranial aneurysms following radiation therapy for pituitary adenoma. Case report. Neurol Med Chir (Tokyo) 1987; 27:224. Cantini R, Giorgetti W, Valleriani AM, et al. Radiation-induced cerebral lesions in childhood. Childs Nerv Syst 1989; 5:135. Flickinger JC, Nelson PB, Taylor FH, Robinson A. Incidence of cerebral infarction after radiotherapy for pituitary adenoma. Cancer 1989; 63:2404. Liwnicz BH, Berger TS, Liwnicz RG, Aron BS. Radiation-associated gliomas: a report of four cases and analysis of postirradiation tumors of the central nervous system. Neurosurgery 1985; 17:436. Schad LR, Gademann G, Knopp M, et al. Radiotherapy treatment planning of basal meningiomas: improved tumor localization by correlation of CT and MR imaging data. Radiother Oncol 1992; 25:56. Sohn JW, Dalzell JG, Suh JH, et al. Dose-volume histogram analysis of techniques for irradiating pituitary adenomas. Int J Radiat Oncol Biol Phys 1995; 32:831. Zaugg M, Adaman O, Pescia R, Landolt AM. External irradiation of macroinvasive pituitary adenomas with telecobalt: a retrospective study with long-term follow-up in patients irradiated with doses mostly between 40 45 Gy. Int J Radiat Oncol Biol Phys 1995; 32:671. Lawrence JH, Linfoot JA. Treatment of acromegaly, Cushing's disease and Nelson's syndrome. West J Med 1980; 133:197. Kjellberg RN, Shintani A, Frantz AG. Proton beam therapy for acromegaly. N Engl J Med 1968; 278:689. Minakova YEL, Kirpatovskaya LYE, Lyass FM. Proton beam therapy with pituitary tumors. Med Radiol (Mosk) 1983; 28:7. Konnov B, Melnikov L, Zargarova O. Narrow proton beam therapy for intracranial lesions. In: Heikkinen E, Kiviniitty K, eds. International workshop on proton and narrow photon beam therapy. Oulu, Finland: University of Oulu, 1989:48.

79a. Friedman WA, Foote KD. Linear accelerator radiosurgery in the management of brain tumors. Ann Med 2000; 32:64. 79b. Inove HK, Kohga H, Hirato M, et al. Pituitary adenomas treated with or without gamma knife surgery: experience in 122 cases. Sterotact Funct Neurosurg 1999; 72 (suppl 1):125. 80. Rahn T, Thoren M, Werner S. Stereotactic radiosurgery in pituitary adenomas. In: Faglia G, Peck-Peccoz P, Abrosi B, eds. Pituitary adenomas: new trends in basic and clinical research. New York: Excerpta Medica, 1991:303. 81. Voges J, Sturm V, Deuss U, et al. LINAC-radiosurgery (LINAC-RS) in pituitary adenomas: preliminary results. Acta Neurochir (Wien) 1996; 65:41. 82. Ikeda H, Jokura H, Yoshimoto T. Gamma knife radiosurgery for pituitary adenomas: usefulness of combined transsphenoidal and gamma knife radiosurgery for adenomas invading the cavernous sinus. Radiat Oncol Investig 1998; 6:26. 83. Park YG, Chang JW, Kim EY, et al. Gamma knife surgery in pituitary microadenomas. Yonsei Med J 1996;37:165. 84. Yoon SC, Suh TS, Jang HS, et al. Clinical results of 24 pituitary macroadenomas with linac-based stereotactic radiosurgery. Int J Radiat Oncol Biol Phys 1998; 41:849. 85. Landolt AM, Haller D, Lomax N, et al. Stereotactic radiosurgery for recurrent surgically treated acromegaly: comparison with fractionated radiotherapy. J Neurosurg 1998; 88:1002. 86. Tishler RB, Loeffler JS, Lunsford LD, et al. Tolerance of cranial nerves of the cavernous sinus to radiosurgery. Int J Radiat Oncol Biol Phys 1993; 27:215. 87. Subach BR, Lunsford LD, Kondziolka D, et al. Management of petroclival meningiomas by stereotactic radiosurgery. Neurosurgery 1998; 42:437. 88. Pendl G, Schrottner O, Eustacchio S, et al. Stereotactic radiosurgery of skull base meningiomas. Minim Invasive Neurosurg 1997; 40:87. 89. Nicolato A, Ferraresi P, Foroni R, et al. Gamma knife radiosurgery in skull base meningiomas: preliminary experience with 50 cases. Stereotact Funct Neurosurg 1996; 66:112. 90. Hakim R, Alexander E III, Loeffler JS, et al. Results of linear accelerator-based radiosurgery for intracranial meningiomas. Neurosurgery 1998; 42:446. 91. Leber KA, Bergloff J, Pendl G. Dose-response tolerance of the visual pathways and cranial nerves of the cavernous sinus to stereotactic radiosurgery. J Neurosurg 1998; 88:43. 92. Ganz JC, Schrottner O, Pendl G. Radiation-induced edema after gamma knife treatment for meningiomas. Stereotact Funct Neurosurg 1996; 66:129. 93. Berg E, van den Berge JH, Blaauw G, et al. Intra-cavitary brachytherapy of cystic craniopharyngiomas. J Neurosurg 1992; 77:545. 94. Saaf M, Thoren M, Bergstrand CG, et al. Treatment of craniopharyngiomasthe stereotactic approach in a ten to twenty-three years' perspective. II. Psychosocial situation and pituitary function. Acta Neurochir (Wien) 1989; 99:97. 95. Van den Berge JH, Blaauw G, Breeman WA, et al. Intracavitary brachytherapy of cystic craniopharyngiomas. J Neurosurg 1992; 77:545. 96. Kumar PP, Good RR, Leibrock LG, et al. High activity iodine-125 endocurietherapy for recurrent skull base tumors. Cancer 1988; 61:1518.

CHAPTER 23 NEUROSURGICAL MANAGEMENT OF PITUITARY-HYPOTHALAMIC NEOPLASMS Principles and Practice of Endocrinology and Metabolism

CHAPTER 23 NEUROSURGICAL MANAGEMENT OF PITUITARY-HYPOTHALAMIC NEOPLASMS

DAVID S. BASKIN Pituitary Tumors Classification Radiologic Evaluation Surgical Approaches Prolactin-Secreting Adenomas Clinical Manifestations Laboratory Evaluation Surgical Decision Making Vision Compromise in Prolactinoma Prolactin-Producing Macroadenomas Prolactin-Producing Microadenomas Surgical Results Acromegaly Endocrine Evaluation Surgical Decision Making Surgical Results Cushing Disease Endocrinologic Evaluation Surgical Decision Making Surgical Results Craniopharyngioma Preoperative Evaluation Surgical Results Endocrine-Inactive Pituitary Adenomas Clinical Presentation Surgical Results Conclusion Chapter References

Advances in endocrinology, neuroradiology, and microneuro-surgery have revolutionized the care of patients with pituitary and hypothalamic tumors. With the development of accurate radioimmunoassays and an understanding of the interactions between the hypothalamus and the pituitary, sophisticated endocrine testing can be performed to define the nature of the dysfunction precisely. Neuroradiologic techniques, particularly high-resolution magnetic resonance imaging (MRI) and magnetic resonance angiography (MRA), enable the clinician to localize many structural abnormalities to within 1 to 2 mm. Improvements in neurosurgical technology, including the use of the operating microscope, special microinstrumentation, and minimally invasive techniques such as endoscopy and high-resolution ultrasonography enable the surgeon to secure total removal of the lesion, with preservation of endocrine function and significant lowering of morbidity and mortality. The differential diagnosis of lesions in the sella and hypothalamic region is vast. Two common lesions are considered in this chapter: pituitary adenomas and craniopharyngiomas (see also Chap. 11, Chap. 17 and Chap. 18).

PITUITARY TUMORS
CLASSIFICATION The classic histologic designation of adenomas as chromophobic, eosinophilic, or basophilic on the basis of light microscopy is now obsolete and has been replaced by a system that classifies adenomas according to the hormones they secrete1 (see Chap. 11). The term chromophobe adenoma has been replaced by terminology that designates two types of nonsecreting tumors: the oncocytoma, which is thought to be a neoplasm with transformed epithelial cells without endocrine potential, and the null cell adenoma, which may have an as yet unidentified secretory product.2 Patients with tumors of the pituitary present for treatment either because of endocrinopathy or because of local mass effects. Classification according to the degree of sellar destruction (grade) and extrasellar extension (stage) can assist the physician in determining the surgical prognosis3 (Table 23-1).

TABLE 23-1. Anatomic (Radiographic and Operative) Classification of Pituitary Adenomas

RADIOLOGIC EVALUATION Before MRI scanning reached its present level of sophistication, the diagnosis of a pituitary tumor was based on thin-section polytomography of the sella, pneumoencephalography, computed tomography (CT) scanning, and bilateral carotid angiography. MRI scanning with and without gadolinium enhancement, with magnified images obtained in the axial, coronal, and sagittal planes, has made pneumoencephalography, CT scanning, and polytomography obsolete because it can more accurately diagnose pituitary adenomas in most cases (see Chap. 20). An MRI diagnosis of adenoma is based on the fact that both normal pituitary gland tissue and the adjacent cavernous sinuses enhance at a different time after the administration of intravenous paramagnetic contrast material than does an adenoma, which often enhances poorly, and usually late.4 The timing of imaging is important. Most radiologists recommend imaging as soon as possible after the administration of the contrast. Other criteria suggestive of adenoma include a convex upper border of the pituitary gland (seen, however, in 2% of normal glands); increased height of the gland (>7 mm); lateral deviation of the pituitary stalk; and a focal area of altered attenuation relative to the normal gland, on either contrast or noncontrast studies. Usually, precise delineation between tumor and other important structures in the area can be accomplished (Fig. 23-1).

FIGURE 23-1. Magnetic resonance image without contrast (A) and with contrast (B) with direct coronal scans for a young woman with a pituitary macroprolactinoma. Note the low-density areas in the lesion on both scans. The surrounding tissue enhances after the administration of intravenous contrast, correlating well with the surgical finding of normal glandular tissue, rather than tumor, surrounding the low-density center. The tumor was precisely confined to the low-density area.

High-resolution MRI scanning with and without gadolinium enhancement is recommended for assessment of all suspected pituitary and hypothalamic lesions. Carotid angiography is reserved for those patients in whom an intrasellar aneurysm is suspected after high-quality MRI scans, including MRI angiography, have been performed. The author has not found it necessary to perform carotid angiography on the last 200 patients treated, now that high-resolution MRI angiography is available. CT scanning is no longer routinely performed in these patients. It can occasionally be helpful for patients in whom areas of hemorrhage, bony erosion, or calcification are being assessed or for patients with unusual bony sphenoid sinus anatomy, particularly those who have undergone previous transsphenoidal exploration. SURGICAL APPROACHES Surgical approaches to pituitary adenomas have been described in detail.5,6 The specific morphologic configuration of the neoplasm, rather than the endocrinologic syndrome, determines the choice between the transcranial and the transsphenoidal approach. The transsphenoidal approach is the technique of choice for tumors that occupy the sella, whether or not any extension has occurred into the sphenoid sinus (Fig. 23-2, Fig. 23-3 and Fig. 23-4). Tumors with vertical suprasellar extension without significant lateral extension are also well treated with this approach. The advantage of the transsphenoidal approach is that it usually allows selective excision of tumor with preservation of remaining normal pituitary gland, even when most of the sella is occupied by tumor. The approach involves no retraction of the cortex whatsoever, as opposed to the transcranial approach, in which, at times, considerable brain retraction may be necessary. In addition, the morbidity of the procedure is exceedingly low, and it is well tolerated even by patients who would be considered unacceptable surgical candidates for craniotomy. In experienced hands, only 1% of patients with pituitary tumors require a transcranial operation.

FIGURE 23-2. Bony and cartilaginous anatomy of the base of the skull, sphenoid sinus, and nasal areas. Note that the posterior wall of the sphenoid sinus is the floor of the sella turcica, making the transsphenoidal route uniquely suited for the removal of sellar lesions. (From Tindall GT, Barrow DL. Disorders of the pituitary. St. Louis: CV Mosby, 1986.)

FIGURE 23-3. Diagrammatic summary of the transsphenoidal surgical approach. A, A linear incision is made from canine fossa to canine fossa. The entire surgical field lies within this incision. This provides a cosmetically favorable result because the scar is never visible externally. The nasal mucosa is dissected away from the cartilaginous and bony nasal septum. B, A speculum is then placed to expose the sphenoid sinus, and the posterior wall of the sphenoid sinus (the floor of the sella) is removed. Note the adenoma in the anterior aspect of the gland, where most of these lesions are located. This procedure is performed with the aid of an operating microscope, using a C-arm fluoroscope, which facilitates visualization of the area. C, Removal of the microadenoma. Using magnification and microdissection technique, the adenoma can be removed from the gland, sparing the normal gland tissue. D, Reconstitution of the sella after removal of the tumor or the gland. Fat is placed in the sella to prevent downward migration and herniation of the optic chiasm. A piece of nasal bone is then used to reconstitute the sellar floor, which later calcifies and forms new bone. (A, C, and D from Hardy J. Transsphenoidal operations on the pituitary. Codman and Shurtless, Inc. A division of Johnson and Johnson. 1983; B from Tindall GT, Barrow DL. Disorders of the pituitary. St. Louis, CV Mosby, 1986.)

FIGURE 23-4. Technical details relating to removal of a microadenoma. A, Basic principles of tumor removal. Note the development of a plane between the tumor located laterally and the normal gland located medially. B, Dissection of the pseudocapsule, or the fibrous tissue surrounding the outer aspects of the tumor, which ensures a clean removal. C, Two important principles of microsurgical removal are illustrated. The first is to carefully inspect, or at least palpate, all hidden pockets. One can see tumor hidden in the anterior corner of the sella, which is extracted by the inserted curette. In addition, extracting the surrounding tissue for biopsy to

confirm that it contains only normal gland and, therefore, that all tumor has been removed is usually advisable. (From Hardy J. Transsphenoidal approach to the sella. In: Wilson CB, ed. Neurosurgical procedures: personal approaches to classical operations. Philadelphia: Williams & Wilkins, 1992:30.)

An advance in surgical technology is the introduction of endoscopy (Fig. 23-5) into neurosurgical procedures.6a In selected cases, a transsphenoidal resection can now be performed via one nostril using the endoscope, so that the degree of invasiveness of the operation is even further reduced.7 Moreover, the endoscope now permits the surgeon to look around the corner at angles that were not possible using conventional microsurgical techniques, thereby improving surgical outcome. Patients do not have nasal packing placed, which thus avoids both the numbness in the upper teeth that persists for at least several months and the nasal congestion and stuffiness that often occurs for several weeks after a standard transsphenoidal operation. In many cases, the patient can be discharged as early as the first postoperative day. A few patients have been given surgery on an outpatient basis with excellent outcomes.

FIGURE 23-5. The use of the endoscope for transsphenoidal surgery. A, Diagram demonstrating an endoscopic endonasal approach to a sellar tumor. No septal, alar, or gingival incision is used, and no speculum or retractor is necessary. B, The endoscope is held in the surgeon's hand until an opening is made into the sphenoid sinus. C, The endoscope is mounted on a special holder, which provides the surgeon with a steady video image and frees both hands to use surgical instruments simultaneously. (From Jho HD, Carrau RL. Endoscopic endonasal transsphenoidal surgery: experience with 50 patients. J Neurosurg 1997; 87:44.)

Another advance is the refinement of intraoperative technology to permit the use of ultrasonography in the operating room to localize small tumors that might be otherwise difficult to visualize.8 This is particularly useful for patients with Cushing disease for whom imaging studies have been normal or equivocal. The ultrasonic probe developed for this procedure is pencil thin and can therefore be used in the very small area that constitutes the surgical field.

PROLACTIN-SECRETING ADENOMAS
Prolactin-secreting adenomas comprise the largest group of pituitary tumors. The behavior and relatively benign clinical manifestations of small prolactinomas distinguish them from the tumors that produce Cushing disease and acromegaly, two distinct endocrinopathies that are usually life-threatening. Whereas the clinical necessity of treating patients with either Cushing disease or acromegaly is clear, the indications for immediate treatment of patients harboring a small prolactin-secreting adenoma are less so (see Chap. 13 and Chap. 21). CLINICAL MANIFESTATIONS In 1954, Forbes and colleagues9 first reported that pituitary adenomas could produce amenorrhea and galactorrhea. However, only recently have these tumors been recognized as a frequent cause of secondary amenorrhea and galactorrhea. Among the women in one surgical series5 (Table 23-2), 80% presented with secondary amenorrhea or galactorrhea, 10% with primary amenorrhea, and 10% with either oligomenorrhea and galactorrhea, secondary amenorrhea without galactorrhea, or secondary amenorrhea only. Among men, prolactinomas usually remain undetected until a large tumor produces either significant panhypopituitarism or compression and invasion of the parasellar structures. In the previously mentioned series,5 only seven men had symptomatic hyperprolactinemia without abnormalities of additional pituitary hormones; either thyroid, adrenal, or gonadotropic function, or a combination of the three, was usually impaired as well. Many of these patients experience impotence early in the course of their disease, but this problem often does not lead to an investigation of the prolactin level. The author can recall seeing in his practice a 35-year-old man who received electroshock therapy for 10 years as treatment for his impotence; the patient presented with a prolactin level of >1000 ng/mL.

TABLE 23-2. Clinical Features of 121 Patients with Prolactinoma Treated by Transsphenoidal Surgery*

Hyperprolactinemia secondary to a pituitary adenoma has extragonadal manifestations. Recent rapid weight gain is a frequent complaint of hyperprolactinemic women and occurs with a frequency that suggests a correlation. Correction of hyperprolactinemia, either by surgery or by medical therapy, has been followed by impressive weight loss in many cases, despite no apparent change in dietary habits. Equally impressive is the incidence of emotional lability, which is often dramatically reversed after the correction of hyperprolactinemia. Studies demonstrate that the estrogen deficiency secondary to hyperprolactinemia causes bone demineralization, sometimes producing secondary complications.10 LABORATORY EVALUATION The first step in the evaluation of a patient with suspected hyperprolactinemia is to obtain a fasting serum prolactin level. The administration of thyrotropin-releasing hormone (TRH) does not consistently distinguish between functional hyperpro-lactinemia and actual prolactinoma11 (see Chap. 13). In men whose basal prolactin values exceed 100 ng/mL, establishing a prolactinoma as the cause of the hyperprolactinemia is not difficult. In women, hyperprolactinemia (>200 ng/mL) almost invariably indicates a tumor. Caution must be exercised, because prolactin levels as high as 662 ng/mL have been observed to occur in nonsecreting tumors, presumably due to pronounced pressure on the pituitary stalk, which inhibits the transport of prolactin inhibitory factor to the pituitary gland.12 In the author's experience, the diagnosis of prolactinoma in a patient with basal prolactin levels <200 ng/mL requires radiographic identification of an intrasellar lesion. Even with unequivocal radiographic demonstration of such a lesion, transsphenoidal exploration occasionally reveals either a diffusely enlarged anterior lobe (pituitary hyperplasia) or a nonneoplastic intrasellar cyst, most often involving the pars intermedia. In such cases, the presence of the radiographic abnormality is usually not related to the elevated prolactin level. Patients must, therefore, be cautioned about this possibility, particularly now that the resolution of MRI imaging has become so good.

SURGICAL DECISION MAKING As experience with the use of the dopamine agonists (i.e., bromocriptine and cabergoline) accumulates, the indications for surgery in patients with prolactin-secreting tumors are changing. Currently, most patients who are referred for surgery have large and invasive tumors. Although a role exists for surgery in these patients, they are rarely cured by an operation alone, and usually require supplemental drug therapy, radiation therapy, or both. The young and healthy patient with a microadenoma has an excellent chance to obtain a long-term remission with surgery; however, most are treated with medication and are never referred to a surgeon. In the author's opinion, surgery has a role in the treatment of these patients, as long as they are properly counseled regarding the options of medical and surgical therapy, and the pros and cons of either approach. Surgery provides the possibility of cure, obviating the need to take medication for years or for the patient's lifetime. Patients should be informed, however, that medical therapy is a perfectly acceptable treatment modality that in most cases provides adequate control of the tumor without the need for subsequent surgical intervention. VISION COMPROMISE IN PROLACTINOMA Many would agree that medical therapy (bromocriptine or cabergoline) is indicated for the management of prolactinomas that are of sufficient size and invasiveness to produce serum prolactin levels of >600 ng/mL. (For such lesions, the cure rate with surgery, even in the most experienced hands, is only 10%.10) Medical therapy is usually effective for long-term control, with normalization of serum prolactin levels. On the other hand, the presence of vision loss complicates the management of such patients. This is because of the concern that such therapy either may fail or may take too long to produce sufficient reduction of tumor volume to relieve the compression of the vision system, which could result in further irreversible vision damage during the trial of medical treatment. Because vision compromise can reverse after surgical treatment, even when the compression is longstanding, some believe that vision compromise is not a contraindication to a trial of medical therapy. Substantial tumor shrinkage can occur within days, leading to improved vision.13,14 Others,5,6 including the author, believe that surgical intervention is indicated in these patients if they are otherwise healthy, because a risk exists of further permanent vision damage with the less rapid decompression provided by medical therapy. If medical therapy is selected for patients with vision compromise, careful monitoring of vision is essential.13,14 PROLACTIN-PRODUCING MACROADENOMAS For macroadenomas, operative removal is recommended if visual compromise is present and if the patient's overall medical condition justifies the small risks of surgical intervention. For macroadenomas without compression of the optic apparatus, surgery may be considered if the tumor is <2 cm and the prolactin level is <600 ng/mL, because surgical cure and a low complication rate are reasonable expectations under these conditions.13 For tumors >2 cm and prolactin levels >600 ng/mL, medical therapy is recommended initially. A desire for pregnancy complicates matters, because pregnant patients with macroadenomas may develop complications related to accelerated tumor growth. Because of this concern, such patients may be candidates for surgery, even if no visual compromise is present. PROLACTIN-PRODUCING MICROADENOMAS For microadenomas, opinions differ among surgeons concerning initial treatment. Some believe that all patients should be treated medically, except for those who develop unacceptable side effects to medical therapy or whose tumors are resistant to dopamine agonists.15 Others believe that surgery should be the initial treatment for healthy patients with microadenomas and that bromocriptine, cabergoline, and irradiation should be reserved for cases of surgical failure or for those in whom the risk of surgery is high.5,6 Surgery does not always cure prolactin-producing microadenomas; in particular, tumors with higher prolactin levels have a greater likelihood of surgical treatment failure. Serious surgical complications can occur, although, in experienced hands, the complication rate is <1%. Still, medical management is not a perfect solution. Treatment with medication is a lifelong commitment. At effective doses, a significant number of patients have unpleasant side effects, although they may be able to tolerate these symptoms and continue to take the drug. Use of the dopamine agonist cabergoline has reduced some of the undesirable side effects associated with medical therapy. Some tumors are relatively resistant to bromocriptine, as indicated by either inadequate lowering of the prolactin levels or continued growth of the tumor. In such cases, the tumor may have a mixed population of cells, some of which are not responsive to bromocriptine (pseudoprolactinoma).16 Well-documented cases have been seen of progressive enlargement of pituitary tumors, pituitary apoplexy, and even metastases of adenomas during bromocriptine therapy.17,18 and 19 If nonsurgical management is chosen, a progressive elevation in prolactin levels approaching 200 ng/mL, or an enlargement of the tumor detected on MRI, represents a secondary indication for surgical intervention. Some believe that long-term treatment with medical therapy reduces the success of subsequent surgery; this effect has been reported to be due to the perivascular fibrosis that occurs when the tumor shrinks during drug treatment.20 SURGICAL RESULTS The surgical treatment of prolactin-secreting microadenomas results in a remission rate of 72% to 97%.21,22,23,24 and 25 Remissions were experienced by 88% of patients who had preoperative prolactin levels of <100 ng/mL, whereas 50% of patients with prolactin levels of >100 ng/mL experienced remissions.25 In long-term studies, remissions were observed in 93% of patients with microadenomas as well as in 88% of patients with tumors (regardless of size) confined to the sella or extending only moderately into the suprasellar region.5 By contrast, remissions were seen in only 37% of female patients and in 15% of male patients who had significant extrasellar extension of tumor and were treated with surgery alone. In a series of patients receiving surgery for microprolactinoma who were followed for a mean of 70 months, only a 3% recurrence was observed.22 Moreover, beyond 10 years after surgery, the cost of medical therapy exceeds that of surgery. The value of surgery in treating macroadenomas without chiasmal compression in patients who have no special circumstances such as a desire for or presence of pregnancy, medication intolerance, or poor response to dopamine agonists is questionable. Surgery combined with dopamine agonists has been compared to use of dopamine agonists alone in the long-term treatment of macroprolactinomas.23 Clearly, for this group of patients surgery did not improve the ultimate outcome for patients who responded to dopamine agonists. The situation is quite different for microadenomas. A study of surgery for microadenomas at an experienced center indicates a remission rate of 84%, even in patients followed for a median of 15.6 years. A 97% remission rate was seen in patients who received surgery using an improved microsurgical technique at a mean follow-up of 3.2 years.24 Even when the procedure is performed by experienced surgeons, prolactinomas have a higher recurrence rate than that of any other pituitary adenoma after primary successful surgery. Recurrence rates vary from 14% to 50% in published series26,27 (Table 23-3). In part, the high recurrence rate may arise from the cause of the prolactin-secreting tumor, which is related to hypothalamic dysregulation. This dysregulation is unaltered after total resection of neoplasm, which leads to the growth of a new tumor rather than regrowth of residual tumor cells. In the author's experience, serum prolactin levels measured in the early postoperative period have prognostic value. A value of <5 ng/mL virtually assures a cure, whereas a prolactin level of >15 ng/mL is presumptive evidence of residual tumor and implies a probable regrowth.

TABLE 23-3. Recurrences after Transsphenoidal Surgery for Prolactinomas: A Survey of the Recent Literature

ACROMEGALY
By the time a patient with a growth hormone (GH)secreting pituitary adenoma seeks the advice of a neurosurgeon, the diagnosis is usually obvious, and the need for

therapy is urgent. Left untreated, this type of tumor produces metabolic disturbances with deleterious consequences and is associated with increased mortality. The surgical options for the treatment of GH-secreting pituitary tumors include craniotomy or transsphenoidal surgery,28,29 either for removal of the entire pituitary gland or for selective removal of adenomatous tissue. Irradiation, first used for the treatment of acromegaly in 1909, still provides an alternative to surgical intervention (see Chap. 22). ENDOCRINE EVALUATION The evaluation of a patient with suspected acromegaly begins with a determination of the fasting GH level along with a complete assessment of anterior pituitary function (see Chap. 17). Additional testing should include the measurement of serum insulin-like growth factor-I (IGF-I, somatomedin C). Failure of suppression of GH levels to <3 ng/mL after oral administration of glucose can be helpful diagnostically, especially in patients with normal fasting GH or IGF-I levels who appear to be clinically symptomatic. A paradoxical rise in GH levels is seen after the intravenous administration of TRH in half of the patients. This test can be included in the preoperative evaluation. SURGICAL DECISION MAKING First-line therapy for patients with acromegaly should be transsphenoidal surgery, especially if a microadenoma or small macroadenoma without invasion of the cavernous sinus is present.30 Medical therapy as a primary treatment modality should be reserved for those in whom surgery cannot be safely undertaken. SURGICAL RESULTS In the postoperative evaluation of the patient with acromegaly, the goals are a fasting GH level of <5 ng/mL and normalization of the IGF-I level. The restoration of normal GH dynamics, with a glucose-induced suppression of the serum GH level to <3 ng/mL after an oral glucose load, is desirable; however, in the author's experience, the absence of such normalization does not necessarily imply residual tumor or recurrence. Importantly, normalization of GH and IGF-I levels may first occur as long as 3 to 6 months after surgery. Thus, if these values are elevated in the immediate postoperative period, a failure of surgery is not inevitable. In experienced hands, transsphenoidal surgery is a highly effective treatment for the GH-secreting pituitary adenoma. Among 103 patients who underwent surgery during a 17-year period, the overall success rate of surgery alone was 82.4% in previously untreated patients, compared with 75% in patients with prior bromocriptine therapy and 63.6% in patients with prior radiation therapy.31 Tumor stage was a strong predictor of outcome, with greater suprasellar extension associated with less favorable results. The tumor grade and the preoperative GH and IGF-I levels were also significant predictors of surgical success, with higher grade and higher preoperative serum hormonal levels associated with a less favorable outcome. In a study of 175 patients treated by transsphenoidal surgery during a period of 11 years, probability of remission after 5 years was 62.7%.32 Moreover, the tumor size and the preoperative basal GH level correlated with outcome; a larger tumor size and a higher GH level were associated with less favorable results. Findings from the University of California in San Francisco28,29 demonstrated long-term remission rates ranging from 80% to 90%, depending on the exact biochemical criteria used. Suprasellar extension, higher preoperative GH levels, and higher grade and stage also correlated with a less favorable outcome. Similar results were reported in another study,33 which noted a 78% cure rate for enclosed adenomas in 100 acromegalic patients. This finding has been supported by a long-term study34 in which patients were followed for at least 5 years. The cure rate for microadenomas was 88%. Similar results have been reported elsewhere with a long-term remission of 83%.35 Some authors have concluded that early postoperative GH dynamics do not clearly correlate with either long-term cure or recurrence. Thus, recurrence of acromegaly has been reported in three patients with normal postoperative GH dynamics.36 Others have found that, although not all patients who have normal early postoperative GH levels but exhibit abnormal dynamics eventually experience relapse, such patients are nevertheless more likely to relapse than those who experience a complete biochemical remission. Clearly the recurrence rate for acromegalic patients after selective adenomectomy is much lower than the rate that has been reported for prolactinomas. Although the reasons are not entirely clear, the suggestion has been made that most GH-secreting adenomas arise de novo in the pituitary gland and are not the result of underlying hypothalamic dysregulation. Table 23-4 summarizes reported recurrence rates seen after transsphenoidal surgery for acromegaly. Reoperation in these patients is successful in 88% if the MRI scan demonstrates a focal tumor recurrence within the sella.37 Note that, for the endocrine-active tumors, the results of surgery are best in these circumstances.

TABLE 23-4. Recurrences after Transsphenoidal Surgery for Acromegaly: A Survey of the Recent Literature

Although radiation therapy is highly effective as adjunctive treatment in patients with GH-secreting adenomas, its value as a primary therapeutic modality is less clear. A comparison of response rates indicates that the results obtained with transsphenoidal surgery are superior to those obtained with irradiation alone. Moreover, the response to transsphenoidal surgery is immediate, whereas a period of several years after the completion of radiation therapy is required before remission is obtained. During this time, the metabolic abnormalities associated with acromegaly continue unabated. Bromocriptine may be a useful adjunct in the treatment of some GH-secreting tumors, but except in rare cases, it is not effective as a primary mode of therapy19 (see Chap. 21). The results of the International Multicenter Acromegaly Study Group using the somatostatin analog octreotide38 do not compare favorably with the results of surgery. Although 88% of patients did show reduction in serum GH and IGF-I levels associated with clinical improvement, strict criteria used to define biochemical cure after surgery were met in less than half of the patients. Furthermore, unlike in the treatment of prolactinomas with bromocriptine, only 44% of patients treated with octreotide showed a decrease of tumor size, and only by 20%. Surgical results have been compared with octreotide treatment results in a series of patients who were first treated medically and then underwent surgery.39 The effects on the levels of high-affinity GHbinding protein, GH, IGF-I, and insulin-like growth factorbinding protein-3 were determined. These markers did not normalize with octreotide therapy but did normalize after surgical removal of the tumor. Therefore, although somatostatin analogs hold promise for the future, they are presently indicated only as adjunctive treatment for patients for whom surgery, radiation therapy, or both have failed (see Chap. 169). Of the therapeutic modalities available for the treatment of GH-secreting pituitary adenomas, only transsphenoidal surgery offers the unique combination of a low morbidity rate, a low incidence of postoperative hypopituitarism, and immediate remission.

CUSHING DISEASE
Cushing disease has fascinated neurosurgeons ever since Harvey Cushing described the clinical syndrome and predicted that one cause of the disorder was a small pituitary adenoma.40 ENDOCRINOLOGIC EVALUATION No endocrinologic findings yield a 100% accuracy rate in the diagnosis of Cushing disease. The evaluation of patients suspected of having this condition should start with verification of sustained hypercortisolism. Morning and evening levels of serum cortisol as well as a 24-hour urine collection with measurement of urinary free cortisol levels should be determined. The diagnosis can generally be established by the demonstration of nonsuppressibility of corticosteroids in serum or urine after administration of low-dose dexamethasone but at least 50% suppressibility after administration of high-dose (8 mg) dexamethasone. A lack of suppression with high-dose dexamethasone can be seen in cases of macroadenoma and thus does not necessarily rule out an adrenocorticotropic hormone (ACTH)secreting

adenoma.41 An elevated serum ACTH level in the presence of hypercortisolism provides additional supportive evidence. Percutaneous transfemoral selective inferior petrosal venous sampling for ACTH before and after the administration of corticotropin-releasing factor has also been of help in equivocal cases.42 If the diagnosis is uncertain and a paraneoplastic (ectopic) source of ACTH secretion is suggested, selective venous sampling to determine a cephalic plasma ACTH gradient has proved to be diagnostically definitive.5 When the samples have been obtained from the inferior petrosal sinus, no misleading results have been found. Sampling of another pituitary hormone along with ACTH is helpful to reduce confusion about results from various sites (see Chap. 74, Chap. 75 and Chap. 219). Such sampling can be taken directly from the cavernous sinus, using a superselective microcatheterization technique.43 Experience indicates a high likelihood of cure after surgery if the serum cortisol values, measured at 8:00 a.m. and at least 24 hours after discontinuation of oral hydrocortisone therapy, either are unmeasurable or, if they are within the normal range, are shown to suppress fully during low-dose dexamethasone testing.44,45 and 46 Some investigators have reported that Cushing disease has not recurred in any patient with undetectable serum cortisol in the early postoperative period,44 although a normal cortisol level during this time does not necessarily imply failure. Measurement of ACTH levels during the operation does not accurately predict complete tumor resection.47 SURGICAL DECISION MAKING Like patients with acromegaly, patients with Cushing disease require swift and efficacious therapeutic intervention (i.e., selective pituitary adenectomy). In adults, a total hypophysectomy is recommended if no abnormal tissue is identified during the transsphenoidal exploration. If no tumor is found in a child or in a young adult, postoperative radiation therapy or medical therapy is recommended instead of hypophysectomy. The rationale for this approach is that panhypopituitarism excludes procreation (except in rare cases after gonadotropin therapy) and that, even with synthetic GH, severe growth limitations occur in children. SURGICAL RESULTS Once the pituitary has been established as the cause of a patient's hypercortisolism, the treatment of choice is transsphenoidal exploration. Patients who have negative imaging studies require careful systematic exploration of the intrasellar contents by an experienced pituitary surgeon.48 Microadenomas secreting ACTH can often be small and located deep within the gland, so that the tumor is not immediately visualized when the dura is opened. Experience with systematic exploration, including microsurgical dissection and the ability to visualize areas that are technically difficult to reach, is essential to achieve optimal surgical results.49 Although false localizing results that were based on sampling of the inferior petrosal sinus have been described, the author always first explores the side with the higher gradient in patients with negative imaging studies and no obvious tumor. In most cases, such exploration has been fruitful. The results of surgical exploration in experienced hands are good, with remission rates ranging from 75% to 92%.50,51,52 and 53 In a large surgical series, an 86% long-term remission rate was achieved for adenomas confined to the intrasellar compartment, whereas only 46% of patients were successfully treated once the tumors had extrasellar extension.50 Similarly, the recurrence rate was only 1.2% in those with intrasellar adenomas, whereas it was 15% in those with extrasellar adenomas. A high grade and stage of the tumor were associated with a low initial remission and a high recurrence rate. The results of this series are summarized in Table 23-5.

TABLE 23-5. Surgical Outcome in Relation to Tumor Size, Extension, and Histologic Confirmation in 216 Patients with Cushing Disease

Unlike with other endocrine-active tumors, in 10% to 15% of patients with a preoperative diagnosis of Cushing disease, no tumor is found, even in the most experienced surgical hands. Some of these patients later are discovered to have paraneoplastic (ectopic) tumors, whereas others may be discovered to have a more obvious pituitary adenoma. Another small subgroup has an entity known as pituitary hyperplasia (hyperplasia of ACTH-secreting cells without obvious neoplasm). Such patients are candidates for either total hypophysectomy or, later, surgical (or medical) adrenalectomy. Pituitary irradiation also may be considered if the pituitary exploration does not reveal an adenoma. The final decision must be individualized. Young patients who desire pregnancy probably are better served by adrenalectomy, despite the risk of subsequent Nelson syndrome. Patients in whom the ability to conceive is not an issue and for whom total endocrine replacement does not produce an excessive burden are better served by total hypophysectomy. In patients with ACTH-secreting pituitary adenomas, the recurrence rate ranges from 3% to 15% for the most experienced surgeons (Table 23-6). As with prolactin- and GH-secreting tumors, the most significant factor contributing to surgical failure is suprasellar or extrasellar extension of tumor. The absolute level of serum ACTH has not proved to be as reliable a prognostic indicator as are the prolactin or GH levels for the previously discussed adenomas. This may be attributed to variability in the values obtained with the radioimmunoassay for ACTH. Considering the morbidity and mortality associated with this disease, surgical treatment is a reasonable option.

TABLE 23-6. Recurrences after Transsphenoidal Surgery for Cushing Disease

CRANIOPHARYNGIOMA
Craniopharyngiomas have challenged neurologists and neurosurgeons since these tumors were first described in 1903.54 With advances in microsurgical technique, the thought was that the total removal of these lesions would be achieved more readily. Although the results of such surgery have improved, most surgeons continue to find this procedure extremely difficult and fraught with morbidity and mortality. Subtotal removal often is the best that can be achieved, even with advanced microsurgical technology and superlative surgical skill.55 Considerable debate exists about the surgical philosophy regarding craniopharyngioma: radical total removal in most cases versus a procedure that is more conservative and attempts to remove as much tumor as is possible, with the surgeon terminating the procedure when dense adherence to important neural or vascular structures is encountered. The problem is further complicated by the fact that adult and childhood craniopharyngiomas probably are different in this regard. That is, total removal is more easily achieved in children because the tumor has been present for a shorter period of time and, therefore, produces less inflammatory and gliotic reaction. In some cases total removal can be achieved with minimal morbidity, whereas in others such removal is not possible without damaging surrounding vital structures. Some of these neoplasms are densely adherent to the optic nerves, hypothalamus, and internal carotid artery;

in these cases, an aggressive approach can produce a devastating outcome, even in the most experienced hands. Radiation therapy plays an important adjunctive role in the treatment of this lesion. The MRI scan has been beneficial in resolving the controversy concerning which patients should undergo radiation therapy. Those patients with residual or recurrent tumor usually can be distinguished from those who have a clean surgical removal. Surgical intervention plus radiation therapy yield superior results compared with surgery alone in patients with obvious recurrent or residual tumor. This may be because the cells producing the secretions that form the cyst of this tumor have their secretory character altered by therapy. Therefore, radiation therapy is mandatory for optimal treatment of most incompletely removed tumors (see Chap. 22). Unlike pituitary adenomas, many craniopharyngiomas require a transcranial approach because most of these lesions are located in the suprasellar region and often do not even extend into the sella. Furthermore, many of these tumors have a dense gliotic scar around them, which makes delivery of the suprasellar component of the tumor into the sella difficult without producing unacceptable damage to the suprasellar structures, including the hypothalamus and/or the optic apparatus. A transsphenoidal approach can be performed in a minority of patients, usually those in whom the tumor is confined largely to the sella. In these patients, the outcome is good, with less morbidity and frequent total tumor removal.56,57 PREOPERATIVE EVALUATION A complete assessment of the hypothalamic-pituitary axis is required for patients with a craniopharyngioma because often significant hypofunction of one or more of the pituitary hormones is present. Unlike with the pituitary adenomas, diabetes insipidus may be a presenting symptom for which patients should be evaluated. Detailed assessment of visual function is critical. These tumors tend to be regionalized to the optic nerve, chiasm, and tract, and significant defects in visual function are common58 (see Chap. 19). SURGICAL RESULTS In 61 children with craniopharyngiomas who were treated with radiation therapy, surgery, or both, the addition of radiation therapy to surgical treatment or the use of radiation therapy alone provided a higher likelihood of regression of disease than did surgery alone.59 In a study of 27 patients, total resection with no recurrence was achieved in 10 patients.60 A 7% recurrence rate was reported among 144 patients, of whom 90% had complete resections.61 Most of 173 patients treated at the Royal Marsden Hospital underwent incomplete or partial excision followed by treatment with radiation therapy.62 The 10- and 20-year progression-free survival rates were 83% and 79%, respectively, supporting the concept that subtotal removal and radiation therapy can achieve excellent long-term tumor control and patient survival with low morbidity. In a series of 168 consecutive patients undergoing surgery between 1983 and 1997, total tumor removal was achieved in 45%.57 Even with this modest attempt at total removal, mortality was 1.1% in primary cases and 10.5% in cases of tumor recurrence. No patient who underwent transsphenoidal surgery died. The rate of recurrence-free survival after total removal was 86% at 5 years and 81% at 10 years. The authors concluded that total tumor removal was preferable, but this was not always possible without hazardous intraoperative manipulation. In a retrospective analysis63 evaluating the neurologic and behavioral sequelae in 32 patients who underwent surgery, total excision was associated with greater immediate mortality and morbidity as well as a higher incidence of subsequent behavioral disability. In a series55 of 74 patients with craniopharyngiomas treated during a 15-year period, an attempt was made to achieve a total removal. When this could not be achieved, the patients were treated with postoperative irradiation. Total removal was achieved in seven patients, six of whom have had no recurrence. However, of the entire series, 91% of treated patients were in long-term remission and fully functional. Although total removal of a craniopharyngioma is possible in some patients in the adult population, this cannot always be achieved. A subtotal removal of tumor followed by radiation therapy has proved to be a satisfactory approach and has led to remission in most patients. With continued careful endocrine replacement therapy and monitoring, these patients can resume a normal functional life.

ENDOCRINE-INACTIVE PITUITARY ADENOMAS

Endocrine-inactive pituitary adenomas produce clinical effects only as a direct consequence of their growth. Resulting clinical syndromes reflect either damage to the anterior pituitary gland and subsequent hypopituitarism, or compression of suprasellar and parasellar structures with development of associated neurologic deficits. Although these tumors have the necessary genetic information and cytoplasmic organelles to manufacture hormones and subunits, the products are secreted either in subclinical concentrations or in bioinactive forms; thus, these neoplasms are termed endocrine-inactive. When immunocytochemical and electron microscopic characteristics are taken into consideration, these tumors can be divided into null cell adenomas, oncocytomas, and glycoprotein-secreting adenomas.64 CLINICAL PRESENTATION Patients with endocrine-inactive pituitary adenomas generally present with visual impairment or hypopituitarism. Headaches are common and can be severe, relating to dural pressure. A variety of visual field defects can be seen, the most common being bitemporal hemianopia or bitemporal superior quadrantanopia. Extraocular dysfunction, related to lateral growth into the cavernous sinus, can occur, with consequent compression of the oculomotor, trochlear, or abducens nerves. Dementia and other symptoms associated with hydrocephalus also can occur when these tumors compress the third ventricular outflow, producing obstructive hydrocephalus. These neoplasms can occasionally occur in children and adolescents. Presenting symptoms differ and usually include pubertal and growth delay and/or primary amenorrhea.65 The tumor characteristics in this group of patients do not differ from those in the adult population. SURGICAL RESULTS Surgical resection of these adenomas does not differ appreciably from the surgery for endocrine-active tumors, except that the tumors are more likely to be macroadenomas, with higher grades and stages. This results in a lower cure rate with surgery alone, which is related to size and extrasellar extension rather than to any innate biologic difference. The goals of surgery include establishment of a diagnosis, decompression of adjacent structures, and an attempt at total surgical removal. Seventy-five percent to 80% of patients with visual compromise experience significant recovery of vision after transsphenoidal resection. Cure rates are low, ranging from 20% to 30%.64 Although some recommend postoperative radiation therapy in all patients because of the low cure rate, others recommend careful serial MRI scanning, with radiation therapy reserved until obvious evidence is found of tumor regrowth. In support of the former approach is the reported 21% recurrence rate in a series of 126 patients with endocrine-inactive adenomas. All of these tumors except one were macroadenomas.66 Glycoprotein-secreting adenomas that may be associated with clinical symptoms (those producing thyroid-stimulating hormone, luteinizing hormone, and follicle-stimulating hormone) are discussed in Chapter 15, Chapter 16 and Chapter 42.

CONCLUSION
In contrast to the results obtained in the past, the outcome of neurosurgical intervention for pituitary and hypothalamic neoplasms has improved greatly, due to both improvements in surgical technique and the use of the surgical microscope along with specialized microinstrumentation. In most cases, the disabling and often life-threatening consequences of disease in this area can be improved significantly by surgical intervention. For pituitary adenomas, the development of the transsphenoidal technique and its subsequent refinement using endoscopy (see Fig. 23-5) and high-resolution ultrasonography has converted an operation with major morbidity and mortality into one with few complications and very good results. CHAPTER REFERENCES
1. 2. 3. 4. 5. 6. Kovacs K, Horvath E, Ezrin E. Pituitary adenomas. Pathol Annu 1977; 2:341. Kovacs K, Horvath E. Atlas of tumor pathology, series 2, fasc 21. Washington: Armed Forces Institute of Pathology, 1986. Wilson CB. Neurosurgical management of large and invasive pituitary tumors. In: Tindall GT, Collins WF, eds. Clinical management of pituitary disorders. New York: Raven Press, 1979:335. Davis PC, Gokhale KA, Joseph GJ, et al. Pituitary adenoma: correlation of half-dose gadolinium-enhanced MR imaging with surgical findings in 26 patients. Radiology 1991; 180:779. Wilson CB. A decade of pituitary microsurgery: the Herbert Olivecrona Lecture. J Neurosurg 1984; 61:814. Wilson CB. Role of surgery in the management of pituitary tumors. Neurosurg Clin North Am 1990; 1:139.

6a. Badie B, Nguyen P, Preston JK. Endoscopic-guided direct endonasal approach for pituitary surgery. Surg Neurol 2000; 53:168.

7. Jho HD, Carrau R. Endoscopic endonasal transsphenoidal surgery: experience with 50 patients. J Neurosurg 1997; 87:44. 8. Watson JC, Shawker TH, Nieman LK, et al. Localization of pituitary adenomas by using intraoperative ultrasound in patients with Cushing's disease and no demonstrable pituitary tumor on magnetic resonance imaging. J Neurosurg 1998; 89:927. 9. Forbes AP, Henneman PH, Griswold GC, Albright F. Syndrome characterized by galactorrhea, amenorrhea and low urinary FSH: comparison with acromegaly and normal lactation. J Clin Endocrinol Metab 1954; 14:265. 10. Klibanski A, Neer RM, Beitins IZ, et al. Decreased bone density in hyperprolactinemic women. N Engl J Med 1980; 303:1511. 11. Chang RJ, Keye WR Jr, Monroe SE, Jaffe RB. Prolactin-secreting pituitary adenomas in women. IV. Pituitary function in amenorrhea associated with normal or abnormal serum prolactin and sellar polytomography. J Clin Endocrinol Metab 1980; 51:830. 12. Albuquerque FC, Hinton DR, Weiss MH. Excessively high prolactin level in a patient with a nonprolactin-secreting adenoma. J Neurosurg 1998; 89:1043. 13. Weiss M. Pituitary tumors: an endocrinological and neurosurgical challenge. Clin Neurosurg 1992; 39:114. 14. Mbanya JCN, Mendelow AD, Crawford PJ, et al. Rapid resolution of visual abnormalities with medical therapy alone in patients with large prolactinomas. Br J Neurosurg 1993; 7:519. 15. Colao A, Annunziato L, Lombardi G. Treatment of prolactinomas. Ann Med 1998; 30:452. 16. Smith MV, Laws ER Jr. Magnetic resonance imaging measurements of pituitary stalk compression and deviation in patients with nonprolactin-secreting intrasellar and parasellar tumors: lack of correlation with serum prolactin levels. Neurosurgery 1994; 34:834. 17. Alhajje A, Lambert M, Crabbe J. Pituitary apoplexy in an acromegalic patient during bromocriptine therapy: case report. J Neurosurg 1985; 63:288. 18. Martin NA, Hales M, Wilson CB. Cerebellar metastasis from a prolactinoma during treatment with bromocriptine: case report. J Neurosurg 1981; 55:615. 19. Yamaji T, Ishibashi M, Kosaka K, et al. Pituitary apoplexy in acromegaly during bromocriptine therapy. Acta Endocrinol (Copenh) 1981; 98:171. 20. Landolt AM, Keller PJ, Froesch ER, Mueller J. Bromocriptine: does it jeopardize the result of later surgery for prolactinomas? Lancet 1982; 2 (8299):657. 21. Hardy J. Transsphenoidal microsurgery of prolactinomas. In: Black PM, Zervas NT, Ridgway EC, Martin JB, eds. Secretory tumors of the pituitary gland. New York: Raven Press, 1984:73. 22. Turner HE, Adams CB, Wass JA. Transsphenoidal surgery for microprolactinoma: an acceptable alternative to dopamine agonists? Eur J Endocrinol 1999; 140:43. 23. Hofle G, Gasser R, Mohsenipour I, et al. Surgery combined with dopamine agonists versus dopamine agonists alone in long-term treatment of macroprolactinoma: a retrospective study. Exp Clin Endocrinol Diabetes 1998; 106:211. 24. Tyrrell JB, Lamborn KR, Hannegan LT, et al. Transsphenoidal microsurgical therapy of prolactinomas: initial outcomes and long term results. Neurosurgery 1999; 44:254. 25. Randall RV, Laws ER, Abboud CF, et al. Transsphenoidal microsurgical treatment of prolactin-producing pituitary adenomas. Mayo Clin Proc 1983; 58:108. 26. Buchfelder M, Fahlbusch R, Schott W, Honegger J. Long-term follow-up results in hormonally active pituitary adenomas after primary successful transsphenoidal surgery. Acta Neurochirurgica 1991; (Suppl) 53:72. 27. Serri O, Rasio E, Beauregard H, et al. Recurrence of hyperprolactinemia after selective transsphenoidal adenomectomy in women with prolactinoma. N Engl J Med 1983; 309:280. 28. Baskin DS, Boggan JE, Wilson CB. Transsphenoidal microsurgical removal of growth hormone-secreting pituitary adenomas: a review of 137 cases. J Neurosurg 1982; 56:634. 29. Ross DA, Wilson CB. Results of transsphenoidal microsurgery for growth hormonesecreting pituitary adenoma in a series of 214 patients. J Neurosurg 1988; 68:854. 30. Melmed S, Jackson I, Kleinberg D. Current treatment guidelines for acromegaly. J Clin Endocrinol Metab 1998; 83:2646. 31. Tindall GT, Oyesiku NM, Watts NB, et al. Transsphenoidal adenomectomy for growth hormone-secreting pituitary adenomas in acromegaly: outcome analysis and determinants of failure. J Neurosurg 1993; 78:205. 32. Davis DH, Laws ER Jr, Ilstrup DM, et al. Results of surgical treatment for growth hormonesecreting pituitary adenomas. J Neurosurg 1993; 73:70. 33. Grisoli F, Leclercq T, Jaquet P, et al. Transsphenoidal surgery for acromegaly: long-term results in 100 patients. Surg Neurol 1985; 23:513. 34. Freda PU, Wardlaw SL, Post KD. Long term endocrinological follow-up evaluation in 115 patients who underwent transsphenoidal surgery for acromegaly. J Neurosurg 1998; 89:353. 35. Swearingen B, Barker FG II, Katznelson L, et al. Long-term mortality after transsphenoidal surgery and adjunctive therapy for acromegaly. J Clin Endocrinol Metab 1998; 83:3419. 36. Serri O, Somma M, Comtois R, et al. Acromegaly: biochemical assessment of cure after long term follow-up of transsphenoidal selective adenomectomy. J Clin Endocrinol Metab 1985; 61:1185. 37. Abe T, Ludecke DK. Recent results of secondary transnasal surgery for residual or recurring acromegaly. Neurosurgery 1998; 42:1013. 38. Vance ML, Harris AG. Long-term treatment of 189 acromegalic patients with the somatostatin analog octreotide. Arch Intern Med 1991; 151:1573. 39. Hernandez I, Soderlund D, Espinosa-de-los Monteros L, et al. Differential effects of octreotide treatment and transsphenoidal surgery on growth hormone-binding protein levels in patients with acromegaly. Neurosurg 1999; 90:647. 40. Cushing H. The basophil adenomas of the pituitary body and their clinical manifestations (pituitary basophilism). Bull Johns Hopkins Hosp 1932; 50:137. 41. Katznelson L, Bogan J, Thob J, et al. Biochemical assessment of Cushing's disease in patients with corticotroph macroadenomas. J Clin Endocrinol Metab 1998; 83:1619. 42. Oldfield EH, Chrousos GP, Schulte HM, et al. Preoperative lateralization of ACTH-secreting pituitary microadenomas by bilateral and simultaneous inferior petrosal venous sinus sampling. N Engl J Med 1985; 312:100. 43. Teramoto A, Yoshida Y, Sanno N, Nemoto S. Cavernous sinus sampling in patients with adrenocorticotrophic hormone dependent Cushing's syndrome with emphasis on inter and intracavernous adrenocorticotrophic hormone gradients. J Neurosurg 1998; 89:762. 44. Trainer PJ, Lawrie HS, Verheist J, et al. Transsphenoidal resection in Cushing's disease: undetectable serum cortisol as the definition of successful treatment. Clin Endocrinol 1993; 38:73. 45. McCance DR, Gordon DS, Fannin TF, et al. Assessment of endocrine function after transsphenoidal surgery for Cushing's disease. Clin Endocrinol 1993; 38:79. 46. Partington M, Davis D, Laws E, Scheithauer B. Pituitary adenomas in childhood and adolescence: results of transsphenoidal surgery. J Neurosurg 1994; 80:209. 47. Graham KE, Samuels MH, Raff H, Barnwell SL, Cook DM, et al. Intraoperative adrenocorticotropin levels during transsphenoidal surgery for Cushing's disease do not predict cure. J Clin Endocrinol Metab 1997; 82:1776. 48. Chandler WF, Schteingart DE, Lloyd RV, McKeever PE. Surgical treatment of Cushing's disease. J Neurosurg 1987; 66:204. 49. Chandler WF. Surgical treatment of pituitary adenomas. In: Lloyd RV, ed. Surgical pathology of the pituitary gland. Philadelphia: WB Saunders, 1993:235. 50. Mampalam TJ, Tyrrell JB, Wilson CB. Transsphenoidal microsurgery for Cushing's disease: a report of 216 cases. Ann Intern Med 1988; 109:487. 51. Tindall GT, Herring CJ, Clark RV, et al. Cushing's disease: results of transsphenoidal microsurgery with emphasis on surgical failures. J Neurosurg 1990; 72:363. 52. Fahlbusch R, Buchfelder M, Muller OA. Transsphenoidal surgery for Cushing's disease. J Royal Soc Med 1986; 79:262. 53. Nakane T, Kuwayama A, Watanabe M, et al. Long term results of transsphenoidal adenomectomy in patients with Cushing's disease. Neurosurgery 1987; 31:218. 54. Erdheim J. Zur normalen und pathologischen histologie der glandula thyreoidea parathyreoidea, und hypophysis. Beitr Path Anat Allge 1903; 33:158. 55. Baskin DS, Wilson CB. Surgical management of craniopharyngiomas: a review of 74 cases. J Neurosurg 1986; 65:22. 56. Abe T, Ludecke DK. Transnasal surgery for infradiaphragmatic craniopharyngiomas in pediatric patients. Neurosurgery 1999; 44:957. 57. Fahlbusch R, Honegger J, Paulus W, et al. Surgical treatment of craniopharyngiomas: experience with 168 patients. J Neurosurg 1999; 90:237. 58. Repka MX, Miller NR, Miller M. Visual outcome after surgical removal of craniopharyngiomas. Ophthalmology 1989; 96:195. 59. Hetelekidis S, Barnes PD, Tao ML, et al. 20-year experience in childhood craniopharyngioma. Int J Radiat Oncol Biophys 1993; 27:189. 60. Tomita T, McLone DG. Radical resections of childhood craniopharyngiomas. Pediatr Neurosurg 1993; 19:6. 61. Yasargil MG, Curcic M, Kis M, et al. Total removal of craniopharyngiomas: approaches and long-term results in 144 patients. J Neurosurg 1990; 73:3. 62. Rajan B, Ashley S, Gorman C, et al. Craniopharyngioma: long-term results following limited surgery and radiotherapy. Radiother Oncol 1993; 26:1. 63. Colangelo M, Ambrosio A, Ambrosio C. Neurological and behavioral sequelae following different approaches to craniopharyngioma: long term follow-up review and therapeutic guidelines. Childs Nerv Sys 1990; 6:379. 64. Wilson CB. Endocrine-inactive pituitary adenomas. Clin Neurosurg 1992; 38:10. 65. Abe T, Ludecke DK, Saeger W. Clinically nonsecreting pituitary adenomas in childhood and adolescence. Neurosurgery 1998; 42:744. 66. Comtois R, Beauregard H, Somma M, et al. The clinical and endocrine outcome to transsphenoidal microsurgery of nonsecreting pituitary adenoma. Cancer 1991; 68:860. 67. Faglia G, Monondo P, Travaglini P, Giovanelli MA. Influence of previous bromocriptine therapy on surgery for microprolactinoma. Lancet 1983; 1:133. 68. Rodrnan EF, Molitch ME, Post KD, et al. Long-term follow-up of transsphenoidal selective adenomectomy for prolactinoma. JAMA 1984; 252:921. 69. Schlechte JA, Sherman BM, Chapter FK, VanGilder J. Long term follow-up of women with surgically treated prolactin-secreting pituitary tumors. J Clin Endocrinol Metab 1986; 62:1296. 70. Maira G, Anile C, Do Mannis L, Barbanno A. Prolactin-secreting adenomas: surgical results and long-term follow-up. Neurosurgery 1989; 24:738. 71. Ciccarelli E, Ghigo E, Miola C, et al. Long-term follow-up of cured' prolactinoma patients after successful adenomectomy. Clin Endocrinol (Oxf) 1990; 32:583. 72. Roelfsema F, van Dutken H, Frolich M. Long-term results of transsphenoidal pituitary microsurgery in 60 acromegalic patients. Clin Endocrinol (Oxf) 1985; 23:555. 73. Arafah BM, Rosenzweig JL, Fenstermaker R, et al. Value of growth hormone dynamics and somatomedin C (insulin-like growth factor 1) levels in predicting the long-term benefit after transsphenoidal surgery for acromegaly. J Lab Clin Med 1987; 109:346W. 74. Artia N, Mon S, Saitoh Y, et al. Transsphenoidal surgery for acromegalyfollow-up results. In: Landolt AM, et al, eds. Progress in pituitary adenoma research. London: Pergammon Press, 1988:265. 75. Landolt AM, Illig R, Zapf J. Surgical treatment of acromegaly. In: Lamberts SWJ, ed. Sandostatin in the treatment of acromegaly. Berlin Heidelberg New York: Springer, 1988:22. 76. Lose M, Qeckter R, Schopohi J, et al. Evaluation of selective transsphenoidal adenomectomy by endocrinological testing and somatomedin-C measurement in acromegaly. J Neurosurg 1989; 70:561. 77. Buchfelder M, Brockmeier S, Fahlbusch R, et al. Recurrence following transsphenoidal surgery for acromegaly. Horm Res 1991; 35:113. 78. Abosch A, Tyrrell JB, Lamborn KR, et al. Transsphenoidal microsurgery for growth hormone-secreting pituitary adenomas: initial outcome and long-term results. J Clin Endocrinol Metab 1998; 83:3411. 79. Hardy J. Presidential address: XVII Canadian Congress of Neurological Sciences. Cushing's disease: 50 years later. Can J Neurol Sci 1982; 9:375. 80. Derome PJ, Delalande O, Bisot A, et al. Short and long term results after transsphenoidal surgery for Cushing's disease: incidence of recurrences. In: Landolt AM, et al, eds. Progress in pituitary adenoma research. London: Pergammon Press, 1988:375. 81. Gudhaume B, Bertagna X, Thomsen M, et al. Transsphenoidal pituitary surgery for the treatment of Cushing's disease: results in 64 patients and long term follow-up studies. J. Clin Endocrinol Metab 1988; 66:1056. 82. Pieters GF, Hermus AR, Meijer E, et al. Predictive factors for initial cure and relapse rate after pituitary surgery for Cushing's disease. J Clin Endocrinol Metab 1989; 69:1122. 83. Knappe UJ, Ludecke DK. Transnasal microsurgery in children and adolescents with Cushing's disease. Neurosurgery 1996; 39:484. 84. Bochicchio D, Lose M, Buchfelder M. Factors influencing the immediate end late outcome of Cushing's disease treated by transsphenoidal surgery: a retrospective study by the European Cushing's Disease Survey Group. J Clin Endocrinol Metab 1995; 80:3114.

CHAPTER 24 PITUITARY TUMORS: OVERVIEW OF THERAPEUTIC OPTIONS Principles and Practice of Endocrinology and Metabolism

CHAPTER 24 PITUITARY TUMORS: OVERVIEW OF THERAPEUTIC OPTIONS

PHILIPPE CHANSON Clinical Classification General Therapeutic Principles of Surgery and Irradiation Surgery Irradiation Therapy Therapeutic Options and Results by Type of Pituitary Adenoma Growth HormoneSecreting Pituitary Adenomas Prolactin-Secreting Tumors Adrenocorticotropic HormoneSecreting Adenomas Thyroid-Stimulating HormoneSecreting Adenomas Clinically Nonfunctioning Pituitary Adenomas Prognosis of Pituitary Adenomas and Therapeutic Perspectives Prognosis Conclusions and Perspectives Concerning Treatment of Pituitary Adenomas Chapter References

Chapter 21, Chapter 22 and Chapter 23 discuss, respectively, medical, radiologic, and surgical therapies of pituitary tumors. This chapter reviews the current literature and draws on the author's own experience to systematically discuss and compare the different therapeutic options, their applicabilities, and their limitations. Whenever possible, an attempt has been made to categorize, in a statistical manner, response rates and recurrences. Pituitary tumors are relatively common neoplasms of the adenohypophyseal cells; they represent ~15% of all intracranial tumors.1,2 The prevalence of clinically recognizable pituitary adenomas is ~200 cases per million, and new cases number 15 per million per year.3 However, asymptomatic adenomas (mostly microadenomas) are found in 6% to 20% (mean, 11%) of presumably normal pituitary glands. This has been demonstrated by autopsy studies4,5 and by systematic magnetic resonance imaging (MRI) studies.6 Thus, pituitary adenomaswhether detected by clinical manifestations or by pituitary imaging that is performed for unrelated reasons (i.e., incidentalomas)are becoming more frequently diagnosed. Therapeutic recommendations are based on previously reported studies that have evaluated results of various treatment strategies. However, few large comparative trials have been undertaken. Thus, therapeutic guidelines often are somewhat subjective. An accurate classification of pituitary adenomas depends on immunocytochemical studies that are performed on tumoral tissue removed at surgery.2 However, not all pituitary tumors require surgery. Thus, for practicality, and for simplification of therapeutic decisions, a clinical classification is used, which is based on the presence or absence of a hypersecretion syndrome. Generally, the clinical classification and the histopathologic classification concur (Fig. 24-1).

FIGURE 24-1. Distribution of the various types of pituitary adenomas in a surgical series. (TSH, thyroid-stimulating hormone; GH, growth hormone; ACTH, adrenocorticotropic hormone.) (From P. Derome, Hpital Foch, Suresnes, France.)

CLINICAL CLASSIFICATION
1. Growth hormone (GH)-secreting adenomas are responsible for acromegaly. On immunocytochemical analysis, these adenomas may be purely GH-secreting or mixed (secreting GH and prolactin [PRL] or GH and a subunit). 2. PRL-secreting adenomas are responsible for hyperprolactinemia. Immunocytochemically, they can be composed of pure PRL-secreting adenomatous cells or can be mixed (secreting PRL and GH, or PRL and a subunit). 3. Corticotropin-secreting adenomas are associated with Cushing syndrome (when Cushing syndrome is due to a pituitary microadenoma, the condition is termed Cushing disease). On immunocytochemical analysis, these ade nomatous cells stain for adrenocorticotropic hormone (ACTH), either alone or in association with other peptides. 4. Thyrotropin (thyroid-stimulating hormone, or TSH)secreting pituitary adenomas are responsible for thyrotoxicosis, which occurs in conjunction with an inappropriate secretion of TSH. The TSH immunostaining may be isolated or may be associated with a subunit, GH, or PRL. 5. Clinically nonfunctioning pituitary adenoma (NFPA) is the preferred term for designating a pituitary adenoma that is not hormonally active (i.e., not associated with one of the above clinical syndromes); patients with NFPAs do not have acromegaly, a hyperprolactinemic syndrome, Cushing syndrome, or hyperthyroidism. Usually, the tumor has been found because of a tumor-mass effect or has been an incidental discovery. The term clinically NFPA is preferred to chromophobe adenoma (the latter term was used before the routine use of immunocytochemical staining); it also is preferable to the term nonsecreting adenoma, because immunocytochemical analysis has demonstrated that the majority of these lesions are, indeed, able to secrete one or more of the pituitary hormones. However, they may not be associated with increased plasma levels of the hormone (silent adenoma), or they may be associated with increased levels of a hormone that often does not produce a recognizable clinical hypersecretion syndrome (e.g., FSH, LH, or free a subunit). Whatever an immunocytochemical study of such surgically removed NFPA may reveal, the therapeutic management is the same. The classification above is used in this chapter.

GENERAL THERAPEUTIC PRINCIPLES OF SURGERY AND IRRADIATION

The management of pituitary tumors commences with a careful definition of the location and extent of the lesion and the endocrinologic abnormalities. The principles of therapy are based primarily on these factors and their clinical sequelae. Any direct effect of the mass (e.g., vision impairment) must be addressed, and any endocrinologic dysfunction must be corrected. Certain hormonal deficiencies, particularly those of cortisol or thyroid hormone, should be corrected immediately. Possible complications of therapy, most prominently hypopituitarism or tumor recurrence, can occur many years (up to 30 years) after therapy and must be carefully considered. The choice of treatment modality is determined by several factors: (a) the need for immediate relief of a mass effect, (b) the need to relieve an endocrine abnormality, (c) the potential for obtaining long-term control, and (d) the character and frequency of possible associated morbidity. Surgery and radiotherapy are the two main available tools for radical treatment of all types of pituitary adenomas. Their common technical characteristics and side effects are presented in this section. More specific medical modalities, applicable to certain types of pituitary tumors, are detailed in the later section, which deals with therapeutic options according to the type of pituitary adenoma. SURGERY

The purpose of surgery in the management of pituitary adenomas includes, according to the circumstances, the histologic confirmation of the diagnosis; the correction of tumor-mass effect; the complete excision of a microadenoma and, if possible, of a macroadenoma; and the reduction of the tumor bulk of an invasive adenoma. Whatever the immunocytochemical type, the greater the extent to which an adenoma is small, is enclosed, and is noninvasive, the better are the results of surgical removal. In most cases, either a transsphenoidal or a subfrontal transcranial approach is used7,8 and 9 (see Chap. 23). TRANSSPHENOIDAL TECHNIQUE The transsphenoidal approach, via a sublabial incision, is now preferred by most pituitary neurosurgeons for the vast majority (>95%) of patients with pituitary tumors. This technique allows entry into the facial portion of the sphenoid sinus, through which access is gained to the pituitary fossa. Binocular surgical microscopy is coupled with fluoroscopic monitoring to obtain direct visualization of the surgical field. The transsphenoidal approach offers the capability of tumor destruction by resec tion, by coagulation, or by freezing. This technique is indicated for removal of a tumor that is confined to the sella turcica, removal of a tumor associated with cerebrospinal fluid (CSF) rhinorrhea or of a pituitary apoplexy, removal of a tumor with sphenoidal extension, removal of a tumor with only modest suprasellar extension, or removal of a tumor which has fluid characteristics that allow the suprasellar part of the adenoma to flow down by gravity into the sellar cavity. Contraindications to use of the transsphenoidal approach include the presence of dumbbell-shaped tumors with constriction at the diaphragma sellae, massive suprasellar tumors, lateral suprasellar extension of a tumor, and an incompletely pneumatized sphenoid bone. Morbidities include transient diabetes insipidus and, rarely, meningitis or persistent CSF rhinorrhea (Table 24-1). When a selective resection of an adenoma is possible, subsequent hypopituitarism is uncommon.

TABLE 24-1. Mortality and Morbidity of Transsphenoidal Surgery

SUBFRONTAL APPROACH The subfrontal approach, via craniotomy, is limited to surgery for dumbbell-shaped lesions for which removal by a transsphenoidal approach is impossible; for tumors with suprasellar extension involving the chiasm; or for tumors involving the surrounding vascular structures. The disadvantages of the subfrontal approach include a high morbidity (increased duration of hospitalization, seizures, memory loss), and increased mortality resulting from damage to vital structures. Furthermore, a substantial incidence of postoperative hypopituitarism and diabetes insipidus is seen. COMBINED TRANSSPHENOIDAL AND SUBFRONTAL APPROACHES When combined transsphenoidal and subfrontal approaches are used, due to the lower risk of side effects and complications with the transsphenoidal operation as compared with the subfrontal approach, transsphenoidal surgery may be used as the primary surgery, and the removal of a tumor remnant may be performed, when necessary, by a subsequent subfrontal route. IRRADIATION THERAPY10,11,12,13,14,15,16,17,18,19,20,21 and 22 Before the advent of transsphenoidal surgery, radiation therapy was extensively proposed as the primary treatment for pituitary adenomas. In general, such irradiation is external mega-voltage photon therapy, although proton beams, cobalt-knife radiosurgery, and linear accelerator focal stereotactic technology are all potentially applicable to pituitary adenomas.21,23,24 (The implantation of pellets containing radioactive isotopes [e.g., yttrium-90, gold-198] has been completely abandoned.) The treatment techniques are variable. The dosage is tailored to the tumor volume, with a minimum dose being delivered to adjacent structures. Optimal techniques for conventional external radiotherapy include use of bilateral coaxial wedge fields plus a vertex field, moving arc fields, and 360 rotational fields12,21,25 (see Chap. 22). Although the use of two bilateral opposed fields is occasionally necessary for large, asymmetric tumors, it is to be discouraged because of the large dose that is delivered to the temporal lobes. Modern radiotherapy simulator facilities together with current generation computed tomography (CT) or MRI scanning allow the fields to be accurately molded around the tumor volume (conformational irradiation). The day-to-day reproducibility of the field setup should be within 2.0 mm. Optimum doses are based on evidence that doses <40 Gy provide a lower probability of tumor control; doses >50 Gy or fractions >2 Gy per day are associated with higher complication rates, including injury to the optic nerves or chiasm, and hypopituitarism. Thus, a dose of 505 Gy is advisable. If the bulk of the tumor is completely removed surgically, leaving only a minimum residue, and if the patient is young, 45 to 50 Gy is advisable. However, if a large residual lesion is present after surgical resection, if recurrent tumor is found, or if significant suprasellar extension is present, a higher dose (50 to 55 Gy depending on tumor bulk) is recommended.12,21,25 As is the case with surgery, radiotherapy must be performed by skilled radiotherapists who have accumulated considerable experience in pituitary irradiation. The major problem after pituitary irradiation (particularly when used as an adjunct to surgery) is the development of partial hypopituitarism or panhypopituitarism. Panhypopituitarism develops in approximately half of patients.13 The variably quoted figures are 30% to 45% for ACTH deficiency, 40% to 50% for gonadotropin deficiency, and 5% to 20% for TSH deficiency.12,13 and 14 The prevalence of deficiencies increases with the duration of follow-up: 100% of patients are GH deficient, 96% are gonadotropin deficient, 84% are ACTH deficient, and 49% are TSH deficient after a mean follow-up of 8 years.16 Hypothalamic-pituitary dysfunction may take up to 20 years to develop.20 The sensitivity of the hypothalamus and pituitary to the effects of radiation is well illustrated by the very frequent occurrence of endocrine dysfunction that is observed in patients irradiated for nasopharyngeal, extracranial, or primary brain tumors, even though these lesions were anatomically distinct from the hypothalamic-pituitary region.19 Accordingly, prolonged and repeated assessment of pituitary function is mandatory after irradiation therapy. This should permit a precise detection of pituitary deficiencies and the selection of appropriate replacement therapy. Nevertheless, one should emphasize that hormonal side effects of irradiation therapy, if diagnosed early, are easily managed. Other disadvantages of radiotherapy include a delayed therapeutic benefit for patients who have hormonally active tumors, irradiation-induced optic neuropathy, cortical injury, and, rarely, irradiation-induced malignancies (e.g., meningioma, astrocytoma).15,26 The term radiosurgery is applied to high-precision localized irradiation, given in one session. The gamma knife is one of these techniques and uses cobalt sources arranged in a hemisphere and focused onto a central target. High-precision stereotactic radiosurgery may also be delivered by adjusted linear accelerators. The aim of radiosurgery is to deliver a high dose of irradiation that is more localized than would be achieved with conventional radiotherapy. However, this is possible only for relatively small adenomas (<34 cm in diameter) and when the margins of the tumor are distant by >5 mm from the optic chiasm or optic nerves (due to the risk of irradiation-induced optic neuropathy that causes visual impairment). The long-term results of radiosurgery on hypersecretion, as on subsequent tumor growth, are presently unknown27 (also see Chap. 22).

THERAPEUTIC OPTIONS AND RESULTS BY TYPE OF PITUITARY ADENOMA

The recommendations for treatment of the different types of adenomas are summarized in Table 24-2.

TABLE 24-2. Multidisciplinary Treatment Decisions Based on Tumor Type in Pituitary Adenomas

GROWTH HORMONESECRETING PITUITARY ADENOMAS THERAPEUTIC OPTIONS Surgery and Radiotherapy. Surgery and radiotherapy, as previously summarized, are commonly used in the treatment of acromegaly. Medical Treatment. Bromocriptine and other dopamine agonists are able to improve symptoms of acromegaly in a few patients and to decrease GH secretion.28,29 Somatostatin, the hypothalamic GH-release inhibitory factor and its analogs, SMS 201-995 (octreotide) and BIM 23014 (lan-reotide), are able to reduce GH secretion. The native somatostatin peptide has a half-life that is too short for it to be administered easily. However, octreotide, given subcutaneously three times daily, has been shown to control GH hypersecretion and to decrease tumor volume in a significant proportion of patients with acromegaly with relatively few side effects.30,31,32,33,34,35 and 36 The availability of a long-acting form of octreotide allows once-monthly intramuscular injections with the same efficacy.37 Another somatostatin analog, lanreotide, when encapsulated in microspheres, has a prolonged release; it has proved to be effective in lowering levels of GH and insulin-like growth factor-I (IGF-I), and often in decreasing the tumor mass of acromegalic patients, in a manner comparable to that of octreotide.38,39,40 and 41 The side effects of somatostatin analogs are benign. Digestive problems (i.e., abdominal cramps, diarrhea, flatulence) are minor and most often transitory. Cholelithiasis occurs in 10% to 55% of patients, with the incidence related to the duration of the study.30,33,34,36 Generally, it is asymptomatic and is treated conservatively. Despite the reduction in insulin secretion due to the use of somatostatin analogs, glucose-tolerance alterations are of minor significance. Somatostatin analogs are very expensive drugs and need to be given for the remainder of life. Importantly, scintigraphy after administration of labeled octreotide (somatostatin-receptor scintigraphy) allows for the visualization of pituitary tumors.42 The resulting images are thought to reflect the concentration of the somatostatin receptors that are present at the surface of the tumor cells. However, scintigraphic findings are poor predictors of long-term results of treatment with somatostatin analog, regardless of the type of pituitary adenoma43 (also see Chap. 169). CRITERIA OF CURE OF ACROMEGALY The results of the various modes of therapy for acromegaly should be analyzed according to stringent criteria. Currently, cure (or good control) of acromegaly is defined by plasma GH levels: the mean of sequential sampling or the nadir after oral glucose administration should be <2.5 g/L and the IGF-I level should be normal.44,45 Indeed, when these goals are achieved, the life expectancy of patients with acromegaly seems comparable to that of the general population.46,47,48,49,50 and 51 In the future, even more stringent criteria (nadir GH after oral glucose administration of <1 g/L, and age- and sex-normalized IGF-I levels without clinical indications of activity) will probably be proposed for defining good therapeutic control of acromegaly.52 In the interim, the following section, using the currently accepted criteria for good control (plasma GH of <2.5 g/L and normal IGF-I level), compares the effects of the different treatments as indicated by several studies. RESULTS OF TREATMENT Transsphenoidal Surgery. According to the stringent criteria indicated above, 42% to 62% of patients can be considered to have their disease well controlled by surgery alone.8,45,48,50,53,54,55,56,57 and 58 (Table 24-3). The results depend on the size of the tumor: surgery is able to cure 61% to 91% of patients with microadenomas and 26% to 60% of those with macroadenomas. When the macroadenoma is very large, or when parasellar or sphenoid sinus invasion has occurred, the cure rate decreases to 17% and 40%, respectively.8 Also, the success rate of surgery in patients with acromegaly varies according to preoperative GH levels: surgical treatment is successful in ~70% of patients with preoperative GH levels of <10 g/L, 43% to 55% of those with GH levels of 10 to 50 g/L, and 18% to 40% of those with GH levels of >50 g/L.8,54 The relapse rate after surgical cure is <3%53,54,59,60 (also see Chap. 12 and Chap. 23).

TABLE 24-3. Results of Transsphenoidal Surgery for Treatment of Acromegaly with Stringent Criteria of Cure

Irradiation Therapy. Irradiation therapy is able to decrease GH levels in a large proportion of patients. Mean plasma GH levels of <5 g/L are obtained in ~50% of patients (40% to 80%, depending on the length of follow-up).18,22,61,62,63,64,65 and 66 However, when more stringent criteria for cure (as stated earlier) are applied, radiation therapy leads to cure of the disease in only 5% to 38% of the cases after a median follow-up of ~7 years22,65,66 (Table 24-4). Irradiation is almost always followed by hypopituitarism, however, and the full impact on GH hypersecretion is delayed for many years. Preliminary results with radiosurgery are now available, but the follow-up is short. In one study,67 at 20 months, 20% of patients had normalized GH and IGF-I levels after radiosurgery, and the mean delay for normalization of hormonal parameters was reduced (1.4 years as opposed to 7.1 years after conventional radiotherapy).68 Studies involving a higher number of patients followed for a longer period of time and assessed with more stringent criteria of cure are needed before one can conclude that radiosurgery is superior to conventional radiotherapy.

TABLE 24-4. Success Rate of Radiation Therapy for Acromegaly with Different Criteria of Cure

Therapy with Bromocriptine and Other Dopaminergic Agonists. Treatment with bromocriptine or other dopaminergic agonists produces improvement in clinical symptoms of acromegaly in half of the patients. These drugs substantially decrease GH levels in some patients but only rarely normalize GH and IGF-I levels (i.e., in <10% of cases).28 Better results seem to be obtained with cabergoline than with bromocriptine, however; in a multicenter study, nearly 40% of patients acromegaly treated with cabergoline were reported to have normalized IGF-I levels.29 Therapy with Somatostatin Analogs. Somatostatin-analog therapy has now gained wide acceptance in the medical treatment of acromegaly. GH levels are decreased in 50% to 80% of patients treated with octreotide subcutaneously three times daily.30,33,34,35 and 36,45,69,70 With this treatment, up to 50% of acromegalic patients may be considered as cured (GH plasma levels of <2 g/L [2030% of cases] and/or normal IGF-I [2060% of cases]) (Table 24-5). Similar results are obtained with lanreotide LAR (long-acting release), 30 mg administered intramuscularly every 10 or 14 days (GH plasma levels of <2 g/L [3070% of cases] and/or normal IGF-I [4070% of patients]), 38,39,40 and 41 or with octreotide LAR. This latter drug has been administered intramuscularly every month at a dose of 20 to 30 mg (yielding GH plasma levels of <2 g/L [5060% of patients] and/or normal IGF-I [6090% of cases])37,71,72 and 73 (see Table 24-5). Such variations in the data obtained from one study to another is probably explained by differences in the methods used for IGF-I assay and by differences in the inclusion criteria used. Thus, in some of the studies assessing the efficiency of long-acting forms of somatostatin analogs, patients were included if they had previously been shown to be responsive to subcutaneous octreotide, whereas in others, patients were entered blindly, without any knowledge of whether or not they were responsive. As demonstrated by a multicenter prospective study, the efficacy of octreotide as primary treatment (in 26 previously untreated patients) proved to be equivalent to that of secondary treatment (in 81 patients previously treated with surgery and/or radiotherapy).70

TABLE 24-5. Effects of Long-Term Treatment of Acromegaly with the Somatostatin Analogs Octreotide and Lanreotide with Various Modes of Preparation and Administration and Different Criteria of Cure

A small reduction in tumor volume may be observed (in general at the level of the suprasellar expansion) in 15% to 70% of patients with acromegaly30,33,34,35,36 and 37,39,40 and 41,45,69,70,71,72 and 73 (see Table 24-5). Administration of a combination of a dopamine agonist and a somatostatin analog may be beneficial for some patients, but long-term studies assessing this therapeutic association are not currently available. GH-receptor antagonist therapy may be effective in the control of the clinical symptoms of acromegaly and normalizes IGF-I levels in a substantial number of patients. SUMMARY OF RECOMMENDATIONS FOR TREATMENT OF GROWTH HORMONESECRETING ADENOMAS Table 24-2 summarizes the treatment decisions for GH-secreting adenomas. 1. Transsphenoidal surgery is the first-line therapy, except when a macroadenoma is extremely large or when surgery is contraindicated. 2. Postoperative radiation therapy (50 to 55 Gy) is performed for partially resected tumors or when GH levels remain elevated after a trial of a somatostatin analog. 3. Somatostatin analogs are best used when surgery is contraindicated or when the surgery has failed to normalize GH levels. These drugs are also used when waiting for the delayed effects of radiation therapy. Somatostatin analogs may be a reasonable primary therapeutic modality if the possibility of surgical cure is low (as in patients with large and/or invasive tumors), provided that the tumor does not threaten vision or neurologic function. PROLACTIN-SECRETING TUMORS THERAPEUTIC OPTIONS Occasionally, surgery and, more rarely, radiotherapy may be used. Their techniques and side effects have been detailed in the first section (see earlier). In the vast majority of cases, however, medical therapy with dopamine agonists is chosen. Bromocriptine or other ergot derivatives with dopaminergic properties (e.g., pergolide, quinagolide, cabergoline) are used. Bromocriptine effectively reduces elevated serum PRL levels in most patients with PRL-secreting pituitary adenomas.74,75 and 76 Thus, it is the primary treatment for microadenomas. Furthermore, macroprolactinomas have also been successfully managed medically.74,77,78 and 79 Bromocriptine not only decreases PRL levels but also is able to produce a dramatic reduction in tumor volume. This antitumor effect of bro-mocriptine may be very rapid. In the presence of chiasmal compression by a macroadenoma, visual improvement often occurs within the first hours after the initiation of bromocriptine therapy. Ovulatory and menstrual cycles may resume in 75% to 80% of female patients.74,78 Patients with macroade-nomas who become pregnant are at risk for complications related to tumor growth. For these patients, pregnancy should be delayed by contraceptive methods, while shrinkage is attempted with bromocriptine. Definitive ablative therapy of macroadenomas may be performed before pregnancy.74,78 However, patients with intrasellar microadenomas or macro-adenomas of <12 mm are at <6% risk for any substantial tumor growth or associated complications.80,81 and 82 Bromocriptine should be discontinued when pregnancy is confirmed, although early concerns regarding the teratogenicity of this drug have not, to date, been substantiated.74 Quinagolide and cabergoline are two other dopamine agonists now available in the United States and/or in Europe. Quinagolide is not an ergot derivative; it binds specifically to dopamine type 2 receptors. It is at least as effective as bromocriptine83 and may be useful in rare cases of resistance84,85 and 86 or intolerance86 to bromocriptine. It offers the advantage of once-daily administration. Cabergoline is notable for its very long duration of action (half-life is ~70 hours). This allows a once- or twice-weekly oral administration.87 The efficacy of cabergoline is at least equal to that of bro-mocriptine and tolerance to it is better.87,88 In patients with macroprolactinoma, cabergoline is quite effective in reducing PRL levels and in producing tumor shrinkage.89,90 and 91 Furthermore, cabergoline normalizes serum PRL levels in a significant number of patients who are resistant to bromocriptine and/or quinagolide90 (see Chap. 13).

RESULTS OF TREATMENT Transsphenoidal Surgery MICROADENOMA. In a compilation of various studies involving a total of 1224 patients, transsphenoidal resection was found to normalize PRL levels in 71.2% of patients.74 Cure rates vary according to preoperative serum PRL level; when the PRL level is >200 g/L, the cure rate decreases to 13%. After surgery, up to 88% of women desiring conception conceived within 1 year.74,92 Initially, the recurrence rate after surgical success was thought to be as high as 50%; however, more recent studies report a recurrence rate from 5% to 18% after a 10-year follow-up. 74,92,93 Nonetheless, even if hyperprolactinemia should relapse, it tends to remain mild and usually without any clinical consequences. Ten years after surgery, 55% to 73% of patients have normal serum PRL levels, and 75% have normal menstrual cycles.92,93
and 94

MACROADENOMA. The success rate for transsphenoidal surgery is much lower for macroadenomas. According to a compilation of 1256 such patients from reported series, the mean cure rate was 31.8%.74 Moreover, after apparent initial cure, 18% of patients relapsed. Surgical results were proportional to the pre-operative size of the tumor and to the initial level of serum PRL. When the adenoma size was >20 mm and the serum PRL levels were >200 g/L, the cure rate was <15%.74 Medical Therapy. Dopamine agonists are able to decrease PRL levels in 73% to 95% of patients, whatever the size of their adenomas; moreover, the drugs produce a decrease in tumor volume in 77% of patients with a macroprolactinoma.74,77,79,83,89,91 This tumor shrinkage may be dramatic (>50% of initial volume in half of the patients). Treatment with dopamine agonists is prolonged and often lifelong. Indeed, after withdrawal of treatment, the tumor usually returns to its original size, often within days to weeks, and the serum PRL levels again rise. In patients who are resistant to bromocriptine, other dopamine agonists such as quinagolide or cabergoline may be a useful alternative; quinagolide allows normalization of PRL levels in 16% to 20% of cases, and cabergoline in 80% of patients.84,85 and 86,90 In cases of intolerance of bromocriptine, PRL may be normalized by quinagolide in more than half of patients.86 Tolerance of cabergoline has proven to be better than tolerance of bromocriptine, so that larger doses can be used in cases of incomplete response.87,88,89,90 and 91 Irradiation Therapy. Irradiation therapy is indicated only in rare cases of surgical failure in which the patient has subsequently been found to be intolerant of or resistant to dopamine agonists 20% to 30%,12,74,95,96 and only half of such patients have normal serum PRL levels 3 to 4 years and 8 years after irradiation therapy, respectively.97 SUMMARY OF RECOMMENDATIONS FOR TREATMENT OF PROLACTIN-SECRETING PITUITARY ADENOMAS Table 24-2 summarizes the treatment decisions for prolactin-secreting pituitary adenomas. 1. Selected hyperprolactinemic patients with microade-nomas may be carefully followed without treatment if regular ovulatory cycles occur, or if the patient is post-menopausal. 2. Patients with microadenomas are usually treated with dopamine agonists. In the uncommon circumstance that transsphenoidal surgery was the initial choice, dopamine agonists may be useful if the serum PRL levels remain elevated postoperatively. When medical therapy with dopamine agonists is the primary treatment, secondary surgery may occasionally be proposed in cases of resistance to or intolerance of these drugs. In dopamine-treated microprolactinoma patients, pregnancy can be permitted, and the dopamine agonists are generally interrupted as soon as the pregnancy is confirmed. 3. In patients with macroadenomas, even in the presence of a chiasmatic syndrome, dopamine agonists are the best primary treatment. Improvement in the visual disturbance is often very rapid, and tumor shrinkage is usually very significant. Thus, the results provided by dopamine agonists are generally much better than those obtained with surgery, even when it is performed by a highly skilled surgeon. If dopamine agonists are not rapidly effective, however, surgery is recommended. In such cases, if serum PRL levels remain high postoperatively (which is likely the case for macroadenomas, particularly when they are large), dopamine agonists are given. When pregnancy is planned in patients with macroprolactinomas, one might propose to pursue treatment with dopamine agonists during the entire pregnancy. Alternatively, before pregnancy, surgery may be selected with the goal of removing much of or the entire lesion. Thereafter, dopamine agonists are given to those patients with remaining hyperprolactinemia until the onset of pregnancy; at that time, the drugs are halted. 4. Currently, radiotherapy is an option that is very seldom used. This treatment must be limited to the rare patients with large, incompletely resected tumors who continue to have high serum PRL levels, are resistant to dopamine agonists, and experience amenorrhea or mass effects of the lesion. ADRENOCORTICOTROPIC HORMONESECRETING ADENOMAS Most ACTH-secreting adenomas (90%) are microadenomas. The principal therapeutic difficulty caused by these lesions are their small size, which may hamper their preoperative visualization by imaging studies and their subsequent surgical resection. When a rare macroadenoma occurs, it usually is large and invasive; in such a case, total surgical resection may be difficult. Due to the poor prognosis of persistent hypercortisolism, adjuvant therapy with drug therapy is frequently required. THERAPEUTIC OPTIONS Transsphenoidal surgery and/or radiotherapy are the main therapeutic options; their techniques and side effects have been described. Medical treatment has an important place as an adjuvant therapy. Although a few case reports of cyproheptadine-induced remission of Cushing disease have been described, in the vast majority of patients no beneficial effect or only a moderate improvement has been observed with cyproheptadine administration,98 and its use has been abandoned. Octreotide and lanreotide have no established place in the treatment of Cushing disease, although a few patients may respond with a decrease in ACTH secretion.98 Medical therapy in Cushing disease is currently limited to the use of compounds directed to the adrenal that are able to inhibit steroidogenesis (i.e., mitotane, ketoconazole, metyrapone, etomidate, trilostane, or aminoglutethimide).98,99 The adrenolytic agent mitotane (1,1-dichloro-2-[o-chlorophenyl]-2-[p-chlorophenyl]-ethane or o,p'-DDD) inhibits 11b-hydroxylase and cholesterol side-chain cleavage enzymes and destroys adrenocortical cells. The drug is given orally at a dosage of 6 to 12 g per day. Principal side effects of mitotane are gastrointestinal symptoms (nausea, loss of appetite), hypercholesterolemia, and gynecomastia. Neurologic side effects and skin rashes rarely occur. True drug-induced hepatitis (with increased levels of serum transaminases), which requires discontinuation of the drug, is rare; this must be distinguished from increased levels of g-glutaryl transferase, which is frequent and benign and occurs in the setting of a fatty liver associated with the obesity of Cushing syndrome. Due to its marked liver enzymeinducing effect, increased production of several binding proteins (in particular transcortin, the cortisol-binding globulin) occurs during mitotane therapy. Thus, the efficacy of the drug needs to be monitored; assessment should not be based on total plasma cortisol levels (which remain artifactually increased) but on free urinary cortisol or on salivary cortisol. These latter tests reflect the serum levels of the free, unbound cortisol. Doses of hydrocortisone given for replacement therapy, if required, may for the same reason show superior effects to those generally used for the replacement therapy of adrenal insufficiency. With prolonged administration at high doses, mitotane is able to produce permanent destruction of adrenocortical cells and result in glucocorticoid insufficiency; this may be associated with hypoaldosteronism (see Chap. 75). Ketoconazole, an imidazole-derivative antimycotic agent, inhibits various cytochrome P450 enzymes, including the side-chain cleavage enzymes 17,20 lyase, 11b-hydroxylase, and 17-hydroxylase. Ketoconazole is administered at a dosage of 600 to 800 mg twice daily. Reported side effects are hepatitis, gynecomastia, and gastrointestinal symptoms. Another imidazole derivative, the anesthetic drug etomidate, when given intravenously at a nonhypnotic dose (3 mg/kg body weight per hour) has an immediate cortisol-lowering effect. Metyrapone, administered at a dosage of 2 to 4 g per day, inhibits 11b-hydroxylase and also may be used in the treatment of hypercortisolism. Nausea, vomiting, and neurologic symptoms are the most commonly reported side effects of metyrapone therapy. Aminoglutethimide (750 mg per day) inhibits several cytochrome p450 steroidogenic enzymes and efficiently controls hypercortisolism, but it may cause sedation, nausea, and rash; the same is true for trilostane (2001000 mg per day), a carbonitrile derivative that inhibits conversion of pregnenolone to progesterone. Except in the rare cases in which medical treatment is poorly tolerated or produces severe side effects (hepatitis), bilateral adrenalectomy is no longer used as therapy for Cushing disease because of substantial morbidity and mortality, the need for permanent replacement therapy with corticosteroids, and the subsequent (albeit low) risk for development of Nelson syndrome.

RESULTS OF TREATMENT In the hands of skilled neurosurgeons, the apparent cure rate achieved by transsphenoidal surgery for Cushing disease due to microadenoma is 70% to 80%.98,100,101,102,103,104,105,106,107,108,109,110 and 111 The criteria for cure are an undetectable morning serum cortisol concentration and a plasma ACTH concentration of <5 pg/mL 4 to 7 days after surgery. 101 Less strict criteria result in higher rates of apparent cure but a higher rate of recurrence (Table 24-6).

TABLE 24-6. Results of Transsphenoidal Surgery for Treatment of Cushing Disease

The recurrence rate of Cushing disease after surgery for a microadenoma is more frequent than was previously thought; a progressive upward trend is seen with time.103 In most series, the recurrence rate is 10% to 20% (see Table 24-6). Only 50% of patients with macroadenomas achieve remissions after initial pituitary microsurgery. In children, the rate of cure after transsphenoidal surgery is 70% to 86%104,112,113 and 114 and even reached 97% for one team.115 As is the case with adults, however, after extended follow-up and using the more stringent criteria now applied, these results will probably prove to be less favorable, probably in the range of 50% to 75%.113,114 In case of surgical failure, secondary irradiation therapy is able to produce remission in 56%116 to 83%117 of patients with Cushing disease after a median follow-up of ~3 years. Primary radiotherapy for Cushing disease is effective in 50% to 60% of adults and 85% of children.21,98,99,118 During the months or years required to achieve maximal benefits from irradiation, hypercortisolism usually can be controlled with adrenal enzyme inhibitors.98,99 After 8 months of therapy, remission of hypercortisolism is achieved in 83% of patients treated with mitotane. Combined with pituitary irradiation, mitotane produces remission of hypercortisolism in 80% to 100% of patients, but discontinuation of the drug leads to recurrence of the hypercortisolism in 30% to 50% of cases. Use of ketoconazole rapidly normalizes serum cortisol levels, but an escape from the effect of the steroidogenic agent is generally observed after a few months of treatment; this prevents long-term administration. Metyrapone is able to rapidly normalize plasma cortisol in 50% to 75% of patients with Cushing disease. Its interruption leads to recurrence of hypercortisolism. Aminoglutethimide only appears to be effective in controlling hypercortisolism in <50% of cases. SUMMARY OF RECOMMENDATIONS FOR TREATMENT OF ADRENOCORTICOTROPIC HORMONESECRETING ADENOMAS Table 24-2 summarizes the treatment decisions for Adrenocorti-cotropic HormoneSecreting Adenomas. 1. The primary therapy for children and adults is generally transsphenoidal surgery. 2. Radiotherapy (50 Gy) is reserved for patients who have undergone only subtotal resection or who remain hyper-secretory after surgery. 3. While one waits for the effects of radiotherapy, or if it is contraindicated, adrenal steroidogenesis inhibitors (in particular, mitotane) may be indicated. THYROID-STIMULATING HORMONESECRETING ADENOMAS THERAPEUTIC OPTIONS Surgery and radiotherapy are the usual therapeutic tools for treating TSH-secreting adenomas. They have been described in detail earlier in this chapter. Medical adjuvant therapy with somatostatin analogs has proven to be very useful. Octreotide and lanreotide are quite effective in reducing TSH levels in patients with TSH-secreting adenomas, thus allowing normalization of plasma thyroid hormone levels in a high proportion of patients.119,120 The associated side effects are similar to those of somatostatin analogs used in the treatment of acromegaly. RESULTS OF TREATMENT Forty percent of patients with TSH-secreting adenomas are not cured by surgery, even when it is combined with radiation therapy. In these patients, thyrotoxicosis persists.121 Octreotide, the somatostatin analog, is able to reduce serum TSH and a subunit levels in 91% of patients and normalizes thyroid hormone levels in 73% of patients.119 The reduction in tumor volume is similar to that obtained in acromegalic patients treated with octreotide. Lanreotide has very similar effects.120 SUMMARY OF RECOMMENDATIONS FOR TREATMENT OF THYROID-STIMULATING HORMONESECRETING ADENOMAS Treatment decisions for TSH-secreting adenomas are summarized as follows: 1. The primary therapy is transsphenoidal surgery, regardless of the size of the tumor. 2. Irradiation therapy (4050 Gy) is generally proposed only in the case of incomplete resection, particularly when the remnant is invasive. 3. Somatostatin analogs are indicated in cases of persistent postoperative hyperthyroidism, while the effects of radiotherapy are being awaited. CLINICALLY NONFUNCTIONING PITUITARY ADENOMAS The majority of NFPAs are gonadotrope cell adenomas. Indeed, among all the types of macroadenomas, NFPAs are the most frequent. Despite their gonadotrope nature, as demonstrated by immunocytochemical staining, NFPAs are only rarely associated with increased levels of dimeric LH or FSH. However, increased levels of uncombined subunits (free a subunit primarily, LH-b subunit more rarely) are more frequently encountered but are generally modest in titer. The posttreatment assessment of cure usually is based on morphologic changes; hormonal monitoring is not usually helpful for this assessment. The main problems raised by NFPAs are the mass effects, mainly optic chiasm compression and/or deficient hormone secretion resulting from compression of normal anterior pituitary cells. THERAPEUTIC OPTIONS Surgery and radiotherapy are the only really effective therapeutic options available for NFPA; technical aspects and side effects have been discussed earlier. Various medical treatments have been tried (i.e., administration of dopamine agonists, gonadotropin-releasing hormone [GnRH] analogs), but none has proven to be sufficiently effective as a reasonable therapeutic option. In occasional patients, the somatostatin analog octreotide may minimally improve visual defects due to

chiasmal compression.122,123 RESULTS OF TREATMENT The strategy of observation only for patients with incidentally discovered pituitary adenomas (incidentalomas) may be appropriate provided the tumor is well delimited, small, and has no suprasellar or lateral extension that risks neurologic or visual chiasm compression, and a meticulous hormonal work-up has ruled out the possibility that the hormonal hypersecretion is sufficient to produce a clinical syndrome.124 In all other cases, clinically evident, but apparently inactive, pituitary adenomas require surgery. Transsphenoidal surgery allows improvement in visual disturbances due to chiasmal syndrome in 44% to 70% of patients.125,126,127,128 and 129 When the NFPA is a gonadotropin-secreting adenoma associated with supranormal levels of gonadotropins or free subunits, surgery almost always reduces supranormal plasma FSH and/or a subunit levels and normalizes them,129 despite the persistence of a tumor remnant. When an NFPA is responsible for pituitary failure, surgery is able to improve pituitary function in 15% to 50% of cases. On the other hand, surgery may aggravate the preoperative pituitary deficiency. After surgery alone, nearly 30% (from 10% to 69%) of patients relapse within 5 to 10 years.10,25,125,127,129,130,131,132,133,134 and 135 variability in these data reflect differences in neurosurgical expertise, as well as the differing quality of the imaging techniques used postoperatively to assess the extent of tumor excision (Table 24-7).

TABLE 24-7. Recurrence Rate According to Therapy for Nonfunctioning Pituitary Adenomas

Radiotherapy is proposed either as a systematic adjunct or if a significant remnant persists. Use of systematic radiation therapy is supported by the low relapse rate (mean, 11%; range, 6% 21%) that is observed when irradiation is routinely combined with surgery10,20,25,125,129,130,131,132,133,134 and 135 (see Table 24-7). However, irradiation is almost always followed by hypopituitarism, and several epidemiologic studies have demonstrated that postirradiation hypopituitarism might be associated with a reduction in life expectancy, despite appropriate replacement therapy136,137 and 138 (see Prognosis section). Results of medical treatment are disappointing. Bromocriptine is not very effective in reducing levels of supranormal gonadotropins and free subunits, and only rarely produces a minimal tumor shrinkage.79,139 Somatostatin analogs are able to improve visual problems minimally in 20% to 40% of cases, but reports of reduction in tumor volume is anecdotal.122,123 Use of GnRH agonists is generally ineffective139 and may be hazardous.140 Prolonged administration of a GnRH antagonist to a small number of patients with a secreting gonadotrope cell adenoma has been reported to reduce supranormal gonadotropin levels but did not produce any change in tumor size.141 THERAPEUTIC RECOMMENDATIONS FOR TREATMENT OF CLINICALLY NONFUNCTIONING ADENOMAS Table 24-2 summarizes recommendations for the treatment of NFPAs. Transsphenoidal surgery with or without postoperative radiation therapy (50 Gy) is performed for almost all patients, irrespective of whether they have visual consequences of their tumor. Selected patients with small, incidentally discovered microadenomas may be carefully followed without immediate therapy.

PROGNOSIS OF PITUITARY ADENOMAS AND THERAPEUTIC PERSPECTIVES

PROGNOSIS Prognosis depends on the type of tumor and a combination of other factors, including (a) the severity of the endocrinologic disturbance or mass-related symptoms and signs, (b) the size and extent of the tumor as indicated earlier, (c) the success of therapy in reversing these abnormalities, (d) morbidities due to therapy, and (e) permanence of the therapeutic response. Although pituitary tumors are generally benign, the failure to provide adequate therapy can lead to severe functional deficits and death. Optimal therapy can greatly improve quality and duration of life. ACROMEGALY If patients with acromegaly remain untreated, they have a 10-year reduction in life expectancy, in particular due to cardiovascular and respiratory problems and to the increased risk of neoplasms. The standardized mortality ratio (SMR, the observed mortality divided by the expected mortality in a sex- and age-matched control population) is from 1.8 to 3.46,48,49 and 50 The most important predictive factor of mortality is the final posttherapeutic serum GH level.46,47,48,49 and 50 When the final serum GH level is <2.5 g/L, the mortality is not significantly different from that of the general population. When the final serum GH level is >2.5 g/L, the SMR is between 1.4 and 2, which signifies a statistically different mortality from that of the general population. This indicates that application of the more stringent definition of cure or successful outcome in acromegaly which is currently used (i.e., serum GH levels of <2.5 g/L) is probably associated with better survival than were the criteria previously used (i.e., serum GH <5 g/L or <10 g/L). CUSHING DISEASE The mortality in patients with Cushing disease is higher than that expected for the control population (SMR, 3.8; 95% confidence limits, 2.517.9). The most common cause of death is vascular disease. Advanced age, persistence of hypertension, and abnormalities in glucose metabolism after treatment are independent predictors of mortality.142 HYPOPITUITARISM Hypopituitarism, particularly after surgery and/or radiation therapy for NFPA, is also associated with an increase in mortality (SMR, 1.702.10), according to the very concordant results of three studies.136,137 and 138 Death was reported to be due mainly to an increase in cardiovascular deaths in some studies136,137 and 138 but not in others137 (Table 24-8). GH deficiency was cited as a potential cause of the increased mortality in these patients; presumably, they were given adequate replacement therapy for other pituitary hormones but had not received GH-replacement therapy. This assertion remains highly questionable, however, because GH deficiency was not proven in all the patients in these studies and because often evidence was found of a long-term lack of pituitary hormones or of an inadequacy of hormonal substitution in many of them. GH treatment in patients with GH deficiency has been reported to be associated with numerous, albeit modest, increases in lean body mass, improved quality of life, amelioration of lipid abnormalities, and increased bone mineral content.143 However, whether or not a substitutive therapy with GH would improve the prognosis for hypopituitarism in adults is presently unknown.

TABLE 24-8. Epidemiologic Studies Demonstrating an Increased Mortality in Patients with Hypopituitarism

PROLONGED (LIFELONG) SURVEILLANCE Regular assessment for many years of the pituitary function of patients treated for pituitary adenomas is essential to allow rapid adequate replacement therapy in patients who develop hypopituitarism. Hypopituitarism may occur up to 30 years after radiation therapy. CONCLUSIONS AND PERSPECTIVES CONCERNING TREATMENT OF PITUITARY ADENOMAS Although progress in imaging techniques has greatly improved the ease of diagnosing and localizing pituitary lesions, it has also been responsible for more frequent detection of incidental and clinically innocent lesions. The appropriate management of such pituitary incidentalomas, taking into account the cost, possible benefits, and complications of potential therapy, needs to be determined more precisely. The early recognition of acromegaly, a condition that is responsible for important morbidity and cosmetic consequences, needs to be improved. The management of pituitary adenomas is currently well defined. According to the type of tumor and the clinical situation, options include surgery, irradiation, and/or drugs.144 Surgical resection by the transsphenoidal route carries a very low morbidity and mortality. In skilled hands, it often allows a complete cure of the disease. Improvement is needed in surgical techniques to obtain a better removal of invasive tumors. Whether endoscopic techniques or the use of intraoperative MRI or intraoperative computer-assisted neuronavigation with robots would improve the outcome is presently unknown. Diagnostic challenges are rare, except in the case of Cushing disease. Here, Cushing syndrome due to occult ectopic ACTH-secreting tumors (often bronchial carcinoids) may be misconstrued as being due to a nonvisible corticotropic microadenoma. This may delay appropriate treatment. Further improvements in imaging techniques are unlikely to allow better definition of these pituitary microadenomas. When Cushing disease is suspected, bilateral inferior petrosal sinus catheterization generally confirms or rules out a pituitary causation. Histopathologic techniques (particularly immunocytochemistry) have greatly assisted the classification of pituitary adenomas. However, no markers of invasiveness or aggressiveness have been identified that can be routinely applied to tumor specimens removed at surgery. In this respect, prospective studies of molecular genetic alterations that have been demonstrated in pituitary adenomas144 will determine whether they are truly predictive of subsequent behavior and can be used to aid clinical management in a manner not possible with current histologic criteria. Conceivably, this may assist in decisions regarding which patients might need postoperative irradiation therapy. CHAPTER REFERENCES
1. Kovacs K, Horvath E. Tumors of the pituitary gland. In: Hartman WH, ed. Atlas tumor of pathology, 2nd series, Fascicle XXI. Washington: Armed Forces Institute of Pathology, 1986:1. 2. Horvath E, Kovacs K. The adenohypophysis. In: Kovacs K, Asa L, eds. Functional endocrine pathology. Boston: Blackwell Science, 1998:247. 3. Ambrosi B, Faglia G. Epidemiology of pituitary tumors. In: Faglia G, Beck-Peccoz P, Ambrosi B, Travaglini P, Spada A, eds. Pituitary adenomas : new trends in basic and clinical research. Amsterdam: Excerpta Medica, 1991:159. 4. Molitch ME, Russell EJ. The pituitary incidentaloma. Ann Intern Med 1990; 112:925. 5. Teramoto A, Hirakawa K, Sanno N, Osamura Y. Incidental pituitary lesions in 1,000 unselected autopsy specimens. Radiology 1994; 193:161. 6. Hall WA, Luciano MG, Doppman JL, et al. Pituitary magnetic resonance imaging in normal human volunteers: occult adenomas in the general population. Ann Intern Med 1994; 120:817. 7. Laws ER Jr, Thapar K. Surgical management of pituitary adenomas. Baillires Clin Endocrinol Metab 1995; 9:391. 8. Fahlbusch R, Honegger J, Buchfelder M. Surgical management of acromegaly. Endocrinol Metab Clin North Am 1992; 21:669. 9. Wilson CB. Extensive clinical experience: surgical management of pituitary tumors. J Clin Endocrinol Metab 1997; 82:2381. 10. Chun M, Masko GB, Hetelekidis S. Radiotherapy in the treatment of pituitary adenomas. Int J Radiat Oncol Biol Phys 1988; 15:305. 11. Grigsby PW, Stokes S, Marks JE, Simpson JR. Prognostic factors and results of radiotherapy alone in the management of pituitary adenomas. Int J Radiat Oncol Biol Phys 1988; 15:1103. 12. Sheline G, Tyrrell J. Pituitary tumors. In: Perez C, Brady L, eds. Principles and practice of radiation oncology. Philadelphia: JB Lippincott, 1987:1108. 13. Nelson PB, Goodman ML, Flickenger JC, et al. Endocrine function in patients with large pituitary tumors treated with operative decompression and radiation therapy. Neurosurgery 1989; 24:398. 14. Snyder PJ, Fowble BF, Schatz NJ, et al. Hypopituitarism following radiation therapy of pituitary adenomas. Am J Med 1986; 81:457. 15. Brada M, Ford D, Ashley S, et al. Risk of second brain tumour after conservative surgery and radiotherapy for pituitary adenoma. BMJ 1992; 304:1343. 16. Littley MD, Shalet SM, Beardwell CG, et al. Hypopituitarism following external radiotherapy for pituitary tumours in adults. Q J Med 1989; 262:145. 17. Linfoot JA. Heavy ion therapy. In: Linfoot JA, ed. Recent advances in the diagnosis and treatment of pituitary tumors. New York: Raven Press, 1979:245. 18. Eastman RC, Gorden P, Glatstein E, Roth J. Radiation therapy of acromegaly. Endocrinol Metab Clin North Am 1992; 21:693. 19. Constine LS, Woolf PD, Cann D, et al. Hypothalamic-pituitary dysfunction after radiation for brain tumors. N Engl J Med 1993; 328:87. 20. Brada M, Rajan B, Traish D, et al. The long-term efficacy of conservative surgery and radiotherapy in the control of pituitary adenomas. Clin Endocrinol (Oxf) 1993; 38:571. 21. Plowman PN. Radiotherapy for pituitary tumours. Baillires Clin Endocrinol Metab 1995; 9:407. 22. Barkan AL, Halasz I, Dornfeld KJ, et al. Pituitary irradiation is ineffective in normalizing plasma insulin-like growth factor I in patients with acromegaly. J Clin Endocrinol Metab 1997; 82:3187. 23. Thorn M, Rhn T, Guo WY, Werner S. Stereotactic radiosurgery with the cobalt-60 gamma unit in the treatment of growth hormone-producing pituitary tumors. Neurosurgery 1991; 29:663. 24. Pollock BE, Kondziolka D, Lunsford LD, Flickinger JC. Stereotactic radio-surgery for pituitary adenomas: imaging, visual and endocrine results. Acta Neurochir Suppl (Wien) 1994; 62:33. 25. Sheline GE. Proceedings: treatment of nonfunctioning chromophobe adenomas of the pituitary. Am J Roentgenol Radium Ther Nucl Med 1974; 120:553. 26. Tsang RW, Laperriere NJ, Simpson WJ, et al. Glioma arising after radiation therapy for pituitary adenoma. A report of four patients and estimation of risk. Cancer 1993; 72:2227. 27. Brada M, Cruickshank G. Radiosurgery for brain tumours. BMJ 1999; 318:411. 28. Jaffe CA, Barkan AL. Treatment of acromegaly with dopamine agonists. Endocrinol Metab Clin North Am 1992; 21:713. 29. Abs R, Verhelst J, Maiter D, et al. Cabergoline in the treatment of acromegaly: a study in 64 patients. J Clin Endocrinol Metab 1998; 83:374. 30. Chanson P, Timsit J, Harris AG. Clinical pharmacokinetics of octreotide. Therapeutic applications in patients with pituitary tumours. Clin Pharma-cokinet 1993; 25:375. 31. Lamberts SW, Hofland LJ, de Herder WW, et al. Octreotide and related somatostatin analogs in the diagnosis and treatment of pituitary disease and somatostatin receptor scintigraphy. Front Neuroendocrinol 1993; 14:27. 32. Lamberts SW, van der Lely AJ, de Herder WW, Hofland LJ. Octreotide. N Engl J Med 1996; 334:246. 33. Sassolas G, Harris AG, James-Deidier A. Long term effect of incremental doses of the somatostatin analog SMS 201-995 in 58 acromegalic patients. French SMS 201-995 Acromegaly Study Group. J Clin Endocrinol Metab 1990; 71:391. 34. Vance ML, Harris AG. Long-term treatment of 189 acromegalic patients with the somatostatin analog octreotide. Results of the International Multi-center Acromegaly Study Group. Arch Intern Med 1991; 151:1573. 35. Ezzat S, Snyder PJ, Young WF, et al. Octreotide treatment of acromegaly. A randomized, multicenter study. Ann Intern Med 1992; 117:711. 36. Newman CB, Melmed S, Snyder PJ, et al. Safety and efficacy of long-term octreotide therapy of acromegaly: results of a multicenter trial in 103 patientsa clinical research center study. J Clin Endocrinol Metab 1995; 80:2768. 37. Gillis JC, Noble S, Goa KL. Octreotide long-acting release (LAR). A review of its pharmacological properties and therapeutic use in the management of acromegaly. Drugs 1997; 53:681. 38. Heron I, Thomas F, Dero M, et al. Pharmacokinetics and efficacy of a long-acting formulation of the new somatostatin analog BIM 23014 in patients with acromegaly. J Clin Endocrinol Metab 1993; 76:721. 39. Morange I, De Boisvilliers F, Chanson P, et al. Slow release lanreotide treatment in acromegalic patients previously normalized by octreotide. J Clin Endocrinol Metab 1994; 79:145. 40. Giusti M, Gussoni G, Cuttica CM, Giordano G. Effectiveness and tolerability of slow release lanreotide treatment in active acromegaly: six-month report on an Italian multicenter study. Italian Multicenter Slow Release Lanreotide Study Group. J Clin Endocrinol Metab 1996; 81:2089. 41. Caron P, Morange-Ramos I, Cogne M, Jaquet P. Three year follow-up of acromegalic patients treated with intramuscular slow-release lanreotide. J Clin Endocrinol Metab 1997; 82:18. 42. Krenning EP, Kwekkeboom DJ, Bakker WH, et al. Somatostatin receptor scintigraphy with [111In-DTPA-D-Phe1]- and [123 I-Tyr 3]-octreotide: the Rotterdam experience with more than 1000 patients. Eur J Nucl Med 1993; 20:716. 43. Chanson P. Predicting the effects of long-term medical treatment in acromegaly. At what cost? For what benefits? Eur J Endocrinol 1997; 136:359. 44. Frohman LA. Acromegaly: what constitutes optimal therapy? J Clin Endocrinol Metab 1996; 81:443. 45. Melmed S, Ho K, Klibanski A, et al. Clinical review 75: recent advances in pathogenesis, diagnosis, and management of acromegaly. J Clin Endocrinol Metab 1995; 80:3395. 46. Bates AS, Van't Hoff W, Jones JM, Clayton RN. An audit of outcome of treatment in acromegaly. Q J Med 1993; 86:293. 47. Rajasoorya C, Holdaway IM, Wrightson P, et al. Determinants of clinical outcome and survival in acromegaly. Clin Endocrinol (Oxf) 1994; 41:95. 48. Abosch A, Tyrrell JB, Lamborn KR, et al. Transsphenoidal microsurgery for growth hormone-secreting pituitary adenomas: initial outcome and long-term results. J Clin Endocrinol Metab 1998;

83:3411. 49. Orme SM, McNally RJ, Cartwright RA, Belchetz PE. Mortality and cancer incidence in acromegaly: a retrospective cohort study. United Kingdom Acromegaly Study Group. J Clin Endocrinol Metab 1998; 83:2730. 50. Swearingen B, Barker FGN, Katznelson L, et al. Long-term mortality after transsphenoidal surgery and adjunctive therapy for acromegaly. J Clin Endocrinol Metab 1998; 83:3419. 51. Melmed S. Tight control of growth hormone: an attainable outcome for acromegaly treatment. J Clin Endocrinol Metab 1998; 83:3409. 52. Giustina A, Barkan A, Casanueva FF, et al. Criteria for cure in acromegaly: a consensus statement. J Clin Endocrinol Metab 2000; 85:526. 53. Roelfsema F, van Dulken H, Frolich M. Long-term results of transsphenoidal pituitary microsurgery in 60 acromegalic patients. Clin Endocrinol (Oxf) 1985; 23:555. 54. Sheaves R, Jenkins P, Blackburn P, et al. Outcome of transsphenoidal surgery for acromegaly using strict criteria for surgical cure. Clin Endocrinol (Oxf) 1996; 45:407. 55. Yamada S, Aiba T, Takada K, et al. Retrospective analysis of long-term surgical results in acromegaly: preoperative and postoperative factors predicting outcome. Clin Endocrinol (Oxf) 1996; 45:291. 56. Colao A, Ferone D, Cappabianca P, et al. Effect of octreotide pretreatment on surgical outcome in acromegaly. J Clin Endocrinol Metab 1997; 82:3308. 57. Colao A, Merola B, Ferone D, Lombardi G. Acromegaly. J Clin Endocrinol Metab 1997; 82:2777. 58. Freda PU, Wardlaw SL, Post KD. Long-term endocrinological follow-up evaluation in 115 patients who underwent transsphenoidal surgery for acromegaly. J Neurosurg 1998; 89:353. 59. Buchfelder M, Brockmeier S, Fahlbusch R, et al. Recurrence following transsphenoidal surgery for acromegaly. Horm Res 1991; 35:113. 60. Losa M, Oeckler R, Schopohl J, et al. Evaluation of selective transsphenoidal adenomectomy by endocrinological testing and somatomedin-C measurement in acromegaly. J Neurosurg 1989; 70:561. 61. Lamberg BA, Kivikangas V, Vartianen J, et al. Conventional pituitary irradiation in acromegaly. Effect on growth hormone and TSH secretion. Acta Endocrinol 1976; 82:267. 62. Feek CM, McLelland J, Seth J, et al. How effective is external pituitary irradiation for growth hormone-secreting pituitary tumors? Clin Endocrinol (Oxf) 1984; 20:401. 63. Kliman B, Kjellberg RN, Swisher B, Butler W. Long-term effects of protonbeam therapy for acromegaly. In: Robbins RJ, Melmed S, eds. Acromegaly. New York: Plenum Press, 1986:221. 64. Speirs CJ, Reed PI, Morrison R, et al. The effectiveness of external beam radiotherapy for acromegaly is not affected by previous pituitary ablative treatments. Acta Endocrinol (Copenh) 1990; 122:559. 65. Chanson P, Grellier-Fouqueray P, Young J, et al. Comment et avec quelle efficacit traite-t-on l'acromgalie en 1997? Enqute transversale sur une population de 74 acromgales. (Abstract). Ann Endocrinol (Paris) 1997; 58(Suppl 2):2S69. 66. Thalassinos NC, Tsagarakis S, Ioannides G, et al. Megavoltage pituitary irradiation lowers but seldom leads to safe GH levels in acromegaly: a long-term follow-up study. Eur J Endocrinol 1998; 138:160. 67. Morange-Ramos I, Regis J, Dufour H, et al. Gamma-knife surgery for secreting pituitary adenomas. Acta Neurochir (Wien) 1998; 140:437. 68. Landolt AM, Haller D, Lomax N, et al. Stereotactic radiosurgery for recurrent surgically treated acromegaly: comparison with fractionated radiotherapy. J Neurosurg 1998; 88:1002. 69. Arosio M, Macchelli S, Rossi CM, et al. Effects of treatment with octreotide in acromegalic patientsa multicenter Italian study. Italian Multicenter Octreotide Study Group. Eur J Endocrinol 1995; 133:430. 70. Newman CB, Melmed S, George A, et al. Octreotide as primary therapy for acromegaly. J Clin Endocrinol Metab 1998; 83:3034. 71. Flgstad AK, Halse J, Bakke S, et al. Sandostatin LAR in acromegalic patients: long-term treatment. J Clin Endocrinol Metab 1997; 82:23. 72. Lancranjan I, Bruns C, Grass P, et al. Sandostatin LAR: a promising therapeutic tool in the management of acromegalic patients. Metabolism 1996; 45:67. 73. Stewart PM, Kane KF, Stewart SE, et al. Depot long-acting somatostatin analog (Sandostatin-LAR) is an effective treatment for acromegaly. J Clin Endocrinol Metab 1995; 80:3267. 74. Molitch M. Prolactinoma. In: Melmed S, ed. The pituitary. Boston: Blackwell Science, 1995:443. 75. Thorner M, Vance ML, Laws ER Jr, et al. The anterior pituitary. In: Wilson JD, Foster DW, Kronenberg HM, Larsen PR, eds. Williams textbook of endocrinology. Philadelphia: WB Saunders, 1998:249. 76. Vance ML, Evans WS, Thorner MO. Drugs five years later. Bromocriptine. Ann Intern Med 1984; 100:78. 77. Liuzzi A, Dallabonzana D, Oppizzi G, et al. Low doses of dopamine agonists in the long-term treatment of macroprolactinomas. N Engl J Med 1985; 313:656. 78. Molitch ME, Thorner MO, Wilson C. Management of prolactinomas. J Clin Endocrinol Metab 1997; 82:996. 79. Bevan JS, Webster J, Burke CW, Scanlon MF. Dopamine agonists and pituitary tumor shrinkage. Endocr Rev 1992; 13:220. 80. Molitch ME. Pregnancy and the hyperprolactinemic woman. N Engl J Med 1985; 312:1364. 81. Kupersmith MJ, Rosenberg C, Kleinberg D. Visual loss in pregnant women with pituitary adenomas. Ann Intern Med 1994; 121:473. 82. Gemzell C, Wang CF. Outcome of pregnancy in women with pituitary adenoma. Fertil Steril 1979; 31:363. 83. van't Verlaat JW, Croughs RJ, Brownell J. Treatment of macroprolactinomas with a new non-ergot, long-acting dopaminergic drug, CV 205-502. Clin Endocrinol (Oxf) 1990; 33:619. 84. Duranteau L, Chanson P, Lavoinne A, et al. Effect of the new dopaminergic agonist CV 205-502 on plasma prolactin levels and tumour size in bromocriptine-resistant prolactinomas. Clin Endocrinol (Oxf) 1991; 34:25. 85. Brue T, Pellegrini I, Gunz G, et al. Effects of the dopamine agonist CV 205-502 in human prolactinomas resistant to bromocriptine. J Clin Endocrinol Metab 1992; 74:577. 86. Vilar L, Burke CW. Quinagolide efficacy and tolerability in hyperprolactinaemic patients who are resistant to or intolerant of bromocriptine. Clin Endocrinol (Oxf) 1994; 41:821. 87. Rains CP, Bryson HM, Fitton A. Cabergoline: a review of its pharmacological properties and therapeutic potential in the treatment of hyperprolactinmia and inhibition of lactation. Drugs 1995; 49:255. 88. Webster J, Piscitelli G, Polli A, et al. A comparison of cabergoline and bro-mocriptine in the treatment of hyperprolactinemic amenorrhea. Caber-goline Comparative Study Group. N Engl J Med 1994; 331:904. 89. Biller BM, Molitch ME, Vance ML, et al. Treatment of prolactin-secreting macroadenomas with the once-weekly dopamine agonist cabergoline. J Clin Endocrinol Metab 1996; 81:2338. 90. Colao A, Di Sarno A, Sarnacchiaro F, et al. Prolactinomas resistant to standard dopamine agonists respond to chronic cabergoline treatment. J Clin Endocrinol Metab 1997; 82:876. 91. Colao A, Di Sarno A, Landi ML, et al. Long-term and low-dose treatment with cabergoline induces macroprolactinoma shrinkage. J Clin Endocrinol Metab 1997; 82:3574. 92. Feigenbaum SL, Downey DE, Wilson CB, Jaffe RB. Transsphenoidal pituitary resection for preoperative diagnosis of prolactin-secreting pituitary adenoma in women: long term follow-up. J Clin Endocrinol Metab 1996; 81:1711. 93. Thomson JA, Davies DL, McLaren EH, Teasdale GM. Ten year follow up of microprolactinoma treated by transsphenoidal surgery. BMJ 1994; 309:1409. 94. Serri O, Hardy J, Massoud F. Relapse of hyperprolactinemia revisited. N Engl J Med 1993; 329:1357. 95. Grossman A, Cohen BL, Charlesworth M, et al. Treatment of prolactinomas with megavoltage radiotherapy. BMJ 1984; 288:1105. 96. Nabarro JD. Pituitary prolactinomas. Clin Endocrinol (Oxf) 1982; 17:129. 97. Tsagarakis S, Grossman A, Plowman PN, et al. Megavoltage pituitary irradiation in the management of prolactinomas: long-term follow-up. Clin Endocrinol (Oxf) 1991; 34:399. 98. Miller JW, Crapo L. The medical treatment of Cushing's syndrome. Endocr Rev 1993; 14:443. 99. Bertagna X, Raux-Demay MC, Guilhaume B, et al. Cushing's disease. In: Melmed S, ed. The pituitary. Boston: Blackwell Science, 1995:478. 100. Arnott RD, Pestell RG, McKelvie PA, et al. A critical evaluation of trans-sphenoidal surgery in the treatment of Cushing's disease: prediction of outcome. Acta Endocrinol (Copenh) 1990; 123:423. 101. Orth DN. Cushing's syndrome. N Engl J Med 1995; 332:791. 102. Bochicchio D, Losa M, Buchfelder M. Factors influencing the immediate and late outcome of Cushing's disease treated by transsphenoidal surgery: a retrospective study by the European Cushing's Disease Survey Group. J Clin Endocrinol Metab 1995; 80:3114. 103. Guilhaume B, Bertagna X, Thomsen M, et al. Transsphenoidal pituitary surgery for the treatment of Cushing's disease: results in 64 patients and long term follow-up studies. J Clin Endocrinol Metab 1988; 66:1056. 104. Mampalam TJ, Tyrrell JB, Wilson CB. Transsphenoidal microsurgery for Cushing disease. A report of 216 cases. Ann Intern Med 1988; 109:487. 105. Burke CW, Adams CB, Esiri MM, et al. Transsphenoidal surgery for Cushing's disease: does what is removed determine the endocrine outcome? Clin Endocrinol (Oxf) 1990; 33:525. 106. Post KD, Habas JE. Comparison of long term results between prolactin secreting adenomas and ACTH secreting adenomas. Can J Neurol Sci 1990; 17:74. 107. Tindall GT, Herring CJ, Clark RV, et al. Cushing's disease: results of trans-sphenoidal microsurgery with emphasis on surgical failures. J Neurosurg 1990; 72:363. 108. Robert F, Hardy J. Cushing's disease: a correlation of radiological, surgical and pathological findings with therapeutic results. Pathol Res Pract 1991; 187:617. 109. Tahir AH, Sheeler LR. Recurrent Cushing's disease after transsphenoidal surgery. Arch Intern Med 1992; 152:977. 110. Trainer PJ, Lawrie HS, Verhelst J, et al. Transsphenoidal resection in Cushing's disease: undetectable serum cortisol as the definition of successful treatment. Clin Endocrinol (Oxf) 1993; 38:73. 111. Ram Z, Nieman LK, Cutler GB Jr, et al. Early repeat surgery for persistent Cushing's disease. J Neurosurg 1994; 80:37. 112. Styne DM, Grumbach MM, Kaplan SL, et al. Treatment of Cushing's disease in childhood and adolescence by transsphenoidal microadenomec-tomy. N Engl J Med 1984; 310:889. 113. Dyer EH, Civit T, Visot A, et al. Transsphenoidal surgery for pituitary adenomas in children. Neurosurgery 1994; 34:207. 114. Leinung MC, Kane LA, Scheithauer BW, et al. Long term follow-up of transsphenoidal surgery for the treatment of Cushing's disease in childhood. J Clin Endocrinol Metab 1995; 80:2475. 115. Magiakou MA, Mastorakos G, Oldfield EH, et al. Cushing's syndrome in children and adolescents. Presentation, diagnosis, and therapy. N Engl J Med 1994; 331:629. 116. Howlett TA, Plowman PN, Wass JA, et al. Megavoltage pituitary irradiation in the management of Cushing's disease and Nelson's syndrome: long-term follow-up. Clin Endocrinol (Oxf) 1989; 31:309. 117. Estrada J, Boronat M, Mielgo M, et al. The long-term outcome of pituitary irradiation after unsuccessful transsphenoidal surgery in Cushing's disease. N Engl J Med 1997; 336:172. 118. Tyrrell JB, Wilson CB. Cushing's disease. Therapy of pituitary adenomas. Endocrinol Metab Clin North Am 1994; 23:925. 119. Chanson P, Weintraub BD, Harris AG. Octreotide therapy for thyroid-stimulating hormone-secreting pituitary adenomas. A follow-up of 52 patients. Ann Intern Med 1993; 119:236. 120. Gancel A, Vuillermet P, Legrand A, et al. Effects of a slow-release formulation of the new somatostatin analogue lanreotide in TSH-secreting pituitary adenomas. Clin Endocrinol (Oxf) 1994; 40:421. 121. Beck-Peccoz P, Brucker-Davis F, Persani L, et al. Thyrotropin-secreting pituitary adenomas. Endocr Rev 1996; 17:610. 122. Warnet A, Timsit J, Chanson P, et al. The effect of somatostatin analogue on chiasmal dysfunction from pituitary macroadenomas. J Neurosurg 1989; 71:687. 123. Warnet A, Harris AG, Renard E, et al. A prospective multicenter trial of octreotide in 24 patients with visual defects caused by nonfunctioning and gonadotropin-secreting pituitary adenomas. Neurosurgery 1997; 41:786. 124. Molitch ME. Evaluation and treatment of the patient with a pituitary incidentaloma. J Clin Endocrinol Metab 1995; 80:3. 125. Ebersold MJ, Quast LM, Laws ER Jr, et al. Long-term results in transsphenoidal removal of nonfunctioning pituitary adenomas. J Neurosurg 1986; 64:713. 126. Harris PE, Afshar F, Coates P, et al. The effects of transsphenoidal surgery on endocrine function and visual fields in patients with functionless pituitary tumours. Q J Med 1989; 71:417. 127. Comtois R, Beauregard H, Somma M, et al. The clinical and endocrine outcome to transphenoidal microsurgery of non secreting pituitary adenomas. Cancer 1991; 68:860. 128. Sassolas G, Trouillas J, Treluyer C, Perrin G. Management of non-functioning pituitary adenomas. Acta Endocrinol 1993; 129(Suppl 1):21. 129. Chanson P. Les adnomes hypophysaires non-fonctionnels. Paris: John Libbey Eurotext, 1998. 130. Hayes TP, Davis RA, Raventos A. The treatment of pituitary chromophobe adenomas. Radiology 1971; 98:149. 131. Ciric I, Mikhael M, Stafford T, et al. Transsphenoidal microsurgery of pituitary macroadenomas with long-term follow-up results. J Neurosurg 1983; 59:395. 132. Vlahovitch B, Reynaud C, Rhiati J, et al. Treatment and recurrences in 135 pituitary adenomas. Acta Neurochir 1988; 42(Suppl):120. 133. Jaffrain-Rea ML, Derome P, Bataini JP, et al. Influence of radiotherapy on long-term relapse in clinically non-secreting pituitary adenomas. A retrospective study (19701988). Eur J Med 1993; 2:398. 134. Bradley KM, Adams CBT, Potter CPS, et al. An audit of selected patients with non-functioning pituitary adenoma treated by transsphenoidal surgery without irradiation. Clin Endocrinol (Oxf) 1994; 41:655. 135. Gittoes NJL, Bates AS, Tse W, et al. Radiotherapy for non-functioning pituitary tumours. Clin Endocrinol (Oxf) 1998; 48:331. 136. Rosen T, Bengtsson A-G. Premature mortality due to cardiovascular disease in hypopituitarism. Lancet 1990; 336:285. 137. Bates AS, Van't Hoff W, Jones PJ, Clayton RN. The effects of hypopituitarism on life expectancy. J Clin Endocrinol Metab 1996; 81:1169. 138. Blow B, Hagmar L, Mikoczy Z, et al. Increased cerebrovascular mortality in patients with hypopituitarism. Clin Endocrinol (Oxf) 1997; 46:75. 139. Liuzzi A, Dallabonzana D, Oppizzi G, et al. Is there a real medical treatment for non-secreting pituitary adenoma? In: Faglia G, Beck-Peccoz P, Ambrosi P, Travaglini P, Spada A, eds. Pituitary adenomas: new trends in basic and clinical research. Amsterdam: Excerpta Medica, 1991:383. 140. Chanson P, Schaison G. Pituitary apoplexy caused by GnRH-agonist treatment revealing gonadotroph adenoma. J Clin Endocrinol Metab 1995; 80:2267. 141. McGrath GA, Goncalves RJ, Voupa JK, et al. New technique for quantitation of pituitary adenoma size: use in evaluating treatment of gonadotroph adenomas with a gonadotropin-releasing hormone antagonist. J Clin Endocrinol Metab 1993; 76:1363. 142. Etxabe J, Vazquez JA. Morbidity and mortality in Cushing's disease: an epidemiological approach. Clin Endocrinol (Oxf) 1994; 40:479. 143. Carroll PV, Christ ER, Bengtsson BA, et al. Growth hormone deficiency in adulthood and the effects of growth hormone replacement: a review. Growth Hormone Research Society Scientific Committee. J Clin Endocrinol Metab 1998; 83:382.

144. Orrego JJ, Barkan AL. Pituitary disorders. Drug treatment options. Drugs 2000; 59:93. 145. Bates AS, Farrell WE, Bicknell EJ, et al. Allelic deletion in pituitary adenomas reflects aggressive biological activity and has potential value as a prognostic marker. J Clin Endocrinol Metab 1997; 82:818.

CHAPTER 25 PHYSIOLOGY OF VASOPRESSIN, OXYTOCIN, AND THIRST Principles and Practice of Endocrinology and Metabolism

CHAPTER 25 PHYSIOLOGY OF VASOPRESSIN, OXYTOCIN, AND THIRST

GARY L. ROBERTSON Anatomy of the Neurohypophysis Gross Features Microscopic Features Chemistry Biosynthesis and Release of Vasopressin and Oxytocin Regulation of Vasopressin Secretion Regulation of Oxytocin Secretion Distribution and Clearance of Vasopressin and Oxytocin Biologic Actions Vasopressin Oxytocin Thirst Mechanism Water Homeostasis Volume, Composition, Distribution, and Balance Osmoregulation Hemodynamic Influences Chapter References

ANATOMY OF THE NEUROHYPOPHYSIS

GROSS FEATURES The neurohypophysis, an extension of the ventral hypothalamus, attaches to the dorsal and caudal surface of the adenohypophysis1 (Fig. 25-1). In adult men and women, it weighs ~100 mg. It is divided by the diaphragma sellae into an upper part, called the infundibulum or median eminence, and a lower part, known as the infundibular process or pars nervosa. The two parts are supplied with blood by branches from the superior and inferior hypophyseal arteries. In the pars nervosa, the arterioles break up into localized capillary networks that drain directly into the jugular vein through the sellar, cavernous, and lateral venous sinuses. In the infundibulum, the primary capillary networks coalesce into another system, the portal veins, which perfuse the adenohypo-physis before discharging into the systemic circulation.

FIGURE 25-1. The neurohypophysis and its principal regulatory afferents are illustrated. (pvn, paraventricular nucleus; or, osmoreceptor; son, supraoptic nucleus; oc, optic chiasm; ds, diaphragma sellae; ah, adenohypophysis; nh, neurohypophysis; br, volume and baroreceptors; ap, area postrema [emetic center]; nts, nucleus tractus solitarii.) (From Robertson GL. Disorders of the posterior pituitary. In: Stein JH, ed. Internal medicine. Boston: Little, Brown and Company, 1983:1728.)

MICROSCOPIC FEATURES On microscopic examination, the neurohypophysis appears as a densely interwoven network of capillaries, pituicytes, and non-myelinated nerve fibers containing many electrondense neuro-secretory granules. These neurosecretory neurons terminate as bulbous enlargements on capillary networks located at all levels of the neurohypophysis, including the stalk and infundibulum. The vasopressin-containing neurosecretory neurons that form the pars nervosa originate primarily in the supraoptic nuclei2 and probably provide most, if not all, of the vasopressin and oxytocin in the peripheral plasma. Those that terminate in the median eminence originate primarily in the paraventricular or other hypothalamic nuclei,2 probably releasing their hormones into the portal blood supply of the anterior pituitary. Other, smaller groups of vasopressinergic neurons project from the paraventricular nucleus to the medulla, amygdala, spinal cord, and the walls of the lateral and third ventricles.2 The latter may secrete directly into the cerebrospinal fluid (CSF).3 Oxytocinergic cell bodies appear to be less numerous than those containing vasopressin.2 They are found primarily in discrete areas in or around the paraventricular nuclei and, to a lesser extent, the supraoptic nuclei. Most oxytocinergic neurons project to the pars nervosa, but many also terminate in the organum vasculosum or the median eminence. In addition, a relatively large paraventricular division runs parallel to the vasopressinergic fibers that connect to the medulla and spinal cord (see Chap. 8 and Chap. 9).

CHEMISTRY
Vasopressin and oxytocin are the only hormones known to be secreted in significant amounts by the neurohypophysis. As first shown by du Vigneaud and coworkers,4 the two hormones have similar structures (Fig. 25-2), being nonapeptides that contain six-membered disulfide rings and the same amino acid residues in seven of the nine positions. They are stored in neurosecretory granules as insoluble complexes with specific carrier proteins known as neurophysins.5 The neurophysins associated with vasopressin and oxytocin also have similar structures, each having ~100 amino-acid residues with extensive areas of homology. Binding of the hormones to their neurophysins has a pH optima (5.25.8) and dissociation constant (Kd) (~5 105) that favor association in neurosecretory granules but ensures almost complete dissociation in plasma and other body fluids.

FIGURE 25-2. Structure of oxytocin, vasopressin, and 1-deamino-(8-D-arginine)-vasopressin (DDAVP). (S, sulfur.)

BIOSYNTHESIS AND RELEASE OF VASOPRESSIN AND OXYTOCIN Vasopressin and oxytocin are synthesized through protein precursors encoded by single-copy genes that are located near each other on chromosome 20 in humans6,7,8 and 9 (Fig. 25-3). The gene for the vasopressin precursor, known as propressophysin or vasopressin neurophysin II, is ~2-kb long. It contains three exons that encode, respectively, (a) a signal peptide, vasopressin, and the variable amino-terminal end of neurophysin; (b) the highly conserved middle portion of neurophysin; and (c) the variable, carboxy-terminal end of neurophysin and copeptin, a glycosy-lated peptide of unknown function. The gene for the oxytocin precursor is similar except that exon C is shorter and codes only for the variable carboxy terminus of neurophysin and a single histidine residue. The copeptin moiety is absent.

FIGURE 25-3. Structure of the vasopressin and oxytocin prohormones and the genes that encode them. (AVP, arginine vasopressin; G, glycosylation.)

In humans and other mammals, the vasopressin and oxytocin genes are expressed in different magnocellular neurons. After transcription and translation in cell bodies within the supraoptic and paraventricular nuclei, the preprohormones are translocated into the endoplasmic reticulum, where the signal peptide is removed and the prohormones fold and self-associate before moving through the Golgi apparatus and on into the neurosecretory granules. There they are transported down the axons and cleaved into the intact hormone, neurophysin, and, in the case of vasopressin, copeptin. This process is critically dependent on correct folding and self-association of the prohormone in the endoplasmic reticulum, because genetic mutations predicted to alter amino acids important for correct folding severely disrupt transport and destroy the neurons, producing the autosomal dominant form of familial neurohypophyseal diabetes insipidus. Release of the hormones and their associated neurophysins occurs via a calcium-dependent exocytotic process similar to that described for other neurosecretory systems.10 An electrical impulse propagated along the neuron depolarizes the cell membrane, causing an influx of calcium, fusion of secretory granules with the outer cell membranes, and extrusion of their contents. REGULATION OF VASOPRESSIN SECRETION OSMOTIC REGULATION The secretion of vasopressin is influenced by a number of variables.11,12,13,14 and 15 The most important under physiologic conditions is the effective osmotic pressure of plasma. This influence is mediated by specialized cells called osmoreceptors. These osmoreceptors appear to be concentrated in the anterolateral hypothalamus,16,17 and 18 in an area that is near, but separate from, the supraoptic nuclei (see Fig. 25-1). This area is supplied with blood by small perforating branches of the anterior cerebral or communicating arteries.1 The basic structure, modus operandi, and organization of the individual osmoreceptors have not yet been determined. However, the system as a whole functions like a discontinuous or set-point receptor (Fig. 25-4). Thus, at plasma osmolalities below a certain minimum or threshold level, plasma vasopressin is suppressed to low or undetectable concentrations. Above this set-point, plasma vasopressin rises steeply in direct proportion to plasma osmolality. The slope of the relationship indicates that a change in plasma osmolality of only 1% alters plasma vasopressin by an average of 1 pg/mL, an amount sufficient to significantly affect the urinary concentration and flow (Fig. 25-5).

FIGURE 25-4. The relationship of thirst and plasma vasopressin to plasma osmolality in healthy adults under varying conditions of water balance. (From Robertson GL. Thirst and vasopressin function in normal and disordered states of water balance. J Lab Clin Med 1983; 101:351.)

FIGURE 25-5. The relation of urine osmolality to plasma vasopressin in healthy adults under varying conditions of water balance. (AVP, arginine vasopressin.) (From Robertson GL. Thirst and vasopressin function in normal and disordered states of water balance. J Lab Clin Med 1983; 101:351.)

The sensitivity and set of the osmoregulatory system varies considerably among healthy adults. The interindividual differences in sensitivity are large (up to 10-fold), are constant over long periods of time, and appear to be genetically determined.19 However, they can be altered slightly by a variety of pharmacologic or pathologic influences.20 The interindividual differences in setpoint are not as large (275290 mOsm/kg) or as constant over time but also appear to have a significant genetic component.19 They are more subject to alteration by a variety of physiologic factors, including posture, pregnancy, and the phase of the menstrual cycle,20,21 all of

which lower the osmotic threshold or setpoint. The sensitivity of the osmoregulatory system also varies for different plasma solutes (Fig. 25-6). Sodium and its anions, which normally contribute >95% of the osmotic pressure of plasma, are the most potent solutes known in terms of their capacity to stimulate vasopressin release.20 Certain sugars, such as sucrose and mannitol, appear to be nearly as potent. However, a rise in plasma osmolality secondary to urea or glucose causes little or no increase in plasma vasopressin in healthy adults or animals. These differences in response to various plasma solutes are independent of any recognized nonosmotic influence and probably reflect some property of the osmoregulatory mechanism. Precisely how the osmoreceptor discriminates so effectively between different kinds of plasma solutes still is unresolved. According to current concepts, the signal that stimulates the osmoreceptor is an osmotically induced decrease in the water content of the cell. If this hypothesis is correct, the capacity of a given solute to stimulate vasopressin secretion should be inversely related to the rate at which it passes from plasma into the osmoreceptor. This concept agrees well with the observed inverse relationship between the stimulatory effect of certain solutes, such as sodium, mannitol, and glucose, and the rate at which they penetrate the bloodbrain barrier. However, urea is an exception because it penetrates the bloodbrain barrier slowly yet is a relatively weak stimulus for thirst and vasopressin. This singular disparity suggests that most, if not all, of the osmoreceptors are located outside of the bloodbrain barrier and that another factor (most likely, the permeability of the osmoreceptor cell itself) determines the solute specificity of the system.

FIGURE 25-6. The relationship of plasma vasopressin to plasma osmolality in healthy adults during the infusion of hypertonic solutions of different solutes. (From Robertson GL. Disorders of the posterior pituitary. In: Stein JH, ed. Internal medicine. Boston: Little, Brown and Company, 1983:1728.)

The solute specificity of the osmoregulatory system is also subject to change. Thus, its sensitivity to stimulation by glucose increases when insulin is deficient.20 This change probably results from decreased permeability of the osmoreceptors to glucose and indicates that these cells are insulin dependent. It may also explain the hyperdipsia and at least part of the hyper-vasopressinemia that occurs in many patients with uncontrolled type 1 diabetes mellitus. HEMODYNAMIC REGULATION The secretion of vasopressin is also affected by changes in blood volume, pressure, or both.11,12,14,15,22 These hemodynamic influences are mediated largely, if not exclusively, by neurogenic afferents that arise in pressure-sensitive receptors in the heart and large arteries and that travel by way of the vagal and glos-sopharyngeal nerves to primary synapses in the nucleus tractus solitarius in the brainstem (see Fig. 25-1). From there, postsyn-aptic pathways project to the region of the paraventricular and supraoptic nuclei. At least one of the links in the afferent chain for volume control involves opioid receptors in the lateral parabrachial nucleus, because administration of selective as well as nonselective antagonists in this area almost totally inhibits the vasopressin response to an acute hypovolemic stimulus.23,24 and 25 The functional properties of the baroregulatory system also differ from those of the osmoregulatory mechanism (Fig. 25-7). In healthy adults and animals, acutely lowering blood pressure increases plasma vasopressin in proportion to the degree of hypotension achieved. However, this stimulus-response relationship follows a distinctly exponential pattern. Thus, small decreases in blood pressure of 5% to 10% usually have little effect on plasma vasopressin, whereas decreases in blood pressure of 20% to 30% result in hormone levels many times those required to produce maximal antidiuresis. The vasopressin response to changes in blood volume has not been well defined but appears to be quantitatively and qualitatively similar to the vasopressin response to changes in blood pressure. An acute rise in blood volume or pressure appears to suppress vasopressin secretion.

FIGURE 25-7. Schematic representation of the relationship between plasma vasopressin and percentage of change in plasma osmolality, blood volume, or blood pressure in healthy adults. (From Robertson GL. Diseases of the posterior pituitary. In: Felig P, Baxter J, Brodus A, Frohman L, eds. Endocrinology and metabolism. New York: McGraw-Hill, 1981:251.)

The failure of small changes in blood volume and pressure to alter vasopressin secretion contrasts markedly with the extraordinary sensitivity of the osmoregulatory system (see Fig. 25-7). The recognition of this difference is essential for understanding the relative contribution of each system to the control of the hormone under both physiologic and pathologic conditions. Because day-to-day variations of total body water rarely exceed 2% to 3%, their effect on vasopressin secretion must be mediated largely, if not exclusively, by the osmoregulatory system. For this reason, patients with destruction of the osmoreceptor exhibit a markedly subnormal vasopressin response to changes in water balance, even though baroregulatory mechanisms are completely intact. On the other hand, baroregulatory input appears to mediate the effects of a large number of pharmacologic agents and pathologic conditions (Table 25-1). Among these are diuretics, isoproterenol, nicotine, prostaglandins, nitroprusside, trimethaphan camsylate, histamine, morphine, and bradykinin, all of which stimulate vasopressin secretion, at least in part, by lowering blood volume or pressure. In addition, norepinephrine and aldosterone suppress vasopressin secretion by raising blood volume, pressure, or both. In addition, upright posture, sodium depletion, congestive failure, cirrhosis, and nephrosis stimulate vasopressin secretion, probably by reducing total or effective blood volume, whereas orthostatic hypotension, vasovagal reactions, and other forms of syncope markedly stimulate secretion of the hormone by reducing blood pressure. This list probably could be extended to include almost every other hormone, drug, and condition known to affect blood volume or pressure. The only recognized exception is a form of orthostatic hypotension associated with the loss of afferent baroregulatory function.26

TABLE 25-1. Variables That Influence Vasopressin Secretion

Changes in blood volume or pressure that are large enough to affect vasopressin secretion do not necessarily interfere with osmoregulation of the hormone.12,13,14 and 15 Instead, they appear to act by shifting the setpoint of the system in such a way as to increase or decrease the effect on vasopressin of a given osmotic stimulus (Fig. 25-8). This kind of interaction ensures that, even in the presence of hemodynamic stimuli, the capacity to osmoregulate is not lost. How this integration occurs is unknown, but it probably involves one or more interneurons that link the osmoreceptor to neurosecretory neurons.

FIGURE 25-8. The relationship between plasma vasopressin and plasma osmolality in the presence of different states of blood volume or pressure. The oblique heavy line, labeled N, represents normovolemic, normotensive conditions. Lines labeled with negative numbers (to the left) or positive numbers (to the right) indicate, respectively, the percentage of decrease or increase in blood volume or pressure. (From Robert-son GL. Disorders of the posterior pituitary. In: Stein JH, ed. Internal medicine. Boston: Little, Brown and Company, 1983:1728.)

EMESIS Nausea is an extremely potent stimulus for vasopressin secretion in humans.15 The pathway that mediates this effect probably involves the chemoreceptor trigger zone in the area postrema of the medulla (see Fig. 25-1). It can be activated by a variety of drugs and conditions, including apomorphine, morphine, nicotine, alcohol, and motion sickness.15 Its effect on vasopressin secretion is instantaneous and extremely potent (Fig. 25-9). Increases in vasopressin of 100 to 1000 times basal levels are not unusual, even when the nausea is transient and unaccompanied by vomiting or changes in blood pressure. Pretreatment with fluphenazine, haloperidol, or promethazine in doses sufficient to prevent nausea completely abolishes the vasopressin response.27 The inhibitory effect of these dopamine antagonists is specific for emetic stimuli because they do not alter the vasopressin response to hyperosmolality, hypovolemia, or hypotension.

FIGURE 25-9. Effect of nausea on plasma vasopressin in a healthy adult. (APO, apomorphine; PRA, plasma renin activity.) (From Robert-son GL. The regulation of vasopressin function in health and disease. Recent Prog Horm Res 1977; 33:333.)

Water loading blunts, but does not abolish, the effect of nausea on vasopressin release, a finding which suggests that osmotic and emetic influences interact in a manner similar to osmotic and hemodynamic pathways.27 Emetic stimuli probably mediate many pharmacologic and pathologic effects on vasopressin secretion. For example, emetic stimulation may be at least partially responsible for the increase in vasopressin secretion that has been observed with intravenous administration of cyclophosphamide, vasovagal reactions, ketoacidosis, acute hypoxia, and motion sickness. Because nausea and vomiting are frequent side effects of many other drugs and diseases, additional examples of emetically mediated vasopressin secretion doubtlessly could be demonstrated. OTHER STIMULI Hypoglycemia. Acute hypoglycemia is a relatively weak stimulus for vasopressin release.28 The receptor and pathway that mediate this effect are unknown but must be separate from those of other recognized stimuli because hypoglycemia stimulates vasopressin secretion in patients who have lost the capacity to respond selectively to osmotic, hemodynamic, or emetic stimuli. However, the vasopressin response to hypoglycemia is accentuated by dehydration and is abolished by water loading. Thus, glucopenic stimuli probably act in concert with osmotic influences, even though the osmoreceptors are unnecessary for the response. Vasopressin release may be triggered by an intracellular deficiency of glucose or one of its metabolites because 2-deoxyglucose is also an effective stimulus.29 Angiotensin. The renin angiotensin system has also been implicated in the control of vasopressin secretion.30 The precise site and mechanism of action have not been defined, but central receptors are likely to be involved because angiotensin is most effective when injected directly into brain ventricles or cranial arteries. The levels of plasma renin or angiotensin required to stimulate vasopressin release have not been determined but probably are high. When administered intravenously, pressor doses of angiotensin increase plasma vasopressin twofold to fourfold. The magnitude of the vasopressin response may depend on the concurrent osmotic stimulus, because angiotensin increases the sensitivity of the osmoregulatory system.31 This dependency on osmotic influences resembles that seen with glucopenic stimuli and may account for the inconsistency of the vasopressin response to exogenous angiotensin. STRESS, TEMPERATURE, AND HYPOXIA Nonspecific stress caused by pain, emotion, or physical exercise has long been thought to cause the release of vasopressin.32 However, this effect now appears likely to be secondary to other stimuli, such as hypotension or nausea, which usually accompanies stress-induced vasovagal reactions. In the absence of hypotension or

nausea, pain sufficient to stimulate the pituitary-adrenal axis has no effect on vasopressin secretion in humans.33 Acute hypoxia or hypercapnia also stimulates vasopressin release.34 In conscious humans, however, the stimulatory effect of moderate hypoxia is inconsistent and appears to occur only in subjects who develop nausea or hypotension.35 Severe hypoxia probably has a greater effect on vasopressin secretion and may be responsible for the osmotically inappropriate hormonal elevations noted in some patients with acute respiratory failure. Whether or not hypercapnia has similar effects on vasopressin secretion in conscious persons is not known. OROPHARYNGEAL INFLUENCES Vasopressin secretion is inhibited by drinking before any detectable decrease in plasma osmolality is seen.36 This inhibition can override a moderately strong osmotic stimulus but is not sustained unless it is followed by a prompt decline in plasma osmolality. The mechanism has not been determined, but it probably involves some kind of oropharyngeal receptor. OTHER HORMONES AND DRUGS Many hormones and drugs influence vasopressin secretion.37 Those that have a stimulatory effect include acetylcholine, nicotine, apomorphine, morphine (high doses), epinephrine, isoproterenol, histamine, bradykinin, prostaglandins, b-endorphin, intravenous cyclophosphamide, vincristine, insulin, 2-deoxy-glucose, angiotensin, lithium, and possibly chlorpropamide and clofibrate. Those that have an inhibitory effect include norepinephrine, fluphenazine, haloperidol, promethazine, oxilorphan, butorphanol, morphine (low doses), alcohol, carbamazepine, glucocorticoids, clonidine hydrochloride, muscimol, and possibly phenytoin. Many stimulants, such as isoproterenol, nicotine, and high doses of morphine, undoubtedly act by lowering blood pressure or producing nausea. Others, such as substance P, prostaglandin, endorphin, and other opioids, also probably exert their influence by one or both of the same mechanisms. Insulin and 2-deoxyglucose appear to act by producing intracellular glucopenia, whereas angiotensin has an undefined but probably independent central effect. Vincristine may act by exerting a direct effect on the neurohypophysis or on peripheral neurons involved in the regulation of vasopressin secretion. Lithium, which antagonizes the antidiuretic effect of vasopressin, also increases secretion of the hormone. This effect is independent of changes in water balance and appears to result from an increase in sensitivity of the osmoregulatory system. The stimulatory effects of chlorpropamide and clofibrate are still controversial. Carbamazepine inhibits vasopressin secretion by diminishing the sensitivity of the osmoregulatory system. This effect occurs independently of changes in blood volume, blood pressure, or blood glucose levels and suggests that the ability of carbamazepine to produce antidiuresis in patients with neurogenic diabetes insipidus is the result of action on the kidney. Vasopressor drugs, such as norepinephrine, inhibit vasopressin secretion indirectly by raising arterial pressure. Dopaminergic antagonists, such as fluphenazine, haloperidol, and promethazine, probably act by suppressing the emetic center because they inhibit the vasopressin response to emetic stimuli only, not to osmotic or hemodynamic stimuli. In low doses, a variety of opioids, including morphine, butorphanol, and oxilorphan, inhibit vasopressin secretion, apparently by increasing the osmotic threshold for vasopressin release. The inhibitory effect of alcohol may be mediated by endogenous opiates; this effect also may be attributable to an elevation in the osmotic threshold for vasopressin release and can be blocked in part by treatment with naloxone hydrochloride. Other drugs that can inhibit vasopressin secretion include clonidine, which appears to act through both central and peripheral adrenoreceptors, and muscimol, which is postulated to act as a g-aminobutyric acid antagonist. Vasopressin and oxytocin may also exert a feedback effect, inhibiting or facilitating their own secretion. In the case of vasopressin, feedback inhibition occurs after systemic or central administration of relatively large doses of the hormone. REGULATION OF OXYTOCIN SECRETION In humans, the only stimulus known to reproducibly increase plasma oxytocin is suckling or other stimulation of the nipple in lactating women.38 This stimulus may also cause the release of oxytocin in nonlactating women, but the effect is less consistent. No recognized stimulus has been found for oxytocin secretion in men. In rats, but not in humans, oxytocin secretion is induced by osmotic, hemodynamic, and emetic stimuli, which indicates that this hormone is regulated quite differently in the two species.

DISTRIBUTION AND CLEARANCE OF VASOPRESSIN AND OXYTOCIN

In healthy adults, vasopressin and oxytocin distribute rapidly into a space roughly equivalent in volume to the extracellular compartment.12,39 This initial mixing phase has a half-time of 4 to 8 minutes and is virtually complete in 10 to 15 minutes. This rapid mixing phase is followed by a second, slower decline that probably corresponds to the metabolic or irreversible phase of clearance. The half-time of the metabolic phase varies considerably from person to person but is in the range of 10 to 20 minutes. The metabolic clearance rate determined by steady-state as well as nonsteady-state methods is largely independent of the plasma concentration within the physiologic range (ranging from 5 to 20 mL/kg per minute for vasopressin12,39 and from 10 to 23 mL/kg per minute for oxytocin).40 In pregnant women, the metabolic clearance rate of vasopressin is increased threefold to fourfold.21 Many tissues have the capacity to inactivate vasopressin in vitro, but most metabolism in vivo probably occurs in the liver and kidney. The plasma of pregnant women contains an enzyme that is capable of rapidly degrading the hormones in vitro, and it may also be active in vivo.21 Vasopressin and oxytocin are also excreted in urine, but the amounts are generally <10% of the total clearance.12,40,41 The mechanisms involved in the excretion of vasopressin probably involve filtration at the glomerulus and variable reabsorption at one or more sites along the tubule. The latter process may be linked in some way to the handling of sodium in the proximal nephron because the urinary clearance of vasopressin varies by as much as 20-fold in a direct relationship with solute clearance. Consequently, measurements of urinary vasopressin do not provide a reliable index of changes in plasma vasopressin unless glomerular filtration and solute clearance are normal. The dependence of urinary oxytocin excretion on solute clearance has not been determined but is probably similar. Vasopressin is also secreted into CSF3 and the portal venous system of the anterior pituitary.2 The concentration of vasopressin in the lumbar cistern is usually lower than that in plasma, but the two values tend to change in a parallel manner, a finding that suggests that they are subject to most, if not all, of the same regulatory influences. The two compartments must receive the hormone from different groups of neurons, however, because patients with neurogenic diabetes insipidus often have normal or increased CSF concentrations of vasopressin. The concentration of vasopressin in adenohypophyseal portal blood is much higher than that in peripheral veins but appears to be subject to some of the same regulatory influences (e.g., hypotension and hypovolemia).

BIOLOGIC ACTIONS
VASOPRESSIN RENAL ACTION The most important action of vasopressin is to conserve body water by reducing the rate of urinary, solute-free water excretion.39 This antidiuretic effect is achieved by promoting the reabsorption of solute-free water from urine as it passes through the distal or collecting tubules of the kidney (Fig. 25-10) (see Chap. 206). In the absence of vasopressin, the membranes lining this portion of the nephron are impermeable to water as well as to solutes. Hence, hypotonic filtrate formed in the more proximal part of the neph-ron passes unmodified through the distal tubule and collecting duct. In this condition, which is known as water diuresis, urine osmolality and flow in a healthy adult usually approximate 40 to 60 mOsm/kg and 15 to 20 mL per minute, respectively. In the presence of vasopressin, the hydroosmotic permeability of the distal and collecting tubules increases, which allows water to back-diffuse down the osmotic gradient that normally exists between tubular fluid and the isotonic or hypertonic milieu of the renal cortex and medulla. Because water is reabsorbed without solute, the urine that remains has an increased osmotic pressure as well as a decreased volume or flow rate. The degree of urinary concentration is proportional to the plasma vasopressin concentration, and in healthy adults, it is usually maximal at hormone concentrations of 5 pg/mL or less (see Fig. 25-5).

FIGURE 25-10. Schematic representation of the effect of vasopressin (AVP) on the formation of urine by the nephron. The osmotic pressure of tissue and tubular fluid is indicated by the density of the shading. The numbers within the lumen of the nephron indicate typical rates of flow in milliliters per minute. Arrows indicate reabsorption of sodium (Na) or water (H2O) by active (solid) or passive (broken) processes. Note that vasopressin acts only on the distal nephron, where it increases the hydroosmotic permeability of tubular membranes. The fluid that reaches this part of the nephron normally amounts to 10% to 15% of the total filtrate and is hypotonic, owing to selective reabsorption of sodium in the ascending limb of the Henle loop. In the absence of vasopressin, the membranes of the distal nephron remain relatively impermeable to water as well as to solute, and the fluid issuing from the Henle loop is excreted essentially unmodified as urine. With maximum vasopressin action, all but 5% to 10% of the water in this fluid is reabsorbed passively down the osmotic gradient that normally exists with the surrounding tissue. (From Robertson GL. Diseases of the posterior pituitary. In: Felig P, Baxter J, Brodus A, Frohman L, eds. Endocrinology and metabolism. New York: McGraw-Hill, 1986:351.)

The effect of vasopressin on urinary concentration and flow can be influenced markedly by changes in the volume of filtrate presented to the distal tubule. If the intake of salt is high, or if a poorly reabsorbed solute, such as mannitol, urea, or glucose, is filtered in increased amounts, the resultant decreased reabsorption in the proximal tubule may overwhelm the limited capacity of the distal nephron to reabsorb water and electrolytes. As a consequence, urine osmolality decreases, and the rate of flow rises, even in the presence of supranormal levels of vasopressin. This type of polyuria is referred to as solute diuresis to distinguish it from that resulting from a deficiency of vasopressin action. Conversely, in clinical conditions, such as congestive failure, in which the proximal nephron reabsorbs increased amounts of filtrate, the capacity to excrete solute-free water is greatly reduced, even in the absence of vasopressin. The antidiuretic effect of vasopressin also may be inhibited by the dissipation of the medullary concentration gradient. The latter may result from such diverse causes as chronic water diuresis, reduced medullary blood flow, or protein deficiency. However, probably because the bulk of the fluid issuing from the Henle loop can still be reabsorbed isotonically in the distal convoluted tubule or proximal collecting duct, the loss of the medullary concentration gradient alone rarely results in marked polyuria. The cellular receptors that mediate the antidiuretic effect of vasopressin are located on the serosal surface of renal tubular epithelia in the collecting ducts. They are known as V2 receptors and have a structure similar to that of other G proteincoupled receptors.42 Binding of these receptors activates adenylate cyclase, which in turn increases the hydroosmotic permeability of the mucosal surface by inserting preformed water channels composed of a protein known as aquaporin-2,43 a nonpeptide antagonist that binds selectively to V2 receptors and blocks the antidiuretic action of vasopressin in humans that has been developed44a,44b and may prove useful in treating clinical disorders of water balance resulting from osmotically inappropriate secretion of vasopressin (see Chap. 27). A number of different mutations in the genes that encode the V2 receptor protein or the aquaporin-2 proteins impair the urinary concentration and results in the clinical syndrome of congenital nephrogenic diabetes insipidus.45 EXTRARENAL ACTION Vasopressin has been implicated in the control of other physiologic functions such as blood pressure, temperature, insensible water loss, adrenocorticotropic hormone (ACTH) secretion, glycogenolysis, platelet function, CSF formation, and memory. Most of these effects are thought to be mediated by different receptors, known as V1a and V1b46,47 which are present in several parts of the body, including the brain.48,49 For the most part, however, these extrarenal effects of vasopressin have been demonstrated only at relatively high concentrations of the hormone in experimental animals, and their putative role in human physiology or pathophysiology is still uncertain. OXYTOCIN The major physiologic action of oxytocin is to facilitate nursing by stimulating the contraction of myoepithelial cells in the lactating mammary gland (see Chap. 106). Oxytocin may also aid in parturition by stimulating contraction of the uterus (see Chap. 109).49a These effects are mediated via a specific oxytocin receptor,50 which may be up-regulated during pregnancy. Whether the hormone has any significant physiologic role in men is unknown. At supraphysiologic concentrations approaching those achieved during the infusion of Pitocin (oxytocin) to induce labor, oxytocin exerts a significant antidiuretic effect51 in humans, probably by stimulating vasopressin V2 receptors.52

THIRST MECHANISM
The thirst mechanism provides an indispensable adjunct to the antidiuretic control of water balance in humans (see Chap. 26 and Chap. 27). Thirst is stimulated by many of the same variables that cause vasopressin release,53 the most potent of which appears to be hypertonicity. In healthy adults, a rise in effective plasma osmolality to 2% to 3% above basal levels produces a strong desire to drink. The absolute level of plasma osmolality at which a desire for water is first perceived may be termed the osmotic threshold for thirst. This threshold varies appreciably, but among healthy adults, it averages ~295 mOsm/kg (see Fig. 25-4). This level is higher than the osmotic threshold for vasopressin release and closely approximates the level at which the amount of hormone secreted is sufficient to produce maximal concentration of the urine (see Fig. 25-5). The osmoreceptors that regulate thirst appear to be located in the anterolateral hypothalamus near, but not totally coincident with, those responsible for vasopressin release.54 The sensitivity and solute specificity of the thirst and vasopressin osmoreceptors also appear to be similar. Thus, the intensity of thirst and the amount of water ingested increase rapidly in direct proportion to plasma sodium or osmolality. As with vasopressin secretion, thirst is not stimulated in healthy adults when the rise in plasma osmolality is secondary to urea or glucose. However, thirst, as well as vasopressin release, is stimulated by hypergly-cemia in insulin-deficient diabetics, probably because insulin is necessary for uptake of glucose by both types of osmoreceptor. Hypovolemia and hypotension are also dipsogenic.55 The degree of hypovolemia or hypotension required to produce thirst appears to be greater than the degree at which vasopressin release is affected. The pathways by which hypovolemia and hypotension produce thirst are uncertain, but they probably are similar, if not identical, to those that mediate the baroregulation of vasopressin. Hemodynamic stimuli also reset the osmotic threshold for thirst, just as they do for vasopressin.56

WATER HOMEOSTASIS
VOLUME, COMPOSITION, DISTRIBUTION, AND BALANCE Water is by far the largest constituent of the human body. In lean, healthy adults, it constitutes 55% to 65% of body weight, and in infants and young children, it represents an even larger proportion.57 Approximately two-thirds of body water is intracellular. The rest is extracellular and is divided further into the intravascular (plasma) and extravascular (interstitial) compartments. Plasma is much the smaller of the two, constituting only approximately one-fourth of the total extracellular volume. The solute composition of intracellular and extracellular fluid differs markedly because most cell membranes possess an array of transport systems that actively accumulate or expel specific solutes.58 However, the total solute concentration of the extracellular and the intracellular fluid is always the same because most cell membranes are freely permeable to water. Thus, distribution of water between the intracellular and extracellular compartments is determined by osmotic pressure resulting from differences in the solute content of the two compartments. If the total solute concentration of one compartment changes, the difference in osmotic pressure induces a rapid efflux or influx of water from the neighboring compartments until osmotic equilibrium is restored.59,60 Similarly, the distribution of extracellular water between the intravascular and interstitial compartments is determined largely by the balance of hemodynamic and oncotic pressure.

The total amount of water in the body is determined by the balance between intake and loss to the environment. The latter occurs via two routes; urination and evaporation, mostly from skin and lungs. The amounts lost via either route can vary markedly depending on antidiuretic function, solute load, physical activity, and temperature. However, even when conservation is maximum, the total amount of water lost by a healthy 70-kg adult cannot be reduced below ~1000 mL a day. Part of this obligatory loss can be replaced by the metabolism of fat (~300 mL per day in the average adult). The rest must come from the ingestion of water either as food or beverage. Thus, the mechanisms for ensuring an adequate intake of water are the most important for maintaining normal hydration. OSMOREGULATION Despite large daily variations in sodium intake and water output, plasma osmolality and its principal determinantplasma sodium concentrationnormally are maintained within a remarkably narrow range (Fig. 25-11). The only perceptible changes occur after meals when plasma osmolality rises transiently as a result of the absorption of sodium, glucose, and other solutes. This stability is achieved largely by keeping total body water in balance with sodium through the osmoregulation of thirst and vasopressin secretion. Thus, a reduction in osmotic pressure of only 1% or 2% inhibits vasopressin secretion, thereby decreasing the urine concentration and increasing the urine flow. Concomitantly, fluid intake is reduced,53 apparently because a sense of satiety develops. Conversely, a rise in the osmotic pressure of body fluids of 1% to 2% stimulates vasopressin secretion and thirst, thereby decreasing urinary water excretion and increasing oral water intake.

FIGURE 25-11. Circadian pattern of urine output, plasma vasopressin, and its recognized influences in healthy young adults. Each value represents the mean standard error of the mean of nine subjects. Note that urine volume decreases by ~50% during sleep, owing largely to a decrease in the rate of solute excretion and a resultant rise in urine osmolality. Plasma vasopressin changes relatively little throughout the day except for transient small increases that occur after meals, coincident with small increases in plasma osmolality and sodium. The 15% to 20% fall in mean arterial pressure that occurs during sleep has no appreciable effect on vasopressin secretion, possibly because this stimulus is counteracted by an increase in plasma volume that results from a net influx of fluid from the interstitial space. (From Robertson GL. The regulation of vasopressin secretion. In: Seldin DW, Giebisch G, ed. The kidney: physiology and pathophysiology. Philadelphia: LippincottRaven, 2000; in press.)

The ability of the thirst or vasopressin mechanisms to effect very large changes in the rate of water intake or excretion provides almost insurmountable barriers to excessive overhydration or underhydration even in certain conditions in which one or the other control mechanism malfunctions. Thus, if plasma osmolality falls enough to maximally inhibit vasopressin secretion (and renal function and solute excretion are normal), the rate of water excretion rises to levels that can equal or surpass all but the most pathologically excessive rates of water intake (as in many patients with severe primary polydipsia). In this situation, the osmotic threshold for vasopressin secretion effectively determines the lower limit to which the osmotic pressure of body fluids can be depressed. If the diuretic control system is inoperable (as in patients treated with antidiuretic hormone), the thirst mechanism can compensate by down-regulating water intake to keep it in balance with even minimal rates of urine output. On the other hand, if plasma osmolality rises sufficiently to stimulate thirst (and access to fresh water is unrestricted), the rate of water intake can rise to levels sufficient to replace all but the most extraordinary rates of loss (as in many patients with severe pituitary or nephrogenic diabetes insipidus). In this situation, the osmotic threshold for thirst effectively determines the highest levels to which plasma osmolality is allowed to rise. However, if the thirst mechanism fails, the antidiuretic mechanism cannot compensate because it cannot generate water to offset even minimal obligatory losses caused by urination and evaporation. HEMODYNAMIC INFLUENCES In humans, blood pressure and volume vary appreciably throughout the day (see Fig. 25-11). However, because the stimulus response curve is curvilinear (see Fig. 25-7), these hemodynamic changes are usually too small to affect thirst or vasopressin secretion. Even if the hemodynamic changes reach levels sufficient to affect thirst or vasopressin secretion, the fundamental nature of the osmoregulatory system is not compromised because they merely raise or lower the setpoint a few percent, depending on whether blood pressure, effective blood volume, or both are rising or falling. This constant resetting has the effect of slightly widening the range over which plasma osmolality is allowed to fluctuate, but it does not jeopardize the essential osmoregulatory system. Consequently, plasma vasopressin as well as plasma osmolality remain relatively constant throughout the day except for the small, transient increases that occur after meals (see Fig. 25-11). Urine osmolality and flow, on the other hand, show considerable circadian variation owing largely to changes in the rate of solute excretion throughout the day. The contributions of vasopressin and thirst to the regulation of blood volume and pressure are trivial and occur largely as an indirect consequence of efforts to preserve osmolality. Indeed, in situations in which total body sodium is increased abnormally, thirst and vasopressin act in such a way as to aggravate, instead of ameliorate, the underlying hypervolemia. The responsibility for coping with disturbances in volume rests primarily with those elements of the renal and endocrine systems that regulate sodium excretion. This distinction is useful to bear in mind when considering the pathogenesis of clinical disorders of salt and water balance. CHAPTER REFERENCES
1. Haymaker W. Hypothalamo-pituitary neural pathways and the circulatory system of the pituitary. In: Haymaker W, et al., eds. The hypothalamus. Springfield: Charles C Thomas, 1969:219. 2. Zimmerman EA. The organization of oxytocin and vasopressin pathways. In: Martin JB, Reichlin S, Bick KL, eds. Neurosecretion and brain peptides. New York: Raven Press, 1981:63. 3. Luerssen TG, Robertson GL. Cerebrospinal fluid vasopressin and vasotocin in health and disease. In: Wood JH, ed. Neurobiology and cerebrospinal fluid, vol 1. New York: Plenum Publishing, 1980:613. 4. du Vigneaud V, Shorr E, Bing RJ, et al. Hormones of the posterior pituitary gland: oxytocin and vasopressin. In: Harvey Lectures, 19541955. New York: Academic Press, 1956. 5. Breslow E. The neurophysins. Adv Enzymol 1974; 40:271. 6. Richter D. Molecular events in expression of vasopressin and oxytocin and their cognate receptors. Am J Physiol 1988; 255:F207. 7. Sausville E, Carney D, Battey J. The human vasopressin gene is linked to the oxytocin gene and is selectively expressed in a cultured lung cancer cell line. J Biol Chem 1985; 260:10236. 8. Riddell DC, Mallonee R, Phillips JA, et al. Chromosomal assignment of human sequences encoding arginine vasopressin-neurophysin ii and growth hormone. Somat Cell Mol Genet 1985; 11:189. 9. Hansen L, Rittig S, Robertson GL. The genetic basis of familial neurohypo-physeal diabetes insipidus. Trends Endocrinol Metab 1997; 8:363. 10. Douglass WW. How do neurons secrete peptides? Exocytosis and its consequences including synaptic vesicle formation in the hypothalamo-neurohypophyseal system. Prog Brain Res 1973; 39:21. 11. Verney EB. Antidiuretic hormone and the factors which determine its release. Proc R Soc Lond [Biol] 1947; 135:25. 12. Robertson GL. The regulation of vasopressin function in health and disease. Recent Prog Horm Res 1977; 33:333. 13. Robertson GL, Athar S, Shelton RL. Osmotic control of vasopressin function. In: Andreoli TE, Grantham JJ, Rector FC, eds. Disturbances in body fluid osmolality. Bethesda, MD: American Physiological Society, 1977:125. 14. Schrier RW, Berl T, Anderson RJ. Osmotic and nonosmotic control of vasopressin release. Am J Physiol 1979; 236:F321. 15. Robertson GL. Thirst and vasopressin function in normal and disordered states of water balance. J Lab Clin Med 1983; 101:351. 16. Jewell PA, Verney EB. An experimental attempt to determine the site of the neurohypophyseal osmoreceptors in the dog. Philos Trans R Soc Lond [Biol] 1957; 240:197. 17. Andersson B. Thirst and brain control of water balance. Am Sci 1971; 59:408. 18. Oldfield BJ, Miselis RR, McKinley MJ. Median preoptic nucleus projections to vasopressin-containing neurones of the supraoptic nucleus in sheep: a light and electron microscopic study. Brain Res 1991; 542:193. 19. Zerbe RL, Miller JZ, Robertson GL. The reproducibility and heritability of individual differences in osmoregulatory function in normal human subjects. J Lab Clin Med 1991; 117:51. 20. Robertson GL. Physiology of ADH secretion. Kidney Int 1987; 32:3. 21. Lindheimer MD, Davison JM. Osmoregulation, the secretion of arginine vasopressin and its metabolism during pregnancy. Eur J Endocrinol 1995; 132:133. 22. Schrier RW, Bert T, Anderson RJ, McDonald KM. Nonosmolar control of renal water excretion. In: Andreoli TE, Giantham JJ, Rector FC, eds. Disturbances in body fluid osmolality. Bethesda, MD: American Physiological Society, 1977:149. 23. Iwasaki Y, Gaskill MB, Boss CA, Robertson GL. The effect of the nonselective opioid antagonist diprenorphine on vasopressin secretion in the rat. Endocrinology 1994; 134:48. 24. Iwasaki Y, Gaskill MB, Robertson GL. The effect of selective opioid antagonists on vasopressin secretion in the rat. Endocrinology 1994; 134:55. 25. Iwasaki Y, Gaskill MB, Fu R, et al. Opioid antagonist diprenorphine micro-injected into parabrachial nucleus selectively inhibits vasopressin response to hypovolemic stimuli in the rat. J Clin Invest 1993; 92:2230. 26. Zerbe RL, Henry DP, Robertson GL. Vasopressin response to orthostatic hypotension: etiological and clinical implications. Am J Med 1983; 74:265. 27. Rowe JW, Shelton RL, Helderman JH, et al. Influence of the emetic reflex on vasopressin release in man. Kidney Int 1979; 16:729. 28. Baylis PH, Zerbe RL, Robertson GL. Arginine vasopressin response to insulin-induced hypoglycemia in man. J Clin Endocrinol Metab 1981; 53:935.

29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43.

Thompson DA, Cambell RG, Lilavivat U, et al. Increased thirst and plasma arginine vasopressin levels during 2-deoxy-D-glucose-induced glucoprivation in humans. J Clin Invest 1981; 67:1083. Mouw D, Bonjour JP, Malvin RL, Vander A. Central action of angiotensin in stimulating ADH release. Am J Physiol 1971; 220:239. Shimizu K, Share L, Claybaugh JR. Potentiation of angiotensin II of the vasopressin response to an increasing plasma osmolality. Endocrinology 1973; 93:42. Rydin H, Verney EB. The inhibition of water-diuresis by emotional stress and by muscular exercise. Q J Exp Physiol 1938; 27:343. Edelson JT, Robertson GL. The effect of the cold pressor test on vasopressin secretion in man. Psychoneuroendocrinology 1985; 11:307. Rose CE Jr, Anderson RJ, Carey RM. Antidiuresis and vasopressin release with hypoxemia and hypercapnia in conscious dogs. Am J Physiol 1984; 247:R127. Heyes MP, Farber MO, Manfredi F, et al. Effect of hypoxia on renal and endocrine function in normal humans. Am J Physiol 1982; 243:R265. Thompson CJ, Burd JM, Baylis PH. Acute suppression of plasma vasopressin and thirst after drinking in hypernatremic humans. Am J Physiol 1987; 252:R1138. Robertson GL, Berl T. Water metabolism. In: Brenner BM, Rector FC, eds. The kidney, 3rd ed. Philadelphia: WB Saunders, 1985:385. Amico JA, Finley BE. Breast stimulation in cycling women, pregnant women and a woman with induced lactation: pattern of release of oxytocin, prolactin and luteinizing hormone. Clin Endocrinol 1986; 25:97. Lauson HD. Metabolism of neurohypophysial hormones. In: Handbook of physiology, vol 6, section 7. Endocrinology. Bethesda, MD: American Physiologic Society, 1971:287. Amico JA, Ulbrecht JS, Robinson AG. Clearance studies of oxytocin in humans using radioimmunoassay measurements of the hormone in plasma and urine. J Clin Endocrinol Metab 1987; 64:340. Berliner BW, Levinsky NG, Davidson DG, Eden M. Dilution and concentration of the urine and the action of antidiuretic hormone. Am J Med 1958; 24:730. Bimbaumer M, Seibold A, Gilbert S, et al. Molecular cloning of the receptor for human antidiuretic hormone. Nature 1992: 357:333. Knepper MA. Molecular physiology of urinary concentrating mechanism: regulation of aquaporin water channels by vasopressin. Am J Physiol 1997; 272:F3.

44a. Ohnishi A, Orita Y, Okahara R, et al. Potent aquaretic agent: a novel non-peptide selective vasopressin 2 antagonist (OPC-3 1260) in men. J Clin Invest 1993; 92:2653. 44b. Tahara A, Saito M, Sugimoto T, et al. Pharmacological characterization of YM087, a potent, nonpeptide human vasopressin V 1a and V2 receptor antagonist. Naunyn-Schmiedeberg's Arch Pharmacol 1998; 357:63. 45. 46. 47. 48. 49. Bichet DG, Fujiwara MT. Diversity of nephrogenic diabetes insipidus mutations and importance of early recognition and treatment. Clin Exp Nephrol 1998; 2:253. Morel A, O'Carrol AM, Brownstein MJ, Lolait S. Molecular cloning and expression of a rat V 1a arginine vasopressin receptor. Nature 1992; 356:523. Sugimoto T, Saito M, Mockzuki S, et al. Molecular cloning and functional expression of a cDNA encoding the human V1b receptor. J Biol Chem 1994; 269:27088. Vaccari C, Lolait S, Ostrowski NL. Comparative distribution of vasopressin V 1b and oxytocin receptor messenger ribonucleic acids in brain. Endocrinology 1998; 139:5015. 49. Hurbin A, Boissin-Agasse L, Orcel H, et al. The V1a and V1b but not the V2 vasopressin receptor genes are expressed in the supraoptic nucleus of the rat hypothalamus and the transcripts are essentially colocalized in the vasopressingeric magnocellular neurons. Endocrinology 1998; 139:4701.

49a. Voutsos L, Cantor D. Randomized, double-masked comparison of oxytocin dosage in induction and augmentyation of labor. Obstet Gynecol 2000; 95:472. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. Kimura T, Tanizawa O, Mon K, et al. Structure and expression of a Human Oxytocin Receptor. Nature 1992; 356:526. Kelly S, Robertson GL, Amico J. Antidiuretic action of oxytocin in humans. Clin Res 1992; 40:711A. Chou CL, DiGiovanni SR, Luther A, et al. Oxytocin as an antidiuretic hormone ii: role of V2 vasopressin receptor. Am J Physiol 1995; 269:F78. Robertson GL. Disorders of thirst in man. In: Ramsay DJ, Booth DA, eds. Thirst: physiological and psychological aspects. London: Springer-Verlag, 1991:453. McKinley MJ. Osmoreceptors for thirst. In: Ramsay DJ, Booth DA, eds. Thirst: physiological and psychological aspects. London: Springer-Verlag, 1991:77. Thrasher TN. Volume receptors and the stimulation of water intake. In: Ramsay DJ, Booth DA, eds. Thirst: physiological and psychological aspects. London: Springer-Verlag, 1991:91. Kozlowski S, Szczepanska-Sadowska E. Antagonistic effects of vasopressin and hypervolemia on osmotic reactivity of the thirst mechanism in dogs. Pflugers Arch 1975; 353:59. Altman PL, Dittmer DS, eds. Blood and other body fluids. Washington, DC: American Society for Experimental Biology, 1961. Wolf AV, McDowell ME. Apparent and osmotic volumes of distribution of sodium, chloride, sulfate and urea. Am J Physiol 1954; 176:207. Darrow DC, Yanett H. Changes in distribution of body water accompanying increase and decrease in extracellular electrolytes. J Clin Invest 1935; 14:266. Leaf A, Chatillon JY, Wrong O, Tuttle EP Jr. The mechanism of the osmotic adjustment of body cells as determined in vivo by the volume of distribution of a large water load. J Clin Invest 1954; 33:1261.

CHAPTER 26 DIABETES INSIPIDUS AND HYPEROSMOLAR SYNDROMES Principles and Practice of Endocrinology and Metabolism

CHAPTER 26 DIABETES INSIPIDUS AND HYPEROSMOLAR SYNDROMES

PETER H. BAYLIS AND CHRISTOPHER J. THOMPSON Definitions Diabetes Insipidus Etiology Clinical Features Differential Diagnosis Thirst in Hypothalamic Diabetes Insipidus Diabetes Insipidus and Pregnancy Treatment Hyperosmolar Syndromes Etiology Clinical Features Osmoregulatory Defects in Chronic Hypernatremia Treatment Chapter References

Blood osmolality in healthy persons is maintained within narrow limits by a series of mechanisms that are described in detail in Chapter 25. Adjustments in water balance determine the constancy of blood osmolality, which is mediated by delicate alterations in thirst appreciation (with consequent promotion of drinking) plus the enormous capacity of the kidney to alter urine flow rates and urine osmolality in response to relatively small changes in the plasma vasopressin concentration (see Chap. 206). Thus, healthy humans are able to conserve their osmotic internal milieu despite extremes in climatic conditions, sustained severe exertion, or, to a certain degree, an inadequate supply of water. Aberrations of the intricate mechanisms involved in maintaining osmoregulation can lead to the inappropriate accumulation of water, which is recognized as one of the hypoosmolar states, or to the loss of renal water, which usually is clinically apparent as polyuria but also may be manifested as one of the hyperosmolar syndromes. This chapter is concerned with clinical situations associated with polyuria and, on occasion, abnormalities of thirst appreciation.

DEFINITIONS
Diabetes insipidus refers to the passage of copious volumes of dilute urine and is synonymous with polyuria. In adults, the urine volume exceeds 2.5 L per 24 hours (>40 mL/kg per 24 hours), while in children the output is greater (>100 mL/kg per 24 hours). Three pathophysiologic conditions result in diabetes insipidus. An absolute or partial deficiency of vasopressin secretion from the neurohypophysis in response to normal osmotic stimulation is termed hypothalamic diabetes insipidus. This disorder is also known as cranial, central, or neurogenic diabetes insipidus. Patients with hypothalamic diabetes insipidus generally have normal thirst sensation. Their basic abnormality is insufficient circulating antidiuretic activity, which is the principal, but not the sole, cause of their polyuria. Diabetes insipidus secondary to decreased renal sensitivity to the antidiuretic effect of vasopressin circulating in normal or high concentrations is usually called nephrogenic diabetes insipidus. Again, these patients rely on normal thirst sensation to regulate water balance. The third mechanism leading to diabetes insipidus is the ingestion of excessive volumes of fluid, which results in suppression of vasopressin release and consequent polyuria. This condition is referred to as dipsogenic diabetes insipidus, sometimes termed primary polydipsia. A decrease in maximal urine-concentrating ability occurs after prolonged periods of polyuria, regardless of the primary cause. The passage of large amounts of dilute urine through the distal nephron removes solute from the renal medullary interstitium, a process known as the washout phenomenon.1,2 The osmotic gradient across the collecting tubular cell, which is essential for the antidiuretic action of vasopressin, is decreased. Thus, any of the three pathophysiologic mechanisms responsible for diabetes insipidus may lead to an additional defect that complicates the interpretation of diagnostic tests based on indirect assessment of the antidiuretic action of vasopressin. In contrast to hypothalamic diabetes insipidus, which is usually the result of a loss of neurosecretory neurons, the chronic hyperosmolar syndromes are frequently the consequence of a defective thirst mechanism. Thirst osmoreceptors may fail to respond to hypertonicity, which results in hypodipsia. Because the putative thirst osmoreceptors are believed to be in proximity to the osmoreceptors that regulate vasopressin secretion, a defect in osmotically mediated vasopressin release is often associated with hypodipsia. Polyuria is rarely a feature of hyperosmolar syndromes, because many patients secrete small amounts of vasopressin that are sufficient to concentrate urine to some extent, and nonosmotic factors regulating vasopressin secretion often remain intact. In view of this characteristic difference between diabetes insipidus and hyperosmolar syndromes, these conditions are discussed separately.

DIABETES INSIPIDUS
ETIOLOGY In theory, any of a series of defects in the vasopressin neurosecretory process can be implicated as the cause of hypothalamic diabetes insipidus.3 Abnormalities may arise in the osmoreceptor that controls vasopressin secretion, even when the thirst osmoreceptor is spared. Alternatively, abnormalities may involve the synthesis and packaging of vasopressin (including genetic defects), damage to the vasopressinergic neurons, or disorders of neurohypophyseal hormone release. Enhanced inactivation of vasopressin by circulating degrading enzymes or antibodies is another potential cause of decreased antidiuretic activity. In practice, however, most cases of permanent hypothalamic diabetes insipidus are caused by damage to the hypothalamo-neurohypophyseal area. The most common causes of this condition are listed in Table 26-1.4

TABLE 26-1. Causes of Diabetes Insipidus

FAMILIAL ABNORMALITIES Studies have revealed exciting data regarding a variety of genetic abnormalities found on chromosome 20 in several kindreds with autosomal dominant hypothalamic diabetes insipidus. The first report on different families showed single nucleotide substitutions in the region coding for neurophysin (glycine to serine at position 57).5 This mutation is presumed to interfere with the normal vasopressin-neurophysin tetramer complex formation that occurs in the packaging and transport of vasopressin to the neurohypophysis. Two groups investigating another extended family discovered a nucleotide substitution in the signal peptide (alanine to threonine at position 1).6,7 The signal peptide directs the prohormone to the endoplasmic reticulum, where it is cleaved. Both groups speculate that the mutation alters the cleavage mechanism, resulting in abnormal processing of the prohormone. After the description of the first genetic abnormalities causing familial hypothalamic diabetes

insipidus, more than 22 different kindreds with unique genetic mistakes have been documented8 (Fig. 26-1). The genetic basis of the DIDMOAD (diabetes insipidus, diabetes mellitus, optic atrophy, deafness), or Wolfram, syndrome is less well understood, although some evidence exists that it is a disorder of mitochondrial DNA.9,10 TRAUMA Closed head trauma or frank damage to the pituitary stalk or hypothalamus as a result of surgical intervention is often the cause of a form of diabetes insipidus that usually presents within 24 hours of injury. In ~50% of cases of post-traumatic diabetes insipidus, the condition resolves spontaneously within a few days. Permanent diabetes insipidus develops in another 30% to 40% of these patients, and the remainder exhibit a triphasic response to injury. In the last group, the onset of polyuria is abrupt and the condition lasts a few days. It is followed by a period of antidiuresis that may last 2 to 14 days before permanent diabetes insipidus develops. This triple response to injury is believed to be attributable to release of the vasopressin that is stored in granules.11 Recognition of this entity by clinicians should help to prevent inappropriate treatment that would result in hyponatremia during the second of the three phases. TUMORS Tumors of the anterior pituitary rarely cause diabetes insipidus. In a series of >100 cases of hypothalamic diabetes insipidus, 13% were attributable to tumors, which included glioma, germinoma, and craniopharyngioma.12 In children, a central tumor is a frequent cause of hypothalamic diabetes insipidus, accounting for ~25% of cases; in this population, the most common intracranial tumor is germinoma.13 Metastatic deposits in the hypothalamus causing diabetes insipidus usually arise from carcinoma of the breast or bronchus.

FIGURE 26-1. Vasopressin gene, vasopressin precursor molecule, and vasopressin with its specific neurophysin. Three exons encode for the precursor molecule, which comprises a signal protein, vasopressin hormone, neurophysin, and a glycoprotein moiety coupled by amino acids. Mutations in the vasopressin gene have been located in all parts of the precursor molecule except vasopressin itself. (Vp, vasopressin.)

GRANULOMATOUS DISEASE Granulomatous disease accounts for only a few cases of diabetes insipidus in adults (i.e., sarcoidosis, tuberculosis). However, in children with granulomatous disease, histiocytosis X may cause as many as 40% of pediatric cases. IDIOPATHIC Idiopathic hypothalamic diabetes accounts for ~25% of all cases.12 One-third of patients with apparent idiopathic disease have circulating antibodies to the vasopressin-producing cells in the hypothalamus, a finding which suggests an autoimmune origin for the disorder.14 Some patients have an acute lymphocytic infiltration of the infundibulum and neurohypophysis that can be demonstrated on open biopsy and subsequently resolves.15 NEPHROGENIC DIABETES INSIPIDUS Mild forms of nephrogenic diabetes insipidus are relatively common (see Table 26-1). Mechanisms responsible for renal resistance to the antidiuretic effect of vasopressin may occur at one or more of the many different sites in the chain of biochemical responses to vasopressin.11,15a Chronic renal disease secondary to numerous conditions, many drugs (e.g., lithium15b), prolonged electrolyte disturbances from hypokalemia, and hypercalcemia account for most cases of nephrogenic diabetes insipidus. The inherited forms of nephrogenic diabetes insipidus are rare, and cause severe polyuria, dehydration, and failure to thrive in the young. With the identification of the V2 (antidiuretic) receptor gene on the X chromosome, a variety of substitutions, mutations, or premature stops have been isolated in kindreds with this disorder that cause defects in the trans-membrane V2 receptor.16 Studies involving three other families with congenital nephrogenic diabetes insipidus have shown autosomal inheritance of the disorder, in contrast to the more common X-linked form due to genetic mutations of the V2 receptor gene localized to the Xq28 region of the long arm of the X chromosome. The autosomal form is due to novel genetic mutations of the gene encoded for the vasopressin-sensitive water channel protein, aquaporin-2, which is located in the collecting tubules.17 DIPSOGENIC DIABETES INSIPIDUS Dipsogenic diabetes insipidus, also called primary polydipsia or habitual water drinking, is often psychogenic in origin. The course of polyuria in psychotic patients is variable, with fluctuations in polydipsia and urine volumes occurring over the years. Occasional patients with hypothalamic diabetes insipidus who are treated with antidiuretic preparations continue to have polydipsia and, consequently, run the risk of developing hyponatremia. Whether these patients continue to drink because of habit or because of a hypothalamic lesion affecting the thirst osmoreceptor is unknown. A few structural abnormalities resulting in increased thirst have been reported. Drugs that cause dryness of the mouth (e.g., thioridazine hydrochloride) may increase drinking but do not result in true polydipsia, in contrast to lithium, which can stimulate thirst directly. CLINICAL FEATURES In adults, the major clinical manifestations of diabetes insipidus include the frequent passage of large volumes of dilute urine (often both day and night), excessive thirst, and increased fluid ingestion. Patients with mild degrees of diabetes insipidus may consider their symptoms to be so minimal that they fail to seek medical attention. However, the severity of diabetes insipidus varies widely, with 24-hour urine volumes ranging from 2.5 to 20 L. Even with the most extreme forms of the disorder, patients maintain their water balance as long as thirst appreciation remains intact and adequate volumes of fluid are ingested. The onset of the disease occurs at any time from the neonatal period to old age, and the sex distribution is approximately equal, although one large review of adults with hypothalamic diabetes insipidus reported a slight male preponderance (60%:40%).12 In infants, diabetes insipidus usually presents with evidence of chronic dehydration, unexplained fever, vomiting, neurologic disturbance, and failure to thrive.13 Enuresis, sleep disturbances, and difficulties at school are the most common presenting complaints in older children. Usually, no growth retardation or failure to enter puberty occurs. Affected children from families with histories of diabetes insipidus often do not complain; they regard their polydipsia and polyuria as the norm. Once hypothalamic diabetes insipidus develops, it rarely goes into remission spontaneously. Patients with hypothalamic diabetes insipidus also may have anterior pituitary dysfunction (see Chap. 17), particularly if their disorder resulted from trauma to or a tumor in the hypo-thalamo-neurohypophyseal area. Even patients with the idiopathic form of the disease frequently have endocrinologic evidence of anterior pituitary dysfunction, which suggests a more generalized hypothalamic disorder.12 Glucocorticoid deficiency secondary to impaired corticotropin secretion or to primary adrenal disease leads to impairment of the ability of the kidneys to excrete a water load and to dilute urine maximally. At least two mechanisms are responsible for this defect. One involves the distal nephron, which remains partially impermeable to water in the absence of glucocorticoid; the other involves persistent vasopressin secretion, possibly secondary to a resetting of the osmostat. A similar abnormality of water excretion has been reported with severe thyroid hormone deficiency (see Chap. 45).

Thus, impairment of anterior pituitary function can mask hypothalamic diabetes insipidus, which becomes apparent only when the hypopituitarism is adequately treated. Routine skull radiographs are rarely helpful in patients with hypothalamic diabetes insipidus but nuclear magnetic resonance imaging (MRI) can be extremely useful. T1-weighted magnetic resonance imaging of the neurohypophysis produces a characteristic hyperintense signal in healthy persons that disappears in most patients with hypothalamic diabetes insipidus.18 The infundibular stalk frequently is thickened in the early phase of the idiopathic form of the disorder, as demonstrated by both MRI and computed tomographic scanning.15 DIFFERENTIAL DIAGNOSIS Polyuria may be defined as the excretion of >2.5 L of urine per 24 hours on two consecutive days, provided that patients are allowed free access to and drink water ad libitum. Once polyuria has been demonstrated, the clinician's first responsibility is to establish the pathophysiologic mechanismdipsogenic diabetes insipidus, hypothalamic diabetes insipidus, or nephrogenic diabetes insipidus. Defining the underlying disease process is then important. INDIRECT TESTS Before the development of plasma assays that were capable of detecting low physiologic concentrations of vasopressin, indirect methods were used to assess the antidiuretic activity of the hormone. The classic diagnostic approach is to measure the responses of urinary osmolality and flow rate to a period of dehydration and, subsequently, to the administration of an exogenous vasopressin preparation. Various dehydration tests have been described in which changes in plasma and urine osmolalities in patients with polyuria were compared to the responses of healthy persons.19,20 and 21 In theory, differentiation between hypothalamic diabetes insipidus, nephrogenic diabetes insipidus, and dipsogenic diabetes insipidus should be readily possible, but the disorders actually can be diagnosed correctly only in certain circumstances. Frequently, tests yield equivocal results. For example, although patients with primary polydipsia might be anticipated to have significantly lower plasma osmolalities and sodium concentrations than patients with hypothalamic or nephrogenic diabetes insipidus, such a distinction is useful diagnostically in only a few cases.22 The reason for the lack of differentiation is the wide variation in the setpoint of the osmoregulatory mechanisms for thirst and vasopressin secretion. Even after fluid deprivation and the administration of exogenous vasopressin, urine osmolality frequently fails to attain normal values (Fig. 26-2) because of the secondary nephrogenic diabetes insipidus induced by prolonged polyuria, as explained earlier. A similar ambiguity arises in the interpretation of results from patients with hypothalamic diabetes insipidus. In theory, these patients, who have low circulating concentrations of vasopressin, should demonstrate a substantial increase in urine osmolality to the normal range in response to exogenous vasopressin. In practice, however, they fail to do so (see Fig. 26-2). Again, the reason for the inadequate urinary response lies in the washout of solute from the renal medullary interstitium. The greater the 24-hour urine volume, the greater is the degree of renal resistance to antidiuretic activity.2 The large overlap in urine osmolality values with these tests is illustrated explicitly in Figure 26-2. The refinement of dehydration tests by calculation of free water clearance adds little to their ability to distinguish among the causes of diabetes insipidus. Other means of evaluating the osmoregulatory system using indirect methods to assess antidiuretic activity in an attempt to identify the cause of polyuria (i.e., infusion of hypertonic saline23) also fail to establish an unequivocal diagnosis for the reasons described earlier, as well as because the saline load induces a solute diuresis. Thus, indirect tests of vasopressin function are considerably limited in their ability to establish the cause of polyuria.

FIGURE 26-2. Urine osmolality is depicted under basal conditions after a period of fluid deprivation (hydropenia) designed to attain maximum urinary concentration, and after intramuscular injection of 5 U of vasopressin (Pitressin). The test group included healthy subjects (stippled area) and patients with dipsogenic diabetes insipidus (primary polydipsia, PP), hypothalamic diabetes insipidus (HDI), or nephrogenic diabetes insipidus (NDI). The bars represent the range of results and the closed circle indicates the mean value for the group. (Adapted from Robertson, GL. Diagnosis of diabetes insipidus. In: Czernichow P, Robinson AG, eds. Diabetes insipidus in man. Frontiers of hormone research, vol 13. Basel: S Karger, 1985:176.)

DIRECT TESTS The introduction of sensitive and specific radioimmunoassays capable of detecting low physiologic concentrations of vasopressin in plasma not only has clarified and simplified the diagnosis of diabetes insipidus, but also has extended the understanding of its underlying pathophysiologic mechanisms. The measurement of plasma vasopressin, plasma osmolality, and urine osmolality under basal conditions affords little diagnostic discrimination. However, after osmotic stimulation by a period of fluid deprivation, an infusion of hypertonic saline, or both, the estimation of these indices provides a precise diagnosis. PLASMA VASOPRESSIN IN THE DIAGNOSIS OF DIABETES INSIPIDUS Hypothalamic Diabetes Insipidus. Hypothalamic diabetes insipidus is recognized by the subnormal plasma concentrations of vasopressin in relation to plasma osmolality that occur in affected persons (Fig. 26-3A). A clear distinction may be made between patients with hypothalamic diabetes insipidus, patients with nephrogenic diabetes insipidus or dipsogenic diabetes insipidus, and healthy persons by assessing plasma osmolality as it is increased by the infusion of hypertonic saline.24 Many patients with diabetes insipidus have detectable plasma vasopressin, which represents the partial form of the disorder. The ability to secrete vasopressin at high plasma osmolality levels partially explains the ability of some patients to generate a concentrated, if submaximal, urine after fluid deprivation. A few patients fail to exhibit detectable immunoreactive plasma vasopressin, despite marked increases in plasma osmolality. However, some of these patients still manage to concentrate their urine to some extent, suggesting that their kidneys are particularly sensitive to very low concentrations of vasopressin. Occasionally, patients with hypothalamic diabetes insipidus clearly demonstrate osmotically regulated vasopressin release (see Fig. 26-3A). In such patients, the theoretic threshold for vasopressin release, obtained from the abscissal intercept of the osmoregulatory line (the function relating plasma, vasopressin to plasma osmolality), is normal, but the slope that defines the sensitivity of osmotically regulated vasopressin release is significantly reduced. Thus, the osmoreceptor controlling vasopressin secretion is probably intact in this group of patients.

FIGURE 26-3. A, Plasma osmolality and vasopressin responses to infusion of 5% hypertonic saline solution in representative patients with hypothalamic diabetes insipidus (HDI, ), nephrogenic diabetes insipidus (NDI, nn), and dipsogenic diabetes insipidus (DDI, ). B, Plasma vasopressin and urine osmolality responses to a period of dehydration in patients with hypothalamic diabetes insipidus (), nephrogenic diabetes insipidus (n), and dipsogenic diabetes insipidus ( ). The responses of healthy subjects are indicated by the stippled areas. The limit of detection of the vasopressin assay (LD) was 0.3 pmol/L (1 pmol/L 1.1 pg/mL).

Some patients who have been treated with parenteral neurohypophyseal extract (Pitressin) have developed antibodies to vasopressin. These patients exhibit vasopressin in their plasma as a direct result of assay interference, but this can be recognized by laboratory testing simply through detection of plasma-binding activity to synthetic vasopressin. Therefore, screening for vasopressin antibodies in all patients treated with Pitressin is wise to prevent spurious plasma results. After hypertonic saline infusion, patients with primary polydipsia or nephrogenic diabetes insipidus have plasma vasopressin and plasma osmolality values that fall within the normal reference range (see Fig. 26-3A). A supranormal plasma vasopressin response to rising plasma osmolality has been demonstrated in a few patients. Whether this response is attributable to a state of chronic underhydration is not known. Thus, hypothalamic diabetes insipidus is clearly distinguishable from other forms of diabetes insipidus by relating plasma vasopressin to plasma osmolality after osmotic stimulation. Nephrogenic Diabetes Insipidus Versus Primary Polydipsia. Analysis of the relationship between plasma vasopressin and urine osmolality after a period of fluid deprivation offers a potential means of differentiating nephrogenic diabetes insipidus from dipsogenic diabetes insipidus (see Fig. 26-3B). Patients with nephrogenic diabetes insipidus have plasma vasopressin concentrations that are inappropriately high in relation to urine osmolality. However, prolonged polyuria from any cause can induce renal resistance to vasopressin, with blunting of the maximal urinary concentration in response to vasopressin. This difficulty can be partially overcome by examining the basal values of plasma vasopressin and urine osmolality. Plasma vasopressin tends to be detectable or even elevated in nephrogenic diabetes insipidus, whereas immunoreactive vasopressin is generally undetectable in primary polydipsia2 (see Fig. 26-3B). The administration of exogenous vasopressin after a period of fluid deprivation does not help to discriminate further between the causes of polyuria. Close examination of the data regarding plasma vasopressin and urine osmolality in patients with partial hypothalamic diabetes insipidus reveals an inappropriately high urine osmolality in relation to the low plasma vasopressin concentrations (see Fig. 26-3B). This observation confirms the earlier impression that the renal tubule may become extraordinarily sensitive to vasopressin. Because sustained polyuria from any cause induces a state of secondary partial nephrogenic diabetes insipidus, the administration of exogenous vasopressin to patients with partial hypothalamic diabetes insipidus fails to induce maximal urinary osmolality. A systematic study to compare the diagnostic efficacy of indirect tests with direct measurement of osmotically stimulated vasopressin release has clearly demonstrated that direct plasma vasopressin measurement methods are superior.25 If the clinician does not have ready access to suitable assays for plasma vasopressin measurement, a satisfactory diagnosis may be established using a closely monitored, prolonged therapeutic trial with desmopressin, administered intramuscularly in daily dosages of 1 g for as long as 7 days, preferably while patients are hospitalized. Patients with hypothalamic diabetes insipidus who undergo this regimen show an improvement in the degree of polyuria and a reduction in polydipsia. At the end of the trial, a standard water deprivation test, followed by the administration of exogenous vasopressin using indirect assessment methods, may demonstrate maximal urinary concentrations that are within the normal reference range. This is because, during the period of the trial, the renal medullary interstitial solute concentration was restored. Patients with primary polydipsia experience progressive hyponatremia and gain weight because they continue to drink fluid despite persistent antidiuresis. Some of these patients run the risk of neurologic disturbances, particularly seizures, secondary to the development of sudden, profound hyponatremia. Finally, patients with nephrogenic diabetes exhibit little, if any, improvement in thirst or polyuria. Even the most carefully conducted therapeutic trial, however, can yield misleading results. For example, a few patients who have had severe hypothalamic diabetes insipidus for many years develop water intoxication when first treated with vasopressin because they initially fail to reduce their water intake appropriately, and therefore they appear to have primary polydipsia. Thus, with the currently available diagnostic investigations, the measurement of plasma vasopressin levels with plasma and urine osmolalities during osmotic stimulation provides the most reliable method for determining the cause of polyuria. THIRST IN HYPOTHALAMIC DIABETES INSIPIDUS The thirst mechanism in patients with hypothalamic and nephrogenic diabetes insipidus generally operates normally.26 Such patients rely on an intact thirst appreciation to maintain water balance. In one study that documented the osmolar threshold for the onset of thirst, no significant difference was found between the mean value for a group of 11 patients with hypothalamic diabetes insipidus and that for 11 healthy persons; however, the range of values was wider in the affected patients. Only a few patients who are treated with vasopressin and who have inappropriate persistent thirst develop episodes of hyponatremia. When thirst appreciation is blunted, hypernatremia develops. DIABETES INSIPIDUS AND PREGNANCY Normal human pregnancy is associated with subtle changes in osmoregulation. Plasma osmolality falls by 8 to 10 mOsm/kg due to a lowering of the osmolar thresholds for both thirst and vasopressin release,27 and a small reduction occurs in maximal urinary concentrating ability. Established hypothalamic diabetes insipidus appears to have little effect on fertility, gestation, delivery, and lactation in humans.28 A few studies have demonstrated that the secretion of oxytocin is normal in patients with diabetes insipidus. By contrast, many pregnant patients with hypothalamic diabetes insipidus notice a worsening of their polyuria and polydipsia.29 The mechanisms responsible for this may include depression of the osmolar thirst threshold to levels at which vasopressin secretion is suppressed further; circulating vasopressin that is degraded by the placental enzyme cystine aminopeptidase; and renal resistance to vasopressin, which is increased. Some patients show no change in their polyuria, whereas a few improve. Transient central and nephrogenic diabetes insipidus associated with pregnancy has been documented in some patients30,31 and has been found to recur in subsequent pregnancies. Whether this is attributable to an exaggeration of normal physiologic adaptation to pregnancy or represents a distinct disease entity remains unresolved. TREATMENT Hypothalamic diabetes insipidus can be treated by the ingestion of adequate volumes of water, with patients relying on the quenching of thirst as the sole indicator of sufficient intake. Some patients who have had severe polyuria since childhood may prefer to manage their symptoms in this manner, thereby avoiding the need for medication. They organize their lives around the inconveniences of frequent micturition and copious drinking. However, good reasons exist why all patients with moderate to severe polyuria should be treated more actively. Prolonged, severe polyuria can result in distention and atonia of the bladder and hydroureter, and, eventually, in hydro-nephrosis with consequent renal damage. Susceptibility to potassium deficiency is another potential risk. Furthermore, untreated patients, if deprived of fluid for any reason, are at risk for the development of life-threatening hypernatremia and dehydration. In young children, withholding therapy may lead to failure to thrive. For these reasons, as well as for the relief of symptoms, antidiuretic treatment is advised for patients with 24-hour urine volumes exceeding 4 L. Hormone replacement therapy using arginine vasopressin, a natural endogenous peptide, is inappropriate for most patients with hypothalamic diabetes, regardless of whether it is administered parenterally or intranasally. The peptide has a short half-life and may be associated with significant pressor side effects that render the use of aqueous vasopressin preparations impractical. However, during the last three decades, considerable advances have been made in the development of synthetic vasopressin analogs with various agonist and antagonist activities to the pressor (V1) and antidiuretic (V2) receptors. One class of analogs with minimal pressor activity but increased antidiuretic potency and some resistance to degradation in vivo has been developed to treat hypothalamic diabetes insipidus. The current drug of choice is desmopressin.32,33 For adults, it can be administered orally, 50 to 400 g one to three times daily; intranasally, 5 to 40 g once or twice daily; or parenterally, 0.5 to 2.0 g daily. A wide variation is seen in individual desmopressin requirements for the control of polyuria. For children, the dosage is halved. Desmopressin is not associated with pressor agonist side effects but carries the potential hazard of dilutional hyponatremia if patients continue to drink inappropriately despite persistent antidiuresis. If desmopressin proves to be too potent, it can be diluted. Alternatively, a shorter-acting preparationlysine vasopressincan be administered intranasally. However, because it possesses pressor activity, it may induce vasoconstriction, angina, or renal and intestinal colic when taken in excess. If desmopressin is unavailable, it still may be possible to obtain vasopressin tannate in oil, a crude extract of bovine neurohypophysis containing arginine vasopressin suspended in peanut oil. Before intramuscular injection, this preparation should be warmed and shaken vigorously until the extract is evenly distributed in the oil. A single dose of 5 to 10 IU provides as much as 72 hours of antidiuresis. However, this agent is associated with pressor side effects similar to those of lysine vasopressin, an erratic absorption rate, and the formation of sterile abscesses. Pitressin also has been administered as a nasal insufflation. Now that desmopressin is established as the drug of choice, little need exists to prescribe the partially effective oral agentschlorpropamide, carbamazepine, clofibrate, or thiazide diureticsbecause all are associated with significant and sometimes dangerous side effects. Rarely, direct treatment of the underlying cause of hypothalamic diabetes insipidus relieves the symptoms. Documented examples include corticosteroid therapy for

hypothalamic sarcoidosis, cyclophosphamide therapy for Wegener granulomatosis, and radiotherapy for metastatic disease of the hypothalamus. Effective treatment of nephrogenic diabetes insipidus still poses problems, except for the forms that are drug induced or related to metabolic disorders (see Table 26-1). The latter are frequently reversible after withdrawal of the drug or correction of the metabolic disturbance. Profound polyuria secondary to the familial forms of this disease is particularly difficult to treat. Restriction of sodium intake, combined with the administration of a thiazide diuretic, reduces urine output by almost 40% in infants. A similar reduction in urine flow may be achieved with the prostaglandin synthetase inhibitor indomethacin when it is administered in dosages of 1.5 to 3.0 mg/kg. The most promising results are achieved with the administration of a combined regimen of thiazide, indomethacin, and desmopressin, which reduces diuresis by as much as 80%.

HYPEROSMOLAR SYNDROMES
Hyperosmolar or hypernatremic syndromes may be defined as plasma osmolality levels and sodium concentrations of >300 mOsm/kg and >145 mEq/L (145 mmol/L), respectively. Although rare, they constitute a major management challenge. ETIOLOGY Transient hyperosmolality may occur after the ingestion of large amounts of salt,34 but most hypernatremic states occur after inadequate water intake. This can occur in any healthy individual in whom the combination of excess fluid lossfrom skin, gastrointestinal tract, lungs, or kidneysand inadequate access to water is found. This occurs most commonly in acute illness in which water intake is compromised by vomiting or impaired consciousness and most vividly in patients with diabetes insipidus, before treatment, or when access to water is denied. In other cases, however, hypernatremia reflects a primary disorder of thirst deficiency (hypodipsia). A number of conditions are associated with hypodipsia (Table 26-2). One of the more common causes of hypodipsic hypernatremia that the authors have seen is ligation of the anterior communicating artery, after subarachnoid hemorrhage from a berry aneurysm. Other centers have reported that neoplasms account for 50% of such cases.35 Craniopharyngiomas are particularly associated with hypodipsic diabetes insipidus, sometimes in conjunction with other hypothalamus-related disorders, such as polyphagia, weight gain, and abnormal thermoregulation. Survivors of diabetic hyperosmolar coma have been shown to have impaired osmoregulated thirst,35 which suggests that hypodipsia contributes to the development of the hypernatremia, which is characteristic of the condition. In almost every case of hypodipsia, associated abnormalities of vasopressin secretion are seen, a finding that reflects the close anatomic proximity of the osmoreceptors for vasopressin secretion and thirst.

TABLE 26-2. Specific Causes of Hypodipsic Hypernatremia

CLINICAL FEATURES In young children and the elderly, hypernatremia may be associated with significant degrees of dehydration.36 Infants are at particular risk, and the mortality is high. In this clinical situation, signs are seen of extracellular fluid loss, decreased skin turgor and elasticity, dry and shrunken tongue, tachycardia, and orthostatic hypotension. Affected infants have depressed fontanelles and tachypnea, and their respirations are deep and rapid. Fever is often present, and the temperature may be as high as 40.5C (105F). Adults with mild hypernatremia may have no symptoms, but as plasma sodium levels rise above 160 mEq/L, neurologic signs become apparent.36,37 Early symptoms include lethargy, nausea, and tremor, which progress to irritability, drowsiness, and confusion. Later features of muscular rigidity, opisthotonus, seizures, and coma reflect generalized cerebral and neuromuscular dysfunction. The most severe neurologic disturbances are seen at both ends of the age spectrum. The severity of such disturbances is also related to the rate at which hypernatremia develops, as well as to the absolute degree of hyperosmolality. Intracerebral vascular lesions are often the cause of death. In contrast to patients with the life-threatening clinical features of hypernatremic dehydration, patients with long-standing, moderate hypernatremia (plasma sodium concentrations of 145 to 160 mEq/L) may have few manifestations of the disorder other than lack of thirst. Hypodipsia is the crucial symptom, but it is often overlooked in the clinical setting because patients fail to complain of lack of thirst. However, careful evaluation of these patients reveals that some have no desire to drink any fluid under any circumstances, which suggests a total loss of the thirst osmoreceptor function. Others have only minimal thirst with marked hypertonicity, whereas a third group eventually experiences a normal thirst sensation, but only at high plasma osmolality levels. The key to recognizing subtle differences in thirst appreciation rests with a satisfactory measure of thirst. Visual analog scales for measuring thirst during dynamic tests of osmoregulation38,39 have been shown to produce highly reproducible results.40 When these scales are used in evaluating healthy persons, a linear increase is noted in the degree of thirst and fluid intake with increase in plasma osmolality, and an osmolar threshold for thirst is seen that is a few milliosmoles per kilogram higher than the osmolar threshold for vasopressin secretion.39 The application of these techniques to patients with chronic hypernatremia has disclosed numerous disorders of osmoregulation. OSMOREGULATORY DEFECTS IN CHRONIC HYPERNATREMIA Chronic hypernatremia is characterized by inappropriate lack of thirst despite increased plasma osmolality and mild hypovolemia. Plasma sodium concentrations are typically elevated (150160 mmol/L) and may reach extremely high concentrations during intercurrent illnesses (e.g., gastroenteritis) in which body water deficits increase. Although adipsic hyper-natremia is uncommon, four distinct patterns of abnormal osmoregulatory function have been described. Type 1 Adipsia. The characteristic abnormalities in type 1 adipsia are subnormal vasopressin levels and thirst responses to osmotic stimulation (Fig. 26-4). The sensitivity of the osmoreceptors is decreased, producing partial diabetes insipidus and relative hypodipsia. Because some capacity remains to secrete vasopressin and experience thirst, such patients are protected from extremes of hypernatremia, as they can produce near-maximal antidiuresis as plasma osmolality increases. Patients with this type of adipsia usually have normal vasopressin responses to hypotension and hypoglycemia, and show suppression of vasopressin secretion, with the development of hypotonic diuresis in response to water loading.

FIGURE 26-4. Thirst and vasopressin responses to osmotic stimulation in adipsic hypernatremia. Type 1: subnormal response of both thirst and vasopressin secretion. Type 2: total lack of response of thirst and vasopressin secretion. Type 3: reset of osmostat for vasopressin release and thirst to the right of normal. Shaded areas

indicate the response ranges in healthy control subjects; the dotted lines are the mean regression lines. (pAVP, plasma arginine vasopressin; LD, limit of detection of the pAVP assay [0.3 pmol/L]).

Type 2 Adipsia. Total ablation of the osmoreceptors produces complete diabetes insipidus and absence of thirst in response to hyperosmolality. This is the pattern of osmoregulatory abnormality seen after surgical clipping of aneurysms of the anterior communicating artery,41,42 and despite the complete absence of osmoregulated thirst and vasopressin release, thirst and vasopressin responses to hypotension and apomor-phine are preserved.41,43 Some patients also develop this type of osmoregulatory dysfunction after surgery for large, suprasellar craniopharyngiomas. Interestingly, these patients also have absent baroregulated thirst and vasopressin secretionpresumably because the extent of surgical injury is such that both the osmoreceptors and the paraventricular and supraoptic nuclei are damaged. Patients with complete adipsic diabetes insipidus have no defense against dehydration, and unless they are closely supervised and trained to drink even in the absence of thirst, they can develop profound hypernatremic dehydration, even in the absence of intercurrent illness. Interest has been shown in the concept that osmoreceptor activity is under bimodal control; that is, a specific stimulus is required to switch off vasopressin secretion in the same way that elevation of plasma osmolality stimulates vasopressin secretion. Patients with complete osmoreceptor ablation clearly are unable to respond to inhibitory inputs; this has been demonstrated in clinical studies in which complete suppression of the secretion of the small quantities of radioimmunoassayable vasopressin or the achievement of maximal free water clearance during water loading was impossible in a patient with this type of osmoregulatory dysfunction.44 Therefore, in some patients vasopressin secretion may not be entirely suppressed during fluid loads, resulting in significant hyponatremia. Type 3 Adipsia. The osmostats for thirst and vasopressin release may be reset to the right of normal (type 3 in Fig. 26-4), such that vasopressin secretion and thirst do not occur until higher plasma osmolalities are reached. Thereafter, the slope of the osmoregulatory lines are normal. This pattern is found in conjunction with a number of cases of essential hypernatremia, although type 1 defects have also been reported.45,46 and 47 Patients also have intact nonosmotic release of vasopressin and increased renal sensitivity to vasopressin, so that renal concentrating ability may be reasonably well maintained. Miscellaneous Causes of Adipsia. Osmoregulatory dysfunction has also been reported in elderly patients, who have diminished thirst in response to hypernatremia.38 Although the defect in thirst appreciation is similar to that in type 1 dysfunction, vasopressin responses have variously been reported as being subnormal, normal, or enhanced. Survivors of diabetic hyper-osmolar, nonketotic coma have also been reported to have hypodipsia with exaggerated vasopressin secretion.35 In addition, a single case has been reported of a young patient who had hypodipsia but a normal osmotically regulated vasopressin release.48 All of these reports lend support to the hypothesis that the osmoreceptors subserving vasopressin release are anatomically and functionally distinct from those controlling thirst. TREATMENT Water replacement is the basic therapy for patients with hyper-osmolar states associated with dehydration. The oral route is preferred, but if the clinical situation warrants urgent treatment, the infusion of hypotonic solutions may be necessary. However, overzealous rehydration with hypotonic fluids may result in seizures, neurologic deterioration, coma, and even death secondary to cerebral edema.34,37 Therefore, the decision to treat with hypotonic intravenous fluids should not be made lightly, and rehydration to a euosmolar state should proceed cautiously over at least 72 hours. As plasma osmolality falls, polyuria indicative of hypothalamic diabetes insipidus may develop; this responds to administration of desmopressin. For patients with chronic hyperosmolar syndromes (see Fig. 26-4), longer-term therapy must be considered. Patients with type 3 defects rarely need specific therapy because their osmoregulatory system is essentially intact but operates around a higher than normal plasma osmolality. Patients with type 1 defects (involving partial destruction of the osmoreceptor) should be treated with a regimen of increased water intake (24 L every 24 hours). If this leads to persistent polyuria, a small dose of desmopressin can be administered, but plasma osmolality or sodium levels must then be monitored regularly. Considerable difficulties arise in treating patients who have complete destruction of their osmoreceptors (type 2 defect), because these patients cannot protect themselves from extremes of dehydration and overhydration. Most patients need between 2 and 4 L of fluid per day, but the precise amount varies according to seasonal climatic changes, and the body weight must be monitored daily to provide an index of fluid balance.44 Regular (usually weekly) measurements of plasma osmolality or sodium are needed to ensure that no significant fluctuations occur in body water, and constant supervision is required to make certain the requisite volume of water is consumed. Despite the most vigorous supervision, such patients are extremely vulnerable to swings in plasma osmolality and are particularly prone to severe hypernatremic dehydration. CHAPTER REFERENCES
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. De Wardener HE, Herxheimer A. The effect of high water intake on the kidneys' ability to concentrate urine in man. J Physiol (Lond) 1957; 139:42. Robertson GL. Diagnosis of diabetes insipidus. In: Czernichow P, Robinson AG, eds. Diabetes insipidus in man. Frontiers of hormone research, vol 13. Basel: S Karger, 1985:176. Maffly RH. Diabetes insipidus. In: Andreoli TE, Grantham JJ, Rector FCJ, eds. Disturbances in body fluid osmolality. Bethesda, MD: American Physiology Society, 1977:285. Robertson GL. Diabetes insipidus. Endocrinol Metab Clin North Am 1995; 24:549. Ito M, Mori Y, Oiso Y, Saito H. A single base substitution in the coding region for neurophysin II associated with familial central diabetes insipidus. J Clin Invest 1991; 87:725. Krishnamani MRS, Philips PA III, Copeland KC. Detection of a novel arginine vasopressin defect by dideoxy fingerprinting. J Clin Endocrinol Metab 1993; 77:596. McLeod JF, Kovacs L, Gaskill MB, et al. Familial neurohypophyseal diabetes insipidus associated with a signal peptide mutation. J Clin Endocrinol Metab 1993; 77:599A. Heppner C, Kotzka J, Bullmann C, et al. Identification of mutations of the arginine vasopressin-neurophsia II gene in two kindreds with familial central diabetes insipidus. J Clin Endocrinol Metab 1998; 83:693. Rotig A, Cormier V, Chatelain P. Deletion of the mitochondrial DNA in a case of early-onset diabetes mellitus, optic atrophy and deafness, Wolfram syndrome (MIM 222300). J Clin Invest 1993; 91:1095. Barrett TG, Bundey SE. Wolfram (DIDMOAD) syndrome. J Med Genet 1997; 34:838. Verbalis JG, Robinson AG, Moses AM. Postoperative and post-traumatic diabetes insipidus. In: Czernichow P, Robinson AG, eds. Diabetes insipidus in man. Frontiers of hormone research, vol 13. Basel: S Karger, 1985:247. Moses AM. Clinical and laboratory observations in the adult with diabetes insipidus and related syndromes. In: Czernichow P, Robinson AG, eds. Diabetes insipidus in man. Frontiers of hormone research, vol 13. Basel: S Karger, 1985:156. Baylis PH, Cheetham T. Diabetes insipidus. Arch Dis Child 1998; 79:84. Scherbaum WA, Bottazzo GF. Autoantibodies to vasopressin cells in idiopathic diabetes insipidus: evidence for an autoimmune variant. Lancet 1983; 1:897. Imura H, Nakao K, Shimatsu A, et al. Lymphocytic infundibuloneurohypo-physitis as a cause of central diabetes insipidus. N Engl J Med 1993; 329:683.

15a.Knoers NV, Monnens LL. Nephrogenic diabetes insipidus. Semin Nephrol 1999; 19:344. 15b.Bendz H, Aurell M. Drug-induced diabetes insipidus. Drug Saf 1999; 21:449. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. Bichet DG, Birnbaumer M, Louergan M, et al. Nature and recurrence of AVPR2 mutations in X-linked nephrogenic diabetes insipidus. Am J Hum Genet 1994; 55:278. Hochberg Z, van Lieburg A, Even L, et al. Autosomal recessive nephrogenic diabetes insipidus caused by an aquaporin-2 mutation. J Clin Endocrinol Metab 1997; 82:686. Sato N, Ishizaka H, Yagi H, et al. Posterior lobe of the pituitary in diabetes insipidus: dynamic MR imaging. Radiology 1993; 186:357. Dashe AM, Cramm RE, Crist CA, et al. A water deprivation test for the differential diagnosis of polyuria. JAMA 1963; 185:699. Miller MT, Dalakos T, Moses AM, et al. Recognition of partial defects in antidiuretic hormone secretion. Ann Intern Med 1970; 73:721. Baylis PH. Diabetes insipidus. Medicine 1997; 25:9. Robertson GL. The regulation of vasopressin function in health and disease. Recent Prog Horm Res 1977; 33:333. Moses A, Streeten D. Differentiation of polyuric states by measurement of responses to changes in plasma osmolality induced by hypertonic saline infusions. Am J Med 1967; 42:368. Baylis PH, Robertson GL. Vasopressin response to hypertonic saline infusion to assess posterior pituitary function. J R Soc Med 1980; 73:255. Zerbe RL, Robertson GL. A comparison of plasma vasopressin measurement with a standard indirect test in the differential diagnosis of polyuria. N Engl J Med 1981; 305:1539. Thompson CJ, Baylis PH. Thirst in diabetes insipidus: clinical relevance of quantitative assessment. Q J Med 1987; 65:853. Davison JM, Gilmore EA, Drr J, et al. Altered osmotic thresholds for vasopressin secretion and thirst in human pregnancy. Am J Physiol 1984; 246:F105. Amico J. Diabetes insipidus in pregnancy. In: Czernichow P, Robinson AG, eds. Diabetes insipidus in man. Frontiers in hormone research, vol 13. Basel: S Karger, 1985:266. Hime MC, Richardson JA. Diabetes insipidus and pregnancy: case report, incidence and review of the literature. Obstet Gynecol Surv 1978; 33:375. Barron WM, Cohen LH, Ulland LA, et al. Transient vasopressin-resistant diabetes insipidus of pregnancy. N Engl J Med 1984; 310:442. Hughes JM, Barron WM, Vance ML. Recurrent diabetes insipidus associated with pregnancy: pathophysiology and therapy. Obstet Gynecol 1989; 73:462. Cobb WE, Spare S, Reichlin S. Diabetes insipidus: management with DDAVP (1-desamino-8- D-arginine vasopressin). Ann Intern Med 1978; 88:183. Williams TDM, Dungar DB, Lyon CC, et al. Antidiuretic effect and pharma-cokinetics of oral 1-desamino-8- D-arginine vasopressin. 1. Studies in adults and children. J Clin Endocrinol Metab 1986; 63:129. Ross EJ, Christie SBM. Hypernatremia. Medicine (Baltimore) 1969; 48:441. McKenna K, Morris AM, Azam H, et al. Subnormal osmotically stimulated thirst and exaggerated vasopressin release in human survivors of hyperosmolar coma. Diabetologia May 1999; 42:538. Robertson GL, Aycinena P, Zerbe RL. Neurogenic disorders of osmoregulation. Am J Med 1982; 72:339. Arieff AL, Guisado R. Effects on the central nervous system of hypernatremic and hyponatremic states. Kidney Int 1976; 10:104. Phillips PA, Rolls BJ, Ledingham JGG, et al. Reduced thirst after water deprivation in healthy elderly men. N Engl J Med 1984; 311:753. Thompson CJ, Thompson J, Burd J, Baylis PH. The osmotic threshold for thirst and vasopressin release are similar in healthy men. Clin Sci 1986; 71:651. Thompson CJ, Selby P, Baylis PH. Reproducibility of osmotic and nonosmotic tests of vasopressin secretion in men. Am J Physiol 1991; 260:R533. Pearce SHS, Argent NB, Baylis PH. Chronic hypernatremia due to impaired osmoregulated thirst and vasopressin secretion. Acta Endocrinol (Copenh) 1991; 125:234. McIver B, Connacher A, Whittle A, et al. Adipsic diabetes insipidus after clipping of anterior communicating artery aneurysm. BMJ 1991; 303:1465.

43. 44. 45. 46. 47. 48.

Teelucksingh S, Steer CR, Thompson CJ, et al. Hypothalamic syndrome and central sleep apnea associated with toluene exposure. Q J Med 1991; 286:185. Ball SG, Vaidja B, Baylis PH. Hypothalamic adipsic syndrome: diagnosis and management. Clin Endocrinol 1997; 47:405. De Rubertis FR, Michelis MF, Beck N, et al. Essential hypernatremia due to ineffective osmotic and intact volume regulation of vasopressin secretion. J Clin Invest 1971; 50:97. Dunger DB, Seckl JR, Lightman SL. Increased renal sensitivity to vasopressin in two patients with essential hypernatremia. J Clin Endocrinol Metab 1987; 64:185. Gill G, Baylis PH, Burn J. A case of essential hypernatremia due to resetting of the osmostat. Clin Endocrinol (Oxf) 1985; 22:545. Hammond DN, Moll GW, Robertson GL, Chelmicks-Schorr E. Hypodipsic hypernatremia with normal osmoregulation of vasopressin. N Engl J Med 1986; 315:433.

CHAPTER 27 INAPPROPRIATE ANTIDIURESIS AND OTHER HYPOOSMOLAR STATES Principles and Practice of Endocrinology and Metabolism

CHAPTER 27 INAPPROPRIATE ANTIDIURESIS AND OTHER HYPOOSMOLAR STATES

JOSEPH G. VERBALIS Frequency and Significance of Hypoosmolality Definition of Hypoosmolality Situations in Which Hyponatremia does not Reflect True Hypoosmolality Influence of Unmeasured Solutes Pathogenesis of Hypoosmolality Solute Depletion Water Retention Cellular Inactivation of Solute Differential Diagnosis of Hypoosmolality Decreased Extracellular Fluid Volume (Hypovolemia) Increased Extracellular Fluid Volume (Edema, Ascites) Normal Extracellular Fluid Volume (Euvolemia) Syndrome of Inappropriate Antidiuresis Diagnostic Criteria Etiology Pathophysiology Clinical Manifestations of Hypoosmolality Therapeutic Approach to Hypoosmolality Acute Treatment of Hypoosmolality Long-Term Treatment of Hypoosmolality Chapter References

FREQUENCY AND SIGNIFICANCE OF HYPOOSMOLALITY

Hypoosmolality of plasma is relatively common in hospitalized patients. The incidence and prevalence of hypoosmolar disorders depend on the nature of the patient population being studied and on the laboratory methods and diagnostic criteria used. Most investigators have used the serum sodium concentration ([Na+]) to determine the clinical incidence of hypoosmolality. When hyponatremia is defined as a serum [Na+] of <135 mEq/L, an incidence as high as 15% to 30% has been observed in patients hospitalized for both short1 and longer2 periods. However, the incidence decreases to a range of 1% to 4% when only patients with a serum [Na+] of <130 to 131 mEq/L are included,1,3 although even when these more stringent criteria are used, incidences of 7% to 30% have been reported in hospitalized geriatric patients.2 Interestingly, many studies have noted a high incidence of iatrogenic or hospital-acquired hyponatremia, which accounts for 40% to 75% of all cases examined.3,4 Furthermore, a frequency analysis of a large group of hospitalized patients demonstrated serum sodium and chloride concentrations that were ~5 mEq/L lower than those in a control group of healthy, nonhospitalized subjects.5 Consequently, the conclusion can be drawn that most cases of hypoosmolality are relatively mild and that it is often acquired during hospitalization. Although this might appear to signify that hypoosmolality generally is of little clinical significance, this conclusion is unwarranted for several reasons. First, severe hypoosmolality (serum [Na+] level of <125 mEq/L), although relatively uncommon, is associated with substantial morbidity and mortality. Second, even relatively mild hypoosmolality can progress to more dangerous levels during the therapeutic management of other disorders. Finally, the observation has been made that mortality is as much as 60-fold higher in patients with even asymptomatic degrees of hypoosmolality than in normonatremic patients.3 Although this probably indicates that hypoosmolality is more an indicator of the severity of many underlying illnesses than an independent factor contributing to mortality, this presumption may not be true in all cases. These considerations, therefore, emphasize the importance of a careful evaluation of all hyponatremic patients, regardless of the clinical setting.

DEFINITION OF HYPOOSMOLALITY
The only clinical abnormality known to result from increased secretion of arginine vasopressin (AVP) is a decrease in the osmotic pressure of body fluids. Hence it is appropriate to begin this chapter by considering how osmotic pressure, or osmolality, is determined and the factors that can influence these measurements. The osmolality of body fluid normally is maintained within narrow limits by osmotically regulated AVP secretion and thirst (see Chap. 25). Although basal plasma osmolality can vary appreciably among individuals, under conditions of normal hydration, the range in the general population lies between 275 and 295 mOsm/kg H2O. Plasma osmolality can be determined by direct measurement using either freezing-point depression or vapor pressure. Alternatively, it can be calculated from the concentrations of the major solutes in serum: pOsm (mOsm/kg H2O) = 2 [Na+] (mEq/L) + glucose (mg/dL)/18 + blood urea nitrogen (mg/dL)/2.8 The two methods produce comparable results under most conditions. Although these methods yield valid measures of total osmolality, however, this is not always equivalent to effective osmolality, sometimes referred to as the tonicity of the fluid.6 Only those solutes that remain relatively compartmentalized within the extracellular fluid (ECF) space (e.g., sodium and chloride) are considered to be effective solutes, because they create osmotic gradients across cell membranes and thus are capable of causing osmotic movement of water from the intracellular fluid (ICF) compartment to the ECF compartment. In contrast, solutes that readily permeate cell membranes (e.g., urea, ethanol, methanol) are not effective solutes, because they do not create osmotic gradients across cell membranes and thus are not capable of causing water movement between fluid compartments. Consequently, only the concentration of effective solutes in serum should be used to define clinically significant hypoosmolality or hyperosmolality of serum, because these are the only solutes that affect cellular hydration. SITUATIONS IN WHICH HYPONATREMIA DOES NOT REFLECT TRUE HYPOOSMOLALITY Because sodium and its accompanying anions represent the major effective serum solutes, hyponatremia and hypoosmolality are usually synonymous. However, two situations exist in which hyponatremia does not reflect true hypoosmolality. The first is pseudohyponatremia, produced by marked elevations of bulky solutes such as lipids or proteins in serum. In such cases, the concentration of sodium per liter of serum water is unchanged, but the concentration of sodium per liter of serum is artifactually decreased because of the increased relative proportion of serum volume that is occupied by the lipids or proteins. Because the increase in lipids or proteins does not appreciably increase the total number of solute particles in solution, however, the measured plasma osmolality is not affected significantly. In addition, measurement of serum [Na+] by ion-specific electrode, a method currently used by most clinical laboratories, is less influenced by high concentrations of lipids or proteins than was measurement of serum [Na+] by flame photometry.7 The second situation in which hyponatremia does not reflect true plasma hypoosmolality occurs when high concentrations of effective solutes other than sodium are present in the ECF, causing relative decreases in serum [Na+ ] despite an unchanged total effective osmolality. This effect usually is seen with hyperglycemia and represents a very common cause of hyponatremia, accounting for 10% to 20% of cases in hospitalized patients.3 Misdiagnosis of hypoosmolality in such cases can again be avoided by measuring plasma osmolality directly or, alternatively, by correcting the measured serum [Na+ ] by 1.6 mEq/L for each 100 mg/dL-increase in serum glucose concentration above normal levels.6 INFLUENCE OF UNMEASURED SOLUTES Plasma osmolality cannot be calculated when the ECF contains significant amounts of unmeasured solutes, such as osmotic diuretics, radiographic contrast agents, and some toxins (ethanol, methanol, and ethylene glycol). In these situations, plasma osmolality is better ascertained by direct measurement, although even this method does not yield an accurate measure of the true effective plasma osmolality if the unmeasured solutes are noneffective solutes, such as ethanol, that freely permeate cells. As a result of these considerations, the ascertainment of true hypoosmolality can sometimes be difficult. Nevertheless, a straightforward and relatively simple approach suffices in most cases. This entails initially calculating the effective plasma osmolality from the measured serum [Na+] and glucose concentrations (2 [Na+] + glucose 18) or, alternatively, simply correcting the serum [Na+] for elevations of serum glucose, as described previously. If the calculated effective osmolality is <275 mOsm/kg H2O, or the corrected serum [Na+] is <135 mEq/L, significant hypoosmolality is present, assuming that significant concentrations of unmeasured solutes or

pseudohyponatremia from hyperlipidemia or hyperproteinemia are not present. If uncertainty exists about either of these latter points, the plasma osmolality should also be measured directly. The absence of a significant discrepancy (<10 mOsm/kg H2O) between the measured and calculated total plasma osmolality (called the osmolar gap) confirms the absence of significant amounts of unmeasured solutes, such as osmotic diuretics, radiocontrast agents, or alcohol. However, if a significant discrepancy between these measures is found, appropriate tests must then be conducted to rule out pseudohyponatremia or to identify possible unmeasured serum solutes. Whether significant hypoosmolality exists in the latter case depends on the nature of the unmeasured solutes. Although this determination may not always be possible, in such cases the clinician at least is alerted to the uncertainty of true hypoosmolality.

PATHOGENESIS OF HYPOOSMOLALITY
The presence of true hypoosmolality always signifies an excess of water relative to solute in the ECF. Because water moves freely across most cell membranes between the ICF and ECF, this usually also implies an excess of total body water relative to total body solute. As shown in Table 27-1, an imbalance between water and solute can be generated either by a depletion of total body solute in excess of concurrent body water losses, or by a dilution of total body solute secondary to increases in total body water. This classification represents an obvious oversimplification, because most hypoosmolar states have components of both solute depletion and water retention. Nonetheless, it represents a valid starting point for understanding the mechanisms underlying the pathogenesis of hypoosmolality and also provides a useful framework for discussing the treatment of hypoosmolar disorders.

TABLE 27-1. Pathogenesis of Hypoosmolar Disorders

SOLUTE DEPLETION Depletion of body solute can result from any significant ECF losses. Whether by renal or nonrenal routes, the fluid loss itself rarely causes hypoosmolality because excreted or secreted fluid usually is isotonic or hypotonic relative to serum. Therefore, when hypoosmolality accompanies ECF losses, it is generally the result of replacement of the fluid losses by hypotonic solutions, causing dilution of the remaining body solutes. This usually occurs when patients drink only water in response to ongoing solute and water losses, but it can also occur when hypotonic fluids are administered intravenously to hospitalized patients. When solute losses are marked, these patients show obvious signs of volume depletion (e.g., addisonian crisis; see Chap. 76). However, such patients often have a more deceptive clinical picture because their volume deficits may be partially replaced. Moreover, they may not manifest signs or symptoms of cellular dehydration because osmotic gradients act to draw water into the relatively hypertonic ICF compartment, causing cellular volume expansion. Consequently, although clinical evidence of hypovolemia strongly supports solute depletion as the cause of plasma hypoosmolality, the absence of clinically evident hypovolemia never completely eliminates this possibility. Although ECF solute loss is responsible for most cases of depletion-induced hypoosmolality, ICF solute loss can also lead to hypoosmolality as a result of osmotic water shifts into the ECF. This mechanism likely contributes to the hypoosmolality that is observed in some cases of diuretic-induced hypoosmolality in which marked depletion of total body potassium occurs.8 WATER RETENTION Despite the importance of solute depletion in some patients, most cases of clinically significant hypoosmolality are caused by increases in total body water rather than by losses of extra-cellular solute. Theoretically, this can occur because of either impaired renal free water excretion or excessive free water intake. However, the former accounts for most hypoosmolar disorders because normal kidneys have sufficient diluting capacity to allow excretion of 20 to 30 L per day of free water.9 Intakes of this magnitude are occasionally seen in psychiatric patients, but not in most patients with primary polydipsia. Furthermore, studies have demonstrated that psychotic patients with polydipsia and hyponatremia also have abnormalities of free water excretion.10 Consequently, dilutional hypoosmolality usually implies an abnormality of renal free water excretion. The renal mechanisms responsible for impairments in free water excretion are commonly classified according to whether the major effect occurs in the proximal or distal nephron (see Table 27-1). Any disorder that leads to a decrease in glomerular filtration rate (GFR) causes increased reabsorption of both sodium and water in the proximal tubule. Consequently, the ability to excrete free water is limited because of decreased delivery of tubular fluid to the distal nephron. Disorders causing solute depletion through nonrenal mechanisms also produce the same effect. Disorders that cause a decreased glomerular filtration rate in the absence of significant ECF fluid losses are, for the most part, edema-forming states associated with decreased effective intravascular volume and secondary hyperaldosteronism. Severe hypothyroidism causes a similar effect, but does so through mechanisms that have not been delineated fully. However, even though these conditions traditionally have been considered to be primarily disorders of excess proximal fluid reabsorption, convincing evidence now exists that water retention also results from increased distal reabsorption caused by hemodynamically stimulated increases in serum AVP levels.11 Distal nephron impairments in free water excretion are characterized by an inability to dilute tubular fluid maximally. Such disorders are usually associated with abnormalities in AVP secretion, as depicted in Table 27-1. Just as depletion-induced hypoosmolar disorders usually include an important component of secondary impairment of free water excretion, so do most dilution-induced hypoosmolar disorders involve a component of secondary solute depletion. This effect has been recognized ever since early studies of the effects of AVP on water retention demonstrated that renal sodium excretion in such subjects was predominantly a result of ECF volume expansion from retained water.12 Thus, after sustained increases in total body water caused by inappropriately elevated AVP levels, sufficient secondary solute losses can occur to lower plasma osmolality further or, more often, to allow the maintenance of a lowered plasma osmolality with lesser degrees of ECF volume expansion from retained water.13 The actual contribution of solute losses to the hypoosmolality of inappropriate antidiuresis is variable and in experimental14,15 and clinical16 studies has been shown to depend on both the rate and the volume of AVP-induced ECF water expansion. Some dilutional disorders do not fit well into either category. Among these is the hyponatremia that sometimes occurs in patients who ingest large volumes of beer with little food intake for prolonged periods, frequently called beer potomania. Even though the volume of fluid ingested may not seem sufficiently excessive to overwhelm renal-diluting mechanisms, in these cases free water excretion is limited by very low urinary solute excretion, thereby causing water retention and dilutional hyponatremia. However, because such patients also have very little food intake, relative depletion of body sodium stores probably also is a contributing factor to the hypoosmolality in some cases. CELLULAR INACTIVATION OF SOLUTE Even the combined effects of water retention plus urinary solute excretion cannot adequately explain the degree of plasma hypoosmolality observed in some patients.17 This observation led to the theory of cellular inactivation of solute. Simply stated, this theory maintains that, as ECF osmolality falls, water moves into cells along osmotic gradients, causing the cells to swell. At some point during volume expansion, these cells osmotically inactivate some of the intracellular solutes as a defense mechanism to prevent continued cellular swelling and subsequent detrimental effects on cell function. With this decrease in intracellular osmolality, some water then shifts back out of the ICF into the ECF, but at the cost of worsening the dilution-induced ECF hypoosmolality. Despite the appeal of this theory, its validity has never been demonstrated conclusively in either human or animal studies. In part, this is because the production of unexplained hypoosmolality has generally been observed only with severe degrees of hypoosmolality (i.e., serum osmolalities of <240 mOsm/kg H2O), which is difficult to study clinically. However, animal studies have failed to

demonstrate this effect even with severe, sustained hypoosmolality.18 An alternative theory has been suggested by in vitro studies of cultured cells, which regulate cell volume in response to hypoosmolar conditions by extrusion of potassium rather than osmotic inactivation of cellular solute.19 Subsequent animal studies confirmed that after long periods of sustained hypoosmolality, the brain regulates its volume back to normal levels20; moreover, this process is accompanied by marked losses of organic solutes, called organic osmolytes,21 as well as electrolytes22 from the brain. Consequently, cellular volume regulation in vivo is more likely to occur predominantly through a process of depletion, rather than inactivation, of a variety of intracellular solutes.13,23

DIFFERENTIAL DIAGNOSIS OF HYPOOSMOLALITY

The diagnostic approach to patients with hypoosmolality includes a careful history-taking (especially with regard to medications); clinical assessment of ECF volume status; a complete neurologic evaluation; measurement of serum electrolytes, glucose, blood urea nitrogen and creatinine; calculated or direct measurements of plasma osmolality, or both; and simultaneous measurement of urinary electrolytes and osmolality. Because of the multiplicity of disorders that can cause hypoosmolality and the fact that many involve more than one pathologic mechanism, a definitive diagnosis is not always possible at the time of presentation. Nonetheless, a relatively simple evaluation with a careful interpretation of laboratory results provides a sufficient categorization of the underlying cause in most cases to allow appropriate decisions regarding initial therapy. Table 27-2 summarizes a method for the differential diagnosis of hypoosmolality based on the commonly used parameters of ECF volume status and spot urine sodium concentration (urine osmolality generally is elevated to varying degrees in most of these disorders and, therefore, is of limited value in differential diagnosis).

TABLE 27-2. Differential Diagnosis of Hypoosmolar Disorders

DECREASED EXTRACELLULAR FLUID VOLUME (HYPOVOLEMIA) The presence of clinically detectable hypovolemia always implies total body solute depletion. If the urine [Na+] is low, a nonrenal cause of solute depletion should be sought. If the urine [Na+] is high despite hypoosmolality, renal causes of solute depletion are likely responsible, most commonly attributable to diuretic therapy. In such cases the possibility of adrenal insufficiency must always be considered as well, especially in the presence of hyperkalemia, hypoglycemia, or clinical signs of Addison disease or hypopituitarism (see Chap. 17 and Chap. 76). Finally, in hypovolemic patients not receiving diuretics and without evidence of adrenal insufficiency, salt-wasting nephropathy is possible (e.g., polycystic kidney disease, chronic interstitial nephritis), and usually requires further, more extensive evaluation of renal function for definitive diagnosis. INCREASED EXTRACELLULAR FLUID VOLUME (EDEMA, ASCITES) The presence of clinically detectable hypervolemia implies a total body sodium excess. In these patients, hypoosmolality results from an even greater expansion of total body water caused by a marked reduction in the rate of water excretion (and sometimes an increased rate of ingestion). The impairment in water excretion is caused by a decreased effective arterial blood volume, which increases the reabsorption of glomerular filtrate not only in the proximal nephron but also in the distal and collecting tubule by stimulating AVP secretion.11 By the time most of these disorders cause significant hypoosmolality, they are usually readily apparent; however, occasionally, hypoosmolality occurs during early stages, with relatively mild clinical manifestations of the underlying disease. These patients have a low urine [Na+] because of secondary hyperaldosteronism, which is also a product of decreased effective intravascular volume. However, under certain conditions, urine [Na+] may be elevated, usually secondary to concurrent diuretic therapy, but also sometimes because of a solute diuresis (e.g., gluco-suria in diabetics) or after successful treatment of the underlying disease (e.g., ionotropic therapy in patients with congestive heart failure). One must recognize that primary polydipsia does not cause signs of hypervolemia, because water ingestion alone, without sodium retention, does not produce clinically apparent degrees of ECF volume expansion. One other important disorder that can produce hypoosmolality and hypervolemia is acute or chronic renal failure with fluid overload. Urine [Na+] in these cases is usually elevated, but it can be variable, depending on the stage of renal failure. This should not represent a diagnostic problem, except perhaps in the early stages of acute tubular necrosis before significant elevations of blood urea nitrogen and creatinine have occurred. NORMAL EXTRACELLULAR FLUID VOLUME (EUVOLEMIA) Hypoosmolality in the clinically euvolemic patient represents the greatest diagnostic challenge, because detecting modest changes in volume status is difficult using standard methods of clinical assessment. In such cases, the determination of urine [Na+ ] is an especially important first step.24 A high urine [Na+] in euvolemic patients usually implies a distally mediated, dilution-induced hypoosmolality, such as the syndrome of inappropriate antidiuresis (SIAD). Glucocorticoid deficiency can mimic SIAD so closely that these two disorders may be indistinguishable in terms of evaluation of water balance. This is because glucocorticoid deficiency can produce elevated serum AVP levels and, in addition, exerts a direct effect on the distal nephron to prevent maximal urinary dilution even in the absence of AVP.25 Clinically, this phenomenon is well known by virtue of the observation that cortisol insufficiency can mask diabetes insipidus in patients who have both disorders. Therefore, both primary and secondary adrenal insufficiency must be ruled out before a diagnosis of inappropriate antidiuresis is established. This is most easily accomplished with a rapid adrenocorticotropic hormone stimulation test (see Chap. 74). A low urine [Na+ ] suggests depletion-induced hypoosmolality secondary to ECF losses with subsequent volume replacement by water or other hypotonic fluids. The solute loss often is nonrenal in origin. An important exception is recent cessation of diuretic therapy, because urine [Na+] can quickly decrease again to low levels within 12 to 24 hours after discontinuation of the diuretic drug. Low urine [Na +] also can be seen in some cases of hypothyroidism, in the early stages of decreased effective intravascular volume (before the development of clinically apparent salt retention and fluid overload), or during the recovery phase of SIAD. Hence, a low urine sodium is less meaningful diagnostically than a high value. Hyponatremia caused by diuretic use can present without clinically evident hypovolemia, and urine [Na+] can be elevated in such cases because of the renal effects of the diuretics. The presence of a low serum [K+] is an important clue to diuretic use, because none of the other disorders that cause hypoosmolality are associated with significant hypokalemia. However, even in the absence of hypokalemia, any hypoosmolar, clinically euvolemic patient taking diuretics should be assumed to have solute depletion and should be treated accordingly. Subsequent failure to correct the hypoosmolality with isotonic saline administration and a persistence of an elevated urine [Na+] after discontinuation of diuretics then changes the diagnosis to a dilution-induced hypoosmolality by exclusion.

SYNDROME OF INAPPROPRIATE ANTIDIURESIS

DIAGNOSTIC CRITERIA The syndrome of inappropriate antidiuresis is the most common cause of euvolemic hypoosmolality. It is also the most prevalent cause of hypoosmolality of all etiologies encountered in clinical practice today, with an incidence of 30% to 40% among all hypoosmolar patients.2,3 The clinical criteria necessary for a diagnosis of SIAD remain basically as set forth by Bartter and Schwartz in 1967.26 A modified summary of these criteria is presented in Table 27-3 along with other findings that support this diagnosis.

TABLE 27-3. Syndrome of Inappropriate Antidiuresis: Diagnostic Criteria

Several of these criteria deserve emphasis or qualification. First, true hypoosmolality must be present. Hyponatremia secondary to pseudohyponatremia or hyperglycemia alone must be excluded. Second, urinary concentration (osmolality) must be inappropriate for plasma hypoosmolality. This does not mean that urine osmolality must be greater than plasma osmolality (a common misinterpretation of this criterion), but simply that the urine osmolality must be greater than maximally dilute (maximal dilution in normal adults is <100 mOsm/kg H2O). One must also remember that urine osmolality need not be elevated inappropriately at all levels of plasma osmolality because in one form of SIAD, the reset osmostat variant, AVP secretion can be suppressed and maximal urinary dilution and free water excretion can occur if plasma osmolality is decreased to sufficiently low levels. Hence, satisfaction of criteria for the diagnosis of SIAD requires only that urine osmolality (or serum AVP) be inadequately suppressed at some level of plasma osmolality below 275 mOsm/kg H2O. Third, clinical euvolemia must be present to establish a diagnosis of SIAD, because both hypovolemia and hypervolemia strongly suggest different causes of hypoosmolality. This does not mean that patients with SIAD cannot become hypovolemic or hypervolemic for other reasons, but in such cases, the underlying inappropriate antidiuresis cannot be diagnosed until the patient is rendered euvolemic and is found to have persistent hypoosmolality. The fourth criterion, renal salt wasting, has probably caused the most confusion over the years regarding the diagnosis of SIAD. This criterion is included primarily because of its usefulness in differentiating between hypoosmolality caused by a decreased effective extracellular volume (in which case renal sodium conservation occurs) and distally mediated dilution-induced disorders, in which urinary sodium excretion is normal or increased secondary to ECF volume expansion. However, two important qualifications limit the utility of urine [Na+] measurement in hypoosmolar patients: (a) urine [Na+] also is high when solute depletion is of renal origin, as with diuretic use or Addison disease, and (b) patients with SIAD can have low urinary sodium excretion if they subsequently become hypovolemic or solute depleted, conditions that sometime follow severe salt and water restriction. Consequently, although high urinary sodium excretion is the rule in most patients with SIAD, its presence certainly does not guarantee this diagnosis, and, conversely, its absence does not necessarily rule out the diagnosis. The fifth criterion emphasizes that, in many ways, SIAD remains a diagnosis of exclusion. Thus, the presence of other potential causes of euvolemic hypoosmolality must always be excluded. This includes not only thyroid and adrenal dysfunction, but also diuretic use, because this also often presents as euvolemic hypoosmolality. Table 27-3 lists three other criteria that support, but are not essential for, the diagnosis of SIAD. The first of these criteria, water loading, is of value when uncertainty exists regarding the cause of modest degrees of hypoosmolality in euvolemic patients (it does not add useful information if the plasma osmolality is already significantly lower than 275 mOsm/kg H2O). An inability to excrete a standard water load normally (with normal excretion defined as a cumulative urine output of at least 90% of the administered water load within 4 hours and suppression of urine osmolality to <100 mOsm/kg H2O27) confirms the presence of an underlying defect in free water excretion. Unfortunately, water loading is abnormal in almost all disorders that cause hypoosmolality, whether dilutional or depletion induced with secondary impairments in free water excretion. Two exceptions to this are primary polydipsia, in which hypoosmolality can rarely be secondary to excessive water intake alone, and the reset osmostat variant of SIAD, in which normal excretion of a water load can occur once plasma osmolality falls below the new setpoint for AVP secretion. The water-load test is also useful to assess water excretion after treatment of an underlying disorder thought to be causing SIAD. For example, after discontinuation of a drug associated with SIAD in a patient who has already achieved a normal plasma osmolality by fluid restriction, a normal water-load test can confirm the absence of persistent inappropriate antidiuresis much more quickly than can simple monitoring of the serum [Na+] during ad libitum fluid intake. Despite its limitations as a diagnostic clinical test, water loading remains an extremely useful tool in clinical research for quantitating changes in free water excretion in response to physiologic or pharmacologic manipulations. The second supportive criterion for a diagnosis of SIAD is an inappropriately elevated AVP level in relation to plasma osmolality. At the time that SIAD was originally described, inappropriately elevated levels of AVP were merely postulated, because the measurement of serum levels of AVP was limited to relatively insensitive bioassays. With the development of sensitive AVP radioimmunoassays capable of detecting the small physiologic concentrations of this peptide that circulate in serum, the hope was that the accurate measurement of serum AVP levels might supplant the classic criteria and become the definitive test for diagnosing SIAD. This has not occurred for several reasons. First, although serum AVP levels are elevated in most patients with this syndrome, the elevations generally remain within the normal physiologic range and are abnormal only in relation to plasma hypoosmolality. This is demonstrated in Figure 27-1.28 Thus, AVP levels can be interpreted only in conjunction with a simultaneous plasma osmolality and a knowledge of the relation between AVP levels and plasma osmolality in normal subjects. Second, 10% to 20% of patients with SIAD do not have measurably elevated AVP levels. As shown in Figure 27-1, many patients have AVP levels that are precisely at, or even below, the limits of detection by radioimmunoassay. Whether these cases are true examples of inappropriate antidiuresis in the absence of AVP, possibly secondary to abnormal regulation of the aquaporin-2 water channels that mediate AVP-stimulated water reabsorption in the collecting ducts,29 or whether they simply represent inappropriate AVP levels that fall below the limits of detection by radioimmunoassay is not clear at this time. For this reason, it is more accurate to use the term SIAD than the originally proposed designation of syndrome of inappropriate antidiuretic hormone secretion (SIADH) to describe this entire group of disorders.30 Third, just as water loading fails to distinguish between various causes of hypoosmolality, so do AVP levels. Many disorders causing solute and volume depletion are associated with elevations of serum AVP levels secondary to nonosmotic hemodynamic stimuli. For similar reasons, patients with disorders that cause decreased effective volume, such as congestive heart failure and cirrhosis, also have elevated AVP levels.11 Even glucocorticoid insufficiency has been associated with inappropriately elevated AVP levels in animal studies, although experience in humans is limited. Thus, many different disorders can cause stimulation of AVP secretion through nonosmotic mechanisms.

FIGURE 27-1. Plasma vasopressin (AVP) levels are shown as a function of plasma osmolality in patients with the syndrome of inappropriate antidiuresis. Each point depicts one patient. The shaded area represents AVP levels in normal subjects over physiologic ranges of plasma osmolality. (From Robertson GL, Aycinena P, Zerbe RL. Neurogenic disorders of osmoregulation. Am J Med 1982; 72:339.)

The final supportive criterion, an improvement in plasma osmolality with fluid restriction but not with volume expansion, can be helpful in differentiating between disorders causing solute depletion and those associated with dilution-induced hypoosmolality. The infusion of isotonic sodium chloride in patients with SIAD is well known to provoke a natriuresis with little correction of osmolality,17,26 whereas fluid restriction allows such patients to achieve solute and water balance gradually through insensible free water losses. In contrast, isotonic saline is the treatment of choice in disorders of solute depletion, because once volume deficits are corrected, the stimulus to AVP secretion and free water retention is eliminated. The diagnostic value of this therapeutic response is limited somewhat by the fact that patients with proximal types of dilution-induced disorders may show a response similar to that found in patients with SIAD. The major drawback of this criterion, however, is that it represents a retrospective test in a situation in which establishing a diagnosis before making a decision regarding treatment options would be preferable. Nonetheless, in difficult cases of euvolemic hypoosmolality, an appropriate therapeutic response can sometimes be helpful in diagnosing SIAD.

ETIOLOGY Disorders associated with SIAD can be divided into several major etiologic groups. These groups are summarized in Table 27-4.

TABLE 27-4. Syndrome of Inappropriate Antidiuresis: Etiologies

TUMORS The most common cause of SIAD is a tumor. Although several different types of tumors are listed in Table 27-4, bronchogenic carcinoma of the lung has been uniquely associated with SIAD since the first description of this disorder in 1957.17 In virtually all cases, the bronchogenic carcinomas causing this syndrome have been of the small cell (or oat cell) variety (a few squamous cell types have been described, but these are rare). The unusually high incidence of small cell carcinoma of the lung in patients with SIAD, together with the relatively favorable therapeutic response of this type of tumor, makes it imperative that all adult patients presenting with an otherwise unexplained SIAD be investigated thoroughly and aggressively for a possible tumor. The evaluation should include a chest computed tomographic scan or magnetic resonance imaging study, with bron-choscopy and cytologic analysis of bronchial washings in any suspicious cases. The reason for this approach is that several studies have reported hypoosmolality that predated any radiographically evident abnormality in patients who developed bronchogenic carcinoma 3 to 12 months later (see Chap. 219). CENTRAL NERVOUS SYSTEM DISORDERS The second major etiologic group of disorders causing SIAD has its origins in the central nervous system (CNS). Despite the diversity of CNS disorders associated with SIAD, no apparent common denominator links them. This is not surprising when one considers the neuroanatomy of neurohypophyseal innervation. As discussed in Chapter 25, the magnocellular AVP neurons receive excitatory input from osmoreceptor cells of the anterior hypothalamus as well as major innervation from brainstem cardiovascular regulatory centers. Although these pathways have yet to be elucidated fully, at least some of them appear to be inhibitory. Any diffuse CNS disorder can potentially disrupt this long and probably multisynaptic pathway from the brain stem to the hypothalamus and, in so doing, can cause AVP hypersecretion and SIAD as a result of a decrease in inhibitory tone originating from these pathways.30a Consequently, a variety of CNS disorders can cause this syndrome, but only destructive lesions involving the hypothalamus or pituitary cause diabetes insipidus. An interesting cause of transient CNS SIAD has been described after pituitary surgery. As many as 33% of patients have been found to develop moderate hyponatremia 5 to 7 days after transsphenoidal resection of pituitary adenomas due to inappropriate AVP secretion as a result of neural lobe/pituitary stalk damage.31 This had largely gone unnoticed in the past, not only because most patients are discharged before this time, but also because, although most patients develop some degree of inappropriate AVP secretion (as documented by impaired water-loading tests),32 only those few who continue to ingest large volumes of fluid actually develop symptomatic hyponatremia and come to medical attention. DRUGS The third major category, drug-induced SIAD, is rapidly becoming the most common cause of hypoosmolality.32a In general, pharmacologic agents cause this syndrome by directly stimulating AVP secretion, by directly activating AVP renal receptors to cause antidiuresis, or by potentiating the effect of AVP on the kidney. Not all of the drug effects associated with inappropriate antidiuresis are fully understood, however; indeed, many agents may work by means of a combination of mechanisms.33 For example, chlorpropamide appears to have both a direct pituitary and a renal stimulatory effect. Newly developed drugs are regularly being added to the list of agents that induce SIAD, particularly psychotropic drugs such as selective serotonin reuptake inhibitors used to treat depression; these have been reported to cause hyponatremia in up to 22% to 28% of elderly patients.34 Agents that cause AVP secretion primarily through solute depletion are best considered to cause depletion-induced hypoosmolality rather than true SIAD. PULMONARY DISORDERS Pulmonary disorders, as a group, represent an often-mentioned but largely misunderstood cause of SIAD. A variety of pulmonary disorders have been associated with this syndrome, but except in tuberculosis, advanced chronic obstructive lung disease, and acute pneumonias, the occurrence of hypoosmolality has been noted only in sporadic case reports. A review of these reports reveals that virtually all cases of nontuberculous pulmonary SIAD have occurred in the setting of acute respiratory failure with marked hypoxia, hypercapnia, or both. When such patients were evaluated serially, the inappropriate AVP secretion was found to be limited to the initial days of hospitalization, when respiratory failure was most marked or when patients were ventilated by mechanical means. Even the cases of tubercular SIAD uniformly occurred in patients with far-advanced, active pulmonary tuberculosis. Therefore, although SIAD may occur fairly commonly with nontumor-related pulmonary disease, two factors should be remembered: (a) in all cases, the pulmonary disease is obvious as a result of severe dyspnea or extensive, radiographically evident infiltrates; and (b) the inappropriate antidiuresis is usually limited to the period of respiratory failure. Once clinical improvement has begun, free water excretion generally improves rapidly. Mechanical ventilation can cause inappropriate AVP secretion, or it can worsen any SIAD caused by other factors. This phenomenon has been associated most often with continuous positive-pressure ventilation, but it can also occur with the use of positive end-expiratory pressure. UNEXPLAINED (IDIOPATHIC) OR MULTIFACTORIAL CAUSES Unexplained or idiopathic SIAD accounts for a small number of cases. Although the cause of the syndrome may not be readily diagnosed initially, the number of patients in whom an apparent cause cannot be established after consistent follow-up over time is relatively low. Hospitalized patients with acquired immunodeficiency syndrome (AIDS) have been recognized to have a high incidence of hyponatremia, as high as 30% to 50% in some series.35 This represents perhaps the most classic example of a truly multifactorial hyponatremia. Approximately half of such patients have a depletional hyponatremia from primary solute losses, both from diarrheal fluid losses and (in some patients) from adrenal insufficiency due to cytomegalovirus adrenalitis (see Chap. 76 and Chap. 214); others have true SIAD from an acute pneumonia, typically due to Pneumocystis carinii; a smaller subset have drug-induced SIAD; and still others with CNS human immunodeficiency virus (HIV) or opportunistic infections have SIAD from disordered AVP regulation. Clearly, many potential etiologies must be considered in the differential diagnosis of hyponatremic patients with HIV infections. Some known stimuli to AVP secretion are notable because of their exclusion from Table 27-4. Despite stimulation of AVP secretion by nicotine (see Chap. 234), cigarette smoking has not been associated with SIAD, presumably because of chronic adaptation. Similarly, although nausea is the most potent known stimulus to AVP secretion, and although the AVP secreted in response to nausea clearly is bioactive,36 in the absence of vomiting, with subsequent ECF and solute depletion, chronic nausea is generally not associated with hypoosmolality. This is probably attributable to the fact that such patients are not inclined to drink fluids under such circumstances (however, hyponatremia can occur when such patients are infused with high volumes of hypotonic fluids). Finally, a causal relation between chronic stress (either physical or mental) and SIAD has not been established. This underscores the fact that stress, independent of nausea or hypotension, has never been clearly shown to cause significantly elevated AVP levels and, in fact, has been found to suppress AVP secretion in animals. PATHOPHYSIOLOGY Disorders that cause inappropriate antidiuresis secondary to elevated serum AVP levels can be subdivided into those associated with either paraneoplastic (ectopic) or pituitary AVP hypersecretion. Most ectopic production is from tumors, and conclusive, cumulative evidence exists that tumor tissue can, in fact, synthesize AVP. Evidence of such synthesis has been derived from several findings: (a) tumor extracts have been found to possess antidiuretic hormone bioactivity and immunologically

recognizable AVP and neurophysin, (b) electron microscopy has revealed that many tumors possess secretory granules, and (c) cultured tumor tissue has been shown to synthesize not only AVP but also the entire AVP prohormone (provasopressin). Although clearly some tumors can produce AVP, whether or not all tumors associated with SIAD do so is not certain, because only approximately half of small cell carcinomas have been found to contain AVP immunoreactivity,37 and many of the tumors listed in Table 27-4 have not been studied as extensively as have bronchogenic carcinomas. Studies of tissue cultures from hyponatremic patients with small cell lung cancers have shown that many tumor cell lines produce mRNA for atrial natriuretic peptide (ANP) in addition to, or instead of, AVP mRNA,38 raising the possibility of ectopic ANP secretion. Clinical studies have verified the importance of elevated serum AVP levels in most hyponatremic patients with small cell lung cancers, but have also demonstrated elevated serum ANP levels, a finding which suggests that ANP might contribute to the hyponatremia in such patients by virtue of aggravating solute depletion as a result of increased natriuresis.39 The only nonneoplastic disorder that has been proven to cause SIAD by means of ectopic AVP production is tuberculosis. However, this conclusion is based on studies of a single patient in whom extracts of tuberculous lung were shown by bioassay to possess antidiuretic activity.40 Studies of serum AVP levels in patients with SIAD during graded increases in plasma osmolality produced by hypertonic saline administration have suggested four patterns of secretion (Fig. 27-2): (a) random hypersecretion of AVP; (b) a reset osmostat, whereby AVP is secreted at an abnormally low threshold of plasma osmolality but otherwise displays a normal response to relative changes in osmolality; (c) inappropriate hypersecretion below the normal threshold for AVP release, but normal secretion in response to osmolar changes within normal ranges of plasma osmolality; and (d) low or undetectable serum AVP levels despite clinical characteristics of SIAD.30

FIGURE 27-2. Schematic summary is shown of different patterns of vasopressin (AVP) secretion observed in patients with the syndrome of inappropriate antidiuresis. Each line (AD) represents the relation between serum AVP and plasma osmolality of individual subjects in whom osmolality was increased via hypertonic saline infusions. The shaded area represents serum AVP levels in normal subjects over physiologic ranges of plasma osmolality. (From Robertson GL, Aycinena P, Zerbe RL. Neurogenic disorders of osmoregulation. Am J Med 1982; 72:339.)

The first pattern simply represents unregulated AVP secretion, which is often observed in patients exhibiting paraneo-plastic AVP production. Nonosmotic resetting of the osmotic threshold for AVP secretion has been well described with volume depletion27 and also has been shown to occur in congestive heart failure, presumably as a result of decreases in effective arterial blood volume. However, many patients with a reset osmostat are clinically euvolemic; therefore, it has been suggested that chronic hypoosmolality itself may reset the intracellular threshold for osmoreceptor firing, although studies in hyponatremic animals have not supported a major role for this mechanism.41 Perhaps the most perplexing aspect of the reset osmostat pattern is its occurrence in some patients with tumors,28,30 which suggests that some of these cases represent tumor-stimulated pituitary AVP secretion rather than ectopic AVP production. The pattern of SIAD that occurs without measurable AVP secretion is not yet understood. This form of the syndrome may be attributable to the secretion of AVP with some bioactivity but altered immunoreactivity, to the presence of other circulating antidiuretic factors, to increased renal sensitivity to very low circulating levels of AVP, to abnormalities of kidney aquaporin-2 regulation,29 or in some cases to ANP-induced natriuresis.38,39 To date, a sufficient number of patients with this form of the disorder have not been studied adequately to form any basis for discrimination among these possibilities (although on theoretic grounds, increased renal sensitivity to low levels of AVP appears most plausible). Despite these well-described patterns of inappropriate AVP secretion in SIAD, however, the fact that no correlation has been found between any of these four patterns and the various causes of the syndrome is surprising.30

CLINICAL MANIFESTATIONS OF HYPOOSMOLALITY

Hypoosmolality is associated with a broad spectrum of neurologic manifestations, ranging from mild nonspecific symptoms (e.g., headache, nausea) to more significant disorders (e.g., disorientation, confusion, obtundation, and seizures). In the most severe cases, death can result from respiratory arrest after tentorial cerebral herniation and brainstem compression. This neurologic symptom complex has been termed hyponatremic encephalopathy and primarily reflects brain edema resulting from osmotic water shifts into the brain because of decreased effective plasma osmolality. Significant symptoms generally do not occur until the serum [Na+] falls below 125 mEq/L, and the severity of symptoms can be roughly correlated with the degree of hypoosmolality4 (Fig. 27-3). Individual variability is marked, however, and for any single patient, the level of serum [Na+ ] at which symptoms appear cannot be predicted with great accuracy. Furthermore, several factors other than the severity of the hypoosmolality also affect the degree of neurologic dysfunction. Most important is the period over which hypoosmolality develops. Rapid development of severe hypoosmolality is frequently associated with marked neurologic symptoms, whereas gradual development over several days or weeks is often associated with relatively mild symptomatology despite profound degrees of hypoosmolality. This is because the brain can counteract osmotic swelling by excreting intracellular solutes (including potassium4 and organic osmolytes22), an adaptive process known as volume regulation.21 Because this is a time-dependent process, rapid development of hypoosmolality can result in brain edema before this adaptation occurs, but with slower development of the same degree of hypoosmolality, brain cells can lose solute sufficiently rapidly to prevent brain edema and neurologic dysfunction.42 This accounts for the much higher incidence of neurologic symptoms as well as higher mortality rates in patients with acute hyponatremia than in those with chronic hyponatremia.4 A striking example of this is the fact that the most dramatic cases of death due to hyponatremic encephalopathy have generally been reported in postoperative patients,43 in whom hyponatremia can develop rapidly as a result of postoperative retention of hypotonic fluid infusions.

FIGURE 27-3. The relationship between serum sodium concentration [Na+ ] and central nervous system (CNS) symptoms in hypoosmolar patients is depicted. Each point represents a different patient. Correlation is expressed between serum [Na+] levels and the level of CNS depression, as indicated by the scale at the top. (From Arieff AJ, Llach F, Massy SG. Neurological manifestations and morbidity of hyponatremia: correlation with brain water and electrolytes. Medicine [Baltimore] 1976; 55:121.)

Underlying neurologic disease also affects the level of hypoosmolality at which CNS symptoms appear; for example, moderate hypoosmolality is generally of little concern in an otherwise healthy patient but can cause morbidity in a patient with an underlying seizure disorder. Nonneurologic metabolic disorders (e.g., hypoxia, hypercapnia, acidosis, hypercalcemia) similarly can affect the level of plasma osmolality at which CNS symptoms occur. Clinical studies have suggested that menstruating women43 and young children44 may be particularly susceptible to the development of neurologic morbidity and mortality during hyponatremia, especially in the acute postoperative setting. The true clinical incidence,45 as well as the underlying mechanisms responsible for these sometimes catastrophic cases, remains to be determined.42 One final consideration is that nausea and vomiting are frequently overlooked as potential signs of increased intracranial pressure in hypoosmolar patients. Because hypoosmolality does not cause any known direct effects on the gastrointestinal tract, the presence of unexplained nausea or vomiting in a

hypoosmolar patient must be assumed to be of CNS origin, and the patient should be treated appropriately for symptomatic hypoosmolality.

THERAPEUTIC APPROACH TO HYPOOSMOLALITY

Despite continuing controversy concerning the rapid correction of osmolality in hypoosmolar patients, relative consensus exists regarding appropriate treatment of this disorder (Fig. 27-4). Once true hypoosmolality is verified, the ECF volume status of the patient should be assessed by careful clinical examination. If fluid retention is present, the treatment of the underlying disease should take precedence over the correction of plasma osmolality. Often, this involves treatment with diuretics, which, by virtue of the excretion of hypotonic urine, should simultaneously improve serum tonicity. If any degree of hypovolemia is present, the patient must be considered to have depletion-induced hypoosmolality, in which case volume repletion with isotonic saline (0.9% sodium chloride) at a rate appropriate for the estimated volume depletion should be initiated immediately. If diuretic use is known or suspected, the isotonic saline should be supplemented with potassium (30 to 40 mEq/L), even if the serum [K+ ] is not low, because of the propensity of such patients to develop total body potassium depletion. Most often, the hypoosmolar patient is clinically euvolemic, in which case the evaluation should include the measurement of urine osmolality and [Na+ ]. Several situations dictate the reconsideration of solute depletion as a potential diagnosis, even in the patient without clinically apparent hypovolemia. These include a decreased urine [Na+], any history of recent diuretic use, and any suggestion of primary adrenal insufficiency. In fact, whenever a reasonable possibility of depletion-induced, rather than dilution-induced, hypoosmolality exists, the patient is most appropriately treated initially with isotonic saline, regardless of whether signs of hypovolemia are present. An improvement in and eventual correction of osmolality support a diagnosis of solute and volume depletion,24 although SIAD may also resolve spontaneously. If the patient has SIAD rather than solute depletion, no harm comes from administration of a limited amount (i.e., 12 L) of isotonic saline, because such patients simply excrete excess sodium chloride without a significant change in plasma osmolality.17

FIGURE 27-4. Schematic summary of the evaluation and treatment of hypoosmolar patients is presented. The dark black arrow indicates that central nervous system (CNS) symptomatology should always be evaluated immediately in all hypoosmolar patients, even while the outlined diagnostic evaluation is still proceeding. (N, no; Y, yes; ECF, extracellular fluid; NSS, normal saline solution [0.9% NaCl]; RX, treat/treatment; P osm, plasma osmolality; D/C, discontinue; AVP, arginine vasopressin; SIAD, syndrome of inappropriate antidiuresis.)

The approach to patients with SIAD varies according to the clinical situation. A patient who meets all essential criteria for the syndrome but has a low urine osmolality should be observed on a trial of modest fluid restriction. If the hypoosmolality is attributable to transient SIAD or severe polydipsia, the urine will remain dilute, and the plasma osmolality will be corrected as free water is excreted. If, however, the patient has the reset osmostat form of the disorder, then the urine will become concentrated at some point before the plasma osmolality and serum [Na+ ] return to normal ranges. If either primary or secondary adrenal insufficiency is suspected, glucocorticoid replacement should be initiated immediately after the completion of a rapid adrenocorticotropic hormone stimulation test. A prompt water diuresis after initiation of glucocorticoid treatment supports a diagnosis of glucocorticoid deficiency. However, the absence of a quick response does not necessarily negate this diagnosis because several days of glucocorticoid replacement are sometimes required for normalization of plasma osmolality. If hypothyroidism is suspected, thyroid function tests should be conducted and a serum thyroid-stimulating hormone level should be obtained; usually, however, thyroid replacement therapy is withheld pending these results unless the patient is obviously myxedematous (see Chap. 45). The presence of renal failure in a patient with hypoosmolality generally requires a more extensive evaluation of renal function before treatment is initiated. ACUTE TREATMENT OF HYPOOSMOLALITY In any significantly hypoosmolar patient, one is immediately faced with the question of how quickly the plasma osmolality should be corrected. Although hyponatremia is associated with a broad spectrum of neurologic symptoms,4 sometimes leading to death in severe cases,43,44 too rapid correction of severe hyponatremia can produce pontine and extrapontine myelinolysis, a brain demyelinating disease that also can cause substantial neurologic morbidity and mortality.46,47 Furthermore, some studies have suggested that the morbidity and mortality associated with even symptomatic hyponatremia may not be as high as previously thought.48 Reviews of clinical and experimental results have suggested that optimal treatment of hyponatremic patients must, therefore, entail balancing the risks of hyponatremia against the risks of correction for each patient. 42,49,50,51 and 52 Although individual variability in response remains great, and consequently which patients will develop neurologic complications from either hyponatremia or its correction cannot always be accurately predicted, consensus guidelines for treating hypoosmolar patients allow a rational approach that minimizes the risks of both these complications. Implicit in these guidelines is the concept that treatment must be tailored to the patient's clinical presentation: appropriate therapy for one patient may be inappropriate for another patient despite equivalent degrees of hypoosmolality. Three major factors should be taken into consideration when making a treatment decision with regard to a hypoosmolar patient: the severity of the hyponatremia, the duration of the hyponatremia, and the patient's neurologic symptomatology. The severity of the hypoosmolality is an important consideration because neither sequelae from hyponatremia itself nor myelinolysis after therapy are likely in patients whose serum [Na+] remains 120 mEq/L (although significant symptoms can develop even at higher serum [Na+] levels if the rate of fall of plasma osmolality is particularly rapid). The importance of duration and symptomatology relate to how well the brain has volume-adapted to the hyponatremia and consequently its degree of risk for subsequent demyelination with rapid correction.42 Cases of acute hyponatremia (defined as 48 hours in duration) are usually symptomatic if the hyponatremia is severe (i.e., 120 mEq/L). These patients are at greatest risk for neurologic complications from the hyponatremia, and their serum [Na+] should be corrected to higher levels promptly. Postoperative patients, particularly young women and children,43,44 appear to be at greatest risk for rapidly progressing hyponatremic encephalopathy and should be evaluated carefully for any neurologic symptomatology. The dark black arrow in Figure 27-4 emphasizes that hypoosmolar patients should always be evaluated quickly for the presence of neurologic symptoms so that appropriate therapy can be initiated, if indicated, even while other results are still pending. Several studies have documented that rapid correction of serum [Na+] in patients with acute hyponatremia carries little, if any, risk of demyelination,53 presumably because sufficient brain volume regulation has not yet occurred to increase brain susceptibility to osmotic dehydration with the correction.42 Conversely, patients with more chronic hyponatremia (defined as >48 hours in duration) who have minimal neurologic symptomatology are at little risk from complications of hyponatremia itself but can develop demyelination after rapid correction because of greater degrees of brain volume regulation through electrolyte and osmolyte losses.54 No indication exists to correct these patients rapidly, regardless of the initial serum [Na+], and they should be treated using slower-acting therapies, such as fluid restriction. Although these extremes have clear treatment indications, the hyponatremia of most patients is of indeterminate duration and produces varying degrees of milder neurologic symptomatology. This group of patients is the most challenging to treat because the hyponatremia will have been present sufficiently long to allow some degree of brain volume regulation, but not long enough to prevent some brain edema and neurologic symptomatology. Most clinicians recommend prompt treatment of such patients because of their symptoms, using methods that allow a controlled and limited correction of their hypoosmolality. Some clinicians have suggested that correction parameters should consist of a maximal rate of correction of serum [Na+] in the range of 1 to 2 mEq/L per hour as long as the total magnitude of correction does not exceed 25 mEq/L over the first 48 hours.55 Other clinicians argue that these parameters should be more conservative, with maximal correction rates of 0.5 mEq/L per hour or less and magnitudes of correction that do not exceed 12 mEq/L in the first 24-hour period and 18 mEq/L in the first 48-hour period.49 A reasonable approach for treatment of individual patients would therefore entail choosing correction parameters within these limits depending on the patient's symptomatology13,42: when patients are only moderately symptomatic, one should proceed at the lower recommended limits of 0.5 mEq/L per hour or less, whereas when patients manifest more severe neurologic symptoms, an initial correction at a rate of 1 to 2 mEq/L per hour would be more appropriate. Regardless of the initial rate of correction chosen, acute treatment should be interrupted as soon as any of the following three end points is reached: (a) the patient's symptoms are abolished, (b) a safe serum [Na+] (generally 120 mEq/L) is achieved, or (c) a total magnitude of correction of 15 to 20 mEq/L is achieved. Once any of these end points is achieved, the active correction should be stopped and the patient treated with slower-acting therapies, such as oral rehydration or fluid restriction, depending on the cause of the hypoosmolality. It follows from these recommendations that serum [Na+ ] levels must be carefully monitored at frequent intervals (at least every 4 hours) during the active phases of treatment to adjust therapy to keep the correction parameters within these guidelines. Controlled limited corrections can be accomplished with either isotonic or hypertonic saline infusions, depending on the cause of the hypoosmolality. Patients with

volume depletion hypoosmolality (e.g., clinical hypovolemia, history of diuretic use, or urine [Na+ ] 30 mEq/L24) usually respond well to isotonic (0.9%) sodium chloride. Patients with euvolemic hypoosmolality (including patients with SIAD) do not respond to isotonic sodium chloride and are best treated with hypertonic (3%) sodium chloride solution given by continuous infusion. An initial infusion rate can be estimated by multiplying the patient's body weight in kilograms by the desired rate of increase in serum [Na+] in milliequivalents per liter per hour (e.g., in a 70-kg patient an infusion of 3% sodium chloride at 70 mL per hour will increase serum [Na+ ] by ~1 mEq/L per hour, whereas an infusion of 35 mL per hour will increase serum [Na+] by ~0.5 mEq/L per hour). Furosemide should be used to treat volume overload, in some cases anticipatorily in patients with known cardiovascular disease. Regardless of the therapy or rate initially chosen, one cannot emphasize too strongly that the plasma osmolality need be corrected acutely only to a safe range rather than completely to normonatremia. Rarely, hyponatremia may be spontaneously corrected by means of a water diuresis. If the hyponatremia is acute (e.g., psychogenic polydipsia with water intoxication), these patients do not appear to be at risk for subsequent demyelination.53 However, in cases in which the hyponatremia has been chronic (e.g., hypocortisolism), such patients are at some risk of demyelination, and intervention (e.g., administration of desmopressin) should be considered to limit the rate and magnitude of correction of serum [Na+], with use of the same end points as described previously for active corrections.13,42 Finally, any potassium deficits in hyponatremic patients should be corrected promptly, because hypokalemic patients appear to be at significantly greater risk for demyelination after correction of hyponatremia.56 LONG-TERM TREATMENT OF HYPOOSMOLALITY If SIAD persists, then investigation of other potential causes should be pursued (see Table 27-4), and long-term therapy should be instituted. The treatment of chronic SIAD entails choosing among several suboptimal therapeutic regimens. Continued fluid restriction represents the least toxic treatment choice and is the preferred treatment for most cases of mild to moderate SIAD. Several points should be remembered when using this approach: (a) all fluids, not only water, must be included in the restriction; (b) the degree of restriction required depends on urine output plus insensible fluid loss (generally, discretionary, i.e., nonfood, fluids should be limited to 500 mL a day below the average daily urine volume); (c) several days of restriction are usually necessary before a significant increase in plasma osmolality occurs; and (d) only fluid, not salt, should be restricted. Because of the ongoing natriuresis, these patients often have a negative total body sodium balance and therefore should be maintained on a relatively high sodium chloride intake unless otherwise contraindicated. However, just as failure to correct a presumed depletion-induced hyponatremia with isotonic saline should lead one to consider the possibility of a dilution-induced hypoosmolality, so should the failure of significant fluid restriction after several days of a confirmed negative fluid balance prompt reconsideration of other possible causes, including solute depletion and clinically inapparent hypovolemia. At the time that fluid restriction is first initiated, any drugs known to be associated with SIAD should be discontinued or changed (e.g., second-generation hypoglycemic agents such as glyburide should be substituted for first-generation agents such as chlorpropamide). Fluid restriction should always be tried as the initial therapy for patients with chronic SIAD, with pharmacologic intervention reserved for refractory cases in which the degree of fluid restriction required to avoid hypoosmolality is so severe that the patient is unable, or unwilling, to maintain it. In such cases, reasonable efforts should be made to ameliorate thirst, such as substituting hard candy or ice chips for fluid drinking. However, alternative pharmacologic management is often necessary. Pharmacologic intervention should also be avoided initially in patients with SIAD that is secondary to tumors, because successful treatment of the underlying malignant lesion often eliminates or reduces the inappropriate AVP secretion. If pharmacologic treatment is necessary, the preferred drug currently is the tetracycline derivative demeclocycline. This agent causes a nephrogenic form of diabetes insipidus, with decreased urine concentration even in the presence of high serum AVP levels. An appropriate dosage of demeclocycline ranges from 600 to 1200 mg per day, administered in divided doses. Treatment must be continued for several days to achieve maximal diuretic effects; consequently, one should wait 3 to 4 days before increasing the dosage. Demeclo-cycline can cause nephrotoxicity, especially in patients with cirrhosis, although this is generally reversible. Renal function should therefore be monitored on a regular basis and the medication discontinued if increasing azotemia is noted. Other agents, such as lithium, have similar effects, but their use is less desirable because of inconsistent results and significant side effects. Urea has been described as an alternative mode of treatment for SIAD.57 Although it has long been recognized that any osmotic diuretic can be used to treat hypoosmolality by virtue of increasing free water excretion, such therapeutic modalities have generally proved impractical for long-term ambulatory use. Urea is an exception because it can be administered orally; furthermore, it corrects hypoosmolality not only by increasing free water excretion but also by decreasing urinary sodium excretion. Dosages of 30 g per day are generally effective. (Dissolving the urea in orange juice or some other strongly flavored liquid to camouflage the taste is advisable.) Even if completely normal water balance is not achieved, the patient can often be allowed to maintain a less strict regimen of fluid restriction while receiving urea. The disadvantages associated with the use of urea include poor palatability, the development of azotemia at higher doses, and unavailability of a convenient form of the agent. Several other drugs that have been described appear to decrease AVP hypersecretion in some cases (e.g., diphenylhy-dantoin, opiates, ethanol), but responses are erratic and unpredictable. Much better therapeutic agents for chronic SIAD will probably soon be available, because the synthesis of a variety of analogs to AVP has yielded potent AVP-receptor antagonists. Particularly promising is the synthesis of several potent non-peptide antagonists that are selective for AVP V2 (antidiuretic) rather than V1 (pressor) receptors,58 or combine both activities.59 Clinical studies using one such agent have confirmed its efficacy in patients with SIAD.60 If such compounds prove to be effective and safe for human use in larger clinical trials, they undoubtedly will become the drugs of choice for euvolemic, and possibly hypervolemic, hypoosmolar patients, rendering all other modes of pharmacologic therapy obsolete. CHAPTER REFERENCES
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. Flear CTG, Gill GV, Burn J. Hyponatremia: mechanisms and management. Lancet 1981; 2:26. Kleinfeld M, Casimir M, Borra S. Hyponatremia as observed in a chronic disease facility. J Am Geriatr Soc 1979; 27:156. Anderson RJ, Chung H, Kluge R, Schrier RW. Hyponatremia: a prospective analysis of its epidemiology and the pathogenetic role of vasopressin. Ann Intern Med 1985; 102:164. Arieff AI, Llach F, Massry SG. Neurological manifestations and morbidity of hyponatremia: correlation with brain water and electrolytes. Medicine (Baltimore) 1976; 55:121. Owen JA, Campbell DG. A comparison of plasma electrolyte and urea values in healthy persons and in hospital patients. Clin Chim Acta 1968; 22:611. Alvis R, Geheb M, Cox M. Hypo- and hyperosmolar states: diagnostic approaches. In: Arieff AI, DeFronzo RA, eds. Fluid, electrolyte, and acid-base disorders, vol 2. New York: Churchill Livingstone, 1985:185. Ladenson JH, Apple FS, Koch DD. Misleading hyponatremia due to hyperlipemia: a method-dependent error. Ann Intern Med 1981; 95:707. Fichman MP, Vorherr H, Kleeman CR, Telfer N. Diuretic-induced hyponatremia. Ann Intern Med 1971; 75:853. Robertson GL. Psychogenic polydipsia and inappropriate antidiuresis. Arch Intern Med 1980; 140:1574. Goldman MB, Luchins DJ, Robertson GL. Mechanisms of altered water metabolism in psychotic patients with polydipsia and hyponatremia. N Engl J Med 1988; 318:397. Schrier RW. Pathogenesis of sodium and water retention in high-output and low-output cardiac failure, nephrotic syndrome, cirrhosis and pregnancy. N Engl J Med 1988; 319:1065. Leaf A, Bartter FC, Santos RF, Wrong O. Evidence in man that urinary electrolyte loss induced by Pitressin is a function of water retention. J Clin Invest 1953; 32:868. Verbalis JG. The syndrome of inappropriate hormone secretion and other hypoosmolar disorders. In: Schrier RW, Gottschalk CW, eds. Diseases of the kidney. Boston: Little, Brown and Company 1998:2393. Verbalis JG. Pathogenesis of hyponatremia in an experimental model of the syndrome of inappropriate antidiuresis. Am J Physiol 1994; 267:R1617. Smith MJ Jr, Cowley AW Jr, Guyton AC, Manning RD Jr. Acute and chronic effects of vasopressin on blood pressure, electrolytes, and fluid volumes. Am J Physiol 1979; 237:F232. Nolph KD, Schrier RW. Sodium, potassium and water metabolism in the syndrome of inappropriate antidiuretic hormone secretion. Am J Med 1970; 49:534. Schwartz WB, Bennett W, Curelop S, Bartter FC. A syndrome of renal sodium loss and hyponatremia probably resulting from inappropriate secretion of antidiuretic hormone. Am J Med 1957; 23:529. Gross PA, Anderson RJ. Effects of DDAVP and AVP on sodium and water balance in conscious rat. Am J Physiol 1982; 243:R512. Grantham JJ. Pathophysiology of hypoosmolar conditions: a cellular perspective. In: Andreoli TE, Grantham JJ, Rector FC Jr, eds. Disturbances in body fluid osmolality. Bethesda, MD: American Physiological Society, 1977:217. Verbalis JG, Drutarosky MD. Adaptation to chronic hypoosmolality in rats. Kidney Int 1988; 34:351. Gullans SR, Verbalis JG. Control of brain volume during hyperosmolar and hypoosmolar conditions. Annu Rev Med 1993; 44:289. Verbalis JG, Gullans SR. Hyponatremia causes large sustained reductions in brain content of multiple organic osmolytes in rats. Brain Res 1991; 567:274. Strange K. Regulation of solute and water balance and cell volume in the central nervous system. J Am Soc Nephrol 1992; 3:12. Chung H-M, Kluge R, Schrier RW, Anderson RJ. Clinical assessment of extracellular fluid volume in hyponatremia. Am J Med 1987; 83:905. Linas SL, Berl T, Robertson GL, et al. Role of vasopressin in the impaired water excretion of glucocorticoid deficiency. Kidney Int 1980; 18:58. Bartter FC, Schwartz WB. The syndrome of inappropriate secretion of antidiuretic hormone. Am J Med 1967; 42:790. Robertson GL. Diseases of the posterior pituitary. In: Felig P, Baxter JD, Frohman LA, eds. Endocrinology and metabolism. New York: McGraw-Hill, 1995:385. Robertson GL, Aycinena P, Zerbe RL. Neurogenic disorders of osmoregulation. Am J Med 1982; 72:339. Knepper MA. Molecular physiology of urinary concentrating mechanism: regulation of aquaporin water channels by vasopressin. Am J Physiol 1997; 272:F3. Zerbe R, Stropes L, Robertson G. Vasopressin function in the syndrome of inappropriate antidiuresis. Annu Rev Med 1980; 31:315.

30a.Liamis G, Elisaf M. Syndrome of inappropriate antidiuresis associated with multiple sclerosis. Neurol Sci 2000; 172:38. 31. Miller M, Moses AM. Drug-induced states of impaired water excretion. Kidney Int 1976; 10:96. 32. Strachan J, Shepherd J. Hyponatremia associated with the use of selective serotonin re-uptake inhibitors. Aust NZJ Psychiatry 1998; 32:295. 32a.Belton K, Thomas SH. Drug-induced syndrome of innapropriate antidiuretic hormone secretion. Postgrad Med J 1999; 75:509. 33. 34. 35. 36. 37. 38. Olson BR, Rubino D, Gumowski J, Oldfield EH. Isolated hyponatremia after transsphenoidal pituitary surgery. J Clin Endocrinol Metab 1995; 80:85. Olson BR, Gumowski J, Rubino D, Oldfield EH. Pathophysiology of hyponatremia after transsphenoidal pituitary surgery. J Neurosurg 1997; 87:499. Bevilacqua M. Hyponatraemia in AIDS. Baillieres Clin Endocrinol Metab 1994; 8:837. Rowe JW, Shelton RL, Helderman JH, et al. Influence of the emetic reflex on vasopressin release in man. Kidney Int 1979; 16:729. Vorherr H, Massry SG, Utiger RD, Kleeman CR. Antidiuretic principle in malignant tumor extracts from patients with inappropriate ADH syndrome. J Clin Endocrinol Metab 1968; 28:162. Gross AJ, Steinberg SM, Reilly JG, et al. Atrial natriuretic factor and arginine vasopressin production in tumor cell lines from patients with lung cancer and their relationship to serum sodium. Cancer Res 1993; 53:67. 39. Johnson BE, Chute BP, Rushin J, et al. A prospective study of patients with lung cancer and hyponatremia of malignancy. Am J Respir Crit Care Med 1997; 156:1669. 40. Vorherr H, Massry SG, Fallet R, et al. Antidiuretic principle in tuberculous lung tissue of a patient with pulmonary tuberculosis and hyponatremia. Ann Intern Med 1970; 72:383.

41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60.

Verbalis JG, Dohanics J. Vasopressin and oxytocin secretion in chronically hypoosmolar rats. Am J Physiol 1991; 261:R1028. Verbalis JG. Adaptation to acute and chronic hyponatremia: implications for symptomatology, diagnosis, and therapy. Semin Nephrol 1998; 18:3. Ayus JC, Wheeler JM, Arieff AI. Postoperative hyponatremic encephalopathy in menstruant women. Ann Intern Med 1992; 117:891. Arieff AI, Ayus JC, Fraser CL. Hyponatraemia and death or permanent brain damage in healthy children. BMJ 1992; 304:1218. Wijdicks EF, Larson TS. Absence of postoperative hyponatremia syndrome in young, healthy females. Ann Neurol 1994; 35:626. Sterns RH, Riggs JE, Schochet SS Jr. Osmotic demyelination syndrome following correction of hyponatremia. N Engl J Med 1986; 314:1535. Karp BI, Laureno R. Pontine and extra pontine myelinolysis: a neurologic disorder following rapid correction of hyponatremia. Medicine (Baltimore) 1993; 72:359. Sterns RH. Severe symptomatic hyponatremia: treatment and outcome. Ann Intern Med 1987; 107:656. Sterns RH, Cappuccio JD, Silver SM, Cohen EP. Neurologic sequelae after treatment of severe hyponatremia: a multicenter perspective. J Am Soc Nephrol 1994; 4:1522. Lauriat SM, Berl T. The hyponatremic patient: practical focus on therapy. J Am Soc Nephrol 1997; 8:1599. Fraser CL, Arieff AI. Epidemiology, pathophysiology, and management of hyponatremic encephalopathy. Am J Med 1997; 102:67. Gross P, Reimann D, Neidel J, et al. The treatment of severe hyponatremia. Kidney Int 1998; 64:S6. Cheng JC, Zikos D, Skopicki HA, et al. Long term neurologic outcome in psychogenic water drinkers with severe symptomatic hyponatremia: the effect of rapid correction. Am J Med 1990; 88:561. Verbalis JG, Gullans SR. Rapid correction of hyponatremia produces differential effects on brain osmolyte and electrolyte reaccumulation in rats. Brain Res 1993; 606:19. Ayus JC, Krothapalli RK, Arieff AI. Treatment of symptomatic hyponatremia and its relation to brain damage: a prospective study. N Engl J Med 1987; 317:1190. Lohr JW. Osmotic demyelination syndrome following correction of hyponatremia: association with hypokalemia. Am J Med 1994; 96:408. Decaux G, Brimioulle S, Fenette F, Mockel J. Treatment of the syndrome of inappropriate secretion of antidiuretic hormone by urea. Am J Med 1980; 69:99. Ohnishi A, Orita Y, Okahara R, et al. Potent aquaretic agent: a novel non-peptide selective vasopressin antagonist (OPC-31260) in men. J Clin Invest 1993; 92:2653. Yatsu T, Tomura Y, Tahara A, et al. Pharmacological profile of ym087, a novel nonpeptide dual vasopressin V1a and V2 receptor antagonist, in dogs. Eur J Pharmacol 1997; 321:225. Saito T, Ishikawa S, Abe K, et al. Acute aquaresis by the nonpeptide arginine vasopressin (AVP) antagonist OPC-31260 improves hyponatremia in patients with syndrome of inappropriate secretion of antidiuretic hormone (SIADH). J Clin Endocrinol Metab 1997; 82:1054.

CHAPTER 28 APPROACH TO THE PATIENT WITH THYROID DISEASE Principles and Practice of Endocrinology and Metabolism

CHAPTER 28 APPROACH TO THE PATIENT WITH THYROID DISEASE

LEONARD WARTOFSKY Medical History of the Thyroid Gland Evaluation of the Patient with Thyroid Disease Examination of the Thyroid Gland Chapter References

The following chapters on the thyroid were written with the clinician in mind. Adequate basic information is provided to serve as a background or rationale for positions taken on controversial issues, as well as to keep the reader abreast of the latest developments. The primary intention of this section, however, is to present expert critical appraisals of all important aspects of each clinical problem in a most current, factual, readable, and comprehensible format.

MEDICAL HISTORY OF THE THYROID GLAND

The naming of the thyroid gland has been attributed to Wharton1 in 1656, but an endocrine function was not proposed until almost 200 years later.2 Appreciation of the clinical disorders affecting the thyroid followed, the earliest descriptions of which appeared in the following order: thyroid cancer in 18113; diffuse toxic goiter by Parry in 1825,4 Graves in 1835,5 and von Basedow in 18406; cretinism in 18717; myxedema in 18748; thyroidectomy for the treatment of toxic goiter in 18849; thyroid extract therapy for myxedema in 189110; Hashimoto disease in 191211; subacute (de Quervain) thyroiditis in 193612; the structure of thyroxine (T 4) in 192613; identification of triiodothyronine (T3) in 195214 the presence of thyroid autoantibodies in Hashimoto disease in 195715; the earliest evidence of the thyroid-stimulating antibodies of Graves disease in 195616; recognition of medullary thyroid carcinoma as a distinct entity in 195917; and reports of cases of postpartum thyroiditis with hypothyroidism18 or thyrotoxicosis19 only since 1976 to 1977. Other subsequent milestones included the first description of resistance to thyroid hormone in 196720; substantiation that circulating T3 was derived largely from peripheral monodeiodination of T4 in 197021; identification of T3-binding receptors in tissues in 197222 and their homology to the viral oncogene erbA in 198623,24; and demonstrations that point mutations in the thyroid-hormone receptor accounted for hormone resistance in 1989 and 1990.25,26 The thyrotropin (TSH) receptor was cloned in 198927,28; studies since have identified both loss of function and gain of function mutations in the TSH receptor that account for specific types of hypothyroidism and hyperthyroidism, respectively.29 The gene for the b subunit of TSH was then cloned,30 facilitating the development of human recombinant TSH (rhTSH); this agent has great clinical utility for the diagnosis and treatment of thyroid cancer.31,32

EVALUATION OF THE PATIENT WITH THYROID DISEASE

The location of the thyroid makes it readily accessible to both inspection and palpation (Fig. 28-1). Often, the casual detection of a lump or mass in the thyroid by either the patient, a spouse, or an acquaintance first brings the patient in for evaluation. The thyroid examination is a mandatory and routine component of any complete history and physical examination. Aspects of the patient's clinical history may suggest thyroid dysfunction, in which case a careful thyroid examination is warranted. Symptoms of thyroid dysfunction may be caused by either hypofunction or hyperfunction of the thyroid, or they may be related to neck symptoms resulting from an enlarging nodule or goiter. The clinician then must assess the significance of the patient's complaints by simultaneously ascertaining both a functional and an anatomic diagnosis. The functional evaluation is based on various components of the history and physical examination, as well as the results of thyroid function tests (which, in general, are outlined in several chapters within this section; specific tests are described in relation to individual disorders that are discussed in separate chapters). The symptoms associated with deficient or excessive thyroid hormone production (Table 28-1) represent the most important and relevant information to be obtained on history-taking. When taking a history, the clinician should be attentive to the quality of the patient's voice: the presence of hoarseness suggests involvement of the recurrent laryngeal nerve, slow husky speech is characteristic of hypothy-roidism, and rapid, more staccato cadences often are present in thyrotoxic individuals. Occasionally, patients may have a serum TSH value that is below or above the normal range, suggestive of hyperthyroidism or hypothyroidism, respectively, but have no complaints or discernible symptoms. Often, this represents early or mild disease, which has not yet progressed to either overt symptoms or abnormalities in the levels of thyroid hormones and has been called chemical, subclinical, or minimally symptomatic thyroid disease.33,34 This entity is discussed further in Chapter 42 and Chapter 45. Important aspects of the clinical history in a patient with a goiter or a thyroid mass are discussed in Chapter 38, Chapter 39, Chapter 40 and Chapter 41.

FIGURE 28-1. Landmarks for the anatomic location of the thyroid gland.

TABLE 28-1. Symptoms and Signs Relevant to Disordered Thyroid Function

EXAMINATION OF THE THYROID GLAND

Some features of the general physical examination are of particular interest in a patient with a suspected thyroid disorder. These include the texture and temperature of the skin, the degree of perspiration, the pulse rate, the deep tendon reflex relaxation time, hoarseness of the voice, facial or periorbital puffiness, ophthalmopathy, edema, apparent weight loss, tremor, and reduced muscle mass or muscle weakness.

Examination of the thyroid is facilitated by the ready availability of water to expedite swallowing. Many patients tend to hyperextend their necks at the beginning of the examination, thinking that this simplifies the palpation when it actually makes it more difficult. The patient's neck should be flexed only slightly, so as to relax the sternocleidomastoid muscles. The neck is inspected for visible goiter or masses, both at rest and during swallowing. Nonthyroidal masses do not move up and down with swallowing because they do not lie within the fascial sheathing of the trachea. That a midline mass may represent a thyroglossal duct cyst (with embryo-logic derivation from the hypoglossal region) may be inferred from its cephalad excursion on protrusion of the tongue. Plethoric suffusion of the face on raising both arms to the sides of the head (Pemberton sign) suggests a retrosternal goiter brought up into a thoracic inlet narrowed by the goiter and causing obstruction to venous return. Palpation of the thyroid is performed best from behind, with the patient in a seated position (Fig. 28-2). The author does not concur with earlier dogma that maintained that a palpable thyroid indicated goiter or disease because the normal gland usually could not be palpated; with patience and experience, a normal gland can be palpated. The size of the normal thyroid varies directly with body size. The average thyroid gland weighs 14 to 18 g. Those physicians who estimate the weight of enlarged glands usually do so on the basis of a comparison to the weight of the normal gland. That is, a gland that is thought to be approximately three times larger than normal would be described as weighing 45 to 55 g.

FIGURE 28-2. Technique for physical examination of the thyroid gland. A, The thyroid is examined from behind, with the patient in a sitting position, avoiding hyperextension of the neck. B, The exploring fingers determine the extent of the gland, after which attention is directed to the size, consistency, and presence of any nodules. C, By using the fingertips alternately to displace the gland to the contralateral side, an appreciation of deeper abnormalities may be gained. This is particularly effective when the sternocleidomastoid muscle is thickened, in which case direct palpation would be difficult. D, Examination for lymphadenopathy is conducted in a routine manner but should be especially thorough in the presence of a thyroid nodule.

Using the fingertips of both hands, the examiner first evaluates the surface of the lateral lobes (see Fig. 28-2A). The fingertips then are moved progressively (see Fig. 28-2B, Fig. 28-2C and Fig. 28-2D) to outline the lateral and medial extent of each lobe, and thence to the connecting thyroid isthmus, which lies just below the cricoid cartilage. A pyramidal lobe most often is identified extending up from the isthmus and, when enlarged, may suggest either Hashimoto thyroiditis or Graves disease. Sometimes, a thyroid gland that appears to be enlarged on inspection proves to be of normal size on palpation. Such cases often represent pseudogoiter,35 which occurs when the gland is located higher than usual over the thyroid cartilage. In addition to size, the examiner should attempt to ascertain the consistency of the gland. The thyroid in a patient with Graves disease typically is soft, spongy, and as malleable as an uncooked sirloin steak but becomes progressively more firm after iodine exposure. This increase in firmness is attributable to involutional changes and an increased ratio of colloid to cellular mass. In general, a continued progression to increasing hardness is seen from Graves disease, to colloid goiter, to early Hashimoto disease with lymphocytic infiltration, to adenomatous and multinodular goiter, to late Hashimoto disease with fibrosis, to infiltrating thyroid malignant disease, and, finally, to Riedel (fibrosing) thyroiditis. An extremely hard consistency in a single thyroid nodule often, but not invariably, signals a malignant lesion; the hard texture may be attributable in part to psammoma bodies in papillary carcinoma or to the dense calcification seen with medullary thyroid cancer. In a patient with an obese or muscular neck, full delineation of the lateral lobes may be difficult. A useful maneuver in such a case is to displace the thyroid cartilage first to one side and then to the other with lateral pressure from the fingertips. This serves to displace the lobe outwardly on the contralateral side, making it more accessible to examination (see Fig. 28-2C). On palpation, the size and position, as well as the consistency, of any thyroid nodules should be noted. Transillumination may be helpful in establishing whether a nodule is cystic; the amount of light transmitted by a penlight pressed against a nodule (in a darkened room) is compared to that transmitted on the contralateral (normal) side. The size of a nodule can be estimated by direct measurement after marking the overlying skin, or by outlining the nodule on a 3- to 4-inch wide strip of adhesive tape placed over the nodule. This tape then can be removed and inserted into the patient's record to be used for future comparisons as a substitute for or adjunct to the precise sizing that may be achieved with ultrasonography. Alternatively, the size of a nodule may be estimated easily on the basis of the width of the examiner's index finger (i.e., usually 1.01.5 cm) by determining how much of the nodule is covered by the examining finger. Uniformly enlarged goiters may be monitored by comparing periodic measurements of neck circumference. This method involves the use of a tape measure and notations regarding fixed landmarks, such as the spine of the seventh cervical vertebra posteriorly and the midpoint of the isthmus anteriorly. Examination of the thyroid also should include a determination of the presence or absence of lymphadenopathy, particularly in the central cricoid area, where an enlarged Delphian node may be the earliest indication of metastatic papillary cancer. Occasionally, lymphadenopathy, which returns to normal after therapy, may be observed in a patient with Graves disease, and enlarged benign nodes may be seen more rarely in a patient with Hashimoto disease. The presence of a palpable thrill over the gland suggests increased vascularity, as seen in Graves disease, and should correlate with an audible bruit on auscultation. Gland tenderness indicates possible subacute thyroiditis, hemorrhage within a nodule, or malignant disease. Extremely rarely, this sign is observed in a patient with Hashimoto thyroiditis; however, it occurs commonly in the rare acute suppurative form of thyroiditis. Imaging studies, such as sonography, radioisotope scanning, or magnetic resonance imaging/computed tomography, may be useful to complement the physical examination (see Chap. 34 and Chap. 35). CHAPTER REFERENCES
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. Rolleston HD. The endocrine organs in health and disease. London: Oxford University Press, 1936. King TW. Observations on the thyroid gland. Guy's Hosp Rep 1836; 1:429. Burns P. Observations on the surgical anatomy of the head and neck. Edinburgh: Bryce, 1811:207. Parry CH. Collections from the unpublished papers of the late Caleb Hillier Parry, vol 2. London: 1825:111. Graves RJ. Clinical lectures. Lond Med Surg J (Part II) 1835; 7:516. von Basedow CA. Exophthalmos durch hypertrophie des Zellgewebes in der Augen hohle. Wochenschr Heilk 1840; 6:197. Fagge CH. On sporadic cretinism occurring in England. BMJ 1871; 1:279. Gull WW. On a cretinoid state supervening in adult life in women. Trans Clin Soc (Lond) 1874; 7:180. Rehn L. Uber die extirpation des Kropfs bei morbus Basedowii. Berlin Klin Wochenschr 1884; 21:163. Murray GR. Note on the treatment of myxoedema by hypodermic injections of an extract of the thyroid gland of a sheep. BMJ 1891; 2:796. Hashimoto H. Zur kenntnis der lymphomatoser veranderung der schildruse (struma lymphomatosa). Arch Klin Chir 1912; 97:219. De Quervain F, Giordanengo G. Die akute und subakute nichteitrige thyreoditis. Mitt Grenzgeb Med Chir 1936; 44:538. Harington CR. Chemistry of thyroxine I. Isolation of thyroxine from the thyroid gland. Biochem J 1926; 20:293. Gross J, Pitt-Rivers R. The identification of 3:5:3'-1-triido-thyronine in human plasma. Lancet 1952; 1:439. Doniach D, Roitt IM. Autoimmunity in Hashimoto's disease and its implications. J Clin Endocrinol Metab 1957; 17:1293. Adams DD, Purves HD. Abnormal responses in the assay of thyrotrophin. Proc Univ Otago Med Sch 1956; 34:11. Hazard JB, Hawk WA, Crile G Jr. Medullary (solid) carcinoma of the thyroid: a clinicopathologic entity. J Clin Endocrinol Metab 1959; 19:152. Amino N, Miyai K, Onishi T, et al. Transient hypothyroidism after delivery in autoimmune thyroiditis. J Clin Endocrinol Metab 1976; 42:296. Ginsberg J, Walfish PG. Post-partum transient thyrotoxicosis with painless thyroiditis. Lancet 1977; 1:1125. Refetoff S, DeWind LT, DeGroot LJ. Familial syndrome combining deafmutism, stippled epiphyses, goiter, and abnormally high PBI: possible target organ refractoriness to thyroid hormone. J Clin Endocrinol Metab 1967; 27:279. Braverman LE, Ingbar SH, Sterling K. Conversion of thyroxine (T 4) to triiodothyronine (T3) in athyreotic human subjects. J Clin Invest 1970; 49:855. Oppenheimer JH, Koerner D, Schwartz HL, Surks MI. Specific nuclear triiodothyronine binding sites in rat liver and kidney. J Clin Endocrinol Metab 1972; 35:330. Sap J, Munoz A, Damm K, et al. The c-erbA protein is a high-affinity receptor for thyroid hormone. Nature 1986; 324:635. Weinberger C, Thompson CC, Ong ES, et al. The c-erbA gene encodes a thyroid hormone receptor. Nature 1986; 324:641. Sakurai A, Takeda K, Ain K, et al. Generalized resistance to thyroid hormone associated with a mutation in the ligand-binding domain of the human thyroid hormone receptor b. Proc Natl Acad Sci U S A 1989; 86:8977. Usala SJ, Tennyson GE, Bale AE, et al. A base mutation of the c-erbA b thyroid hormone receptor in a kindred with generalized thyroid hormone resistance. Molecular heterogeneity in two other kindreds. J Clin Invest 1990; 85:93. Parmentier M, Libert F, Maenhaut C, et al. Molecular cloning of the thyrotropin receptor. Science 1989; 246:1620. Nagayama Y, Kaufman KD, Seto P, Rapoport B. Molecular cloning, sequence and functional expression of the cDNA for the human thyrotropin receptor. Biochem Biophys Res Commun 1989; 165:1184.

29. 30. 31. 32.

Morris JC. The clinical expression of thyrotropin receptor mutations. The Endocrinologist 1998; 8:195. Wondisford FE, Radovick S, Moates JM, et al. Isolation and characterization of the human thyrotropin beta-subunit gene. J Biol Chem 1988; 263:12538. Meier CA, Braverman LE, Ebner SA, et al. Diagnostic use of recombinant human thyrotropin in patients with thyroid carcinoma (Phase I/II study). J Clin Endocrinol Metab 1994; 78:188. Ladenson PW, Braverman LE, Mazzaferri EL, et al. Comparison of administration of recombinant human thyrotropin with withdrawal of thyroid hormone for radioactive iodine scanning in patients with thyroid carcinoma. N Engl J Med 1997; 337:888. 33. Haden ST, Marqusee E, Utiger RD. Subclinical hyperthyroidism. The Endocrinologist 1996; 6:322. 34. Ayala A, Wartofsky L. Minimally symptomatic (subclinical) hypothyroidism. The Endocrinologist 1997; 7:44. 35. Gwinup G, Morton E. The high lying thyroid: a cause of pseudogoiter. J Clin Endocrinol Metab 1975; 40:37.

CHAPTER 29 MORPHOLOGY OF THE THYROID GLAND Principles and Practice of Endocrinology and Metabolism

CHAPTER 29 MORPHOLOGY OF THE THYROID GLAND

VIRGINIA A. LIVOLSI Embryology of the Thyroid Clinical Implications of Thyroid Embryology Nondescent and Maldescent Lateral Aberrant Thyroid Digeorge Syndrome Thyroid Inclusions Gross Anatomy of the Thyroid Electron Microscopy of the Thyroid Immunohistochemistry of the Thyroid Fine-Needle Aspiration Chapter References

EMBRYOLOGY OF THE THYROID

To understand and evaluate the morphologic and clinical aspects of thyroid disease, knowing the embryology, anatomy, cytology, and immunohistochemistry of the thyroid gland is essential. The embryologic development of the human thyroid gland can be understood by distinguishing between the origin of the medial portions of the gland and the lateral lobes (see Chap. 47). The medial portion of the thyroid arises as a bilobed vesicular structure at the foramen cecum of the tongue; it is attached to the tongue by the thyroglossal duct. Beginning in approximately the fifth week of fetal life, as the heart and great vessels move caudally, the thyroid gland descends until it reaches its adult location in the lower anterior neck. This descent of thyroid tissue is accompanied by the elongation of the thyroglossal duct, which normally atrophies. Remnants of thyroid tissue may persist anywhere along this pathway of descent. If the thyroglossal duct does not atrophy, it may become cystic, dilated, or even cancerous.1 The lateral portions of the thyroid are derived embryologi-cally from a portion of the ultimobranchial body, a component of the fifth branchial pouch. From this embryologic structure the parafollicular or C cells are derived. The ultimobranchial bodies fuse with the medial anlage of the thyroid in its upper lateral aspects; hence, in humans, this is the area where the C cells are found (see Chap. 53). In early fetal life the histology of the thyroid is characterized by solid cords or clusters of cells with scant, rough, endoplasmic reticulum. At the 50-mm stage, intercellular clefts form and, eventually, follicles are produced (Fig. 29-1). Rough endoplasmic reticulum and Golgi apparatuses are extensive, lysosomes become abundant, and thyroglobulin synthesis begins.1a,2

FIGURE 29-1. Immunoperoxidase stain with anticalcitonin antisera has been used to stain this C-cell cluster (center). (Immunoperoxidase-diami-nobenzidine [DAB] with hematoxylin nuclear counterstain; 200)

CLINICAL IMPLICATIONS OF THYROID EMBRYOLOGY

NONDESCENT AND MALDESCENT Abnormal migration of the median thyroid anlage may take the form of either nondescent or maldescent. Nondescent of the thyroid produces the clinical syndrome of lingual thyroid, in which the entire tissue is located at the base of the tongue3,4,5 and 6 (see Chap. 47). Maldescent may cause thyroid tissue to remain along the pathway of the thyroglossal duct or in the trachea, larynx, or posterior pharynx.4,5 Elongated descent into the superior mediastinum may lead to substernal, preaortic, or pericardial thyroid.3,7 When the thyroid is diffusely involved by a disease process such as nodular goiter or thyroiditis, the remnants of thyroid along the medial aspects of the gland's descent also may be involved, may enlarge, or may produce nodules or mass lesions that are separate from the thyroid (i.e., in skeletal muscle). Such lesions may be confused clinically and surgically with neoplastic disease. LATERAL ABERRANT THYROID The lateral aberrant thyroid, which has been the subject of controversy for many years, has been defined in several ways. The most acceptable definition is that it is thyroid tissue that is located lateral to the jugular vein. Thyroid tissue that is unassociated with a lymph node and located lateral to the jugular vein may be seen as a component of anomalous development (especially when the thyroid is involved by thyroiditis or nodular goiter), after neck surgery, or after trauma to the neck, in which case dislodged portions of the thyroid may implant in the lateral aspects of the neck. Unfortunately, the term lateral aberrant thyroid has also been used to describe thyroid tissue in lymph nodes lateral to the jugular vein. It is widely recognized that normal-appearing thyroid follicles within lymph nodes lateral to the jugular vein represent metastatic thyroid carcinoma, not a developmental anomaly. The possibility that normal thyroid follicles may be present as an embryologic rest within capsules of medially located lymph nodes remains disputed, although many pathologists agree with this concept. Others, however, insist that any thyroid follicles present within a lymph node, regardless of their location, represent meta-static thyroid cancer to that node.7,8 DIGEORGE SYNDROME Abnormalities in the development of the fifth branchial pouch are associated with partial or total absence of C cells. Frequently accompanying this are abnormal or absent parathyroid glands and thymus, known as DiGeorge syndrome9 (see Chap. 60). THYROID INCLUSIONS The thyroid itself may contain inclusions, especially in its lateral aspects, from numerous branchial or pharyngeal pouches. Thus, intrathyroidal parathyroids, salivary gland remnants, or thymic tissue remnants may be identified. Solid cell nests, probably representing rests of ultimobranchial body, may also be seen.10,11 Solid cell nests are found in the posterolateral or pos-teromedial portion of the lateral lobes of the thyroid. They can be found in up to 21% of adult glands and in between 33% and 89% of fetal and neonatal glands.10,13 These nests are composed of epidermoid cells with palisading at the edges of the nests.10 Mucin may be found in some of these cells. Some of these nests contain groups of C cells. In chronic lymphocytic thyroiditis, especially the fibrous variant, these solid cell nests become prominent.14 In two reported cases, C-cell hyperplasia and elevated serum calcitonin levels were found in lymphocytic thyroiditis.15,16 In addition, the intimate association of the thyroid with the developing mesodermal structures of the neck may lead to the finding of normal thyroid tissue within skeletal muscle or within fat. This is not neoplastic

growth but merely a normal variant of thyroid anatomy.

GROSS ANATOMY OF THE THYROID

The normal adult thyroid gland consists of two lobes connected by an isthmus. Thyroid tissue is light brown, and the cut surface often glistens. The normal gland is surrounded by a delicate fibrous capsule and weighs 15 to 25 g. The normal thyroid is attached loosely to neighboring structures, and the fascial planes are distinct. Nodules, which are common in the adult gland, can be seen grossly in ~10% of the population in the United States.17 The vascular supply to the thyroid is derived from the superior and inferior thyroidal arteries; venous blood drains through thyroidal veins into the external jugular vein. The thyroid contains a rich lymphatic network; intraglandular and subcapsular lymphatics drain into the internal jugular lymph nodes.18 On microscopic examination, the thyroid is divided into lobules composed of 20 to 40 follicles (Fig. 29-2). Each lobule is supplied by an intralobular artery and vein. The follicles are lined by epithelial cells that surround central deposits of colloid (Fig. 29-3). In general, the follicles are uniform in size (ranging from 50 to 500 m); however, the plane of section may give the appearance of small follicles interspersed with large ones. Studies using serial sections disclose that small follicles, or clusters of epithelial cells that apparently lie between follicles, represent the edges of follicles whose maximum diameter is situated at another level.19 Estimates are that, in human adults, ~3 million follicles are present in the thyroid. Follicular cells have a definite polarity; their apices are directed toward the lumen of the follicle, whereas their bases are positioned toward the basement membrane. The colloid found within the follicular lumen also can be seen on the cell surfaces. The periodic acidSchiff stain (which stains glycogen) reveals that the cytoplasm of follicular cells may contain numerous small to large (0.5 to 2.0 m) inclusions, representing colloid (glycoprotein) droplets.

FIGURE 29-2. A portion of a thyroid lobule. Note the relatively uniform size of the individual follicles. Smaller units represent edges of follicles whose maximum diameter lies at another level of sectioning. (Hematoxy-lin and eosin; 100)

FIGURE 29-3. Follicular epithelium is cuboidal in the normal state. Note the interfollicular capillaries (arrows). (The circles in the colloid are artifacts of fixation.) (Hematoxylin and eosin; 250)

ELECTRON MICROSCOPY OF THE THYROID

Ultrastructural studies of normal thyroid disclose that the follicular cells are arranged in a single layer. The apical surface contains microvilli that extend into a central lumen (Fig. 29-4); one to four cilia project from the central portion of each follicular cell. The cilia may alter the physical properties of the colloid, allowing for its ingestion by the apical portion of the follicular cell.20,21 and 22 Within the cytoplasm, lysosomes are prominent and endoplasmic reticulum and small mitochondria are seen. Well-developed desmosomes and terminal bars are found between cells. The basal surface is separated from the interfollicular space by a basement membrane that is 35 to 40 nm thick. In the interstitium, fenestrated capillaries and collagen fibers are noted.

FIGURE 29-4. A normal thyroid follicle at the ultrastructural level. Note the cytoplasm, which is rich in organelles. Microvilli are visible at the cellcolloid interface (arrow) and in the basement membrane (arrowheads). 10,600

Scanning electron microscopic studies indicate that, within each follicle, some individual cells or groups of cells appear to be different from neighboring ones. Thus, the apical morphology and density of microvilli differ; these differences may reflect different functional states. Experiments with cultured dog thyroid indicate that morphologic changes result from thyroid-stimulating hormone stimulation. Scanning electron microscopy has demonstrated that 15 minutes to 4 hours after administration of thyroid-stimulating hormone, thickening of the cell border occurs with subsequent development of cytoplasmic projections and an increase in the number of microvilli.23 In the interfollicular stroma, and occasionally abutting between the follicular epithelial cells, individual or scattered small groups of C cells are present (Fig. 29-5). Usually, on sections stained with hematoxylin and eosin, identification of C cells is difficult; calcitonin immunostaining may be necessary to disclose their location (Fig. 29-6). C cells are large, polygonal elements with centrally placed nuclei and pale-staining granular cytoplasm. In children younger than 6 years of age, the C cells are abundant, whereas in adults, they are scattered individually or appear in groups of three to five cells.24 The normal adult complement of C cells remains debated.24,25 The currently proposed estimate of C-cell number in adult human thyroid is no greater than 40 C cells/mm2.24,25

FIGURE 29-5. A cluster of C cells in the interfollicular stroma. Note the large size and abundant cytoplasm of these cells as compared with normal follicular elements. (Hematoxylin and eosin; 250)

FIGURE 29-6. Immunoperoxidase stain with anticalcitonin antisera has been used to stain this C-cell cluster (center). (Immunoperoxidase-diami-nobenzidine [DAB] with hematoxylin nuclear counterstain; 200)

On electron microscopy, the cytoplasm of C cells is occupied by numerous membrane-bound (neurosecretory) granules. On immunoelectron microscopic examination, these granules have been shown to contain calcitonin.26

IMMUNOHISTOCHEMISTRY OF THE THYROID

Immunoperoxidase techniques have improved our understanding of the substances, especially hormones, that are contained in normal and abnormal thyroid cells.26a Antisera to thyroglobulin demonstrate this protein both within colloid and in follicular epithelium. Immunoelectron microscopic studies indicate that thyroglobulin is present in the rough endoplasmic reticulum, the Golgi apparatus, and the apical cell border.27 The expression and distribution of intermediate filament proteins vary in normal and diseased thyroids.28 Thus, normal follicular epithelium contains cytokeratin and, rarely, epidermal prekeratin. By contrast, cells in papillary carcinoma contain large amounts of prekeratin, as demonstrated by staining with antiprekeratin antibodies.21 C cells that contain neurosecretory granules may be stained by nonspecific silver techniques (Grimelius, Sevier-Munger).29 Specific immune staining is appropriate for calcitonin (see Fig. 29-6).

FINE-NEEDLE ASPIRATION
Fine-needle aspiration cytology of the thyroid has proven to be a safe and effective screening and diagnostic method for the evaluation of thyroid nodules (see Chap. 39). To understand changes seen in pathologic conditions, one must be aware of the range of normal results with such needle aspirations. In aspirates from normal glands, uniform epithelial cells, often forming follicles surrounding drops of colloid, may be seen (Fig. 29-7). When seen en face, a honeycomb pattern is observed. The follicular cells are round to oval, measuring 8 to 10 m in diameter; a central nucleus and pale cytoplasm are present. The nuclear chromatin is uniform and finely granular.30 Variations occur in normal glands as a part of the aging process. Nuclear size is increased significantly in thy-rocytes from normal glands of subjects 60 years of age or older in comparison with the size of those in individuals younger than 30 years of age.31 These variations are important to recognize to avoid the diagnosis of atypical follicular epithelium from fine-needle aspiration samples solely on the basis of size.

FIGURE 29-7. This specimen was obtained by needle aspiration from a normal thyroid lobe after surgery for laryngeal cancer. Note the rounded clusters of uniform follicular cells. The wispy material visible in the smear background and lower right is colloid. (Hematoxylin and eosin; 400)

CHAPTER REFERENCES
1. Yang YJ, Haghir S, Wanamaker JR, Powers CN. Diagnosis of papillary carcinoma in a thyroglossal duct cyst by fine needle aspiration biopsy. Arch Pathol Lab Med 2000; 124:139. 1a. Chan AS. Ultrastructural observations on the formation of follicles in the human fetal thyroid. Cell Tissue Res 1983; 233:693. Garcia-Bunuel R, Anton B, Brandes D. The development of lysosomes in the fetal thyroid in correlation with the onset of functional maturation. Endocrinology 1971; 91:438. Larochelle D, Arcand P, Belzile M, Gagnon NB. Ectopic thyroid tissuea review of the literature. J Otolaryngol 1979; 8:523. Nienas FW, Gorman CA, Devine KD, Woolner LB. Lingual thyroid. Clinical characteristics of 15 cases. Ann Intern Med 1973; 79:205. Wong RJ, Cunningham MJ, Curtin HD. Cervical ectopic thyroid. Am J Oto-laryngol 1998; 19:397. Prasad KC, Bhat V. Surgical management of lingual thyroid: a report of four cases. J Oral Maxillofac Surg 2000; 58:223. LiVolsi VA, Perzin K, Savetsky L. Carcinoma arising in median ectopic thyroid (including thyroglossal duct tissue). Cancer 1974; 34:1303. Rabinov CR, Ward PH, Pusheck T. Evolution and evaluation of lateral cystic neck masses containing thyroid tissue: lateral aberrant thyroid revisited. Am J Otolaryngol 1996; 17:12. Robinson HB. DiGeorge's or the IIIIV pharyngeal pouch syndrome: pathology and a theory of pathogenesis. In: Rosenberg HS, Bolande RP, eds. Perspectives in pediatric pathology, vol 2. Chicago: Year Book Medical Publishers, 1975:173. 10. Harach HR. Solid cell nests of the thyroid. Acta Anat (Basel) 1985; 122:249. 2. 3. 4. 5. 6. 7. 8. 9.

11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26.

Autelitano F, Santeusanio G, Tondo UD, et al. Immunohistochemical study of solid cell nests of the thyroid gland found from an autopsy study. Cancer 1987; 59:477. Bechner ME, Shulz MS, Richardson T. Solid and cystic ultimobranchial body remnants in the thyroid. Arch Pathol Lab Med 1990; 114:1049. Harach HR, Day ES, Franssila KO. Thyroid spindle cell tumor with mucous cysts: an intrathyroid thymoma? Am J Surg Pathol 1985; 8:525. Janzer RC, Weber E, Hedinger C. The relation between solid cell nests and C-cells of the thyroid gland. Cell Tissue Res 1979; 197:295. Biddinger PW, Brennan MR, Rosen PP. Symptomatic C cell hyperplasia associated with chronic lymphocytic thyroiditis. Am J Surg Pathol 1991; 15:599. Libbey NP, Nowakowski KJ, Tucci JR. C-cell hyperplasia of the thyroid in a patient with goitrous hypothyroidism and Hashimoto's thyroiditis. Am J Surg Pathol 1989; 13:71. DeHaven JW, Sherwin RS. The thyroid nodule: approach to diagnosis and therapy. Conn Med 1979; 43:761. Romanes GJ. Cunningham's textbook of anatomy. London: Oxford University Press, 1964:537. Rienhoff WF. Gross and microscopic structure of thyroid gland in man. Contrib Embryol 1930; 21:99. Sobrinho-Simoes M, Johannesson JV. Scanning electron microscopy of the normal human thyroid. J Submicrosc Cytol 1981; 13:209. Echeverria OM, Hernandez-Pando R, Vasquez-Nin GH. Ultrastructural, cytochemical and immunohistochemical study of nuclei and cytoskeleton of thyroid papillary carcinoma cells. Ultrastruct Pathol 1998; 22:185. Gould VE, Johannessen JV, Sobrinho-Simoes M. The thyroid gland. In: Johannessen JV, ed. Electron microscopy in human medicine, vol 10. New York: McGraw-Hill, 1981. Rapoport B, Jones AL. Acute effects of thyroid stimulating hormone on cultured thyroid cell morphology. Endocrinology 1978; 102:175. Guyetant S, Wion-Barbot N, Rousselot MC, et al. C-cell hyperplasia associated with chronic lymphocytic thyroiditis. Hum Pathol 1994; 25:514. Guyetant S, Rousselot M, Durison M, et al. Sex-related C-cell hyperplasia in the normal human thyroid: a quantitative autopsy study. J Clin Endo-crinol Metab 1997; 82:42. Wolfe HF, Voelkel EF, Tashjian AJ. Distribution of calcitonin-containing cells in the normal adult human thyroid gland: a correlation of morphology and peptide content. J Clin Endocrinol Metab 1974; 38:688.

26a. DeMicco C, Kopp F, Vassko V, Grino M. In situ hybridization and immuno-histochemistry study of thyroid peroxidase expression in thyroid tumors. Thyroid 2000: 10:109. 27. Ide M. Immunoelectron microscopic localization of thyroglobulin in the human thyroid gland. Acta Pathol Jpn 1984; 34:575. 28. Miettinen M, Franssila K, Lehto V-P, et al. Expression of intermediate filament proteins in thyroid gland and thyroid tumors. Lab Invest 1984; 50:262. 29. Wilander E, Justti-Berggren L, Lundqvist M, Grimelius L. Staining of rat thyroid parafollicular (C) cells with the Sevier-Munger silver technique. Acta Pathol Microbiol Immunol Scand [A] 1980; 88:339. 30. Kini SR. Guidelines to clinical aspiration cytology: thyroid. New York: Igaku-Shoin, 1987. 31. Ferchter GE, Goerthler K. Age related nuclear size variability of thyrocytes in thyroid aspirates. Anal Quant Cytol 1983; 5:75.

CHAPTER 30 THYROID PHYSIOLOGY: SYNTHESIS AND RELEASE, IODINE METABOLISM, BINDING AND TRANSPORT Principles and Practice of Endocrinology and Metabolism

CHAPTER 30 THYROID PHYSIOLOGY: SYNTHESIS AND RELEASE, IODINE METABOLISM, BINDING AND TRANSPORT
H. LESTER REED Iodine Metabolism Iodide Clearance Iodide Excess Transport of Iodide Thyroid Hormone Synthesis Organification Thyroglobulin Coupling Hormone Storage and Secretion Thyroid Hormone Circulation and Transport Binding Proteins Peripheral Metabolism of Thyroid Hormones Regulation of Thyroid Hormone Economy Neuroendocrine Modulation Hypothalamus Pituitary Gland Peripheral Thyroid Metabolism Thyroid Gland Chapter References

The thyroid hormones, L -3,5,3',5'-tetraiodothyronine (L -thyroxine [T4]) and, to a much less extent, L-3,5,3'-triiodothyronine ( L -triiodothyronine [T3]) are synthesized by the follicular epithelial cells of the thyroid gland. This synthesis requires the availability of iodine and is increased by thyroid-stimulating hormone (thyrotropin; TSH) from the anterior pituitary gland through a specific thyroidal receptor. Small amounts of thyroid hormone are secreted continuously into the blood and are almost entirely bound (in a large circulating reservoir) to plasma proteins (Table 30-1), with a very small percentage remaining unbound or free. The speculation is that the free hormones (free T4 and T3), principally derived from circulating T4, enter the cell and mediate their effects through specific nuclear receptors (see Chap. 31), which are heterogeneously distributed among tissues. The hypo-thalamicpituitarythyroid axis is regulated by autocrine, paracrine, hemocrine, and environmental factors to maintain steady-state hormone economy.

TABLE 30-1. Thyroid HormoneBinding Proteins

IODINE METABOLISM
Iodine is essential for the synthesis of thyroid hormones. With four iodines per molecule of T4, iodine comprises 66% of T4 by weight; with three iodines, T3 is 58% iodine (Fig. 30-1). Normally, ~90 g (~120 nmol) of T4 and 6.5 g (~10 nmol) of T3 are secreted daily by the thyroid gland. Thus, 60 to 80 g (~550 nmol) of iodine must be transported into the gland daily to maintain normal daily hormone production.1 Iodine is not always present in sufficient quantities from dietary sources in the environment. Surprisingly, even in the United States, the daily iodine intake has been declining.2 Accumulation of the absolute iodine requirement for thyroid hormone synthesis is facilitated by an efficient system for concentrating and conserving iodine in the thyroid gland. Between 5000 and 10,000 g of hormonal iodine is stored within the gland.1 This pool constitutes a protective reserve against periods of dietary iodine deficiency.

FIGURE 30-1. Chemical structures of L -thyroxine (T4), L -triiodothyronine (T3), and reverse T3 (rT3). (From Hershman JM. Endocrine pathophysiology. Philadelphia: Lea & Febiger, 1980.)

IODIDE CLEARANCE Within a certain range, the thyroid gland can adjust to variation in dietary iodine with changes in its clearance of iodide from plasma. However, either chronic dietary deficiency of iodine or conditions of severe iodine excess often exceed the capacity for regulation, resulting in disease states (see Chap. 37 and Chap. 38). Virtually all of dietary iodine is reduced to iodide and absorbed in the small intestine. Circulating iodide is cleared from the blood principally by the kidney (80%) and by the thyroid (20%).1 Renal excretion, measured with a 24-hour collection, varies with filtered load and reflects 97% of dietary intake; only ~3% is lost in the stool.1 Renal iodide is passively reabsorbed; thus, the renal clearance depends on glomerular filtration rate and is apparently unaffected by serum iodide concentration. Patients with end-stage renal disease (ESRD) have decreased renal iodide clearance and elevated serum iodide concentrations.3 Iodide excess is a proposed mechanism for the increased incidence of hypothyroidism, thyroid nodules, and goiter, whereas the mechanisms for increased thyroid cancer with ESRD are unknown.3 In contrast,

thyroidal iodide clearance changes inversely with dietary intake and intrathyroidal iodine stores, increasing as much as fivefold under conditions of iodine deficiency.1 In addition to the thyroid gland, the salivary glands, gastric mucosa, choroid plexus, and mammary glands may concentrate iodine, which can be detected with radioiodine scintigraphy1; however, only in the thyroid is this function influenced by TSH. Therefore, although radiation-induced salivary gland inflammation and breast milk contamination are possible complications of treatment with radioactive iodine, they should not be worsened by recombinant human TSH administration.4 IODIDE EXCESS The amount of iodine in the American diet, with its iodine-enriched foods, usually greatly exceeds the recommended 150 g/day of iodine in most areas. Nonetheless, a decline in dietary iodine has been noted over the last 20 years, with ~10% of the U.S. population being at risk for iodine deficiency.2 The American diet contrasts with diets in many other areas of the world, such as central Africa,2,3 where the percentage of people at risk for deficiency is much higher.5,6 Additional dietary iodine from many common agents such as cough medicines, vitamins, kelp, iodinated contrast agents (iopanoic acid, sodium ipodate), antiarrhythmics (amiodarone), antiseptics (iodoform gauze), and water purification tablets (tetraglycine hydroperiodide) is often unexpected (see Chap. 37). In normal individuals, as well as patients with ESRD, an increased iodine load can decrease radioiodine uptake and increase serum TSH and thyroidal size (see Chap. 37).3 High doses of iodine dilute diagnostic radioiodine tracers and thereby reduce fractional thyroidal uptake of the tracer; this causes low radioiodine uptake and diminished ability to image the thyroid by scintis-can. Further, depending on the amount, excess iodine inhibits uptake, organification, and release of the hormones (Fig. 30-2) and decreases thyroidal blood flow.7 An intrinsically abnormal gland is less likely to adjust successfully to these changes in dietary or pharmaceutical iodine and, thus, may present with clinical disease1 (see Chap. 37 and Chap. 38).

FIGURE 30-2. Thyroid hormone synthesis. Iodide (I) is transported by the iodide transporter (IT) from the plasma into the thyroid cell and then I moves to the apical membrane, where it is organified and coupled under the influence of thyroid peroxidase (TPO) to thyroglobulin (Tg), which is synthesized within the cell. Hormone stored as colloid reenters the cell through endocytosis and moves back toward the basal membrane, where T4 and T3 are secreted. Nonhormonal iodide is recycled. Hormone synthesis may be inhibited at: (1) iodine transport by thiocyanate, perchlorate, interleukin-1, and tumor necrosis factor-a; (2) organification and (3) coupling by propylthiouracil (PTU) and methimazole and through decreasing TPO by interferon-g and interleukin-1; (4) endocytosis by lithium and iodide; (5) thyroglobulin proteolysis by iodide; and (6) intrathyroidal deiodination by PTU and tumor necrosis factor-a.

TRANSPORT OF IODIDE Thyroid cells actively transport iodide and other monovalent anions such as perchlorate and pertechnetate (see Chap. 33) from the plasma into their cytoplasm. This adenosine triphosphatase (ATPase)dependent sodium-iodide cotransporter or symporter has been isolated and cloned (see Chap. 34).8 Although the transport system, often called trapping, can be saturated, its capacity is above that of typical plasma iodide levels. Thus, small increases in available iodide are followed by increased transport. This trapping by the symporter is rate-limiting to thyroid hormone synthesis and is regulated by the TSH receptor. It can move iodide against a gradient and increase thyroidal iodide 30-fold over blood concentrations. Iodine is stored in the thyroid gland in thyroglobulin (Tg), a large glycoprotein essential to the synthesis and storage of the thyroid hormones (see Thyroglobulin section later in this chapter).

THYROID HORMONE SYNTHESIS

The major synthetic steps regulated by TSH are transport of iodide, organification, coupling, Tg synthesis, and endocytotic secretion. Under the stimulation of TSH, the thyroid cells remove iodide from the capillary network at the base of the cell (see Fig. 30-2) and move the iodide to the apex of the cell, where it is joined with molecules of tyrosine to make mostly T4 and some T3. T4 and T3 are stored in colloid within the follicles of the gland. They are then released together, or some T4 is further deiodinated to T3 before release (see Fig. 30-2). These two steps are also under the influence of TSH or other proteins that bind to the TSH receptor. Mutations of this TSH receptor are continuously activated in some toxic adenomas and multinodular goiters.9 A small but measurable amount of nonTSH-dependent T4 secretion also occurs normally. ORGANIFICATION Iodide ions transported into the thyroid cell are oxidized before being used for iodinating tyrosyl residues present in the Tg molecules. This process, called organification, takes place at the apex of the cell (see Fig. 30-2) and requires the enzyme thyroid peroxidase (TPO). The decrease in peroxidase activity and organification that follows iodine excess, known as the Wolff-Chaikoff effect, may protect against hyperthyroidism. The effect is transient, and escape occurs within several days. Failure to escape from the Wolff-Chaikoff effect is common in certain diseased glands; therefore, the reduction in organification due to iodine excess is perpetuated, leading to hypothyroidism1 (see Chap. 37). The activity of TPO may be deficient on a congenital basis and is reduced by antimicrosomal antibodies directed against it (see Chap. 46 and Chap. 197) or with administration of thiourea drugs, such as propylthiouracil (PTU), which is commonly used to treat hyperthyroidism (see Chap. 42). THYROGLOBULIN The tyrosine residues to which the iodine is joined in the organification process are part of the Tg molecule, a 660-kDa glycoprotein. Defects in Tg synthesis have been proposed to cause some cases of goiter, although such cases are rare. Thus far, therapeutic interference with the synthesis of Tg has not been possible. Serum values of Tg are extremely useful in the management of well-differentiated thyroid cancer as an indicator of biologic activity or recurrence.10 Iodination of the tyrosyl residues within Tg (i.e., organification of the iodine) involves substitution on the tyrosyl ring. If one iodine replaces a hydrogen, then monoiodotyrosine (MIT) is formed; if two iodines are joined to tyrosine, diiodotyrosine (DIT) is formed. The MIT/DIT ratio depends on iodine availability; during iodine deficiency, for example, relatively more MIT is formed. COUPLING The iodothyronines are formed by coupling of two DITs to form T4 and one MIT and one DIT to form T 3. The coupling process is catalyzed by TPO, blocked by thiourea drugs and excess iodine, and stimulated by TSH (see Fig. 30-2). Congenital defects in MIT and DIT coupling result in decreased hormone synthesis and goiter. Because the MIT/DIT ratio varies inversely with iodine availability, relatively more T3 is formed with iodine deficiency and relatively more T4 with iodine surfeit. HORMONE STORAGE AND SECRETION Thyroglobulin is contained within cytoplasmic bodies called vesicles that fuse with the apical membrane of the cell and extrude their Tg content into the colloid space at the center of the thyroid follicle (see Fig. 30-2). This space is bounded by the apical membranes of adjacent follicle cells and is inaccessible to the circulation. Here, hormone is stored in colloid. Ordinarily, only a very small fraction of stored hormone, ~1%, turns over each day. Under TSH stimulation, hormones are secreted, and

less colloid is stored. The process of hormone secretion moves in the opposite direction back to the basal membrane of the cell. Each step is controlled by TSH. First, pseudopods of the apical cell membrane enclose colloid and form a droplet that merges with the apical membrane and moves into the cell. This process of endocytosis occurs within minutes of TSH stimulation. In the interior of the cell, the colloid droplet is joined by a lysosome, forming a phagolysosome. As the phagolysosome moves toward the base of the cell, the Tg molecule is digested. Amino acids and nonhormone iodine are recovered in the cell for recycling. T4 and T3 are secreted into the capillaries at the base of the cell in proportion to their presence in Tg. Most T3 is derived from extrathyroidal conversion of T4 by 5'monodeiodinase type I (5'D-I) in peripheral tissues. However, a 27-kDa protein with 5'D-I activity has been found within human thyroid glands.1,6 It is regulated by TSH, has higher activity in Graves disease and follicular adenomas, and has lower activity in papillary carcinoma.

THYROID HORMONE CIRCULATION AND TRANSPORT

Approximately 120 nmol of T4, 47 nmol of T3, and 50 nmol of reverse T3 (rT3) appear in the serum each day in a typical 70-kg person11,12 and 13 (Table 30-2). The balance between production and degradation is mediated by, among other things, nutrition, non-thyroidal illness, exercise, pregnancy, and medications. At steady-state conditions, this balance is reflected by serum values that represent only ~29% of T4, 14% of T3, and 14% of rT3 total body hormone stores or pool (Qtot). Less than 1% of the total circulating amount of each hormone is free in the plasma (see Table 30-2). This extraordinary degree of binding to transport proteins has several important consequences: (a) The plasma contains a large capacity to store hormone, serving as a reservoir or a buffer against fluctuations in blood levels; (b) Little hormone is lost through the kidneys by glomerular filtration of free hormone; and (c) Changes in binding proteins affect the serum reservoir but do not affect the amount of free hormone, presumably not altering the physiologic actions of the hormone (see Table 30-1). Either the binding protein capacity, the free hormone concentration, or the binding protein concentration may be measured to clarify the changes in circulating T4 or T3 concentrations that accompany genetic or acquired aberrations in the binding proteins (also see Chap. 33 and ref. 13a).

TABLE 30-2. Kinetic Characteristics of Thyroid Hormones

Table 30-2 shows some of the similarities and differences between the major iodothyronines. T3 has 10% and rT3 has only 2% of Qtot compared with T4. Although ~50% of T4 and rT3 circulates in the blood and other rapidly equilibrating tissues, such as liver and kidney, only ~25% of T3 is found in these same kinetic compartments. Hormone stores are turned over at a rate of ~10% per day for T4, 70% per day for T3, and 350% per day for rT3. Thus, with inhibited production of endogenous hormone (see Table 30-2) or increased appearance of exogenous hormone (e.g., increased dosage), the new steady-state conditions are reached within weeks for T4, within days for T3, and within hours for rT3.11,12 Compared with T3, T4 has a higher binding affinity (see Table 30-1) and slower plasma clearance rate (relative to the distribution volume [VD]), making T4 a stable circulating hormone reservoir and substrate for T3 formation. The rapid plasma clearance rate of T3, relative to its VD, is associated with more labile equilibration or disequilibration, reflecting shorter response times of circulating T3 to environmental and nutritional influences. BINDING PROTEINS T4 and T3 are reversibly bound to several proteins synthesized by the liver (see Table 30-1). The variety of relative serum concentrations, hormone affinity constants (Ka), and dissociation constants (Kd) help provide a buffered, stable plasma pool and a homogeneous distribution of free hormone during capillary transport.14 Transthyretin (TTR), previously known as thyroid-binding prealbumin (TBPA), has a low affinity and rapid Kd; therefore, it plays a greater role in delivering iodothyronines to tissues than do circulating proteins with higher binding affinity.14,15 Specifically, TTR is the major T4-binding protein in the cerebrospinal fluid and is thought to be a major mechanism for homogenous central nervous system hormone distribution.1 Some evidence dates this mechanism to more than 300 million years.15 Thyroxine-binding globulin (TBG), however, with its relatively high binding affinity and slower dissociation, provides a stable hormone reservoir in the plasma. Changes in the glycoprotein nature of TBG slow its removal in certain physiologic conditions such as pregnancy. Albumin, with its relatively high serum concentration, has multiple binding sites with a low binding affinity and, thus, like TTR, may contribute to tissue delivery of the free hormone. Lipoproteins bind small amounts of thyroid hormones, but their physiologic significance is unknown.14 Increased binding to albumin from a hereditary cause is associated with increased production of both T4 and T 3, which suggests increased tissue delivery by this protein deiodination.16 Under normal conditions, only ~25% of TBG and 0.3% of TTR molecules carry T4. The hormone distribution among the carrying proteins is listed in Table 30-1. Over a wide range of binding protein concentrations, steady-state free hormone concentration, and production rates remain nearly constant.14,16 Thus, the measurement of free T4 or free T3 concentrations by equilibrium dialysis or a comparable method is critical in clinical circumstances when changes in the transport proteins are present (see Table 30-1).14 Genetic or induced alterations of TBG are important because this protein carries 75% of T4 and 80% of T3, and because these bound iodothyronines have the greatest impact on measured values of circulating total thyroid hormone by routine analysis. Clearance of TBG is decreased by an increased sialic acid content of the terminal sugar of its polysaccharide chain. Sialic acid content is increased in pregnancy; the resultant decreased clearance of TBG, along with increased production of the protein, contributes to elevated serum TBG. Inherited conditions may affect TBG, TTR, and albumin. Familial increases and decreases of TBG, increased TTR binding of T4, decreased TBG affinity for T4, and abnormal T4 binding to albumin have all been described. Generally, these variations are associated with an asymptomatic euthyroid condition.14,16 Non-thyroidal illness may be associated with circulating inhibitors to hormone binding to these carrier proteins (see Chap. 36).

PERIPHERAL METABOLISM OF THYROID HORMONES

T4 is metabolized by the sequential removal of iodine atoms. Although most human tissues have the ability to perform this enzymatic step, liver and kidney are the most important sites of deiodination. The stepwise removal of iodine atoms from thyroxine has been well described.11 The process is enzymatic and depends on available sulfhydryl compounds. First, a hydrogen atom replaces iodine at one of four positions on T4. Subsequently, further iodine atoms are removed, yielding thyronines with two, one, and, finally, no iodine atoms. Iodine may be removed from the inner (tyrosyl) or outer (phenolic) ring. The entire process is referred to as sequential deiodination and is mediated by a family of enzymes listed in Table 30-3.

TABLE 30-3. Thyroid Hormone Deiodinase Characteristics

Removing either iodine atom from the outer ring (designated as the 3' or 5' position) of T4 yields T3. By convention, this pathway is referred to as 5'-deiodination. Two distinct enzymes are known to catalyze this process, monodeiodinase type I (5'D-I) and type II (5'D-II), both of which are described in Table 30-3. Because T3 has more metabolic activity than T4, this pathway represents activation. Intrapituitary conversion of T4 to T3 is important in the regulation of TSH secretion. Removing either iodine atom from the inner ring (the 3 or 5 position) of T4 yields rT3. Reverse T3 has no known metabolic activity. It does not prevent goiter but is a competitive inhibitor of 5'D-I. Therefore, the 5-deiodination pathway catalyzed both by 5'D-I, at a different pH optima from that for phenolic ring deiodination, and by a type III seleno-enzyme (see Table 30-3) that is found in placenta, brain, and skin, constitutes an inactivation. Substrate conjugation with sulfate can increase the tyrosyl ring deiodination of T4 and T3 by 5'D-I several hundred-fold, thereby inactivating these hormones.17 T3 sulfate is produced in peripheral tissues and does not suppress TSH in normal humans. It may account for facilitated 5'D-I or augmented biliary losses in disease states.18 The sequential deiodination continues to yield thyronine and a family of diiodothyronines and monoiodothyronines that have uncertain bioactivity (see Fig. 30-1). Peripheral T4 deiodination contributes 80% to 85% of the T3 plasma appearance rate (PAR). This conversion occurs mostly in the liver and kidney. The thyroid gland may secrete ~15% of the T3 total PAR directly in the form of T3. The role of thyroidal 5'D-I as a contributing mechanism of peripheral T4 deiodination is presently under investigation. Nearly 100% of rT3 comes from peripheral T4 deiodination, although not necessarily from hepatic or renal sources. 5'D-I is a selenium-dependent enzyme found in human thyroid, liver, and kidney; it has decreased activity in hypothyroidism, undernutrition, nonthyroidal illness, and selenium deficiency, and after PTU administration.6,12,19 When phenolic ring deiodination (5'D-I) activity is decreased, both the production of T3 and the degradation of rT3 decline.12 These changes result in an increase in plasma rT3 and a decrease in T3 concentration as found with fasting and some nonthyroidal illnesses.13,19 Many other situations may also have a similar profile consisting of a low serum level of T3 and high rT3, including drug administration (e.g., glucocorticoids, b-blockers, thionamides, oral cholecystographic dyes), chronic illness (cirrhosis, renal failure, malignancy), and acute illness (e.g., myocardial infarction, sepsis, uncontrolled diabetes mel-litus, severe burns) (see Chap. 36). However, in human immunodeficiency virus (HIV) infection, the rT3 is paradoxically decreased and T3 is normal or increased, suggesting another mechanism during nonthyroidal illness (NTI) to that previously described (see Chap. 36). In central Africa, myxedematous cretinism associated with iodine deficiency predominates, with little incidence of neurologic cretinism.6 Some of these regions have coexistent iodine and selenium deficiency. The resultant hypothyroidism that occurs in these locations is associated with a decreased selenium-dependent activity of 5'D-I that may attenuate the frequency of neurologic cretinism.6 Fetal brain T4 acts as the major source of brain T3 converted by 5'D-II that is a seleno-protein but is resistant to dietary selenium deficiency20 (see Table 30-3). Placental type III deiodinase (see Table 30-3) shares a similar resistance to dietary selenium deficiency that may reflect tissue specific preservation of selenium stores.21 The presence of more circulating fetal T4 that has not been converted to T3 by 5'D-I, an extremely selenium sensitive enzyme, results in increased serum T4 available for uptake by the fetal brain during early pregnancy. Thus, the selenium-mediated decrease in T4 peripheral deiodination may help reduce the incidence of neurologic cretinism. Iodine should be replaced first in this particular circumstance to avoid premature activation of 5'D-I by selenium replacement,6,20 thus predisposing to a further lowering of serum T4. Other pathways for the disposal of thyroid hormones exist. The ether bond may be broken or the alanine side chain of T4 and T3 may be oxidized. Tetraiodoacetic acid (tetrac) and triiodoacetic acid (TRIAC) are formed by this oxidation. These compounds have some metabolic activity, and their clinical role is being investigated. Sulfoconjugation of iodothyronines facilitates tyrosyl (inner ring) deiodination except possibly of rT3, which is already a very favored substrate for phenolic (outer ring) deiodination. Glucu-ronide conjugates of iodothyronines are more hydrophilic and are consequently excreted more readily in the bile. In humans, changing this enterohepatic circulation with an anion exchange resin (cholestyramine) lowers serum T4 in postabsorptive states, after excessive thyroxine ingestion, and in Graves disease.22 The intrapituitary generation of T3 from T4 by 5'D-II helps maintain regulation of TSH release. In contradistinction to 5'D-I, as the substrate level of T4 declines, the activity of this form of phenolic ring deiodination increases, and fasting and PTU do not inhibit its activity (see Table 30-3). Pharmacologic inhibition of 5'D-I conversion of T4 to T3 with the use of iodinated radiographic contrast agents (e.g., iopanoic acid) and thiourea compounds (e.g., PTU) is an effective means of rapidly decreasing T3 production in hyperthyroid patients (see Table 30-3 and Fig. 30-2) (see Chap. 42).

REGULATION OF THYROID HORMONE ECONOMY

Thyroid hormone regulation is directed through the neuroen-docrinehypothalamicpituitarythyroidperipheral tissue axis (Fig. 30-3). This system has autocrine (e.g., enzyme autoregula-tion), paracrine (e.g., somatostatin, thyrotropin-releasing hormone [TRH]), and endocrine (e.g., T4, T3) autoregulation that is also influenced by environmental factors (e.g., energy balance, circadian variation, illness).

FIGURE 30-3. Major steps in thyroid hormone regulation. Thyrotropin-releasing hormone (TRH) from the hypothalamus stimulates and somatostatin (SS) inhibits thyroid-stimulating hormone (TSH) release under the influence of such neurotransmitters as dopamine (DA), serotonin (SR), and tumor necrosis factor (TNF). The pituitary secretes TSH that leads to regulation of the thyroid gland. Peripheral monodeiodination by monodeiodinase type I (DI-I) or type II (DI-II)6,20 yields triiodothyronine (T3) and other iodothyronines that may enter the bowel22 or may be further deiodinated after sulfur (T3S) or other conjugates (T3/4 conjugates) are formed. Thyroid hormones may be actively transported into cells by a transport protein (TP).19 T 3 initiates action in target tissues through a variety of thyroid-hormone receptors (TRs), which differ between tissues. Circulating thyroxine (T4) principally (heavy arrows) and T3 are transported by binding proteins (BPs)14,15 and have inhibitory action, primarily on the pituitary but also on the hypothalamus. (The images used here were obtained from ISMI's master clip collection, 1895 Francisco Blvd East, San Rafael, CA 94901-5506, USA.)

NEUROENDOCRINE MODULATION Serum TSH secretion has a diurnal rhythm: values peak after midnight (nearly 50% over the 24-hr mean or mesor) and are lowest in midafternoon. The core body temperature rhythm is highly associated with this serum TSH rhythm.23 Advancing age may blunt this nocturnal surge in TSH.24 In rodents and human infants, cold stimulates TSH secretion.23 Hypothyroid patients administered a constant T4 dose have a slight decline in serum T4, unchanged T3, and increased TSH response to TRH in winter months.23 Several factors are well recognized to increase the release of TSH (e.g., estrogens, a-adrenergic agonists, dopamine antagonists, sleep deprivation) or decrease it (e.g., thyroid hormones, glucocorticoids, dopamine, growth hormone, somatostatin, tumor necrosis factor, sleep initiation)23,25,26 (see Fig. 30-3). HYPOTHALAMUS To maintain thyroid hormone production, the tonic stimulation by TRH is required. Severe hypothalamic injury or separation of the pituitary from the hypothalamus by stalk section results in hypothyroidism. TRH is a tripeptide (Glu-His-Pro) synthesized in the paraventricular nucleus of the hypothalamus, and its mRNA concentration is inversely related to concentrations of circulating thyroid hormones. TRH is transported down nerve endings to the median eminence and reaches the anterior pituitary through the portal capillary plexus. Somatostatin, found in the anterior periventricular region, inhibits TSH release. Administration of somatostatin decreases the TSH response to TRH, nocturnal TSH rise, and the TSH elevations seen in primary hypothyroidism (see Chap. 15). Tumor necrosis factor-a (TNF-a) and interleukin-1 (IL-1) decrease TSH, and some of that effect is possibly mediated through TRH. TRH binds to high-affinity receptors on the surface of TSH-producing cells (thyrotropes) and leads to a prompt release of stored TSH. With more prolonged stimulation, new synthesis and release of TSH occur. As a diagnostic procedure, the provocation of TSH secretion by injecting TRH has limited utility because free T4 assays and sensitive TSH assays are readily available27 (see Chap. 33). Direct measurement of serum TRH is not yet clinically feasible. PITUITARY GLAND TSH is a glycoprotein synthesized in the anterior pituitary gland. It has a molecular mass of 28 kDa and is composed of two subunits, designated as the a and b chains, whose bioactiv-ity may depend on the attached carbohydrate moiety. This two-subunit structure is similar to that of follicle-stimulating hormone, luteinizing hormone, and human chorionic gonado-tropin (see Chap. 11, Chap. 15 and Chap. 16); in fact, these regulatory hormones all share a common structure for the a subunit. Excess a subunit is synthesized by the pituitary cells. The rate-limiting step in TSH synthesis is production of the b subunit, which also confers specificity in stimulation of target organs. Only intact TSH is routinely measured in the serum, but measurement of subunits may be useful in syndromes of inappropriate TSH secretion. TSH-producing pituitary tumors may be associated with disproportionately elevated serum a-subunit levels (see Chap. 15 and Chap. 42). Two major factors regulate TSH synthesis and secretion: inhibitory, negative feedback at the anterior pituitary, from circulating thyroid hormones and hypothalamic mediators, such as somatostatin and dopamine; and stimulation by TRH from the hypothalamus (see Fig. 30-3). High circulating levels of T4 or T3 suppress TSH synthesis and release by negative feedback. Either excess endogenous hormone from primary hyperthyroidism or exogenous T4 or T3 from hormone administration should lead to low serum TSH levels; otherwise, a measurable TSH concentration in the face of elevated levels of free T4 and T3 is considered inappropriately elevated and may suggest a TSH-secreting pituitary tumor. Physiologically, small changes in the levels of free T4 or T3 lead to a small inverse-logarithmic change in serum TSH concentration, sufficient to return the free hormone level to its prior level.26 These changes in TSH between 0.1 and 10.0 mU/L have been inversely correlated with free T4 and a 15% change in resting energy expenditure.28 Intrapituitary T3 and, thus, TSH regulation are derived principally by 5'D-II (see Table 30-3) from circulating T4. TSH, therefore, may be seen to rise if serum T 4 is decreased, even though serum T3 is in the normal or slightly elevated range. This situation is found during primary hypothyroidism or iodine deficiency, in which the TSH-secretion rate may increase 20-fold (see Chap. 45). However, with excess T3 administration, TSH secretion and thyroidal iodine uptake can be greatly inhibited; this inhibition is the basis for the T3 suppression test (see Chap. 33). Consequently, most T3 used by the pituitary is generated locally. This is in contrast to most other human tissues, which obtain their supplies of T3 from the circulation. A noted exception is the brown adipose tissue that is found in human infants and in both adult and infant rodents, which has a high activity of 5'D-II. THYROID GLAND Production of T3 and T4 by the thyroid gland is regulated primarily by circulating TSH. A small fraction of thyroidal activity and T4 release, however, has been described as nonTSH dependent. TSH binds to specific membrane receptors on the surface of thyroid cells, activates intracellular adenylate cyclase, and mediates most of its action through increased cyclic adenosine mono-phosphate. This receptor may also be engaged by proteins found in autoimmune thyroid disease that activate or block post-receptor effects (see Chap. 196 and Chap. 197). TSH stimulation induces thyroid growth and differentiation and all phases of iodine metabolism from uptake to secretion of T3 and T4. In addition, there is some autoregulation of thyroid cells by iodide; however, the principal regulation of the thyroid gland is from the pituitary. Recombinant human TSH, which has now been commercially manufactured and has a longer half-life than endogenous TSH, increases serum Tg, T4, and T3 after its administration to normal volunteers.4 This difference is probably due to increased sialylation of the recombinant form. When available for clinical use, recombinant human TSH will expand our understanding of thyroid gland responses to TSH and be of major benefit in the management of thyroid cancer patients. PERIPHERAL THYROID METABOLISM Many disease states result in decreased T4 to T3 conversion.11,13 Although T3 levels may be markedly depressed, serum TSH does not rise in this setting. A dissociation of peripheral (hepatic, 5'D-I) and central (pituitary, 5'D-II) generation of T3, as outlined in Table 30-3, or tissue specific hormone uptake,19 may help explain this observation. Very severe NTI may depress total T4 to nearly undetectable levels and elevate free T4, which suggests a decreased binding to carrier proteins (see Chap. 36). Intracellular free T3 binds to several types of specific nuclear receptors. These receptors may activate or inactivate postreceptor responses depending on the isoform of the receptor (see Chap. 31). The intracellular free T3 concentration is mediated by both passive and active tissue uptake, possible conjugation and deiodination1,29 These combined mechlocally, and efflux of T3 from the tissue. anisms allow for a diverse tissue specific delivery and effect of thyroid hormones. One contribution to tissue heterogeneity is the uncoupling protein-3 (UCP-3), which may be linked to some metabolic actions of T 3.30 Therefore, it is not surprising to find thyroid hormone analogs that have augmented hepatic and skeletal effects when compared with equivalent changes in serum TSH caused by thyroxine.31 This type of physiology may be analogous to the tissue specific effects of estrogen receptor agonists and antagonists. Increased tissue requirements of thyroid hormones may occur during pregnancy in hypothyroid women32 and also may occur with extensive exercise.33 These variations in normal physiology depend on energy balance, tissue hormone use, and systemic and local responses, so that thyroid hormone delivery can fluctuate over a wide range (see Chap. 36). CHAPTER REFERENCES
1. Cavalieri RR. Iodine metabolism and thyroid physiology: current concepts. Thyroid 1997; 7:177. 2. Hollowell JG, Staehling NW, Hannon H, et al. Iodine nutrition in the United States. Trends and public health implications: iodine excretion data from National Health and Nutrition Examination Surveys I and III (1971 1974 and 19881994). J Clin Endocrinol Metab 1998; 83:3401. 3. Kaptein EM. Thyroid hormone metabolism and thyroid disease in chronic renal failure. Endocr Rev 1996; 17:45. 4. Ramirez L, Braverman LE, White B, Emerson CH. Recombinant human thyrotropin is a potent stimulator of thyroid function in normal subjects. J Clin Endocrinol Metab 1997; 82:2836. 5. Danforth E, Burger AG. The impact of nutrition on thyroid hormone physiology and action. Annu Rev Nutr 1989; 9:201. 6. Berry MJ, Larsen PR. The role of selenium in thyroid hormone action. Endocr Rev 1992; 13:207. 7. Arntzenius AB, Smit LJ, Schipper J, et al. Inverse relation between iodine intake and thyroid blood flow: color doppler flow imaging in euthyroid humans. J Clin Endocrinol Metab 1991; 73:1051. 8. Dai G, Levy O, Carrasco N. Cloning and characterization of the thyroid iodide transporter. Nature 1996; 379:458. 9. Tonacchera M, Chiovato L, Pinchera A, et al. Hyperfunctioning thyroid nodules in toxic multinodular goiter share activating thyrotropin receptor mutations with solitary toxic adenoma. J Clin Endocrinol Metab 1998; 83:492. 10. Burmeister LA, Goumaz MO, Mariash CN, Oppenheimer JH. Levothyrox-ine dose requirements for thyrotropin suppression in the treatment of differentiated thyroid cancer. J Clin Endocrinol Metab 1992; 75:344.

11. Engler D, Burger AG. The deiodination of the iodothyronines and of their derivatives in man. Endocrine Rev 1984; 5:151. 12. LoPresti JS, Gray D, Nicoloff JT. Influence of fasting and refeeding on 3,3',5'-triiodothyronine metabolism in man. J Clin Endocrinol Metab 1991; 72:130. 13. Kaptein EM, Kaptein JS, Chang EI, et al. Thyroxine transfer and distribution in critical nonthyroidal illnesses, chronic renal failure, and chronic ethanol abuse. J Clin Endocrinol Metab 1987; 65:606. 13a. Schussler GC. The thyroxine binding proteins. Thyroid 2000; 10:141. 14. Robbins J. Thyroid hormone transport proteins and the physiology of hormone binding. In: Braverman LE, Utiger RD, eds. Werner and Ingbar's the thyroid: a fundamental and clinical text, 7th ed. Philadelphia: Lippincott Raven, 1996:96. 15. Schreiber G, Southwell BR, Richardson SJ. Hormone delivery systems to the brain-transthyretin. Exp Clin Endocrinol Diabetes 1995; 103:75. 16. Bianchi R, Iervasi G, Pilo A, et al. Role of serum carrier proteins in the peripheral metabolism and tissue distribution of thyroid hormones in familial dysalbuminemic hyperthyroxinemia and congenital elevation of thyroxine-binding globulin. J Clin Invest 1987; 80:522. 17. Mol JA, Visser TJ. Rapid and selective inner ring deiodination of thyroxine sulfate by rat liver deiodinase. Endocrinology 1985; 117:8. 18. LoPresti JS, Nicoloff JT. 3,5,3'-Triiodothyronine (T3) sulfate: a major metabolite in T3 metabolism in man. J Clin Endocrinol Metab 1994; 78:688. 19. Everts ME, deJong M, Lim C, et al. Different regulation of thyroid hormone transport in liver and pituitary: its possible role in the maintenance of low T3 production during nonthyroidal illness and fasting in man. Thyroid 1996; 6:359. 20. Croteau W, Davey JC, Galton VA, St Germain DL. Cloning of the mammalian type II iodothyronine deiodinase. A selenoprotein differentially expressed and regulated in human and rat brain and other tissues. J Clin Invest 1996; 98:405. 21. Croteau W, Whittemore SL, Schneider MJ, St Germain DL. Cloning and expression of a cDNA for mammalian type III iodothyronine deiodinase. J Biol Chem 1995; 270:16569. 22. Solomon BL, Wartofsky L, Burman KD. Adjunctive cholestyramine therapy for thyrotoxicosis. Clin Endocrinol 1993; 38:39. 23. Reed HL. Environmental influences on thyroid hormone regulation. In: Braverman LE, Utiger RD, eds. Werner and Ingbar's the thyroid: a fundamental and clinical text, 7th ed. Philadelphia: LippincottRaven, 1996:259. 24. Greenspan SL, Klibanski A, Rowe JW, Elahi D. Age-related alterations in pulsatile secretion of TSH: role of dopaminergic regulation. Am J Physiol 1991; 260:E486. 25. Maes M, Mommen K, Hendrickx D, et al. Components of biological variation, including seasonality, in blood concentrations of TSH, TT 3, FT4, PRL, cortisol and testosterone in healthy volunteers. Clin Endocrinol (Oxf) 1997; 46:587. 26. Morley JE. Neuroendocrine control of thyrotropin secretion. Endocr Rev 1981; 2:396. 27. Spencer CA, LoPresti JS, Patel A, et al. Applications of a new chemilumino-metric thyrotropin assay to subnormal measurement. J Clin Endocrinol Metab 1990; 70:453. 28. al-Adsani H, Hoffer LJ, Silva JE. Resting energy expenditure is sensitive to small dose changes in patients on chronic thyroid hormone replacement. J Clin Endocrinol Metab 1997; 82:1118. 29. Ribeiro RC, Cavalieri RR, Lomri N, et al. Thyroid hormone export regulates cellular hormone content and response. J Biol Chem 1996; 271:17147. 30. Freake HC. Uncoupling proteins: beyond brown adipose tissue. Nutr Rev 1998; 56:185. 31. Sherman SI, Ringel MD, Smith MJ, et al. Augmented hepatic and skeletal thyromimetic effects of tiratricol in comparison with levothyroxine. J Clin Endocrinol Metab 1997; 82:2153. 32. Mandel SJ, Larsen PR, Seely EW, Brent GA. Increased need for thyroxine during pregnancy in women with primary hypothyroidism. N Engl J Med 1990; 323:91. 33. Rone JK, Dons RF, Reed HL. The effect of endurance training on serum tri-iodothyronine kinetics in man: physical conditioning marked by enhanced thyroid hormone metabolism. Clin Endocrinol 1992; 37:325.

CHAPTER 31 THYROID PHYSIOLOGY: HORMONE ACTION, RECEPTORS, AND POSTRECEPTOR EVENTS Principles and Practice of Endocrinology and Metabolism

CHAPTER 31 THYROID PHYSIOLOGY: HORMONE ACTION, RECEPTORS, AND POSTRECEPTOR EVENTS

PAUL M. YEN Target Gene Regulation by Thyroid Hormone Thyroid-Hormone Receptors and their structure Thyroid-Hormone Receptor Isoforms Thyroid Hormone Response Elements Thyroid-Hormone Receptor Binding to Thyroid Hormone Response Elements Repression of Basal Transcription/Corepressors Transcriptional Activation by Thyroid Hormone Coactivators Model of Thyroid-Hormone Receptor Action Chapter References

TARGET GENE REGULATION BY THYROID HORMONE

The effects of thyroid hormone (TH; L-thyroxine, T4; L-triiodo-thyronine, T3) on cellular growth, development, and metabolism are legion and diverse. TH exerts its major effects at the genomic level; however, examples have been found of nongenomic action in the cytoplasm, plasma membrane, and mitochondrion.1 Early studies clearly demonstrated that TH could bind to nuclear sites and stimulate transcription and translation of new proteins.2 These early studies led to a model for genomic TH action for which the general features have been confirmed, but for which much refinement has occurred as the molecular details of TH action have been elucidated. As seen in Figure 31-1, circulating free TH enters the cell by passive diffusion, and in some tissues is converted from T4 to the more biologically potent T3 by deiodinase activity. TH then enters the nucleus and specifically binds to nuclear thyroid-hormone receptors (TRs). TRs have been shown to be intimately associated with chromatin, and ligand binding to TRs stimulates transcription of target genes and protein synthesis in a concentration- and time-dependent manner.

FIGURE 31-1. General model for thyroid hormone action in the nucleus. (mRNA, messenger RNA; RXR, retinoid X receptor; TR, thyroid-hormone receptor; T3, triiodothyronine; T4, thyroxine.)

This model of TH action is based on the notion that the amount of T3 binding to nuclear receptors correlates with gene activation. In the euthyroid rat, research has shown that ~50% of T3 nuclear receptors are bound to TH.2 In the hyperthyroid state, a corresponding increase in T3 occupancy is seen. In general, receptor occupancy correlates well with gene activation for those genes that have been characterized. Approximately 30 genes have been described that are regulated by TH.3 Most of these genes are positively regulated, but interestingly, some genes are negatively regulated by TH. A partial list of some of the better-characterized genes, which have contributed to our understanding of TH action, is included in Table 31-1. Transcriptional regulation by TH can occur via effects on the rate of transcription (rate of nuclear RNA synthesis) and/or stabilization of mRNA. In most cases, TH appears to regulate the transcriptional rate directly, although TH can modulate mRNA stability in certain instances (e.g., malic enzyme and thyroid-stimulating hormone b subunit [TSH-b] genes).4,5

TABLE 31-1. Genes Regulated by Thyroid Hormone

THYROID-HORMONE RECEPTORS AND THEIR STRUCTURE

Two related TRs, which have been identified, bind TH with high affinity and specificity6,7; these TRs, initially called c-erbAs, are protooncogenes of a previously described viral oncogene product, v-erbA, that causes erythroblastosis in chicks. On the basis of amino-acid sequence homology and similarity in the organization of functional domains of receptors, TRs belong to a large superfamily of nuclear-hormone receptors that include the steroid hormone, vitamin D, and retinoic acid receptors. Like the other family members, TRs contain a central DNA-binding domain with two zinc finger motifs and a carboxy-terminal ligand-binding domain8,9 (Fig. 31-2). The ligand-binding domain is important for ligand binding and also for transactivation and dimerization. X-ray crystallographic studies of liganded TR have shown that TH is embedded in a hydrophobic pocket flanked by discontinuous stretches of the ligand-binding domain.10 A subregion important for transcriptional activation (AF-2)11,12 also exists at the extreme carboxy terminus. This subregion is highly conserved within the nuclear-hormone receptor superfamily and appears to undergo a major conformational change on ligand binding.11,12 The ligand-binding domain contains at least nine hydrophobic, heptad repeats, which potentially may be involved in TR homo- and heterodimerization.13 Mutations in the ninth heptad region abrogates TR heterodimerization, thus suggesting that this may be a particularly important region for dimerization. The hinge region between the DNA- and TH-binding domains contains a nuclear-localization motif common among nuclear-hormone receptors.8,9 However, unlike steroid-hormone receptors, which associate with cytoplasmic heat shock proteins in the absence of ligand, TRs are localized predominantly in the nucleus and bind DNA even in the absence of ligand.

FIGURE 31-2. General organization of major thyroid-hormone receptor domains and functional subregions.

THYROID-HORMONE RECEPTOR ISOFORMS There are two genes encoding TRs, a and b, which are located on human chromosomes 17 and 3.8,9 These genes encode TR isoforms (TRs a-1, b-1, b-2) that bind T3 with similar affinity (reported dissociation constant [Kd] between 1010 to 1011 mol/L) and mediate TH-regulated gene expression. These TR isoforms range from 400 to slightly more than 500 amino acids in size.8,9 The DNA-binding and ligand-binding domains are highly conserved among the isoforms and also among different mammalian species (Fig. 31-3). By alternative splicing, the TRb gene generates two mRNAs that encode two proteins: TRa-1 and c-erbA a-2, which differ only in the coding region of the last exon (see Fig. 31-3). Because this region is critical for ligand binding, c-erbA a-2 cannot bind TH. In cotransfection experiments, c-erbA a-2 blocks the transcriptional activity of TRs, so this alternative splice product may antagonize TR action.

FIGURE 31-3. Comparison of amino-acid homologies and their functional properties among thyroid-hormone receptor (TR) isoforms; length of receptors is indicated just above receptor diagrams and percentage amino-acid homology with TRb-2 is included within the receptor diagrams. (T3, triiodothyronine.)

The TRb gene encodes two TRs, TRb-1 and TRb-2, that are generated by alternative promoter choice and/or mRNA splicing.8,9 These receptors have different amino-terminal regions but are otherwise identical. They appear to have similar TH binding affinity and transcriptional activity. TRa-1 and TRb-1 are expressed in almost all tissues; in contrast, TRb-2 is selectively expressed in the anterior pituitary gland and specific areas of the hypothalamus, as well as in the developing brain and inner ear.14 The specific roles of these isoforms are poorly understood. Cotransfection and knockout-mouse studies suggest that many TH-regulated effects are redundant and may be regulated by either TRa or TRb. However, some studies suggest that TRb-2 may be involved in the negative regulation of anterior-pituitary genes such as TSH-b,15 and that TRb may play important roles in brain and inner ear development.14 During development, important differences are seen in the timing of expression of these isoforms, as TRa-1 mRNA is expressed in the rat fetal brain whereas TRb-1 mRNA is induced in the brain just before birth.2 The regulation of the TR mRNAs is isoform- and cell-type dependent. In the intact rat pituitary, T3 decreases TRb-2 mRNA, modestly decreases TRa-1 mRNA, and slightly increases rat TRb-1 mRNA.16 This results in a 30% decrease in total T3 binding in the T3-treated rat pituitary. In other rat tissues, T 3 slightly decreases TRa-1 and c-erb a-2 mRNA except in the brain, where c-erbA a-2 levels are not affected. TRb-1 mRNA is affected minimally in non-pituitary tissues. In patients with nonthyroidal illness who have decreased free T3 and T4 serum levels, increased TRa and TRb mRNA levels have been observed in peripheral mononuclear cells and liver biopsy specimens.17 Thus, induction of TR expression may compensate for decreased circulating TH levels in some of these patients.

THYROID HORMONE RESPONSE ELEMENTS

TRs are ligand-dependent transcription factors that bind to distinct DNA sequences, generally in the promoter region of target genes. TRs frequently positively regulate target genes by stimulating gene transcription in the presence of TH; however, they also can negatively regulate transcription of certain genes (e.g., TSH-b, a glycoprotein subunit, and thyrotropin-releasing hormone). In vitro binding and functional analyses of thyroid hormone response elements (TREs) from positively regulated target genes have demonstrated that TREs generally contain a hexamer half-site sequence of AGGT(C/A)A arranged as two or more tandem repeats.3 Similar to steroid-hormone receptors, TRs bind to TREs as dimers. However, unlike steroid-hormone receptors, which bind to two well-conserved palindromic half-sites, TRs bind to TREs that vary in their primary nucleotide sequence as well as the number, spacing, and orientation of their half-sites.3,9,18 In particular, TRs bind to TREs in which half-sites are arranged as direct repeats, inverted palindromes, and palindromes (Fig. 31-4). Of the ~30 natural TREs that have been characterized, most are arranged as direct repeats, followed by inverted palindromes, and then palindromes. In simple TREs containing two half-sites, the optimal spacings of half-sites in these arrangements are four, six, and zero nucleotides, respectively.3,8,9,18 Several studies have shown that flanking sequences around the half-sites also can contribute to transcriptional activity.18

FIGURE 31-4. Half-site orientations and optimal nucleotide spacing between half-sites. N refers to nucleotides and arrows indicate directions to half-sites on the sense strand.

THYROID-HORMONE RECEPTOR BINDING TO THYROID HORMONE RESPONSE ELEMENTS

TRs bind to TREs as monomers, dimers, and heterodimers in vitro.8,9,18 However, the functional roles of these putative complexes in transcriptional regulation are not well understood. Although TRs can form heterodimers with several other family members on TREs, their major heterodimer partner appears to be the retinoid X receptor (RXR), which bears some amino-acid homology with the retinoic acid receptor and binds 9-cis retinoic acid.8,9,18 At least three major isoforms of RXR exist, so possibly different TR/RXR-isoform complexes may have differential affinities for TREs and/or abilities to transactivate target genes. TR/RXR heterodimers are likely to be involved in TH-mediated transcriptional regulation. In in vitro experiments, T3 caused rapid dissociation of TR homodimers from TREs (direct repeats and inverted palindrome) but had little effect on the binding of TR/RXR heterodimer to TREs,9 results which suggest that the latter complexes may be involved in transcriptional activation. On the other hand, both TR homo- and heterodimers bind to TREs in the absence of ligand and, thus, could be involved in repression of basal transcription by TRs (see the following).

REPRESSION OF BASAL TRANSCRIPTION/COREPRESSORS

Unliganded TRs bind to TREs and can repress transcription of positively regulated target genes (Fig. 31-5). Although the physiologic significance of this basal repression is not known, it may influence target-gene expression during early fetal development when the fetal thyroid gland is incapable of producing TH. The basal repression may be due, in part, to interactions with the basal transcription machinery as unliganded TRs can interact with transcription factor IIB (TFIIB), a key component of the basal transcription complex.19 However, several laboratories have cloned proteins that interact with TR and retinoic acid receptor (RAR) in the absence of their cognate ligands.11,12 These proteins repressed basal transcription by TR and RAR and have been termed corepressors. One of these corepressors is a 270-kDa protein termed nuclear-receptor corepressor (N-CoR). It has two transferable repression domains and a carboxy-terminal a-helical interaction domain. A truncated version of N-CoR, N-CoRi, which is missing the repressor region, also has been identified and may represent an alternative-splice variant of N-CoR.20 This protein blocks basal repression by N-CoR and thus may serve as a natural antagonist for N-CoR. Another corepressor, silencing mediator for RAR and TR (SMRT), is a 168-kDa protein that has some homology with N-CoR.11,12

FIGURE 31-5. Model for repression, derepression, and transcriptional activation by thyroid-hormone receptor (TR). (RXR, retinoid X receptor; T3, triiodothyronine.)

The hinge region of TR (located between the DNA- and ligand-binding domains) is important for interactions with corepressors. Mutations in this region decrease interactions with corepressors and abrogate basal repression, without affecting transcriptional activation.11,12 Several groups showed that corepressors can complex with another putative corepressor, mSin3, and histone deacetylase.21,22 These findings suggest that local histone deacetylation may play an important role in basal repression by altering the local chromatin structure near the minimal promoter region (where transcription is initiated). In negatively regulated target genes, ligand-independent activation of transcription occurs in the absence of TH.23,24 This activation may determine a set point from which ligand-dependent negative regulation begins. The mechanism of the ligand-independent activation is not known but may also involve TR interaction with corepressors.20,23

TRANSCRIPTIONAL ACTIVATION BY THYROID HORMONE COACTIVATORS

Many factors potentially can modulate TH-mediated transcription. These include variability among TR isoforms, TR complexes, heterodimerization partners, nature of TREs, and TR phosphorylation state.8,23a,23b These factors likely influence interactions of liganded TR with TR-associated nuclear proteins called coactivators. Several such proteins have been identified that interact with liganded TR and enhance TH-mediated transcriptional activation.11,12 A 160-kDa protein called steroid receptor coactivator-1 (SRC-1)25 has been identified that interacts with steroid-hormone receptors and TR. SRC-1 mRNA undergoes alternative splicing to generate multiple SRC-1 isoforms, although the functional significance of these SRC-1 isoforms currently is not known.26 Several other 160-kDa proteins also interact with liganded nuclear-hormone receptors and TRs, as well as share partial sequence homology with SRC-1. These findings suggest that a family of coactivators related to SRC-1 may exist.11,12 Some studies have suggested that interaction of coactivators with nuclear-hormone receptors involves the carboxy-terminal AF-2 subregion, although other subregions of the TR ligand-binding domain also may be involved. As seen in Figure 31-6, these putative coactivators have several common features. First, multiple putative nuclear-hormone receptor interaction sites are present, which have a signature LXXLL sequence motif.11,12 Several coactivators also have a polyglutamine region, similar to that found in androgen receptors. In addition, a bHLH motif is seen in the amino-terminal region, suggesting that these coactivators may bind to DNA. Also located in this region is the PAS (Per/Arnt/Sim) domain, which is shared by a number of hypothalamic genes that regulate circadian rhythm and may serve as a dimer interface with other cofactors. Finally, in addition to the SRC-1 family, other cofactors may be associated with liganded TR.11,12,27 The functional roles of these proteins are not well understood.

FIGURE 31-6. Comparison of the organization and structure of putative nuclear-hormone receptor coactivators. (SRC-1, steroid receptor coactivator-1; TIF-2, transcription intermediary factor 2; GRIP-1, glucocorti-coid receptor interacting protein; RAC-3, receptor-associated coactivator 3; TRAM-1, thyroid receptor activator molecule 1; ACTR, activator of thyroid receptor; AIB-1, amplified in breast-1; PCIP, p300/CBP cointegrator-associated protein.)

SRC-1 can interact with the cyclic AMP response element binding protein (CREB)binding protein (CBP), the putative coactivator for cyclic adenosine monophosphatestimulated transcription as well as the related protein, p300, which interacts with the viral coactivator E1A.11,12 These proteins might serve as integrator molecules for different signaling inputs such as protein kinase A- and C-pathwaymediated transcription as well as bridge-liganded TRs with other adapter molecules and/or the basal transcriptional machinery11,12,28 (see Fig. 31-6). In addition, TRs, coactivators, CBP, and the histone acetylase, p300, and CBP-associated

factor (PCAF) can form a complex that can remodel local chromatin structure. Indeed, CBP and PCAF as well as SRC-111,12 have intrinsic histone acetylase activity, although the histone substrates are different for these proteins.

MODEL OF THYROID-HORMONE RECEPTOR ACTION

On the basis of these findings, a model for the mechanism of basal repression and transcriptional activation has emerged (Fig. 31-7). When ligand is absent, TR homodimers or TR/RXR heterodimers bind to the TRE and complex with corepressor, which, in turn, interacts with mSin3 or a related protein, and histone deacetylase. This complex may keep surrounding histones deacetylated and maintain chromatin near the TRE in a transcriptionally repressed state. When ligand is present, the TR/corepressor complex dissociates and is replaced by a coactivator complex that likely contains CBP and the histone acetylase p300/CBP associated factor (PCAF). These changes result in remodeling of chromatin structure and nucleosome positioning. The subsequent recruitment of RNA polymerase II to the transcription initiation site within the minimal promoter results in transcriptional activation. This model is probably an oversimplification as other proteins likely exist that form the TR/coactivator complex,29,30 which may interact directly with the basal-transcriptional machinery or other transcription factors. The identities of these proteins and their protein/protein interactions remain to be elucidated.

FIGURE 31-7. Molecular model for basal repression in the absence of triiodothyronine (T3) and transcriptional activation in the presence of T3. X refers to possible additional cofactors that remain to be identified. (See text for details.) (HDAC, histone deacetylase; TAF, TBP-associated factors; TFIIE, transcription factor IIE; RXR, retinoid X receptor; TR, thyroid-hormone receptor; TFIIB, transcription factor IIB; TBP, TATA-binding protein; TRE, thyroid hormone response element; TATA, tumor-associated transplantation antigen; P/CAF, p300/CBP associated factor; CBP, CREB-binding protein.)

Our understanding of TR action has grown at an accelerating pace and has shed light on both nuclear-hormone receptor action and general mechanisms of transcriptional regulation.31a As more of the molecular details of TR action have become known, so has the molecular basis of human diseases that involve TRs and coactivators. As is discussed in the next chapter (Chap. 32), mutations of TRb-1 have been associated with resistance to TH, an autosomal disorder in which patients have elevated serum concentrations of TH and inappropriately normal thyrotropin levels. Almost all of these patients have mutations in the TRb-1 ligand-binding domain that decrease TH-binding affinity but still allow the TR to bind to DNA.31 These receptors block the transcriptional activity of wild-type receptors (dominant negative activity). These mutant receptors may have defects in corepressor release and interactions with coactivators.32,33 In addition, mutations in CBP have been associated with Rubinstein-Taybi syndrome, a congenital neurologic disorder.34 Last, amplification and overexpression of the coactivator AIB-1 may be associated with primary human breast cancer.35 New knowledge of TR action should provide further insight into the molecular basis of these and other diseases involving nuclear-hormone receptors. CHAPTER REFERENCES
1. Davis PJ, Davis FB. Nongenomic actions of thyroid hormone. Thyroid 1996; 6:497. 2. Oppenheimer J, Schwartz H, Strait K. An integrated view of thyroid hormone actions in vivo. In: Weintraub B, ed. Molecular endocrinology: basic concepts and clinical correlations. New York: Raven Press, 1995:249. 3. Williams G, Brent G. Thyroid hormone response elements. In: Weintraub B, ed. Molecular endocrinology: basic concepts and clinical correlations. New York: Raven Press, 1995:217. 4. Leedman PJ, Stein AR, Chin WW. Regulated specific protein binding to a conserved region of the 3'-untranslated region of thyrotropin beta-subunit mRNA. Mol Endocrinol 1995; 9:375. 5. Oppenheimer J, Schwartz H, Mariash C, et al. Advances in our understanding of thyroid hormone action at the cellular level. Endocr Rev 1987; 8:288. 6. Weinberger C, Thompson C, Ong E, et al. The c-erbA gene encodes a thyroid hormone receptor. Nature 1986; 324:641. 7. Sap J, Munoz A, Damm K, Vennstrom B. The c-erbA protein is a high affinity receptor for thyroid hormone. Nature 1986; 324:635. 8. Lazar M. Recent progress in understanding thyroid hormone action. Thyroid Today 1997; 20:1. 9. Yen P, Chin W. New advances in understanding the molecular mechanisms of thyroid hormone action. Trends Endocrinol Metab 1994; 5:65. 10. Wagner R, Apriletti J, McGrath M, et al. A structural role for hormone in the thyroid hormone receptor. Nature 1995; 378:690. 11. Torchia J, Glass C, Rosenfeld M. Co-activators and co-repressors in the integration of transcriptional responses. Curr Opin Cell Biol 1998; 10:373. 12. McKenna N, Lanz R, O'Malley B. Nuclear receptor co-regulators. Cell Mol Biol 1999; 20:321. 13. Forman B, Yang C, Au M, et al. A domain continuing leucine-zipper-like motifs mediate novel in vivo interactions between the thyroid hormone and retinoic acid receptors. Mol Endocrinol 1989; 3:1610. 14. Forrest D, Golarai G, Connor J, Curran T. Genetic analysis of thyroid hormone receptors in development and disease. Recent Prog Horm Res 1996; 51:1. 15. Langlois MF, Zanger K, Monden T, et al. A unique role of the beta thyroid hormone receptor isoform in negative regulation by thyroid hormone. Mapping of a novel amino-terminal domain important for ligand-independent activation. J Biol Chem 1997; 272:24927. 16. Hodin R, Lazar M, Chin W. Differential and tissue-specific regulation of the multiple rat c-erbA mRNA species by thyroid hormone. J Clin Invest 1990; 85:101. 17. Williams G, Franklyn J, Neuberger J, Sheppard M. Thyroid hormone receptor expression in the sick euthyroid syndrome. Lancet 1989; 2:1477. 18. Glass C. Differential recognition of target genes by nuclear receptor monomers, dimers and heterodimers. Endocr Rev 1994; 15:391. 19. Baniahmad A, Ha I, Reinberg D, et al. Interaction of human thyroid hormone receptor b with transcription factor TFIIB may mediate target gene derepression and activation by thyroid hormone. Proc Natl Acad Sci U S A 1993; 90:8832. 20. Hollenberg A, Monden T, Madura J, et al. Function of nuclear co-repressor protein on thyroid hormone response elements is regulated by the receptor A/B domain. J Biol Chem 1996; 271:28516. 21. Alland L, Muhle R, Hou H Jr, et al. Role for N-CoR and histone deacetylase in Sin3-mediated transcriptional repression. Nature 1997; 387:49. 22. Heinzel T, Lavinsky RM, Mullen TM, et al. A complex containing N-CoR, mSin3 and histone deacetylase mediates transcriptional repression. Nature 1997; 387:43. 23. Tagami T, Madison LD, Nagaya T, Jameson JL. Nuclear receptor corepressors activate rather than suppress basal transcription of genes that are negatively regulated by thyroid hormone. Mol Cell Biol 1997; 17:2642. 23a. Takeda T, Nagasawa T, Miyamoto T, et al. Quantitative analysis of DNA binding affinity and dimerization properties of wild-type and mutant thyroid hormone receptor b1. Thyroid 2000; 10:11. 23b. Forrest D, Vennstrm B. Functions of thyroid hormone receptors in mice. Thyroid 2000; 10:11. 24. Hollenberg A, Monden T, Flynn T, et al. The human thyrotropin-releasing hormone gene is regulated by thyroid hormone through two distinct classes of negative thyroid hormone response elements. Mol Endocrinol 1995; 9:540. 25. Onate S, Tsai S, Tsai M, O'Malley B. Sequence and characterization of a coactivator for the steroid hormone receptor superfamily. Science 1995; 270:1354. 26. Kamei Y, Xu L, Heinzel T, et al. A CBP integrator complex mediates transcriptional activation and AP-1 inhibition by nuclear receptors. Cell 1996; 85:403. 27. Lee J, Choi H, Gyuris J, et al. Two classes of proteins dependent on either the presence or absence of thyroid hormone for interaction with the thyroid hormone receptor. Mol Endocrinol 1995; 9:243. 28. Korzus E, Torchia J, Rose DW, et al. Transcription factorspecific requirements for coactivators and their acetyltransferase functions. Science 1998; 279:703. 29. Fondell J, Ge J, Roeder R. Ligand induction of a transcriptionally active thyroid hormone receptor coactivator complex. Proc Natl Acad Sci U S A 1996; 93:8329. 30. Freedman L. Increasing the complexity of coactivation in nuclear receptor signaling. Cell 1999; 97:5. 31. Refetoff S, Weiss R, Usala S. The syndromes of resistance to thyroid hormone. Endocr Rev 1993; 14:348. 31a. Pohlenz J, Manders L, Sadow PM, et al. A novel point mutation in cluster 3 of the thyroid hormone receptor beta gene (P247L) causing mild resistance to thyroid hormone. Thyroid 1999; 9:1195. 32. Yoh SM, Chatterjee VK, Privalsky ML. Thyroid hormone resistance syndrome manifests as an aberrant interaction between mutant T3 receptors and transcriptional corepressors. Mol Endocrinol 1997; 11:470. 33. Liu Y, Takeshita A, Misiti S, et al. Lack of coactivator interaction can be a mechanism for dominant negative activity by mutant thyroid hormone receptors. Endocrinology 1998; 139:4197. 34. Petrij F, Giles R, Dauwerse H, et al. Rubinstein-Taybi syndrome caused by mutations in the transcriptional co-activator CBP. Nature 1995; 376:348. 35. Anzick SL, Kononen J, Walker RL, et al. AIB1, a steroid receptor coactivator amplified in breast and ovarian cancer. Science 1997; 277:965.

CHAPTER 32 THYROID HORMONE RESISTANCE SYNDROMES Principles and Practice of Endocrinology and Metabolism

CHAPTER 32 THYROID HORMONE RESISTANCE SYNDROMES

STEPHEN JON USALA Clinical Forms of Resistance to Thyroid Hormone Genetic Mutations in Thyroid Hormone Resistance Syndromes Homozygous c-erbab Mutations in Refetoff and Bercu Patients Molecular Mechanisms with Dominant Negative b Receptors Clinical Heterogeneity in Thyroid Hormone Resistance Syndromes Treatment Considerations for Thyroid Hormone Resistance Chapter References

CLINICAL FORMS OF RESISTANCE TO THYROID HORMONE

In 1967, a bizarre finding in two young deaf-mute patients was reported by Refetoff and coworkers.1 These patients had delayed bone ages and stippled epiphyses consistent with juvenile hypothyroidism but, paradoxically, significantly elevated levels of circulating thyroid hormones. It was concluded that these patients had a tissue-specific combination of hypothyroidism, hyperthyroidism, and euthyroidism. Because tissues such as bone and brain were judged to lack the appropriate sensitivity to the high levels of thyroid hormone and the patients failed to manifest other common symptoms and signs of hyperthyroidism, the proposal was made that the Refetoff patients were the first examples of thyroid hormone resistance. Time has validated this conclusion. Many kindreds and sporadic patients with thyroid hormone resistance syndromes have been studied. The most common form of thyroid hormone resistance is called generalized resistance to thyroid hormone2 (Fig. 32-1). Patients with this form must satisfy three criteria: elevated serum levels of free thyroid hormones (i.e., free triiodothyronine [T3] and free thyroxine [T4]), inappropriately normal levels of serum thyroid-stimulating hormone (TSH), and clinical euthyroidism. Patients with the syndrome of generalized resistance to thyroid hormone manifest a proportional decrease in sensitivity to thyroid hormone in the pituitary and many of the peripheral tissues. As manifested by the Refetoff patients, the degree of tissue sensitivity to thyroid hormone may vary. In other examples, patients with generalized resistance to thyroid hormone are affected in the brain (i.e., reduced intelligence quotient resulting in part from resistance) but have no evidence of resistance in the bone compartment (i.e., normal final adult height).2 The variable tissue resistances are reflected in the different phenotypes of generalized resistance to thyroid hormone. By definition, however, the patients lack the clinical signs and symptoms of hyperthyroidism, such as weight loss, heat intolerance, nervousness, and hyperkinesis, and the compensatory elevation of thyroid hormone is such that the patients do not have complaints of hypothyroidism. In 108 different kindreds with generalized resistance to thyroid hormone, most patients were identified after screening for goiter.2

FIGURE 32-1. Three clinical forms of thyroid hormone resistance in humans. The first shows blockage of the biologic responses to thyroxine (T4) and triiodothyronine (T3) in the pituitary and peripheral tissues, which is characteristic of generalized resistance to thyroid hormone. The second shows selective blockage at the pituitary, which is characteristic of selective pituitary resistance to thyroid hormone. The third shows selective blockage at the peripheral tissues, as described in one patient with selective peripheral resistance to thyroid hormone. Patients with identical mutations in the c-erbAb thyroid-hormone receptor can present clinically with generalized or selective pituitary resistance. TSH, thyroid-stimulating hormone. (Modified from Usala SJ, Weintraub BD. Familial thyroid hormone resistance: clinical and molecular studies. In: Mazzaferri E, Kreisberg RA, Bar RS, eds. Advances in endocrinology and metabolism, vol 2. Chicago: MosbyYear Book, 1991:59.)

A less common form of resistance is selective pituitary resistance to thyroid hormone, in which the sensitivity to thyroid hormone is decreased in the pituitary relative to peripheral tissues (see Fig. 32-1). Most of the patients with this syndrome have inappropriately normal TSH at elevated levels of free thyroid hormone, as in generalized resistance to thyroid hormone, but with signs and symptoms of hyperthyroidism.2,3 The last criterion is often a subjective one; the assignment of selective pituitary resistance or generalized resistance to thyroid hormone may be less than rigorous. Because the phenotypic borders are often not sharp, differentiating selective pituitary resistance from generalized resistance to thyroid hormone can be a problem.2,3

FIGURE 32-2. Mutations in the c-erbAb thyroid-hormone receptor gene responsible for generalized thyroid hormone resistance. The relative sites of the mutations in the triiodothyronine (T3)-binding domain of the receptor with the alterations in amino acids are indicated. The last three exons of the b gene comprise amino acids 247295, 296381, and 382461. The relative deficiencies in the T3-binding affinities of the mutant receptors measured in vitro (1.0 = no defect, 0.01 = 100-fold reduction in Ka) are shown (Ka mutant/Ka wild-type). The mean total thyroxine (TT4), total T3 (TT3), and thyroid-stimulating hormone (TSH) levels for the patients are listed; the results in brackets are mean values expressed as a percent of the corresponding mean levels of unaffected family members or the mean value for the laboratory. F100, F102, and F108 had prior (inappropriate) ablative therapy and were on T3 therapy. Values for F68 are free T 4 and free T3 given in picomoles per liter. The T3-binding domain of c-erbAb spans amino acids 243461, at the COOH terminus. The black bars represent areas that interact with thyroid hormone auxiliary proteins (RXR) and stabilize RXR-receptor heterodimers. Some of these regions may stabilize homodimers of c-erbAb as well. The 349428 region contains heptad repeats (gray boxes) with hydrophobic amino acids that form a leucine zipper structure, which is involved in receptor dimerization. Some of these heptads appear to be critical for the dominant negative function of mutant c-erbAb 1 receptors, although other domains may be involved. (SD, standard deviation.) (From Refetoff S, Weiss RE, Usala SJ. The syndrome of resistance to thyroid hormones. Endocr Rev 1993; 14:348.)

Another form of resistance is selective peripheral resistance to thyroid hormone, which is of more theoretical than therapeutic significance.2,4 In this form, the sensitivity of peripheral tissues to thyroid hormones is decreased relative to that of the pituitary, to the point of clinical hypothyroidism; thyrotropic resistance to thyroid hormone is nonexistent or inadequate, and the TSH secretion is not increased to compensate for the increased peripheral resistance (see Fig. 32-1). One well-studied patient

taking large doses of exogenous T3 had clinical hypothyroidism and suppressed TSH.4 The possibility exists that selective peripheral resistance to thyroid hormones is more common than indicated by this one patient and escapes diagnosis. Making this diagnosis is difficult, because the markers of thyroid hormone action in humans, with the exception of TSH levels, lack quantitativeness.

GENETIC MUTATIONS IN THYROID HORMONE RESISTANCE SYNDROMES

Generalized resistance to thyroid hormone is a genetic disease. Most reported cases belong to families with multiple affected members, and inspection of these pedigrees reveals that, in all but the original Refetoff kindred, the abnormal trait is transmitted dominantly.2 Compelling evidence that generalized resistance to thyroid hormone is a disease of the c-erbAb thyroid-hormone receptor gene has come from linkage studies.2,5,6 Restriction fragment length polymorphisms (RFLPs) have been used to establish tight linkage between the b receptor gene and the syndrome of generalized resistance to thyroid hormone. Given the dominant inheritance pattern of pedigrees of generalized resistance to thyroid hormone, it was originally postulated that resistance occurred because mutant b receptors inhibited the activity of the normal b and a1 receptors (from one and two alleles, respectively).6 This hypothesis was convincingly proved in humans from the recessive kindred that has been extensively studied by Refetoff and coworkers.7,8 Sixty-five different mutations in the c-erbAb gene have been identified in 115 families with either generalized or selective pituitary resistance to thyroid hormone9 (Fig. 32-2). These mutations exist as single (i.e., heterozygous) alleles, except in the Refetoff and Bercu patients.7,8,10,11 The majority of the c-erbAb mutations is located in the penultimate and final exons of the gene and reduce T3-binding affinity; these mutations cluster within two hot spots in the T3-binding domain between amino acids 310-353 and 429-461.2,9 Presently, only one exception exists to the c-erbAb rule. A kindred has been reported in whom the resistance phenotype is not linked to the c-erbAb gene and who has no c-erbAb1 or c-erbAb2 defect.12 These data suggest that other factors that mediate thyroid hormone action may be defective, resulting in thyroid hormone resistance. HOMOZYGOUS C-ERBAb MUTATIONS IN REFETOFF AND BERCU PATIENTS Two humans with mutations, the Refetoff and Bercu patients, provide important information on the interrelationships of the a and b receptors in mediating thyroid hormone action in humans (Fig. 32-3). These patients with mutations are homozygous for very different abnormalities in c-erbAb and have very different phenotypes.

FIGURE 32-3. The Refetoff patient at 8.5 years of age (A) and the Bercu patient at 3.5 weeks of age (B). The Refetoff patient has a complete absence of functional b receptors. A bird-like facies and a pigeon chest are present. The Bercu patient is homozygous for a dominant negative b receptor (i.e., S receptor). Note the exophthalmos. (From Refetoff S, DeWind LD, DeGroot LJ. Familial syndrome combining deaf-mutism, stippled epiphyses, goiter, and abnormally high PBI: possible target organ refractoriness to thyroid hormone. J Clin Endocrinol Metab 1967; 27:279; and from Ono S, Schwartz ID, Mueller OT, et al. Homozygosity for a dominant negative thyroid hormone receptor gene responsible for generalized thyroid hormone resistance. J Clin Endocrinol Metab 1991; 73:990.)

The resistance syndrome of the Refetoff patient is considerably milder than that of the Bercu patient (see Fig. 32-3). Although the original bone radiographs of the first Refetoff patient suggested juvenile hypothyroidism, the final adult height in affected members was above the parental mean. The intelligence quotients of the Refetoff patients were normal compared with the ranges seen in hearing-impaired individuals.13 The Refetoff patients had a major deletion in both b receptor alleles and had only functional a receptors.7,8 Significantly, the obligate heterozygotes in the Refet-off kindred were phenotypically normal; these heterozygotes had normal TSH and thyroid hormone levels. Several conclusions can be drawn from the Refetoff patients: that only one b-receptor allele is necessary (with two a receptor alleles) for normal thyroid hormone action in humans; most, or at least life-sustaining, thyroid hormone action can be mediated solely through the a receptor; and the heterozygous, mutant c-erbAb alleles in resistance patients act as dominant negative genes. The obligate heterozy-gotes demonstrate that it is not the loss of function of a b-receptor allele that results in thyroid hormone resistance. The Bercu patient (see Fig. 32-3) had a complex pattern of hyperthyroidism and hypothyroidism resulting from homozygosity of a dominant negative allele.10,11 This patient manifested severe pituitary resistance with strikingly elevated TSH levels in the setting of high free thyroid hormone levels. Significant growth retardation and profound bone retardation suggested hypothyroidism. However, this patient was tachycardic and appeared to be hypermetabolic, consistent with hyperthyroidism. The patient was severely mentally retarded, although it is not clear whether this was secondary to hyperthyroidism or hypothyroidism or both conditions in distinct regions of the brain.14 Because the Bercu patient was clinically much more severely affected than the Refetoff patients, it follows that dominant negative forms of c-erbAb inhibit at least some thyroid hormone action mediated through the a receptor. The severity of the Bercu patient phenotype may also be due to the silencing activity (active repression) by the double dose of mutant c-erbAb receptor. The variable thyroidal states in the Bercu patient highlights the fact that different tissues are regulated by different relative concentrations of a and b receptors. MOLECULAR MECHANISMS WITH DOMINANT NEGATIVE b RECEPTORS The molecular genetics of the syndrome of generalized resistance to thyroid hormone indicated that the mutant c-erbAb receptors antagonized the normal regulation of gene expression by wild-type a and b receptors. Most of the mutant b1 receptors that have been cloned and synthesized in vitro have compromised T3 binding and transactivating ability, although a broad spectrum of defects exists.15 For example, the mutant S receptor of the Bercu patient has no detectable T3-binding activity and cannot induce expression from heterologous promoters containing thyroid hormone response element (TREs) with exposure to T3 in transient expression experiments.16 However, other human mutant receptors do have demonstrable T3-binding activity and can up-regulate gene expression at high levels of thyroid hormone.15,16 The DNA-binding function of the mutant receptors appears to be required for dominant negative activity.17 No human DNA-binding mutations are found in the b thyroid-hormone receptor gene, except for the total deletion of the b alleles in the Refetoff patients, which is not dominant negative. The hypothesis has been made that the dominant negative activity of the human mutant receptors may result from competitive inhibition by mutant homodimers or heterodimers with wild-type receptors.2 This would be a form of passive repression. Mutant receptors as homodimers or heterodimers may also block formation of the transcriptional preinitiation complex and silence gene transcription.18,19 and 20 This could result in repression of basal gene expression. Novel mutant c-erbAb receptors with relatively normal T3-binding affinities have been isolated from kindreds with pronounced clinical manifestations of thyroid hormone resistance.21,22 The R243Q and R243W c-erbAb mutations may create thyroid hormone resistance in part by increasing mutant homodimer formation or by reducing the receptor affinity for T3 only after it binds to DNA.21 An R383H mutant receptor has been shown to impair release of a corepressor factor in the context of TSH and TRH promoters.22 CLINICAL HETEROGENEITY IN THYROID HORMONE RESISTANCE SYNDROMES The reasons for clinical heterogeneity in thyroid hormone resistance syndromes are complex. The correlation between the degree of functional impairment of the mutant c-erbAb receptor and the phenotype of resistance to thyroid hormone is not precise. Several explanations have been advocated. In fibroblasts from one kindred, an increased proportion of mutant receptor messenger RNA (mRNA) was found that was associated with short stature.23 In other

kindreds in whom the mutant versus wild-type b1 receptor mRNA was quantitated, levels were approximately equivalent.24,25 Another gene or genes could affect the phenotypic expression of a mutant c-erbAb allele.26 Although the mutant c-erbAb allele might be the most dominant gene, its bioeffects could be modulated by other genes such as those for corepressors and coactivators.27 That genetic factors modulate the expression of the resistance phenotype is suggested by comparison of thyroid function tests in unaffected first-degree relatives and relatives by marriage between kindreds with identical c-erbAb mutations.28 For example, among members of a kindred with an R32OC mutation, higher concentrations of free T4 were necessary to maintain a normal TSH level in affected subjects and unaffected first-degree relatives (siblings and children of affected patients) but not relatives by marriage.28 The molecular analysis of thyroid hormone resistance syndromes has revealed that selective pituitary resistance to thyroid hormone can also result from mutations in c-erbAb.29,30 Identical mutations have been found in different patients classified with generalized resistance and selective pituitary resistance to thyroid hormone.2,3,29,30 This clinical discrepancy may result in part from the difficulty in discriminating selective pituitary resistance from generalized resistance. Patients with nervousness and tachycardia as their only thyrotoxic symptom and sign have been variously diagnosed as having selective pituitary resistance or generalized resistance to thyroid hormone. However, other genetic or acquired factors may play a role in determining the relative levels of resistance between pituitary and peripheral tissues. For example, a bona fide case of selective pituitary resistance to thyroid hormone was seen in the proband of a Kindred G-H with the ARG-316-HIS mutation in c-erbAb; this patient had unusually high thyroid hormone levels compared with other kindred members with the identical mutant allele.29,31 Interestingly, thyroid hormone resistance has been diagnosed prenatally.31a

TREATMENT CONSIDERATIONS FOR THYROID HORMONE RESISTANCE

The diagnosis of generalized resistance or selective pituitary resistance to thyroid hormone does have a bearing on treatment. For the patient with generalized resistance who is clinically euthyroid, antithyroid medications, thyroidectomy, and radioiodine treatment are to be strictly avoided.2 In generalized resistance to thyroid hormone, the patient's own hypothalamicpituitarythyroid axis is allowed to render the appropriate level of thyroid hormone. However, in cases of concomitant primary hypothyroidism with elevation in serum TSH, the patient should be treated with thyroid hormone. Patients with generalized resistance who have been inappropriately treated with radioiodine or thyroidectomy may have striking elevations in TSH; for poorly understood reasons, in these patients completely normalizing the serum TSH level and avoiding signs and symptoms of thyrotoxicosis with T4 may be difficult. A minority of adolescents with generalized resistance demonstrate significant bone resistance and subnormal growth.32 The suggestion has been made that exogenous T4 or T3 may augment growth.2,32 For patients receiving thyroid hormone, it should be administered cautiously, with special attention given to cardiac and neurologic status. The treatment of selective pituitary resistance to thyroid hormone is somewhat controversial. This condition must be differentiated from a TSH-secreting adenoma, because the treatment of the latter is surgical.33,34 Radiologic visualization by magnetic resonance imaging or computed tomographic scan is primarily used in the diagnosis. The molar ratio of a-subunit to TSH is often elevated in patients with TSH-secreting adenoma.33,34 Therapy for pituitary resistance has included b-blockers such as propranolol, bromocriptine, methimazole, or propylthiouracil, 3,5,3'-triiodo-thyroacetic acid (i.e., TRIAC), somatostatin, and thyroidectomy.2,3,34,35,36 and 36a Many cases are mild, and no intervention other than observation is required, because the level of symptomatic thyrotoxicosis can abate with time.3 One severely thyrotoxic patient has been successfully treated for several years with methimazole.36 Attention deficit hyperactivity disorder (ADHD) is commonly found in patients with thyroid hormone resistance syndromes.37,38 and 39 Indicative of learning disabilities, discrepancies often exist between achievement and visual-motor development scores, and intelligence quotients. In one study of 49 affected (generalized resistance syndrome) and 55 unaffected members of kindreds, 50% of affected adults met the criteria for ADHD, and 70% of affected children were diagnosed with ADHD.38 However, the converse study, which examined 277 children with ADHD, found no thyroid hormone resistance, although the prevalence of other thyroid abnormalities was 5.4%.39 Thyroid hormone resistance syndromes are a very rare cause of ADHD, although perhaps TSH and free T4 measurements are advisable in children with ADHD given the high prevalence of other thyroid diseases.39 Children with resistance to thyroid hormone have weaker abilities of perceptual organization and lower school achievement compared to those with ADHD, indicating a more severe neurobehavioral impairment than with ADHD.40 In children with resistance to thyroid hormone and ADHD (but not in unaffected children with ADHD), administration of T3 in supra-physiologic doses (2575 g per day) may be beneficial in reducing hyperactivity and impulsivity.41 CHAPTER REFERENCES
1. Refetoff S, DeWind LT, DeGroot LJ. Familial syndrome combining deaf-mutism, stippled epiphyses, goiter, and abnormally high PBI: possible target organ refractoriness to thyroid hormone. J Clin Endocrinol Metab 1967; 27:279. 2. Refetoff S, Weiss RE, Usala SJ. The syndrome of resistance to thyroid hormones. Endocr Rev 1993; 14:348. 3. Beck-Peccoz P, Chatterjee VKK. The variable clinical phenotype in thyroid hormone resistance syndrome. Thyroid 1994; 4:225. 4. Kaplan MM, Swartz SL, Larsen PR. Partial peripheral resistance thyroid hormone. Am J Med 1981; 70:1115. 5. Usala SJ, Weintraub BD. Familial thyroid hormone resistance: clinical and molecular studies. In: Mazzaferri E, Kreisberg RA, Bar RS, eds. Advances in endocrinology and metabolism, vol 2. Chicago: MosbyYear Book, 1991:59. 6. Usala SJ, Bale AE, Gesundheit N, et al. Tight linkage between the syndrome of generalized thyroid hormone resistance and the human c-erbAb gene. Mol Endocrinol 1988; 1:1217. 7. Takeda A, Balzano S, Sakurai A, et al. Screening of nineteen unrelated families with generalized resistance to thyroid hormone for known point mutations in the thyroid-hormone receptor b gene and the detection of a new mutation. J Clin Invest 1991; 87:496. 8. Takeda K, Sakurai A, DeGroot LJ, Refetoff S. Recessive inheritance of thyroid hormone resistance caused by complete deletion of the protein-coding region of the thyroid hormone receptor-b gene. J Clin Endocrinol Metab 1992; 74:49. 9. Announcement 1994. A registry for resistance to thyroid hormone. Mol Endocrinol 1994; 8:1558. 10. Ono S, Schwartz ID, Mueller OT, et al. Homozygosity for a dominant negative thyroid hormone receptor gene responsible for generalized thyroid hormone resistance. J Clin Endocrinol Metab 1991; 73:990. 11. Usala SJ, Menke JB, Watson TL, et al. A homozygous deletion in the c-erbAb thyroid hormone receptor gene in a patient with generalized thyroid hormone resistance: isolation and characterization of the mutant receptor. Mol Endocrinol 1991; 5:327. 12. Weiss RE, Hayashi Y, Nagaya T, et al. Dominant inheritance of resistance to thyroid hormone not linked to defects in the thyroid hormone receptor a or b genes may be due to a defective cofactor. J Clin Endocrinol Metab 1996; 81:4196. 13. Refetoff S, DeGroot LJ, Bernard B, DeWind LT. Studies of a sibship with apparent hereditary resistance to the intracellular action of thyroid hormone. Metabolism 1972; 21:723. 14. Stein SA, Nici JA, Gamache MP, et al. Preliminary characterization of the neurological and neuropsychological abnormalities in a homozygote and heterozygotes of kindred S with generalized thyroid hormone resistance syndrome. Thyroid 1991; 1(Suppl):5. 15. Hayashi Y, Weiss RE, Sarne DH, et al. Do clinical manifestations of resistance to thyroid hormone correlate with the functional alterations of the corresponding mutant thyroid hormone-b receptors? J Clin Endocrinol Metab 1995; 80:3246. 16. Meier CA, Dickstein BM, Ashizawa K, et al. Variable transcriptional activity and ligand binding of mutant b 1, T 3 receptors from four families with generalized resistance to thyroid hormone. Mol Endocrinol 1992; 6:248. 17. Nagaya T, Madison LD, Jameson JL. Thyroid hormone receptor mutant that causes resistance to thyroid hormone. J Biol Chem 1992; 267:13014. 18. Baniahmad A, Tsai SY, O'Malley BW, Tsai MJ. Kindred S thyroid hormone receptor is an active and constitutive silencer and repressor for thyroid hormone and retinoic acid responses. Proc Natl Acad Sci U S A 1992; 89:10633. 19. Fondell JD, Roy AL, Roeder RG. Unliganded thyroid hormone receptor inhibits formation of a functional preinitiation complex: implications for active repression. Genes Dev 1992; 7:1400. 20. Baniahmad A, Ha I, Reinberg D, et al. Interaction of human thyroid hormone receptor b with transcription factor TFIIB may mediate target gene derepression and activation by thyroid hormone. Proc Natl Acad Sci U S A 1993; 90:8832. 21. Yagi H, Pohlenz J, Hayashi Y, et al. Resistance to thyroid hormone caused by two mutant thyroid hormone receptors b, R243Q and R243W, with impairment of function that cannot be explained by altered in vitro 3,5,3'-triiodothyronine-binding affinity. J Clin Endocrinol Metab 1997; 82:1608. 22. Clifton-Bligh RJ, de Zegher F, Wagner RL, et al. A novel TRb mutation (R383H) in resistance to thyroid hormone syndrome predominantly impairs corepressor release and negative transcriptional regulation. Mol Endocrinol 1998; 12:609. 23. Mixson AJ, Hauser P, Tennyson G, et al. Differential expression of mutant and normal beta T3 receptor alleles in kindreds with generalized resistance to thyroid hormone. J Clin Invest 1993; 91:2296. 24. Hayashi Y, Janssen OE, Weiss RE, et al. The relative expression of mutant and normal thyroid hormone receptor genes in patients with generalized resistance to thyroid hormones determined by estimation of their specific messenger RNA products. J Clin Endocrinol Metab 1993; 76:64. 25. Klann RC, Torres B, Menke JB, et al. Competitive PCR quantitation of c-erbAb1, c-erbAa1, and c-era2 messenger RNA levels in normal heterozygous and homozygous fibroblasts of kindred S with thyroid hormone resistance. J Clin Endocrinol Metab 1993; 77:969. 26. Weiss RE, Marcocci C, Bruno-Bossio G, Refetoff S. Multiple genetic factors in the heterogeneity of thyroid hormone resistance. J Clin Endocrinol Metab 1993; 76:257. 27. Koenig RJ. Thyroid hormone receptor coactivators and corepressors. Thyroid 1998; 8:703. 28. Weiss RE, Tunca H, Knapple WL, et al. Phenotype differences of resistance to thyroid hormone in two unrelated families with an identical mutation in the thyroid hormone receptor b gene (R320C). Thyroid 1997; 7:35. 29. Geffner ME, Su F, Ross NS, et al. An arginine to histidine mutation in codon 311 of the c-erbAb gene results in a mutant receptor which does not mediate a dominant negative phenotype. J Clin Invest 1993; 91:538. 30. Mixson AJ, Renault JC, Ransom S, et al. Identification of a novel mutation in the gene encoding the b-triiodothyronine receptor in a patient with apparent selective pituitary resistance to thyroid hormone. Clin Endocrinol 1993; 38:227. 31. Hao E, Menke JB, Smith AM, et al. Divergent dimerization properties of mutant b1 thyroid hormone receptors are associated with different dominant negative activities. Mol Endocrinol 1994; 8:841.

31a. Asteria C, Rajanayagam O, Collingwood T, et al. Prenatal diagnosis of thyroid hormone resistance. J Clin Endocrinol Metab 1999; 84:405. 32. Gesundheit N, Gyves PW, Okihiro MM, et al. Short stature in children with generalized thyroid hormone resistance: clinical and biochemical features in eight patients from three kindreds. (Abstract). Endocrinol 1988; 121(Suppl):T-36. 33. Magner JA. TSH-mediated hyperthyroidism. Endocrinologist 1993; 3:289. 34. Beck-Peccoz P, Mariotti S, Guillausseau PJ. Treatment of hyperthyroidism due to inappropriate secretion of thyrotropin with the somatostatin analog SMS 201995. J Clin Endocrinol Metab 1989; 68:208. 35. Kunitake JM, Hartman N, Henson LC, et al. 3,5,3'-triiodothyroacetic acid therapy for thyroid hormone resistance. J Clin Endocrinol Metab 1989; 69:461. 36. Dulgeroff AJ, Geffner ME, Koyal SN, et al. Bromocriptine and Triac therapy for hyperthyroidism due to pituitary resistance to thyroid hormone. J Clin Endocrinol Metab 1992; 75:1071. 36a. Weiss RE, Refetoff S. Treatment of resistance to thyroid hormone-primum non nocere. J Clin Endocrinol Metab 1999; 84:401. 37. Cugini CD Jr, Leidy JW Jr, Chertow BS, et al. An arginine to histidine mutation in codon 315 of the c-erbAb thyroid hormone receptor in a kindred with generalized resistance to thyroid hormones results in a receptor with significant 3,5,3'-triiodothyronine binding activity. J Clin Endocrinol Metabol 1992; 74:1164. 38. Hauser P, Zametkin AJ, Martinez P. Attention deficit-hyperactivity disorder in people with generalized resistance to thyroid hormone. N Engl J Med 1993; 328:997. 39. Weiss RE, Skin M, Trommer B, Refetoff S. Attention deficit hyperactivity disorder and thyroid function. J Pediatr 1993; 123:539. 40. Stein MA, Weiss RE, Refetoff S. Neurocognitive characteristics of individuals with resistance to thyroid hormone: comparisons with individuals with attention-deficit hyperactivity disorder. Dev Behav Pediatr 1995; 16:406. 41. Weiss RE, Stein MA, Refetoff S. Behavioral effects of Liothyronine (L-T3) in children with attention deficit hyperactivity disorder in the presence and absence of resistance to thyroid hormone. Thyroid 1997; 7:389.

CHAPTER 33 THYROID FUNCTION TESTS Principles and Practice of Endocrinology and Metabolism

CHAPTER 33 THYROID FUNCTION TESTS

ROBERT C. SMALLRIDGE Iodothyronines Protein Bound Concentrations Free Thyroxine and Triiodothyronine by Equilibrium Dialysis Free Thyroxine: Two-Step and Analog Methods Serum Binding Proteins Triiodothyronine Uptake and Free Thyroxine Index Thyroxine/Thyroxine-Binding Globulin Ratio Effects of Drugs Thyrotropin Thyroglobulin and Thyroglobulin Antibodies Thyroid Peroxidase Antibodies Thyroid Hormone Antibodies Thyrotropin Receptor Antibodies Peripheral Indices of Thyroid Function Blood Tests Other Tests Evaluation of the Patient Chapter References

The determination of circulating levels of thyroid hormones is essential for an accurate assessment of the functional status of patients. By means of specific competitive immunoassays (radioimmunoassays [RIAs], enzyme or chemiluminescent immunoassays [EIAs; CLIAs], or immunoradiometric assays [IRMAs]), small quantities of these and other hormones can be detected, enabling physicians to diagnose even mild degrees of thyroid dysfunction. RIA methodology involves a reaction between a specific antibody and two antigens, one of which is radiolabeled (see Chap. 237). In addition, various other laboratory procedures are available for examining thyroid disorders. Numerous factors may affect the function of the hypothalamicpituitarythyroid axis. These include physiologic factors (age, pregnancy, stress, temperature, genetics), pathologic factors (hyper- or hypothyroidism, systemic illness, surgery, starvation), and pharmacologic factors (Table 33-1). Some of these conditions produce a bewildering array of alterations in serum concentrations of these hormones, and a carefully selected set of tests may be required to establish the correct diagnosis. These tests are discussed in this chapter.

TABLE 33-1. Effects of Some Drugs on Thyroid Hormone Tests

Preanalytical factors affecting thyroid function tests include not only the ones just mentioned, but also variables directly affecting the specimen (e.g., hemolysis, bilirubin, free fatty acids).1 Assay performance is also affected by many analytical characteristics (e.g., bias, specificity, sensitivity and working range, presence of interfering substances, imprecision, recovery, calibration, and parallelism).2

IODOTHYRONINES
PROTEIN BOUND CONCENTRATIONS Serum thyroxine (T4) concentration was, for many years, the most useful first-line test of thyroid function. Historically, T4 was estimated by measuring the serum protein bound iodine or the butanol extractable iodine; these measures are no longer used. (In suspected iodide ingestion, the protein bound iodine is much greater than the T4.) Serum T4 is determined almost exclusively by RIA; values in euthyroid patients range between 5 and 12 g/dL. In hyperthyroidism, higher levels are present, and in hypothyroidism, lower levels are seen. In general, the level of serum triiodothyronine (T3), the most active thyroid hormone, varies in parallel with the T4 level. Healthy subjects have values of 80 to 200 ng/dL, with higher levels occurring in hyperthyroidism. Indeed, the T3 levels may increase despite a normal serum T4 level, a condition known as T3 toxicosis. Whereas T4 is usually low in hypothyroidism, serum T3 is often in the normal range, possibly owing to increased conversion of T4 to T3. Also, the peripheral conversion of T4 to T3 is exquisitely sensitive to acute stresses of many causes. Thus, a low serum T3 level is to be expected in many nonthyroidal illnesses3, whereas serum T4 levels may be low, normal, or elevated. Serum reverse T3 (rT3) ranges from 10 to 60 ng/dL in euthyroid patients. Like T4, serum rT3 level is high in hyperthyroidism and low in hypothyroidism. In nonthyroidal illness, serum rT3 concentrations are often elevated,4 whereas T4 levels are normal or low, and T3 levels are reduced. The discrepant results for T3 and rT3 in such disorders have made rT3 measurements clinically useful in differentiating primary hypothyroidism from acute illness. Although RIAs have been developed for the diiodothyronines,4 monoiodothyronines,5,6 measurement of and T4 and T3 sulfates7,8, these iodothyronine metabolites has no current clinical utility. Several of the iodothyronines have been measured in the urine,9 but their clinical applicability is limited. FREE THYROXINE AND TRIIODOTHYRONINE BY EQUILIBRIUM DIALYSIS To assess the patient's true metabolic status, estimation of the concentration of free T4 or free T3the active hormonesis advisable.The most precise determination is obtained by equilibrium dialysis.10 In this procedure, a diluted serum sample is incubated with T4 labeled with iodine-125 (125I) in a dialysis system. The free fraction of hormone (both the patient's T4 and the 125I-T4 tracer) achieves equilibrium by diffusing across a semipermeable membrane, whereas the protein bound hormone remains inside the dialysis tubing. The small percentage of dialyzable T4 (percentage of free T4) is determined directly, and the total T4 is measured by RIA. Multiplying the total T4 by the percentage of free T 4 yields the free T4 value. Free T3 is measured in an analogous manner. A reliable commercial assay for free T4 by equilibrium dialysis is available2,11 with a normal range of 0.8 to 2.7 ng/dL. Commercial assays for free T3 may be unreliable in patients with nonthyroidal illnesses or those taking

amiodarone.12,13 FREE THYROXINE: TWO-STEP AND ANALOG METHODS A variety of commercial RIA kits have become available for measuring free T4.11,14,15 and 16 A patient's serum is incubated with an anti-T4 antibody, and the T4 is allowed to reach a new equilibrium among endogenous serum binding proteins and the added antibody (Fig. 33-1). The amount of T4 extracted from the serum and bound to the anti-T4 antibody is quantified and related to known serum calibrators, which are similarly extracted. The antibody bound T4 can be quantified in several ways. Some assays require two steps and entail removal of serum and T4 binders. Another method necessitates only a single incubation and includes the addition of a T4 analogue to the incubation system. The two-step procedure for free T4 has proved comparable to equilibrium dialysis in evaluating many clinical disorders. In contrast, the analog method has yielded inaccurate results in situations in which there may be an aberrant increase in T4 binding to albumin or thyroxine-binding prealbumin (TBPA; transthyretin [TTR]). A comparison of the methods for estimating serum free T4 is given in Table 33-2.

FIGURE 33-1. Free thyroxine (T4) measurement by immunoassay. Antibody extraction assays involve the incubation of patient serum samples with anti-T4 antibody, during which time T4 approaches a new equilibrium with all the binders present. (TBG, thyroxine-binding globulin; TBPA, thyroxine-binding prealbumin.) (From Witherspoon LR, Shuler SE. Estimation of free thyroxine concentration: clinical methods and pitfalls. J Clin Immunoassay 1984; 7:192.)

TABLE 33-2. Methods for Estimating Serum Free Thyroxine in Thyroid Diseases

The free hormone hypothesis has been extensively reviewed, as have the various methods available for measurement.17,17a Several assays have been useful in measuring free T4 in patients with abnormal thyroid hormonebinding proteins or with non-thyroidal illness,11,14,15 although low values are reported in some assays. Measurement of free T4 by ultrafiltration, which avoids serum dilution, may be more accurate in certain situations.18,19

SERUM BINDING PROTEINS

Thyroid hormones are bound to three circulating proteins: thyroxine-binding globulin (TBG); TBPA, or TTR; and albumin. Routine serum T4 and T3 determinations measure the total amount of the hormones, of which more than 99% is protein bound. Many factors influence the levels of TBG and TBPA; thus, a total T4 or T3 level may deviate from euthyroid levels. Therefore, examining the status of these binding proteins is important. TRIIODOTHYRONINE UPTAKE AND FREE THYROXINE INDEX Several methods are used to evaluate binding proteins and to estimate the free T4 concentration. A commonly used test is the T3 uptake (T3U) test.10 In this procedure, a patient's serum sample is placed in a test tube to which is added a tracer amount of 125I-T3 and a resin or other solid phase matrix. During incubation, binding sites on the resin and on the patient's serum binding proteins compete for the 125I-T3 until equilibrium is achieved. When the serum has an increased number of unoccupied binding sites, a greater percentage of the 125I-T3 binds to these serum proteins. When the resin is separated and counted for radioactivity, the decreased percentage of 125I-T resin uptake indicates an increased number of unoccupied TBG-binding sites. In contrast, when the patient's TBG-binding sites are nearly saturated, the T U is 3 3 elevated. The free thyroxine index (FT4I) is calculated by multiplying the serum total T4 by the percentage of resin uptake. Several modifications have evolved in FT4I determinations. To permit comparisons of varying methods, a thyroid hormonebinding ratio (THBR) can be derived by dividing the patient's T3U by the mean of a reference population. Multiplying the THBR by the total T4 gives the FT4I. A second change has been the development of nonisotopic automated methods. Rather than using T3, T4 is labeled (enzyme or fluorescein). This assay method is called a T-uptake.20 Table 33-3 illustrates the directional changes observed in the more common derangements of thyroid function. Whenever total T4 and T3U values diverge and the FT4I is normal, binding protein concentrations of TBG are probably abnormal. In these settings, an excellent correlation is found between the FT4I and the free T4 determined by equilibrium dialysis. Unfortunately, numerous conditions may alter the binding proteins to produce a spurious FT4I estimate of the free T4.

TABLE 33-3. Diagnostic Utility of the Free Thyroxine Index Values

A free T3 index, calculated as a product of the total T3 and the T3U ratio, provides a reasonable estimate of serum free T3. It may aid in the diagnosis of T3 toxicosis when the serum T4 is normal21 (see Chap. 42). THYROXINE/THYROXINE-BINDING GLOBULIN RATIO A second method of examining thyroid hormonebinding proteins is to measure TBG by RIA and calculate a T4/TBG ratio. TBG can be measured by sensitive immunoelectrophoresis techniques.22 This method correlates well with the usual clinical state. The FT4I has been calculated using five nonisotopic T-uptake assays, the T3U method, and TBG.20 All methods performed comparably in patients with hyperor hypothyroidism or non-thyroidal illnesses. However, all were affected by sera containing increased concentrations of TBG (see also ref. 22a). EFFECTS OF DRUGS Numerous drugs affect the measurement of total and free thyroid hormones.23 Many of these changes result from altered binding of T4 and T3 to their carrier proteins. Other drugs may act on the hypothalamus, the pituitary, or the thyroid gland. The patient's thyroid status depends on the number and severity of effects evoked. The myriad changes induced by commonly used drugs are summarized in Table 33-1.

THYROTROPIN
Serum thyrotropin (thyroid-stimulating hormone, TSH) has been a reliable indicator of primary hypothyroidism, with levels rising even when thyroid deficiency is mild and the T4 level is still normal (see Chap. 15 and Chap. 45). Distinguishing euthyroid patients from those with mild hyperthyroidism based on a suppressed TSH level has been encumbered by inadequate assay sensitivity. Although RIAs were developed that could detect TSH concentrations of 0.1 to 0.3 U/mL, this sensitivity was achieved by extensive purification of the radioligand and concentration of specimens. Commercially available RIAs have not provided quantitative values below 1 U/mL. Many second-generation assays detect TSH in the range of 0.1 to 0.5 U/mL, and third-generation assays have an even 10-fold greater functional sensitivity24,25 and 26 (Fig. 33-2). The first-generation RIA relies on a radiolabeled antigen and a single polyclonal antibody. The newer IRMAs use two or three antibodies (one or two are radiolabeled). These monoclonal antibodies are directed at the TSH subunits or at some specific epitopes on the TSH molecule (Fig. 33-3). This provides a more stable radioligand and a higher assay sensitivity. EIAs for TSH also use several monoclonal antibodies and have similar performance characteristics. They rely on colorimetry, however, and are affected by lipemic and hemolyzed samples.

FIGURE 33-2. Schematic representation of the generation system of nomenclature designation for thyroid-stimulating hormone (TSH) assays based on interassay precision [percent coefficient of variation (CV)]. Each generation represents a ten-fold improvement in functional sensitivity (20% interassay CV value). The black bars denote the 95% confidence limits of measurement at different TSH levels. (From Nicoloff JT, Spencer CA. The use and misuse of the sensitive thyrotropin assays. J Clin Endocrinol Metab 1990; 71:553.)

FIGURE 33-3. Principles of immunoradiometric assay. In step 1, a radio-labeled monoclonal antibody reacts with a specific antigen. In step 2, a second monoclonal antibody reacts with a different unique antigenic binding site. The second antibody is bound to a solid phase carrier, which is then washed to remove unbound radiolabeled antibody from the reaction. The complex bound to the solid matrix is then counted in a gamma spectrometer, and the concentration of antigen is determined by comparing the results to a simultaneously constructed standard curve.

These assays usually show undetectable TSH levels in hyperthyroid patients (Fig. 33-4) and in patients taking sufficient doses of T4 to suppress the TSH response to thyrotropinreleasing hormone. Thus, a suppressed basal TSH determination may be sufficient to diagnose hyperthyroidism in more than 95% of cases. This obviates the need for thyrotropin-releasing hormone stimulation tests in the diagnosis of mild hyperthyroidism or determination of the adequacy of suppressive thyroid hormone therapy.

FIGURE 33-4. Comparative evaluation of sera from the same group of normal (open symbols) and hyperthyroid (solid symbols) patients measured by different commercially available thyroid-stimulating hormone (TSH) immunometric assays, ordered (left to right) relative to their functional sensitivities (20% interassay). The normal reference ranges (mean 3 standard deviations of log values) are shown by the open box, and the analytic (intraassay) sensitivity limits are depicted by the shaded area. Assay 1 exhibits third-generation functional performance, whereas assays 2 through 9 exhibit second-generation performance. Assays 10 through 12, although cited by their manufacturers as being sensitive TSH assays, only meet first-generation performance criteria and do not qualify as sensitive TSH assays. (From Nicoloff JT, Spencer CA. The use and misuse of the sensitive thyrotropin assays. J Clin Endocrinol Metab 1990; 71:553.)

Assays may vary in their characteristics, and standard criteria for performance have been proposed.26 For clinical use, the recommendation is that the lower limit of the assay be determined by its functional sensitivity; that is, the concentration at which the interassay precision CV = 20%.26 This level is higher than the analytical sensitivity, but provides greater assurance that the TSH value reported is a real number. Although sensitivity is the most critical factor involved in selection of a TSH assay, specificity may be influenced by epitope differences in the TSH antibodies, and interference in the assay by heterophile antibodies.26,27

THYROGLOBULIN AND THYROGLOBULIN ANTIBODIES

Serum thyroglobulin (Tg) measurements are useful in the follow-up of patients with thyroid cancer28,29 (see Chap. 40). Numerous studies have shown that very low or undetectable Tg levels are found in patients who have been successfully treated and have no evidence of residual thyroid tissue. The performance characteristics of the assay used are important. Five problems complicating current Tg assay methods have been evaluated for six different Tg assays.29 First, standardization has been lacking, although a Tg reference standard has now been adopted. Second, functional sensitivity in some assays is suboptimal: Small amounts of residual thyroid tissue may go undetected. Third, poor precision of repeated interassay measurements, over a relevant clinical time frame of 6 to 12 months, may limit the ability to detect cancer recurrences or progression. Fourth, the hook effect may lead to a report of falsely normal or low Tg values in sera with very high Tg concentrations, especially in an immunoradiometric assay (IRMA). Samples should be run undiluted and at a 1:10 dilution to avoid this potentially serious mistake. Fifth, Tg antibodies can overestimate the Tg level in double-antibody RIAs, whereas IRMAs underestimate the true value. Sera should be screened for the presence of Tg antibodies, which should be reported to the clinician.29

THYROID PEROXIDASE ANTIBODIES

Patients with autoimmune thyroid disease commonly develop antibodies to thyroid antigens. Thyroid microsomal antibodies frequently are present in patients with chronic lymphocytic thyroiditis, and Tg antibodies develop in a few patients. Passive hemagglutination tests have often been used clinically. In such tests, sheep erythrocytes are coated with either thyroid microsomal proteins or Tg. These cells are exposed to serial dilutions of serum, and the presence of antibodies is indicated by observing hemagglutination in the test tube. These antibodies may also be detected by several other techniques, including immunofluorescence, complement fixation, EIA, IRMA, and RIA (see Chap. 46). Antithyroid peroxidase antibody measurements are more sensitive and specific for autoimmune thyroid disease30 and are replacing microsomal antibody determinations.

THYROID HORMONE ANTIBODIES

Occasional patients have autoantibodies to T4, T3, or both.27,31,32 Depending on the method used to separate bound from free hormone in the RIA, these patients may have spuriously high or low T4 or T3 values (Fig. 33-5). These antibodies are usually immunoglobulin G and can be identified by a variety of techniques, including affinity chromatography, immunoprecipitation, immunoelectrophoresis, and RIA. The presence of such antibodies should be considered when serum T4 or T3 measurements conflict with the clinical presentation.

FIGURE 33-5. Schematic representations of how antithyroxine (anti-T4) autoantibodies in patient's serum interfere with liquid-phase (single- or double-antibody methods) and solid-phase radioimmunoassays (RIAs). Thus, the bound 125I-T4 would be increased in single-antibody methods, leading to a falsely low value, whereas the liquid-phase RIA using second antibodies or the solid-phase RIA would yield a falsely high value. Identical situations occur in RIAs for triiodothyronine. Open circles represent 125I-labeled T4; solid circles represent cold T4. (PEG, polyethylene glycol.) (From Sakata S, Nakamura S, Miura K. Autoantibodies against thyroid hormones or iodothyronine. Ann Intern Med 1985; 103:579.)

THYROTROPIN RECEPTOR ANTIBODIES

Certain thyroid disorders have been associated with an immunologic abnormality. Early bioassays, such as the long-acting thyroid stimulator and its protector assay, demonstrated that the immunoglobulin of patients with Graves disease stimulates thyroid hormone secretion. Several functional types of antibodies are now recognized to exist (see Chap. 42). Although these antibodies appear to interact with thyroid follicular cell receptors, their effects are varied. Some antibodies promote thyroid gland function (thyroid-stimulating antibodies), some inhibit the binding of TSH to its receptor (thyroid-binding inhibitory immunoglobulins), and some enhance or inhibit thyroid growth. These antibodies are measured by a variety of bioassays and receptor assays.33,34 and 35 Patients may present with hyperthyroidism, hypothyroidism, goiter, or thyroid atrophy. Pregnant women with high titers of these antibodies may give birth to infants with transient hyperthyroidism or hypothyroidism.36 PERIPHERAL INDICES OF THYROID FUNCTION Serum thyroid hormone measurements are essential for making the diagnosis of thyroid dysfunction. Usually, altered levels of T4 and T3 correlate well with the patient's clinical presentation, and the diagnosis of hyperthyroidism or hypothyroidism is readily made. In some situations, however, the patient's metabolic status is not apparent from the thyroid blood tests. Such occurrences are expected in euthyroid hyperthyroxinemia37 and nonthyroidal illnesses3 (see Chap. 36). For instance, patients with thyroid hormone resistance have high serum T4 and T3 levels yet may appear euthyroid. The recommendation is that many target organs be evaluated clinically because tissue responses may be in the hyperthyroid, euthyroid, or hypothyroid range owing to heterogeneity of tissue resistance.38,39 Patients with

nonthyroidal illness have low serum T3 levels and sometimes low T4 levels as well. No consensus yet exists about whether these patients are completely euthyroid or whether their tissues are hypothyroid, perhaps as a protective response to conserve precious energy stores. A variety of tests demonstrate the effects of thyroid hormone on peripheral tissues (Table 33-4). Occasionally, these tests reflect a relatively specific direct effect of thyroid hormone on an organ system. Examples are increased production of specific proteins, enhanced enzyme activity, and augmented cardiac contractility. Other tests are affected because of a change in the metabolism of the particular substance being analyzed, such as an elevated serum creatine phosphokinase (CPK) level in hypothyroidism.

TABLE 33-4. Peripheral Indices of Thyroid Function

Although tests reflecting peripheral action of thyroid hormone are not indicated in the evaluation of most patients, occasionally one or more of these tests may be informative. BLOOD TESTS The influence of thyroid hormone status on several enzymes in the erythrocyte has been extensively studied. Red blood cell Na+/K+ adenosine triphosphatase is decreased in hyperthyroidism, and consequently intracellular sodium is increased; opposite changes occur in hypothyroidism.40 Glucose-6-phosphate dehydrogenase activity is increased in half of hyperthyroid patients and is modestly reduced in hypothyroidism.41 Transketolase and carbonic anhydrase activities are low in hyperthyroidism, and the latter is elevated in hypothyroid patients.42 CPK levels may be increased, sometimes to extremely high values, in myxedema. Isoenzyme studies have shown this CPK to be of skeletal muscle origin,43 and the elevation probably results from impaired clearance. The low levels in hyperthyroidism are the result of an increased disappearance rate. Plasma tyrosine levels, basally and in response to oral tyrosine ingestion, are elevated in hyperthyroidism and decreased in hypothyroidism.44 Cholesterol levels are high in hypothyroid patients and low in hyperthyroid patients. Thyroid dysfunction influences numerous components of the lipoproteins, whether high-density, low-density, or very-low-density lipoproteins. Additionally, changes occur in the phospholipids, apoproteins, and hepatic lipase activities.45,46 Sex hormonebinding globulin, a hepatic protein, is increased in hyperthyroidism and decreased in hypothyroidism. In patients with thyroid hormone resistance, sex hormonebinding globulin may be inappropriately normal,38 suggesting that the liver shares the tissue resistance. Ferritin, an iron-storage protein synthesized by erythrocyte precursors and reticuloendothelial cells, is increased in hyperthyroidism.47 Similar results have been reported for plasma cyclic adenosine monophosphate.48 Angiotensin-converting enzyme activity49 varies directly with the thyroid status, whereas plasma levels of several other endothelium-derived proteins (fibronectin and factor VIIIrelated antigen) are increased in hyperthyroidism.50 Increased serum osteocalcin (bone Gla protein), and urinary pyridinoline cross-link excretion, markers of bone metabolism, are observed with elevated thyroid hormone levels.51 Soluble interleukin-2 receptor is increased in hyperthyroidism and decreases with antithyroid drug therapy.52 OTHER TESTS Measurement of the basal metabolic rate (BMR) takes advantage of the correlation between energy expenditure at rest and the metabolic response to thyroid hormone. When BMR is measured under carefully controlled conditions, low values are compatible with the diagnosis of hypothyroidism and high values with hyperthyroidism. The normal range is 20% to +5%. Test conditions must be rigorously controlled because numerous pathophysiologic conditions and drugs can influence the BMR test. A more contemporary test, the basal oxygen uptake, correlates directly with thyroid status.53 Cardiac contractility is influenced by thyroid hormone. The clinician has two noninvasive measures for examining a patient's cardiac status. One test, the QKd interval, measures the time from the onset of the Q wave of the electrocardiogram to the Korotkoff sound recorded over the brachial artery. Euthyroid patients have a value of 205 12 milliseconds. This interval is shortened in hyperthyroidism and lengthened in hypothyroidism.54 Unfortunately, this test is not available in many hospitals. An alternative estimate of contractility can be obtained using systolic time intervals determined by echocardiography. This procedure measures the preejection period and the left ventricular ejection time; an inverse relation is seen between thyroid function and the preejection period/left ventricular ejection time ratio.55 The Achilles tendon reflex is a measure of the muscular activity of the deep tendon reflexes and may be quantified with the aid of a photoelectric cell and electrocardiogram. A prolonged relaxation time, although not pathognomonic, is a characteristic physical finding in ~60% of patients with myxedema. The instruments are less reliable in documenting the rapid reflexes that occur in hyperthyroidism.56

EVALUATION OF THE PATIENT

The clinician has available a large number of tests for evaluating a patient's thyroid status.57 In many instances, a serum free T4 estimate and a serum TSH level provide sufficient information for adequate diagnosis and management. However, a number of conditions (e.g., drug use, acute illness, dysproteinemias, autoimmunity, resistance disorders) require additional laboratory procedures for confirmation. Fortunately, a broad spectrum of ancillary tests are available to assist the physician. Some authors have proposed TSH as the first-line test. One group has recommended a TSH test only, with free T4 measured when TSH is <;0.1 or >5.0 mIU/L,58 whereas another group59 has shown an economic savings using a TSH algorithm, with a diagnostic failure rate of 2.7%. CHAPTER REFERENCES
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. Keffer JH. Preanalytical considerations in testing thyroid function. Clin Chem 1996; 42:125. Nelson JC, Wilcox RB. Analytical performance of free and total thyroxine assays. Clin Chem 1996; 42:146. De Groot LJ. Dangerous dogmas in medicine: the nonthyroidal illness syndrome. J Clin Endocrinol Metab 1999; 84:151. Engler D, Burger AG. The deiodination of the iodothyronines and of their derivatives in man. Endocr Rev 1984; 5:151. Smallridge RC, Wartofsky L, Green BJ, et al. 3'-L-Monoiodothyronine: development of a radioimmunoassay and demonstration of in vivo conversion from 3',5'-diiodothyronine. J Clin Endocrinol Metab 1979; 48:32. Corcoran JM, Eastman CJ. Radioimmunoassay of 3-L-monoiodothyronine: application in normal human physiology and thyroid disease. J Clin Endocrinol Metab 1983; 57:66. Chopra IJ, Santini F, Hurd RE, Chua Teco GN. A radioimmunoassay for measurement of thyroxine sulfate. J Clin Endocrinol Metab 1993; 76:145. Chopra IJ, Wu S-Y, Chua Teco GN, Santini F. A radioimmunoassay for measurement of 3,5,3'-triiodothyronine sulfate: studies in thyroidal and non-thyroidal diseases, pregnancy, and neonatal life. J Clin Endocrinol Metab 1992; 75:189. Faber J, Busch-Sorensen M, Rogowski P, et al. Urinary excretion of free and conjugated 3',5'-diiodothyronine and 3,3'-diiodothyronine. J Clin Endocrinol Metab 1981; 53:587. Witherspoon LR, Shuler SE. Estimation of free thyroxine concentration: clinical methods and pitfalls. J Clin Immunoassay 1984; 7:192. Nelson JC, Weiss RM, Wilcox RB. Underestimates of serum free thyroxine (T4) concentrations by free T 4 immunoassays. J Clin Endocrinol Metab 1994; 79:76. Sapin R, Schlienger J-L, Kaltenbach G, et al. Determination of free triiodothyronine by six different methods in patients with non-thyroidal illness and in patients treated with amiodarone. Ann Clin Biochem 1995; 32:314. Piketty M-L, d'Herbomez M, Le Guillouzic D, et al. Clinical comparison of three labeled-antibody immunoassays of free triiodothyronine. Clin Chem 1996; 42:933. Wong TK, Pekary AE, Hoo GS, et al. Comparison of methods for measuring free thyroxine in nonthyroidal illness. Clin Chem 1992; 38:720. Docter R, van Toor H, Krenning EP, et al. Free thyroxine assessed with three assays in sera of patients with nonthyroidal illness and of subjects with abnormal concentrations of

thyroxine-binding proteins. Clin Chem 1993; 39:1668. 16. Van Blerk M, Smitz J, Rozenski E, et al. Four radioisotopic immunoassays of free thyroxine compared. Ann Clin Biochem 1996; 33:335. 17. Ekins R. Measurement of free hormones in blood. Endocr Rev 1990; 11:5. 17a. Wang R, Nelson JC, Weiss RM, Wilcox RB. Accuracy of free thyroxine measurements across natural ranges of thyroxine binding to serum proteins. Thyroid 2000; 10:31. 18. Faber J, Waetjen I, Siersbaek-Nielsen K. Free thyroxine measured in undiluted serum by dialysis and ultrafiltration: effects of non-thyroidal illness, and an acute load of salicylate or heparin. Clin Chim Acta 1993; 223:159. 19. Surks MI, DeFesi CR. Normal serum free thyroid hormone concentrations in patients treated with phenytoin or carbamazepine: a paradox resolved. JAMA 1996; 275:1495. 20. Faix JD, Rosen HN, Velazquez FR. Indirect estimation of thyroid hormone-binding proteins to calculate free thyroxine index: comparison of nonisotopic methods that use labeled thyroxine (T-Uptake). Clin Chem 1995; 41:41. 21. Sawin T, Chopra D, Albano J, Azizi F. The free triiodothyronine (T3) index. Ann Intern Med 1978; 88:474. 22. Copping S, Byfield PGH. Thyroxine-binding globulin deficiency reexamined. Clin Endocrinol 1988; 28:45. 22a. Schussler GC. The thyroxine-binding proteins. Thyroid 2000; 10:141. 23. Surks MI, Sievert R. Drugs and thyroid function. N Engl J Med 1995; 333:1688. 24. Nicoloff JT, Spencer CA. The use and misuse of the sensitive thyrotropin assays. J Clin Endocrinol Metab 1990; 71:553. 25. Franklyn JA, Black EG, Betteridge J, Sheppard MC. Comparison of second and third generation methods for measurement of serum thyrotropin in patients with overt hyperthyroidism, patients receiving thyroxine therapy, and those with nonthyroidal illness. J Clin Endocrinol Metab 1994; 78:1368. 26. Spencer CA, Takeuchi M, Kazarosyan M. Current status and performance goals for serum thyrotropin (TSH) assays. Clin Chem 1996; 42:140. 27. Despres N, Grant AM. Antibody interference in thyroid assays: a potential for clinical misinformation. Clin Chem 1998; 44:440. 28. Ozata M, Suzuki S, Miamoto T, et al. Serum thyroglobulin in the follow-up of patients with treated differentiated thyroid cancer. J Clin Endocrinol Metab 1994; 79:98. 29. Spencer CA, Takeuchi M, Kazarosyan M. Current status and performance goals for serum thyroglobulin assays. Clin Chem 1996; 42:164. 30. McLachlan SM, Rapoport B. The molecular biology of thyroid peroxidase: cloning, expression and role as autoantigen in autoimmune thyroid disease. Endocr Rev 1992; 13:192. 31. Sakata S, Nakamura S, Miura K. Autoantibodies against thyroid hormones and iodothyronine. Ann Intern Med 1985; 103:579. 32. Sakata S. Autoimmunity against thyroid hormones. Crit Rev Immunol 1994; 14:157. 33. Vitti P, Elisei R, Tonacchera M, et al. Detection of thyroid-stimulating antibody using Chinese hamster ovary cells transfected with cloned human thyrotropin receptor. J Clin Endocrinol Metab 1993; 76:499. 34. Weetman AP, McGregor AM. Autoimmune thyroid disease: further developments in our understanding. Endocr Rev 1994; 15:788. 35. Costagliola S, Morgenthaler NG, Hoermann R, et al. Second generation assay for thyrotropin receptor antibodies has superior diagnostic sensitivity for Graves' disease. J Clin Endocrinol Metab 1999; 84:90. 36. McKenzie JM, Zakarija M. Fetal and neonatal hyperthyroidism and hypothyroidism due to maternal TSH receptor antibodies. Thyroid 1992; 2:155. 37. Borst GC, Eil C, Burman KD. Euthyroid hyperthyroxinemia. Ann Intern Med 1983; 98:366. 38. Smallridge RC, Parker RA, Wiggs EA, et al. Thyroid hormone resistance in a large kindred: physiologic, biochemical, pharmacologic, and neuropsychologic studies. Am J Med 1989; 86:289. 39. Refetoff S, Weiss RE, Usala SJ. The syndromes of resistance to thyroid hormone. Endocr Rev 1993; 14:348. 40. Dasmahapatra A, Cohen MP, Grossman SD, Lasker N. Erythrocyte sodium/potassium adenosine triphosphatase in thyroid disease and nonthyroidal illness. J Clin Endocrinol Metab 1985; 61:110. 41. Viherkoski M, Lamberg BA. The glucose-6-phosphate dehydrogenase activity (G-6-PD) of the red blood cells in hyperthyroidism and hypothyroidism. Scand J Clin Lab Invest 1970; 25:137. 42. Kiso Y, Yoshida K, Kaise K, et al. Erythrocyte carbonic anhydrase-I concentrations in patients with Graves' disease and subacute thyroiditis reflect integrated thyroid hormone levels over the previous few months. J Clin Endocrinol Metab 1991; 72:515. 43. Docherty I, Harrop JS, Hine KR, et al. Myoglobin concentration, creatine kinase activity, and creatine kinase B subunit concentrations in serum during thyroid disease. Clin Chem 1984; 30:42. 44. Belanger R, Chandramohan N, Misbin R, Rivlin RS. Tyrosine and glutamic acid in plasma and urine of patients with altered thyroid function. Metabolism 1972; 21:855. 45. de Bruin TWA, van Barlingen H, van Linde-Sibenius Trip M, et al. Lipoprotein (a) and apolipoprotein B plasma concentrations in hypothyroid, euthyroid, and hyperthyroid subjects. J Clin Endocrinol Metab 1993; 76:121. 46. Tan KCB, Shiu SWM, Kung AWC. Effect of thyroid dysfunction on high-density lipoprotein subfraction metabolism: roles of hepatic lipase and cholesteryl ester transfer protein. J Clin Endocrinol Metab 1998; 83:2921. 47. Takamatsu J, Majima M, Miki K, et al. Serum ferritin as a marker of thyroid hormone action on peripheral tissues. J Clin Endocrinol Metab 1985; 61:672. 48. Peracchi M, Bamonti-Catena F, Lombardi L, et al. Plasma and urine cyclic nucleotide levels in patients with hyperthyroidism and hypothyroidism. J Endocrinol Invest 1983; 6:173. 49. Smallridge RC, Rogers J, Verma PS. Serum angiotensin-converting enzyme: alterations in hyperthyroidism, hypothyroidism, and subacute thyroiditis. JAMA 1983; 250:2489. 50. Graninger W, Pirich KR, Speiser W, et al. Effect of thyroid hormones on plasma protein concentrations in man. J Clin Endocrinol Metab 1986; 63:407. 51. Garnero P, Vassy V, Bertholin A, et al. Markers of bone turnover in hyperthyroidism and the effects of treatment. J Clin Endocrinol Metab 1994; 78:955. 52. Smallridge RC, Tsokos GC, Burman KD, et al. Soluble interleukin-2 receptor is a thyroid hormonedependent early-response marker in the treatment of thyrotoxicosis. Clin Diagn Lab Immunol 1997; 4:583. 53. Lim VS, Zavala DC, Flanigan MJ, Freeman RM. Basal oxygen uptake: a new technique for an old test. J Clin Endocrinol Metab 1986; 62:863. 54. Young RT, Van Herle AJ, Rodbard D. Improved diagnosis and management of hyperand hypothyroidism by timing the arterial sounds. J Clin Endocrinol Metab 1976; 42:330. 55. Tseng KH, Walfish PG, Persaud JA, Gilbert BW. Concurrent aortic and mitral valve echocardiography permits measurement of systolic time intervals as an index of peripheral tissue thyroid functional status. J Clin Endocrinol Metab 1989; 69:633. 56. Ballantyne GH, Croxson MS. The effect of exercise, thyroid status and insulin-induced hypoglycemia on the Achilles tendon reflex time in man. Eur J Appl Physiol 1981; 46:77. 57. Larsen PR, Alexander NM, Chopra IJ, et al. Revised nomenclature for tests of thyroid hormones and thyroid-related proteins in serum. J Clin Endocrinol Metab 1987; 64:1089. 58. Klee GG, Hay ID. Biochemical testing of thyroid function. Endocrinol Metab Clin North Am 1997; 26:763. 59. Davey RX, Clarke MI, Webster AR. Thyroid function testing based on assay of thyroid-stimulating hormone: assessing an algorithm's reliability. Med J Aust 1996; 164:329.

CHAPTER 34 THYROID UPTAKE AND IMAGING Principles and Practice of Endocrinology and Metabolism

CHAPTER 34 THYROID UPTAKE AND IMAGING

SALIL D. SARKAR AND DAVID V. BECKER Radiopharmaceuticals Iodine-131 Technetium-99M Pertechnetate Iodine-123 Fluorine-18 Fluorodeoxyglucose Thallium-201 Chloride, Technetium-99M Sestamibi, Technetium-99M Tetrofosmin Instrumentation Scintillation Scanner and Camera Thyroid Probe Applications, Techniques, and Interpretation of Radionuclide Imaging Applications of Imaging Techniques of Imaging Interpretation of Images Applications, Techniques, and Interpretation of Thyroid Uptake Applications of Uptake Measurements Techniques of Uptake Measurements Interpretation of Uptakes Pharmacologic Manipulations Triiodothyronine Suppression Test Perchlorate Discharge Test Thyroid-Stimulating Hormone Stimulation Chapter References

Thyroid radionuclide uptake and imaging play an important part in the clinical management of thyroid disease when used in conjunction with a physical examination and laboratory tests. The proper use of radionuclide techniques requires a basic knowledge of their underlying principles and of their clinical advantages and limitations. Major changes have occurred in the last two decades in the choice of radiopharmaceuticals for thyroid imaging. Iodine-131 (131I), used routinely in the past, is associated with a relatively high radiation absorbed dose to the thyroid. Although no evidence exists of any deleterious radiation effect,1 the use of 131I has been replaced by the use of technetium-99m (99m Tc) pertechnetate and iodine-123 (123I) for routine thyroid imaging. Both 99m Tc pertechnetate and 123I deliver a much lower radiation dose per Ci deposited, permitting the administration of larger radiotracer amounts that provide higher count rates and improved image resolution. However, 99m Tc pertechnetate is not optimsally suited for imaging of metastatic differentiated thyroid cancer. Cancer tissue functions relatively poorly compared with normal thyroid tissue, and 99m Tc pertechnetate is usually not accumulated by such tissue in sufficient amounts relative to background for external imaging. Iodine-123, with higher target/background activity ratios, allows imaging of metastases and is being used increasingly as an alternative to 131I, the traditional radiopharmaceutical for the evaluation of metastatic disease. Fluorine-18 fluorodeoxyglucose (18F-FDG) is the newest radiotracer to be used in evaluating thyroid cancer. It may help image metastases not detected with 131I and, therefore, appears to be complementary to the use of radioiodine. For thyroid uptake measurements, 131I continues to be used widely because it is inexpensive, the methodology is simple, and the radiation exposure from the small amount of radioactivity required for this test is low (Table 34-1).

TABLE 34-1. Thyroid Radiopharmaceuticals

Imaging instrumentation in nuclear medicine also has greatly improved. Rectilinear scanners equipped with detector heads that move back and forth across the organ to be imaged have been replaced by stationary scintillation cameras that afford higher count rates and superior image resolution.

RADIOPHARMACEUTICALS
IODINE-131 Iodine-131, produced in a nuclear reactor, is selectively localized by thyroid tissue and is metabolized in the same manner as stable iodine (127I). The biologic half-life of iodine is ~80 days in the normally functioning thyroid gland, and its physical half-life is 8 days. Because of the relatively long half-life and energetic b-emission, the thyroidal radiation absorbed dose is proportionately high.2,3 Therefore, in benign thyroid disease in which every attempt is made to keep the radiation dose as low as possible, the diagnostic use of 131I generally is limited to the measurement of thyroid uptake, a nonimaging test that requires only a small amount of radiotracer, ~0.005 mCi. In thyroid cancer, however, larger amounts (~2 mCi) of 131I are used to improve the ability to detect metastatic foci. Such amounts, however, are associated with a higher radiation dose and may cause stunning of residual thyroid tissue, so that the efficacy of a subsequent ablative 131I dose is diminished. If stunning is a concern, 123I may be used instead for the diagnostic workup (see later).4,5 and 6 Finally, and perhaps most important, 131I still has a key role in the treatment of hyperthyroidism and differentiated thyroid cancer by virtue of its selective accumulation by follicular cells and its high radiation dose to thyroid tissue (see Chap. 40 and Chap. 42). TECHNETIUM-99M PERTECHNETATE Tc pertechnetate is conveniently available from reactor-produced molybdenum-99 generators. 99mTc pertechnetate is trapped by the thyroid, but unlike 131I, it does not participate in hormonogenesis; therefore, its biologic half-life in the thyroid is short. The physical half-life also is short (6 hours), and no b-rays are emitted. Consequently, the radiation absorbed dose to the thyroid is low,7,7a permitting the use of this radiopharmaceutical for routine imaging of the thyroid gland. Furthermore, because of its short half-life, 99m Tc pertechnetate may be used in a lactating patient with only temporary interruption of breast-feeding.8 Maximum uptake of 99m Tc occurs early; therefore, imaging is performed 20 to 30 minutes after tracer administration. The thyroid uptake of 99m Tc pertechnetate is much lower than that of 123I, the alternative radiopharmaceutical for routine thyroid imaging (see later). However, the lower uptake of 99mTc pertechnetate is offset by higher count rates from larger administered radiotracer amounts. As alluded to earlier, 99mTc pertechnetate is suitable primarily for imaging of the in situ thyroid gland because the target/background uptake ratios do not permit imaging of metastases, which accumulate radiotracer poorly relative to normal thyroid tissue.
99m

IODINE-123 I, a cyclotron product, has biodistribution properties identical to those of 131I but a much shorter physical half-life (13 hours). Thus, high thyroid/background ratios are achieved at low radiation doses.2 Overall image quality and detection of hypoactive thyroid lesions with 123I are comparable to those with 99m Tc pertechnetate,9,10 although imaging time per view is longer with 123I because the administered activity is lower. 123I may be considered the ideal physiologic imaging agent for the
123

thyroid, but it is not used universally because of its high cost (roughly 10-fold greater than 99m Tc pertechnetate per patient) and more complex logistics. 123I also may be used for thyroid uptake measurements, particularly when uptake is combined with imaging, although, as noted earlier, 131I is more widely used for this purpose. In thyroid cancer, the diagnostic use of 123I as an alternative to 131I is being investigated. FLUORINE-18 FLUORODEOXYGLUCOSE Physiologic substrates, including glucose, can be labeled with a positron-emitting radionuclide such as fluorine-18. Malignant tumors have increased rates of glycolysis and glucose utilization and, consequently, are associated with increased uptake of 18F-FDG. 18F-FDG undergoes phosphorylation by hexokinase but does not undergo further metabolism. Its biodistribution can be imaged with specialized detectors, and the procedure is referred to as positron emission tomography (PET). Preliminary reports suggest that PET with 18F-FDG may be a valuable adjunct in the management of differentiated thyroid cancer.11,12,13 and 14 PET frequently detects metastases not visualized by routine radioiodine scanning and may be particularly helpful in evaluating patients with negative scans and positive serum thyroglobulin levels. To date, however, the use of this diagnostic modality has been limited because of the high cost of PET detectors. The introduction of lower-cost hybrid cameras capable of both standard gamma imaging and PET is expected to increase the use of 18F-FDG in thyroid cancer cases. THALLIUM-201 CHLORIDE, TECHNETIUM-99M SESTAMIBI, TECHNETIUM-99M TETROFOSMIN Thallium-201 chloride, 99mTc sestamibi, and 99m Tc tetrofosmin, widely used for myocardial perfusion studies, have been used for the evaluation of thyroid cancer. As nonspecific tumor markers, these radiotracers may help detect recurrent or metastatic thyroid cancer not visualized by radioiodine imaging.15,16 However, their routine use as adjuncts to whole body radioiodine imaging is still not widely accepted.

INSTRUMENTATION
SCINTILLATION SCANNER AND CAMERA The rectilinear scanner was the first nuclear imaging instrument devised.17 This highly focused detector views a small area of the organ at a time; it images successive areas by a transverse back-and-forth motion until the entire organ is imaged. Because the detector head points in only one direction, oblique and lateral views of the thyroid are difficult to obtain. Image resolution is relatively poor compared with that of contemporary scintillation cameras. The principal advantage of the rectilinear scanner is the ability to obtain a full-sized 1:1 image, which facilitates precise in vivo location of neck masses. However, this instrument is not readily available presently. The scintillation camera is the standard imaging instrument in nuclear medicine.18 The detector is capable of viewing either a small region of interest such as the thyroid, or a large area such as the abdomen or chest, with the use of appropriate collimators (see later). In addition, the detector can be easily rotated for lateral and oblique views. Incoming gamma rays interact with a crystal in the scintillation detector to produce flashes of light. The crystal is in contact with photomultiplier tubes that detect, amplify, and locate the light signals electronically. Scintillation crystals generally are 1/4 inch or 3/8 inch thick. The 1/4-inch crystal is suitable for only low energy radionuclides such as 99m Tc and 123I, whereas the thicker 3/8-inch crystal provides adequate interactions with higher energy gamma rays such as 131I. In addition, detectors are fitted with a collimator, generally a removable lead plate with many holes or a lead cone with a single small hole at the tip (pinhole collimator). Detector heads and fields of view are smaller for the portable cameras as compared to the standard wide-field-of-view cameras. Collimators with multiple holes that are parallel to each other are routinely used with cameras to view a large area of the body. The lead septa between the holes in a parallel hole collimator allow primarily those rays originating directly below each hole to enter through that hole, thereby narrowing the view of each photomultiplier tube and assisting it in locating the origin of the gamma radiation in the field of view. To prevent septal penetration of gamma rays, a high energy collimator, that is, one with thick septa, is used to image a radionuclide with high energy gamma emissions such as 131I. Conversely, a low energy collimator with thin septa is used for low energy gamma emissions as with 123I. Parallel hole collimators, with the wide-field-of-view camera, are the mainstay of clinical nuclear imaging and are particularly suitable for whole body imaging for thyroid cancer metastases. Pinhole collimators are used with scintillation cameras to provide a detailed picture of small areas and, therefore, are suitable for imaging such organs as the thyroid.19,20 This method may detect areas of localized variations in function within thyroid tissue that often are not palpable as discrete abnormalities. A disadvantage of the pinhole collimator is that it introduces spatial distortion, making it more difficult to correlate the image with physical landmarks. For the same reason, measurement of gland size using a calibrated marker next to the thyroid is unreliable. Pinhole collimator imaging of the thyroid is feasible using cameras with both small and large fields of view. THYROID PROBE The thyroid probe is a small, portable scintillation detector used to measure thyroid uptake. It consists of a scintillation crystal and a single photomultiplier tube housed in a cylindrical lead casing. This probe, placed at a distance from the neck, is capable of measuring or counting radioactivity.21 It does not produce an image of the organ.

APPLICATIONS, TECHNIQUES, AND INTERPRETATION OF RADIONUCLIDE IMAGING

APPLICATIONS OF IMAGING Imaging is of value in distinguishing Graves disease from toxic nodular goiter, assessing the function of nodules, differentiating subacute thyroiditis and related conditions from Graves disease and toxic nodular goiter, defining occult thyroid lesions, determining the location and size of functional thyroid tissue, and detecting metastases from differentiated thyroid cancer. Perhaps the most frequent application of thyroid imaging is in evaluating the function of palpable nodules. A nodule is a palpatory or pathologic finding and cannot be defined by thyroid imaging alone. The imaging of palpable nodules allows normally functioning tissue to be distinguished from hypofunctioning (cold) or hyperfunctioning (hot) lesions. Hyperfunctional nodules (Fig. 34-1) may be associated with clinical or subclinical hyperthyroidism and are unlikely to be cancer. Hypofunctional nodules may be associated with a higher incidence of malignancy (see Chap. 39 and Chap. 40).

FIGURE 34-1. Typical images in (A) Graves disease; (B) toxic multinodular goiter; (C) solitary hyperfunctioning nodule; (D) destructive thyroiditis.

Imaging with or without an uptake measurement may help differentiate hyperthyroidism due to thyroiditis (viral, postpartum, or amiodarone related), from hyperthyroidism due to Graves disease or nodular goiter (see Fig. 34-1). Occult or nonpalpable thyroid lesions may be a clinical consideration in some patients without known primary tumors and in patients exposed to therapeutic x-rays or

other radiation in infancy and childhood, who are at risk of developing a thyroid neoplasm later in life.22,23 Thyroid size generally is estimated by palpation or by ultra-sonography. Size also may be assessed scintigraphically, using a rectilinear scanner that provides a full-sized 1:1 image.24 Estimates of the size of the thyroid gland or thyroid nodule are usually used to calculate the amount of therapeutic radioiodine to be administered in hyperthyroidism and to assess the response to radioiodine or suppressive therapy. The use of ultrasonography to measure thyroid volume is discussed in Chapter 35. Imaging may help establish the presence of a nonpalpable thyroid gland, for instance in hypothyroid infants, and may help localize mediastinal, lingual, or other ectopic thyroid tissue25,26 (Fig. 34-2). Mediastinal imaging is useful to locate aberrant thyroid tissue that might manifest as a substernal mass on a chest radiograph, often with tracheal deviation.

FIGURE 34-2. Upper left, normally located thyroid gland in infant. Upper right, lingual thyroid in infant. Below, large mediastinal thyroid in elderly woman (markers at shoulders and suprasternal notch).

Whole body radioiodine imaging, together with serum thyroglobulin determinations, is used to identify, locate, and monitor differentiated thyroid cancer (see Chap. 40). TECHNIQUES OF IMAGING IMAGING OF THE THYROID GLAND The radiopharmaceuticals of choice are 123I and 99mTc pertechnetate. For mediastinal thyroid tissue, 123I imaging at 24 hours may have advantages over 99m Tc imaging because activity in the heart blood pool on the 99mTc image may preclude optimal visualization of overlying thyroid tissue. Imaging with 99m Tc, however, may suffice for most goiters with substernal extensions and, hence, can be used for the initial evaluation. 131I may be used in lieu of 123I, if the latter is not available. 131I offers the same advantage as 123I, and its more energetic gamma rays are less likely to be absorbed by the overlying sternum. has theoretical advantages over 99mTc for the evaluation of lingual thyroid tissue because of the absence of significant 123I accumulation in the salivary glands, which may interfere with interpretation (the salivary glands trap but do not bind 99m Tc or iodine, so only early imaging, as with 99mTc, shows accumulation of radioactivity). Nonetheless, 99m Tc has been used successfully in assessing ectopic thyroid tissue.26
123I

Imaging is performed 4 or 24 hours after the administration of 0.3 to 0.4 mCi of 123I orally, or 24 hours after the oral ingestion of 0.03 to 0.05 mCi of 131I, or 20 to 30 minutes after the intravenous injection of ~5 mCi of 99mTc pertechnetate. Standard imaging projections are anterior, left anterior oblique, and right anterior oblique. Prior exposure to stable iodine in iodinated contrast radiographic dyes and iodine-rich foods, or to thyroid hormone, may depress the thyroidal uptake of radioiodine or 99m Tc for 2 to 10 weeks, depending on the nature of the interfering substance and thyroid function.27,28 and 29 After prior exposure to stable iodine, it may be possible to obtain an adequate image in 7 to 10 days if the radiopharmaceutical amount is increased (e.g., 10 mCi of 99m Tc). The standard imaging instrument is the scintillation camera with a pinhole collimator. Single photon emission computed tomography with the camera and a parallel hole collimator may allow three-dimensional evaluation of the thyroid, but the clinical application of this technique is limited because of its complexity and cost.30 When hypothyroid infants are imaged, the stomach also should be imaged because there may be a trapping defect in the thyroid associated with a similar defect in the salivary glands and gastric mucosa.31 The imaging time per camera view is ~5 minutes for 99mTc pertechnetate and ~10 minutes for 123I, and varies with the thyroid uptake of radiotracer and administered amount of radiotracer. The location on the scintillation camera image of a palpable mass may be determined by placing a small radioactive source over the center of the mass or at several points around it. EXTENDED BODY IMAGING FOR DIFFERENTIATED THYROID CANCER For optimal imaging of metastatic differentiated thyroid carcinoma, patient preparation is crucial. All patients will have undergone surgical or radioiodine ablation of normal thyroid tissue to remove competing functioning normal tissue. High levels of thyroid-stimulating hormone (TSH) maximize radio-iodine uptake in thyroid tissues and facilitate their visualization.32,33 Sustained elevation of endogenous serum TSH can be achieved by one of two regimens. In the first regimen, all thyroid hormone preparations are withheld for 6 weeks before imaging; alternatively, triiodothyronine (T3) is substituted for thyroxine (T4) in patients receiving long-term T4 therapy 6 weeks in advance and is discontinued 2 to 3 weeks before imaging, thereby shortening the period of symptomatic hypothyroidism.32,34 Recombinant human TSH may be used in selected patients in lieu of prolonged thyroid hormone withdrawal.35,36 The prior use of iodinated contrast radiographic dyes, therapeutic iodide-containing preparations including amiodarone, vitamins, and kelp should be avoided. In addition, an iodine-poor diet instituted a week before the test may enhance radiotracer uptake by thyroid tissue.37,38 Foods rich in iodine include certain seafood (ocean fish, shrimp, lobster, clams, oysters, seaweed), milk, cured and spicy meats (ham, bacon, salami, pastrami), certain commercial food preparations (pizza, chili, potato chips, pretzels, nuts), and canned fruits and vegetables. Diuretics such as hydrochlorothiazide and furosemide have been used to help deplete body iodine stores before imaging. Although radioiodine uptake by tumor tissue may be increased by diuretics, the renal clearance of radioiodine is decreased, resulting in 131I retention and increased total body radiation dose.39 I is the traditional radiopharmaceutical for the imaging of differentiated thyroid cancer, although 123I has been proposed as an alternative. The administered amount of 131I usually is ~2 mCi. Larger amountsup to 10 mCihave been used in an attempt to visualize more metastatic sites.40 Their use, however, is controversial because subsequent uptake of therapeutic 131I may be impaired due to stunning of thyroid tissue.4,5 and 6 If 123I is used, ~2 mCi of the radiotracer may be administered. Imaging after 131I therapy, optimally at 7 to 10 days, often helps discover tumor foci not visualized on the pretherapy scan.41,42
131

Body imaging is best performed with a wide-field-of-view scintillation camera and a parallel hole collimator that is appropriate for the energy of the radionuclide used, that is, a high energy collimator for 131I and low energy collimator for 123I. If a significant amount of radioactivity is present in residual tissue in the thyroid bed, the thyroid bed should be covered with a lead shield during imaging, or the halo or flare from this radioactivity may obscure smaller amounts of abnormal uptake in adjacent areas. Despite this precaution, however, the presence of large thyroid remnants may well preclude optimal visualization of the lungs, necessitating additional evaluation with CT. Anterior and posterior views of the head/neck, thorax, abdomen/pelvis, and proximal femurs are obtained routinely. In addition, images of the anterior neck taken with a pinhole collimator facilitate the evaluation of thyroid remnants and cervical lymph node metastases. Quantitation of uptake in residual thyroid tissue may be used to determine the ablative 131I dose. Such uptakes have been traditionally measured with the thyroid probe but can be derived more accurately from the camera images with the use of suitable regions of interest.43 Quantitative techniques are being developed to allow the prediction of absorbed radiation dose actually delivered to the tumor.44 INTERPRETATION OF IMAGES The assessment of function in a thyroid nodule is generally straightforward. Occasionally, one encounters the entity known as discrepant nodule, a nodule that appears

to be functioning, or hot, on the early 99m Tc image but is hypoactive on the relatively delayed radioiodine image45,46 (Fig. 34-3). This discrepancy may be explained by the preservation within the nodule of the trapping mechanism, but the loss of organification. Discrepant nodules do not represent a specific histopathologic entity and comprise a variety of benign and malignant lesions.46

FIGURE 34-3. Left, 99m Tc image shows decreased uptake in a nodule in the right lobe and increased uptake in a nodule in the isthmus. Right, 131I image shows decreased uptake in the right lobe nodule, as with 99m Tc, and discrepant (decreased) uptake in the isthmic nodule. Both nodules were benign on histopathologic examination.

Subacute thyroiditis is characterized by sharply decreased radiotracer accumulation in the thyroid gland and, therefore, can be easily distinguished from hyperthyroidism due to overproduction of hormone (Graves/toxic nodular disease), which is associated with increased radiotracer uptake (see Fig. 34-1). Mediastinal thyroid tissue occasionally does not collect enough radioactivity to be adequately visible, although prior stimulation with exogenous TSH may improve the likelihood of imaging this tissue. If the pinhole collimator is not positioned directly over the mediastinal tissue, this tissue may appear to be above the suprasternal notch because of parallax error.47 Extension of the neck, as required for imaging, may move the thyroid cephalad so that partly substernal thyroids may appear to lie entirely above the suprasternal notch. Activity adjacent to the thyroid occasionally is seen in a pyramidal lobe or in the esophagus. The former often is observed in Graves disease, whereas the latter is due to radioactivity in swallowed saliva and should not be mistaken for substernal extension of the thyroid. Esophageal activity usually diminishes after the patient drinks water. Visualization of thyroglossal duct cysts is unusual in euthyroid patients, although far more common in thyroidectomized patients with high TSH levels (see later). Rarely, metastatic cervical nodes are detected in a patient with an intact thyroid gland.48 Whole body imaging with radioiodine for metastases (Fig. 34-4) usually shows concomitant uptake in the salivary glands and stomach, reflecting trapping of the radioiodine. In the presence of thyroid tissuebenign or malignantuptake may also be seen in the liver, reflecting accumulation and metabolism of labeled thyroid hormone. A hypertrophied thyroglossal duct remnant is often visualized, because of the hypothyroid state and elevated serum TSH levels. Contamination of clothing or skin by saliva and accumulation of radiotracer in bronchopulmonary secretions may mimic metastases.49,50 and 51 Uncommon sites of uptake worth noting include the thymus, gallbladder, sebaceous cyst, Meckel's diverticulum, breast cyst, and such neoplasms as gastric cancer, lung cancer, and Warthin tumor.49,51 With careful attention to patient preparation and methodology, metastases from differentiated thyroid carcinoma are detected with ~80% accuracy. Anaplastic and medullary thyroid cancers concentrate radioiodine poorly.

FIGURE 34-4. Whole-body images in metastatic differentiated thyroid cancer. Left, 131I image before therapy shows uptake in large thyroid remnant in the neck and no metastases. Middle, image after therapy dose of 131I shows uptake in thyroid remnant, and mediastinal/lung and abdominal metastases. Right, image with 18F-fluorodeoxyglucose shows mediastinal/lung and abdominal metastases.

APPLICATIONS, TECHNIQUES, AND INTERPRETATION OF THYROID UPTAKE

APPLICATIONS OF UPTAKE MEASUREMENTS A thyroid scan primarily provides information as to the location, configuration, and distribution of radioactivity, but does not provide a quantitative estimate of global thyroid function and therefore is not indicated for this purpose. The 24-hour radioiodine thyroid uptake test was an important part of the workup for thyroid dysfunction in the past; however, it is not as critical today because of the availability of sensitive techniques for measuring serum levels of thyroid hormone and TSH. Nevertheless, the uptake test, together with the thyroid scan, plays an important role in the diagnosis of hyperthyroidism due to destructive thyroiditis, which is characterized by very low uptake of radioiodine and 99m Tc pertechnetate. More importantly, the radioiodine uptake is also used in calculating the therapeutic amount of radioiodine to be administered in Graves disease and toxic nodular disease (see Chap. 42). Its occasional use in the T3 suppression and perchlorate discharge tests is described later. TECHNIQUES OF UPTAKE MEASUREMENTS The uptake test consists of the administration of ~0.005 mCi of 131I or 0.100 mCi of 123I orally and measurement of the percentage of the administered amount in the neck (thyroid) 24 hours later, using the thyroid probe.52 Methods using 99mTc are available but require a scintillation camera and are somewhat cumbersome.53 INTERPRETATION OF UPTAKES The normal 24-hour radioiodine uptake ranges from 10% to 35%; regional variations due to differences in iodine intake may occur.54,55 The uptake in hypothyroid patients tends to overlap with the low range of normal, but values in overtly hyperthyroid patients usually are distinctively high. Among hyperthyroid patients, those with Graves disease usually have higher 24-hour uptakes than those with autonomously functioning toxic nodules. Patients with hyperthyroidism due to thyroiditis (viral, postpartum, amiodarone/cytokine related) and those with struma ovarii typically have low thyroid uptakes. In contrast, some disorders, including Hashimoto (autoimmune) disease and iodine deficiency, may be associated with high uptake despite euthyroid or even hypothyroid status.27 As with thyroid imaging, the prior intake of exogenous iodide or thyroid hormone interferes with the thyroid uptake test. However, the results of the uptake test still may be of value. For instance, a low uptake in cases of suspected subacute thyroiditis may not be diagnostic, but a normal or high uptake despite the prior intake of iodine or thyroid hormone makes the diagnosis unlikely.

PHARMACOLOGIC MANIPULATIONS
Tests that use the thyroid scan or uptake, in conjunction with pharmacologic manipulation of the hypothalamicpituitarythyroid axis, may provide additional diagnostic

information in certain disorders. Three such tests involving the use of T3, potassium perchlorate, and TSH are discussed next. TRIIODOTHYRONINE SUPPRESSION TEST The T3 suppression test is based on the finding that the administration of T3 (75100 g per day for 810 days) suppresses the uptake of radioiodine in the normal thyroid gland by at least 50% but does not suppress uptake in Graves disease or toxic nodular goiter because the thyroid glands in the latter are autonomous and not dependent on pituitary TSH.56,57 The T3 suppression test, once popular for the diagnosis of borderline hyperthyroidism, is only occasionally used today in conjunction with imaging to determine thyroid nodular autonomy. Modifications of the standard T3 suppression test are necessary in some conditions. Occasional patients, particularly those with nodular thyroids, may exhibit autonomous increase in thyroid function while receiving long-term replacement therapy with T4.58 This may be confirmed by measuring the thyroid uptake without discontinuing T4 and without the addition of T3. Such uptake measurements normally should be no more than 1% to 2%. The adequacy of suppressive thyroid hormone therapy can be evaluated in the same manner, although the ultrasensitive TSH assays and the thyrotropin-releasing hormone stimulation test offer alternative methods (see Chap. 15). PERCHLORATE DISCHARGE TEST Impaired organic iodination within the thyroid in certain disorders results in increased unbound iodide. The administration of potassium perchlorate, a competitive inhibitor of iodide trapping, to patients with these disorders causes discharge of the unbound iodide. In healthy subjects, the thyroidal iodide content does not change significantly after perchlorate administration because most of the iodide is bound and cannot be discharged. A change in thyroidal iodide content after perchlorate administration can be inferred from a change in the thyroidal content of previously administered 131I. The test involves baseline measurement of thyroid uptake 2 hours after 131I administration. Then, 600 to 1000 mg of potassium perchlorate is given orally, followed 1 hour later by a second measurement of thyroid uptake. A 5% or greater decrease in the thyroid counts suggests impaired organic iodination.59 The test may be enhanced by use of carrier iodide.60 THYROID-STIMULATING HORMONE STIMULATION The intramuscular or subcutaneous injection of TSH increases the uptake of 99m Tc or radioiodine by the thyroid gland. Recombinant human TSH could potentially be used to visualize hypo-active retrosternal thyroid tissue, enhance radiotracer uptake in nodular goiters before 131I therapy, and distinguish primary from secondary hypothyroidism.61 It is also very useful in the management of differentiated thyroid cancer.35,35a,36 CHAPTER REFERENCES
1. 2. 3. 4. 5. 6. 7. Becker DV. Choice of therapy for Graves' hyperthyroidism. N Engl J Med 1984; 311:464. MIRD/Dose Estimate Report no. 5: summary of current radiation dose estimates to human from I-123, I-124, I-125, I-126, I-130, I-131 and I-132 as sodium iodide. J Nucl Med 1975; 16:857. Dunning DE, Schwartz G. Variability of human thyroid characteristics and estimates of dose from ingested I-131. Health Phys 1981; 40:661. Jeevanram RK, Shah DH, Sharma SM, et al. Influence of initial large dose on subsequent uptake of therapeutic radioiodine in thyroid cancer patients. Nucl Med Biol 1986; 13:277. Park HM, Park YH, Zhou XH. Detection of thyroid remnant/metastases without stunning: an ongoing dilemma. Thyroid 1997; 7:277. Kao CH, Yen TC. Stunning effects after a diagnostic dose of iodine-131. Nuklearmedizin 1998; 37:30. MIRD/Dose Estimate Report no. 8: summary of current radiation dose estimates to normal humans from Tc-99m as sodium pertechnetate. J Nucl Med 1976; 17:74.

7a.Sfakianakis GN, Ezuddin SH, Sanchez JE, et al. Pertechnetate scintigraphy in primary congenital hypothyroidism. J Nucl Med 1999; 40:799. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. Romney B, Nickoloff EL, Esser PD. Excretion of radioiodine in breast milk. J Nucl Med 1989; 30:124. Ryo UY, Vaidya P, Schneider AB, et al. Thyroid imaging agents: a comparison of I-123 and Tc-99m pertechnetate. Radiology 1983; 148:819. Arnold JE, Pinsky S. Comparison of Tc-99m and I-123 for thyroid imaging. J Nucl Med 1976; 17:361. Sisson JC, Ackerman RJ, Meyer MA, et al. Uptake of 18-fluoro-2-deoxy-D-glucose by thyroid cancer: implication for diagnosis and therapy. J Clin Endocrinol Metab 1993; 77:1090. Feine U, Lietzenmayer R, Hanke JP, et al. Fluorine-18-FDG and Iodine-131 iodide uptake in thyroid cancer. J Nucl Med 1996; 37:1468. Dietlein M, Scheidhauer K, Voth E, et al. Fluorine-18-fluorodeoxyglucose positron emission tomography and iodine-131 whole-body scintigraphy in the followup of differentiated thyroid cancer. Eur J Nucl Med 1997; 24:1342. Altenvoerde G, Lerch H, Kuwert T, et al. Positron emission tomography with F-18-deoxyglucose in patients with differentiated thyroid carcinoma, elevated thyroglobulin levels, and negative iodine scans. Langenbacks Archives of Surgery 1998; 383:160. Ugur O, Kostakoglu L, Caner B, et al. Comparison of 201 Tl, 99mTc-MIBI and 131 I imaging in the followup of patients with well-differentiated thyroid carcinoma. Nucl Med Commun 1996; 17:373. Gallowitsch HJ, Mikosch P, Kresnik E, et al. Thyroglobulin and low-dose Iodine-131 and Technetium-99m-Tetrofosmin whole-body scintigraphy in differentiated thyroid carcinoma. J Nucl Med 1998; 39:870. Brownell GL, Aronow S, Hine GJ. Radioisotope scanning. In: Hine GJ, ed. Instrumentation in nuclear medicine, vol 1. New York: Academic Press, 1967:381. McIntyre WJ, Saha GB, Go RT. Planar imaging with single-head large-field-of-view cameras: are they still the workhorse? Semin Nucl Med 1994; 24(1):11. Sarkar SD. In vivo thyroid studies. In: Gottschalk AA, Hoffer PB, Potchen EJ, eds. Diagnostic nuclear medicine, 2nd ed. Baltimore: Williams & Wilkins, 1988:756. Pinsky SM, Ryo UY. Technique and utility of thyroid scans. In: DeGroot L, Frohman LA, Kaplan EL, Refetoff S, eds. Radiation associated thyroid carcinoma. New York: Grune & Stratton, 1977:297. Hine GJ, Williams JB. Thyroid radioiodine uptake measurements. In: Hine GJ, ed. Instrumentation in nuclear medicine, vol 1. New York: Academic Press, 1967:327. Schneider AB, Ron E, Lubin J, et al. Dose-response relationships for radiation-induced thyroid cancer and thyroid nodules: evidence for the prolonged effects of radiation on the thyroid. J Clin Endocrinol Metab 1993; 77:362. Cicerio G, Frohman LA, Bekerman C, et al. Scintigraphic thyroid abnormalities after radiation: a controlled study with Tc-99m pertechnetate scanning. Ann Intern Med 1982; 97:55. Mandart G, Erbsmann F. Estimation of thyroid weight by scintigraphy. Int J Nucl Med Biol 1975; 2:185. Schoen EJ, Dos Remedios LV, Backstrom M. Heterogeneity of congenital primary hypothyroidism: the importance of thyroid scintigraphy. J Perinat Med 1987; 15:137. Verelst J, Chanoine JP, Delange F. Radionuclide imaging in primary permanent congenital hypothyroidism. Clin Nucl Med 1991; 16:652. Grayson RR. Factors which influence the radioactive iodine thyroidal uptake test. Am J Med 1960; 28:397. Krugman LC, Hershman JM, Chopra IJ, et al. Patterns of recovery of the hypothalamic-pituitary-thyroid axis in patients taken off chronic thyroid therapy. J Clin Endocrinol Metab 1975; 41:70. Laurie AJ, Lyon SG, Lasser EC. Contrast material iodides: potential effects on radioactive iodine thyroid uptake. J Nucl Med 1992; 33:237. Chen JJS, LaFrance ND, Allo MD, et al. Single photon emission computed tomography of the thyroid. J Clin Endocrinol Metab 1988; 66:1240. Lever EG, Medeiros-Neto GA, DeGroot LJ. Inherited disorders of thyroid metabolism. Endocr Rev 1983; 4:213. Beierwaltes WH. The treatment of thyroid carcinoma with radioactive iodine. Semin Nucl Med 1978; 3:79. Sarkar SD, Torres MA, Manalili E, et al. Iodine-131 effects on thyroid remnant function: influence of serum TSH levels. Radiology 1998; 209P:404. Schneider AB, Line BR, Goldman JM, Robbins J. Sequential serum thyroglobulin determinations, I-131 scans, and I-131 uptakes after triiodothyronine withdrawal in patients with thyroid cancer. J Clin Endocrinol Metab 1981; 53:1199. Ladenson PW, Braverman LE, Mazzaferri EL, et al. Comparison of administration of recombinant human thyrotropin with withdrawal of thyroid hormone for radioactive iodine scanning in patients with thyroid carcinoma. N Engl J Med 1997; 337:888.

35a.Haugen BR, Pacini F, Reiners C, et al. A comparison of recombinant human thyrotropin and thyroid hormone withdrawal for the detection of thyroid remnant or cancer. J Clin Endocrinol Metab 1999; 84:3877. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. Adler ML, Macapinlac HA, Robbins RJ. Radioiodine treatment of thyroid cancer with the aid of recombinant human thyrotropin. Endocr Pract 1998; 4:282. Lakshmanan M, Schaffer A, Robbins J, et al. A simplified low iodine diet in I-131 scanning and therapy of thyroid cancer. Clin Nucl Med 1988; 13:866. Maxon HR, Thomas SR, Boehringer A, et al. Low iodine diet in I-131 ablation of thyroid remnants. Clin Nucl Med 1983; 8:123. Maruca J, Santner S, Miller K, Santen J. Prolonged iodine clearance with a depletion regimen for thyroid carcinoma. J Nucl Med 1984; 25:1089. Halpern SE, Preisman R, Hagan PL. Scanning dose and the detection of thyroid metastases. J Nucl Med 1979; 20:1099. Sherman SI, Tielens ET, Sostre S, et al. Clinical utility of posttreatment radioiodine scans in the management of patients with thyroid carcinoma. J Clin Endocrinol Metab 1994; 78:629. Pacini F, Hippi F, Formica N, et al. Therapeutic doses of iodine-131 reveal undiagnosed metastases in thyroid cancer patients with detectable serum thyroglobulin levels. J Nucl Med 1987; 28:1888. Sarkar SD, Leveque FC, Palestro CJ, Afriyie MO. Comparison of the thyroid probe and scintillation camera for uptake measurements in thyroid remnants. J Nucl Med 1999; in press. J Nucl Med 1999; 40:208P. Maxon HR, Thomas SR, Samaratunga RC, et al. Dosimetric considerations in the radioiodine treatment of macrometastases and micrometastases from differentiated thyroid cancer. Thyroid 1997; 7:183. Demeester-Mirkine N, Van Sande J, Corvilain J, Dumont J. Benign thyroid nodule with normal iodide trap and defective organification. J Clin Endocrinol Metab 1975; 41:1169. Kusic Z, Becker DV, Sanger EL, et al. Comparison of technetium-99m and iodine-123 imaging of thyroid nodules: correlation with pathologic findings. J Nucl Med 1990; 31:393. McKitrick WL, Park HM, Kosegi JE. Parallax error in pinhole thyroid scintigraphy: a critical consideration in the evaluation of substernal goiters. J Nucl Med 1985; 26:418. Ryo UY, Stachura ME, Schneider AB, et al. Significance of extrathyroidal uptake of Tc-99m and I-123 in the thyroid scan. J Nucl Med 1981; 22:1039. Brucker-Davis F, Reynolds JC, Skarulis MC, et al. False-positive iodine-131 whole-body scans due to cholecystitis and sebaceous cyst. J Nucl Med 1996; 37:1690. Kappes RS, Sarkar SD, Har-El G, et al. Iodine-131 therapy of thyroid cancer: extensive contamination of the hospital room in a patient with tracheostomy. J Nucl Med 1994; 35:2053. Serafini A, Sfakianakis G, Michalakis G, et al. Breast cyst simulating metastases on Iodine-131 imaging in thyroid carcinoma. J Nucl Med 198; 39:1910. Becker DV, Charkes ND, Dworkin H, et al. Procedure guideline for thyroid uptake measurement. J Nucl Med 1996; 37:1266. Smith JJ, Croft BY, Brookeman VA, Teates CD. Estimation of 24-hour thyroid uptake of I-131 sodium iodide using a 5-minute uptake of technetium-99m pertechnetate. Clin Nucl Med 1990; 15:80. Robertson JS, Nolan NG, Wahner HW, McConahey WM. Thyroid radioiodine uptakes and scans in euthyroid patients. Mayo Clin Proc 1975; 50:79. Hooper PL, Turner JR, Conway MJ, Plymate SR. Thyroid uptake of I-123 in a normal population. Arch Intern Med 1980; 140:757. Burke G. The triiodothyronine suppression test. Am J Med 1967; 42:600. Blum M, Seltzer TF, Campbell CC, Burroughs VJ. Evaluation of euthyroid solitary autonomous nodule of the thyroid gland: importance of scintillation scanning and thyrotropin releasing hormone

58. 59. 60. 61.

testing. JAMA 1982; 247:1991. Dymling JF, Becker DV. Occurrence of hyperthyroidism in patients receiving thyroid hormone. J Clin Endocrinol Metab 1967; 27:1487. Baschieri L, Benedetti G, DeLuca F, Negri M. Evaluation and limitations of the perchlorate test in the study of thyroid function. J Clin Endocrinol Metab 1963; 23:786. Stewart RDH, Murray IPC. Effect of small doses of carrier iodide upon the organic binding of radioactive iodine by the human thyroid gland. J Clin Endocrinol Metab 1967; 27:500. Ramirez L, Braverman LE, White B, et al. Recombinant human thyrotropin is a potent stimulator of thyroid function in normal subjects. J Clin Endocrinol Metab 1997; 82:2836.

CHAPTER 35 THYROID SONOGRAPHY, COMPUTED TOMOGRAPHY, AND MAGNETIC RESONANCE IMAGING Principles and Practice of Endocrinology and Metabolism

CHAPTER 35 THYROID SONOGRAPHY, COMPUTED TOMOGRAPHY, AND MAGNETIC RESONANCE IMAGING

MANFRED BLUM Sonography (Echography) of the Thyroid Normal Thyroid Gland Thyroid Nodule Goiter Lymphadenopathy Sonography in the Patient with Known Thyroid Cancer Sonography in Conjunction with Needle Biopsy Sectional Thyroid Imaging Magnetic Resonance Imaging of the Thyroid Computed Tomography of the Thyroid Conclusion Chapter References

Thyroid goiters and nodules are common disorders, whereas thyroid malignancies are relatively rare. The main purpose of sonography (ultrasonography), computed tomography (CT), and magnetic resonance imaging (MRI) of the thyroid gland is to help to confirm the diagnosis and to assist in planning therapy. Imaging should not be used routinely in screening procedures because it is not efficient in establishing the likelihood of cancer in a thyroid nodule. Although percutaneous needle biopsy is the most specific diagnostic tool for this purpose, its limitations are evident. Decision analysis has indicated that no single test is best; therefore, several tests continue to be needed.1 Appropriate clinical diagnosis and management depend on a familiarity with these tests and their applicability. Technologic advances have provided several methods that image the thyroid gland and surrounding tissues. In an era of cost containment, the physician must choose diagnostic tests carefully, selecting only those that contribute meaningfully to management.2 The imaging findings complement information gained from the history and physical examination; these techniques do not supplant good clinical skills. Indeed, the results of these tests may be confusing and may lead to an erroneous diagnosis when taken out of clinical context; however, when appropriately used and interpreted, they are extremely useful. All imaging techniques represent anatomic mapping. The scintiscan examines the accumulation by thyroid tissues of iodine or its partial analog, pertechnetate, to provide a functional map of where iodine is taken up and metabolized (see Chap. 34). Ultrasonography, CT, and MRI, whose suggested use is shown in Table 35-1, provide topographic maps; none of these techniques differentiates benign from malignant lesions. Diagnosing histopathology is not possible with these methods, except by inference and by correlation with other data. Evidence of gross fixation of structures and loss of tissue planes may be demonstrated on an image, but invasion that is obvious at surgery may have escaped detection by the best of imaging studies.

TABLE 35-1. Suggested Clinical Use of Nonisotopic Thyroid Imaging

SONOGRAPHY (ECHOGRAPHY) OF THE THYROID

In ultrasonography, high-frequency sound waves enter the body and are transmitted or reflected by tissue interfaces to produce a photographic image of the internal structure. With current high-resolution equipment, nodules as small as 2 to 3 mm can be identified by using a signal with a frequency of 7.5 to 10 MHz that penetrates <5 cm. The deeper penetration that is achieved by lower-frequency sound waves seldom is necessary for the evaluation of thyroid structures.3,4,4a Patients are examined in the supine position with the neck hyperextended. For full imaging of the thyroid gland and appropriate landmarks, the neck must be surveyed in the sagittal, transverse, and oblique planes. The images are produced quickly and assembled in rapid sequence in real time, much as in a motion picture. Swallowing may facilitate identification of the esophagus; swallowing also elevates the thyroid so that its lower poles can be examined. A useful advance in thyroid ultrasonography is color flow Doppler imaging, which adds dynamic flow information to a static gray-scale image.4b Color-encoded signals differentiate a fluid-filled cystic space and vasculature, indicating the direction and the velocity of blood flow and the degree of vascularity.5 The assignment of color is arbitrary; arterial signals can be assigned the color red and the companion venous signals, blue, assuming that venous flow is parallel to, but in the opposite direction of, arterial flow. However, because vessels may be tortuous, portions of the same vessel may display in different colors depending on the direction of flow in relation to the transducer, even if the true direction of flow has not changed. The shade of a color is related roughly to flow velocity. The ultrasonography operator must be experienced and must be aware of the clinical question that has been posed by the clinician to provide an appropriate answer. Routine protocols for scanning are unsatisfactory. Ultrasonography is safe and relatively inexpensive. No ionizing radiation is present, and damage to tissues has not been reported. The procedure permits continuation of suppressive therapy with thyroid hormone. Use of contrast material and patient preparation are unnecessary. Ultrasonography is of limited value adjacent to the trachea, is hard to interpret in the upper jugular region, and is not useful just behind the air-filled trachea or substernally. NORMAL THYROID GLAND The thyroid gland (Fig. 35-1) is characterized by homogeneous echoes, which give the gland a uniform ground-glass appearance. The surrounding muscles are of lower echogenicity. Medial to the thyroid lobes is the air-filled trachea, which does not transmit ultrasonic waves and is, therefore, poorly imaged. A calcified tracheal ring is represented anteriorly by dense echoes. The tubular carotid artery is echo free. Lateral and anterior to the carotid is the jugular vein, which can be identified when it is distended during a Valsalva maneuver. Behind the thyroid and anteromedial to the longus colli muscle is the esophagus, which can be identified as the patient swallows. The neurovascular bundle containing the inferior thyroid artery and recurrent laryngeal nerve may sometimes be identified, and blood vessels as small as 1 to 2 mm can occasionally be seen on the surface of the thyroid.

FIGURE 35-1. Sonogram of normal thyroid gland showing a transverse section of the lower anterior neck. Both lobes of the thyroid gland (T) are seen; they have a uniform, ground-glass appearance. The isthmus (I) is located anterior to the echo-dense tracheal ring (TR); behind that are artifactual reverberations of the sound waves. The carotid artery (C) and jugular veins (J) are found laterally. The sternocleidomastoid muscles (SCM) are anterior, and the longus colli muscles (LC) are posterior. The esophagus (E) is anteromedial to the left longus colli muscle.

THYROID NODULE Thyroid nodules distort the uniform echo pattern of the normal gland. Seventy-five percent of nodules are of lower echogenicity and 15% are of higher echogenicity than normal thyroid tissue, but this characteristic has limited diagnostic value because most thyroid cancers are less echo-dense than normal tissue, and benign nodules may be more or less dense.4 The nodules may contain echo-dense deposits of calcium. Some nodules have a sonolucent rim called a halo (Fig. 35-2). Within small nodules, the echo texture tends to be uniform, but nodules larger than 2.5 cm usually have irregular, echo-free zones that represent cystic or hemorrhagic degeneration. These degenerative changes in a nodule may almost completely replace the solid structure, and careful examination in various planes is necessary to discern internal echoes that represent septa or small solid regions (Fig. 35-3). These complex cystic nodules are extremely common and must be differentiated from the rare true thyroid cyst, which is encountered in ~1 in 500 to 1000 nodules4 (Fig. 35-4). A thyroid cyst is globular, smooth walled, and without internal echoes.

FIGURE 35-2. Sonogram of solid thyroid nodule showing the right thyroid lobe (T) of a patient who had a cold nodule (N). The nodule is of slightly lower echogenicity than the rest of the lobe and is surrounded by a halo (arrows) that is almost sonolucent. The lesion was a follicular adenoma. Asterisks indicate the anterior border of the thyroid lobe. (TR, a portion of a tracheal ring.)

FIGURE 35-3. Sonogram of complex thyroid nodule. Sonograms in the transverse (TRV) and sagittal (SAG) planes of the right thyroid lobe (T) in a patient with a cold right thyroid nodule on radionuclide imaging. Previous aspiration yielded nondiagnostic debris. The sonograms show extensive degeneration (D) of the nodule. The lesion could easily be mistaken for a cyst in the transverse view. However, in the sagittal view, the solid component (S) is seen at the lower pole. This view was used as a guide to insert a needle into the solid component. The aspirate was consistent with a benign nodular thyroid. A second, low-echogenicity, 2 3 mm nonpalpable nodule higher in the right lobe in the sagittal view (arrows) suggests that this is a multinodular goiter. The increased acoustic shadowing behind the fluid-filled space in both views is caused by enhanced transmission of sound waves through fluid. The carotid artery (C), thyroid isthmus (I), and trachea (TR) are seen. The patient was given suppressive therapy after the fluid was aspirated. The solid component shrank, and the fluid did not reaccumulate. (From Blum M. Practical application of modern technology in thyroid evaluation. In: Van Middlesworth L, ed. The thyroid gland: practical clinical treatise. Chicago: Year Book Medical Publishers, 1986:58.)

FIGURE 35-4. Sonogram showing a cyst (C) in the thyroid lobe (T). Dense echoes behind the cyst (*) signify enhanced transmission of the sound waves through the fluid-filled structure. No internal echoes are seen within the cyst space; it is globular in shape and has smooth walls.

Ultrasonography cannot differentiate benign from malignant solitary nodules. However, a true simple cyst is almost invariably benign. Because almost 90% of all solitary nodules are benign, the discovery of a solitary nodule should not cause undue alarm. Nevertheless, ~10% of solitary nodules in an otherwise normal gland are malignant, and they can be solid or can have degenerative changes. Large deposits of calcium signify healing of an old degenerative change or hemorrhage and may be seen in benign or malignant nodules. Punctate calcifications frequently represent psammoma bodies in papillary cancer, whereas more dense calcification may represent medullary carcinoma. A low echo halo that is seen around certain nodules signifies a boundary, perhaps a capsule. Doppler technique may demonstrate blood vessels within this rim. A halo may be seen in either benign or malignant conditions.4 Improved technology has permitted the detection of nodules in the millimeter range (Fig. 35-5); although this is an enormous advance, it is also the source of a problem.3 As many as 20% of all people have nonpalpable micronodules of indeterminate significance.3,6 Indeed, some of these lesions may represent occult thyroid cancer, the incidence of which varies from a few percent in this country to as much as 20% in other parts of the world. Clinical experience has indicated that these lesions are of little biologic consequence in most patients, and their discovery during echography may cause needless concern and therapy. Yet some small carcinomas do metastasize and may ultimately cause death.7 Therefore, the finding of a minute tumor cannot be dismissed. Future investigation must determine the proper treatment for these patients, including the role of periodic ultrasonography to monitor changes in the size of the nodule and the appropriateness of suppressive therapy with thyroid hormone. Identification of a nonpalpable nodule may have greatest clinical value in a patient who underwent therapeutic irradiation in youth (see Fig. 35-5). Another clinical challenge raised by ultrasonography pertains to the discovery of nonpalpable nodularity of the thyroid gland when only a solitary nodule is palpable (Fig. 35-6B). This is a common occurrence in patients with a palpable solitary nodule and should be approached in the same way as a dominant nodule in a

patient with clinical multinodularity. By current consensus, fine-needle aspiration biopsy of the dominant nodule appears to be the most cost-effective approach when the nodule is cold on isotope scanning.

FIGURE 35-5. Sonogram of nonpalpable thyroid nodule. Imaging findings in an asymptomatic 46-year-old man who underwent radiation therapy for acne during his midteens. The thyroid gland was normal to palpation. An iodine-123 scintillation scan was normal in the anterior, left anterior oblique (LAO), and right anterior oblique (RAO) projections. Inset, 10-MHz thyroid sonogram showing a 2 4 mm hypoechoic nodule (arrow) in the right lobe of the thyroid gland (T). (SCM, sterno-cleidomastoid muscle; M, strap muscles; C, carotid artery.) (From Blum M. Practical application of modern technology in thyroid evaluation. In: Van Middlesworth L, ed. The thyroid gland: practical clinical treatise. Chicago: Year Book Medical Publishers, 1986:59.)

FIGURE 35-6. The use of periodic ultrasonography in a patient with a thyroid nodule, showing the evolution of a nodular goiter. A 50-year-old man had annual sonograms because of a small nodule in the left thyroid lobe. He was euthyroid. Antithyroglobulin and antithyroid peroxidase antibodies were not detected. The concentrations of thyroglobulin and calcitonin were not elevated. The patient had no history of exposure to therapeutic radiation, but he had a strong family history of goiter. Fine-needle aspiration biopsy showed colloid and unremarkable thyroid cells, consistent with an adenomatous goiter. The patient elected observation without intervention. A, An image in the sagittal plane from the sonogram shows a 1.7 1.5 1.2 cm solitary nodule in the posterior portion of the left thyroid lobe (arrow). Activity throughout the rest of the thyroid was uniform. The sonogram was repeated after a 7-month interval and was unchanged. However, another sonogram (B) obtained 1 year later showed several new nodules in the left lobe (arrows). The lower part of the original nodule now has an area of cystic degeneration (lower X). Only this nodule was clinically palpable. The rest of the lobe is sonographically heterogeneous. Suppressive therapy with L -thyroxine was advised, and continued observation is planned.

Ultrasonography can be used with great accuracy and objectivity to assess the size of the thyroid gland8 or of a nodule periodically during the course of therapy. Because patients are likely to change physicians over the years, an objective assessment of thyroid size may be invaluable to facilitate continuity of care. Comparison of serial ultrasonographic scans may lead to surgery if a nodule grows despite suppressive therapy. Figure 35-6 shows evolution of a nodular goiter. Resolution of a nodules is shown in Figure 35-7.

FIGURE 35-7. The use of periodic ultrasonography in a patient with a thyroid nodule showing resolution of the nodule. A 38-year-old euthyroid woman was found to have a small palpable nodule in the inferior portion of the right thyroid lobe. Iodine-123 scan showed that the nodule was cold and a needle biopsy showed evidence of adenomatous hyperplasia. Ultrasonography that was performed in the transverse plane (A) and in the sagittal plane (B) showed a 0.9 0.9 cm nodule in the lower portion of the right thyroid lobe (+, X, +). Suppressive therapy with L -thyroxine was given. Ultrasonography was repeated 5.5 months later and used multiple cuts through the thyroid region in the transverse and sagittal planes (C and D); a nodule was not evident, and none was palpated clinically. (L, thyroid lobe; T, trachea and associated artifacts; C, carotid artery; SM, strap muscles.)

GOITER Sonography can identify a goiter by showing an enlarged thyroid gland and heterogeneity of the echo pattern. These findings are not specific for any particular type of pathology. However, generalized low echogenicity may suggest Hashimoto thyroiditis9; subacute thyroiditis usually exhibits very low echogenicity.10 Examination of a goiter with Doppler technique may demonstrate enhanced blood flow, especially in Graves disease. In a uniform goiter, the sonogram can show a region whose echo pattern is different from that of the rest of the gland, suggesting a second lesion, especially if it is surrounded by a halo. The author has observed several patients with Hashimoto disease in whom such a focal lesion led to the demonstration by needle biopsy of a neoplasm (Fig. 35-8). Because ultrasonography can be used to measure the size of a goiter, it also may be useful in calculating the dose of iodine-131 for the treatment of hyperthyroidism.

FIGURE 35-8. Sonogram of a nodule in Hashimoto thyroiditis. Sonogram of a patient with an enlarged multinodular thyroid. The antithyroid microsomal antibody titer was 1:25,600, indicative of Hashimoto (chronic lymphocytic) thyroiditis. An isotope scintiscan showed generalized heterogeneity of activity. The sonogram shows an echo-dense focal area (M) surrounded by a halo (arrows) in the right thyroid lobe (T). A needle biopsy of this mass revealed follicular carcinoma, which was confirmed at surgery. The rest of the thyroid has a low-echogenicity irregular echo pattern that is typical of Hashimoto thyroiditis and is enlarged. (TR, trachea.)

LYMPHADENOPATHY Normal lymph nodes are rarely seen on sonograms. When nodes are slightly enlarged due to inflammation, they are thin and oval in shape and have a central echogenic hilus.11 In distinction, nodes involved with malignancy more commonly are rounded, having a widened cortex and a narrowed or absent hilus. Thus, the ratio of the longitudinal to transverse diameter (L/T ratio) has been reported as <1.5 in 62% of nodes containing metastases and as >2 in 79% of benign reactive nodes.12 However, no significant differences in size are seen between benign and malignant nodes. When color and spectral Doppler techniques are used, malignant nodes frequently demonstrate a high ratio of systolic to diastolic blood flow (resistive index), with a mean of 0.92, whereas benign reactive nodes tend to be characterized by an index of <0.6.13 Therefore, ultrasonography is useful in the diagnosis of patients who have lymphadenopathy (especially if they are suspected of having thyroid cancer, if they have a history of therapeutic irradiation in youth, or if they have had a thyroidectomy for cancer or an adenoma.) However, even in patients with thyroid cancer, enlarged benign nodes are more common than malignant ones. SONOGRAPHY IN THE PATIENT WITH KNOWN THYROID CANCER Sonography is helpful in the management of a patient with thyroid carcinoma.4 Periodic ultrasonographic studies may disclose recurrent, nonpalpable carcinoma in the thyroid bed or in lymph nodes (Fig. 35-9). Ultrasonography can be performed during suppressive therapy, so that the inconvenience and risks of hypothyroidism associated with radionuclide scintigraphy are avoided. In patients with cervical adenopathy due to thyroid cancer but with a nonpalpable primary lesion, ultrasonography may disclose a mass in the thyroid gland, even if the scintiscan is normal.

FIGURE 35-9. Routine follow-up of a patient with a history of thyroid cancer and a normal examination with ultrasonography. Magnetic resonance imaging (MRI), a more costly and cumbersome procedure, was used to confirm a suspicious finding, leading to surgery. This 28-year-old woman had a left hemithyroidectomy and isthmusectomy for a 1.2-cm encapsulated papillary carcinoma without adenopathy and was taking suppressive therapy with L -thyroxine. Thyroid-stimulating hormone was undetectable by third-generation assay, and the concentration of thyroglobulin was <1.5 ng/mL. The examination did not reveal a nodule. Annual sonography (A, in the transverse plane; B, the same with Doppler added; and C, longitudinal plane) showed a new nodule in the bed of the left lobe. B, with Doppler added, highlights the carotid artery (bright circle), whereas the image of the nodule remains unchanged. The patient was reluctant to undergo another operation because a nodule was not palpable by several physicians and the concentration of thyroglobulin remained low. Therefore, MRI of the neck was performed 1.5 months after the initial ultrasonography to see if the sonographic findings could be verified with another imaging modality. The MRI confirmed the masses. D, A T1-weighted MRI scan that shows a 1-cm nodule whose appearance is more intense than muscle. It remained markedly hyperintense on the prolonged repetition time (RT) sequences and short tau inversion images (a technique that is used to enhance contrast by suppressing the signal that is derived from fat), which is not usual for normal thyroid tissue. A total thyroidectomy was performed, revealing that the nodule was papillary carcinoma. (Arrowhead on MRI scan, cancerous nodule; SCM, sternocleidomastoid muscle; SM, strap muscles; T, trachea and associated artifacts; C, carotid artery; J, jugular vein; L, thyroid lobe; E, esophagus; SAM, scalenus anticus muscle; LCM, longus colli muscle.)

SONOGRAPHY IN CONJUNCTION WITH NEEDLE BIOPSY Fine-needle aspiration biopsy of thyroid nodules has become a major diagnostic tool. For most needle biopsies, prior ultrasonography is not needed. The procedure is safe and inexpensive, and its accuracy may be as high as 97% when an expert performs the puncture and an experienced cytologist interprets the specimen. However, the accuracy declines to 50% with inexperience.2 Other factors that detract from accuracy include a low yield in complex, degenerated nodules, when fluid and necrotic debris are aspirated rather than the cellular component. Particularly when the nodule is deep or in a goiter, sampling errors are common in nodules <1 cm or in nodules >2.5 cm, which usually have degenerated zones. In an attempt to reduce these sources of error, and for non-palpable nodules, ultrasonography-guided biopsy has been used.14,15 and 16 A special echographic transducer to guide the needle is available but is not essential. Rather, it is easier to locate the nodule or node freehand, applying the transducer from one angle and inserting the needle from another direction. Ultra-sonography-guided biopsy does not negate the importance of multiple punctures of a nodule or goiter to improve the diagnostic yield. For complex nodules, the solid component is identified and the needle inserted into that area (Fig. 35-10; see Fig. 35-3). Thereafter, the fluid component can be aspirated to collapse the cystic space if desired. Aspirating the fluid first and then sampling the solid component is less satisfactory because the remnant may be small and difficult to find.

FIGURE 35-10. The use of ultrasonography to help with a differential diagnosis in a patient who had a hemithyroidectomy for adenoma and a new clinical finding. This 24-year-old woman had a left thyroid lobectomy 2 years earlier because of a toxic autonomous thyroid nodule, which was reported to be a follicular adenoma. She was not given suppressive therapy and remained euthyroid. Enlargement of the right lobe was detected. Ultrasonography was performed to try to differentiate compensatory hypertrophy from a focal lesion. A, The sonogram showed a uniform-appearing right thyroid lobe (L) with a 3.1 2.2 1.4 cm nodule (N) that was mainly cystic (CY) and had a small solid component. B, A flow Doppler study showed a vascular halo at the periphery of the nodule. All of these findings suggested tumor and not hypertrophy. Iodine-123 scan did not show a hyperfunctioning area in the zone that was solid tissue. Fine-needle aspiration biopsy of a solid portion of the new nodule, using the sonogram as a guide, showed sheets of uniform follicular cells without colloid. Surgery disclosed a follicular adenoma.

SECTIONAL THYROID IMAGING

Sectional imaging, which includes CT and MRI, is used to depict the regional anatomy of the neck and the superior mediastinum, but it plays no role in the diagnosis of the patient with a solitary thyroid nodule or goiter.17,18 and 19 Nevertheless, when these tests are done for other indications, the asymptomatic and unsuspected thyroid lesions that are often revealed need to be managed. As is shown in Table 35-1, however, some circumstances exist for which these imaging techniques are the only means available for the proper clinical evaluation of patients. In selected instances, sectional imaging may be used to assess the extent of an unusually large or obstructive thyroid gland or nodule, a cancer (Fig. 35-11), or a substernal goiter (Fig. 35-12), especially before surgery. Furthermore, when the postsurgical clinical findings are confusing, sectional imaging can provide accurate and essential management information. A detailed knowledge of the regional anatomy and experience with the techniques are required for the optimal interpretation of the images. Consultation between the clinician and the radiologist is important in this situation, as with all other imaging methods. Although the early identification of a recurrence of a tumor using this technique is possible, sonography is more sensitive for small lesions, less costly, more convenient, and safer. The periodic repetition of sectional imaging as a screening tool is rarely warranted.17,19,20

FIGURE 35-11. Computed tomographic (CT) scan of neck containing extensive thyroid cancer. CT scan with contrast agent enhancement. Anaplastic thyroid cancer (CA) encased the right carotid artery and jugular vein (open arrows). The trachea (TR) is displaced and indented, the right wall of the esophagus (E) is involved, and muscle tissue planes are lost. (SCM, left sternocleidomastoid muscle; T, left thyroid lobe; J, left jugular vein; C, left carotid artery.) (From Blum M. Practical application of modern technology in thyroid evaluation. In: Van Middlesworth L, ed. The thyroid gland: practical clinical treatise. Chicago: Year Book Medical Publishers, 1986:61.)

FIGURE 35-12. The selective use of magnetic resonance imaging (MRI) to show the extent of a large goiter. This 40-year-old woman had a large goiter and clinical evidence of mild obstruction of the thoracic inlet only when she raised her arms and flexed her head forward. MRI of the neck and superior mediastinum was performed to evaluate for evidence of tracheal involvement and substernal extension. This image shows a large goiter (G) at the level of the thoracic inlet. The thyroid capsule was intact. Slight compression of the trachea (T) was present at the level of the clavicular heads (CH), but no tracheal invasion was seen. Other images (not shown here) demonstrated extension of the goiter to the level of the aortic arch. The thyroid surgeon alerted a thoracic surgeon to be available if the goiter could not be delivered through a low collar incision.

MAGNETIC RESONANCE IMAGING OF THE THYROID

MRI depicts the interactions that occur between the hydrogen atoms (protons) in different tissues in the patient's body and an externally applied magnetic field (radiowaves of a specific frequency). Different properties (relaxation times) of the hydrogen atoms, termed T1 and T2, can be selectively emphasized. Because the hydrogen atoms of different tissues have different T1 and T2 properties, a computer-assisted analysis of T1-weighted and T2-weighted signals is used to differentiate the thyroid gland from skeletal muscle, blood vessels, or regional lymph nodes. Normal thyroid tissue tends to be slightly more intense than muscle on a T1-weighted image, and tumors frequently appear even more intense. The intravenous administration of noniodinated contrast agents, such as gadolinium-DTPA (pentetic acid), may further enhance the characterization of tissue and organs, as may suppressing the signal that is derived from fat (short tau inversion, or STIR). The test is uncomfortable; claustrophobic reactions occur, and considerable noise is inherent to the technique. Generally, the patient's entire body must be inserted into a large cylinder. Advances such as the use of special surface electromagnets over the neck provide impressive images and, thus, may increase the usefulness of MRI for thyroid evaluation. The equipment is costly and in great competitive demand for other types of examinations. The guidelines for use were discussed earlier in Sectional Thyroid Imaging. Figure 35-9 shows the evaluation of a suspicious ultrasonographic finding, and Figure 35-12 demonstrates the use of MRI to show the extent of a large obstructive goiter and to help plan a surgical approach. Occasionally, in this setting, sectional imaging may help the physician to choose between medical management and surgery. In Figure 35-13, an MRI shows the size and extent of postsurgery cancer. Figure 35-14 demonstrates the use of MRI to examine enlarged upper cervical lymph nodes in a patient with thyroid cancer. Figure 35-15 shows an MRI that helped the physician to diagnose a complex clinical problem. Several studies have suggested new directions for future research, including correlation of images with the biochemistry and histopathology of tissue.21,22 Qualitative and quantitative similarities of the proton response of thyroid tissue in the neck and chest have been demonstrated,23 and differences in these characteristics in malignant and benign thyroid tissues have been studied in vitro.24

FIGURE 35-13. The use of magnetic resonance imaging (MRI) to show the extent of a residual thyroid cancer after surgery, as a guide to possible additional therapy. MRI of a 72-year-old man with a history of several myocardial infarctions, coronary artery surgery, and a right thyroid lobectomy because of an invasive multifocal Hrthle cell carcinoma of the thyroid. An attempt to remove the lesion resulted in the sacrifice of the right laryngeal nerve, but the tumor could not be completely excised. The left lobe (LL) was not removed for technical reasons. Iodine-131 ( 131I) whole body scan showed that the mass did not take up radioiodine in the presence of the left lobe. 131I therapy was rejected by the patient. The concentration of thyroglobulin has remained stable in the 10- to 17-ng/mL range for 2 years. Thyroid-stimulating hormone level has remained low, consistent with cardiovascular tolerance. Clinical examination revealed an ill-defined and deep right paratracheal mass that had not grossly changed. Sonography in the past had not clearly revealed the extent of the mass. MRI was performed to evaluate the size of the mass and possible progression, which would lead to radiation therapy. The MRI scan shows a right paratracheal mass (M) that is unchanged in size and extent from the previous MRI study. The arrow points to the region where the cancer partly encases the carotid artery (C), as before. The management was not changed. Note that on this T1-weighted image, the cancer is characteristically brighter than the left thyroid lobe (LL) or muscle tissue (SCM). (T, trachea; E, esophagus.)

FIGURE 35-14. Magnetic resonance imaging (MRI) scan of the neck that was made to obtain information about possible tumor involvement of lymph nodes that had been detected with a sonogram. Ten years previously, this 57-year-old woman had a right hemithyroidectomy and removal of the isthmus because of a 1.5-cm encapsulated noninvasive papillary carcinoma of the thyroid. Her serum thyroglobulin was not elevated, and a postoperative sonogram showed a normal-appearing left thyroid lobe without adenopathy. Suppressive therapy has been maintained, and the thyroid-stimulating hormone has been undetectable. The serum thyroglobulin has not risen (1.7 ng/mL). The current examination suggested left upper jugular adenopathy but no nodule in the thyroid region. No clinical evidence of an infection was present, but the patient had considerable disease of the gingiva and a flare of her psoriasis. Ultrasonography showed two new anterior cervical lymph nodes, each 1.5 cm in size. The left thyroid lobe was <1.5 cm in length and was uniform. The right thyroid bed showed no nodules or nodes. MRI of the neck was performed to better define the uniformity of the lymph nodes. This T1-weighted coronal image shows upper internal jugular chain lymph nodes (arrows) with a homogeneous display of signal pattern on all pulse sequences. The nodes are not brighter than muscle (SCM), which is more compatible with hyperplastic nodes than with cancer. Six months later, clinical examination and sonography failed to disclose the nodes. Clinical follow-up is in progress. (P, parotid gland.)

FIGURE 35-15. Magnetic resonance image (MRI) of neck at thoracic inlet showing a thyroid tumor in the neck and primary cancer of the lung rather than a single primary lesion and metastases. The patient had a large left thyroid tumor that extended substernally. Clinically, the lesion was continuous across the superior mediastinum with a mass that extended into the right superior hemithorax. Computed tomography could not exclude a connection between the two sides of the mass. This MRI shows two separate and distinct masses. (T, thyroid mass; L, mass of higher signal intensity in the anterior segment of the right upper lobe of the lung that extends through the chest wallan adenocarcinoma of the lung.) The difference in the signal characteristics of the two masses connotes tissue differences and suggests two tumors. (Courtesy of Dr. Deborah L. Reede.)

COMPUTED TOMOGRAPHY OF THE THYROID

The characterization of tissue and organs by CT is accomplished by computer-assisted analysis of the attenuation of x-rays that are transmitted through the patient. For some purposes, the images may be examined directly because the thyroid gland is somewhat more radiopaque than the rest of the soft-tissue structures of the neck due to its high iodine content. More often, the contrast must be enhanced by the intravenous administration of iodinated material. Unfortunately, however, the iodine in the dye may be counterproductive to further diagnostic testing and to the treatment of thyroid disorders. Therefore, in most centers, CT has assumed a role in the diagnosis and management of thyroid problems that is complementary to that of MRI.14,17 CT should be used selectively and only in response to a specific clinical question that cannot be answered in a more cost-effective way. CT is more sensitive than MRI in detecting small metastases to cervical or mediastinal lymph nodes (stage I, 1.5 cm in diameter),25 is more reliable than MRI in the detection of small pulmonary nodules,26 is more appropriate for the unstable or claustrophobic patient, and is the only study possible in those patients with cardiac pacemakers or other biomechanical devices, who require sectional imaging.27 Pragmatically, access to CT is greater than access to MRI because of the larger number of CT scanners available and because the cost of a CT examination is lower than that of an MRI examination. Therefore, in some centers, CT is used in a manner similar to that described for MRI. An example of CT imaging for a patient with extensive thyroid cancer is shown in Figure 35-11.

CONCLUSION
No best method exists for evaluating thyroid masses. The history, physical examination, imaging techniques, and percutaneous biopsy all contribute to management.28 The diagnostic tests must be used selectively, with good judgment, and in a cost-effective fashion to answer specific questions that are posed by the clinical circumstance. The imaging methods provide information about gross anatomyin the case of ultrasonography, down to the millimeter range. Ultrasonography is the primary imaging method used to identify the solid component of a complex nodule for performing guided fine-needle aspiration, to determine the comparative size of nodules in patients who are under observation (especially when they are taking thyroid-stimulating hormonesuppressive therapy), to detect a small nodule in patients who were exposed to therapeutic irradiation of the head or neck, and to search for a recurrence of thyroid cancer after surgery.

CHAPTER REFERENCES
1. 2. 3. 4. Molitch ME, Beck JR, Dreisman M, et al. The cold thyroid nodule: an analysis of the diagnostic and therapeutic options. Endocr Rev 1984; 5:185. Van Herle AJ, Rich P, Ljung BME, et al. The thyroid nodule. Ann Intern Med 1982; 96:221. Leopold GR. Ultrasonography of superficially located structures. Radiol Clin North Am 1980; 18:161. Butch RJ, Simeone JF, Mueller PR. Thyroid and parathyroid ultrasonography. Radiol Clin North Am 1985; 23:57.

4a.Barraclough BM, Barraclough BH. Ultrasound of the thyroid and parathyroid glands. World J Surg 2000; 24:158. 4b.Hiromatsu Y, Ishibashi M, Miyaki I, et al. Color doppler ultrasonography in patients with subacute thyroiditis. Thyroid 1999; 9:1189. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. Foley WD. Color Doppler flow imaging. Boston: Andover Medical Publishers, 1991. Tan GH, Gharib H. Thyroid incidentalomas: management approaches to nonpalpable nodules discovered incidentally on thyroid imaging. Ann Intern Med 1997; 126:226. Boehm TM, Rothouse L, Wartofsky L. Occult follicular carcinoma of the thyroid with a solitary slowly growing metastasis. JAMA 1976; 235:2420. Hegedus L. Decreased thyroid gland volume in alcoholic cirrhosis of the liver. J Clin Endocrinol Metab 1984; 58:930. Marcocci C, Vitti P, Cetani F, et al. Thyroid ultrasonography helps to identify patients with diffuse lymphocytic thyroiditis who are prone to develop hypothyroidism. J Clin Endocrinol Metab 1991; 72:209. Blum M, Passalaqua AM, Sackler J, Pudiowski R. Thyroid echography of subacute thyroiditis. Radiology 1977; 124:795. Solbiati L, Rizzatto G, Bellotti E, et al. High resolution sonography of cervical lymph nodes in head and neck cancer: criteria for differentiation of reactive versus malignant nodes (abstract). Radiology 1988; 169(P):113. Vassallo P, Wernecke K, Roos N, Peters PE. Differentiation of benign from malignant superficial lymphadenopathy: the role of high-resolution US. Radiology 1992; 183:215. Choi M, Lee JW, Jang KJ. Distinction between benign and malignant causes of cervical, axillary, and inguinal adenopathy: value of Doppler spectral waveform analysis. AJR Am J Roentgenol 1995; 165:981. Blum M. Practical application of modern technology in thyroid evaluation. In: Van Middlesworth L, ed. The thyroid gland: practical clinical treatise. Chicago: Year Book Medical Publishers, 1986:47. Boland GW, Lee MJ, Mueller PR, et al. Efficacy of sonographically guided biopsy of thyroid masses and cervical lymph nodes. AJR Am J Roentgenol 1993; 161:1053. Gharib H, Goellner JR, Johnson DA. FNA cytology of the thyroid: a 12 year experience with 11,000 biopsies. Clin Lab Med 1995; 13:699. Blum M, Reede DL, Seltzer TF, et al. Computerized axial tomography in the diagnosis and management of thyroid and parathyroid disorders. Am J Med Sci 1984; 287:34. Blum M, Holliday RA, Yee JM. Non-isotopic imaging of the thyroid. In: Wagner HN Jr, ed. Nuclear medicine. Philadelphia: WB Saunders, 1995:595. Bahist B, Ellis K, Gold RP. Computed tomography of intrathoracic goiters. AJR Am J Roentgenol 1983; 140:455. Higgins CB, Auffermann W. MR imaging of thyroid and parathyroid glands: a review of current status. AJR Am J Roentgenol 1988; 151:1095. Charkes ND, Maurer AH, Siegel JA, et al. MR imaging in thyroid disorders: correlation of signal intensity with Graves disease activity. Radiology 1987; 164(2):491. Mountz JM, Glazer GM, Dmuchowski C, Sisson JC. MR imaging of the thyroid: comparison with scintigraphy in the normal and diseased gland. J Comput Assist Tomogr 1987; 11(4):612. Sandler MP, Putton JA, Sacks GA, et al. Evaluation of intrathoracic goiter with I-123 scintigraphy and nuclear magnetic resonance imaging. J Nucl Med 1984; 25:874. Tennvall J, Biorklund A, Moller T, et al. Studies of MRI relaxation times in malignant and normal tissues of the human thyroid gland. Prog Nucl Med 1984; 8:142. Yousem DM, Som PM, Hackney DB, et al. Central nodal necrosis and extra-capsular neoplastic spread in cervical lymph nodes: MR imaging versus CT. Radiology 1992; 182:753. Webb WR, Sostman HD. MR imaging of thoracic disease: clinical uses. Radiology 1992; 182:621. Shellock FG. MR imaging of metallic implants and materials: a compilation of the literature. AJR Am J Roentgenol 1988; 151:811. Ghaarib H. Fine-needle aspiration biopsy of thyroid nodules: advantages, limitations, and effects. Mayo Clin Proc 1994; 69:44.

CHAPTER 36 ABNORMAL THYROID FUNCTION TEST RESULTS IN EUTHYROID PERSONS Principles and Practice of Endocrinology and Metabolism

CHAPTER 36 ABNORMAL THYROID FUNCTION TEST RESULTS IN EUTHYROID PERSONS

HENRY B. BURCH Effects of Illness on Indices of Thyroid Function Changes in Individual Parameters of Thyroid Function Triiodothyronine and Reverse Triiodothyronine Total Thyroxine Free Thyroxine Thyroid-Stimulating Hormone Disease-Specific Variation Caloric Deprivation Renal Failure Hepatic Disorders Acquired Immunodeficiency Syndrome and Acquired Immunodeficiency SyndromeRelated Complex Psychiatric Disturbance Pathophysiology Sites of Nonthyroidal Illness Interaction with Thyroid Economy Differentiation From Primary Thyroid Dysfunction Nonthyroidal Illness Simulating Thyroid Dysfunction Nonthyroidal Illness Masking Thyroid Dysfunction Prognostic Implications and Role of Thyroid Hormone Therapy Euthyroid Hyperthyroxinemia Drugs Affecting Parameters of Thyroid Function Chapter References

EFFECTS OF ILLNESS ON INDICES OF THYROID FUNCTION

The frequency with which thyroid disease is encountered in the general population, the ease with which it is excluded in otherwise healthy persons, and the poor specificity of the individual signs and symptoms of thyroid dysfunction all have contributed to the practice of screening for thyroid disease in patients with nonthyroidal illness (NTI), in the search for a readily reversible component to the presenting illness. The resultant recognition of multiple abnormalities in parameters of thyroid function in patients who ultimately are deemed to be euthyroid has engendered the term euthyroid sick syndrome to describe these persons.1,2 Although early cross-sectional studies in such patients had revealed a baffling collection of low, normal, or high values for total and free thyroid hormones, subsequent work has suggested a continuum of change, within which a given patient's status is determined by the severity and duration of illness as well as the presence of mitigating influences that are associated with the specific underlying disorder. Hence, an early and dramatic reduction in serum total and free triiodothyronine (T3) levels occurs, followed, in cases of sufficient severity, by a depression in serum total thyroxine (T4) and variable changes in serum free thyroxine (FT4) levels. Concurrent with decrements in levels of circulating thyroid hormones, a seemingly inept response is seen at the hypothalamic-pituitary level; this persists until the recovery phase of the underlying illness, at which time a rise in thyroid-stimulating hormone (TSH) to even supranormal levels may be seen that coincides with a rapid recovery in T4 and T3 levels3 (Fig. 36-1).

FIGURE 36-1. Influence of the stage and the severity of nonthyroidal illness (NTI) on the parameters of thyroid function. Early in the course of illness, marked decreases in triiodothyronine (T3) are accompanied by reciprocal changes in reverse T3 (rT3). The free thyroxine (FT4) and thyroid-stimulating hormone (TSH) generally remain within normal limits but may transiently exceed this range, particularly during the recovery phase. The total thyroxine (T4), which generally is normal during mild to moderate NTI, falls progressively with increasing disease severity. Recovery is accompanied by gradual normalization of T4, followed by normalization of T3 and reverse T3, as well as of TSH. (Adapted from Nicoloff JT, LoPresti JS. Extrinsic and intrinsic variables: nonthyroidal illness. In: Braverman LE, Utiger RD, eds. Werner and Ingbar's the thyroid. A fundamental and clinical text, 7th ed. Philadelphia: JB Lippincott, 1996:289.)

CHANGES IN INDIVIDUAL PARAMETERS OF THYROID FUNCTION

TRIIODOTHYRONINE AND REVERSE TRIIODOTHYRONINE The most prevalent and pronounced anomaly in the parameters of thyroid function during NTI is a depression in the serum total and free T3 levels. This manifestation, which is present in 70% or more of hospitalized patients, may be considered, with a few notable exceptions (discussed later), the sine qua non of the euthyroid sick syndrome.4,5 The likelihood of finding a depression of serum T3 varies with the severity of the NTI; in one study, it occurred in 23% of patients on regular hospital wards, 56% of those in intensive care units, 76% of those undergoing coronary artery bypass, and 86% of those receiving heart transplants.5 The depression in serum T3 may be pronounced, achieving levels below those seen in hypothyroid persons.6,7 Concomitant reductions in thyroid hormonebinding proteins account for less than one-third of the observed decrease in serum T3 level4; much of the remainder is attributable to a decreased peripheral monodeiodination of serum T4 (discussed later). Serum free T3, as measured by equilibrium dialysis, also is decreased in as many as 50% of patients with NTI. Given the frequency and magnitude of the T3 depression during severe NTI, the finding of a normal or elevated T3 concentration in this setting occasionally is the first indication of a coexistent thyro-toxicosis.8 Serum reverse T3 (rT3), a metabolically inert metabolite of T4 obtained through 5-deiodination, has long been valued as a diagnostic tool in the distinction of sick euthyroid patients (elevated rT3) from patients with hypothyroidism (depressed rT3), although this is not always reliable.9,10 As with T4 and T3, serum rT3 is predominantly protein bound in the circulation. Beyond its role as the alternate metabolite of T4, rT3 has the ability to act as both substrate and inhibitor of type I and type II 5'-monodeiodinases, thereby potentially contributing to the decreased production of T3 seen in NTI.11 TOTAL THYROXINE Thyroxine levels enjoy a greater degree of stability than do T3 levels during NTI of mild or moderate severity. However, with increasing gravity of the underlying illness, the serum total T4 levels may be subnormal in as many as 20% to 50% of hospitalized patients.5,6 and 7,12,13 Extremely low serum T4 levels portend a fatal outcome to an extent that rivals traditional prognostic indices (see later). Conversely, a rise in T4 levels accompanies the recovery phase of illness in patients who survive critical illness (see Fig. 36-1). Hyperthyroxinemia may be seen occasionally in NTI, although this generally occurs in a predictable fashion in association with certain hepatic disorders or acute psychiatric illnesses, or in conjunction with specific medications (see later).

FREE THYROXINE In view of the described abnormalities in serum T4 and T3, an accurate assessment of circulating FT4 levels assumes a critical role in the metabolic evaluation of patients with NTI. Likewise, the measurement of serum FT4 is essential to the understanding of the apparently euthyroid status that is observed in these patients despite the profound changes in other parameters of thyroid function. Unfortunately, several features inherent to NTI serve to undermine attempts to measure and interpret serum FT4 levels accurately. These include postulated circulating inhibitors to T4 binding, the effects of commonly used drugs on hormone binding and on the hypothalamicpituitary thyroid axis, and the frequent finding of spuriously low serum FT4 values during NTI when these are assessed using FT4 indices or are measured by several of the indirect FT4 techniques in common use.14 Serum FT4 index calculations, which depend on the resin uptake of labeled T3 or T4, generally underestimate serum FT4 levels during NTI, perhaps as a result of the presence of circulating inhibitors to thyroid hormone binding. This artifact may be particularly pronounced in patients who have low total T4 levels.12 Serum FT4 assays that use synthetic T4 analogs also may yield falsely low FT4 levels because of undesired protein binding of the analog; this occurs despite attempts to design agents that minimize such effects.14 As a result, these methodologies are (paradoxically) dependent on concentrations of thyroid hormonebinding proteins as well as on serum dilution.15 When assessed by equilibrium dialysis, serum FT4 has been found to be normal or minimally elevated in most patients with NTI.7,14,16 However, the high cost and technical demands that are associated with equilibrium dialysis generally have limited the utility of this measurement in clinical practice. Measurement of serum FT4 by direct equilibrium dialysis using undiluted serum3,17 has revealed normal FT4 values in most patients with NTI.18 This method and similar techniques ultimately may serve to increase the feasibility and accuracy of serum FT4 measurement in patients with NTI. THYROID-STIMULATING HORMONE The response of serum TSH to the marked reduction in levels of free and total T3 has been viewed alternatively as either inappropriately blunted or indicative of normal intracellular thyroid hormone effects. A normal serum TSH level in the setting of NTI frequently is offered as evidence for euthyroidism. However, current experimental evidence suggests the presence of a complicity at the hypothalamic-pituitary level in the thyroidal alterations that occur in NTI. For example, an abrupt decline in serum TSH occurs in acutely ill or stressed patients, such as during fasting,19 after major surgery, 5 or after bone marrow transplantation.13 Furthermore, the subsequent rise in serum TSH that coincides with the normalization of serum T4 suggests an amelioration of a relative central hypothyroidism in these patients.3 Moreover, although the responsiveness of the pituitary thyrotrope to exogenous thyrotropin-releasing hormone (TRH) generally is maintained during NTI,16,20,21 an inability to augment this response appropriately has been demonstrated,22 and a relative blunting of the serum TSH peak occurs in critical illness23 and with acute caloric deprivation.24 The nocturnal surge that occurs in serum TSH levels in healthy persons also is diminished during NTI.20,25 Alterations in serum TSH levels in hospitalized patients with NTI have been examined using a second-generation sensitive TSH assay.26 Of 1580 hospitalized patients, 33% were found to have abnormal serum TSH values, and 15% had levels more than 3 standard deviations above or below the mean for control subjects. When a subgroup of these patients was followed longitudinally, 84% had NTI or recent glucocorticoid use as the sole explanation for their abnormal serum TSH value; this included 76% of patients who had an undetectable serum TSH and 50% of patients who had a serum TSH of more than 20 mIU/mL (Fig. 36-2). Interestingly, 86% of patients with NTI who had undetectable serum TSH levels by a second-generation assay (<0.1 mIU/mL) had detectable levels (>0.01 mIU/mL) using a third-generation TSH assay.27 These results have been duplicated elsewhere.28

FIGURE 36-2. Etiology of abnormal values in sensitive assays for thyroid-stimulating hormone (TSH) values in hospitalized patients. A cohort of 329 individuals selected from 1580 hospitalized patients and categorized on the basis of TSH level were reevaluated after recovery from nonthyroidal illness to determine the true thyroid status. The percentage of patients in each TSH category who were ultimately found to have thyroid disease (TD), nonthyroidal illness (NTI), or glucocorticoid use (GC) as the explanation for the baseline TSH is shown. Patients in each category included: TSH <0.1 mIU/mL (n = 45); 0.10.34 (n = 70); 0.350.59 (n = 54); 0.604.1 (n = 70); 4.26.8 (n = 33); 6.920.0 (n = 35); and >20 mIU/mL (n = 22). Most notable is the high percentage of patients with TSH <0.1 mIU/mL or >20 mIU/mL who were ultimately found to be free of thyroid dysfunction (76% and 50%, respectively). (Data derived from Spencer CA, Eigen A, Shen D, et al. Specificity of sensitive assays of thyrotropin used to screen for thyroid disease in hospitalized patients. Clin Chem 1987; 33:1391.)

Although cross-sectional analysis reveals a high frequency of abnormalities in thyroid function test results during NTI, changes beyond the reference range for serum TSH and FT4 when present, generally are small and may be best accounted for by a broadening of the range accepted as normal during NTI5,26 (as discussed later).

DISEASE-SPECIFIC VARIATION
A wide variety of severe illnesses, including poorly controlled diabetes mellitus, infection, malignancy, acute myocardial infarction, stroke, major surgery or injury, and malnutrition, may lead to characteristic patterns of altered thyroid hormone indices. This section reviews selected nonthyroidal disorders, with an emphasis on caloric deprivation as both a model and a common component of diverse forms of NTI. In addition, disorders in which the usual pattern of NTI has been modified because of the superimposition of effects unique to the underlying disease state are discussed. CALORIC DEPRIVATION The effect of prolonged fasting on the parameters of thyroid function is a well-studied model for NTI. With this model, the myriad of interfering influences that are present in many hospitalized ill patients are avoided. During acute caloric deprivation, serum total and free T3 levels begin to descend within the first 24 hours and reach a plateau within 1 to 2 weeks. The concentration of serum rT3 rises, and it plateaus at nearly double basal levels during this period.29 Serum T4 levels generally remain stable during fasting, and FT4 levels are either normal or slightly increased. Serum TSH levels decrease slightly during the initial fasting period and return to baseline within 96 hours.19 The pituitary response to TRH is blunted during a fast relative to the fed state, but generally remains in a range that is appropriate to the true thyroid state.24,30 Interestingly, the effect of fasting on thyroid indices is ameliorated when as little as 50 g of glucose or protein is given, but not when fat is the sole source of the caloric supplement.1,30 The concept that these changes are homeostatic in nature has been strengthened by the observation that the administration of T3 in this setting increases gluconeogenesis at the expense of muscle protein catabolism and thereby promotes depletion of lean body mass.31 RENAL FAILURE Patients with end-stage renal disease have a higher incidence of both goiter and primary hypothyroidism than do hospitalized control patients.32 Like patients with other serious NTIs, patients with chronic renal failure have depressions in serum T3 and T4 levels, normal FT4 levels, normal TSH levels, and elevated free rT3 levels.32,32a However, in contrast to other disorders, serum total rT3 levels generally are normal rather than elevated in renal failure. Multiple differences in rT3 handling account for this observation, including a normal rather than a decreased rT3 metabolic clearance rate, a decreased protein binding of rT3, an enhanced exit rate of rT3 from the serum into tissue, and enhanced tissue rT3 binding.32 Patients with the nephrotic syndrome have enhanced urinary T4 and T3 losses; much of this is accounted for by the concurrent loss of thyroid-binding globulin (TBG) in the urine. Despite this finding, these patients usually have changes in their thyroid indices similar to those of patients with chronic renal failure. Shortly after the initiation of dialysis, patients with renal failure may experience improvement in serum T4 and T3 levels; this may be

only temporary, because patients undergoing long-term dialysis generally have persistent decrements in these levels.1,32 HEPATIC DISORDERS The liver supports several key aspects of thyroid function, including the biosynthesis of each of the three principal binding proteins, the peripheral conversion of T4 to T3, and the provision of a reservoir for T 4 and T3. Therefore, the fact that disorders of hepatic function compound the defects in thyroid economy that typically are observed in NTI is not unexpected. Patients with hepatic cirrhosis generally have normal or decreased serum T4 values and depressed T3 levels. The decrease in serum T4, when present, does not appear to be a result of reductions in serum TBG levels (which generally are normal) but may be due in part to decreases in thyroxine-binding prealbumin and albumin concentrations, both of which are lower than normal in this population. Serum FT4 and free T3 levels usually are within the normal range in cirrhosis. TSH levels frequently are normal but occasionally may be increased in patients with cirrhosis.6,33 An elevation, rather than a depression, in T3 and T4 levels distinguishes patients with acute infectious hepatitis, chronic active hepatitis, and primary biliary cirrhosis from those with the pattern that is more typical of NTI.33,34 An increased synthesis and release of TBG occurs in these disorders, resulting in elevations of both T4 and T3. The serum FT4 generally remains in the normal range. Occasionally, the serum TSH may be elevated in patients with chronic active hepatitis or primary biliary cirrhosis, but this probably is related to the higher prevalence of Hashimoto thyroiditis and associated primary thyroid failure in patients with these conditions.34 ACQUIRED IMMUNODEFICIENCY SYNDROME AND ACQUIRED IMMUNODEFICIENCY SYNDROMERELATED COMPLEX Persons infected with the human immunodeficiency virus (HIV) have an atypical pattern of change in the indices of thyroid function that varies with the severity of illness.35,36 An unexplained rise in serum TBG levels occurs with HIV infection; this is associated with a rise in serum total T4 values, but normal FT4 values and normal rather than decreased T3 levels in ambulatory patients. With hospitalization or progressive illness, however, serum T3 levels fall.36 In addition, patients with acquired immunodeficiency syndrome (AIDS) may have unexpectedly normal or even low rT3 values relative to healthy control subjects.35,37,38 The serum TBG rises progressively in HIV infection as the disease advances from the asymptomatic carrier stage to AIDS-related complex and, finally, to AIDS. The TBG levels have been found to correlate inversely with the CD4 count in these persons.38 The serum TSH may be higher than in healthy control subjects, but remains within the normal range and responds with a normal or slightly exaggerated response to the injection of TRH.37 The paradoxically normal or even high serum T3 level in HIV-infected patients has been postulated to contribute to the extreme cachexia that is seen with this disorder, but no correlation has been found between weight loss and relative serum T3 elevation in these patients.36 Patients with advanced AIDS ultimately may manifest decreases in serum T3 and T4, which signify a preterminal state.35 PSYCHIATRIC DISTURBANCE Thyroid hormone levels have been examined extensively in patients hospitalized with acute psychiatric illness. In various studies, between 10% and 33% of such patients have elevated serum T4 levels at the time of admission.39,40 and 41 The serum FT4 index is elevated in 3% to 18% of such patients, and the FT4 value is either normal or supranormal.39 Serum T3 concentrations generally are normal39 or minimally elevated,41 and rT3 levels are increased. The serum TSH, as assessed using a sensitive assay, is elevated in as many as 17% of such cases, prompting the speculation that hyperthyroxinemia in these patients is driven centrally.41 However, little correlation exists between the serum total T4 or FT4 index and the TSH level in patients who experience hyperthyroxinemia during their psychiatric illness.40,41 TRH responsiveness was found to be blunted in more than one-half of the patients in one study,39 but in none of four patients in another report.41 The response to TRH in depressed patients has been found to be directly proportional to the basal serum TSH level and provides little additional diagnostic value.42 Rapid improvement in each of these parameters has been demonstrated repeatedly in most hospitalized psychiatric patients. Therefore, the utility of thyroid function testing in these patients at the time of hospital admission has been legitimately questioned. However, in most cases, with the combined use of a sensitive TSH assay and a measured or estimated serum FT4, true thyroid dysfunction should be readily distinguished from the transient effects of the psychiatric disturbance. The exceptions notwithstanding, diverse NTIs have a surprisingly predictable influence on parameters of thyroid economy. The severity of the illness rather than the type of illness appears to determine the presence and magnitude of the observed change. The perturbations in indices of thyroid function observed in these and other common medical conditions are outlined in Table 36-1.

TABLE 36-1. The Effect of Various Clinical Conditions or Illnesses on Indices of Thyroid Function

PATHOPHYSIOLOGY
SITES OF NONTHYROIDAL ILLNESS INTERACTION WITH THYROID ECONOMY Every accessible point in thyroid hormone homeostasis has been examined or considered as a potential contributor to the thyroidal effects of NTI (Fig. 36-3). A number of studies have examined the role of cytokines in the mediation of these effects (discussed later). At the hypothalamic-pituitary level, decreases in mRNA for TRH have been noted in the hypothalamus obtained at autopsy from patients succumbing to an NTI.43 In these patients, TRH gene expression was inversely correlated with antemortem serum T3 levels. Other endogenous modulators of the hypothalamicpituitarythyroid axis that potentially are activated during NTI include cortisol, somatostatin, dopamine, and the b-endorphins. Within the pituitary, several potential defects, including altered tissue levels of T444 or of the T3 receptor45; enhanced activity of the type II 5'-monodeiodinase enzyme13; and diminished synthesis,46 release,19 or bio-activity47,48 of TSH, have been considered as possible explanations for the apparently blunted pituitary responsiveness that occurs in NTI. Multiple disturbances, rather than any single defect, probably account for the diverse changes in the hypotha-lamic-pituitary control of thyroid function observed in NTI.

FIGURE 36-3. Potential sites of interaction between thyroid economy and nonthyroidal illness. Interaction at multiple levels of regulation has been demonstrated or

hypothesized (see text). (TRH, thyrotropin-releasing hormone; T4, thyroxine; T3, triiodothyronine; TSH, thyroid-stimulating hormone; rT3, reverse T3; T2, diiodothyronine.)

Much has been learned about thyroid hormone production, binding defects, and peripheral metabolism in NTI. Decrements in circulating thyroid hormone levels are well-established features of NTI, yet the production and release of serum T4 is normal in these disorders. Furthermore, a decreased production or an accelerated loss (nephrotic syndrome) of binding proteins accounts for only a small portion of the observed decreases in serum T4 and T3 levels. Circulating inhibitors of thyroid hormone binding have been postulated to contribute to the lowering of serum total T4 levels12 and the decreased T4/TBG ratio49 that occurs in NTI in the face of normal or high serum FT4 concentrations. Free fatty acids have been investigated in this regard.10 However, the nature and even the existence of these inhibitors remains controversial, and alternative explanations have been proposed for the binding abnormality in NTI.50 Defects in TBG itself, as evidenced by relative increases in an isoform that has an altered electrophoretic motility, have been reported to affect thyroid hormone levels in NTI.51 Altered metabolism of T4 resulting from decreased activity of type I 5'-monodeiodinase activity is the best known, yet still incompletely understood, phenomenon noted in NTI. The result of this enzymatic inhibition is the striking decrease in serum T3 and elevation in rT3 that is characteristic of the euthyroid sick syndrome. Diminished hepatic uptake of T4 may also contribute to decreased peripheral production of T3 in patients with NTI.52 Interestingly, rT3 may compete with T4 as a substrate for both type I and type II 5'-monodeiodinases, compounding the defect in T4-to-T3 conversion.11 Other nonpharmacologic mechanisms contributing to the inhibition of 5'-monodeiodination that occurs during NTI remain to be elucidated. The cloning of the complementary DNA for human type I 5'-monodeiodinase53 has allowed the demonstration of decreased expression of this gene in an animal model of NTI.54 Tissue levels of thyroid hormones are likely to be important determinants of the true metabolic state in NTI but are not readily measured. The existing studies suggest that decreased tissue levels of thyroid hormones are present in patients who succumb to a wasting illness as compared to patients who experience sudden death.45 The uptake of T4 into cultured rat hepatocytes52 or human hepatoma cells55 is lower in the presence of sera from patients with NTI than in control sera. T3 receptor mRNA is increased in peripheral leukocytes and liver cells from patients with NTI, and the increments in hepatic T3 receptor are partially reversed after liver transplantation.56 These findings suggest a mechanism for an intact cellular response to T3 in the face of decreases in cellular uptake and tissue levels of thyroid hormones. Finally, the ultimate assessment of the integrity of thyroid hormone economy during NTI is measurement of the transcription and translation products from thyroid hormoneresponsive genes. Circulating markers of thyroid hormone action, such as levels of TSH, ferritin, osteo-calcin, sex hormonebinding protein, and angiotensin-converting enzyme, are generally normal in NTI,57 as is the production rate of rT 3, a finding consistent with euthyroidism at the target organ level.15 Similarly, clinical indicators of thyroid status, such as basal metabolic rate, QKd interval (a measure of cardiac contractility), and deep tendon reflex relaxation times, are most consistent with a euthyroid state in these patients.15 The role of cytokines in the modulation of the thyroid hormone changes seen in NTI has received considerable attention.58 Interpretation of these studies is hampered by the complexity and redundancy of the cytokine network, the flu-like illness (and accompanying NTI-related thyroid changes) associated with experimental infusion of these agents, and the difficulty in discerning causation from mere association when concurrent changes are observed in both thyroid hormone and cytokine levels. Further compounding this difficulty is the lack of measurable changes in circulating levels of those cytokines that act through paracrine mechanisms, such as TNF-a. Despite these obstacles, several interesting findings have emerged. At the hypothalamic-pituitary level, interleukin-1b (IL-1b) decreases the levels of TRH mRNA in the rat,59 and TNF-a, IL-1b, and interleukin-6 (IL-6) decrease serum levels of TSH in human subjects. At the thyrocyte level, TNF-a, interleukin-1a (IL-1b, IL-1b, IL-6, and interferon-g display inhibitory effects in vitro, and TNF-a, IL-1b, and IL-6 decrease thyroid hormone levels in experimental animals.58,60,61 and 62 A strong correlation has been noted between serum IL-6 levels and T3 levels (inversely correlated) and rT3 levels,63,64 but as noted earlier, this does not prove causation. For example, although IL-6 knock-out mice have diminished NTI-related thyroid hormone changes,65 the reduction of IL-6 after mouse liver macrophage depletion does not prevent the development of NTI changes in thyroid hor-mones.66 Likewise, although IL-6 infusion in humans acutely decreases both T3 and TSH, these latter changes remit with chronic IL-6 administration.67 Lastly, incubation of rat hepato-cytes with the cytokines and IL-6, TNF-a, or IL-1b inexplicably increases 5'-monodeiodinase activity in rat hepatocytes.61 Hence, although cytokines are likely to contribute to the thy-roid hormone changes associated with NTI, the extent of this participation remains to be determined.

DIFFERENTIATION FROM PRIMARY THYROID DYSFUNCTION

NONTHYROIDAL ILLNESS SIMULATING THYROID DYSFUNCTION A timely distinction between the alterations in thyroid indices that are due to NTI alone and those that are related to true thyroid dysfunction is critical both to the prevention of inappropriate intervention and to the identification of patients with myxedema or thyrotoxicosis, in whom metabolic correction may have a profound influence on disease outcome. A knowledge of the previously outlined patterns of change typically seen in NTI, as well as an awareness of the pitfalls associated with strict reliance on normal ranges for thyroid function tests in NTI, better equips the clinician to make this distinction. Advances in serum TSH assays have greatly increased the sensitivity and specificity of these assays for detecting thyroid hormone excess and deficiency states. However, applying the normal range designed to distinguish euthyroidism from thyroid hormone excess or thyroid hormone deficiency to patients with NTI results in a loss of specificity.26 This is particularly true for patients with serum TSH values only slightly above or below the normal range. When the euthyroid range for serum TSH is broadened to encompass values between 0.1 and 20.0 mIU/mL in patients with NTI, the specificity of this test is enhanced to 97%.26 However, the fact that even patients with NTI who have serum TSH values that are undetectable or greater than 20 mIU/mL are still likely to be found ultimately to be euthyroid underscores the peril of relying on any single thyroid function test in these cases.26 The combined measurement of serum FT4 and TSH is prudent in patients with NTI in whom thyroid dysfunction is suspected. Although deviation from the reference range commonly occurs when either of these assays is used in patients with NTI, the concurrence of a low serum FT4 and a high TSH (hypothy-roid pattern) or of a high serum FT4 and a low TSH (hyperthyroid pattern) is distinctly uncommon as a manifestation of NTI alone5,68 (Fig. 36-4). Clinicians must be familiar with the techniques used at their institutions to measure or to estimate FT4. Indices, as well as analog methodologies, are likely to underestimate serum FT4 values during NTI. Newer, simplified methods for measuring FT4 by direct equilibrium dialysis should provide the needed supplement to a sensitive TSH assay in this setting. For patients for whom diagnostic uncertainty remains, the measurement of significant antibody titers against thyroid peroxidase (microsomal antigen) appears to enhance the likelihood of finding underlying thyroid disease.26 This possibility also is greater in patients who are older than 60 years of age and in those who have other autoimmune disorders.15 Similarly, an abnormal TSH response to TRH further increases the likelihood that true thyroid dysfunction is present. A diagnostic approach encompassing these concepts in hospitalized patients with NTI is illustrated in Figure 36-5. A comparison of thyroid function indices found in NTI to those observed in primary hypothyroidism is provided in Table 36-2.

FIGURE 36-4. Distinguishing true thyroid dysfunction from the effects of nonthyroidal illness (NTI). Although both the free thyroxine (T4) and thyroid-stimulating hormone (TSH) values may deviate from the reference range during NTI, the combined use of these parameters generally allows for the accurate distinction from primary thyroid dysfunction. Patients with NTI who have abnormal values for either FT4 or TSH (or both) are likely to occupy positions above and to the right of the reference range as shown in this nomogram. This is distinct from the pattern seen in patients experiencing either primary hypothyroidism (above and to the left) or hyperthyroidism (below and to the right). (Adapted from Kaptein EM. Clinical application of free thyroxine determinations. Clin Lab Med 1993; 13:653.)

FIGURE 36-5. An approach to thyroid function testing in hospitalized patients. The combined use of free thyroxine (FT4) and a sensitive thyroid-stimulating hormone (TSH) assay suggests primary thyroid dysfunction when these two parameters deviate in opposite directions from the reference range. Conversely, thyroid dysfunction is unlikely when both FT4 and TSH are normal. When only one of these parameters is normal, further information or testing is needed to distinguish non-thyroidal illness (NTI) from true thyroid dysfunction. A normal FT 4 but subnormal TSH in patients receiving glucocorticoids or dopamine is likely to be explicable as a drug effect alone. In the absence of these medications, this pattern may represent a subclinical hyperthyroidism that will yield a blunted TSH response to thyrotropin-releasing hormone (TRH) testing. A normal FT4 but high TSH may be seen in both subclinical hypothyroidism and NTI. The presence of significant (>1:400) microsomal antibody (MA) titers increases the likelihood of the former; negative MA titers increase the likelihood that the TSH elevation is the result of NTI alone. Testing should be performed again to verify this interpretation after recovery from illness. (From Spencer CA. Serum TSH measurement: a 1990 status report. Thyroid Today 1990; 13:1.)

TABLE 36-2. Comparison of Laboratory Findings in Nonthyroidal Illness and Primary Hypothyroidism

This discussion has dealt primarily with distinguishing NTI from primary thyroid dysfunction. Other, rarer causes of diagnostic uncertainty in critically ill patients include central hypothyroidism (in which other clinical evidence that is suggestive of hypothalamic-pituitary dysfunction may be evident) and disorders of inappropriate TSH secretion, including TSH-secreting pituitary adenomas and syndromes of thyroid hormone resistance. NONTHYROIDAL ILLNESS MASKING THYROID DYSFUNCTION Occasionally, the superimposition of NTI on an established thyroid disorder results in the transient normalization of thyroid indices, thereby masking true thyroid dysfunction. The most common example of this occurrence is the depression of serum TSH from a clearly hypothyroid value toward one that might easily be attributed to NTI alone.5,69 Another example is a decrease of serum T3 and T4 in a thyrotoxic patient to the euthyroid range during an NTI, with subsequent unmasking of the toxic state after recovery from the disease.8,70 These examples point to the need for careful follow-up of patients who are found to have incongruous thyroid indices or who manifest equivocal clinical findings during the course of an NTI. Thus, values should be measured again 2 to 4 weeks after recovery from the disease, or at any time during the course of an illness that the clinical findings become suggestive of true thyroid dysfunction.

PROGNOSTIC IMPLICATIONS AND ROLE OF THYROID HORMONE THERAPY

In critically ill patients, a strong correlation has been found between the serum T4 level at the time of hospital admission and the mortality rate. Given that the magnitude of depression in thyroid hormones corresponds to the severity of the underlying NTI, the fact that patients who have the greatest thyroid hormone depression have the worst prognosis is not unexpected. A particularly high mortality rate has been noted in critically ill patients with levels of serum T4 and TSH that remain depressed for >1 week after admission to hospital.5 The serum T4 depression correlates negatively with survival in a manner that parallels the traditional Acute Physiology and Chronic Health Evaluation II (APACHE-II) prognostic score. Critically ill elderly patients with high serum IL-6 values and low serum albumin levels are more likely to have NTI-related thyroid alterations and have a higher likelihood of dying from their acute illness.71 Although the thyroid changes associated with NTI are generally believed to be adaptive in nature and to conserve energy for the organism as a whole, the possibility exists that organ or tissue-specific hypothyroidism is present,72 the correction of which might lead to clinical improvement. Several studies have assessed the utility of replacement therapy in critically ill patients or laboratory animals who had depressed levels of thyroid hormones. Early controlled studies in humans, using either T3 or T4, failed to demonstrate beneficial effects on survival. Two prospective randomized trials have examined the use of T3 to treat patients undergoing coronary artery bypass surgery.73,74 Despite improvement in cardiac index and decreased vascular resistance in one of these trials,74 no differences were observed in either study with respect to multiple outcome variables, including the requirement for inotropic agents, the incidence of postoperative arrhythmia, time in the intensive care unit, or mortality rate (Fig. 36-6). On the basis of these studies, little role appears to exist for the administration of T3 or T4 to patients with a NTI.

FIGURE 36-6. Lack of effect of triiodothyronine (T3) treatment on cardiovascular parameters in patients undergoing coronary artery bypass graft surgery. Values shown represent readings taken at 4 hours after g/kg intravenous bolus followed by 0.12 g/kg/hour continuous infusion) or dopamine (5 g/kg/minute continuous infusion) or normal saline alone (placebo). Cardiac index (CI) shown as 10 (L/min)/m2; stroke volume (SV) shown in mL; mean arterial blood pressure (MAP) shown as mm Hg; heart rate (HR) shown as beats per minute; and systemic vascular resistance (SVR) shown as 0.1 dyne s cm5. (Adapted from Bennett-Guerrero E, Jimenez JL, White WD, D'Amico EB, Baldwin BI, Schwinn DA. Car-diovascular effects of intravenous triiodothyronine in patients under-going coronary artery bypass graft surgery. A randomized, double-blind, placebo-controlled trial. Duke T 3 study group. JAMA 1996; 275:687.)

In patients being evaluated for possible myxedema, in whom the mitigating influences of an NTI, the use of medication, and an equivocal clinical assessment prevent the exclusion of hypothyroidism, the judicious administration of thyroid hormone while awaiting the results of serum FT4 and TSH measurements is unlikely to cause acute harm in patients with NTI and significantly improves the outcome in patients with myxedema.

EUTHYROID HYPERTHYROXINEMIA
Numerous disorders result in elevations of circulating T4 levels without true thyrotoxicosis. These can be grouped into four large categories, including abnormalities in the concentration or function of thyroid hormonebinding proteins, the effects of pharmacologic agents, syndromes of generalized or peripheral thyroid hormone resistance, and variants of NTI. These conditions are summarized in Table 36-3. Among patients with disorders involving thyroid hormone binding proteins, those with familial dysalbuminemic hyper-thyroxinemia and serum thyroxine-binding prealbumin elevations have a falsely elevated serum FT4 index resulting from normal binding of the radiolabeled T3 that is used in most thyroid hormonebinding resin or T3-resin uptake assays. In addition, patients with familial dysalbuminemic hyperthyroxinemia,74a as well as persons with T4 autoantibodies, may have falsely elevated serum FT4 values, when determined using some FT4 methodologies. The elevation in serum T4 levels that is seen after the use of oral contrast agents and amiodarone in healthy persons must be distinguished from the true hyperthyroidism that may occur through jodbasedow mechanisms in the subset of patients who are susceptible to this phenomenon (see Chap. 37).

TABLE 36-3. Causes of Euthyroid Hyperthyroxinemia

Generalized resistance to thyroid hormone, a disorder caused by mutations in the c-erbA b gene for the T3 receptor, is characterized by a heterogeneity in tissue responsiveness to the elevated levels of thyroid hormone; this may impede attempts to exclude thyroid dysfunction (see Chap. 32). Despite the variable tissue response to thyroid hormone in this disorder, most patients with generalized resistance to thyroid hormone are asymptomatic and have similarly affected family members, which facilitates the establishment of the correct diagnosis. Patients with pituitary-selective thyroid hormone resistance are truly hyperthyroid and, therefore, are not included in this section. (Common NTIs in which patients may present with elevatedrather than normal or depressedthyroid hormone levels have been discussed previously.) The true metabolic state in hyperemesis gravidarum is somewhat controversial75; some argue for a human chorionic gonadotropinmediated thyrotoxicosis in this common disorder.

DRUGS AFFECTING PARAMETERS OF THYROID FUNCTION

Many medications may exert unwanted effects at one or more levels of thyroid hormone economy.76 Of particular importance are those agents that are commonly used in the diagnosis or treatment of critically ill patients, such as iodinated contrast agents, dopamine, corticosteroids, heparin, and furosemide. In addition, a number of common medications such as calcium carbonate,77 ferrous sulfate,78 aluminum hydroxide,79 sucralfate,80 and choles-tyramine81 may adversely affect gastrointestinal absorption of L -thyroxine. The effects of NTI on thyroid economy are compounded by the addition of these drugs and make the distinction from true thyroid dysfunction even more challenging to discern. Although certain of these changes are purely in vitro phenomena, such as the heparin-induced interference with thyroid hormone binding, other manifestations, such as the impairment of pituitary responsiveness that is caused by dopamine therapy, can lead to decreases in FT4 to hypothyroid levels in the setting of critical illness. Table 36-4 provides a list of common medications known to affect thyroid economy, as well as the known or hypothesized mechanisms underlying these interactions.

TABLE 36-4. Potential Effects of Various Drugs on Parameters of Thyroid Function

CHAPTER REFERENCES
1. Wartofsky L, Burman KD. Alterations in thyroid function in patients with systemic illness: the euthyroid sick syndrome. Endocr Rev 1982; 3:164. 2. Wartofsky L. The low T3 of euthyroid sick syndrome: update 1994. In: Braverman LE, Refetoff S, eds. Endocrine review monographs. 3. Clinical and molecular aspects of diseases of the thyroid. Bethesda, MD: Endocrine Society Monographs, 1994:248. 3. Hamblin PS, Dyer SA, Mohr VS, et al. Relationship between thyrotropin and thyroxine changes during recovery from severe hypothyroxinemia of critical illness. J Clin Endocrinol Metab 1986; 62:717. 4. Bermudez F, Surks MI, Oppenheimer JH. High incidence of decreased serum triiodothyronine concentration in patients with nonthyroidal dis-ease. J Clin Endocrinol Metab 1975; 41:27. 5. Bayer MF, Macoviak JA, McDougall IR. Diagnostic performance of sensitive measurements of serum thyrotropin during severe nonthyroidal illness: their role in the diagnosis of hyperthyroidism. Clin Chem 1987; 33:2178. 6. Melmed S, Geola FL, Reed AW, et al. A comparison of methods for assess-ing thyroid function in nonthyroidal illness. J Clin Endocrinol Metab 1982; 54:300. 7. Surks MI, Hupart KH, Pan C, Shapiro LE. Normal free thyroxine in critical nonthyroidal illnesses measured by ultrafiltration of undiluted serum and equilibrium dialysis. J Clin Endocrinol Metab 1988; 67:1031. 8. Lum SMC, Kaptein EM, Nicoloff JT. Influence of nonthyroidal illness on serum thyroid hormone indices in hyperthyroidism. West J Med 1983; 138:670. 9. Burmeister LA. Reverse T3 does not reliably differentiate hypothyroid sick syndrome from euthyroid sick syndrome. Thyroid 1995; 5:435. 10. Chopra IJ. Clinical review 86. Euthyroid sick syndrome: is it a misnomer? J Clin Endocrinol Metab 1997; 82:329. 11. Hurd RE, Chopra IJ. Reverse T3. Endocrinologist 1993; 3:365. 12. Chopra IJ, Solomon DH, Hepner GW, Morgenstein AA. Misleadingly low free thyroxine index and usefulness of reverse triiodothyronine measure-ment in nonthyroidal illness. Ann Intern Med 1979; 90:905. 13. Wehmann RE, Gregerman RI, Burns WH, et al. Suppression of thyrotropin in the low-thyroxine state of severe nonthyroidal illness. N Engl J Med 1985; 312:546. 14. Hay ID, Bayer MF, Kaplan MM, et al. American Thyroid Association assessment of current free thyroid hormone measurements and guidelines for future clinical assays. Clin Chem 1991; 37:2002. 15. Kaptein EM. Clinical application of free thyroxine determinations. Clin Lab Med 1993; 13:653. 16. Faber J, Kirkegaard C, Rasmussen B, et al. Pituitary-thyroid axis in critical illness. J Clin Endocrinol Metab 1987; 65:315. 17. Nelson JC, Tomei RT. Direct determination of free thyroxine in undiluted serum by equilibrium dialysis/radioimmunoassay. Clin Chem 1988; 34:1737. 18. Chopra IJ. Simultaneous measurement of free thyroxine and free 3,5,3'-tri-iodothyronine in undiluted serum by direct equilibrium dialysis/radioim-munoassay: evidence that free triiodothyronine

and free thyroxine are normal in many patients with the low triiodothyronine syndrome. Thyroid 1998; 8:249. 19. Spencer CA, Lum SMC, Wilber JF, et al. Dynamics of serum thyrotropin and thyroid hormone changes in fasting. J Clin Endocrinol Metab 1983; 56:883. 20. Romijn JA, Wiersinga WM. Decreased nocturnal surge of thyrotropin in nonthyroidal illness. J Clin Endocrinol Metab 1990; 70:35. 21. Spencer CA, Schwarzbein D, Guttler RB, et al. Thyrotropin-releasing hor-mone stimulation test responses employing third and fourth generation TSH assays. J Clin Endocrinol Metab 1993; 76:494. 22. Maturlo SJ, Rosenbaum RL, Surks MI. Variable thyrotropin response to thyrotropin-releasing hormone after small decreases in plasma free thyroid hormone concentrations in patients with nonthyroidal illness. J Clin Invest 1980; 66:451. 23. Sumita S, Ujike Y, Namiki A, et al. Suppression of the thyrotropin response to thyrotropin-releasing hormone and its association with severity of criti-cal illness. Crit Care Med 1994; 22:1603. 24. Borst GC, Osborne RC, O'Brian JT, et al. Fasting decreases thyrotropin respon-siveness to thyrotropin-releasing hormone: a potential cause of misinterpretation of thyroid function tests in the critically ill. J Clin Endocrinol Metab 1983; 57:380. 25. Adriaanse R, Romijn JA, Brabant G, et al. Pulsatile thyrotropin secretion in nonthyroidal illness. J Clin Endocrinol Metab 1993; 77:1313. 26. Spencer C, Eigen A, Shen D, et al. Specificity of sensitive assays of thyrotro-pin used to screen for thyroid disease in hospitalized patients. Clin Chem 1987; 33:1391. 27. Spencer CA, LoPresti JS, Patel A, et al. Applications of a new chemilumino-metric thyrotropin assay to subnormal measurement. J Clin Endocrinol Metab 1990; 70:453. 28. Franklyn JA, Black EG, Betteridge J, Sheppard MC. Comparison of second and third generation methods for measurement of serum thyrotropin in patients with overt hyperthyroidism, patients receiving thyroxine therapy, and those with nonthyroidal illness. J Clin Endocrinol Metab 1994; 78:1368. 29. LoPresti JS, Gray D, Nicoloff JT. Influence of fasting and refeeding on 3,3',5'-triiodothyronine metabolism in man. J Clin Endocrinol Metab 1991; 72:130. 30. Burman KD, Dimond RC, Harvey GS, et al. Glucose modulation of alterations in serum iodothyronines induced by fasting. Metabolism 1979; 28:291. 31. Burman KD, Wartofsky L, Dinterman RE, et al. The effect of T3 and reverse T 3 on muscle protein catabolism during fasting as measured by 3-methyl-histidine excretion. Metabolism 1979; 28:805. 32. Kaptein EM. Thyroid hormone metabolism and thyroid diseases in chronic renal failure. Endocr Rev 1996; 17:45. 32a. Acker CG, Singh AR, Flick RP, et al. A trial of thyroxine in acute renal failure. Kidney Int 2000; 57:293. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. John R, Evans PE, Scanlon MF, Hall R. Clinical value of immunoradiomet-ric assay of thyrotropin for patients with nonthyroidal illness and taking various drugs. Clin Chem 1987; 33:566. Schussler GC, Schaffner F, Korn F. Increased serum thyroid hormone binding and decreased free hormone in chronic active liver disease. N Engl J Med 1978; 299:510. LoPresti JS, Fried JC, Spencer CA, Nicoloff JT. Unique alterations of thyroid hormone indices in the acquired immunodeficiency syndrome. Ann Intern Med 1989; 110:970. Sellmeyer DE, Grunfeld C. Endocrine and metabolic disturbances in human immunodeficiency virus infection and the acquired immune deficiency syndrome. Endocr Rev 1996; 17:518. Hommes MJT, Romijn JA, Endert E, et al. Hypothyroid-like regulation of the pituitary-thyroid axis in stable human immunodeficiency virus infection. Metabolism 1993; 42:556. Lambert M, Zech F, De Nayer P, et al. Elevation of serum thyroxine-binding globulin (but not of cortisol-binding globulin and sex hormonebind-ing globulin) associated with the progression of human immunodeficiency virus infection. Am J Med 1990; 89:748. Spratt DI, Pont A, Miller MB, et al. Hyperthyroxinemia in patients with acute psychiatric disorders. Am J Med 1982; 73:41. Nader S, Warner MD, Doyle S, Peabody CA. Euthyroid sick syndrome in psychiatric inpatients. Biol Psychiatry 1996; 40:1288. Chopra IJ, Solomon DH, Huang T-S. Serum thyrotropin in hospitalized patients: evidence for hyperthyroxinemia as measured by an ultrasensitive thyrotropin assay. Metabolism 1990; 39:538. Maes M, Schotte C, Vandewounde M, et al. TSH responses to TRH as a function of basal serum TSH: relevance for unipolar depression in femalesa multivariate study. Pharmacopsychiatry 1992; 25:136. Fliers E, Guldenaar SE, Wiersinga WM, Swaab DF. Decreased hypotha-lamic thyrotropin-releasing hormone gene expression in patients with non-thyroidal illness. J Clin Endocrinol Metab 1997; 82:4032. Arem R, Wiener GJ, Kaplan SG, et al. Reduced tissue thyroid hormone levels in fatal illness. Metabolism 1993; 42:1102. Hupart KH, De Fesi CR, Katz CP, et al. Decreased anterior pituitary T3 nuclear receptors in a Walker 256 carcinomabearing rat model of nonthy-roidal disease. Acta Endocrinol (Copenh) 1989; 121:811. Boado R, Chopra IJ, Huang T-S, et al. A study of pituitary thyrotropin, its subunits, and messenger ribonucleic acids in nonthyroidal illness. Metabolism 1988; 37:395. Lee H-Y, Suhl J, Pekary AE, Hershman JM. Secretion of thyrotropin with reduced concanavalin-Abinding activity in patients with nonthyroidal illness. J Clin Endocrinol Metab 1987; 65:942. Magner J, Roy P, Fainter L, et al. Transiently decreased sialylation of thy-rotropin (TSH) in a patient with the euthyroid sick syndrome. Thyroid 1997; 7:55. Pandian MR, Morgan C, Nelson JC, Fisher DA. Differentiating various abnormalities of thyroxine binding to serum proteins by radioelectro-phoresis of thyroxine and immunoassay of binding proteins. Clin Chem 1990; 36:457. Mendel CM, Laughton CW, McMahon FA, Cavalerri RR. Inability to detect an inhibitor of thyroxine-serum protein binding in sera from patients with nonthyroid illness. Metabolism 1991; 40:491. Reilly CP, Wellby ML. Slow thyroxine binding globulin in the pathogenesis of increased dialysable fraction of thyroxine in nonthyroidal illness. J Clin Endocrinol Metab 1983; 57:15. Lim CF, Docter R, Visser TJ, et al. Inhibition of thyroxine transport into cultured rat hepatocytes by serum of nonuremic critically ill patients: effects of bilirubin and nonesterified fatty acids. J Clin Endocrinol Metab 1993; 76:1165. Mandel SJ, Berry MJ, Kieffer JD, et al. Cloning and in vitro expression of the human selenoprotein type I iodothyronine deiodinase. J Clin Endo-crinol Metab 1992; 75:1133. Boelen A, Platvoetter Schiphorst MC, Wiersinga WM. Immunoneutraliza-tion of interleukin-1, tumor necrosis factor, interleukin-6 or interferon does not prevent the LPS-induced sick euthyroid syndrome in mice. J Endo-crinol 1997; 153:115. Sarne DH, Refetoff S. Measurement of thyroxine uptake from serum by cultured human hepatocytes as an index of thyroid status: reduced thyroxine uptake from serum of patients with nonthyroidal illness. J Clin Endo-crinol Metab 1985; 61:1046. Williams GR, Neubauer JM, Franklin JA, Sheppard MC. Thyroid hormone receptor expression in the sick euthyroid syndrome. Lancet 1989; 2:1477. Seppel T, Becker A, Lippert F, Schlaghecke R. Serum sex hormonebinding globulin and osteocalcin in systemic nonthyroidal illness associated with low thyroid hormone concentrations. J Clin Endocrinol Metab 1996; 81:1663. Bartalena L, Bogazzi F, Brogioni S, et al. Role of cytokines in the pathogenesis of the euthyroid sick syndrome. Eur J Endocrinol 1998; 138:603. Kakucska I, Romero LI, Clark BD, et al. Suppression of thyrotropin-releasing hormone gene expression by interleukin-1-beta in the rat: implications for nonthyroidal illness. Neuroendocrinology 1994; 59:129. Poth M, Tseng YC, Wartofsky L. Inhibition of TSH activation of human thyroid cells by tumor necrosis factor: an explanation for decreased thyroid function in systemic illness? Thyroid 1991; 1:235. Davies PH, Sheppard MC, Franklyn JA. Inflammatory cytokines and type I 5'-deiodinase expression in phi1 rat liver cells. Mol Cell Endocrinol 1997; 129:191. Jones TH, Wadler S, Hupart KH. Endocrine-mediated mechanisms of fatigue during treatment with interferon-alpha. Semin Oncol 1998; 25(Suppl 1):54. Boelen A, Platvoet-Ter Schiphorst MC, Wiersinga WM. Soluble cytokine receptors and the low 3,5,3'-triiodothyronine syndrome in patients with nonthyroidal disease. J Clin Endocrinol Metab 1995; 80:971. Bartalena L, Brogioni S, Grasso L, et al. Relationship between the increased serum interleukin-6 concentration to changes in thyroid function in non-thyroidal illness. J Endocrinol Invest 1994; 17:269. Boelen A, Maas MA, Lowik CW, et al. Induced illness in interleukin-6 (IL-6) knock-out mice: a causal role of IL-6 in the development of the low 3,5,3'-triiodothyronine syndrome. Endocrinology 1996; 137:5250. Boelen A, Platvoet-Ter Schiphorst MC, van Rooijen N, Wiersinga WM. Selective macrophage depletion in the liver does not prevent the development of the sick euthyroid syndrome in the mouse. Eur J Endocrinol 1996; 134:513. Stouthard JM, van der Poll T, Endert E, et al. Effects of acute and chronic interleukin-6 administration on thyroid hormone metabolism in humans. J Clin Endocrinol Metab 1994; 79:1342. Midgley JEM, Sheehan CP, Christofides ND, et al. Concentrations of free thyroxine and albumin in serum in severe nonthyroidal illness: assay artifacts and physiologic influences. Clin Chem 1990; 36:765. Van den Berghe G, de Zegher F, Lauwers P. Dopamine and the sick euthy-roid syndrome in critical illness. Clin Endocrinol (Oxf) 1994; 41:731. 70. Archambeaud-Mouveroux F, Dejax C, DeBuhan B, Bonnaud F. Hyperthy-roidism without elevated levels of thyroxine and triiodothyronine in a patient with pulmonary TB. South Med J 1989; 82:907. Girvent M, Maestro S, Hernandez R, et al. Euthyroid sick syndrome, associated endocrine abnormalities, and outcome in elderly patients undergoing emergency operation. Surgery 1998; 123:560. Brennan MD, Bahn RS. Thyroid hormone and illness. Endocr Pract 1998; 4:396. Bennett-Guerrero E, Jimenez JL, White WD, et al. Cardiovascular effects of intravenous triiodothyronine in patients undergoing coronary artery bypass graft surgery. A randomized, double-blind, placebo-controlled trial. Duke T 3 study group. JAMA 1996; 275:687. Klemperer JD, Klein I, Gomez M, et al. Thyroid hormone treatment after coronary-artery bypass surgery. N Engl J Med 1995; 333:1522.

74a. Tang KT, Yang HJ, Choo KB, et al. A point mutation in the albumin gene in a Chinese patient with familial dysalbuminemic hyperthyroxinemia. Eur J Endocrinol 1999; 141:374. 75. 76. 77. 78. 79. 80. 81. Burrow GN. Thyroid function and hyperfunction during gestation. Endocr Rev 1993; 14:194. Surks MI, Sievert R. Drugs and thyroid function. N Engl J Med 1995; 333:1688. 77. Schneyer CR. Calcium carbonate and reduction of levothyroxine efficacy. (Letter). JAMA 1998; 279:750. Campbell NRC, Hasinoff BB, Stalts H, et al. Ferrous sulfate reduces thyroxine efficacy in patients with hypothyroidism. Ann Intern Med 1992; 117:1010. 79. Liel Y, Sperber AD, Shany S. Nonspecific intestinal adsorption of levothy-roxine by aluminum hydroxide. Am J Med 1994; 97:363. Sherman SI, Tielens ET, Ladenson PW. Sucralfate causes malabsorption of L-thyroxine. Am J Med 1994; 96:531. Solomon BL, Wartofsky L, Burman KD. Adjunctive cholestyramine therapy for thyrotoxicosis. Clin Endocrinol (Oxf) 1993; 38:39.

CHAPTER 37 ADVERSE EFFECTS OF IODIDE Principles and Practice of Endocrinology and Metabolism

CHAPTER 37 ADVERSE EFFECTS OF IODIDE

JENNIFER A. NUOVO AND LEONARD WARTOFSKY Population Exposure to Iodide Adverse Effects of Iodide Extrathyroidal Effects Intrathyroidal Effects Iodide Sources Average Iodide Intake Radiographic Contrast Agents Iodide-Containing Drugs Mechanism of Action of Excess Iodide Results of Chronic Iodide Excess in Healthy Persons Preexisting Factors Iodine-Induced Goiter and Hypothyroidism Effects During Pregnancy Neonatal Period Coast Goiter Predisposition to Induced Hypothyroidism Clinical Features of Iodide-Induced Hypothyroidism Iodide-Induced Thyrotoxicosis Clinical Features of Iodide-Induced Thyrotoxicosis Treatment of Iodide-Induced Thyrotoxicosis Iodine-Induced Autoimmunity Chapter References

POPULATION EXPOSURE TO IODIDE

Iodine is a requisite part of the thyroid hormone molecule, and a deficiency or excess of iodine can dramatically affect thyroid gland function. In 1820, iodide therapy was being used in Europe for the treatment of goiter; shortly thereafter, cases of probable thyrotoxicosis were reported in a few patients receiving this treatment. Although iodide supplementation has been adopted worldwide in the past 50 years, concern remains over the danger this poses to certain susceptible individuals.1,2 Iodine deficiency and goiter remain major health problems in many parts of the world. During the two decades from 1965 to 1985 in the United States, however, iodine supplementation resulted in an increased prevalence of disordered thyroid regulation resulting from iodine excess.1,2 and 3 Although the normal thyroid gland usually adapts readily to extremes of iodine intake, some persons are vulnerable to the induction of goiter, hypothyroidism, or hyperthyroidism with exposure to excess iodine. This risk may be diminishing, for it appears that iodine intake in North America is on the decline, conceivably secondary to a reduction in its use as a preservative and disinfectant in the food and dairy industry.4 The physiology of iodine metabolism is discussed in Chapter 30 and has been reviewed.5 Normal thyroid economy can be maintained with an iodine intake of only 50 to 70 g per day because some iodine is available internally from the leak of nonhormonal iodine from the thyroid and from the peripheral deiodination of thyroxine (T4) and triiodothyronine (T3). Urinary iodide excretion fairly accurately reflects dietary (and non-dietary) intake. Urinary iodide excretion may vary from 45 to 700 g per 24 hours in areas of iodide sufficiency but be as low as 3 g per 24 hours in regions of endemic goiter.6 In addition to dietary sources, large amounts of iodine are available in drugs, intravenous and oral radiologic contrast agents, and supplements from health-food stores. With increasing frequency, such exposures to iodine excess are being recognized as having adverse effects on thyroid function. Speculation also exists that the chronically high levels of iodine intake observed in Western countries may contribute to the increasing prevalence of autoimmune thyroid disease.

ADVERSE EFFECTS OF IODIDE

EXTRATHYROIDAL EFFECTS The adverse effects of iodides may be classified as extrathyroidal or intrathyroidal. Extrathyroidal effects are uncommon; the most frequently encountered is sialoadenitis, which occurs after the administration of large doses of iodide. (Therapeutic doses of radioactive iodine also may cause sialoadenitis; however, this depends on the radiation dose rather than on the extremely small amount of stable iodine administered.) The acute, painful swelling of the parotid and other salivary glands resolves shortly after iodide therapy is discontinued. Severe skin eruptions have been reported after topical and enteral iodine exposure. In addition, iodism may occur, which is marked by symptoms of metallic taste, acneiform skin lesions, mucous membrane irritation, nausea, and salivary gland swelling. These symptoms subside on withdrawal of iodide. Various allergic reactions, including a syndrome similar to polyarteritis nodosa, eosinophilia, and fever, also have been reported. Drug sensitivity rashes, including dermatitis herpetiformis and hypocomplementemic vasculitis, can occur after iodide exposure but are rare.7 INTRATHYROIDAL EFFECTS Iodide exposure causes three primary types of intrathyroidal effects. The first is iodide-induced thyroiditis, which sometimes causes painful inflammation of the thyroid that occurs after the receipt of large doses of iodide and usually resolves rapidly and spontaneously within a few days after the exposure8; however, experimental evidence exists for an iodine-induced autoimmune thyroiditis as well.9,10 The second intrathyroidal iodide effect is iodide goiter with hypothyroidism or, more commonly, goiter alone without hypothyroidism. This usually occurs after prolonged exposure to dietary or drug-containing iodides.11,12 Ultrasonographic techniques, however, have detected significant increases in thyroid size after only 4 weeks of excess iodine supplementation.13 The third effect of iodide on the thyroid is iodide-induced thyrotoxicosis, which is seen after excess iodine exposure in four patient groups: patients from areas of endemic (iodide-deficiency) goiter; patients not from endemic areas with known goiters; euthyroid patients with autonomous hyperfunctioning nodules or histories of prior hyperthyroidism or Graves disease; and, rarely, patients with no underlying thyroid disorders.14,15

IODIDE SOURCES
AVERAGE IODIDE INTAKE The normal thyroid adapts easily to iodide excess or deficiency. Hyperthyroidism or hypothyroidism resulting from abnormal amounts of dietary iodine is unusual. In the United States, dietary iodide intake is estimated to average 350 to 650 g per day, although it may exceed 1 mg per day in some areas. Some of the major sources are iodinated feed used in the dairy industry (secreted as iodine in cow's milk), iodates used as dough conditioners in bread baking, iodide in drinking water, and iodide supplements in salt.15a An average of 10 g of iodine-supplemented table salt per day in the American diet provides 76 g of iodine.16 In some parts of Africa, South America, Europe, the Middle East, and Asia, where iodide-deficiency goiter is prevalent, iodine intake may still average <50 g per day. By contrast, in Japan and other Asian nations, a diet rich in fish and often supplemented with iodide-rich seaweed (kelp) results in an intake of several milligrams per day.11 Elemental iodine or inorganic iodide salts, once ingested, are reduced to iodide, thereafter becoming available for uptake by the thyroid. Iodine is a substantial component of many commonly available oral and topical medications, as well as oral and intravenous contrast agents. Table 37-1 lists commonly used iodide-containing compounds, and the iodine dose expected with routine use.

TABLE 37-1. Iodine Content of Iodinated Radiologic Dyes and Some Iodine-Containing Drugs

RADIOGRAPHIC CONTRAST AGENTS Between 180,000 and 320,000 g of iodine is derived from the radiographic contrast agent given for routine oral cholecystography, and intravenous pyelography presents a dose of >106 g of iodine. Although most of the iodine in radiologic contrast media is in the form of organic iodine, large amounts of free iodide also are present. These organic iodine compounds may be taken up by peripheral tissues, particularly the liver, where deiodination generates free iodide that can be trapped by the thyroid. Much of the iodine becomes protein bound and circulates in serum, remaining available to the thyroid for prolonged periods after a single exposure. Blood iodide content may increase to 200 times baseline levels after oral cholecystography or intravenous pyelography and remains elevated for days to weeks.14,17 Although the thyroidal uptake of even a small amount of the circulating iodide load after a dye study may be significant, normal thyroidal autoregulation results in a reduced fractional uptake and iodide clearance; this serves to regulate thyroid function and maintain homeostasis. In this context, therefore, it is not surprising that, despite the 4 million intravenous pyelography procedures and 4 million oral cholecystography procedures that are done yearly in the United States, clinically significant iodine-induced alteration in thyroid function is uncommon. Some radiologic contrast agents (those containing ipodate) have additional specific actions, such as inhibiting peripheral conversion of T4 to T3. IODIDE-CONTAINING DRUGS Common iodide-containing drugs are listed in Table 37-1. Until recently, reports of iodine-induced goiter and hyperthyroidism were related exclusively to iodide-containing cough syrups and expectorants. Potassium iodide, calcium iodide, and iodinated glycerol usually are combined with antihistamines, decongestants, or bronchodilators as mucolytic agents in cough preparations and asthma medications. These antitussives contain 15 to 325 mg of iodine per teaspoon and are taken in doses of 1 to 2 teaspoons every 4 hours, imposing a typical exogenous iodine dose of 90 to 3900 mg per day. Iodochlorhydroxyquin is compounded in creams as a topical antifungal and bacteriostatic agent. Povidone-iodine solution (Betadine) is widely used as a topical antiseptic. When this agent has been applied directly to open wounds or burns, or has been used as a vaginal douche, high blood iodide levels have been reported that occasionally cause iodide toxicity.14,18,19 and 20 The anti-amebic drugs with the active compound iodoquinol contain 64% organic iodine and, in typical dosages, provide 19.5 g of iodine daily, usually for a 20-day course. Amiodarone is an antiarrhythmic agent, which is widely used in both Europe and the United States. The drug contains 37.2% iodine and is prescribed initially in dosages of 300 to 1200 mg per day, providing 100 to 400 mg of iodine and an estimated 900 g of free iodide.14 Amiodarone can have both acute and chronic effects on thyroid function, as well as the unique effect of inhibiting peripheral conversion of T4 to T3; the drug has been implicated in both hypothyroidism and hyperthyroidism.21,22 Increased serum reverse T3 levels and decreased T3 levels may be seen, as well as increased or decreased thyroid gland T4 production and release.23 Amiodarone causes problems of clinical significance more commonly than do the other drugs mentioned. Reports from France describe a particularly high incidence of abnormalities, perhaps because of widespread use, and also perhaps because the dietary iodine intake is borderline low, resulting in a high incidence of endemic goiter and a susceptibility to thyroid abnormalities after iodine exposure.14,24 The use of benziodarone, a uricosuric available in Europe, has prompted similar reports of iodine-related thyroid abnormalities.

MECHANISM OF ACTION OF EXCESS IODIDE

When a small amount of iodide (100250 g per day) is administered along with the usual dietary intake, there is no appreciable change in iodine-131 (131I) uptake, and no change in thyroidal organic iodine or in the amount of iodotyrosines and iodothyronines produced.25 After larger, single iodide doses (several milligrams), the formation of organic iodine within the thyroid decreases and the monoiodotyrosine/diiodotyrosine ratio16,26 This increases, resulting in decreased T4 and T3 formation. inhibitory effect of iodide on organification and thyroid hormone formation is known as the Wolff-Chaikoff effect. A high concentration of inorganic iodide within the thyroid is necessary for this effect to occur. This inhibition serves as temporary protection against the production of excess amounts of thyroid hormone in the period after the excess ingestion. The inhibitory effect is temporary because the intrathyroidal iodine concentration ultimately decreases as a result of reduced iodide transport; the result is resumption of organic iodine formation and escape from the Wolff-Chaikoff effect. Larger doses, including those in normal daily intake of iodide,27 have an additional independent effect on the thyroid gland to block the release of thyroid hormone. Intrathyroidal excess iodide appears to inhibit proteolysis of thyroglobulin, thereby inhibiting the release of stored T4 and T3 into the circulation. Although this effect may be seen in healthy persons, the diminished hormonal release does not result in hypothyroidism because a decreasing peripheral T4 level stimulates a rise in serum thyroid-stimulating hormone (TSH) and augmented T4 production from the thyroid. Large doses of stable iodine induce an acute reduction in T4 release from the thyroid gland and are used as effective short-term therapy for Graves hyperthyroidism (see Chap. 42). This acute inhibitory effect of iodide is of therapeutic value when a rapid reduction in thyroid hormone secretion is desirable, as in severe hyperthyroid states, in preparation for surgery, or in the treatment of thyroid storm. Five drops of SSKI (190 mg of iodine) or 15 drops of Lugol solution (135 mg of iodine), given two to three times a day, quickly and dramatically decreases thyroid hormone release.28 Although these dosages also block organification, the major effect is on hormone secretion. The use of iodides as the only antithyroidal drug in hyperthyroidism is not recommended, because they are only partially and transiently effective in reducing serum hormone levels. A period of 10 to 14 days is required before the thyroid gland adapts, and the production and release of T4 resume. RESULTS OF CHRONIC IODIDE EXCESS IN HEALTHY PERSONS The long-term administration of iodides results in escape from or adaptation to the iodide concentration. Organic binding is no longer inhibited, and iodothyronine formation proceeds at the expected accelerated rate because of the excess iodine present. Despite excess organic iodine, the serum T4 level usually does not increase measurably in healthy persons, although such an effect has been described.15 In the normal gland, an inhibition of T4 and T3 release by acute, large doses of iodide is more likely to be seen, although it will be overcome by an increase in TSH. Thus, as peripheral iodothyronine levels decrease, hormone release normalizes. Probably because the gland remains responsive to TSH, hypothyroidism is prevented in most populations with high iodide ingestion.14,15 PREEXISTING FACTORS Patients may be exposed to excess iodine in health foods, seaweed, expectorants, amiodarone, and iodine-containing intravenous and oral contrast dyes. Although the incidence of goiter, hyperthyroidism, and hypothyroidism caused by this exposure is small, it should be considered carefully whenever goiter or altered thyroid function is seen in patients given iodide diagnostically or therapeutically. Although iodine-induced thyroid disease probably occurs in the setting of a preexisting thyroid abnormality that prevents the normal adaptation to excess iodine, iodide-induced thyroid abnormalities have been noted in seemingly healthy persons.

IODINE-INDUCED GOITER AND HYPOTHYROIDISM

Iodine-induced goiter was first reported in 1938; since then, >200 cases have been described. Of 154 cases reviewed, euthyroid goiter accounted for 39%, hypothyroidism without goiter for 17%, and hypothyroidism with goiter for the remaining cases.11 Such goiters, often associated with hypothyroidism, have been reported sporadically in the United States for many years and more commonly in Europe. In the past, the problem was seen most often after the use of

iodide-containing expectorants. Amiodarone use has been associated with hypothyroidism and, in several cases, goiter has developed with benziodarone use. The high incidence of antithyroidal antibodies in patients with goiter development suggests that these patients have underlying Hashimoto thyroiditis with abnormal hormone synthesis and a unique sensitivity to the effects of iodine.29 It may be that iodide-induced goiter or hypothyroidism occurs only in the presence of thyroid dysfunction. The true incidence of goiter in patients exposed to iodides is unknown. The incidence is higher in women than in men, and most patients have received large amounts of iodide (several milligrams to >1 g) daily for long periods.3 Goiter also has been seen after therapy with small amounts of iodide, however.12 The interval from the institution of therapy with iodine-containing compounds to the appearance of goiter or hypothyroidism varies widely, from a few months to several years. EFFECTS DURING PREGNANCY Iodides cross the placenta easily; therefore, one should expect that goiter may be seen in neonates whose mothers received iodide-containing drugs during pregnancy. Congenital goiter and neonatal hypothyroidism resulting from long-term maternal iodide use are rare and have similar manifestations as in adults, but the consequences for infants are severe and include asphyxiation from tracheal compression and potential mental retardation if the condition is not treated. Death from asphyxiation was reported in 8 of 22 infants born with goiter.11 Surgical intervention may be warranted in selected cases, although the goiters resolve over a few months in asymptomatic infants. Generally, the mothers have had no manifestations of goiter or hypothyroidism. NEONATAL PERIOD The frequency or magnitude of the problem of neonatal iodide-induced goiter or hypothyroidism is unknown. Iodides are secreted into breast milk, and hypothyroidism or goiter may be seen in breast-fed infants of mothers chronically exposed to high doses of iodides.11,30 Prenatal vitamins usually contain 150 g of iodine, which should not constitute a hazard but does pose an additional risk in pregnant women who already are ingesting iodides from other sources. In an intensive care unit setting, within 1 month after the extensive topical use of 1% iodine and 3.9% iodine solutions (Polyvidone) for 4 to 8 days, goiter and hypothyroidism developed in 5 of 30 neonates.31 The affected infants had significantly increased ioduria, whereas the other infants did not, which suggests a difference in skin permeability. In all cases, the goiter resolved after brief therapy with thyroid hormone, and neither hypothyroidism nor goiter recurred subsequently. These observations suggest inadequate or undeveloped ability of the neonatal thyroid gland to autoregulate or control iodide uptake and binding, resulting in saturation and suppression of hormone synthesis and secretion. COAST GOITER Coast goiter, the endemic iodide goiter seen in the northern Japanese island of Hokkaido, is the most common form of iodide goiter.12,32 It occurs in 6% to 12% of that population, most commonly among the seaweed harvesters and their families, who may consume 50 to 200 mg of iodide daily. In a survey of schoolchildren, 508 goiters were found, 330 in girls and 178 in boys. Nearly all the glands (97.4%) were diffusely enlarged, and a histologic diagnosis of colloid goiter was rendered in the seven glands that underwent biopsy. Despite having extremely large goiters, these patients all have been euthyroid. When seaweed was withdrawn from the diet, the elevated plasma and urinary iodide levels returned toward normal, resulting in a rebound increase of 131I uptake 1 to 2 weeks later.32 In the few patients studied, the goiters disappeared or significantly decreased in size after seaweed withdrawal (Fig. 37-1). In 75% of 50 patients treated with thyroid hormone who maintained their usual dietary habits, the goiters decreased in size or resolved.33 The mechanism underlying the absence of hypothyroidism in these patients is not well understood, but they appear to reach a new steady state, manifested by a normal serum T4 level and goiter, presumably maintained by TSH. The erratic availability of seaweed also may be important because seaweed is a significant part of the diet only during good harvesting weather.12 A similar incidence of hyperthyroidism in five coastal regions of Japan has been reported, along with a widely varying incidence of hypothyroidism (09.7%). High urinary iodine concentration correlated with antibody-negative hypothyroidism.34

FIGURE 37-1. A 14-year-old patient from Hok-kaido, Japan, with usual daily diet of 50 to 80 mg of iodine derived from seaweed (kelp). A, Appearance at time of first presentation with coast goiter. B, Appearance 3 weeks after removal of seaweed from the diet. C, Appearance after 2 months of treatment with thyroid hormone. The goiter has regressed further. (From Suzuki H, Higuchi T, Sawa K, et al. Endemic coast goiter in Hokkaido, Japan. Acta Endocrinol [Copenh] 1965; 50:161.)

PREDISPOSITION TO INDUCED HYPOTHYROIDISM Congenital and acquired defects in intrathyroidal organification increase the risk of induced hypothyroidism with iodine exposure. Patients with simple goiter resulting from organification defects appear to be highly susceptible to iodide-induced hypothyroidism. A prospective study of biochemically euthyroid patients with Hashimoto thyroiditis who were treated with iodides revealed a high incidence of iodide-related hypothyroidism that was reversible on cessation of iodides.29 These thyroid glands probably fail to escape from the acute inhibitory effect of iodide on the organification process and manifest a persistence of the Wolff-Chaikoff effect. The increasing incidence of Hashimoto thyroiditis with hypothyroidism or goiter in the United States may be the result of an unmasking of the organification defect with a chronically high dietary iodide intake. Hypothyroidism also has been seen in euthyroid patients given iodides who previously underwent 131I or surgical therapy for Graves disease.11,35 With withdrawal of iodides, the euthyroid state was restored, and TSH levels normalized. The speculation is that this phenomenon is the result of an organification defect caused by the radiation therapy. Although hypothyroidism after iodide exposure was seen in all patients treated with 131I, it also was seen in 40% of surgically treated patients.35 Thus, an underlying defect related to Graves disease that can be compounded by a radiation-induced sensitivity of the thyroid gland to an iodide load appears to exist. CLINICAL FEATURES OF IODIDE-INDUCED HYPOTHYROIDISM The clinical presentation of iodide-induced hypothyroidism is variable. Goiter alone is much more common than is hypothyroidism, and the thyroid gland may be modestly or significantly enlarged. Generally, the goiter is diffuse, but in the context of antithyroidal antibodies, a nodular gland may be present. A history of previous Hashimoto disease, Graves disease, or goiter should raise the suspicion of an iodide-induced abnormality of thyroid function. The symptoms and signs of hypothyroidism may be subtle or absent.10,34,36 In the presence of normal to marginal values for serum T4 and basal TSH, the serum TSH response to thyrotropin-releasing hormone affords the most sensitive indicator of hypothyroidism. Iodine-131 uptake may not be a helpful measure, depending on the amount of exogenous iodide ingested and the functional capabilities of the thyroid gland. In a setting of high iodide intake, a low uptake of radioiodine is expected. Low, normal, and elevated uptakes have all been seen, however, with iodide-induced goiter. An early peak in 131I uptake has been described, with the 4-hour or 6-hour uptake higher than the 24-hour uptake, a typical finding in patients with organification defects. The best evidence that goiter or hypothyroidism is related to excess iodides is return to normal function and size of the gland on withdrawal of exogenous iodide. Patients at highest risk are those with underlying thyroid abnormalities, Hashimoto disease, or previously treated Graves disease. A Japanese study of iodine-induced hypothyroidism indicated that patients with hypothyroidism that was reversible after iodine withdrawal had underlying focal lymphocytic thyroiditis, whereas those with irreversible hypothyroidism had more severe destruction of the thyroid gland.37

IODIDE-INDUCED THYROTOXICOSIS
The occurrence of hyperthyroidism after iodide administration has been reported throughout the world since this element was first used to treat endemic iodine-deficient goiter. The first case report of thyrotoxicosis after iodine supplementation was published in 1821. Iodide-induced hyperthyroidism subsequently has been referred to as

the jodbasedow phenomenon. Thyrotoxicosis occurring after iodide supplementation for endemic goiter in iodine-deficient areas is frequently reported. The mechanism underlying this phenomenon may relate to the rapid iodination and proteolysis of previously iodine-poor thyroglobulin, or to the presence of a subpopulation of patients with goiters who have gross or microscopic autonomous areas of functioning tissue. The latter hypothesis is supported by the observation that hyperthyroidism may develop after an iodine load in euthyroid patients with solitary autonomous nodules.38 Four populationsin Holland, Tasmania, Yugoslavia, and Australiawere observed closely for the prevalence of thyrotoxicosis before and after iodide supplementation. The incidence of thyrotoxicosis after iodide supplementation ranged from .01% to .04%.14 The incidence of thyrotoxicosis in a population rises 6 months after long-term iodine exposure begins, peaks at 1 to 3 years, and returns to baseline by 6 to 10 years. Although several reports have indicated no increased incidence of thyrotoxicosis in some populations given iodine prophylactically, these follow-up studies may have been conducted too late to detect transient hyperthyroidism. Reviews have addressed the prevention of iodine-induced thyrotoxicosis39 and its epidemiology in Europe and worldwide.40 Iodine-induced thyrotoxicosis has been reported from areas of endemic iodine deficiency in Germany after the use of iodine-containing drugs or contrast media, in much the same manner as was seen when iodine was added as a dietary supplement.41 The reported 5% to 30% incidence in this area of endemic goiter in Germany is much higher than that seen in iodine-sufficient areas. Follow-up studies on patients exposed to iodine-containing drugs or dyes indicate that serum iodothyronine concentrations increase rapidly in relation to iodine dose, with a doubling of free T4 index and a 53% increase in serum T3 at a time when urinary iodine excretion is increased 5-fold to 10-fold over baseline.42 Additional exposure to excess iodine intake further increased serum iodothyronine concentrations, albeit less dramatically.42 In addition, the incidence of T3 toxicosis relative to that of T4 toxicosis declined with increasing iodine intake. In the United States, concern exists that iodine-induced thyrotoxicosis may occur in persons without any underlying goiter or thyroid function abnormality. The incidence of goiter is ~3% in this country, and no areas of iodine deficiency are found. Dozens of cases of drug-induced thyrotoxicosis have been reported in the United States. Far more cases have come from Europe, although these may be from areas of relative dietary iodine deficiency. Several of the patients described in the United States have had preexisting thyroid disorders, most have had multinodular goiters, and most have been women. Drug-induced thyrotoxicosis has been seen after the use of potassium iodide, oral and intravenous contrast agents, the iodine-containing antiarrhythmic amiodarone (see preceding), and the topical antiinfective iodochlorhydroxyquin. Hyperthyroidism has been observed to develop 4 days to 4 months after iodine exposure. The thyrotoxicosis may persist for 1 to 6 months and usually is self-limited. In one study, thyrotoxicosis developed in four of eight patients with goiter after a 180-mg daily dose of potassium iodide, and the suggestion was made that thyrotoxicosis may be a common outcome in patients from nonendemic areas with goiter who are exposed to iodides.43 The authors recommended that caution be exercised in administering iodine-containing drugs or performing radiographic contrast procedures in persons with goiter. Patients with multinodular goiter are at some increased risk for iodine-induced thyrotoxicosis. The risk may be overstated, however; despite the high prevalence of goiter and the several million dye studies that are done yearly in this country, an epidemic of thyrotoxicosis has not occurred. Thus, most patients with goiter are not clinically affected by iodine exposure, and goiter should not be considered a contraindication to the use of necessary iodine-containing drugs or contrast media. In a study of 10 women who were in remission after a prior history of Graves thyrotoxicosis, the administration of SSKI caused recurrence of thyrotoxicosis in two patients and blunting of the thyrotropin-releasing hormone stimulation test in two others.44 Iodine-induced thyrotoxicosis also has occurred with metastatic thyroid carcinoma after earlier thyroidectomy.45 CLINICAL FEATURES OF IODIDE-INDUCED THYROTOXICOSIS The clinical features of iodide-induced thyrotoxicosis in endemic goiter are similar to those seen in Graves disease with thyrotoxicosis. In the former case patients tend to be older, however, and to have long histories of iodide deprivation preceding the addition of iodine. Some predisposition may be present with advancing age, because thyrotoxicosis occurred in none of 50,000 children in this country who were treated with iodides in the 1920s. The ratio of men to women affected may also be different from the ratio in Graves disease. In Yugoslavia, the ratio of men to women with thyrotoxicosis increased from 1:10 to 1:6 after iodine supplementation. In Holland, 15% of the cases of hyperthyroidism occurred in men before iodine supplementation, compared to 31% after iodine supplementation. Exophthalmos is not encountered, and thyroid antibodies are not detected. Among patients with iodide-induced thyrotoxicosis from iodine-deficient areas, most patients had long-standing nodular goiter; however, 15% to 30% of patients had minimal to no goiter, and many persons with goiter did not have nodules. The clinical manifestations of drug-induced thyrotoxicosis are similar to those of Graves hyperthyroidism and include tremor, tachycardia, palpitations, weight loss, heat intolerance, concentration difficulties, and mental status changes.19 There is no tenderness of the thyroid gland such as is seen in granulomatous thyroiditis; there is no exophthalmos such as is seen in Graves disease. The serum T4 and free T4 values and the free T4 index are elevated. Typically, the 24-hour 131I uptake is reduced, and the technetium thyroid scan may show reduced or patchy uptake. Generally, the T3 level also is elevated; however, with amiodarone and the oral cholecystographic dye ipodate, the peripheral conversion of T4 to T3 is blocked, and the serum T3 level may be disproportionately low compared to the elevated serum T4 level. TREATMENT OF IODIDE-INDUCED THYROTOXICOSIS The treatment of thyrotoxicosis is discussed in Chapter 42. The major difference in the approach to treating iodide-induced hyperthyroidism as opposed to Graves disease relates to the self-limited nature of the former, with resolution occurring after removal of the source of excess iodine and clearance of iodine from the circulation. The thyrotoxicosis, although usually self-limited, may persist for weeks to months because of the increased thyroidal iodine stores and increased plasma iodide. Mildly symptomatic patients may be treated with b-blockade therapy alone; more specific antithyroid drugs may be necessary in severely affected patients, but ablative therapy should not be required. Amiodarone iodine-induced thyrotoxicosis may be resistant to the usual antithyroidal therapy.46 A study of 58 patients with amiodarone-induced thyrotoxicosis showed persistent thyrotoxicosis at 6 to 9 months in more than one-half of those who were not treated. The use of methimazole in high doses proved ineffective in most patients. Euthyroidism was restored, however, in all patients who were treated with a combination of methimazole and potassium perchlorate (1 g per day). Other authors have found prolonged persistent thyrotoxicosis even with the addition of potassium perchlorate to propylthiouracil.47 Successful resolution of the hyperthyroidism may require prolonged therapy. Of note for those cases in which the underlying cardiac dysrhythmia precludes safe discontinuation of the drug, the thyrotoxicosis was treated successfully in five patients by adding only thiourea.48 Although most cases of amiodarone-induced thyrotoxicosis are related to the provision of its rich iodine content to patients who have thyroid glands with underlying autonomous function (e.g., hyperfunctioning nodules), a variant of the syndrome that is akin to destructive thyroiditis exists.49 Patients with this condition, who may have low radioiodine uptake and elevated serum interleukin-6 levels,50 should be treated with glucocorticoids rather than methimazole.51

IODINE-INDUCED AUTOIMMUNITY
Investigators have speculated that increased iodine intake may contribute to the prevalence of autoimmune thyroid disease. Individuals with Hashimoto disease are well known to be sensitive to the effects of iodine and vulnerable to iodine-induced hypothyroidism. Less well known is that iodine exposure may induce a lymphocytic thyroiditis in a variety of experimental animals (rat, hamster, dog) as well as in humans.9,52 A few studies have examined the greater prevalence of thyroid antibodies after increases in iodine exposure.53 Specific histologic features of iodine-induced goiter have been described in one study of 28 patients, with half the biopsies also showing typical lymphocytic infiltration.54 Persons with preexisting thyroid disease who live in areas of iodide deficiency may be predisposed to the development of iodine-induced thyrotoxicosis.52 Iodine-rich thyroglobulin has been shown to be more immunogenic,55 and a greater proliferation of T cells is seen in patients with chronic thyroiditis after exposure to normally iodinated thyroglobulin, whereas noniodinated thyroglobulin elicits no response.56 In some unknown manner, iodine may relate to the development of thyroid growth-stimulating immunoglobulins.57 In this regard, iodide has been shown to enhance immunoglobulin G release into the media of cultured human lymphocytes,58 but other mechanisms of immune stimulation may exist.56 Controversy exists as to whether amiodarone may induce autoimmunity and, if so, whether the drug itself or its iodine content is responsible. Thirty percent to 50% of patients who become hypothyroid during amiodarone treatment have antithyroid antibodies in their serum. In one prospective study, antibodies developed in 6 of 13 patients given amiodarone, compared to none of 22 control subjects.59 The proposal has been made that amiodarone alters T-cell subsets with enhanced T-cell expression of Ia antigen, consistent with precipitation of organ-specific autoimmunity in susceptible persons.60 CHAPTER REFERENCES
1. 2. 3. 4. Braverman LE. Iodine and the thyroid: 33 years of study. Thyroid 1994; 4:351. Koutras DA. Control of efficiency and results and adverse effects of excess iodine administration on thyroid function. Ann Endocrinol (Paris) 1996; 57:463. Khan LK, Li R, Gootnick D, et al. Thyroid abnormalities related to iodine excess from water purification units. Lancet 1998; 352:1519. Hollowell JG, Staehling NW, Hannon WH, et al. Iodine nutrition in the United States. Trends and public health implications: iodine excretion data from national health and nutrition examination surveys I and III (1971 1974 and 19881994). J Clin Endocrinol Metab 1998; 83:3401. 5. Cavalieri RR. Iodine metabolism and thyroid physiology: current concepts. Thyroid 1997; 7:177. 6. Ermans AM. Disorders of iodine deficiency. In: Ingbar SH, Braverman LE, eds. The thyroid: a fundamental and clinical text. Philadelphia: JB Lippincott, 1986:707. 7. Peacock I, Davison H. Observations of iodide sensitivity. Ann Allergy 1958; 16:158.

8. 9. 10. 11. 12. 13. 14. 15.

Edmunds HT. Acute thyroiditis from potassium iodide. BMJ 1955; 1:354. Bagchi N, Brown TR, Urdanivia E, Sundick RS. Induction of autoimmune thyroiditis in chickens by dietary iodine. Science 1985; 230:325. Sato K, Okamura K, Hirata T, et al. Immunological and chemical types of reversible hypothyroidism; clinical characteristics and long-term prognosis. Clin Endocrinol 1996; 45:519. Vagenakis AG, Braverman LE. Adverse effects of iodides on thyroid function. Med Clin North Am 1975; 59:1075. Wolff J. Iodide goiter and the pharmacologic effects of excess iodide. Am J Med 1969; 47:101. Namba H, Yamashita S, Kimura H, et al. Evidence of thyroid volume increase in normal subjects receiving excess iodide. J Clin Endocrinol Metab 1993; 76:605. Fradkin JE, Wolff J. Iodide induced thyrotoxicosis. Medicine (Baltimore) 1983; 62:1. Savoie J-C, Massin JP, Thomopoulos P, Leger F. Iodine induced thyrotoxicosis in apparently normal glands. J Clin Endocrinol Metab 1975; 4:685.

15a. Jooste PL, Weight MJ, Lombard CJ. Short-term effectiveness of mandatory iodization of table salt, at an elevated iodine concentration, on the iodine and goiter status of school children with endemic goiter. Am J Clin Nutr 2000; 71:75. 16. Park YK, Harland BF, Vandervein JE, et al. Estimation of dietary iodine intake of Americans in recent years. J Am Diet Assoc 1981; 79:17. 17. Burgi H, Wimpfheimer C, Burger A, et al. Changes of circulating thyroxine, triiodothyronine, and reverse triiodothyronine after radiographic contrast agents. J Clin Endocrinol Metab 1976; 43:1203. 18. Nagataki S. Effect of excess quantities of iodide. In: Greer MA, Solomon DH, eds. Handbook of physiology, section 7, Endocrinology, VIII. Baltimore: Williams & Wilkins, 1974:329. 19. Safran M, Braverman LE. Effect of chronic douching with polyvinyl-pyrrolidone-iodine on iodine absorption and thyroid function. Obstet Gynecol 1982; 60:35. 20. Rajatanavin R, Safran M, Stoller WA, et al. Five patients with iodine-induced hyperthyroidism. Am J Med 1984; 77:378. 21. Loh KC. Amiodarone-induced thyroid disorders: a clinical review. Post-grad Med J 2000; 76:133. 22. Harjai KJ, Licata AA. Effects of amiodarone on thyroid function. Ann Intern Med 1997; 126:63. 23. Iudica-Souza C, Burch HB. Amiodarone-induced thyroid dysfunction. The Endocrinologist 1999; in press. 24. Leger AF, Massin JP, Laurent ME, et al. Iodine-induced thyrotoxicosis: analysis of eighty-five consecutive cases. Eur J Clin Invest 1984; 14:449. 25. Paul T, Myers B, Witorsch RJ, et al. The effect of small increases in dietary iodine on thyroid function in euthyroid subjects. Metabolism 1988; 37:121. 26. Wolff J, Chaikoff IL. Plasma inorganic iodide as a homeostatic regulator of thyroid function. J Biol Chem 1948; 174:555. 27. Gardner DF, Centor RM, Utiger RD. Effects of low dose oral iodide supplementation on thyroid function in normal men. Clin Endocrinol (Oxf) 1988; 28:283. 28. Wartofsky L, Ransil BJ, Ingbar SH. Inhibition by iodine of the release of thyroxine from the thyroid glands of normal subjects and patients with thyrotoxicosis. J Clin Invest 1970; 49:78. 29. Braverman LE, Ingbar SH, Vagenakis AG. Enhanced susceptibility to iodide myxedema in patients with Hashimoto's disease. J Clin Endocrinol Metab 1971; 32:515. 30. DeLange F, Chanoine JP, Abrassart C, Bourdoux P. Topical iodine, breast-feeding, and neonatal hypothyroidism. Arch Dis Child 1988; 63:106. 31. Chabrolle JP, Rossier A. Goiter and hypothyroidism in the newborn after cutaneous absorption of iodine. Arch Dis Child 1978; 53:495. 32. Suzuki H, Higuchi T, Sawa K, et al. Endemic coast goiter in Hokkaido, Japan. Acta Endocrinol (Copenh) 1965; 50:161. 33. Higuchi T. The study of endemic seashore goiter in Hokkaido. Folia Endo-crinol Jpn 1964; 40:982. 34. Konno N, Makita H, Yuri K, et al. Association between dietary iodine intake and prevalence of subclinical hypothyroidism in the coastal regions of Japan. J Clin Endocrinol Metab 1994; 78:393. 35. Braverman LE, Walker KA, Ingbar SH. Induction of myxedema by iodide in patients euthyroid after radioiodine or surgical treatment of diffuse toxic goiter. N Engl J Med 1969; 281:816. 36. Lesher JL Jr, Fitch MH, Dunlap DB. Subclinical hypothyroidism during potassium iodide therapy for lymphocutaneous sporotrichosis. Cutis 1994; 53:128. 37. Tajiri J, Higashi K, Morita M, et al. Studies of hypothyroidism in patients with high iodine intake. J Clin Endocrinol Metab 1986; 63:412. 38. Ermans AM, Camus M. Modification of thyroid function induced by chronic administration of iodide in the presence of autonomous thyroid tissue. Acta Endocrinol (Copenh) 1972; 70:463. 39. Dunn JT, Semigran MJ, Delange F. The prevention and management of iodine-induced hyperthyroidism and its cardiac features. Thyroid 1998; 8:101. 40. Stanbury JB, Ermans AE, Bourdoux P, et al. Iodine-induced hyperthyroidism: occurrence and epidemiology. Thyroid 1998; 8:83. 41. Kallee E, Wahl R, Bohner J, et al. Thyrotoxicosis induced by iodine-containing drugs. J Mol Med 1980; 4:221. 42. Emrich D, Karkavitsas N, Facorro U, et al. Influence of increasing iodine intake on thyroid function in euthyroid and hyperthyroid states. J Clin Endocrinol Metab 1982; 54:1236. 43. Vagenakis AG, Wang C, Burger A, et al. Iodide-induced thyrotoxicosis in Boston. N Engl J Med 1972; 287:524. 44. Roti E, Gardini E, Minelli R, et al. Effects of chronic iodine administration on thyroid status in euthyroid subjects previously treated with antithyroid drugs for Graves' hyperthyroidism. J Clin Endocrinol Metab 1993; 76:928. 45. Yoshinari M, Tokuyama T, Okamura K, et al. Iodide-induced thyrotoxicosis in a thyroidectomized patient with metastatic thyroid carcinoma. Cancer 1988; 61:1674. 46. Martino E, Aghini-Lombardi F, Mariotti S, et al. Amiodarone: a common source of iodine induced thyrotoxicosis. Horm Res 1987; 26:158. 47. Newnham HH, Topliss BJ, LeGrand BA, et al. Amiodarone-induced hyperthyroidism: assessment of the predictive value of biochemical testing and response to combined therapy using propylthiouracil and potassium perchlorate. Aust N Z J Med 1988; 18:37. 48. Davies PH, Franklyn JA, Sheppard MC. Treatment of amiodarone induced thyrotoxicosis with carbimazole alone and continuation of amiodarone. BMJ 1992; 305:224 49. Smyrk TC, Goellner JR, Brennan MD, Carnei JA. Pathology of the thyroid in amiodarone-associated thyrotoxicosis. Am J Surg Pathol 1987; 11:197. 50. Bartalena L, Grasso L, Brogioni S, et al. Serum interleukin-6 in amiodarone-induced thyrotoxicosis. J Clin Endocrinol Metab 1994; 78:423. 51. Bartalena L, Brogioni S, Grasso L, et al. Treatment of amiodarone-induced thyrotoxicosis, a difficult challenge: results of a prospective study. J Clin Endocrinol Metab 1996; 81:2930. 52. Harach HR, Escalante DA, Onativia A, et al. Thyroid carcinoma and thyroiditis in an endemic goitre region before and after iodine prophylaxis. Acta Endocrinol (Copenh) 1985; 108:55 53. Koutras DA, Karaiskos KS, Evangelopoulou K, et al. Thyroid antibodies after iodine supplementation. In: Drexhage HA, Wiersinga WM, eds. The thyroid and autoimmunity. Amsterdam: Excerpta Medica, 1986:211. 54. Mizukami Y, Michigishi T, Nonomura A, et al. Iodine-induced hypothyroidism: a clinical and histological study of 28 patients. J Clin Endocrinol Metab 1993; 76:466. 55. Sundick RS, Herdegen DM, Brown TR, Bagchi N. The incorporation of dietary iodine into thyroglobulin increases its immunogenicity. Endocrinology 1987; 120:2078. 56. Rose NR, Saboori AM, Rasooly L, Burek CL. The role of iodine in autoimmune thyroiditis. Crit Rev Immunol 1997; 17:511. 57. Medeiros-Neto GA, Halpern A, Cozzi ZS, et al. Thyroid growth immunoglobulins in large multinodular endemic goiters: effect of iodized oil. J Clin Endocrinol Metab 1986; 63:644. 58. Weetman AP, McGregor AM, Campbell H, et al. Iodide enhances IgG synthesis by human peripheral blood lymphocytes in vitro. Acta Endocrinol (Copenh) 1983; 103:210. 59. Monteiro E, Galvao Teles A, Santos ML, et al. Antithyroid antibodies as an early marker for thyroid disease induced by amiodarone. BMJ 1986; 292:227. 60. Rabinowe SL, Larsen PR, Antman EM, et al. Amiodarone therapy and autoimmune thyroid disease. Am J Med 1986; 81:53.

CHAPTER 38 NONTOXIC GOITER Principles and Practice of Endocrinology and Metabolism

CHAPTER 38 NONTOXIC GOITER

PAUL J. DAVIS AND FAITH B. DAVIS Pathogenesis of Nontoxic Goiter Specific Etiologies of Goiter Congenital Goiter Endemic Goiter Sporadic Goiter Clinical Features of Nontoxic Goiter History and Physical Findings Nonthyroidal Anterior Cervical Neck Masses Laboratory Evaluation of the Patient with Goiter Establishment of the Cause of Nontoxic Goiter Prognosis of Nontoxic Goiter Treatment of Nontoxic Goiter Substernal Goiter Chapter References

Goiter (L. guttur, throat) is a clinical term denoting thyroid gland enlargement to twice normal size or larger (see also ref. 1). The term implies no specific etiology, although the mechanisms of certain toxic and nontoxic (eumetabolic) goitrous states are understood. The prevalence of nontoxic goiter in the United States is ~5%,1a with a predominance among females. In the Framingham, Massachusetts, study, the prevalence of non-toxic goiter was 4.2%, with a 4:1 female to male ratio.2

PATHOGENESIS OF NONTOXIC GOITER

Nontoxic goiter is the result of an interplay of factors intrinsic and extrinsic to the thyroid. In most patients, subtle limitations on hormonogenesisimposed, for example, by iodine insufficiency or constitutional biochemical abnormalities in the glandlead to nontoxic goiter when permissive normal or increased circulating levels of thyroid-stimulating hormone (TSH, thyrotropin) foster gland enlargement. Impaired hormonogenesis, TSH-dependent goiter, and eumetabolism may also reflect the presence of Hashimoto thyroiditis or the ingestion of goitrogens. In the absence of TSH, as in hypopituitarism, nontoxic goiter is rare. Some have argued that the mechanisms of diffuse and nodular nontoxic goiter are different, with the latter reflecting an autonomous state independent of the permissive effect of TSH.3 Although gland autonomy is present in certain patients with nontoxic nodular goiters, it occurs in few such patients, and whether autonomy develops early or late in the course of goitrogenesis is unknown.4 The suggestion has also been made that the cyclic hypersecretion of TSH in response to periodic underproduction of hormone by the thyroid gland is essential to the development of nodular goiter. Formation of fibrous strands or scar tissue (consequent to cyclic follicular hyperplasia, microscopic hemorrhage, and necrosis) has been postulated to impose the anatomical limits within which nodules form.5 Nevertheless, the possibility seems likely that subtle constitutional derangements of hormonogenesis and the permissive action of circulating TSH act together to produce both diffuse and nodular nontoxic goiter. Various abnormalities of hormone production have been described in cases of nontoxic goiter (Table 38-1).6,7,8,9,10,11 and 12,12a These biochemical abnormalities, or altered sensitivity to TSH, might be expressed only in certain follicles (biochemical follicular heterogeneity), which would explain the occurrence of nodular goiter. Excluded from Table 38-1 are genetically determined biochemical lesions that exist in familial goiter with hypothyroidism (see Chap. 47).

TABLE 38-1. Biochemical Findings Described in Nontoxic Goitrous Tissue

Other mechanisms leading to the development of goiter are infiltrative diseases of the thyroid, such as amyloidosis or granulomatous diseases, and the action of nonpituitary thyrotropic substances. Among the latter are polypeptides that circulate in pregnancy (human chorionic gonadotropin [hCG] and, less importantly, a placental TSH-like substance13) and thyroid-stimulating antibodies (see Chap. 197). Some of these thyroid-stimulating immunoglobulins (hTSI) have been identified in human serum14,15,16,17,18,19,20,21 and 22 and act either through the thyroid cell membrane TSH receptor or by enhancing thyroid cell activity independently of the TSH receptor. The hTSIs that interact with the TSH receptor are recognized in vitro in a competitive binding assay involving radiolabeled TSH and thyroid cell membranes (TSH-binding inhibitory immunoglobulin [TBII] assay). Thyroid-stimulating antibodies that are inactive in the TBII assay are recognized by their growth effects, measured in vitro by sensitive histocytochemical methods or other methods. The hTSIs that cross-react with an antigenic site at the TSH receptor are probably unimportant in the pathogenesis of nontoxic goiter,15 although they play a role in diffuse toxic goiter. On the other hand, thyroid-stimulating antibodies that act independently of the TSH site are found in the sera of more than 50% of patients with nontoxic goiter. Although these observations support an autoimmune contribution to the pathogenesis of nontoxic goiter, a search for helper T lymphocytes sensitized to thyroid membrane antigens in the peripheral blood of patients with nontoxic nodular goiter yielded negative results.15 Hence, the importance of hTSIs in the formation of nontoxic goiter is not yet clear. Certain cytokines have been implicated in the pathogenesis of nontoxic goiter. For example, the content of insulin-like growth factor-I (IGF-I) is increased more than two-fold in thyroid nodules compared with normal tissue obtained from the same gland.17 Thyroid glands from patients with Graves disease and Hashimoto thyroiditis show no increase in IGF-I. Transforming growth factor-b (TGF-b) has been proposed to be an autocrine growth inhibitor in thyroid follicular cells, at least in the setting of iodide deficiency.18 TGF-b blocks the growth stimulation, measured as thymidine incorporation, imposed on thyroid follicular cells in vitro by IGF-I, epidermal growth factor, and TGF-a, and endogenous TGF-b levels are decreased in iodide-deficient nontoxic goiter.18 Although these studies suggest that cytokines can promote or mediate goiter formation, evidence exists that at least in goiter due to iodine-insufficiency, thyroid growthpromoting factors are not detectable in serum.23 The goiter of pregnancy may be a response both to placental TSHs and to maternal iodide insufficiency. The latter is a complex function of increased maternal renal iodide clearance, iodide parasitism by the fetus, and maternal losses during lactation.

SPECIFIC ETIOLOGIES OF GOITER

CONGENITAL GOITER Congenital goiter is either familialthat is, an expression of genetic disorders of intrathyroidal hormonogenesis (see Chap. 47)or sporadic. In the neonate, sporadic congenital goiter reflects intrauterine iodide insufficiency or fetal exposure to goitrogens. Maternal consumption of naturally occurring goitrogens is rarely the cause of thyroid enlargement in the neonate; this contrasts with endemic goiter, in which community-wide ingestion of a goitrogen occurs. Use of antithyroid medications is a significant cause of sporadic congenital goiter (i.e., thionamide administration to pregnant thyrotoxic women). Thionamides cross the placental barrier and limit fetal hormonogenesis after the 12th week of gestation. The requisites for such a congenital goiter include fetal TSH production, maturation of the fetal thyroid to the point of TSH responsiveness, and sensitivity of the pituitary to low levels of thyroid hormone. Because insubstantial quantities of maternal thyroxine (T4) or triiodothyronine (T3) cross the placenta, the maternal ingestion of T4 or T3 concomitantly with thionamides does not prevent fetal goiter. A specialized metabolically active analog,

3,5-dimethyl-3-isopropyl-L -thyronine (DIMIT),24 does cross the placenta, but it is not available for clinical use. Other iatrogenic goitrogens, such as aminoglutethimide and carbutamide, do not appear to cause fetal goiter, and neither agent is likely to be prescribed to pregnant women. Lithium and amiodarone, both potential causes of fetal goiter, are contraindicated during pregnancy. ENDEMIC GOITER Iodine deficiency has been extensively studied as a cause of endemic goiter.25 Before the introduction of iodized salt in 1925, goiter belts of low iodine content in water supplies existed in the United States near the Great Lakes, in the Appalachian region, and elsewhere.26 During World War I, a 5% incidence of goiter was seen among U.S. Army inductees,27 presumably reflecting widespread iodine deficiency. By World War II, the nationwide effort to increase dietary iodine had reduced the incidence of goiter in military recruits to 0.06%. Dietary iodine insufficiency no longer is a significant cause of euthyroid goiter in the United States but persists in enclaves in South America, Africa, and Asia,1,28 and as many as 200 million people worldwide are estimated to experience iodine-deficiency goiter. The excessive dietary intake of iodine is an occasional but nonendemic cause of goiter in the United States. In a localized region in Japan, however, high iodine intake has caused endemic goiter. The inhibition of thyroid hormonogenesis by excess iodine is termed the Wolff-Chaikoff effect. Severe fetal iodine deprivation culminates in endemic and sporadic cretinism (see Chap. 47), with distinctive clinical findings. Moderate fetal iodine insufficiency results in euthyroid fetal goiter. Iodide crosses the placenta, and parasitism of maternal iodide by the fetus contributes to the goiter of pregnancy in women whose dietary iodine intake is marginal. Both well-characterized and poorly characterized dietary goitrogens have been reported in a variety of epidemiologic studies. Water supplies have been a frequent source.29 Certain vegetables (e.g., cabbage and other members of the Brassica familyturnip, kale, rape) contain thionamide-like substances.1 Cassava is an important goitrogen in Africa, but no evidence exists for dietary goitrogens as a cause of endemic goiter in the United States. Most patients with endemic goiter have no constitutional abnormalities of thyroid hormonogenesis. The goiter formation reflects either extrinsic iodine deficiency or goitrogen ingestion. SPORADIC GOITER As with endemic goiter, sporadic goiter may reflect variations in iodide intake30 and dietary goitrogen content. Although iodine insufficiency rarely causes thyroid enlargement in the United States, iodine excess can lead to goiter in two settings. First, although the thyroid gland escapes from the Wolff-Chaikoff effect in normal subjects after several weeks of excessive iodine intake, the thyroid in some patients does not. Resulting defective hormonogenesis and release cause increased pituitary TSH secretion and goiter formation. A new steady state is achieved in which the enlarged, inefficient gland produces sufficient hormone to maintain euthyroidism. This proposed pathogenetic sequence, based on increased TSH production, seems likely, but this hypothesis has not been validated by prospective application of sensitive serum TSH assays to populations at risk. Patients at risk for goiter, due to failure to escape from iodine inhibition of thyroid hormonogenesis, are those with Hashimoto thyroiditis or with other gland damage syndromes, such as those due to thyroid surgery or radiation of the neck.31 A second setting in which goiter occurs with excess iodine administration is jodbasedow,32 a syndrome of latent hyperthyroidism that is rare in the United States (see Chap. 37). Increasing iodine intake promotes goiter and, nearly concomitantly, thyrotoxicosis. The source of increased iodine intake in goitrous patients is either medicinalfor example, saturated solution of potassium iodide (SSKI) continues to be used by some physicians as a mucolytic agent in patients with chronic lung disease, despite its clinical ineffectivenessor paramedicinalfor example, from kelp ingestion. Radiologic contrast media and skin antiseptics contain iodine, but, because of their short-term use, are not causes of goiter. Long-term iodine excess followed by the acute interruption of iodine intake can have a distinct effect on thyroid function tests. A variety of medications have antithyroid and goitrogenic effects (Table 38-2).33,34,35,36,37,38 and 39 The mechanisms of the antithyroid actions of these agents can be interference with hormonogenesis at the level of organification of iodide or inhibition of hormone release or both. Lithium, a monovalent cation, has an effect on the thyroid similar to that of the anion iodide: hormone release is inhibited after the drug is concentrated in the gland.36,40 Amiodarone, an antiarrhythmic drug, has been associated with the induction of hyperthyroidism and hypothyroidism.37 Most patients who receive the drug remain euthyroid. The effects of amiodarone appear to be largely due to its iodine content, and this agent appears to promote hyperthyroidism in patients with limited, possibly insufficient, iodine intake. Aminoglutethimide35 and oral sulfonylureas39 interfere with hormonogenesis. Among the sulfonylureas, only carbutamide, which is used in Europe, causes goiter. Aminoglutethimide, an antiadrenal agent used as medical treatment of states of excess adrenocortical hormone production, can cause hypothyroidism and goiter.

TABLE 38-2. Drugs Reported to Cause Goiter in Humans

Thyroiditis syndromes may be associated with goiter (see Chap. 46). The thyroid gland in these conditions is frequently symptomatic locally, and the glandular enlargement may be due to inflammatory infiltration rather than to TSH. The differential diagnosis of nontoxic goiter must always include early hyperthyroidism or, particularly in elderly subjects, monosystemic or apathetic hyperthyroidism. In these states, clinical hypermetabolism may be absent and isolated features such as myopathy, cardiomyopathy, or weight loss may be encountered41 (see Chap. 199). Pregnancy is also a cause of nontoxic goiter, which usually remits postpartum. After multiple pregnancies, substantive thyromegaly may occur. The possible contributions of other environmental factors to the development of nontoxic goiter are more difficult to evaluate. For example, cigarette smokingthe inhalation of thiocyanatehas been implicated in goitrogenesis,42 and serum thyroglobulin levels appear to be higher in smokers than in nonsmokers. The incidence of goiter in patients with end-stage renal disease is variable43 and unlikely to be explained by a single factor. The variability in incidence appears to be regional and may relate to diet or to the use in dialysis baths of tap water that contains ions or other factors that are goitrogenic.

CLINICAL FEATURES OF NONTOXIC GOITER

HISTORY AND PHYSICAL FINDINGS Patients with nontoxic goiter have no systemic symptoms, and the goiter is usually found on routine physical examination or is detected by the patient as a neck mass. Local symptoms occasionally include those due to tracheal compression (stridor or respiratory distress) and/or displacement of the esophagus (dysphagia). Rarely, substernal extension of nontoxic goiter may lead to recurrent laryngeal nerve entrapment and hoarseness. Spontaneous hemorrhage into a goiter manifests as local pain and tenderness in one thyroidal lobe. When Hashimoto thyroiditis causes nontoxic thyroidal enlargement, the gland is usually without local symptoms but diffuse tenderness or adenopathy occasionally occurs (see Chap. 46). History-taking from patients with nontoxic goiter includes a focused review of possible environmental or genetic factors that may be causative. The emphasis is on dietary or medicinal sources of excessive iodine intake, antithyroid drugs, or, in regions of the world where iodine deficiency or naturally occurring dietary goitrogens

are common, on establishment of a possible role for these factors. Goiter during pregnancy or the enhancement of established thyromegaly in successive pregnancies should raise the possibility of marginal iodine deficiency. Low-dose head-and-neck irradiation in childhood increases the risk of nodular goiter and thyroid carcinoma.44 Systemic illnesses such as sarcoidosis or amyloidosis raise the rare possibility of infiltrative disease of the thyroid. Isolated thyroid nodules in the euthyroid patient often indicate adenomas or cysts (see Chap. 39). Cancer originating in other organs may (rarely) metastasize to the thyroid.45 At autopsy, infection of the thyroid is relatively common in patients who have succumbed to septicemia, but such involvement is seldom clinically apparent before death.46 Physical examination of the nontoxic goiter should include the following: size, consistency, the absence or presence of nodularity, bruit, tenderness, and concomitant anterior cervical lymphadenopathy. The clinical and etiologic implications of each of these factors are controversial, but their accurate description is useful in the subsequent evaluation of the patient. Extreme firmness of a diffuse goiter is suggestive of infiltrative disease (amyloidosis, Riedel struma) or intense TSH stimulation (Hashimoto thyroiditis); firmness of a nodule has been suggested to be a sign of thyroid cancer. However, in the individual patient these generalizations are frequently incorrect. Tenderness of the thyroid gland can be a useful clinical finding, indicating inflammatory disease or intrathyroidal hemorrhage, usually into a preexisting cyst. Tenderness may also reflect acute thyroiditis contiguous with intercurrent tracheitis and occasionally is encountered in Hashimoto thyroiditis. Granulomatous or subacute thyroiditis is usually associated with exquisite thyroidal tenderness and pain radiating to the mandible. In all age groups, anterior cervical and supraclavicular adenopathy is a hallmark of locally metastatic papillary thyroid carcinoma (see Chap. 40). Because a centrally located node immediately above the isthmus of the thyroid gland drains both lobes, it has been ascribed oracular qualities (delphian node), usually reflecting the presence of thyroid cancer. However, this node is sometimes enlarged in patients with Hashimoto thyroiditis. Dullness to percussion over the upper sternum may be noted when a large substernal goiter is present. Instructive examples of nontoxic goiters are shown in Figure 38-1, and radiologic findings associated with substernal goiter are depicted in Figure 38-2.

FIGURE 38-1. Examples of nontoxic goiter. Top left, Sporadic multinodular goiter in an 85-year-old woman. No local symptoms were present. Top right, 65-year-old woman with a 40-year history of nontoxic goiter and thyrotoxicosis of recent onset. Multiple ablative doses of 131I controlled the hyperthyroidism but had no appreciable effect on thyroid gland size. Bottom right, Multinodular goiter with clinical findings of thoracic inlet obstruction and tracheal compression syndrome in a 70-year-old woman. The goiter was treated surgically to relieve obstructive symptoms. Bottom left, Endemic goiter in a 45-year-old Peruvian woman.

FIGURE 38-2. Radiographic and imaging studies in substernal goiter. Patient was a 28-year-old woman who presented with acute bronchitis. Left, Posteroanterior chest roentgenogram revealed right paratracheal mass contiguous with the aortic arch. Middle, Lateral roentgenogram of the chest showed tracheal compression (arrow) by anterior mediastinal mass. Right, Transverse superior mediastinal computed tomographic scan. The trachea (arrow) is enveloped and compressed by goiter, which extended posteriorly to the vertebral body.

NONTHYROIDAL ANTERIOR CERVICAL NECK MASSES

The thyroidal origin of neck masses is usually apparent from their location and their movement with the thyroid gland on deglutition. Previous neck surgery, obesity, and hypertrophied bellies of sternocleidomastoid muscles can obscure the thyroidal nature of neck masses. Included in the differential diagnosis of neck lesions are branchial cleft cyst and thyroglossal duct cyst (Fig. 38-3). Branchial cleft cysts are anterolateral, whereas thyroglossal duct cysts are midline and are found at any point between the isthmus and the area above the laryngeal prominence (cartilage). When thyroglossal duct cysts remain contiguous with the base of the tongue (foramen caecum linguae), they move upward with voluntary protrusion of the tongue. Cystic hygroma (diffuse, fluid-filled, multiloculated lymphangioma that is present at birth) arises from the supraclavicular fossa and should not be confused with goiter.

FIGURE 38-3. Abnormal nonthyroidal cervical structures included in the differential diagnosis of goiter. Left, Infected branchial cleft cyst in a 28-year-old woman with a neck mass and fever. The lateral, superior location in the neck is typical for this lesion. Right, Thyroglossal duct cyst in an asymptomatic 42-year-old woman. Lesion was smooth and moved with deglutition and with protrusion of the tongue. The midline or near-midline location and cystic consistency are typical for this lesion. The thyroid gland was not enlarged.

LABORATORY EVALUATION OF THE PATIENT WITH GOITER

The principles of diagnostic evaluation of the patient with non-toxic goiter are: (a) confirmation of eumetabolism; (b) establishment of the cause, insofar as doing so is cost effective; and (c) determination of the impact of the lesion on important non-thyroidal structures in the neck or upper mediastinum. Confirmation of the euthyroid state is essential, regardless of the chronicity of the goiter and its historically eumetabolic nature. The development of hyperthyroidism in the setting of long-standing nontoxic goiter is frequentthyrotoxicosis may develop in as many as one-third of patients with a large non-toxic goiter of 30 years' duration. Nevertheless, in a large series of elderly subjects with hyperthyroidism,41 patients with long-standing goiter comprised <10% of the subject population. Because diffuse nontoxic goiter, and occasionally nodular goiter, can reflect the presence of Hashimoto thyroiditis, the hypothyroid state may emerge insidiously during the course of a euthyroid goiter. Tests of serum thyroid function in the patient with nontoxic goiter should include TSH measurement. In patients with normal serum T4 concentrations, mild T3 toxicosis should be ruled out. Thyrotropin-releasing hormone testing is only rarely required to demonstrate euthyroidism in goitrous patients with nondiagnostic serum free T4 and total T3 values and low, but not fully suppressed, serum TSH levels. A substantial minority (39%) of patients with nontoxic goiter have abnormal esophageal transit times.47 This observation is not an indication for routine evaluation of esophageal function in goiter patients, because symptoms or medical consequences of the transit abnormality appear to be minimal,47 unless the goiter is very large.

ESTABLISHMENT OF THE CAUSE OF NONTOXIC GOITER

Serum antithyroid antibody titers should be measured in all patients with nontoxic goiter. Elevated antibody titers (e.g., anti-thyroglobulin and antithyroid peroxidase [anti-TPO, microsomal] antibodies) are prima facie evidence of an autoimmune basis for goiter formation in the euthyroid patient (e.g., those with Hashimoto thyroiditis). Elevated titers occur in up to 90% of such patients.48 In antibody-negative patients with nontoxic goiter, thyroidal radioiodide uptake and scan (iodine-123) may be helpful diagnostically. An elevated uptake is consistent with increased TSH action (e.g., antibody-negative Hashimoto thyroiditis or iodide deficiency) or with the action of non-TSH thyroid stimulators. Extremely low or undetectable radioiodide uptake by the thyroid is an important finding in the goitrous patient; the differential diagnosis includes excessive iodide intake, lymphocytic thyroiditis (painless thyroiditis, usually with transient hyperthyroidism), and granulomatous thyroiditis. Struma ovarii and factitious hyperthyroidism are also associated with extremely low or undetectable radioiodide uptake, but no goiter is present. The normalization of a previously high iodine intake in patients with iodine-induced hypothyroidism may result in an acute elevation of thyroidal radioiodide uptake, although serum T4 values remain low. A patchy uptake pattern on scan is frequently encountered in Hashimoto thyroiditis. Localized areas of reduced or absent uptake are difficult to interpret. They are associated with a 0% to 30% risk of carcinoma in various reports but most commonly represent adenoma or cyst. Areas of increased uptake are associated with an extraordinarily low risk of cancer (see Chap. 34). The estimation of iodide intake by measurement of 24-hour urinary iodide excretion is seldom necessary in the United States. Serum thyroglobulin levels are elevated in one-third of patients with nontoxic goiter and are not helpful in establishing a cause. Plasma calcitonin levels are normal in patients with benign nontoxic goiter.49 Ultrasonography of the thyroid is useful in distinguishing cystic from solid nodules in nontoxic goiter (see Chap. 35). Needle biopsy and aspiration cytology of the thyroid gland are valuable techniques for establishing certain pathologic states of the thyroid gland (see Chap. 39).50 In antibody-negative patients with Hashimoto thyroiditis, needle biopsy is useful in establishing the diagnosis. Both needle biopsy and aspiration cytology are important tools in assessing a nontoxic goiter for the possible presence of cancer. In patients with large goiters or local symptoms attributable to goiter, the extent of the goiter can be determined in several ways. Routine chest radiographs reveal a paratracheal mass and tracheal deviation (posteroanterior view) and obliteration of the superior retrosternal space (lateral view). Radionuclide scintigraphy of the thyroid usually shows the retrosternal extent of the gland. Iodine-123 is inadequate for substernal scanning; therefore, iodine-131 (131I) should be requested when a substernal goiter is a possibility. Technetium-99m pertechnetate has also been used, although less successfully, to evaluate the extent of substernal glands. Flow-loop studies of the upper airway51 and barium pharyngoesophagography are valuable in detecting tracheal and esophageal obstruction due to goitrous encroachment. Flow-loop studies do not necessarily correlate with clinical symptoms, because the latter may reflect the patient's anxiety. The marked distensibility of the esophagus minimizes, as noted earlier, the clinical impact of goiter on swallowing.

PROGNOSIS OF NONTOXIC GOITER

In young or middle-aged patients, particularly those with non-toxic gland enlargement due to Hashimoto thyroiditis, goiter may remit spontaneously. Goiter due to severe Hashimoto thyroiditis is thought to lead to progressive thyroid atrophy (remission of goiter) and eventual hypothyroidism. Mild thyroiditis usually has no metabolic sequelae. As many as one-third of autonomous (warm or hot) solitary nodules may remit spontaneously (burn out).52 The natural course of nontoxic goiter, however, ordinarily is further enlargement of the gland and the appearance of benign nodularity, estimated at 4.5% increase in size annually.53 Indeed, even normal thyroids, over the life span, often develop micronodularity or macronodularity. Except in the case of patients who have received low-energy head-and-neck irradiation, the risk of development of thyroid carcinoma in a nontoxic goiter is low (<1%). The overall risk of thyroid cancer in patients who have had head-and-neck irradiation may be as high as 8%.54 As indicated earlier, long-term follow-up of patients with large nontoxic goiter reveals a strong incidence (33%) of development of thyrotoxicosis. The rate of hyperthyroidism in patients with previously nontoxic goiter depends on duration of follow-up, for example, 15% at 12 years55 compared with the higher incidence cited earlier at 30 years. Thus, the prognostic spectrum of nontoxic goiter is broad, and the histopathology is usually benign. The interesting possibility has been raised that a substantial number of patients with nontoxic goiter have (extrapituitary) tissue overexposure to thyroid hormone.56 That is, the lower the level of circulating TSH in such patients, the more likely that indices of thyroid hormone action on bone or liver are elevated. The indices measured were, respectively, plasma concentration of bone g-carboxyglutamic acidcontaining protein (Gla protein) and plasma sex hormonebinding globulin.56 The long-term implications of the presence of low-grade, minimally excessive levels of thyroid hormone, as inferred from Gla protein and sex hormonebinding globulin levels, are unknown, although patients may be at increased risk for atrial fibrillation.57

TREATMENT OF NONTOXIC GOITER

The approach to managing nontoxic goiter involves (a) prevention of progressive enlargement of the thyroid gland; (b) anticipation, where appropriate, of hyperthyroidism or hypothyroidism; (c) reduction of the risk of carcinoma in the gland; and (d) relief of local symptoms attributable to goiter. Annual clinical evaluation of the patient with nontoxic goiter is a cornerstone of management, as is routine instruction of the patient to contact the treating physician if local symptoms emerge or change, or if systemic symptoms develop that are consistent with hypothyroidism or hyperthyroidism. Because nontoxic goiter is a state in which the permissive or directorial role of TSH is expressed, suppression of endogenous TSH secretion with exogenous thyroid hormone (L -thyroxine, T4) prevents further enlargement of the gland.58 Two-thirds of sporadic nontoxic goiters shrink when endogenous TSH is suppressed by exogenous thyroid hormone.59 This rate of success is less likely for nodular goiters than for diffuse goiters.59 Modest goiter, particularly that associated with Hashimoto thyroiditis, predictably responds to TSH suppression with a return of the gland to normal or near-normal size. Nodular goiter responds to suppression therapy in a nonpredictable manner. Sometimes normal thyroid tissue located between multiple nodules atrophies when endogenous TSH is suppressed, bringing the nodules into greater relief. This change can be misinterpreted as an increase in nodule size in the face of suppression. The effectiveness of the strategy of TSH suppression in management of nodular goiter is under review by several groups. Nodular or nonnodular, very large nontoxic goiters are unlikely to respond to suppressive treatment, and aggressive suppression of TSH places such patients at risk for iatrogenic hyperthyroidism. Although the suggestion has been made that T3 may be more effective than T4 in reducing goiter size,60 use of T4 is preferable because T4 has a long half-life and it does not lead to the transient episodes of elevated circulating T3 levels that occur with daily T3 therapy. TSH is suppressed in hypothyroid patients by 1.8 g/kg body weight per day, and by 2.2 to 2.5 g/kg per day in thyroid cancer patients.61 Although a definitive study of the suppressibility by T4 of TSH in nontoxic goiter patients using a sensitive TSH assay has not been conducted, a dose of 2 to 2.5 g/kg per day suffices in these patients. Avoidance of overtreatment with exogenous thyroid hormone is desirable, particularly in the elderly. The increasing availability of sensitive TSH assays that readily distinguish between suppressed and normal levels of TSH should help in determining the lowest suppressive dose of T4.62 The risk of bone demineralization in the course of T4 administration to suppress endogenous TSH secretion has been widely discussed, although its clinical impact, particularly in the setting of treatment of

nontoxic goiter, is not clear.63 Another apparent complication of long-term T4 treatment of nontoxic goiter patients may be increased left ventricular cardiac mass and diastolic dysfunction.64 The frequency and significance of these echocardiographic findings have not yet been determined. Autonomous nodular goiter should not be treated with T4; the additional hormonal load may result in clinical hyperthyroidism. In clinically euthyroid patients with goiter whose serum T4 levels are in the lower 25% of the normal range and whose serum TSH concentrations are moderately elevated, hypothyroidism is incipient (subclinical hypothyroidism) and should be managed expectantly with replacement doses of T4 (<0.10 to 0.15 mg per day). 65 A distinction is made here between replacement therapy and suppressive therapy. The latter is intended to remove the thyroid gland from the influence of endogenous TSH and to raise serum T4 levels to the normal range but above the pretreatment level. However, in incipient hypothyroidism with goiter, the intent is to raise the low serum thyroid hormone levels to the patient's normal range and to normalize serum TSH levels. The first .025-mg (25-g) increment (at 2- to 4-week intervals) of daily T4 replacement that normalizes serum TSH defines full replacement.65,66 and 67 The risk of occurrence of carcinoma in nontoxic goiter appears to be low. Suppression of endogenous TSH reduces the possibility even further. TSH is a permissive factor in the emergence of thyroid cancer, particularly, papillary carcinoma.68 The suppression of endogenous TSH in patients who have had prior head-and-neck irradiation has been presumed to reduce the risk of the emergence of thyroid cancer. However, patients who develop palpable abnormalities of the thyroid after irradiation should undergo thyroidectomy.69 Interest is increasing in the management of sporadic nontoxic goiter in older patients with 131I. In one study, 27 patients who received 5073700 MBq of 131I experienced a mean reduction in thyroid gland volume of 34% 1 year posttreatment.70 A review of multiple studies of this strategy confirmed the 1-year success rate and concluded that a 50% to 60% reduction in gland volume was achieved by 3 to 5 years after isotope administration.71 Estimates of the volume changes of the thyroid gland in such studies have been made ultrasonographically and with magnetic resonance imaging. The complications of this approach include hypothyroidism (20% to 30% incidence at 5 years after treatment) and radiation thyroiditis shortly after treatment. Radiation thyroiditis may be associated with transient hyperthyroidism. The risk of fatal and nonfatal thyroid cancer is also present but has not been precisely established. Thus, the approach is not endorsed in younger patients. A particularly interesting, but infrequent, complication of 131I treatment of nontoxic goiter is transient Graves-like hyperthyroidism associated with circulating levels of TSH receptor antibody.72 This syndrome appears to be distinct from radiation thyroiditisrelated release of stored thyroid hormone from the gland, but measurement of anti-TPO titers before 131I administration appeared to predict both radiation thyroiditis and hyperthyroidism associated with TSH receptor antibody. Large nontoxic goiters may induce substantial local symptoms and may be cosmetically unacceptable. In long-standing goiters of large size, fibrosis and hemorrhage may significantly reduce the amount of TSH-responsive glandular epithelium. Hence, in these patients, the suppression of endogenous TSH with exogenous thyroid hormone has little effect on gland size. Similarly, attempts to reduce the size of large nontoxic goiters by 131I ablation usually are disappointing, at least in cosmetic terms. For radioablation to succeed, residual glandular epithelium must comprise a significant proportion of the goiter. Treatment of large nontoxic goiters with 131I can relieve tracheal compression,71 as confirmed by magnetic resonance imaging; this may be the case even when the extent of gland shrinkage in a large goiter is only 5% to 10%.73 Large goiters accompanied by substantial local symptoms or unacceptable cosmesis should usually be treated surgically. Surgery may be difficult if a large and vascular goiter has involved vital structures in the neck. The incidence of tracheomalacia in large goiters has been overemphasized. When it does occur, however, it is a life-threatening complication and must be promptly managed with tracheal intubation or tracheotomy. Long-term serial measurement of serum thyroglobulin in one surgical series predicted growth activity of thyroid remnants in patients who had undergone resection of nontoxic goiter.74

SUBSTERNAL GOITER
Substernal goiter refers to thyroidal enlargement, the major portion of which is inferior to the thoracic inlet.75 Most substernal goiters are contiguous with an enlarged cervical thyroid; only very rarely is ectopic goitrous thyroid found in the mediastinum in the absence of cervical thyroid tissue. The clinical features of substernal goiter are summarized in Table 38-3 and illustrated in Figure 38-2. It most often occurs in older women and is frequently asymptomatic, in which case it appears on routine chest radiographs as a tracheal deviation or as a right paratracheal soft-tissue mass. Symptomatic substernal goiter may be associated with dyspnea or dysphagia. The severity of dyspnea may not correlate well with objective measurements of airway patency by flow-loop studies. Occasionally, superior vena caval obstruction may result from a substernal goiter.76 Patients with substernal goiter are usually euthyroid77 but occasionally are thyrotoxic.41 The mediastinal extent of goiter may be demonstrated by radio-nuclide study or by computed tomography (see Fig. 38-2).

TABLE 38-3. Clinical Features of Patients with Substernal Goiter

The management of large substernal goiter is surgical. Virtually all such goiters may be removed through a standard cervical collar incision without a sternotomy.78 Ablative radioactive iodine administration offers little to the euthyroid patient with substernal goiter. Therapy with L -thyroxine to suppress endogenous TSH may prevent further gland enlargement but does not reduce gland size, except in the very rare patient with hypothyroidism. An important issue in the therapy of substernal goiter is the role and timing of surgery. Symptomatic patients and those with known tracheal, esophageal, or vena caval encroachment by the goiter are candidates for early surgery. The surgical literature indicates some bias for routine surgical excision of all substernal goiters because of the small risk of progression of the lesion to significant airway obstruction and because of a variable risk of there being cancer in the gland,77,78 compounded by the inaccessibility of the gland to diagnostic needle aspiration. Surgical removal of substernal goiter through a cervical collar incision has an extraordinarily low complication rate when performed by experienced surgeons.78,79 However, in otherwise healthy patients in the sixth and seventh decades of life, such surgery is recommended only when computed tomography or flow-loop studies indicate the presence of significant tracheal narrowing. CHAPTER REFERENCES
1. Peterson S, Sanga A, Eklf H, et al. Classification of thyroid size by palpation and ultrasonography in field surveys. Lancet 2000; 355:106. 1a. Matovinovic J. Endemic goiter and cretinism at the dawn of the third millennium. Annu Rev Nutr 1983; 3:341. 2. Vander JB, Gaston EA, Dawber TR. The significance of nontoxic thyroid nodules. Ann Intern Med 1968; 69:537. 3. Toft AD, Irvine WJ, Hunter WM. A comparison of plasma TSH levels in patients with diffuse and nodular non-toxic goiter. J Clin Endocrinol Metab 1976; 42:973. 4. Bregengard C, Kirkegaard C, Faber J, et al. Relationships between serum thyrotropin, serum free thyroxine (T 4), and 3,5,3-triiodothyronine (T3) and the daily T4 and T3 production rates in euthyroid patients with multinodular goiter. J Clin Endocrinol Metab 1987; 65:258. 5. Romelli F, Stoder H, Bruggisser D. Pathogenesis of thyroid nodules in multinodular goiter. Am J Pathol 1982; 109:215. 6. Peter HJ, Studer H, Forster R, Gerber H. The pathogenesis of hot and cold follicles in multinodular goiters. J Clin Endocrinol Metab 1982; 55:941. 7. Niepomniszcze H, Altschuler N, Korob MH, Degrossi OJ. Iodide-peroxidase activity in human thyroid: I. Studies on non-toxic nodular goiter. Acta Endocrinol (Copenh) 1969; 62:193. 8. Field JB, Larsen PR, Yamashita K, et al. Demonstration of iodide transport defect but normal iodide organification in nonfunctioning nodules of human thyroid glands. J Clin Invest 1973; 52:2404. 9. Sugenoya A, Yamada Y, Kaneko G, et al. In vitro study on release of thyroid hormone in solitary autonomously functioning thyroid nodules using cell culture method. Endocrinol Jpn 1984; 31:749. 10. Gheri RG, Berrelli D, Cicchi P, et al. Thyroxine and triiodothyronine levels in thyroid vein blood and in thyroid tissue of patients with autonomous adenomas. Clin Endocrinol (Oxf) 1981; 15:485. 11. Toccafondi R, Rotella CM, Tanini A, et al. Effects of TSH on cAMP levels and thyroid hormone release in human thyroid autonomous nodules: relationship with iodothyronine and iodine content

in thyroglobulin. Clin Endocrinol (Oxf) 1982; 17:537. 12. Thomas CG Jr, Combest W, McQuade R, et al. Biological characteristics of adenomatous nodules, adenomas, and hyperfunctioning nodules as defined by adenylate cyclase activity and TSH receptors. World J Surg 1984; 8:445. 12a. Schuppert F, Ehrenthal D, Frilling A, et al. Increased major histocompatibility complex (MHC) expression in nontoxic goiters. J Clin Endocrinol Metab 2000; 85:858. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. Cooper DS, Ridgway EC, Maloof F. Unusual types of hyperthyroidism. Clin Endocrinol Metab 1978; 7:199. Drexhage HA, Bolcazzo GF, Doniach D, et al. Evidence for thyroid growth stimulating immunoglobulins in some goitrous thyroid diseases. Lancet 1980; 2:287. van der Gaag RD, Drexhage HA, Wiersinga WM, et al. Further studies on thyroid growth-stimulating immunoglobulins in euthyroid nonendemic goiter. J Clin Endocrinol Metab 1985; 60:972. Aguayo J, Sakatsume Y, Jamieson C, et al. Nontoxic nodular goiter and papillary thyroid carcinoma are not associated with peripheral blood lymphocyte sensitization to thyroid cells. J Clin Endocrinol Metab 1989; 68:145. Minuto F, Barreca A, Del Monte P, et al. Immunoreactive insulin-like growth factor I (IGF-I) and IGF-I binding protein content in human thyroid tissue. J Clin Endocrinol Metab 1989; 68:621. Grubeck-Loebenstein B, Buchan G, Sadeghi R, et al. Transforming growth factor b regulates thyroid growth. Role in the pathogenesis of nontoxic goiter. J Clin Invest 1989; 83:764. Smyth PPA, Neylan D, O'Donovan DK. The prevalence of thyroid-stimulating antibodies in goitrous disease assessed by cytochemical section bio-assay. J Clin Endocrinol Metab 1982; 54:357. Valente WA, Vitti P, Rotella CM, et al. Antibodies that promote thyroid growth: a distinct population of thyroid-stimulating autoantibodies. N Engl J Med 1983; 309:1028. Rapoport B, Greenspan FS, Filetti S, Pepitone M. Clinical experience with a human thyroid cell bioassay for thyroid-stimulating immunoglobulin. J Clin Endocrinol Metab 1984; 58:332. Etienne Decerf J, Winand RJ. A sensitive technique for determination of thyroid-stimulating immunoglobulin (TSI) in unfractionated serum. Clin Endocrinol (Oxf) 1981; 14:83. Vitti P, Chiovato L, Tonacchera M, et al. Failure to detect thyroid growth-promoting activity in immunoglobulin G of patients with endemic goiter. J Clin Endocrinol Metab 1994; 78:1020. Comite F, Burrow GN, Jorgensen EC. Thyroid hormone analogs and fetal goiter. Endocrinology 1978; 102:1670. Henneman G. Non-toxic goiter. Clin Endocrinol Metab 1979; 8:167. Oddie TH, Fisher DH, McConahey WM, Thompson CS. Iodine intake in the United States: a reassessment. J Clin Endocrinol Metab 1970; 30:659. Kelly FC, Snedden WW. Prevalence and geographic distribution of endemic goitre. In: Endemic goitre. Monograph series no. 44. Geneva: World Health Organization, 1960:27. Stanbury JB, Kroc RL, eds. Human development and the thyroid gland. Adv Exp Med Biol 1972; 30:3. Langer P, Greer MA. Antithyroid substances and naturally occurring goitrogens. Basel: S Karger, 1977:79. Agerboek H. Non-toxic goitre: the role of iodine deficiency in goitre formation in a non-endemic area. Acta Endocrinol (Copenh) 1974; 76:74. Braverman LE, Woeber KA, Ingbar SH. Induction of myxedema by iodide in patients euthyroid after radioiodine or surgical treatment of diffuse toxic goiter. N Engl J Med 1969; 281:816. Vagenakis AG, Wang C, Burger A, et al. Iodide-induced thyrotoxicosis in Boston. N Engl J Med 1972; 287:523. Wartofsky L, Ransil BJ, Ingbar SH. Inhibition by iodine of release of thyroxine from the thyroid glands of patients with thyrotoxicosis. J Clin Invest 1970; 49:78. Braverman LE. Therapeutic considerations. Clin Endocrinol Metab 1978; 7:221. Rallison MI, Kumagai LF, Tyler FH. Goitrous hypothyroidism induced by aminoglutethimide, anticonvulsant drug. J Clin Endocrinol Metab 1967; 27:265. Spaulding SW, Burrow GN, Bermudez F, Himmelhoch JM. The inhibitory effect of lithium on thyroid hormone release in both euthyroid and thyrotoxic patients. J Clin Endocrinol Metab 1972; 35:905. Singh BN, Nademanee K. Amiodarone and thyroid function: clinical implications during antiarrhythmic therapy. Am Heart J 1983; 106:857. Day TK, Powell-Jackson PR. Fluoride, water hardness, and endemic goitre. Lancet 1972; 1:1135. Brown J, Solomon DH. Effects of tolbutamide and carbutamide on thyroid function. Metabolism 1956; 5:813. Martino E, Placidi GF, Sardano G, et al. High incidence of goiter in patients treated with lithium carbonate. Ann Endocrinol (Paris) 1982; 43:269. Davis PJ, Davis FB. Hyperthyroidism in patients over the age of 60 years: clinical features in 85 patients. Medicine (Baltimore) 1974; 53:161. Borup Christensen S, Ericsson U-B, Janzon L, et al. Influence of cigarette smoking on goiter formation, thyroglobulin, and thyroid hormone levels in women. J Clin Endocrinol Metab 1984; 58:615. Spector DA, Davis PJ, Helderman JH, et al. Thyroid function and metabolic state in chronic renal failure. Ann Intern Med 1976; 85:724. Favus MJ, Schneider AB, Stachura ME, et al. Thyroid cancer occurring as a late consequence of head-and-neck irradiation. N Engl J Med 1976; 294:1019. Ivy HK. Cancer metastatic to the thyroid: a diagnostic problem. Mayo Clin Proc 1984; 59:856. Hazard JB. Thyroiditis: a review. Am J Clin Pathol 1955; 25:289. Glinoer D, Verelst J, Ham HR. Abnormalities of esophageal transit in patients with sporadic nontoxic goiter. Eur J Nucl Med 1987; 13:239. Strakosch CR, Wenzel BE, Row VV, Volpe R. Immunology of autoimmune thyroid diseases. N Engl J Med 1982; 307:1499. Levine GA, Hershman JM, Van Herle AJ, et al. Thyroglobulin and calcitonin in patients with nontoxic goiter. JAMA 1978; 240:2282. Tani EM, Skoog L, Lowhagen T. Clinical utility of fine needle aspiration cytology of the thyroid. Annu Rev Med 1988; 39:255. Jauregui R, Lilker ES, Bayley A. Upper airway obstruction in euthyroid goiter. JAMA 1977; 238:2163. Silverstein GE, Burke G, Cogan R. The natural history of the autonomous hyperfunctioning thyroid nodule. Ann Intern Med 1967; 67:539. Berghout A, Wiersinga WM, Smits NJ, Touber JL. Interrelationships between age, thyroid volume, thyroid nodularity and thyroid function in patients with sporadic nontoxic goiter. Am J Med 1990; 89:602. Refetoff S, Harrison J, Karanfilski BT, et al. Continuing occurrence of thyroid carcinoma after irradiation to the neck in infancy and childhood. N Engl J Med 1975; 292:171. Wiersinga WM. Determinants of outcome in sporadic nontoxic goiter. Thyroidol Clin Exp 1992; 4:41. Faber J, Perrild H, Johansen JS. Bone Gla protein and sex hormonebinding globulin in nontoxic goiter: parameters for metabolic status at the tissue level. J Clin Endocrinol Metab 1990; 70:49. Sawin CT, Geller A, Wolf PA, et al. Low serum thyrotropin concentrations as a risk factor for atrial fibrillation in older persons. N Engl J Med 1994; 331:1249. Astwood EB, Cassidy CE, Aurbach GD. Treatment of goiter and thyroid nodules with thyroid. JAMA 1960; 174:459. Ross DS. Thyroid hormone suppressive therapy of sporadic nontoxic goiter. Thyroid 1992; 2:263. Shimaoka K, Sokal JE. Suppressive therapy of nontoxic goiter. Am J Med 1974; 57:576. Burmeister LA, Goumaz MO, Mariash CN, Oppenheimer JH. Levothyroxine dose requirements for thyrotropin suppression in the treatment of differentiated thyroid cancer. J Clin Endocrinol Metab 1992; 75:344. Spencer CA, LoPresti JS, Patel A, et al. Applications of a new chemiluminometric thyrotropin assay to subnormal measurement. J Clin Endocrinol Metab 1990; 70:453. Mller CG, Bayley TA, Harrison JE, Tsang R. Possible limited bone loss with suppressive thyroxine therapy is unlikely to have clinical relevance. Thyroid 1995; 5:81. Fazio S, Biondi B, Carella C, et al. Diastolic dysfunction in patients on thyroid-stimulating hormone suppressive therapy with levothyroxine: beneficial effect of beta-blockade. J Clin Endocrinol Metab 1995; 80:2222. Davis FB, LaMantia RS, Spaulding SW, et al. Estimation of a physiologic replacement dose of levothyroxine in elderly patients with hypothyroidism. Arch Intern Med 1984; 144:1752. Hennessey JV, Evaul JE, Tseng Y-C, et al. L-thyroxine dosage: a reevaluation of therapy with contemporary preparations. Ann Intern Med 1986; 105:11. Wartofsky L. Diffuse nontoxic and multinodular goiter. In: Current therapy in endocrinology and metabolism 19851986. Toronto: BC Decker, 1986:75. Rojeski MT, Gharib H. Nodular thyroid disease: evaluation and management. N Engl J Med 1985; 313:428. McConahey WM, Hayles AB. Radiation and thyroid neoplasia. Ann Intern Med 1976; 84:749. deKlerk JM, van Isselt JW, van Kijk A, et al. Iodine-131 therapy in sporadic nontoxic goiter. J Nucl Med 1997; 38:372. Huysmans D, Hermus A, Edelbroek M, et al. Radioiodine for nontoxic multinodular goiter. Thyroid 1997; 7:235. Nygaard B, Knudsen JH, Hegedus L, et al. Thyrotropin receptor antibodies and Graves' disease, a side effect of 131 I treatment in patients with nontoxic goiter. J Clin Endocrinol Metab 1997; 82:2926. Kay TWH, d'Eurden MC, Andrews JT, Martin FR. Treatment of nontoxic multinodular goiter with radioactive iodine. Am J Med 1988; 84:19. Date J, Feldt-Rasmussen U, Blichert-Toft M, et al. Long-term observation of serum thyroglobulin after resection of nontoxic goiter and relation to ultra-sonographically demonstrated relapse. World J Surg 1996; 20:351. Katlic MR, Wang CA, Grillo HC. Substernal goiter. Ann Thorac Surg 1985; 39:391. McKellar DP, Verazin GT, Lim KM, et al. Superior vena cava syndrome and tracheal obstruction due to multinodular goiter. Head Neck 1994; 16:72. Allo MD, Thompson NW. Rationale for the operative management of sub-sternal goiters. Surgery 1983; 94:969. Katlic MR, Grillo HC, Wang CA. Substernal goiter: analysis of 80 patients from Massachusetts General Hospital. Am J Surg 1985; 149:283. Melliere D, Saada F, Etienne G, et al. Goiter with severe respiratory compromise: evaluation and treatment. Surgery 1988; 103:367.

CHAPTER 39 THE THYROID NODULE Principles and Practice of Endocrinology and Metabolism

CHAPTER 39 THE THYROID NODULE

LEONARD WARTOFSKY AND ANDREW J. AHMANN Prevalence of Thyroid Nodules Differential Diagnosis of Thyroid Nodules Irradiation-Related Thyroid Neoplasia Diagnostic Evaluation of the Thyroid Nodule History and Physical Examination Laboratory Tests Thyroid Scanning Thyroid Hormone Suppression Thyroid Needle Aspiration and Biopsy Integrated Approach to Management of Thyroid Nodules Management of Thyroid Cysts Management of Autonomous Nodules Chapter References

The clinical management of nodular thyroid disease is controversial.1,2,3,4,5,6,7,8,9,10,11 and 12 The finding of a thyroid nodule poses considerable anxiety for patient and physician alike, despite the modest probability of malignancy and the likelihood that a tumor, if present, will follow a relatively benign course. In the evaluation of thyroid nodules, the physician has the opportunity to use a growing number of diagnostic tools, no single one of which provides the desired accuracy in identifying all cases of carcinoma requiring surgical intervention. Some consensus has been reached, however, on guidelines for the diagnosis and management of thyroid nodules.13

PREVALENCE OF THYROID NODULES

The prevalence of thyroid nodules must be considered from several perspectives. As many as 1.5% of adolescents may have palpable thyroid nodules, and the frequency increases linearly with age.11 In large population studies, nodules have been reported in 3% to 7% of adults, with a 5:1 female/male ratio. Many of these reports have failed to distinguish solitary nodules from multiple thyroid nodules; on histologic examination, many single, palpable nodules are found to be within a multinodular thyroid gland. Some autopsy series report thyroid nodules in more than one-half of consecutive cases, although many may be small, nonpalpable lesions, and multinodularity occurs in up to 75% of these.1,3,4 High-resolution ultrasonography has identified nodules in 13% to 40% of patients undergoing evaluation for nonthyroidal problems.14,15,16 and 17 Thus, a discrepancy exists between the true prevalence of thyroid nodules and the number that are apparent by physical examination. Normally, nodules must approach 1 cm in diameter to be consistently recognized on palpation, although this size varies according to the location of the nodule within the gland and the anatomy of the neck. Nonpalpable nodules, such as those incidentally seen on ultrasonographic studies for nonthyroidal neck evaluation, are of unknown clinical significance.

DIFFERENTIAL DIAGNOSIS OF THYROID NODULES

The differential diagnosis of apparent thyroid nodules covers a wide range of pathology1,2,3,4 and 5 (Table 39-1). The usual diagnostic challenge is to distinguish malignant from benign lesions. Most true intrathyroidal nodules are colloid adenomas (2760%) or simple follicular adenomas (2640%), although fully reliable figures are not available.18 Follicular adenomas introduce significant uncertainty in the diagnosis and treatment because no available measures short of surgical resection with detailed histologic examination allow the separation of well-differentiated follicular carcinomas from adenomas. Instead, the distinction depends on the presence or absence of capsular or vascular invasion by the tumor; cellular detail is of little value.19 Hrthle cell tumors are a subset of follicular lesions composed of oval to polygonal cells with dense, granular, acidophilic cytoplasm and a prominent macronucleolus. The malignant potential of histologically benign Hrthle cell tumors and the natural history of Hrthle cell malignancies continue to be subjects of controversy.20,21

TABLE 39-1. Differential Diagnosis of Apparent Thyroid Nodules

Although the functional activity of thyroid adenomas varies, most appear to have defects in the iodine symporter resulting in reduced iodide accumulation and/or organification; this may be demonstrated by reduced trapping of radionuclide, leading to the designation of cold nodules. Five percent to 15% of thyroid nodules are classified as hot on the basis of a relatively increased ability to trap iodide. Most of these hot nodules are autonomously functioning (see Chap. 34 and Chap. 42). The patients are usually clinically euthyroid, although more than one-half of those older than 60 years of age may be hyperthyroid. Whether or not hyperthyroidism is present is related to the size of the hyperfunctioning nodules. Nodules >3 cm in diameter are associated with hyperthyroidism in 20% of cases, whereas <2% of smaller lesions produce a toxic state.22,23 The thyrotoxicosis associated with a hot nodule may be transient, arising from acute hemorrhage, or, more commonly, after exposure to a high iodine load (see Chap. 37). Although most toxic, autonomous nodules secrete both thyroxine (T4) and triiodothyronine (T3), occasionally they may elevate serum T3 alone or, rarely, serum T4 alone. Moreover, even when levels of circulating thyroid hormones are normal, a blunted response of thyroid-stimulating hormone (TSH) to thyrotropin-releasing hormone (TRH) stimulation is common, suggesting supraphysiologic iodothyronine production. Progression to a relatively mild toxic state is usually insidious. Autonomous adenomas have no unique histologic characteristics. Clinically, the functional distinction is important because of the natural history of these nodules: hot nodules have a much lower malignancy potential than hypofunctioning lesions. Constitutive activating mutations of the TSH receptor gene have been identified as the probable basis for many hot nodules.24,25,26 and 27 The absence of function in a thyroid nodule should arouse a greater suspicion of cancer. More than 60% of malignant lesions of the thyroid are papillary carcinomas, with follicular and anaplastic lesions being the next most common histologic types (see Chap. 40). Medullary thyroid carcinomas28,29 account for 5% of malignant thyroid nodules, and other categories represent <5% of such nodules.3,7 A cold thyroid nodule occasionally represents metastasis. The malignancies most likely to metastasize to the thyroid are malignant melanoma, bronchogenic carcinoma, renal cell carcinoma, breast carcinoma, and lymphoma. Presently, ~17,200 new thyroid cancer cases are diagnosed annually in the United States, and 1200 thyroid cancerrelated deaths occur.30 Although the probability at birth of eventually dying from thyroid cancer is <0.1%, the management of potentially malignant thyroid nodules may lead to 50,000 thyroid operations yearly. Studies have revealed occult thyroid cancer in 6% of autopsy cases in North American series; clinically significant cancers are found in 10% to 14% of patients with palpable thyroid nodules, and the incidence has increased in the past 50 years.1,3,7 As with benign nodules and many other thyroid diseases, thyroid cancer is more common in women than in men; however, among men with nodules, the probability of cancer is higher. The incidence of thyroid cancer increases with age, but a greater percentage of nodules in patients younger than 20 years of age represents malignancy.31,32 and 33 Although data are conflicting on the increased risk of malignancy in new nodules in patients older than 60 years of age, the incidence of anaplastic carcinoma is higher in this group.

IRRADIATION-RELATED THYROID NEOPLASIA

Many causative factors in the development of thyroid nodules, especially thyroid carcinomas, are uncertain. One undisputed factor, however, is radiation exposure. Animal studies have consistently confirmed radiation induction of thyroid neoplasia, which can be further promoted by TSH. By 1950, many children with thyroid

carcinoma were known previously to have received radiation to the neck as treatment for benign conditions, including enlarged thymus, chronic tonsillitis, acne, and cervical adenitis. Numerous studies have now confirmed a linear relationship between radiation dose (for doses up to 1500 rad) and the incidence of thyroid nodules and thyroid cancer. Some 40 million Ci of iodine-131 (131I) and another 100 million Ci of shorter-lived isotopes of iodine were released into the atmosphere by the nuclear reactor accident at Chernobyl in the Belarus in April 1986. The number of new cases of thyroid cancer appearing in patients who were exposed to the fallout as young children in 1986 has shown a dramatic increase between 1989 and the present.34,35 Similar associations were demonstrated for radioactive fallout after the World War II experience in Japan and the atomic test explosions in the Marshall Islands. Nasopharyngeal radium implants appear to add little risk of cancer. Childhood radiation exposure is more likely to produce thyroid neoplasms than similar exposure at a later age, possibly because of greater cellular mitotic activity at the earlier age. In the United States, palpable nodules are found in 16% to 29% of persons who received low-dose head and neck irradiation in childhood, and carcinoma is found in one-third of these patients with nodules.36 Lesions can occur within several years of exposure, but the peak incidence after external radiation exposure for both benign and malignant nodules occurs after a latency period of 20 to 30 years, and the risk may exist for more than 35 years. Approximately 85% of radiation-induced cancers are pure papillary or papillary-follicular lesions and have a favorable prognosis; anaplastic lesions are rare. At least one-third of the carcinomas associated with earlier irradiation to the head and neck are <1 cm in diameter. The lymph nodes are involved in 25% to 30% of patients, including many with nonpalpable primary lesions. The irradiated thyroid gland often presents with multiple nodules, and at surgery, the lesion of initial concern may prove to be benign, whereas one or more carcinomas may be found elsewhere in the gland. Thus, multicentricity is common, and most cancers coexist with benign adenomas. Contrary to usual circumstances, those nodules associated with previous radiation therapy do not demonstrate a marked reduction in cancer risk when the thyroid contains multiple nodules. The thyroid cancers seen after the Chernobyl accident have appeared after a much shorter latency period, but they, too, were largely papillary in nature. The consequences of high-dose external head and neck irradiation (>2000 rad) are not well defined because of relatively short follow-up in reported series. Although low-dose irradiation does not produce hypothyroidism, larger doses, such as those used for Hodgkin disease, are associated with a high rate of thyroid dysfunction, as manifested by elevated serum TSH levels, with or without hypothyroxinemia.37 Despite a postulated inducer effect of elevated TSH levels, and despite some reports of nodular thyroid disease, including carcinoma, in patients who have received high-dose irradiation, no definitive evidence exists of an increased incidence of carcinoma in this group. Consideration of isodose curves, penumbra effect, backscatter, and specifics of the employed field suggests that in many of the reported cases thyroid lesions occurring after higher dose radiation therapy actually have been associated with low-dose exposure of the thyroid.38 As with high-dose external irradiation, 131I therapy does not appear to be causally related to subsequent thyroid carcinoma. In both cases, the high-dose thyroid exposure with attendant cell destruction, fibrosis, and hypothyroidism may serve to attenuate any carcinogenic effect.

DIAGNOSTIC EVALUATION OF THE THYROID NODULE

Because most thyroid nodule morbidity is related to cancerous lesions, clinical evaluation focuses on the identification of malignant nodules. Numerous methods have evolved to better define the probability of cancer in a given lesion. HISTORY AND PHYSICAL EXAMINATION The single most important historical risk factor for cancer is irradiation. The age at exposure, the type and site of therapy, and, if possible, the radiation dose to the thyroid should be determined. A family history of pheochromocytoma, hypercalcemia, mucosal abnormalities, or medullary thyroid carcinoma should raise suspicion of the latter diagnosis as part of the multiple endocrine neoplasia syndrome (see Chap. 188). A family history of benign goiter may be reassuring, although the rare Pendred syndrome (familial goiter and deafmutism) is associated with a higher cancer risk.1,3 Nodules in men are more likely to be malignant than those in women; nodules in children are more likely to be malignant than those in adults. The diagnosis of thyroid lymphoma should be considered in patients with a previous diagnosis of Hashimoto thyroiditis, especially in women older than 50 years of age. The following clinical findings are thought to be more common in malignant than benign nodules: a history of radiation therapy, a family history of multiple endocrine neoplasia type 2A or 2B, rapid growth, hoarseness, pain, dysphagia, respiratory obstructive symptoms, and growth of the nodule despite T4 medication; and physical examination findings of cervical lymphadenopathy, firmness, documented nodule growth, vocal cord paralysis, fixation, and Horner syndrome. In general, signs and symptoms alone are not sufficiently sensitive or specific to allow selection of candidates for surgery. However, patients with advanced disease may present with lymphadenopathy, recent growth of hard nodules, and vocal cord paralysis, all of which suggest malignancy; the presence of obstructive symptoms is less reliable.1,39 As indicated, identification of a single nodule renders the patient at higher risk than does multinodularity.40 LABORATORY TESTS The commonly used blood thyroid function tests often are of limited value in the evaluation of thyroid nodules. Functional thyroid parameters, including serum T3 and T4 or TRH stimulation (see Chap. 15 and Chap. 33) are useful in evaluating possible toxic adenomas. Thyroglobulin levels may be elevated in patients with thyroid malignancy but do not differentiate malignant adenomas from benign adenomas or from thyroiditis.41,42 Serum levels of antithyroglobulin and antimicrosomal antibodies also are of limited value.1,3 An increased basal plasma calcitonin level is reasonably sensitive and specific for medullary thyroid carcinoma in the setting of a thyroid nodule, although abnormal calcium and pentagastrin stimulation tests or detection of mutations in the ret protooncogene provide the greatest sensitivity43 (see Chap. 40, Chap. 53 and Chap. 188). In one large series, an elevated plasma calcitonin level was found in 0.5% of all patients with thyroid nodules and in 15.7% of those with thyroid carcinomas, a finding that led the authors to recommend routine measurement in the evaluation of thyroid nodules.44 Serum carcinoembryonic antigen levels are also elevated in most patients with medullary thyroid carcinoma, but they may be increased in patients with other thyroid carcinomas as well.1,4,7,45 None of these tests intended to detect medullary thyroid carcinoma is cost-effective for the initial evaluation of the nodular thyroid. THYROID SCANNING RADIONUCLIDE UPTAKE Based on the pattern of radionuclide uptake (see Chap. 34), nodules may be classified as hot (hyperfunctioning) or cold. Hyperfunctioning nodules rarely represent malignancy, non-delineated or hypofunctioning lesions carry an intermediate risk, and nonfunctioning (cold) nodules have the highest risk.2,4,7 More than 80% of nonfunctioning nodules, however, still represent benign pathology. Therefore, radioisotope scans are of low diagnostic specificity despite their high sensitivity for nodules >1 cm in diameter. Scanning usually is done with iodine-123 (123I) or technetium-99m (99m Tc) pertechnetate. Despite some limitations, the advantages of low radiation dose, low cost, short scanning time, and reliability have led to the wide use of 99m Tc. 123I also delivers low radiation and is the preferred iodine-scanning agent. In addition to functional information, scans may reveal multinodularity in up to one-third of clinically palpable solitary lesions, a finding that decreases the likelihood of malignancy. Nuclear imaging with radioiodine and other isotopes plays an important role in the follow-up evaluation and treatment of patients proven to have thyroid cancer.46 FLUORESCENT THYROID SCANNING Fluorescent thyroid scanning uses americium-241, which can excite thyroidal iodine, causing the release of x-rays; this release correlates quantitatively with iodine content of the imaged tissue. Fluorescent scanning is rapid and provides minimal radiation exposure, thereby offering special advantages for children and pregnant women. The procedure is nearly 100% sensitive but only 64% specific when areas of low iodine content are taken as positive results. However, a large series has indicated that the converse analysis may be more reliable because areas with an iodine content ratio above 0.60 yielded 63% sensitivity and 99% specificity for benign lesions.47 Unfortunately, the required equipment is not widely available, and accumulated data remain too limited for this technique to be used routinely. The use of other radionuclides, including thallium-201, selenomethionine-75, gallium-67, and cesium-131, for differentiating benign from malignant lesions of the thyroid has been investigated. None of these has proved to be a reliable indicator of malignancy. OTHER THYROID IMAGING TECHNIQUES Radiography. Routine radiographic techniques have long been used in an attempt to identify calcifications characteristic of malignancy, but they are of limited value. Shell-like calcifications are most typical of benign cysts, stippled (psammomatous) calcifications have high specificity for papillary carcinoma, and flocculent deposits are more characteristic of medullary carcinoma, but other descriptive terms of calcification have minimal predictive value.1,2

Ultrasonography. Ultrasonography is frequently used to evaluate the thyroid. Conventional ultrasonographic techniques are most valuable in differentiating solid from cystic lesions.48 This differentiation is important because solid lesions have a higher malignant potential and require different therapy. However, cystic lesions larger than 4 cm in diameter may pose a significant cancer risk.1 Modern high-resolution ultrasonography provides detailed information and can demonstrate lesions as small as 2 mm; this increased sensitivity has revealed that pure simple cysts are exceedingly rare because most cystic lesions contain some solid components.49 Although high-resolution ultrasonography discloses multinodularity in up to 40% of glands presumed to have a single palpable nodule, whether this subclinical multinodularity has the same favorable connotation as multinodularity detected with the less sensitive techniques of physical examination or radionuclide scanning is unclear (see Chap. 35). However, it does appear that the majority of small incidentalomas of the thyroid detected by high resolution ultrasonography are benign.50 Ultrasonography may be used for the following purposes: to differentiate solid from cystic nodules, to detect multinodularity, to detect occult thyroid malignancy in cases of metastatic cervical lymphadenopathy from an unknown primary, to monitor the size of a nodule (including any response to suppressive therapy), to differentiate solid from hemorrhagic expansion in fast-growing thyroid lesions, to guide needle biopsy in selected cases, to monitor irradiated thyroids, and to monitor the local recurrence of thyroid carcinoma. Unfortunately, the value of ultrasonography in specifically identifying malignancy is limited. Carcinomas are usually hypoechoic, but many adenomas have a similar echogenicity. The appearance of a 1- to 2-mm, decreased echogenic halo surrounding nodules strongly favors follicular adenoma as the diagnosis, but this also has been found in a few carcinomas.49,51 Other Imaging Techniques. Two other imaging techniques deserve mention. Computed tomography (CT) can provide descriptive information similar to that achieved with ultra-sonography but offers no advantages except in cases of mediastinal extension.51 CT is associated with high doses of ionizing radiation, is more expensive, and requires iodine contrast media for maximal visualization. Ultrasonography is preferable to CT as a thyroid imaging method in most cases. Magnetic resonance imaging of the thyroid does not require iodine contrast or involve radiation exposure and provides superior vascular imaging and substernal views as compared with CT, but both CT and magnetic resonance imaging are useful for substernal goiter and for both initial presentation and recurrence of thyroid malignancies (see Chap. 35). Use of a number of other isotopes and the technique of positron emission tomography scanning have been applied to the assessment of thyroid nodules and thyroid malignancies.52,53

THYROID HORMONE SUPPRESSION

Thyroid hormone has been used for many years to reduce the size of thyroid lesions thought to be dependent on TSH stimulation. In using hormone administration as a diagnostic test for thyroid nodules, the assumption is that benign lesions will show preferential reduction in size. Typically, patients are given a 3- to 6-month trial of T4 at a dose titrated to result in undetectable serum TSH in an ultrasensitive assay or unresponsiveness of TSH to TRH stimulation, but without induction of clinical hyperthyroidism. Continued growth of a nodule or lack of reduction in size during therapy increases the suspicion of malignancy. However, an absolute reduction in size sufficient to denote benignancy has not been determined, and various investigators have adopted different criteria for response, which renders the interpretation of results more difficult. A partial response to treatment is not particularly reassuring because of the possibility of inconsistency in palpated size and/or a misleading apparent reduction in size of a nodule due to regression of surrounding normal tissue. Seven percent to 16% of nonresponding lesions harbor carcinoma, and an occasional carcinoma has exhibited partial regression.2 Largely uncontrolled studies have suggested that a complete response to suppressive therapy occurs in <10% of cases, whereas a 50% reduction in size is seen in ~30% of cases.7 However, one double-blind controlled study found no significant difference in sonographically measured colloid nodule size reduction between a group of patients treated with suppressive doses of T4 and those treated with placebo.54 Although 21% of thyroid hormonetreated nodules decreased >30% in volume over 6 months, equivalent size reductions were noted in nodules in the placebotreated group. Some confirmation of these findings was provided by the failure to observe significant shrinkage of nodules on T4 therapy in two other studies.55,56 On the other hand, other workers continue to suggest that suppressive therapy may shrink both nontoxic goiter57 and thyroid nodules.58,59,60 and 61 One review of relevant studies on the efficacy of L -thyroxine in suppressing thyroid nodules concluded that therapy provided benefit for only 10% to 20% of lesions proven benign by fine-needle aspiration cytology.9 Results of other studies suggest that the smaller lesions in younger patients are the ones that are more likely to shrink with suppression.62,63 But in spite of decades of such therapy and numerous published reports, one can conclude that additional carefully controlled studies of large numbers of patients are required to clarify the management of thyroid nodules with suppression therapy.10 The increasing concern about possible risk of osteopenia after long-term suppressive doses of thyroid hormone has been allayed somewhat by careful analyses of the data.64 Use of prudent suppression doses of T4 has not been shown to contribute to osteopenia.65 Moreover, supplementation of T4 with estrogen replacement in postmenopausal women may obviate any potential risk of osteopenia.66,67 Failure of suppression appears to minimally increase the probability of cancer, whereas successful suppression reduces the probability by ~25%.7 An unusual category of lesions that may carry an exceptionally high risk of carcinoma comprises those that respond initially to thyroid hormone but later grow.7 A trial of thyroid hormone suppression therapy alone is neither sensitive nor specific but may have utility as an adjunct to other modalities of evaluation. This therapy requires periodic and regular follow-up.

THYROID NEEDLE ASPIRATION AND BIOPSY

Obtaining tissue or cells from a thyroid nodule by some form of biopsy technique is the best single method to identify malignancy.68,69 and 70 Biopsies can be performed by fine-needle (22- to 27-gauge) aspiration (FNA), large-needle (16- to 18-gauge) aspiration, or cutting-needle biopsy (14-gauge Tru-Cut or Silverman needle). Most clinicians have found FNA to provide the highest incidence of successful samples and the lowest incidence of complications while yielding a diagnostic precision equal to or better than that of other methods.2 This technique is shown in Figure 39-1. In selected large lesions for which FNA has failed to provide a satisfactory diagnosis, some other aspiration method may be used.

FIGURE 39-1. Technique for fine-needle aspiration biopsy. A, Aspirate 1 to 2 mL of air into syringe to loosen plunger and facilitate expiration of contents after aspiration. B, Insert needle into lesion without aspiration. C, Retract syringe piston to provide maximum suction. D, Make several passes at different angles, withdrawing needle to near the surface before redirecting. E, Release suction passively. F, Withdraw needle. G, Express needle contents onto a glass slide.

FNA has been popular in Europe for decades and now has attained similar acceptance in the United States. The ability to obtain adequate samples improves with experience, and the success rate approaches 95%. However, the collecting technique appears less critical than the ability of the pathologist to interpret the cytologic specimens correctly. Because of these two primary factors, this procedure is best limited to situations in which the operator and pathologist each have considerable experience.71,72 and 73 Ultrasonography-guided FNA has proven a useful technique for biopsy of low-lying nodules or those that are detected by ultra-sonography but are difficult to palpate.74,75 FNA results may yield a variety of descriptive diagnoses.76,77 The most common and important categories are typically benign, suspicious for malignancy, malignancy, and inadequate for diagnosis. Examples of cytologic and histologic features are shown in Figure 39-2. At least 60% of aspirates are categorized as benign. The determination of sensitivity and specificity depends on whether suspicious lesions are considered positive. This latter group is primarily composed of highly cellular lesions, which represent ~20% of aspirations, of which 20% are cases of malignancy.77 When all suspicious lesions are taken to surgery, the sensitivity of FNA exceeds 90%, with specificity approximating 70%.68,69 and 70 The specificity of FNA for the malignant aspirates probably exceeds 95%. The use of FNA has decreased the frequency of surgery by 50% while doubling the rate of finding carcinoma in patients undergoing surgery. The relative economic advantage of this technique has been

documented.3,4,6 Difficulties in interpretation of FNA biopsy specimens involve the differentiation of cellular specimens from possible Hrthle cell tumors, low-grade carcinomas, cellular adenomas, and Hashimoto thyroiditis. A few clinicians have performed quantitative DNA analyses on aspirated thyroid cells in an attempt to distinguish benign from malignant lesions.78 DNA measurements have been unsuccessful, although the technique may provide prognostic indications of malignant tumor aggressiveness, with correlation to outcome and survival.79 Another approach being evaluated is to perform immunohistochemical analysis on the FNA sample.80

FIGURE 39-2. A, Fine-needle aspirate (FNA) of follicular neoplasm showing hypercellular specimen consisting of follicular epithelium. 150 B, Microfollicular structures with centrally located colloid (arrow). Follicular cell nuclei are enlarged and pleomorphic (FNA of follicular neoplasm, 540). C, Follicular carcinoma (histology). Note well-defined capsule with invasion (arrow). Residual normal thyroid gland is present at top of the field. 95 D, Papillary carcinoma (FNA). Specimen is hypercellular, with numerous papillary clusters. 150 E, Papillary cluster with diagnostic intranuclear inclusions at arrow. 960 F, Hashimoto thyroiditis (FNA). Note chronic inflammatory background with reactive lymphocytes. A follicular epithelial group showing oxyphilic metaplasia is seen in the background. Note enlarged nuclei and granular cytoplasm of these cells (arrow). 480 (Slides and legend text courtesy of Dr. Sanford Robbins. Photography courtesy of Charles Brown, M.S. From the Department of Pathology, Walter Reed Army Medical Center, Washington, DC.)

Hashimoto disease also may be difficult to differentiate from lymphoma or from carcinoma with background Hashimoto changes. Anaplastic carcinomas may resemble poorly differentiated lymphoma or granulomatous thyroiditis. Cysts pose several diagnostic difficulties.48,51 Attempts to correlate characteristics of aspirated fluid with pathologic diagnosis have proved unreliable, although the occasional finding of clear, colorless fluid is typical of parathyroid cysts.2 Cystic fluid may be difficult to aspirate unless a large needle is used. Cells obtained by centrifugation of cyst fluid may provide insufficient or confusing information.76 Any solid remnant that follows initial cyst aspiration should be subjected to cautious follow-up and consideration for repeat aspiration. Finally, solid nodules <1 cm or >4 cm in diameter may be associated with an increased sampling error. FNA carries minimal risk when properly performed. Concerns have been raised as to the possibility of severe local hemorrhage with airway obstruction and recurrent laryngeal nerve damage, but such complications are not documented in large reported series of FNA.2,70 Bleeding complications are unusual and almost always self-limited. However, core biopsy has been associated with occasional severe bleeding and rare reports of tumor seeding.

INTEGRATED APPROACH TO MANAGEMENT OF THYROID NODULES

The clinical challenge in thyroid nodule management is to define a diagnostic protocol that will produce the most accurate and cost-effective use of the various diagnostic methods. The implementation of decision analysis appears to highlight the difficulty in attaining this goal because consideration of all possible combinations of tests would require unmanageable decision trees, and the calculation of individual occurrence probabilities depends on an inconsistent literature. Limited decision analysis has suggested that FNA biopsy, the most accurate single evaluation technique, provides a minimal advantage in quality-adjusted life expectancy over the use of the usually inexact trial of thyroid suppression in patients already known to have solid, cold thyroid nodules. Either of these latter approaches appears slightly superior to a decision tree selecting immediate surgery when life expectancy is used as the gauge, but the differences may not be significant.7 On the other hand, analysis of the selection of radionuclide scanning indicates questionable benefit because initial aspiration can provide equal sensitivity and specificity at a lower cost.3,4,6,69 Figure 39-3 suggests an algorithm that may be useful in a practice in which FNA biopsy is frequently used and experienced cytopathologic support is available. Many factors may alter the protocol for individual patients or clinical practice situations. At present, no right or wrong approach exists, and many physicians prefer radionuclide imaging as the initial step. However, this may increase costs because only 5% to 10% of scans obviate the need for aspiration, whereas 60% of 80% of FNA biopsies eliminate scan requirements. The use of scans in identifying multinodularity is largely negated by the high incidence of benign cytology results in these cases. Aspiration and biopsy identify predominantly cystic lesions, which rarely are simple cysts; therefore, ultrasonography has limited use initially but may be of significant value in monitoring the results of suppressive therapy, especially in patients with suspicious lesions and in practices in which the patient may see different clinicians on follow-up. The argument can be made that the 20% to 25% cancer risk in suspicious lesions should lead to immediate surgery in all cases, but a trial of suppressive therapy may still be warranted in selected patients. In many borderline clinical situations, the patient's strong preference or surgical risk factors may be important considerations. When a trial of suppressive therapy is undertaken, the approximate dosage required is >1.7 g/kg per day of T4.81 The dose is increased incrementally by 0.025 mg per day every 5 to 6 weeks with TSH monitoring until TSH levels are reduced to the desired range. Usually the degree of TSH suppression need not be to an undetectable level but rather to a level of 0.1 to 0.3 mU/L. This should be the case particularly in postmenopausal women, in whom possible adverse effects of a fully suppressive dosage on bone mineral density are to be avoided. Nodules are assessed for change in size by physical examination (or ultra-sonography if required) every 6 weeks for the first 6 months. The follow-up intervals may be more prolonged when significant decreases in size are observed, extending eventually to annual follow-up.

FIGURE 39-3. Approach to the solitary nodule. *22- to 25-gauge needle, repeated with 18-gauge needle if fluid is obtained. Evidence of carcinoma (papillary, medullary, poorly differentiated, follicular) or lymphoma. For example, sheets of follicular cells. For example, small groups of uniform follicular cells with little colloid.|| Selected patients with above-normal surgical risk or other extenuating circumstances may be followed closely with suppressive therapy. Changes in cytologic findings redirect clinician to the appropriate arm of the algorithm. (FNA, fine-needle aspiration; T4, thyroxine.)

FNA biopsy should be repeated immediately when a nodule is found to be enlarging on suppressive therapy, with surgical exploration likely to be inevitable unless cystic fluid or hemorrhage with benign cytology is found. FNA should also be repeated if no significant reduction in nodule size is obtained after 6 to 12 months of suppressive therapy. Of repeat FNA biopsies, 98% confirm the original diagnosis.82 Lesions with benign cytology on initial FNA biopsy may be identified as malignancies in only a very few cases on subsequent repeat aspiration.72,82 However, the

prognosis for neoplasms discovered during a later evaluation is unlikely to be different from the prognosis for neoplasms discovered earlier. The authors believe that reaspiration is preferable to the requirement of nodule size reduction as a confirmation of benignity in such cases and that use of this management approach significantly reduces the frequency of unnecessary surgery. Patients with a history of irradiation present a special situation.83 Historically, these patients have been immediately referred for surgery because of their high cancer rate. Some clinicians now advocate FNA biopsy in the management of these patients, although sufficient evidence for the reliability of benign results is still lacking. The surgical approach to thyroid nodules varies. An ipsilateral lobectomy and isthmusectomy are commonly used for single nodules when the preoperative diagnosis is uncertain. Frequently, histologic evaluation of frozen sections is inconclusive or unreliable, and the final diagnosis requires careful examination of permanent sections. When the ultimate diagnosis is carcinoma, the customary practice is to complete a near-total thyroidectomy within 1 week of the first operation.84 Near-total thyroidectomy is the initial procedure of choice in patients with thyroid nodules and a history of thyroid irradiation (see Chap. 44). When only a lobectomy, or lobectomy and isthmusectomy, is adequate (i.e., with benign histopathology), a question of indication for postoperative T4 therapy often arises. In one group of patients with a history of neck irradiation who underwent partial thyroidectomy for nodules, the T4-treated patients had a nodule recurrence rate of 8.4%, compared with a rate of 35.8% in untreated patients.85 However, T4 therapy may not prevent postoperative recurrence of nontoxic goiter.86,87 The importance of referral to a surgeon highly skilled and experienced in thyroid surgery cannot be overemphasized.88 An alternative nonsurgical approach popularized in Europe is the percutaneous injection of 95% ethanol into thyroid nodules previously proved benign by FNA. This approach has been applied to benign thyroid cysts and hyperfunctioning nodules89,90; it also has been applied to cold nodules.91,92,93,94 and 95 The procedure can be very painful for the patient and has been associated with transient increases in serum thyroglobulin and the thyroid hormones, with self-limited thyrotoxicosis. Fever, local pain, hematomata, and vocal cord paralysis are also possible in inexperienced hands. MANAGEMENT OF THYROID CYSTS Simple thyroid cysts are often successfully drained by thyroid aspiration. In many cases, a single aspiration is curative; in other cases, cysts recur even after multiple aspirations. If the initial aspiration leaves no solid residual, the risk of cancer is very low; an additional therapeutic effect can be obtained in recurrent lesions by instilling sclerosing agents such as tetracycline96 or ethanol.89 No convincing evidence exists that thyroid hormone therapy reduces the likelihood of recurrence. Surgery is usually reserved for cysts >4 cm in diameter or for recurrent cysts that produce local symptoms. In cysts defined only by needle aspiration of at least 1 mL of fluid, it may be prudent to consider surgery for lesions that recur after two or more aspirations, especially if the fluid is hemorrhagic, if the nodules are >3 cm in diameter, or if a residual abnormality remains after maximum drainage.97 MANAGEMENT OF AUTONOMOUS NODULES Hot nodules are commonly identified by thyroid scanning.98 The need for definitive treatment is determined by the degree of functional activity of the nodules rather than by their malignant potential. Nontoxic hot nodules are most often followed clinically. Toxic nodules may be treated with antithyroid drugs temporarily but require definitive therapy with 131I or surgery. The choice between these two modalities remains controversial. Surgery is recommended for those with a history of radiation, for children, and for women of childbearing age. The use of 131I usually reduces the nodule's functional activity, but palpable nodules often persist.99 For patients with toxic nodules, doses of 131I (520 mCi) appear to be successful in rendering patients euthyroid, but this may depend on the ratio of dose to nodular area100; higher dosesespecially when given to euthyroid patients with autonomous nodulesproduce hypothyroidism101 (see Chap. 45). Percutaneous injection of ethanol (see earlier) has also been used with some success.90,102 CHAPTER REFERENCES
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. Ashcraft MW, VanHerle AJ. Management of thyroid nodules. I. History and physical examination, blood tests, x-ray tests, and ultrasonography. Head Neck Surg 1981; 3:216. Ashcraft MW, VanHerle AJ. Management of thyroid nodules. II. Scanning techniques, thyroid suppressive therapy, and fine needle aspiration. Head Neck Surg 1981; 3:297. Mazzaferri EL. Management of a solitary thyroid nodule. N Engl J Med 1993; 328:553. Burch HB. Evaluation and management of the solid thyroid nodule. Endocrinol Metab Clin North Am 1995; 24:663. Sheppard MC, Franklyn JA. Management of the single thyroid nodule. Clin Endocrinol 1992; 37:398. Ridgway EC. Clinical review 30: clinician's evaluation of a solitary thyroid nodule. J Clin Endocrinol Metab 1992; 74:231. Molitch ME, Beck JR, Dreisman M, et al. The cold thyroid nodule: an analysis of diagnostic and therapeutic options. Endocr Rev 1984; 5:185. Hermus AR, Huysmans DA. Treatment of benign nodular thyroid disease. N Engl J Med 1998; 338:1438. Gharib H, Mazzaferri EL. Thyroxine suppressive therapy in patients with nodular thyroid disease. Ann Intern Med 1998; 128:386. Ridgway EC. Medical treatment of benign thyroid nodules: have we defined a benefit? Ann Intern Med 1998; 128:403. Rallison ML, Dobyns BM, Meikle AW, et al. Natural history of thyroid abnormalities: prevalence, incidence, and regression of thyroid diseases in adolescents and young adults. Am J Med 1991; 91:363. Jackson SG, Wartofsky L. The thyroid nodule. In: Monaco F, ed. Thyroid diseases: clinical fundamentals and therapy. Boca Raton, FL: CRC Press, 1993:chap VIIA,253. Singer PA, Cooper DA, Daniels GH, et al. Treatment guidelines for patients with thyroid nodules and well differentiated thyroid cancer. Arch Intern Med 1996; 156:2165. Brander A, Viikinkoski P, Nickels J, Kivisaari L. Thyroid gland: US screening in a random adult population. Radiology 1991; 181:683. Carroll BA. Asymptomatic thyroid nodules: incidental sonographic detection. AJR Am J Roentgenol 1982; 138:499. Horlocker TT, Hay JE, James EM. Prevalence of incidental nodular thyroid disease detected during high resolution parathyroid ultrasonography. In: Medeiros-Neto G, Gaitan E, eds. Frontiers in thyroidology, vol 1. New York: Plenum Press, 1986:1309. Brander A, Viikinkoski P, Nickels J, Kivisaari L. Thyroid gland: ultrasound screening in middle aged women with no previous thyroid disease. Radiology 1989; 173:507. Walfish AG, Hazani E, Strawbridge HTG, et al. Combined ultrasound and needle aspiration cytology in the assessment and management of hypo-functioning thyroid nodules. Ann Intern Med 1977; 87:270. Brown CL. Pathology of the cold nodule. Clin Endocrinol Metab 1981; 10:235. Cooper DS, Schneyer CR. Follicular and Hrthle cell carcinoma of the thyroid. Endocrinol Metab Clin North Am 1990; 19:577. Goldman ND, Coniglio JU, Falk SA. Thyroid cancers I: papillary, follicular, and Hrthle cell. Otolaryngol Clin North Am 1996; 29:593. Hamburger JI. Evolution of toxicity in solitary nontoxic autonomously functioning thyroid nodules. J Clin Endocrinol Metab 1980; 50:1089. Hamburger JI. The autonomously functioning thyroid nodule: Goetsch's disease. Endocr Rev 1987; 8:439. Porcellini A, Ciollo I, Laviola L, et al. Novel mutations of thyrotropin receptor gene in thyroid hyperfunctioning adenomas. J Clin Endocrinol Metab 1994; 76:657. Paschke R, Ludgate M. The thyrotropin receptor in thyroid diseases. N Engl J Med 1997; 337:1675. Duprez L, Hermans J, Van Sande J, et al. Two autonomous nodules of a patient with multinodular goiter harbor different activating mutations of the thyrotropin receptor gene. J Clin Endocrinol Metab 1997; 82:306. Russo D, Arturi F, Suarez HG, et al. Thyrotropin receptor gene alterations in thyroid hyperfunctioning adenomas. J Clin Endocrinol Metab 1996; 1548. Marsh DJ, Learoyd DL, Robinson BG, et al. Medullary thyroid carcinoma: recent advances and management update. Thyroid 1995; 5:407. Heshmati HM, Gharib H, van Heerden JA, Sizemore GW. Advances and controversies in the diagnosis and management of medullary thyroid carcinoma. Am J Med 1997; 103:60. Landis SH, Murray T, Bolden S, Wingo PA. Cancer statistics, 1998. CA Cancer J Clin 1998; 48:6. McHenry C, Smith M, Lawrence AM, et al. Nodular thyroid disease in children and adolescents: a high incidence of carcinoma. Am Surg 1988; 54:444. Gorlin JB, Sallan SE. Thyroid cancer in childhood. Endocrinol Metab Clin North Am 1990; 19:649. Hung W. Nodular thyroid disease and thyroid carcinoma. Pediatr Ann 1992; 21:50. Robbins J. Lessons from Chernobyl: the event, the aftermath fallout: radioactive, political, social. Thyroid 1997; 7:189. Becker DV, Robbins J, Beebe GW, et al. Childhood thyroid cancer following the Chernobyl accident. A status report. Endocrinol Metab Clin North Am 1996; 25:197. Sarne D, Schneider AB. External radiation and thyroid neoplasia. Endocrinol Metab Clin North Am 1996; 25:181. Smith RE, Adler RA, Clark P, et al. Thyroid function after mantle irradiation in Hodgkin's disease. JAMA 1981; 245:46. Rosen IB, Simpson JA, Sutcliff S, et al. High-dose radiation and the emergence of thyroid nodular disease. Surgery 1984; 96:988. Blum M, Rothschild M. Improved nonoperative diagnosis of the solitary cold thyroid nodule. JAMA 1980; 243:242. Belfiore A, LaRosa GL, LaPorta GA, et al. Cancer risk in patients with cold thyroid nodules: relevance of iodine intake, sex, age, and multinodularity. Am J Med 1992; 93:363. Spencer CA, Wang C-C. Thyroglobulin measurement: techniques, clinical benefits, and pitfalls. Endocrinol Metab Clin North Am 1995; 24:841. Torrens JI, Burch HB. Serum thyroglobulin measurement: utility in clinical practice. Endocrinologist 1996; 6:125. Vierhapper H, Raber W, Bieglmayer C, et al. Routine measurement of plasma calcitonin in nodular thyroid disease. J Clin Endocrinol Metab 1997; 82:1589. Pacini F, Fontanelli M, Fugazzola L, et al. Routine measurement of serum calcitonin in nodular thyroid diseases allows the preoperative diagnosis of unsuspected sporadic medullary thyroid carcinoma. J Clin Endocrinol Metab 1994; 78:826. Saad MF, Fritsche HA Jr, Samaan NA. Diagnostic and prognostic values of carcinoembryonic antigen in medullary carcinoma of the thyroid. J Clin Endocrinol Metab 1984; 58:889. Cavalieri RR. Nuclear imaging in the management of thyroid carcinoma. Thyroid 1996; 6:485. Patton JA, Sandler MP, Partain CL. Prediction of benignancy of the solitary cold thyroid nodule by fluorescent scanning. J Nucl Med 1985; 26:461. de los Santos ET, Keyhani-Rofagha S, Cunningham JJ, Mazzaferri EL. Cystic thyroid nodules: the dilemma of malignant lesions. Arch Intern Med 1990; 150:1422. Simeone JF, Daniels GH, Mueller PR, et al. High-resolution real-time sonography of the thyroid. Radiology 1982; 145:431. Tan GH, Gharib H. Thyroid incidentalomas: management approaches to nonpalpable nodules discovered incidentally on thyroid imaging. Ann Intern Med 1997; 126:226. Blum M. Non-isotopic imaging of the neck in patients with thyroid nodules or cancer. In: Thyroid cancer: clinical management. Totowa, NJ: Humana Press, 2000, 9. Bloom AD, Adler LP, Shuck JM. Determination of malignancy of thyroid nodules with positron emission tomography. Surgery 1993; 114:728. Wartofsky L. Management of patients with scan negative, thyroglobulin positive differentiated thyroid carcinoma. In: Wartofsky L, Sherman SI, Gopal J, Schlumberger M, Hay ID, eds. The use of radioactive iodine in patients with papillary and follicular thyroid cancer. J Clin Endocrinol Metab 1998; 83:4195. Gharib H, James EM, Charboneau JW, et al. Suppressive therapy with levothyroxine for solitary nodules. N Engl J Med 1987; 317:70. Cheung PSY, Lee JMH, Boey JH. Thyroxine suppressive therapy of benign solitary thyroid nodules: a prospective randomized study. World J Surg 1989; 13:818. Reverter JL, Lucas A, Salinas I, et al. Suppressive therapy with levothyroxine for solitary thyroid nodules. Clin Endocrinol 1992; 36:25. Berghout A, Wiersinga WM, Drexhage HA, et al. Comparison of placebo with L-thyroxine alone or with carbimazole for treatment of sporadic non-toxic goitre. Lancet 1990; 336:193. Celani MF, Mariani M, Mariani G. On the usefulness of levothyroxine suppressive therapy in the medical treatment of benign solitary, solid, or predominantly solid thyroid nodules. Acta Endocrinol 1990; 123:603. Papini E, Bacci B, Panunzi C, et al. A prospective randomized trial of levothyroxine suppressive therapy for solitary thyroid nodules. Clin Endocrinol 1993; 38:507.

60. LaRosa GL, Lupo L, Giuffrida D, et al. Levothyroxine and potassium iodide are both effective in treating benign solitary solid cold nodules of the thyroid. Ann Intern Med 1995; 122:1. 61. Zelmanovitz F, Genro S, Gross JL. Suppressive therapy with levothyroxine for solitary thyroid nodules: a double-blind controlled clinical study and cumulative metaanalyses. J Clin Endocrinol Metab 1998; 83:3881. 62. Papini E, Petrucci L, Guglielmi R, et al. Long-term changes in nodular goiter: a 5-year prospective randomized trial of levothyroxine suppressive therapy for benign cold thyroid nodules. J Clin Endocrinol Metab 1998; 83:780. 63. Lima N, Knobel M, Cavaliere H, et al. Levothyroxine suppressive therapy is partially effective in treating patients with benign, solid thyroid nodules and multinodular goiters. Thyroid 1997; 7:691. 64. Wartofsky L. Does replacement L-thyroxine therapy cause osteoporosis? Adv Endocrinol Metab 1993; 4:157. 65. Marcocci C, Golia F, Bruno-Bossio G, et al. Carefully monitored levothyroxine suppressive therapy is not associated with bone loss in premenopausal women. J Clin Endocrinol Metab 1994; 78:818. 66. Schneider DL, Barrett-Connor EL, Morton DJ. Thyroid hormone use and bone mineral density in elderly women. JAMA 1994; 271:1245. 67. Franklyn J, Betteridge J, Holder R, Sheppard MC. Effect of estrogen replacement therapy upon bone mineral density in thyroxine treated post-menopausal women with a past history of thyrotoxicosis. Thyroid 1995; 5:359. 68. Oertel YC. Fine needle aspiration and the diagnosis of thyroid cancer. Endocrinol Metab Clin North Am 1996; 25:69. 69. Gharib H, Goellner JR. Fine needle aspiration biopsy of the thyroid: an appraisal. Ann Intern Med 1993; 118:282. 70. Gharib H, Goellner JR, Johnson DA. Fine-needle aspiration cytology of the thyroid. A 12-year experience with 11,000 biopsies. Clin Lab Med 1993; 13:699. 71. Hall TL, Layfield LJ, Philippe A, Rosenthal DL. Sources of diagnostic error in fine needle aspiration of the thyroid. Cancer 1989; 63:718. 72. Hamburger JI, Husain M, Nishiyama R, et al. Increasing the accuracy of fine-needle biopsy for thyroid nodules. Arch Pathol Lab Med 1989; 113:1035. 73. Caraway NP, Sneige N, Samaan NA. Diagnostic pitfalls in thyroid fine-needle aspiration: a review of 394 cases. Diagn Cytopathol 1993; 9:345. 74. Leenhardt L, Hejblum G, Franc B, et al. Indications and limits of ultrasound-guided cytology in the management of nonpalpable thyroid nodules. J Clin Endocrinol Metab 1999; 84:24. 75. Hatada T, Okada K, Ishii H, et al. Evaluation of ultrasound-guided fine-needle aspiration biopsy for thyroid nodules. Am J Surg 1998; 175:133. 76. Wartofsky L, Oertel Y. Fine needle aspiration biopsy of thyroid nodules. In: Van Nostrand D, ed. Nuclear medicine atlas. Philadelphia: JB Lippincott, 1988:193. 77. Tani EM, Skoog L, Lwhagen T. Clinical utility of fine-needle aspiration cytology of the thyroid. Annu Rev Med 1988; 39:255. 78. Bckdahl M, Wallin G, Lwhagen T, et al. Fine-needle biopsy cytology and DNA analysis. Surg Clin North Am 1987; 67:197. 79. Pasielka JL, Zedenius J, Auer G, et al. Addition of nuclear DNA content to the AMES-risk group classification for papillary thyroid cancer. Surgery 1992; 112:1154. 80. Davila RM, Bedrossian CWM, Silverberg AB. Immunocytochemistry of the thyroid in surgical and cytologic specimens. Arch Pathol Lab Med 1988; 42:51. 81. Hennessey JV, Evaul JE, Tseng YL, et al. L-Thyroxine dosage: a reevaluation of therapy with contemporary preparations. Ann Intern Med 1986; 105:11. 82. Erdogan MF, Kamel N, Aras D, et al. Value of re-aspirations in benign nodular thyroid disease. Thyroid 1998; 8:1087. 83. DeGroot LJ. Clinical review 2: diagnostic approach and management of patients exposed to irradiation to the thyroid. J Clin Endocrinol Metab 1989; 69:925. 84. Brooks JR, Starnes HF, Brooks DC, Pelkey JN. Surgical therapy for thyroid carcinoma: a review of 1249 solitary thyroid nodules. Surgery 1988; 104:940. 85. Fogelfeld L, Wiviott MBT, Shore-Freedman E, et al. Recurrence of thyroid nodules after surgical removal in patients irradiated in childhood for benign conditions. N Engl J Med 1989; 320:835. 86. Anderson PE, Hurley PR, Rosswick P. Conservative treatment and long term prophylactic thyroxine in the prevention of recurrence of multinodular goiter. Surg Gynecol Obstet 1990; 171:309. 87. Hegedus L, Hansen JM, Veiergang D, Karstrup S. Does prophylactic thyroxine treatment after operation for non-toxic goitre influence thyroid size? BMJ 1987; 294:801. 88. Sosa JA, Bowman HM, Tielsch JM, et al. The importance of surgeon experience for clinical and economic outcomes from thyroidectomy. Ann Surg 1998; 228:320. 89. Lippi F, Ferrari C, Manetti L, et al. Treatment of solitary autonomous thyroid nodules by percutaneous ethanol injection: results of an Italian multicenter study. J Clin Endocrinol Metab 1996; 81:3261. 90. Cho YS, Lee HK, Ahn IM, et al. Sonographically guided ethanol sclerotherapy for benign thyroid cysts: results in 22 patients. AJR Am J Roentgenol 2000; 174:213. 91. Papini E, Pacella CM, Verde G. Percutaneous ethanol injection (PEI): what is its role in the treatment of benign thyroid nodules? Thyroid 1995; 5:147. 92. Caraccio N, Goletti O, Lippolis PV, et al. Is percutaneous ethanol injection a useful alternative for the treatment of the cold benign thyroid nodule? Five years experience. Thyroid 1997; 7:699. 93. Bennedbk FN, Nielsen LK, Hegedus L. Effect of percutaneous ethanol injection therapy versus suppressive doses of L-thyroxine on benign solitary solid cold thyroid nodules: a randomized trial. J Clin Endocrinol Metab 1998; 83:830. 94. Goletti O, Monzani F, Lenziardi M, et al. Cold thyroid nodules: a new application of percutaneous ethanol injection treatment. J Clin Ultrasound 1994; 22:175. 95. Zingrillo M, Collura D, Ghiggi MR, et al. Treatment of large cold benign thyroid nodules not eligible for surgery with percutaneous ethanol injection. J Clin Endocrinol Metab 1998; 83:3905. 96. Treece GL, Georgitis WJ, Hofeldt FD. Resolution of recurrent thyroid cysts with tetracycline instillation. Arch Intern Med 1983; 140:2285. 97. Rosen IB, Provias JP, Walfish PG. Pathologic nature of cystic thyroid nodules selected for surgery by needle aspiration biopsy. Surgery 1986; 100:606. 98. Hamburger JI. The autonomously functioning thyroid nodule: Goetsch's disease. Endocr Rev 1987; 8:439. 99. Goldstein R, Hart IR. Follow-up of solitary autonomous thyroid nodules treated with 131 I. N Engl J Med 1983; 309:1473. 100. Estour B, Millot L, Vergely N, et al. Efficacy of low doses of radioiodine in the treatment of autonomous thyroid nodules: importance of dose/area ratio. Thyroid 1997; 7:357. 101. Ross DS, Ridgway EC, Daniels GH. Successful treatment of solitary toxic thyroid nodules with relatively low-dose iodine 131, with low prevalence of hypothyroidism. Ann Intern Med 1984; 101:488. 102. Tarantino L, Giorgio A, Mariniello N, et al. Percutaneous ethanol injection of large autonomous hyperfunctioning thyroid nodules. Radiology 2000; 214:143.

CHAPTER 40 THYROID CANCER Principles and Practice of Endocrinology and Metabolism

CHAPTER 40 THYROID CANCER

ERNEST L. MAZZAFERRI Radiation and Thyroid Carcinoma Papillary Thyroid Carcinoma Prevalence Familial Papillary Carcinoma Pathology Diagnosis Factors Influencing Prognosis Staging Systems Consensus Views Concerning Therapy Surgery Radioiodine Therapy Thyroid Hormone Therapy Other Therapy Follow-Up Follicular Thyroid Carcinoma Prevalence Pathology Diagnosis Factors Influencing Prognosis Therapy Medullary Thyroid Carcinoma Multiple Endocrine Neoplasia Type 2 Syndromes Prevalence Pathology Diagnosis Factors Influencing Prognosis Therapy Follow-Up Family Screening Chapter References

Approximately 18,100 new cases of thyroid carcinoma were diagnosed in the United States in 1999, ranking it 22nd in incidence among the major malignancies.1 Its frequency varies with gender and age and is highest among women between the ages of 30 and 70 years. The lifetime risk of being diagnosed with thyroid cancer is ~0.64% for women and 0.25% for men. 2 In 1998, ~1200 of the 135,000 persons with thyroid carcinoma in the United States died of their disease.1 Between 1973 and 1992, its incidence among all ages rose steadily (almost 28%), whereas its mortality rates dropped more than 23%.2 The declining mortality rates are largely due to early diagnosis and effective therapy applied at an early tumor stage when it is most amenable to surgery and 131I therapy.3

RADIATION AND THYROID CARCINOMA

Thyroid carcinoma may be caused by exposure to ionizing radiation. It was the first solid tumor found to have a significantly increased incidence in A-bomb survivors, but only among those younger than 20 years of age at the time of exposure.4 External Radiation. The risk of developing papillary thyroid cancer after therapeutic external radiation, used in the past to treat children with benign head and neck conditions, is well known.5 Exposure before the age of 15 years poses a major risk that becomes progressively greater with increasing amounts of radiation between 0.10 Gy (10 rad) and 10 Gy (1000 rad). This increases the incidence of thyroid carcinoma within 5 years of exposure, and this increased incidence continues for 30 years, at which time it begins to decline.5 Girls are only slightly more likely than boys to develop thyroid carcinoma after irradiation.5 Radioiodine-Induced Thyroid Carcinoma. Until recently, most studies have suggested that 131I is less effective than external gamma radiation in inducing thyroid cancer.4 A slightly elevated risk of thyroid cancer found in a large Swedish population exposed to diagnostic doses of 131I was attributed to cancers found among those examined for a suspected thyroid tumor.4 Another study from the United States reported a slight elevation of thyroid cancer mortality after treatment of hyperthyroidism with 131I, but the absolute risk was small, and the underlying thyroid disease appeared to play a role.6,6a However, most of these studies involved adults. When a large number of children developed thyroid carcinoma after being exposed to radioiodine fallout from the Chernobyl nuclear reactor accident in 1986, it became clear that 131I and other short-lived radioiodines were potent thyroid carcinogens in children, particularly those exposed when younger than age 10 years.7 Nuclear Weapon Fallout. Nuclear weapons that were tested above ground in Nevada between 1951 and 1963 resulted in radiation exposure to individuals across the continental United States. The chance of a significant exposure was highest in the 1950s for children who routinely drank milk from a backyard cow or goat that had ingested grass contaminated with radioactive fallout. The average cumulative thyroid dose from radioiodine fallout was 0.02 Gy (2 rad) for the American population collectively and 0.1 Gy (10 rad) for those younger than age 20 years at the time of exposure in the range known to cause thyroid carcinoma in children.5,8 Approximately 50,000 cases of thyroid cancer are likely to result from these exposuresnearly half of which have probably already appearedbut these are highly uncertain estimates.9 Nevertheless, a large study found an association between thyroid carcinoma and exposure to radioiodine fallout from nuclear tests in Nevada for children exposed at younger than 1 year of age and for those in the 1950 to 1959 birth cohort.10 Screening for thyroid disease that is due to fallout exposure is not recommended, but physicians should discuss the risk and palpate the neck of concerned patients.9

PAPILLARY THYROID CARCINOMA

PREVALENCE Thyroid carcinoma is classified into four major types, which in decreasing order of frequency are papillary, follicular, medullary, and anaplastic carcinomas. Papillary and follicular carcinomasoften termed differentiated thyroid carcinomasarise from follicular cells, synthesize thyroglobulin (Tg), and tend to be slow growing. Medullary carcinoma, which originates from thyroidal C cells that secrete calcitonin, may be sporadic or familial. Anaplastic carcinoma usually arises from well-differentiated thyroid tumors and is almost invariably fatal within a brief period (see Chap. 41). Papillary thyroid carcinoma accounts for ~80% of all thyroid cancers in the United States,11 including those induced by radiation.12 It is diagnosed in ~1 in 17,000 people in the United States annually, although occult microscopic papillary thyroidcarcinoma is found in 10% or more of autopsy and surgical thyroid specimens from men and women throughout adult life.13 In contrast, clinically manifest papillary thyroid carcinoma has a peak incidence in the fourth decade and is three times more frequent in women than in men. In children, the sex ratio is nearly equal, which may reflect radiation-induced disease. FAMILIAL PAPILLARY CARCINOMA Approximately 5% of papillary carcinomas are familial tumors, which are sometimes inherited as an autosomal dominant trait.14,15,15a Others are inherited as a component of familial adenomatous polyposis (Gardner syndrome), occurring at a young age as bilateral, multicentric tumors with an excellent prognosis, particularly those with ret-PTC activation.16,17 Cowden disease is an autosomal dominant syndrome characterized by multiple mucocutaneous hamartomas, keratoses, fibrocystic breast disease or breast cancer, and well-differentiated thyroid cancer.18,19 Carney complexa syndrome with spotty skin pigmentation, myxomas, schwannomas, and multiple neoplasia that affects multiple glandsalso includes thyroid carcinoma.20 PATHOLOGY

Gross Tumor. Papillary carcinoma usually is an unencapsulated and invasive tumor with ill-defined margins that is 2 to 3 cm when diagnosed, but this varies widely (Fig. 40-1A).3 Approximately 10% extend through the thyroid capsule into surrounding neck tissues, and another 10% are fully encapsulated.21 They typically are firm and solid, but some develop chronic hemorrhagic necrosis, yielding a thick brownish fluid on needle biopsy that may be mistaken for a benign cyst.22 Small tumors as large as 1.0 cm that often have a stellate appearance, termed microcarcinomas, usually are found by serendipity but rarely are invasive or metastasize (see Fig. 40-1B).23

FIGURE 40-1. Papillary thyroid carcinoma: gross pathology. A, Large papillary thyroid carcinoma completely replacing right thyroid lobe and extending beyond the thyroid capsule. Such lesions are associated with high recurrence and mortality rates. B, Occult papillary thyroid carcinoma found incidentally at surgery. This lesion almost invariably has a benign clinical course.

Architectural Features. Most papillary carcinomas contain complex branching papillae with a fibrovascular core covered by a single layer of tumor cells, intermingled with follicular structures (Fig. 40-2A), although some have a pure follicular or trabecular appearance.24 The term mixed papillaryfollicular carcinoma has no clinical value because the follicular component does not alter 131I uptake or prognosis.25 Moreover, a tumor's nuclear features are more important than its architectural appearance in establishing a diagnosis of papillary carcinoma. Those with cellular features of papillary carcinoma but a pure follicular pattern are termed follicular variant papillary carcinoma (see Fig. 40-2D), which some argue comprise most tumors diagnosed as follicular carcinoma.26

FIGURE 40-2. Papillary thyroid carcinoma: histology. A, Microscopic papillary thyroid carcinoma (arrow) showing a mixed papillary and follicular architecture, and encapsulation. Isolated microscopic papillary carcinomas, found in ~10% of the general population, are incidental findings at surgery or autopsy and almost never are of clinical significance. 20 B, Papillary thyroid carcinoma with typical papillae containing a fibrovascular core, and cells with large, pale-staining nuclei that appear crowded and overlapping. 100 C, Fine-needle aspiration biopsy specimen of papillary thyroid carcinoma showing typical cytologic features of the tumor. 400 D, High-power magnification of follicular variant papillary thyroid carcinoma showing characteristic cellular features, including nuclear inclusions, irregularly placed, pale-staining nuclei, and abundant cytoplasm. 400

Cytology Features. Papillary carcinoma has cellular features that distinguish it from other thyroid neoplasms regardless of its architecture, permitting an accurate diagnosis by fine-needle aspiration (FNA) (see Fig. 40-2C). The cells are large and contain pink to amphophilic finely granular cytoplasm and large pale nuclei (Orphan Annie eye nuclei) with inclusion bodies and nuclear grooves that identify the tumor as papillary carcinoma (see Fig. 40-2D). Psammoma bodies (Fig. 40-3)the ghosts of infarcted papillae that are virtually pathognomonic of papillary carcinomaare calcified, concentric lamellated spheres found in approximately half the cases within or near the tumor and lymph nodes or in cytology specimens.27 Lymphocytic infiltrationranging from focal areas of lymphocytes and plasma cells to classic Hashimoto disease (see Chap. 46)is often seen in papillary carcinoma. Papillary carcinomas are commonly found in multiple sites within the thyroid gland, which are generally thought to be intraglandular metastases. Multiple microscopic tumor foci are found in as many as 80% of cases when the gland is examined in considerable detail, but in routine clinical practice their frequency is as high as 45%, and they usually are bilateral (Fig. 40-4).11

FIGURE 40-3. Psammoma bodies of papillary carcinoma (arrows) are calcified, dark-staining, lamellated spheres that are virtually pathognomonic of papillary thyroid carcinoma. 400

FIGURE 40-4. Microscopically multifocal papillary thyroid carcinoma (arrows) usually represents intraglandular metastases. 20

Lymph Node Metastases. Papillary carcinoma commonly metastasizes to lymph nodes in the lateral neck, central neck compartment, and mediastinum. Gross lymph node metastases are present in approximately half the cases at the time of diagnosis, while even moreas many as 85% in studies from Japan28,29 have microscopic nodal metastases.25 Their number rises as primary tumor size increases. Nodal metastases tend to be bilateral when the isthmus or both thyroid lobes are involved with tumor, or they may extend into the mediastinum or extend beyond the lymph node capsule into soft tissues, which are all poor prognostic signs.25,30 Distant Metastases. Fewer than 5% of patients have distant metastases at the time of diagnosis, and another 5% develop them later.11 In a review of 1231 patients with distant metastases, 49% were in the lung, 25% were in bone, 15% were in lung and bone, and 12% were in the central nervous system or in multiple organs.11 Pulmonary metastases may be large discrete nodules or may have a snowflake appearance caused by diffuse lymphangitic tumor spread (Fig. 40-5A, Fig. 40-5B and Fig. 40-5C). Pulmonary metastases not seen on radiographs may be detected only with whole body scanning done after therapeutic doses of 131I have been administered.31,32,33 and 34

FIGURE 40-5. Distant metastases of papillary thyroid carcinoma. A, Papillary thyroid carcinoma with diffuse lymphangitic spread of tumor giving a typical snowflake appearance on the chest radiograph. Such tumors typically concentrate 131I in younger patients and tend to have a good prognosis. B, Papillary thyroid carcinoma with diffuse nodular infiltrates that concentrated 131I poorly. This patient had pulmonary metastases at the time of initial diagnosis. C, Papillary thyroid carcinoma (left) in a young woman that did not concentrate 131I and grew steadily over a 4-year period (right).

Thyroglossal Duct. Papillary thyroid carcinoma arising within a thyroglossal duct is almost always small and usually has a benign course.35 Papillary Microcarcinoma. Different histologic variants or subtypes of papillary carcinoma have distinct biologic behaviors (Table 40-1). Five histologic subtypes or variants were recognized in the 1988 World Health Organization classification,24 and at least five others have been described since then.36,37 Papillary microcarcinoma is a tumor 1.0 cm or smaller. Approximately 20% are multifocal, and as many as 60% have cervical lymph node metastases,29 which may be their presenting feature.23 Lung metastases occur rarely with multifocal tumors that have bulky cervical metastases; these are the only micro-carcinomas with significant morbidity and mortality.23,37 Otherwise, the recurrence and cancer-specific mortality rates are near zero.13,23

TABLE 40-1. Tumor Histologic Variants That Influence Prognosis in Papillary and Follicular Thyroid Carcinoma

Encapsulated Papillary Carcinoma. Encapsulated papillary carcinoma is an otherwise typical tumor completely surrounded by a fibrous capsule. It accounts for ~10% of papillary carcinomas and is approximately half as likely to metastasize as is typical papillary carcinoma.37 Tumor recurrence is lower than usual, and death that is due to cancer almost never occurs with these tumors.27,37 Follicular Variant Papillary Carcinoma. Follicular variant papillary carcinoma accounts for ~10% of papillary carcinomas.37 Usually not encapsulated, its microfollicular architectural pattern is otherwise indistinguishable from follicular carcinoma, but the typical nuclear features of papillary carcinoma identify its true nature, which can be diagnosed by FNA cytology.27 Its metastases often have psammoma bodies and may show the typical features of papillary carcinoma. Some claim that its prognosis is similar to the usual papillary carcinoma, and others suggest that distant metastases are more likely and that long-term outcome may be unfavorable with this tumor.36,37 and 38 Diffuse Macrofollicular Variant Papillary Carcinoma. Diffuse macrofollicular variant papillary carcinoma is an uncommon tumor that can be confused with goiter or macrofollicular adenoma on frozen section.36 It occurs predominantly in women with goiter, approximately one-third of whom have hyperthyroidism. Most have distant metastases with very high mortality rates.36 Tall Cell Variant. Approximately 10% of papillary thyroid carcinomas show extensive papillae formation with cells twice as tall as they are wide that comprise at least 30% of the tumor.36 Compared with typical papillary carcinoma, tall cell tumors tend to be diagnosed approximately two decades later (in the mid-50s), are larger tumors associated with significantly more invasion into extrathyroidal soft tissues, and have more distant metastases.18,36,37 The tumor, which can be identified on a cytology specimen from fine-needle biopsy, often expresses the p53 suppressor oncogene, perhaps accounting for its frequent loss of 131I uptake and long-term mortality rates that are two- to threefold those of typical papillary carcinomas.18,37 Columnar Cell Variant. Columnar cell variant is a rare variant possibly related to tall cell carcinoma. It occurs mainly in men and is composed of rectangular cells with clear cytoplasm.36 More than 90% develop distant metastases, which usually are unresponsive to 131I therapy or chemotherapy and result in death.36,37 When encapsulated, it has a much better prognosis.21 Diffuse Sclerosis Variant. Approximately 5% of papillary carcinomas are of the diffuse sclerosis variant type, which is even more common in children and adolescents.36,37 Approximately 10% of the tumors in the children of Chernobyl are of this type.39 Usually involving both lobes, diffuse sclerosis variant presents as a goiter with extensive squamous metaplasia, sclerosis, abundant psammoma bodies, and lymphatic invasion involving the whole thyroid gland. Lymph node metastases are almost always present, and ~25% have lung metastases.36,37 Cytology specimens reveal squamous metaplasia, inflammatory cells, and psammoma bodies, but

may be difficult to differentiate from thyroiditis. Although it has a higher incidence of local and pulmonary metastases than typical papillary carcinomas, there is some disagreement about whether its long-term prognosis is worse than usual.36,37 Oxyphilic (Hrthle Cell) Variant. Approximately 2% of papillary carcinomas have nuclei resembling those of Hrthle cell follicular carcinomas.40 Multiple oxyphilic thyroid tumors and a familial occurrence have been noted in some cases.41 This tumor cannot be identified by FNA cytology but is recognized by its papillary architecture on the final histologic sections. Compared with typical papillary carcinomas, oxyphilic carcinomas have fewer neck nodal metastases at diagnosis but have higher rates of recurrence and cause-specific mortality; in this respect, they resemble oxyphilic follicular carcinoma.37,40 Solid or Trabecular Variant. A tumor with a predominantly (>75%) solid architectural pattern that maintains the typical nuclear features of papillary carcinoma, the solid or trabecular variant has a propensity for extrathyroidal spread and lung metastases.36,37 Some, however, find it to be more common in children and report that its prognosis is the same as with typical papillary carcinoma.42 Insular Carcinoma. Approximately 5% of all thyroid carcinomas show solid clusters of cells with small follicles that contain Tg but resemble pancreatic islet cells. This is often categorized as a variant of follicular carcinoma, but it may show papillary differentiation, and it has been suggested that it should be considered a separate entity derived from follicular epithelium.43 Insular carcinomas usually are large and highly invasive tumors that grow through the tumor capsule and into tumor blood vessels. Compared with differentiated thyroid carcinoma, insular carcinoma presents at an older age (54 vs. 36 years); with larger tumors (4.7 vs. 2.5 cm); with fewer neck metastases (36% vs. 50%) but more distant metastases (26% vs. 2%); and has a worse 30-year cancer-specific mortality rate (25% vs. 8%). 18 Insular carcinoma also displays aggressive behavior in children.44 DIAGNOSIS History. In the past, many papillary carcinomas were large and invasive when first diagnosed; however, with the use of FNA, most are small tumors found at an early stage as an asymptomatic thyroid nodule on routine examination or in screening programs of patients with head and neck irradiation.3 The timeliness of diagnosis affects the prognosis (see Prognosis).3 Occasionally, it causes pain, hoarseness, dysphagia, or hemoptysis, or it infiltrates surrounding structures or grows rapidlyfindings that are associated with high mortality rates. However, only a few patients have such symptoms. A history of irradiation is important, but only 30% of patients develop palpable thyroid nodules after irradiation, and of these, only one-third are malignant.45 Thyroid nodules that appear after radiation should undergo an evaluation similar to that used for nodules that occur spontaneously. Physical Examination. Papillary carcinoma usually is manifest as a palpably discrete thyroid nodule that moves upward when the patient swallows, but a cervical lymph node metastasis may be its only sign. A midline mass above the thyroid isthmus may be a metastatic lymph (Delphian) node or may be carcinoma within a thyroglossal duct that can be identified from its upward movement when the tongue is protruded. Papillary carcinoma may be a firm, nontender, discrete mass, but many are soft and cystic or diffusely infiltrate one lobe or the entire thyroid. Distant metastases are found less often at the time of diagnosis of papillary carcinoma than follicular carcinoma, but when they are found, the primary tumor is almost invariably large (see Fig. 40-1A and Fig. 405B). Fine-Needle Aspiration. Approximately 10% of thyroid nodules show clear evidence of malignancy, such as vocal cord paralysis, signs of invasion, or bulky lymph node metastases, whereas the others appear benign on examination. Neither the history nor the physical examination offers sufficient evidence of a nodule's benign nature that further testing can be deferred (see Chap. 34, Chap. 35 and Chap. 39). The first test in a clinically euthyroid patient is FNA.46 Other tests are too nonspecific to be used first. Except for hyperfunctional (hot) nodules that are rarely malignant, most thyroid noduleswhether single in an otherwise normal gland or a dominant nodule in a multinodular goiterrequire FNA for diagnosis. Although large-needle cutting biopsies can be done, FNA is considerably safer and is highly effective in obtaining sufficient cytology to identify the distinctive features of papillary carcinoma, including most of its variants46 (see Chap. 39). Nodules that yield diagnostic or suspicious cytology are excised; the others are scanned to exclude hot nodules, whereas benign nodules are simply observed or sometimes treated with thyroxine-suppressive therapy. Thyroxine suppression should not be done without prior FNA and should not be used as a diagnostic test, because thyroid carcinomas may appear to shrink.46 Prior Neck Irradiation. FNA should be the first test performed on a palpable thyroid nodule in a euthyroid patient with a history of head-and-neck irradiation. An asymptomatic patient with a palpably normal thyroid gland who has history of head and neck irradiation should not undergo ultrasonography because nearly 90% of such patients have thyroid nodules found by this test, most of which are benign.9,47 Ultrasonography reveals thyroid lesions <1 cm in ~50% of the healthy middle-aged population.48,49 Thyroid hormone does not prevent the appearance of thyroid nodules or cancer in previously irradiated patients with a palpably normal thyroid gland,45 but it does reduce the recurrence of nodules in irradiated patients who have undergone thyroid surgery for benign thyroid nodules.50 FACTORS INFLUENCING PROGNOSIS Once the diagnosis is established, the prognosis is determined by an interaction of three variables: tumor stage, the patient's age, and therapy. The cancer-specific mortality rate is <10% over three decades, but distant metastases or serious local recurrences can occur many years after initial therapy.3,11 Delay in Diagnosis. The author of this chapter has found that the median time from the first manifestation of thyroid cancernearly always a neck massto initial therapy was 4 months in patients who survived and 18 months in those who died of cancer (P <.001).3 The 30-year cancer mortality was nearly twice as high when therapy was delayed for longer than a year than when it was done within 12 months of a nodule's discovery (13% vs. 6%, P <.001).3 Age at Time of Diagnosis. Age is the most important prognostic factor. The adverse effect of age appears at ~40 years and becomes progressively worse thereafter, increasing at a steep rate after age 60 years, when men have the worst prognosis.3,11 The response to therapy is most favorable in younger patients whose tumors concentrate 131I.51,52 and 53 Prognosis in Children. Outcome is usually favorable in children, although their tumors are typically at a more advanced stage, with more local and distant metastases than are those of adults at the time of diagnosis.51,53,54 and 55 Recurrence rates in children are ~40% over several decades compared with 20% in adults.3,53 The rate of pulmonary metastases in children is almost twice that in adults, reaching more than 20% in some series.3,11,54 Their prognosis for survival is, however, excellent, with or without a history of irradiation, except for those younger than age 10 years, who have high mortality rates.51,52 and 53 Nonetheless, despite the relatively good prognosis in children, one study found that children with thyroid cancer had, as a group, poorer long-term survival rates than did normal children.56 Gender. Men tend to have higher recurrence and cancer-specific mortality rates than do women.37 I found that the 30-year cancer-specific mortality rate for men with papillary carcinoma was nearly twice that of women (7.8% vs. 4%, P <.01) and that gender was an independent prognostic factor.3 Although estrogen and progesterone receptors are expressed in as many as 50% of papillary carcinomas, this does not explain the divergent risks imposed by gender. Tumor Size. The tumor features affecting prognosis are summarized in Table 40-1 and Table 40-2. Primary tumors <1.5 cm in diameter rarely recur or cause death, whereas those >4.5 cm are associated with high mortality rates.3,11 In the author's study, tumors stratified as <1.5 cm, 1.5 to 4.4 cm, and 4.5 cm, respectively, were associated with distant metastases in 4%, 10%, and 17% of patients and had 30-year cancer-specific mortality rates of 0.5%, 8%, and 22%.3 The 20-year cause-specific mortality rates in the large Mayo Clinic series for tumors 2 to 3.9 cm, 4 to 6.9 cm, and 7 cm were 6%, 16%, and 50%; respectively.57

TABLE 40-2. Factors Influencing the Prognosis of Papillary and Follicular Thyroid Cancer

Multicentricity. Among those undergoing completion thyroidectomy for recurrent cancer, multifocal disease in the thyroid lobe excised first is almost always associated with bilateral thyroid cancer.58 Nonetheless, some report almost no tumor recurrences in the thyroid remnant,59,60 while others find that the rate of recurrence of locally persistent disease is significantly higher after less than near-total thyroidectomy.61,62,63,64 and 65 One study reported a 1.7-fold higher risk of recurrence in multifocal compared with unifocal tumors.66 Another study found that the only two parameters significantly influencing tumor recurrence of papillary microcarcinomas were the number of histologic foci (P <.002) and the extent of initial thyroid surgery.23 Lymph Node Metastases. When lymph nodes are meticulously examined, as many as 85% of cases contain metastatic deposits that correlate with primary tumor size and the presence of multicentric tumor, and reflect aggressive tumor behavior.37 Some find that metastatic lymph nodes have no impact on recurrence or survival.59,67 Others report an increased risk for local tumor recurrence when cervical lymph node metastases are present.3,62,68,69 The prognosis is less favorable with bilateral metastases, mediastinal lymph node metastases, or when tumor is invading through the lymph node capsule into surrounding tissues; these findings correlate with high cancer-specific mortality rates.3,30,70,71,72 and 73 Thyroid Capsular Invasion and Extrathyroidal Extension. As many as one-third of the papillary carcinomas may invade the thyroid capsule, which in its most severe form results in tracheal or spinal cord invasion and compromise of the major vessels. When this occurs, the mortality rate is ~20% at 5 years, which is a 10-fold increase over that of noninvasive tumors.3,11 Distant Metastases. Distant metastases are the main cause of death from papillary carcinoma. The 5-year survival rate of 1231 patients with distant metastases was 53%.11 Long-term survival is common in children and young adults with pulmonary metastases and is most favorable when they are discovered early, are small, and concentrate 131I (see Fig. 40-5).54,74 For example, 10-year survival rates are ~80% in young patients with micronodular pulmonary metastases that concentrate 131I, but are only ~20% with macronodular lung or bone metastases.75 Early scintigraphic diagnoses before the metastases are apparent on chest roentgenograms and their treatment with 131I appear to be the most important elements in prolonging disease-free survival and improving the survival rate.76 Widespread distant metastases may cause thyrotoxicosis, which usually has a poor prognosis. Other Tumor Factors. The histologic variants of this tumor affect prognosis (see Table 40-1 and Table 40-2).18,36 Coexistent Hashimoto thyroiditis is associated with a low tumor stage and may be an independent predictor of a favorable prognosis.77,78 Ana-plastic transformation occurs in well-differentiated thyroid carcinoma, dramatically altering its course and resulting in aggressive local invasion of tumor and widespread, rapidly fatal metastases that do not concentrate 131I.37 Irradiation-Induced Papillary Carcinomas. Often, although large multicentric tumors are associated with more frequent recurrences than are spontaneously occurring papillary carcinomas, their cancer-specific mortality rates are similar.45,79 Graves Disease. The serum of patients with Graves disease stimulates thyroid follicular cells in vitro and can produce progression of thyroid carcinoma.80 One study of papillary carcinoma associated with Graves disease found the tumors were more often multifocal and that the rate of distant metastases was four times higher than usual.81 Other studies have failed to show this effect.57 However, on balance the literature suggests that thyroid cancer occuring with Graves disease is more aggressive than usual.18a Oncogenes. Several oncogenes, including the novel PTC oncogenes that are specific for papillary thyroid carcinoma, have been identified.82,83 and 84 The PTC oncogene, which has been found in many papillary carcinomas in the children of Cherno-byl,85 induces the formation of papillary thyroid carcinoma in transgenic mice, showing that it is specific to the development of this tumor.86,87 To date, no convincing evidence exists that prognosis can be predicted on the basis of a tumor's genetic makeup. STAGING SYSTEMS Several staging systems and prognostic scoring systems have been devised to discriminate between low-risk patients who are anticipated to have a good outcome, thus requiring less aggressive therapy, and higher-risk patients who require the most aggressive therapy to avoid morbidity and/or mortality from thyroid carcinoma.88 Most do not identify the variants of papillary and follicular carcinoma, which have remarkably different behaviors. Their greatest utility is in epidemiology studies and as tools to stratify patients for prospective therapy trials.37 They are least useful in determining treatment for individual patients. CONSENSUS VIEWS CONCERNING THERAPY Three studies have shed light on the current practice. At an international symposium held in 1987 in the Netherlands, 160 specialists from 13 countries recommended total thyroidectomy followed by postoperative 131I thyroid remnant ablation for most patients with differentiated thyroid carcinoma, regardless of their age.89 A second study was based on the responses of 157 thyroid experts to a questionnaire regarding the management of a hypothetical patient with a solitary thyroid nodule.90 The majority recommended total or near-total thyroidectomy followed by 131I ablation of the thyroid remnant; most did not recommend altering this for different tumor histologic types. The third study was based on a survey of the clinical members of the American Thyroid Association who were queried about their long-term management of a hypothetical patient with papillary thyroid carcinoma.91 Most recommended near-total thyroidectomy and 131I ablation, and almost everyone preferred long-term levothyroxine (T 4) therapy in doses sufficient to lower the thyroid-stimulating hormone (TSH) levels to 0.01 to 0.5 U/mL. The majority did not alter therapy for patients with a history of radiation, extremes of age, the presence of a nodule <1 cm, multiple foci in the contralateral lobe, or capsular invasion of the nodule. SURGERY When papillary thyroid carcinoma is diagnosed on FNA, total thyroidectomy can be done without frozen section because of the high specificity of FNA for papillary carcinoma. If the FNA is suspicious for papillary carcinoma, frozen section of the tumor should be done at surgery. When either gross findings or frozen sections suggest malignancy, total thyroidectomy can be performed, because almost all such cases have cancer. If frozen section is not diagnostic of malignancy, a thyroid lobectomy with or without isthmusectomy is recommended, because ~75% are benign lesions.92 Extent of Thyroid Resection. Some prefer lobectomy and regional lymph node dissection as the initial surgery for nearly all patients.93 Others advise total or near-total thyroidectomy for most patients.94,95 Guidelines from the National Cancer Center Consortium advise total thyroidectomy for most patients with papillary and follicular thyroid cancer.95a In most cases, more extensive thyroid surgery should be done, because disease-free survival is improved, even in children and adults with low-risk tumors.95,96,97 and 98 In addition to removing multifocal and bilateral carcinoma, total thyroidectomy affords the opportunity to ablate residual uptake in the thyroid bed with small doses of 131I, facilitating subsequent follow-up. Subtotal Thyroidectomy. Resection of less than a thyroid lobe often done as a nodulectomy is inadequate therapy and is not the current standard of practice.37,99 Even microscopic thyroid carcinoma requires more surgery than subtotal lobectomy.13,23,100 Ipsilateral lobectomy and isthmusectomy may be adequate for microcarcinomas discovered postoperatively on study of the final histologic sections, providing they are unifocal tumors confined to the thyroid in a patient who has not been exposed to significant radiation.3,13,23,100 Complications with this procedure are few, and survival in this group is virtually assured.3,13,23,100 Tumor Recurrence after Subtotal Thyroidectomy. Tumors treated by lobectomy alone have a recurrence rate in the opposite lobe of as much as 10% and have the highest frequency (11%) of subsequent pulmonary metastases, compared with 1% recurrence rates after total thyroidectomy and 131I therapy.11,65 Higher recurrence rates are observed with cervical node metastases and multicentric tumors, justifying a more aggressive surgical approach, particularly in those older than 40 years (see Chap. 43).3 Near-Total or Total Thyroidectomy. At least near-total thyroidectomy (ipsilateral total lobectomy, isthmusectomy, and nearly total contralateral lobectomy) should be performed for tumors that are either 1.5 cm in diameter or multicentric (any size), or are either metastatic or invade the thyroid capsule. Modified neck dissection that preserves the sternocleidomas-toid muscle is done for involved lateral cervical lymph nodes. Radical neck dissection is only done for tumors extensively invading the

strap muscles.25 Completion Thyroidectomy. If only partial lobectomy has been performed, it is best to consider completion thyroidectomy for lesions that are anticipated to have the potential for recurrence, because large thyroid remnants are difficult to ablate with 131I.101 Completion thyroidectomy has a low complication rate and is appropriate to perform routinely for tumors 1 cm, because as many as 40% of patients have residual carcinoma in the contralateral thyroid lobe.102,103 When there has been a local or distant tumor recurrence, carcinoma is found in >60% of the excised contralateral lobes.58 A study of children from Chernobyl found that completion thyroidectomy allowed for the diagnosis and treatment of recurrent cancer and lung or lymph node metastases in 61% of patients in whom residual carcinoma was not preoperatively recognized.97 In another study, patients who underwent completion thyroidectomy within 6 months of their primary operation developed significantly fewer lymph node and hematogenous recurrences and survived significantly longer than those in whom the second operation was delayed for longer than 6 months.104 Radioiodine-Assisted Surgery. The completeness of surgical excision of recurrent or persistent disease can be improved by giving 100 mCi 131I to patients with functioning lymph node metastases and locating the tumor with the aid of an intraoperative probe.105 In one study, it detected both suspected and unsuspected lesions in 56% of the patients, although ~25% had nodal metastases that were undetected by this and other techniques.105 Surgical Complications. Hypoparathyroidism and damage to the recurrent laryngeal nerve are the main surgical complications, which are highest after total thyroidectomy. Rates of hypoparathyroidism as high as 5% are reported in adults,106 and even higher rates are reported in children97,107 undergoing total thyroidectomy. However, one study reported a 5.4% rate of hypocalcemia after total thyroidectomy that persisted in only 0.5% of the patients 1 year after surgery.108 In a review of seven published surgical series, the mean rates of permanent recurrent laryngeal nerve injury and hypoparathyroidism, respectively, were 3% and 2.6% after total thyroidectomy and 1.9% and 0.2% after subtotal thyroidectomy.109 Hypoparathyroidism occurs at a lower rate when experienced surgeons perform the surgery and the posterior thyroid capsule is left intact on the contralateral side. A study of 5860 patients found that surgeons who performed more than 100 thyroidectomies a year had the lowest complication rates (4.3%), which were four-fold lower than those who performed <10 cases annually.110 Thyroidectomy During Pregnancy. Thyroid carcinoma during pregnancy may occasionally progress rapidly, perhaps because of high maternal b-human chorionic gonadotropin levels that have a TSH-like effect.111 Nonetheless, most tumors are slow growing and have an excellent prognosis during pregnancy, and surgery can be delayed until delivery in most women.112 RADIOIODINE THERAPY Sodium-Iodide Symporters. Iodide is concentrated much less avidly by differentiated thyroid carcinoma than by normal thyroid tissue, perhaps because of abnormalities in the sodium-iodide symporters (NIS) that have been identified in differentiated thyroid carcinomas. One study found increased NIS activity in papillary carcinoma,113 and others have found reduced NIS activity and heterogeneous immunohistochemical NIS staining in differentiated thyroid carcinoma.114,115 Preparation for 131I Therapy. In preparation for 131I therapy, all women of childbearing age must have a pregnancy test unless they have undergone a tubal ligation or hysterectomy. Recombinant human TSH (rhTSH) can be given to raise the serum TSH sufficiently to perform a postoperative total-body 131I scan (see Follow-Up) but is not yet approved for use in 131I therapy. However, in special cases in which thyroid hormone withdrawal is not possible or is ineffective, rhTSH may be used sucessfully in the preparation of patients for 131I therapy.115a,115b Radioiodine is given ~6 weeks after surgery when serum TSH levels have risen enough (>30 U/mL) to stimulate neoplastic and normal thyroid tissues to concentrate 131I maximally. Tri-iodothyronine (Cytomel), 1 g/kg per day (~25 g orally, two or three times daily) is given for the first 4 weeks after surgery, then is discontinued for 2 weeks, and serum TSH usually rises above 30 U/mL.116 The patient must avoid iodine during this period, especially in the form of drugs and iodine-rich foods. During the last 2 weeks, a low-iodine diet should be ingested117 and the serum TSH and Tg levels should be measured just before performing the diagnostic 131I scan. Lithium. Tumor 131I retention is enhanced by lithium, which decreases the release of iodine from normal thyroid and tumor cells.118 Given at a dosage of 400 to 800 mg daily in divided doses (10 mg/kg) for 7 days, it increases 131I uptake in metastatic lesions while only slightly increasing uptake in normal tissue.118 Blood lithium levels should be measured daily and maintained between 0.8 and 1.2 nmol/L, which prolongs the biologic half-life of 131I without altering the amount of whole-body radiation. Lithium was found to increase 131I retention in 24 of 31 metastatic lesions and in 6 of 7 thyroid remnants. Comparing 131I retention during lithium treatment with that during the control period showed that the mean increase in the retention half-life was 50% in tumors and 90% in remnants. An increase in the accumulated 131I and the lengthening of the effective half-life (biologic turnover and isotope decay) combined to increase the estimated 131I radiation dose in metastatic tumor by an average of more than two-fold, which was greatest in tumors with initially low 131I uptake.118a Radiation to tumors with a short biologic half-life (<6 days) is maximized without increasing radiation to other organs. Whole-Body 131I Scan and the Stunning Effect. A whole-body scan is obtained 24 to 72 hours after giving 2 to 4 mCi of 131I. Larger scanning doses should not be given, because focal abnormalities not seen with 2 to 4 mCi doses are unlikely to be ablated successfully.119 Moreover, 131I doses as small as 3 mCi diminish the subsequent uptake of therapeutic doses of 131I.120 Termed the stunning effect, this is presumably due to follicular cell damage induced by large scanning doses of 131I that decrease uptake in the thyroid remnant or metastases for several weeks, thus impairing the therapeutic efficacy of 131I.121 After a large scanning dose of 131I is given, there is an increase in serum Tg that is associated with a higher rate of incomplete ablation, perhaps reflecting a stunning effect on the thyroid remnant.122 However, 2- or 3-mCi doses of 131I or the use of 131I that avoids the stunning effect are slightly less sensitive than larger scanning 131I doses in identifying thyroid remnants.121,122 Thyroid Remnant Ablation. Because it is nearly impossible to remove all thyroid tissue with routine surgery, uptake of 131I is almost always seen in the thyroid bed postoperatively, which must be ablated before 131I will optimally concentrate in metastatic deposits.97,123 There are three compelling reasons to ablate a thyroid remnant. First, a large thyroid remnant can obscure 131I uptake in cervical or lung metastases because they accumulate 131I optimally only in the absence of normal thyroid tissue.123 Second, high levels of circulating TSH are necessary to enhance tumor 131I uptake, which cannot be achieved with a large thyroid remnant.116 Third, serum Tg measurements are the most sensitive test for carcinoma when they are measured during hypothyroidism after the thyroid bed uptake has been ablated.124 Nonetheless, there continues to be debate concerning the use of 131I to ablate uptake in the thyroid bed after near-total thyroidectomy.57,64 Recurrence after 131I Ablation. Patients with tumors who have the potential for recurrence are given 131I postoperatively to ablate the thyroid remnant.64 A large and growing number of studies demonstrate decreased recurrence of papillary carcinoma and decreased disease-specific mortality attributable to 131I therapy (Fig. 40-6).3,63,64,68,125,126,127 and 128 The lowest incidence of pulmonary metastases occurs after total thyroidectomy and 131I. In one study, recurrences in the form of pulmonary metastases, analyzed as a function of initial therapy of papillary or follicular carcinoma, was reported to be as follows: thyroidectomy plus 131I (ablation dose of 100 mCi), 1.3%; thyroidectomy alone, 3%; partial thyroidectomy plus 131I, 5%; and partial thyroidectomy alone, 11%.65

FIGURE 40-6. Differentiated thyroid carcinoma recurrence rates after different types of medical therapy. (Adapted from data published in Mazzaferri EL, Jhiang SM. Long term impact of initial surgical and medical therapy on papillary and follicular thyroid cancer. Am J Med 1994; 97:418.)

Dose for Thyroid Remnant Ablation. Thyroid remnant ablation usually can be achieved with a dose of 30 to 50 mCi of 131I, which appears to be as effective as larger doses in preventing tumor recurrence.3,125,129 This has been a popular way to avoid hospitalization, but is no longer necessary in most states because of a
131I

change in federal regulations that permits the use of much larger 131I doses in ambulatory patients.130 Nonetheless, considering the large differences in cost and radiation exposure and the fact that doses >50 mCi do not substantially improve the rate of successful ablation, it is reasonable to use a 30- to 50-mCi dose for remnant ablation.129 Some use a dosimetry calculation (see Quantitative Tumor Dosimetry) that delivers 30,000 rad to the thyroid remnants; in one study, this was achieved with a mean 131I dose of almost 87 mCi that completely ablated 86% of the remnants.101 Increasing the 131I dose to deliver >30,000 rad does not increase the success rate.101,129 Response rates are significantly lower when patients have less than a total or near-total thyroidectomy or have a thyroid remnant calculated to be >2 g.101 Therapeutic 131I for Residual or Recurrent Carcinoma. Residual or recurrent carcinoma should be treated surgically whenever possible; however, only ~50% to 75% of differentiated thyroid carcinomas and their metastases and approximately one-third of Hrthle cell carcinomas concentrate 131I.131,132 and 133 One study found that two-thirds of 283 patients with lung or bone metastases had tumors that concentrated 131I.134 This is crucial to survival. For example, 10-year survival rates in one study were 83% or 0%, respectively, depending on whether pulmonary metastases did or did not concentrate 131I.75 The lung metastases that concentrate 131I best are the smallest lesions found in young patients.75 Empiric Fixed Doses. There are basically three approaches to therapy: empiric fixed doses, upper bound limits set by blood dosimetry, and quantitative dosimetry.130 With the first, a fixed amount of 131I is given based on what is being treated. For example, 30 to 50 mCi are given to ablate thyroid remnants, 150 to 175 mCi for residual carcinoma in cervical nodes or neck tissues, and 200 mCi or more for distant metastases. Tumor 131I uptake in amounts adequate for imaging with 4-mCi scanning doses is usually sufficient for 131I therapy, using empiric doses from 30 to 200 mCi.135 Upper Bound Limits Set by Blood Dosimetry. Blood dosimetry is done to establish an upper limit on the amount of 131I that can be given safely, which is generally considered to be 200 rad to the whole blood from a single dose.136 Quantitative Tumor Dosimetry. The third approach is to calculate the dose of 131I that is required to deliver 30,000 rad to ablate the thyroid remnant or 8000 to 12,000 rad to treat nodal or discrete soft tissue metastases. For pulmonary metastases, the amount of 131I is administered that delivers 200 rad to whole blood with no more than 80 mCi of whole blood retention at 48 hours.130 The two most important factors in determining success are the mass of residual tissue and the effective half-time of 131I in that tissue.130 An 80% response was found in tumor deposits that received at least 8000 rad.101 Lesions that receive <3000 to 4000 rad from 150 to 200 mCi 131 I should be considered for alternative therapy. Repeat 131I Treatments. As long as metastatic deposits concentrate 131I, treatment should be continued every 6 to 12 months until the tumor has been ablated or adverse effects are seen. Repeat doses of 131I should not be given until the bone marrow has fully recovered from the previous dose. Few adverse effects occur with this approach to 131I therapy.137 Immediate Risks of 131I Therapy. There are few immediate serious risks of 131I therapy, except when brain or spinal cord metastases are present that can undergo edema and hemorrhage 12 hours to 2 weeks after treatment.138 Severe radiation thyroiditis can occur within a week of administering a large dose of 131I to a patient who has undergone only lobectomy, causing pain, swelling, and (rarely) airway compromise that may require prednisone therapy.139 Thyroid storm, a rare occurrence, may appear ~2 to 10 days after administering a therapeutic dose of 131I to a patient with a large functioning tumor burden.130 Some experience acute bone pain after being treated with 131I. Approximately 4 to 12 hours after the oral administration of 200 mCi or more of 131I, two-thirds of patients develop mild radiation sickness characterized by headache, nausea, and occasional vomiting that resolves in ~24 hours.130 This is rarely seen with 131I doses <200 mCi. Patients with extensive tumor may rarely develop transient vocal cord paralysis.130 Facial nerve paralysis has been reported in one patient given a very high dose of 131I.130 Radiation cystitis does not occur if the patient is well hydrated. Mild radiation sialadenitis, leukopenia, and a slight drop in the number of platelets often occur ~6 weeks after therapy, but ordinarily these effects are mild and usually transient.140 Parotid Dysfunction. Having the patient suck on hard lemon candy increases salivary flow, which decreases but does not prevent the effects of salivary gland 131I radiation. Transient parotid swelling reminiscent of Stensen duct obstruction may occur for nearly a year after 131I therapy. In one study, ~60% of patients reported side effects lasting >3 months, including sialoadenitis (33%) and transient loss of taste or smell (27%).141 More than a year after the last 131I treatment, 43% experienced reduced salivary gland function, and more than 4% had complete xerostomia, both of which were related to the cumulative dose of 131I.141 Nearly 23% of the patients reported chronic or recurrent conjunctivitis.141 Radiation Pneumonitis. Pulmonary fibrosis may occur after 131I therapy for widespread pulmonary metastases, but this is rare when the whole body retention is <80 mCi 48 hours after treatment. Most diffuse pulmonary metastases can be treated with 150 mCi 131I without risking pulmonary fibrosis.142 Smaller 131I doses ~100 mCi can be given when there is diffuse and intense uptake of the scanning dose in the lungs.33 Leukemia and Other Bone Marrow Effects. There is a small risk of developing acute myelogenous leukemia after 131I therapy, estimated at 3 to 22 excess cases per 1000 patients treated with 131I, depending on the cumulative dose.130,137 However, the lower estimate seems more likely based on a Swedish study that found two leukemia cases among 834 thyroid carcinoma patients treated with 131I, which was not statistically significant.143 When 131I doses are given at 12-month intervals and total cumulative doses are limited to 500 mCi in children and 800 mCi in adults, long-term effects on the bone marrow are minimal, and few cases of leukemia occur.137,144,145 In practice, this is usually not a problem, because tumor tissue that concentrates 131I is likely to be ablated by several treatments, leaving either no residual tumor or metastases that do not concentrate 131I. Cumulative doses >800 mCi are given to patients with extensive metastatic disease, because the risk posed by the thyroid cancer outweighs that due to radiation. Trivial uptake of 131I in the neck or elsewhere that cannot be ablated is not a reason for administering large cumulative 131I doses. Cancer Caused by 131I Therapy. Small increases in the incidence of colon, breast, bladder, and salivary cancer have been found in some studies of 131I therapy for thyroid carcinoma, but not in others.130,143 This underscores the need for laxatives and hydration after 131I treatment, especially for hypothyroid patients. Infertility and Gonadal Failure. Large doses of 131I are a risk to gonadal damage, but it is infrequently observed.130,146 A European study of 2113 pregnancies in women with thyroid carcinoma who had been treated with surgery and 131I found that the miscarriage rate increased from 11% before surgery to 20% after surgery and remained at this level after 131I therapy.147 Miscarriages were more frequent in women who were treated during the year preceding conception, but whether this was related to gonadal irradiation or to insufficient control of hormonal thyroid status was uncertain.147 Testicular germinal cell function may be transiently impaired when men are given 131I therapy for thyroid carcinoma. 148 Because this occasionally is permanent, it seems prudent to advise young men to bank their sperm before therapy. THYROID HORMONE THERAPY Levothyroxine Thyroid-Stimulating Hormone Suppression. Differentiated thyroid carcinomas contain TSH receptors, and TSH stimulates their growth.80 Therapy with T4 significantly reduces recurrence rates and cancer-specific mortality rates (see Fig. 40-6).3 The T4 dose needed to attain serum TSH levels in the euthyroid range is greater in those with thyroid cancer (2.11 g/kg per day) than in patients with primary hypothyroidism associated with nonmalignant disease (1.62 g/kg per day).149 For patients who have undergone total thyroid ablation for thyroid carcinoma, the T4 dosage necessary to achieve an undetectable basal serum TSH level that does not increase after thyrotropin-releasing hormone administration is 2.7 0.4 (SD) g/kg per day.150 One study found that a constantly suppressed TSH (.05 U/mL) was associated with a longer relapse-free survival than serum TSH levels that were always 1 U/mL, and that the degree of TSH suppression was an independent predictor of recurrence.151 However, a prospective study of 617 patients in the National Thyroid Cancer Treatment Cooperative Study found that disease stage, patient age, and 131I therapy independently predicted disease progression, but the degree of TSH suppression did not.152 Hence, these data do not support the concept that great degrees of TSH suppression are required to prevent disease progression. As a practical matter, the most appropriate dose of T4 for most patients with thyroid carcinoma is that which reduces the serum concentration to just below the lower limit of the normal range for the assay being used. Complications. Cardiovascular abnormalities, which are well recognized in overt thyrotoxicosis, also occur in those taking suppressive doses of T4.153 Among the cardiovascular problems associated with subclinical thyrotoxicosis are an increased risk of atrial fibrillation,154 a higher 24-hour heart rate, more atrial premature contractions per day, and not only increased cardiac contractility but also ventricular hypertrophy.153,155,156 Patients with thyroid carcinoma treated with suppressive doses of T4 have a high rate of bone turnover that decreases acutely after withdrawing treatment.157 This is of most concern in postmenopausal women, but using the smallest T4 dose necessary to suppress TSH has been shown to have no significant effects on bone

metabolism and bone mass in men or women with thyroid carcinoma.158 OTHER THERAPY Retinoic Acid. A few patients may benefit from retinoic acid, a drug that in vitro partly redifferentiates follicular thyroid carcinoma. In one study, retinoic acid given orally (1.18 0.37 mg/kg) for at least 2 months to 12 patients with differentiated carcinoma that could not be treated with other modalities induced significant 131I uptake in two patients.159 This response was associated with a rise in serum Tg concentration, suggesting tumor redifferentiation. Surgical Excision of Metastases. Focal lesions that do not concentrate 131I adequately and isolated skeletal metastases should be considered for surgical excision or external irradiation.68 Life-threatening tumor refractory to all other forms of therapy may be palliatively treated with doxorubicin (Adriamycin), although the response rate is poor.11 FOLLOW-UP Follow-up consists of examination; neck ultrasonography; chest radiography; and, if postoperative thyroid 131I ablation has been performed, whole-body 131I scans and serum Tg determinations (Fig. 40-7). Patients with clinically significant tumors (>1.5 cm) should be evaluated every 6 to 12 months for 10 years. Whole-body 131I scan performed 12 months after surgery and 131I ablation often can document complete absence of tumor. Thereafter, scanning can be done at infrequent intervals unless there is a change in the examination or a rise in Tg.

FIGURE 40-7. Use of recombinant human thyroid-stimulating hormone (TSH) in the management of differentiated thyroid carcinoma. Top shows method of recombinant human TSH (rhTSH) administration, and bottom shows algorithm for its use. (CT, computed tomography; IM, intra-muscularly; MRI, magnetic resonance imaging; Tg, thyroglobulin.)

RECOMBINANT HUMAN THYROID-STIMULATING HORMONE Periodic withdrawal of thyroid hormone therapy is required during follow-up. This causes symptomatic hypothyroidism to raise the serum TSH concentrations sufficiently to stimulate thyroid tissue so that 131I scanning and serum Tg measurements can be obtained. Intramuscular administration of rhTSH stimulates thyroidal 131I uptake and Tg release while the patient continues thyroid hormone suppression therapy, thus avoiding symptomatic hypothyroidism.160,161 Now approved for diagnostic use, rhTSH has been tested in two large international multicenter studies. The first study found that whole-body 131I scan results done after two 0.9-mg doses of rhTSH (while thyroid hormone therapy was continued) were of good quality; they were equivalent to the scans obtained after thyroid hormone withdrawal in 66% of the patients, superior in 5%, and inferior in 29%.161 This study found that rhTSH stimulates 131I uptake for whole-body scanning, but the sensitivity of scanning after rhTSH administration was less than after the withdrawal of thyroid hormone.161 Scanning with rhTSH was associated with significantly fewer symptoms and dysphoric mood states. A second multicenter international study was done to test two dosing schedules of rhTSH on the results of whole-body scans and serum Tg levels compared with those obtained after thyroid hormone withdrawal. The whole-body 131I scanning method was more carefully standardized in the second study than in the first.162 The scans in this study were concordant in 89% of the patients, with superior whole-body scans seen in 4% of the subjects after rhTSH and in 8% after thyroid hormone withdrawal, differences that were not statistically significant. The combination of whole-body scanning and serum Tg measurements detected 93% of the patients with disease or tissue limited to the thyroid bed and detected 100% of the patients with metastatic carcinoma.162 Although not yet approved for preparation of patients for 131I therapy, rhTSH has been used successfully for this purpose.163 Recombinant human TSH, 0.9 mg, is given intramuscularly every day for 2 days, followed by a minimum of 4 mCi of 131I on the third day and a whole-body scan and Tg measurements on the fifth day (see Fig. 40-7). Whole-body 131I images are acquired after 30 minutes of scanning or after obtaining 140,000 counts. A serum Tg of 2.0 ng/mL obtained 72 hours after the last rhTSH injection indicates that thyroid tissue or thyroid carcinoma is present, which almost always can be identified on the rhTSH-stimulated whole-body scan using the indicated scanning method.162 The drug is well tolerated, with mild headache and nausea being its main adverse effects. Serum Thyroglobulin Measurement. Serum Tg determinations and whole-body 131I imaging together detect recurrent or residual disease in most patients who have undergone total thyroid ablation. The serum Tg concentration reflects the mass of normal thyroid tissue or differentiated thyroid carcinoma, the degree of thyroid physical damage or inflammation, and the level of TSH receptor stimulation.164 Serum anti-Tg antibodies should be measured in the sample obtained for serum Tg assay because these antibodies, which are found in as many as 25% of patients with thyroid carcinoma, invalidate serum Tg measurements in most assays.124,165 Tg measurement is more sensitive when T4 has been stopped or rhTSH is given to elevate the serum TSH.31,162 Under these circumstances, serum Tg has a lower false-negative rate than whole-body 131I scanning.31,162 Detecting serum Tg by a newly introduced Tg mRNA method is a more sensitive marker of residual thyroid tissue or cancer than measuring Tg by immunometric assay, particularly when Tg mRNA is detected during T4 treatment or with circulating anti-Tg antibodies.166 The author uses a sensitive Tg immunometric assay with a detection limit of 0.5 ng/mL. A serum Tg above this level during T4 therapy after total or near-total thyroidectomy and 131I ablation has been achieved is a sign of persistent normal tissue or differentiated thyroid carcinoma, which is an indication for repeat scanning when there is no other evidence of disease (see Fig. 40-7). If serum Tg rises above 10 ng/mL after T 4 is discontinued or rises above 2.5 ng/mL after rhTSH is administered, normal or malignant thyroid tissue is usually present, even if the 2- to 4-mCi 131I diagnostic scan is negative (i.e., <1% 131I uptake).94,162,164 In this case, neck ultrasonography, magnetic resonance imaging, or other scans are performed to detect occult tumor that can be excised. However, if tumor is not found and the serum Tg is >10 ng/mL, the author gives a therapeutic dose of 131I, usually 100 to 150 mCi, and perform a posttreatment scan. In the author's experience, ~20% of such patients have lung metastases. Others use different cutoff values and different doses of 131I, but the Tg level to trigger treatment has gradually come down.34 Scanning Patients with High Serum Thyroglobulin Concentrations and Negative 131I Scans. Several radionuclide scanning techniques may be used to identify the location of tumor in patients with high serum Tg levels, negative diagnostic 131I scans, and negative neck ultrasonography. Thallium-201. Thallium-201 (201Tl) scintigraphy may identify metastases or uptake in the neck when the serum Tg is elevated but the diagnostic whole-body 131I scan is negative.167 In a study comparing 201Tl and 131I scans, the sensitivities and specificities, respectively, were 94% and 96% for 201Tl, and 29% and 100% for 131I.168 However, another study found that 131I is much more sensitive than 201Tl in demonstrating residual thyroid tissue after surgery (100% and 33%, respectively).169 Others also find that 131I scintigraphy is more sensitive and more specific than 201Tl scintigraphy in identifying both residual neck uptake and metastases.170 In a study of 36 paired whole-body 131I and 201Tl scan results, residual uptake in the neck was seen on all the 131I scans but on only 17% of the 201Tl scans.170 In this study, multiple metastatic lesions identified on 14 131I scans were interpreted as negative or nonspecific, or as showing fewer lesions on the corresponding 201Tl scans. Sixteen 201Tl scans gave false-positive results. The authors concluded that 131I scintigraphy is more sensitive and more specific than 201Tl scintigraphy for detecting distant metastases and for identifying residual activity in the neck after thyroidectomy.170 Other studies have found the two scanning techniques approximately equally useful in detecting local recurrences or distant metastases.171 Technetium-99m. Technetium-99m (99m Tc) may localize differentiated thyroid carcinoma. One study found that the sensitivities of 201Tl, 99m Tc-tetrofosmin, and 131I in identifying distant metastases were comparable (85%, 85%, and 78%, respectively).169 However, 131I was much more sensitive than 99mTc-tetrofosmin for demonstrating remnant thyroid tissue after surgery (100% and 33%, respectively).169 Scanning with 99m Tc-methoxyisobutyl isonitrile (99m Tc-MIBI) also detects

metastases of thyroid carcinoma and may be useful in the postoperative follow-up. One large study found increased accumulation of 99m Tc-MIBI in 75% of the patients with lung metastases, in 100% of those with lymph node metastases, and in 94% of the patients with bone metastases.172 The 99m Tc-MIBI scans identified more lesions in the lung than did 201Tl or 131I scans, which, respectively, were positive in 80% and 85% of patients with lung metastases, in 100% and 42% of patients with lymph node metastases, and in 90% and 87% of patients with bone metastases.172 Whole-Body Positron Emission Tomography. Whole-body positron emission tomography (PET) with 18F-fluorodeoxyglucose (FDG) may identify differentiated thyroid carcinoma metastasis that cannot be identified by scintigraphy with 131I or 99m Tc. Although PET has better sensitivity, resolution imaging, and spatial localization, this has to be balanced with its higher cost when compared with thallium scintigraphy.173 False-positive 18F-fluorodeoxyglucose uptake may occur with benign lung disease.174 FDG-PET scans are useful in predicting survival in differentiated thyroid cancer. In one study, multivariate analysis demonstrated that the single strongest predictor of survival was the volume of disease that displayed avidity for FDG-PET. The probability of surviving 3 years with FDG volumes of 125 mL or less was 96% compared with 18% in those with a FDG volume >125 mL. All of the 10 patients with distant metastases and negative PET scans were alive and well at the end of the study, whereas those with positive PET scans were more likely to die of disease.174a False-Positive 131I Scans. Body secretions, transudates, inflammation, nonspecific mediastinal uptake, and neoplasms of nonthyroidal origin may concentrate 131I.175 This also can be seen with physiologic secretion of 131I from the nasopharynx, salivary and sweat glands, stomach, genitourinary tract, and from skin contamination with sputum or tears.176 Pathologic pulmonary transudates and inflammation that is due to cysts, as well as lung lesions caused by fungal and other inflammatory disease, may produce false-positive scans. Diffuse hepatic uptake of 131I is rarely due to occult liver metastases but usually is from hepatic clearance of Tg labeled with 131 I by functioning thyroid remnants or extrahepatic thyroid cancer metastases. The more 131I uptake that appears in the residual thyroid tissue, the more it appears in the liver. In one large study,177 diffuse hepatic 131I uptake was seen in 60% of 399 patients undergoing 131I scans and in nearly 36% of 1115 131I scintigraphy studies. Twelve percent of the diagnostic scans in this study showed uptake in the liver. The frequency of hepatic uptake in the posttherapy scans was related to the dose of 131I as follows: 39% with 30 mCi; 61.5% with 75 to 100 mCi; and 71.3% with 150 to 200 mCi.177 Patients whose 131I scans show hepatic uptake without uptake in the thyroid bed or in extrahepatic metastases, however, often have occult liver metastases.177 I Treatment of Patients with Negative 131I Scans and High Serum Thyroglobulin. When the serum Tg level is elevated and a tumor cannot be found by localizing techniques, pulmonary metastases are sometimes found only after administrating therapeutic doses of 131I.34,178 The Tg cutoff level has been gradually coming down for treating patients with large doses of 131I whose only evidence of disease is an elevated serum Tg. It was ~30 or 40 ng/mL 10 years ago, but now is closer to 10 ng/mL when the patient is off thyroid hormone.34 There is increasing evidence that this treatment is beneficial. Multivariate analyses have shown the prognostic importance of the size of pulmonary metastases at the time of discovery.179 Another multivariate analysis of prognostic factors in 134 patients with pulmonary metastases showed that early (normal chest roentgenogram) scintigraphic diagnosis and 131I therapy for lung metastases were the most important elements in obtaining both a significant improvement in survival rate and a prolonged disease-free time interval.76
131

Two studies32,33 found beneficial effects of 131I therapy for patients with high serum Tg levels, a normal chest roentgenogram, and a negative diagnostic 131I scan. Of the patients reported in these two series, 80% achieved a negative 131I on posttherapy scan, 60% had a serum Tg <5 ng/mL off thyroid hormone, six of eight patients had a normalization of the computed tomography scan, and two patients had negative lung biopsies.32,33 This may occur with one or two 131I treatments, although complete resolution of pulmonary metastases after 131I therapy is difficult to achieve.54 When a partial response with reduction of metastatic disease is achieved, patients usually have a good quality of life with no further disease progression and a low mortality rate.54

FOLLICULAR THYROID CARCINOMA

PREVALENCE Follicular thyroid carcinoma accounts for ~5% to 10% of all thyroid cancers in the United States.11 It occurs at a slightly older age than papillary carcinoma but in recent years has been diagnosed earlier; in some studies, almost half of patients are younger than 40 years of age at the time of diagnosis.180 This tumor is rare in children, occurs infrequently after head and neck irradiation, and is not commonly found at autopsy as an occult tumor. PATHOLOGY Follicular carcinoma, which is usually encapsulated even when it is aggressive, may initially appear benign on FNA, gross tumor inspection, and frozen-section study; this must be differentiated from follicular variant papillary carcinoma and follicular adenoma.11 Aggressive tumors show extensive vascular invasion or gross capsular penetration and satellite tumor foci around the periphery of the neoplasm (Fig. 40-8A). Most are single lesions that usually do not show necrotic degeneration. Large aggressive tumors extend into the opposite lobe or may lie in adjacent cervical tissue; they should not be mistaken for lymph node metastases, which ordinarily occur only in the more advanced cases with distant metastases.

FIGURE 40-8. Follicular thyroid carcinoma. A, Large follicular thyroid carcinoma (arrow) with satellite lesions above the primary. The primary tumor contains areas of hemorrhagic necrosis. B, Follicular thyroid carcinoma showing capsular invasion. The tumor, at the top of the photomicrograph, shows a finger-like extension into the capsule and small areas of tumor just below the capsule near its center. 100 C, Follicular thyroid carcinoma showing capsular and vascular invasion. There is a large vessel engorged with tumor in the center of the photo. 100 D, High-power view of follicular thyroid carcinoma containing microfollicles and cells with small, fairly uniform nuclei, which on high power show multiple mitotic figures. 400

Microscopically, these are usually compact, highly cellular tumors composed of microfollicles, trabeculae, and solid masses of cells (see Fig. 40-8B, Fig. 40-8C and Fig. 40-8D). Less often, follicular carcinoma has medium-sized or large follicles and such low invasive characteristics that it is difficult to differentiate from a benign adenoma. Such carcinomas have an excellent prognosis.11 Follicular carcinoma has compact, dark-staining, round nuclei that are more uniform in shape, size, and location than the nuclei of papillary carcinoma and are difficult or impossible to identify as carcinoma by FNA.11,46 Hrthle Cell Carcinoma. Oxyphilic cells, termed Hrthle or Askanazy cells, which contain increased amounts of acidophilic cytoplasm with numerous mitochondria on electron microscopy, may constitute most or all of a follicular carcinoma. Some consider Hrthle cell neoplasms to be a distinct clinicopathologic entity; others consider them to be variants of follicular thyroid cancer.26 Regardless of their classification, Hrthle cell carcinomas have a less favorable prognosis than nonoxyphilic follicular carcinomas, although they initially may not appear less differentiated or more invasive.11 Classification. Follicular carcinoma can be classified as minimally or highly invasive. Minimally invasive tumors show just enough evidenceslight penetration of the tumor capsuleto make a diagnosis of carcinoma. Others contain multiple foci of vascular and capsular penetration but are fairly discrete masses. Highly invasive tumors have satellite nodules. Distant Metastases. Follicular carcinoma tends to metastasize to lung, bone, the central nervous system, and other soft tissues with greater frequency than does papillary carcinoma, and the metastases often avidly concentrate 131I (Fig. 40-9 and Fig. 40-10). Also, unlike papillary carcinoma, small lesions can metastasize widely,

although tumors >4 cm are associated with a much higher mortality rate.181

FIGURE 40-9. A, Follicular thyroid carcinoma with a large isolated nodular pulmonary metastasis. B, Hrthle cell carcinoma with multiple nodular pulmonary and bone metastases that did not concentrate 131I.

FIGURE 40-10. Follicular carcinoma metastatic to pelvis. A, Radiograph showing large osteolytic lesions of pubic bones. B, Follicular carcinoma, 131I uptake in lesions shown in A. Note that 131I uptake is more extensive than might be predicted from the radiograph.

DIAGNOSIS Follicular carcinoma may present as an asymptomatic neck mass, usually without palpable cervical lymph nodes, or a distant metastasis may be the first sign of tumor. Metastases may appear as large discrete pulmonary nodules, osteolytic bone lesions causing pathologic fractures, or as central nervous system tumors with serious neurologic sequelae, which are rarely seen in the absence of a palpable thyroid lesion.11 Bulky metastatic lesions may be functional and can cause thyrotoxicosis, which can be triiodothyronine toxicosis (see Chap. 42). The diagnostic evaluation is similar to that for papillary thyroid carcinoma, except a problem is usually encountered in differentiating benign Hrthle cell tumors and follicular adenomas from their malignant counterparts by FNA or by study of the frozen sections at surgery.95a Typically, such tumors are simply designated as follicular neoplasms, because their benign or malignant character cannot be determined. Large-needle aspiration biopsies and cutting-needle biopsies usually yield results similar to those of FNA but have more serious complications.46 Some physicians suggest that large-needle biopsy, if it can be done efficiently and safely, may be diagnostically useful for follicular nodules 3 cm that demonstrate suspicious cytology on FNA. However, most recommend surgical excision of thyroid nodules that are suspicious on FNA46,95a (see Chap. 39). FACTORS INFLUENCING PROGNOSIS The 10-year mortality from follicular carcinoma is ~10%182 but ranges to 50%, depending on the degree of vascular invasion and capsular penetration by the tumor.11,182,183 and 184 Hrthle cell carcinomas have the worst prognosis.183 Tumor recurrence in distant sites is seen more often with follicular than with papillary carcinomas and occurs most frequently with highly invasive tumors, when the primary lesions are >4 cm, and in Hrthle cell carcinomas.11,180,181,183 Marked cellular atypia or frank ana-plastic transformation is also associated with a poor prognosis. Women with this tumor may fare better than men do, but age has the most important influence on prognosis. Patients younger than 40 years of age at the time of diagnosis have the best prognosis, whereas older patients tend to have higher recurrence and mortality rates from tumors that often do not concentrate 131I.11 Ten- and 15-year survival rates with distant metastases (~30% and 10%, respectively) are lowest among patients older than 40 years of age.11,181 THERAPY Appropriate treatment can improve survival with this tumor. Relapse, which occurs less often with surgery followed by 131I therapy,11,180 is seen least often in distant sites after total thyroidectomy and 131I therapy.65 Survival rates are best if the tumor initially concentrates 131I and can be completely ablated with the initial surgery and 131 I therapy.11 The argument for total or near-total thyroidectomy for follicular carcinoma is stronger for tumors that are more highly invasive, for Hrthle cell carcinomas, and for primary tumors >4 cm in diameter. 11,181,184 Less extensive thyroid surgery may be adequate for tumors that are minimally invasive, because such tumors ordinarily have an excellent prognosis95a,185; however, because this distinction cannot be made preoperatively, most clinicians prefer total thyroidectomy for follicular carcinoma. When the diagnosis of follicular carcinoma is not initially appreciated (frozen-section histologic analysis is notoriously inaccurate with this tumor), only lobectomy and isthmusectomy are usually done. If the permanent histologic sections show the tumor to be malignant, then complete thyroidectomy should be done within 2 to 3 days.11 In most clinics, patients with moderately to extensively invasive follicular carcinomas are routinely given 131I postoperatively to ablate residual 131I uptake in the thyroid bed. When the tumor is metastatic, the principal treatment is 131I, as described for papillary carcinoma, provided the tumor concentrates the isotope. If it does not, treatment with external radiation or doxorubicin may be effective.95a Thyroid hormone suppression is always given unless the patient is thyrotoxic from metastatic tumor tissue, which produces thyroid hormone.

MEDULLARY THYROID CARCINOMA

Medullary thyroid carcinoma (MTC), first recognized in 1959, is a pleomorphic neoplasm with amyloid struma that arises from the calcitonin-secreting C cells of the thyroid. Approximately 20% are familial tumors that are transmitted as an autosomal dominant trait and are often associated with other endocrine neoplasms. The genes responsible for the familial forms of MTC map to the pericentromeric region of chromosome 10.186 MULTIPLE ENDOCRINE NEOPLASIA TYPE 2 SYNDROMES MTC may occur in four settings. Sporadic MTC is a unilateral thyroid tumor with no somatic lesions other than the neoplasm. Multiple endocrine neoplasia type 2A (MEN2A) is familial, bilateral MTC with hyperparathyroidism and bilateral pheochromocytoma. Multiple endocrine neoplasia type 2B (MEN2B) may be sporadic or familial and features bilateral MTC, bilateral pheochromocytoma, an abnormal phenotype with multiple mucosal ganglioneuromas, and musculoskeletal abnormalities

suggestive of the marfanoid habitus. Familial non-MEN MTC (FMTC) comprises bilateral MTC with no other endocrine tumors or somatic abnormalities.187 MEN2A is always inherited as an autosomal dominant trait, whereas MEN2B may be similarly transmitted or can occur sporadically. FMTC is an autosomal dominant trait that is the least common form, and manifests at a later age.187 Medullary Thyroid Carcinoma. The MTC in both MEN2 syndromes is generally bilateral and multicentric, as opposed to sporadic MTC, which is usually unilateral. The C cells are normally located in the upper and middle thirds of the lateral thyroid lobes. In MEN2, the C cells initially undergo hyperplasia, which precedes the development of MTC. Parathyroid Disease. One-third to one-half of the patients with the MEN2A syndrome have hyperparathyroidism, most of whom (85%) have parathyroid hyperplasia.188 In contrast, almost no patients with MEN2B have parathyroid disease.189 Patients with MEN2A seldom present with symptoms of hypercalcemia but often form kidney stones. Parathyroid hyperplasia is discovered during MTC surgery in most MEN2A patients, even those without clinical or biochemical evidence of hyperparathyroidism.189 Pheochromocytoma. Adrenal medullary disease, ranging from diffuse or nodular hyperplasia to large bilateral multilobular pheochromocytomas, occurs in both MEN2 syndromes. Pheochromocytomas or medullary hyperplasia is typically bilateral and occurs in ~40% of patients, with a range of 6% to 100% in different kindreds.188 The symptoms are typically subtler than those encountered with sporadic pheochromocytoma.190 The diagnosis is usually established by demonstrating high urinary or plasma catecholamine levels, but the total urinary catecholamines may be normal, with only an increased epinephrine:norepinephrine ratio, particularly in those with adrenal medullary hyperplasia190 (see Chap. 86). Multiple Endocrine Neoplasia Type 2B. MEN2B is characterized by a constellation of somatic abnormalities consisting of ganglioneuromas of the tarsal plates and the anterior third of the tongue, and a marfanoid habitus. Nodules may occur in the lips, causing them to appear lumpy and patulous. Ganglio-neuromas in the alimentary tract may be associated with constipation, diarrhea, and megacolon. The marfanoid characteristics include long limbs, hyperextensible joints, scoliosis, and anterior chest deformities, but not ectopic lens or cardiovascular abnormalities189 (see Chap. 188). Point mutations of the RET protooncogene have been identified in germline and tumor DNA of unrelated patients from kindreds with MEN2A, MEN2B, and FMTC.186,191,192 A relationship exists between specific RET protooncogene mutations and the disease phenotype in MEN2.192 Although several different and independent point mutations in the genomic sequence of the RET protooncogene have been identified, all involving codons for cysteine residues, the normal function of RET is not yet known, and its role in the development of these inherited neoplasms remains unclear. Nevertheless, the identification of point mutations provides an important direct means of identifying affected MEN and FMTC kindreds. PREVALENCE MTC accounts for ~10% of all thyroid malignancies. Approximately 80% to 90% of MTC cases occur sporadically, and 10% to 20% are inherited. MTC may occur at any age, but sporadic disease is diagnosed later in life than is familial MTC. The median age of patients seen at the Mayo Clinic with sporadic MTC was 51 years, compared with 21 years for those with familial tumors.190 MEN2A and FMTC are both characterized by the development of bilateral MTC, but the age at onset of FMTC is usually later, ranging from 40 to 50 years, as compared with an average of 20 to 30 years for MEN2A. Familial MTC occurs with equal frequency in both sexes, while sporadic MTC has a female-male ratio of 1.5:1. PATHOLOGY Sporadic and familial MTC are histologically similar, although a wider spectrum of appearances is encountered in familial tumors, ranging from isolated hypertrophied C cells to large bilateral multicentric tumors that are usually in the superior portions of the thyroid lobes. Typically, the tumor is composed of fusiform or polygonal cells surrounded by irregular masses of amyloid and abundant collagen (Fig. 40-11). Calcifications are present in approximately one-half of the tumors, and occasionally trabecular bone formation is seen. Calcitonin can usually be demonstrated in the tumor by immunohistochemical studies.

FIGURE 40-11. Medullary thyroid carcinoma showing trabecular architecture with spindle and round cells and abundant stroma that stained positively with Congo red, for amyloid. 100

Cervical node metastases occur early in the disease and can be seen with primary lesions as small as several millimeters. Tumors >1.5 cm in diameter are more likely to metastasize to distant sites. The pattern of tissue calcitonin staining may differentiate virulent from less aggressive tumors. In one study, patients with primary tumors that showed intense homogeneous calcitonin staining were all clinically well on follow-up examination, whereas patients whose tumors showed patchy localization of calcitonin either developed metastatic disease or died of cancer within 6 months to 5 years of initial surgery.193 DIAGNOSIS Clinical Features. Patients with sporadic disease or previously unrecognized familial MTC usually present with one or more painless thyroid nodules in an otherwise normal gland, but the tumor may cause pain, dysphagia, and hoarseness. The dominant sign may be enlarged cervical lymph nodes or, occasionally, distant metastases, most commonly to the lung, followed in frequency by metastases to the liver, bone (osteolytic or osteoblastic lesions), and brain. In approximately one-half of patients with sporadic MTC, cervical lymph node metastases are present at the time of diagnosis. The thyroid nodule may be cold or normal on radionuclide imaging and is usually solid on echography and malignant on FNA. Radiographs may show dense, irregular calcifications of the primary tumor and cervical nodes, and mediastinal widening that is due to metastases. Rarely, an abnormal phenotype may correctly identify the tumor, or a paraneoplastic syndrome may be the presenting manifestation. Hormonal Features. In addition to calcitonin, MTC may synthesize calcitonin gene-related peptide, L -dopa decarboxylase, serotonin, prostaglandins, adrenocorticotropin, histaminase, carcinoembryonic antigen, nerve growth factor, and substance P.190 Elevated calcitonin, histaminase, L -dopa decar-boxylase, and carcinoembryonic antigen serum levels occur frequently in MTC patients, but not in other thyroid cancers. Approximately 10% of MTC patients have episodes of flushing often induced by alcohol ingestion, calcium infusion, and pentagastrin injection; these episodes may be due to tumor release of prostaglandins and serotonin.190 The only recognized clinical manifestation of high circulating calcitonin levels is a secretory diarrhea that occurs in ~30% of patients and is usually seen only with advanced tumors. Because of the indolent course of MTC, the secretion of adrenocorticotropin by the tumor may cause typical Cushing syndrome (see Chap. 75 and Chap. 219). Calcitonin Determination. Assay of plasma calcitonin is helpful in the early diagnosis of C-cell hyperplasia and MTC and has influenced the management of these disorders, particularly the MEN2 syndromes. Elevated basal plasma calcitonin levels are found in almost all patients with a palpable MTC and correlate directly with tumor mass.190 Basal calcitonin may not be elevated in patients with small tumors and is almost invariably normal in those with C-cell hyperplasia; however, the plasma calcitonin levels increase to abnormally high levels after stimulation with calcium or pentagastrin. Pentagastrin, 0.5 g/kg, is given intravenously over 5 seconds, and blood samples are taken at 0, 2, and 5 minutes. Depending on the calcitonin antibody used, this stimulus causes calcitonin levels to rise approximately three- to fivefold

with MTC or C-cell hyperplasia.190 Calcium and pentagastrin infused together is more reliable than either used alone because some patients respond to one but not the other. In the combined test, 2 mg/kg of elemental calcium is infused over 60 seconds, followed by 0.5 g/kg of pentagastrin infused over 5 to 10 seconds, and plasma calcitonin is measured before and 1, 2, 3, 5, and 10 minutes after infusion. Depending on the calcitonin antibody used, the plasma calcitonin level rises approximately five-fold with MTC or C-cell hyperplasia if the basal calcitonin level is minimally elevated. Diagnosis of familial cases is now possible with genetic screening long before the thyroid tumor is clinically manifest or calcitonin levels are elevated (see Factors Influencing Prognosis). Hypercalcitoninemia is not absolutely diagnostic of MTC because it occurs in other conditions.190 When the differential diagnosis of a high plasma calcitonin value is between MTC and another malignancy, a higher calcitonin value and a palpable thyroid tumor usually can identify patients with MTC. In addition, calcitonin secretion by other cancers is poorly stimulated by pentagastrin190 (see Chap. 53). FACTORS INFLUENCING PROGNOSIS MTC is much more aggressive than papillary and follicular carcinoma and has a cancer-specific mortality of ~20% at 10 years.182 Also, a significant number of deaths occur from pheo-chromocytoma.194 The survival rate is substantially worse with sporadic tumors or when metastases are found at the time of diagnosis, with the MEN2B phenotype, and among patients older than 50 years of age at the time of diagnosis. However, the most important prognostic factors are age and tumor stage at the time of diagnosis and the presence of residual disease postoperatively.195,196,196a Prognosis is best with FMTC and MEN2A. Early detection and treatment has a profound impact on the clinical course of MTC. The 10-year survival rates are nearly similar to those in unaffected subjects when nodal metastases are not present, but fall to ~45% with nodal metastases.190 Patients operated on during the first decade of life generally have no evidence of residual disease postoperatively.190 However, persistent or recurrent disease is seen in approximately one-third of patients operated on in the second decade and gradually increases in frequency with age until the seventh decade, when approximately two-thirds of patients have persistent disease after surgery.190 This is largely due to the clinical stage of the disease at the time of surgery.195,196,196a Before 1970, MTC was usually diagnosed in the fifth or sixth decade. With periodic calcitonin screening, affected patients with MEN kindred have been diagnosed at a much earlier age, usually in the second decade or earlier, when they have C-cell hyperplasia or microscopic carcinoma confined to the thyroid.197 Now, with the availability of genetic testing, affected patients can be identified at birth.198 However, the sensitivity of genetic screening for MEN2A offered by diagnostic laboratories that limit RET analysis to exons 10 and 11 is ~83%.199 Genetic testing that includes RET exon 14 results in a more complete and accurate analysis with a sensitivity approaching 95%.199 It is recommended that clinicians confirm the comprehensiveness of a laboratory's genetic screening approach for MEN2A to ensure thoroughness of sample analysis. THERAPY Initial Surgery. Surgery offers the only chance for cure and should be performed as soon as the disease is detected.195,196 Before thyroidectomy, however, pheochromocytoma must be rigorously searched for and excised. The treatment of MTC confined to the neck is total thyroidectomy because the disease is often multicentric, even in patients with a negative family history who are often unsuspected relatives of affected MEN2 kindred.195,196 Because cervical node metastases occur early and adversely influence survival, all patients with palpable MTC or clinically occult disease that is visible on cut section of the thyroid should undergo routine dissection of lymph nodes in the central neck compartment. The lateral lymph nodes should be dissected when they contain tumor, but radical neck dissection is not recommended unless the jugular vein, accessory nerve, or sternocleidomastoid muscle is invaded by tumor.107 Residual or Recurrent Medullary Thyroid Carcinoma. Residual or recurrent MTC, manifested by elevated calcitonin levels, occurs commonly after primary treatment of the tumor. Reoperation in appropriately selected patients is the only treatment that consistently and reliably reduces calcitonin levels and may result in excellent local disease control. Although reoperative neck microdissections can normalize calcitonin levels when metastatic MTC is confined to regional lymph nodes, there is no curative therapy for widely metastatic disease. Improved results are reported with surgical management of recurrent MTC, mainly through better preoperative selection of patients and the institution of routine laparoscopic liver examination preoperatively, which identifies distant metastases in patients with normal computed tomography and magnetic resonance imaging.200 Patients with widely metastatic MTC often live for years, but many develop symptoms secondary to tumor persistence or progression. Judicious palliative, reoperative resection of discrete, symptomatic lesions provides significant long-term relief of symptoms with minimal operative mortality and morbidity.201 Patients with metastatic MTC causing significant symptoms or physical compromise may respond to palliative reoperative resection despite the presence of widespread incurable metastatic disease.201 Inoperable Disease. Patients with inoperable disease are often given palliative treatment, with external radiation for localized disease, or with doxorubicin or other chemotherapy combinations for widespread, life-threatening disease, which is of limited benefit. Radiolabeled metaiodobenzylguanidine and111 In-octreotide are potentially useful in palliative care. Octreotide does not improve the natural course of advanced stages of MTC.202 Initial reports of the aggressive use of radioimmunotherapy with radiolabeled monoclonal antibodies against carcinoembryonic antigen in patients with far advanced disease appear hopeful. The phase I studies, which show the safety of administering high myeloablative doses of 131I-MN-14 F(ab)2 labeled carcinoembryonic antigen, are encouraging but require confirmation.202a FOLLOW-UP The efficacy of surgery for MTC is assessed postoperatively by measurement of plasma calcitonin levels, which may require as long as 6 months to normalize. A normal basal and provoked calcitonin level after surgery indicates cure in most cases. Persistent modest basal calcitonin elevation is often seen after surgery in patients who may remain well for many years, particularly those from a MEN2A kindred who should be observed without further aggressive therapy. When plasma calcitonin levels are extremely high or when diarrhea occurs postoperatively, metastases can be localized by neck palpation, chest radiography, computed tomography, isotope bone (but not liver) scans, liver biopsy, angiography, and venous catheterization with calcitonin measurement. Both 99m Tc-dimercaptosuccinic acid and111 In-octreotide studies have similar sensitivity to localized primary MTC; however, these scans do not detect small lymph node involvement (micrometastases) before initial surgery.203 Unfortunately, both scans have no clinical implication for preoperative MTC staging. FAMILY SCREENING Genetic Screening. Genetic screening should be done in all first-degree relatives of any patient who tests positive for a MEN2 or FMTC mutation. One study found that without genetic testing, even when the mean age at the time of thyroidectomy was ~10 years, a significant number of patients (21%) with MEN2A or MEN2B have persistent or recurrent MTC over a follow-up time of approximately a decade.204 Prophylactic Total Thyroidectomy. In 1993, when mutations of the RET protooncogene were found to account for hereditary MTC, surgeons gained the opportunity to prophylactically operate on patients at an asymptomatic stage or before the disease became clinically manifest. With this approach, microscopic or grossly evident MTC is often present in the excised thyroid glands, but almost none of the patients have metastasis of their MTC to regional lymph nodes at the time of surgery.198 Almost all patients are biochemically cured with prophylactic total thyroidectomy, which can be performed safely in experienced centers.107 Prophylactic total thyroidectomy done by an experienced surgeon is recommended at age 6 years for patients who test genetically positive for MEN2A.107,205 Thyroidectomy should be done at an even earlier age for children who test positive for MEN2B, because of its aggressive behavior.204,206 Central neck lymph node dissection should be included when calcitonin levels are elevated or if patients are older than 10 years.107 Surgical Management of Hyperparathyroidism. Surgical management of hyperparathyroidism in those with MEN2A is controversial. Some advocate total parathyroidectomy and heterotopic autotransplantation for all MEN2A patients with hyperparathyroidism.207 Others advocate preservation of the parathyroids, which obviates the potential morbidity associated with total parathyroidectomy and autotransplantation.208,209 Screening for Pheochromocytoma. Before thyroidectomy is performed, those with MEN2 also should have catecholamine and metanephrine measurements. All patients with MTC require postoperative follow-up of serum calcitonin levels. CHAPTER REFERENCES
1. Landis SH, Murray T, Bolden S, Wingo PA. Cancer statistics, 1999. CA Cancer J 1999; 49:8.

2. 3. 4. 5. 6.

Kosary CL, Ries LAG, Miller BA, et al. SEER cancer statistic review, 1973 1992: tables and graphs. Bethesda, MD: National Cancer Institute. NIH Pub. No. 96-2789, 1995. Mazzaferri EL, Jhiang SM. Long-term impact of initial surgical and medical therapy on papillary and follicular thyroid cancer. Am J Med 1994; 97:418. Hall P, Holm LE. Radiation-associated thyroid cancer: facts and fiction. Acta Oncol 1998; 37:325. Ron E, Lubin JH, Shore RE, et al. Thyroid cancer after exposure to external radiation: a pooled analysis of seven studies. Radiat Res 1995; 141:259. Ron E, Doddy MM, Becker DV, et al. Cancer mortality following treatment for adult hyperthyroidism. JAMA 1998; 280:347.

6a.Franklyn JA, Maisonneuve P, Sheppard M, et al. Cancer incidence and mortality after radioiodine treatment for hyperthyroidism: a population-based cohort study. Lancet 1999; 353:2111. 7. Jacob P, Goulko G, Heidenreich WF, et al. Thyroid cancer risk to children calculated. Nature 1998; 392:31. 8. National Cancer Institute. Estimated exposures and thyroid doses received by the American people from iodine-131 in fallout following Nevada atmospheric nuclear bomb tests: a report from the National Cancer Institute. Washington, DC: US Department of Health and Human Services, National Institutes of Health, National Cancer Institute, 1997. 9. Institute of Medicine. Committee on exposure of the American people to I-131 from the Nevada atomic-bomb test: implications for public health. Washington, DC: National Academy Press, 1999. 10. Gilbert ES, Tarone R, Bouville A, Ron E. Thyroid cancer rates and 131 I doses from Nevada atmospheric nuclear bomb tests. Am J Med 1998; 90:1654. 11. Mazzaferri EL. Thyroid carcinoma: papillary and follicular. In: Mazzaferri EL, Samaan N, eds. Endocrine tumors. Cambridge, England: Blackwell Scientific Publications Inc., 1993:278. 12. Robbins J, Schneider AB. Radioiodine-induced thyroid cancer: studies in the aftermath of the accident at Chernobyl. Trends 1998; 9:87. 13. Moosa M, Mazzaferri EL. Occult thyroid carcinoma. Cancer J 1997; 10:180. 14. Burgess JR, Duffield A, Wilkinson SJ, et al. Two families with an autosomal dominant inheritance pattern for papillary carcinoma of the thyroid. J Clin Endocrinol Metab 1997; 82:345. 15. Malchoff CD, Sarfarazi M, Tendler B, et al. Familial papillary thyroid carcinoma is genetically distinct from familial adenomatous polyposis coli. Thyroid 1999; 9:247. 15a.Lupoli G, Vitale G, Caraglia M, et al. Familial papillary thyroid microcarcinoma: a new clinical entity. Lancet 1999; 353:37. 16. Bell B, Mazzaferri EL. Familial adenomatous polyposis (Gardner's syndrome) and thyroid carcinoma: a case report and review of the literature. Dig Dis Sci 1993; 38:185. 17. Cetta F, Olschwang S, Petracci M, et al. Genetic alterations in thyroid carcinoma associated with familial adenomatous polyposis: clinical implications and suggestions for early detection. World J Surg 1998; 22:1231. 18. Burman KD, Ringel MD, Wartofsky L. Unusual types of thyroid neoplasms. Endocrinol Metab Clin North Am 1996; 25:49. 19. Dahia PLM, Marsh DJ, Zheng ZM, et al. Somatic deletions and mutations in the Cowden disease gene, PTEN, in sporadic thyroid tumors. Cancer Res 1997; 57:4710. 20. Stratakis CA, Courcoutsakis NA, Abati A, et al. Thyroid gland abnormalities in patients with the syndrome of spotty skin pigmentation, myxomas, endocrine overactivity, and schwannomas (Carney complex). J Clin Endocrinol Metab 1997; 82:2037. 21. Evans HL. Encapsulated columnar-cell neoplasms of the thyroid: a report of four cases suggesting a favorable prognosis. Am J Surg Pathol 1996; 20:1205. 22. de los Santos ET, Keyhani-Rofagha S, Cunningham JJ, Mazzaferri EL. Cystic thyroid nodules: the dilemma of malignant lesions. Arch Intern Med 1990; 150:1422. 23. Baudin E, Travagli JP, Ropers J, et al. Microcarcinoma of the thyroid gland: the Gustave-Roussy Institute experience. Cancer 1998; 83:553. 24. Hedinger C, Williams ED, Sobin LH. The WHO histological classification of thyroid tumors: a commentary on the second edition. Cancer 1989; 63:908. 25. Mazzaferri EL, Young RL, Oertel JE, et al. Papillary thyroid carcinoma: the impact of therapy in 576 patients. Medicine (Baltimore) 1977; 56:171. 26. LiVolsi V, Asa SL. The demise of follicular carcinoma of the thyroid gland. Thyroid 1994; 4:233. 27. LiVolsi VA. Papillary lesions of the thyroid. In: LiVolsi VA, ed. Surgical pathology of the thyroid. Philadelphia: WB Saunders, 1990:136. 28. Noguchi M, Yamada H, Ohta N, et al. Regional lymph node metastases in well-differentiated thyroid carcinoma. Int Surg 1987; 72:100. 29. Sugino K, Ito K Jr, Ozaki O, et al. Papillary microcarcinoma of the thyroid. J Endocrinol Invest 1998; 21:445. 30. Yamashita H, Noguchi S, Murakami N, et al. Extracapsular invasion of lymph node metastasis is an indicator of distant metastasis and poor prognosis in patients with thyroid papillary carcinoma. Cancer 1997; 80:2268. 31. Pacini F, Lari R, Mazzeo S, et al. Diagnostic value of a single serum thyroglobulin determination on and off thyroid suppressive therapy in the follow-up of patients with differentiated thyroid cancer. Clin Endocrinol 1985; 23:405. 32. Pineda JD, Lee T, Ain K, et al. Iodine-131 therapy for thyroid cancer patients with elevated thyroglobulin and negative diagnostic scan. J Clin Endocrinol Metab 1995; 80:1488. 33. Schlumberger M, Arcangioli O, Piekarski JD, et al. Detection and treatment of lung metastases of differentiated thyroid carcinoma in patients with normal chest X-rays. J Nucl Med 1988; 29:1790. 34. Schlumberger M, Mancusi F, Baudin E, Pacini F. 131-I therapy for elevated thyroglobulin levels. Thyroid 1997; 7:273. 35. Yoo KS, Chengazi VU, O'Mara RE. Thyroglossal duct cyst with papillary carcinoma in an 11-year-old girl. J Pediatr Surg 1998; 33:745. 36. LiVolsi VA. Unusual variants of papillary thyroid carcinoma. In: Mazza-ferri EL, Kreisberg RA, Bar RS, eds. Advances in endocrinology and metabolism, 6th ed. St. Louis: Mosby-Year Book, Inc., 1995:39. 37. Ain KB. Papillary thyroid carcinoma etiology, assessment, and therapy. Endocrinol Metab Clin North Am 1995; 24:711. 38. Sobrinho-Simoes M, Soares J, Carneiro F. Diffuse follicular variant of papillary carcinoma of the thyroid: report of eight cases of a distinct aggressive type of tumor. Surg Path 1990; 3:189. 39. Nikiforov Y, Gnepp DR. Pediatric thyroid cancer after the Chernobyl disaster: pathomorphologic study of 84 cases (19911992) from the Republic of Belarus. Cancer 1994; 74:748. 40. Herrera MF, Hay ID, Wu PS, et al. Hrthle cell (oxyphilic) papillary thyroid carcinoma: a variant with more aggressive biologic behavior. World J Surg 1994; 16:669. 41. Katoh R, Harach HR, Williams ED. Solitary, multiple, and familial oxyphil tumours of the thyroid gland. J Pathol 1998; 186:292. 42. Rosai J. Papillary carcinoma. Monogr Pathol 1993; 138. 43. LiVolsi VA. Follicular lesions of the thyroid. In: LiVolsi VA, ed. Surgical pathology of the thyroid. Philadelphia: WB Saunders, 1990:173. 44. Hassoun AAK, Hay ID, Goellner JR, Zimmerman D. Insular thyroid carcinoma in adolescents: a potentially lethal endocrine malignancy. Cancer 1997; 79:1044. 45. Schneider AB, Shore-Freedman E, Ryo UY, et al. Radiation-induced tumors of the head and neck following childhood irradiation: prospective studies. Medicine (Baltimore) 1985; 64:1. 46. Mazzaferri EL. Management of a solitary thyroid nodule. N Engl J Med 1993; 328:553. 47. Schneider AB, Bekerman C, Leland J, et al. Thyroid nodules in the follow-up of irradiated individuals: comparison of thyroid ultrasound with scanning and palpation. J Clin Endocrinol Metab 1997; 82:4020. 48. Ezzat S, Sarti DA, Cain DR, Braunstein GD. Thyroid incidentalomas: prevalence by palpation and ultrasonography. Arch Intern Med 1994; 154:1838. 49. Tan GH, Gharib H. Thyroid incidentalomas: management approaches to nonpalpable nodules discovered incidentally on thyroid imaging. Ann Intern Med 1997; 126:226. 50. Fogelfeld L, Wiviott MBT, Shore-Freedman E, et al. Recurrence of thyroid nodules after surgical removal in patients irradiated in childhood for benign conditions. N Engl J Med 1989; 320:835. 51. Dottorini ME, Vignati A, Mazzucchelli L, et al. Differentiated thyroid carcinoma in children and adolescents: a 37-year experience in 85 patients. J Nucl Med 1997; 38:669. 52. Harach HR, Williams ED. Childhood thyroid cancer in England and Wales. Br J Cancer 1995; 72:777. 53. Hung W. Well-differentiated thyroid carcinomas in children and adolescents: a review. Endocrinologist 1994; 4:117. 54. Samuel AM, Rajashekharrao B, Shah DH. Pulmonary metastases in children and adolescents with well-differentiated thyroid cancer. J Nucl Med 1998; 39:1531. 55. Zimmerman D, Hay ID, Gough IR, et al. Papillary thyroid carcinoma in children and adults: long-term follow-up of 1039 patients conservatively treated at one institution during three decades. Surgery 1988; 104:1157. 56. Schlumberger M, De Vathaire F, Travagli JP, et al. Differentiated thyroid carcinoma in childhood: long term follow-up of 72 patients. J Clin Endocrinol Metab 1987; 65:1088. 57. Hay ID. Papillary thyroid carcinoma. Endocrinol Metab Clin North Am 1990; 19:545. 58. Pasieka JL, Thompson NW, McLeod MK, et al. The incidence of bilateral well-differentiated thyroid cancer found at completion thyroidectomy. World J Surg 1992; 16:711. 59. Shaha AR, Loree TR, Shah JP. Prognostic factors and risk group analysis in follicular carcinoma of the thyroid. Surgery 1995; 118:1131. 60. Sanders LE, Cady B. Differentiated thyroid cancer: reexamination of risk groups and outcome of treatment. Arch Surg 1998; 133:419. 61. Sarda AK, Bal S, Kapur MM. Near-total thyroidectomy for carcinoma of the thyroid. Br J Surg 1989; 76:90. 62. DeGroot LJ, Kaplan EL, McCormick M, Straus FH. Natural history, treatment, and course of papillary thyroid carcinoma. J Clin Endocrinol Metab 1990; 71:414. 63. Taylor T, Specker B, Robbins J, et al. Outcome after treatment of high-risk papillary and non-Hrthle-cell follicular thyroid carcinoma. Ann Intern Med 1998; 129:622. 64. Mazzaferri EL. Thyroid remnant 131 I ablation for papillary and follicular thyroid carcioma. Thyroid 1997; 7:265. 65. Massin JP, Savoie JC, Garnier H, et al. Pulmonary metastases in differentiated thyroid carcinoma: study of 58 cases with implications for the primary tumor treatment. Cancer 1984; 53:982. 66. Loh KC, Greenspan FS, Gee L, et al. Pathological tumor-node-metastasis (pTNM) staging for papillary and follicular thyroid carcinomas: a retrospective analysis of 700 patients. J Clin Endocrinol Metab 1997; 82:3553. 67. Cady B. Staging in thyroid carcinoma. Cancer 1998; 83:844. 68. Tsang TW, Brierley JD, Simpson WJ, et al. The effects of surgery, radioiodine, and external radiation therapy on the clinical outcome of patients with differentiated thyroid carcinoma. Cancer 1998; 82:375. 69. Simpson WJ, McKinney SE, Carruthers JS, et al. Papillary and follicular thyroid cancer: prognostic factors in 1578 patients. Am J Med 1987; 83:479. 70. Kurozumi K, Nakao I, Nishida T, et al. Significance of biologic aggressiveness and proliferating activity in papillary thyroid carcinoma. World J Surg 1998; 22:1237. 71. Sugitani I, Yanagisawa A, Shimizu A, et al. Clinicopathologic and immunohistochemical studies of papillary thyroid microcarcinoma presenting with cervical lymphadenopathy. World J Surg 1998; 22:731. 72. Akslen LA, Varhaug JE. Oncoproteins and tumor progression in papillary thyroid carcinoma: presence of epidermal growth factor receptor, c-erbB-2 protein, estrogen receptor related protein, p21-ras protein, and proliferation indicators in relation to tumor recurrences and patient survival. Cancer 1995; 76:1643. 73. Sellers M, Beenken S, Blankenship A, et al. Prognostic significance of cervical lymph node metastases in differentiated thyroid cancer. Am J Surg 1992; 164:578. 74. Sisson JC, Giordano TJ, Jamadar DA, et al. 131 I treatment of micronodular pulmonary metastases from papillary thyroid carcinoma. Cancer 1996; 78:2184. 75. Nmec J, Zamrazil V, Pohunkov D, Rhling S. Radioiodide treatment of pulmonary metastases of differentiated thyroid cancer: results and prognostic factors. Nuklearmedizin 1979; 18:86. 76. Casara D, Rubello D, Saladini G, et al. Different features of pulmonary metastases in differentiated thyroid cancer: natural history and multivariate statistical analysis of prognostic variables. J Nucl Med 1993; 34:1626. 77. Kashima K, Yokoyama S, Noguchi S, et al. Chronic thyroiditis as a favorable prognostic factor in papillary thyroid carcinoma. Thyroid 1998; 8:197. 78. Schffler A, Palitzsch KD, Seiffarth C, et al. Coexistent thyroiditis is associated with lower tumour stage in thyroid carcinoma. Eur J Clin Invest 1998; 28:838. 79. Viswanathan K, Gierlowski TC, Schneider AB. Childhood thyroid cancer: characteristics and long-term outcome in children irradiated for benign conditions of the head and neck. Am J Dis Child 1994; 148:260. 80. Filetti S, Belfiore A, Amir SM, et al. The role of thyroid-stimulating antibodies of Graves' disease in differentiated thyroid cancer. N Engl J Med 1988; 318:753. 81. Belfiore A, Garofalo MR, Giuffrida D, et al. Increased aggressiveness of thyroid cancer in patients with Graves' disease. J Clin Endocrinol Metab 1990; 70:830. 81a.Mazzaferri EL. Thyroid cancer and Graves' disease: the controversy ten years later. Endocrine Practice 2000; 6:221. 82. Fusco A, Grieco M, Santoro M, et al. A new oncogene in human thyroid papillary carcinomas and their lymph-nodal metastases. Nature 1987; 328:170. 83. Donghi R, Sozzi G, Pierotti MA, et al. The oncogene associated with human papillary thyroid carcinoma (PTC) is assigned to chromosome 10 q11-q12 in the same region as multiple endocrine neoplasia type 2A (MEN2A). Oncogene 1989; 4:521. 84. Smanik PA, Furminger TL, Mazzaferri EL, Jhiang SM. Breakpoint characterization of the ret/PTC oncogene in human papillary thyroid carcinoma. Hum Mol Genet 1995; 4:2313. 85. Nikiforov YE, Nikiforova M, Fagin JA. Radiation-induced post-Chernobyl pediatric thyroid carcinomas. Oncogene 1998; 17:1983. 86. Jhiang SM, Sagartz JE, Tong Q, et al. Targeted expression of the ret/PTC1 oncogene induces papillary thyroid carcinomas. Endocrinology 1996; 137:375. 87. Santoro M, Chiappetta G, Cerrato A, et al. Development of thyroid papillary carcinomas secondary to tissue-specific expression of the RET/PTC1 oncogene in transgenic mice. Oncogene 1996; 12:1821. 88. American Joint Committee on Cancer. Head and neck tumors: thyroid gland. In: Beahrs OH, Henson DE, Hutter RVP, Myers MH, eds. Manual for staging of cancer, 4th ed. Philadelphia: JB Lippincott, 1992:53. 89. Van De Velde CJH, Hamming JF, Goslings BM, et al. Report of the consensus development conference on the management of differentiated thyroid cancer in the Netherlands. Eur J Cancer Clin Oncol 1988; 24:287. 90. Baldet L, Manderscheid JC, Glinoer D, et al. The management of differentiated thyroid cancer in Europe in 1988: results of an international survey. Acta Endocrinol (Copenh) 1989; 120:547. 91. Solomon BL, Wartofsky L, Burman KD. Current trends in the management of well differentiated papillary thyroid carcinoma. J Clin 1996; 81:333.

92. Chen H, Zeiger MA, Clark DP, et al. Papillary carcinoma of the thyroid: can operative management be based solely on fine-needle aspiration? J Am Coll Surg 1997; 184:605. 93. Cady B. Our AMES is true: how an old concept still hits the mark: or, risk group assignment points the arrow to rational therapy selection in differentiated thyroid cancer. Am J Surg 1997; 174:462. 94. Schlumberger MJ. Medical progress: papillary and follicular thyroid carcinoma. N Engl J Med 1998; 338:297. 95. Mazzaferri EL. Treating differentiated thyroid carcinoma: where do we draw the line? Mayo Clin Proc 1991; 66:105. 95a.Mazzaferri EL. NCCN thyroid carcinoma practice guidelines. Oncology 1999; 13:391. 96. Newman KD, Black T, Heller G, et al. Differentiated thyroid cancer: determinants of disease progression in patients <21 years of age at diagnosis: a report from the Surgical Discipline Committee of the Children's Cancer Group. Ann Surg 1998; 227:533. 97. Miccoli P, Antonelli A, Spinelli C, et al. Completion total thyroidectomy in children with thyroid cancer secondary to the Chernobyl accident. Arch Surg 1998; 133:89. 98. Hay ID, Grant CS, Bergstralh EJ, et al. Unilateral total lobectomy: is it sufficient surgical treatment for patients with AMES low-risk papillary thyroid carcinoma? Surgery 1998; 124:958. 99. Pasieka JL, Rotstein LE. Consensus conference on well-differentiated thyroid cancer: a summary. Can J Surg 1999; 36:298. 100. Tourniaire J, Bernard MH, Bizollon-Roblin MH, et al. Thyroid papillary microcarcinoma: 179 cases since 1973. Presse Med 1998; 27:1467. 101. Maxon HR, Englaro EE, Thomas SR, et al. Radioiodine-131 therapy for well-differentiated thyroid cancer: a quantitative radiation dosimetric approach: outcome and validation in 85 patients. J Nucl Med 1992; 33:1132. 102. Chao TC, Jeng LB, Lin JD, Chen MF. Completion thyroidectomy for differentiated thyroid carcinoma. Otolaryngol Head Neck Surg 1998; 118:896. 103. DeGroot LJ, Kaplan EL. Second operations for completion of thyroidectomy in treatment of differentiated thyroid cancer. Surgery 1991; 110:936. 104. Scheumann GFW, Seeliger H, Musholt TJ, et al. Completion thyroidectomy in 131 patients with differentiated thyroid carcinoma. Eur J Surg 1996; 162:677. 105. Travagli JP, Cailleux AF, Ricard M, et al. Combination of radioiodine ( 131 I) and probe-guided surgery for persistent or recurrent thyroid carcinoma. J Clin Endocrinol Metab 1998; 83:2675. 106. Burge MR, Zeise TM, Johnsen MW, et al. Risks of complication following thyroidectomy. J Gen Intern Med 1998; 13:24. 107. Dralle H, Gimm O, Simon D, et al. Prophylactic thyroidectomy in 75 children and adolescents with hereditary medullary thyroid carcinoma: German and Austrian experience. World J Surg 1998; 22:744. 108. Pattou F, Combemale F, Fabre S, et al. Hypocalcemia following thyroid surgery: incidence and prediction of outcome. World J Surg 1998; 22:718. 109. Udelsman R, Lakatos E, Ladenson P. Optimal surgery for papillary thyroid carcinoma. World J Surg 1996; 20:88. 110. Sosa JA, Bowman HM, Tielsch JM, et al. The importance of surgeon experience for clinical and economic outcomes from thyroidectomy. Ann Surg 1998; 228:320. 111. Driggers RW, Kopelman JN, Satin AJ. Delaying surgery for thyroid cancer in pregnancy: a case report. J Reprod Med 1998; 43:909. 112. Moosa M, Mazzaferri EL. Outcome of differentiated thyroid cancer diagnosed in pregnant women. J Clin Endocrinol Metab 1997; 82:2862. 113. Saito T, Endo T, Kawaguchi A, et al. Increased expression of the sodium/iodide symporter in papillary thyroid carcinomas. J Clin Invest 1998; 101:1296. 114. Arturi F, Russo D, Schlumberger M, et al. Iodide symporter gene expression in human thyroid tumors. J Clin Endocrinol Metab 1998; 83:2493. 115. Jhiang SM, Cho JY, Ryu K-Y, et al. An immunohistochemical study of Na+/I symporter in human thyroid tissues and salivary gland tissues. Endocrinology 1998; 139:4416. 115a.Rotman-Pikielny P, Reynolds JC, Barker WC, et al. Recombinant human thyrotropin for the diagnosis and treatment of a highly functional meta-static struma ovarii. J Clin Endocrinol Metab 2000; 85:237. 115b.Rudavsky AZ, Freeman LM. Treatment of scan-negative, thyroglobulin-positive metastatic thyroid cancer using radioiodine Endocrinol Metab 1997; 82:11.
131

I and recombinant human thyroid stimulating hormone. J Clin

116. Goldman JM, Line BR, Aamodt RL, Robbins J. Influence of triiodothyronine withdrawal time on 131-I uptake postthyroidectomy for thyroid cancer. J Clin Endocrinol Metab 1980; 50:734. 117. Lakshmanan M, Schaffer A, Robbins J, et al. A simplified low iodine diet in I-131 scanning and therapy of thyroid cancer. Clin Nucl Med 1988; 2:866. 118. Pons F, Carrio I, Estorch M, et al. Lithium as an adjuvant of iodine-131 uptake when treating patients with well-differentiated thyroid carcinoma. Clin Nucl Med 1987; 8:644. 118a.Koong SS, Reynolds JC, Movius EG, Keenan AM, et al. Lithium as a potential adjuvant to 131 I therapy of metastatic, well-differentiated thyroid carcinoma. J Clin Endocrinol Metab 1999; 84:912. 119. Maxon HR, Thomas SR, Hertzberg VS, et al. Relation between effective radiation dose and outcome of radioiodine therapy for thyroid cancer. N Engl J Med 1983; 309:937. 120. Park HM, Perkins OW, Edmondson JW, et al. Influence of diagnostic radioiodines on the uptake of ablative dose of iodine. Thyroid 1994; 4:49. 121. Leger FA, Izembart M, Dagousset F, et al. Decreased uptake of therapeutic doses of iodine-131 after 185-MBq iodine-131 diagnostic imaging for thyroid remnants in differentiated thyroid carcinoma. Eur J Nucl Med 1998; 25:242. 122. Muratet JP, Giraud P, Daver A, et al. Predicting the efficacy of first iodine-131 treatment in differentiated thyroid carcinoma. J Nucl Med 1997; 38:1362. 123. Vassilopoulou-Sellin R, Klein MJ, Smith TH, et al. Pulmonary metastases in children and young adults with differentiated thyroid cancer. Cancer 1993; 71:1348. 124. Spencer CA, Takeuchi M, Kazarosyan M, et al. Serum thyroglobulin autoantibodies: prevalence, influence on serum thyroglobulin measurement, and prognostic significance in patients with differentiated thyroid carcinoma. J Clin Endocrinol Metab 1998; 83:1121. 125. Hodgson DC, Brierley JD, Tsang RW, Panzarella T. Prescribing 131 iodine based on neck uptake produces effective thyroid ablation and reduced hospital stay. Radiother Oncol 1998; 47:325. 126. Cunningham MP, Duda RB, Recant W, et al. Survival discriminants for differentiated thyroid cancer. Am J Surg 1990; 160:344. 127. DeGroot LJ, Kaplan EL, Straus FH, Shukla MS. Does the method of management of papillary thyroid carcinoma make a difference in outcome? World J Surg 1994; 18:123. 128. Samaan NA, Maheshwari YK, Nader S, et al. Impact of therapy for differentiated carcinoma of the thyroid: an analysis of 706 cases. J Clin Endocrinol Metab 1983; 56:1131. 129. Bal C, Padhy AK, Jana S, et al. Prospective randomized clinical trial to evaluate the optimal dose of 131 I for remnant ablation in patients with differentiated thyroid carcinoma. Cancer 1996; 77:2574. 130. Brierley J, Maxon HR. Radioiodine and external radiation therapy. In: Fagin JA, ed. Thyroid cancer, 1st ed. Boston/Dordrecht/London: Kluwer Academic Publishers, 1998:285. 131. Simpson WJ, Panzarella T, Carruthers JS, et al. Papillary and follicular thyroid cancer: impact of treatment in 1578 patients. Int J Radiat Oncol Biol Phys 1988; 14:1063. 132. Samaan NA, Schultz PN, Haynie TP, Ordonez NG. Pulmonary metastasis of differentiated thyroid carcinoma: treatment results in 101 patients. J Clin Endocrinol Metab 1985; 60:376. 133. Ruegemer JJ, Hay ID, Bergstralh EJ, et al. Distant metastases in differentiated thyroid carcinoma: a multivariate analysis of prognostic variables. J Clin Endocrinol Metab 1988; 67:501. 134. Schlumberger M, Tubiana M, De Vathaire F, et al. Long-term results of treatment of 283 patients with lung and bone metastases from differentiated thyroid carcinoma. J Clin Endocrinol Metab 1986; 63:960. 135. Mazzaferri EL. Carcinoma of follicular epithelium: radioiodine and other treatment outcomes. In: Braverman LE, Utiger RD, eds. The thyroid: a fundamental and clinical text, 7th ed. Philadelphia: Lippincott-Raven, 1996:922. 136. Benua RS, Cicale NR, Sonenberg M, et al. The relation of radioiodine dosimetry to results and complications in the treatment of metastatic thyroid cancer. AJR Am J Roentgenol 1962; 87:171. 137. Maxon H III, Smith HS. Radioiodine-131 in the diagnosis and treatment of metastatic well differentiated thyroid cancer. Endocrinol Metab Clin North Am 1990; 19:685. 138. Datz FL. Cerebral edema following iodine-131 therapy for thyroid carcinoma metastatic to the brain. J Nucl Med 1986; 27:637. 139. DiRusso G, Kern KA. Comparative analysis of complications from I-131 radioablation for well-differentiated thyroid cancer. Surgery 1994; 116:1024. 140. Allweiss P, Braunstein GD, Katz A, Waxman A. Sialadenitis following I-131 therapy for thyroid carcinoma: concise communication. J Nucl Med 1984; 25:755. 141. Alexander C, Bader JB, Schaefer A, et al. Intermediate and long-term side effects of high-dose radioiodine therapy for thyroid carcinoma. J Nucl Med 1998; 39:1551. 142. Brown AP, Greening WP, McCready VR, et al. Radioiodine treatment of metastatic thyroid carcinoma: the Royal Marsden Hospital experience. Br J Radiol 1984; 57:323. 143. Hall P, Holm L-E. Cancer in iodine-131 exposed patients. J Endocrinol Invest 1995; 18:147. 144. Van Nostrand D, Neutze J, Atkins F. Side effects of rational dose iodine-131 therapy for metastatic well-differentiated thyroid carcinoma. J Nucl Med 1986; 27:1519. 145. De Vathaire F, Schlumberger M, Delisle MJ, et al. Leukaemias and cancers following iodine-131 administration for thyroid cancer. Br J Cancer 1997; 75:734. 146. Schlumberger M, De Vathaire F, Ceccarelli C, et al. Outcome of pregnancy in women with thyroid carcinoma. J Endocrinol Invest 1995; 18:150. 147. Schlumberger M, De Vathaire F, Ceccarelli C, et al. Exposure to radioactive iodine-131 for scintigraphy or therapy does not preclude pregnancy in thyroid cancer patients. J Nucl Med 1996; 37:606. 148. Pacini F, Gasperi M, Fugazzola L, et al. Testicular function in patients with differentiated thyroid carcinoma treated with radioiodine. J Nucl Med 1994; 35:1418. 149. Burmeister LA, Goumaz MO, Mariash CN, Oppenheimer JH. Levothyroxine dose requirements for thyrotropin suppression in the treatment of differentiated thyroid cancer. J Clin Endocrinol Metab 1992; 75:344. 150. Bartalena L, Martino E, Pacchiarotti A, et al. Factors affecting suppression of endogenous thyrotropin secretion by thyroxine treatment: retrospective analysis in athyreotic and goitrous patients. J Clin Endocrinol Metab 1987; 64:849. 151. Pujol P, Daures JP, Nsakala N, et al. Degree of thyrotropin suppression as a prognostic determinant in differentiated thyroid cancer. J Clin Endocrinol Metab 1996; 81:4318. 152. Cooper DS, Specker B, Ho M, et al. Thyrotropin suppression and disease progression in patients with differentiated thyroid cancer: results from the National Thyroid Cancer Treatment Cooperative Registry. Thyroid 1999; 8:737. 153. Biondi B, Fazio S, Carella C, et al. Cardiac effects of long term thyrotropin-suppressive therapy with levothyroxine. J Clin Endocrinol Metab 1993; 77:334. 154. Sawin CT, Geller A, Wolf PA, et al. Low serum thyrotropin concentrations as a risk factor for atrial fibrillation in older persons. N Engl J Med 1994; 331:1249. 155. Fazio S, Biondi B, Carella C, et al. Diastolic dysfunction in patients on thyroid-stimulating hormone suppressive therapy with levothyroxine: beneficial effect of b-blockade. J Clin Endocrinol Metab 1995; 80:2222. 156. Shapiro LE, Sievert R, Ong L, et al. Minimal cardiac effects in asymptomatic athyreotic patients chronically treated with thyrotropin-suppressive doses of L-thyroxine. J Clin Endocrinol Metab 1997; 82:2592. 157. Toivonen J, Tahtela R, Laitinen K, et al. Markers of bone turnover in patients with differentiated thyroid cancer with and following withdrawal of thyroxine suppressive therapy. Eur J Endocrinol 1998; 138:667. 158. Marcocci C, Golia F, Vignali E, Pinchera A. Skeletal integrity in men chronically treated with suppressive doses of L-thyroxine. J Bone Miner Res 1997; 12:72. 159. Grnwald F, Menzel C, Bender H, et al. Redifferentiation therapy-induced radioiodine uptake in thyroid cancer. J Nucl Med 1998; 39:1903. 160. Meier CA, Braverman LE, Ebner SA, et al. Diagnostic use of recombinant human thyrotropin in patients with thyroid carcinoma (phase I/II study). J Clin Endocrinol Metab 1994; 78:188. 161. Ladenson PW, Braverman LE, Mazzaferri EL, et al. Comparison of administration of recombinant human thyrotropin with withdrawal of thyroid hormone for radioactive iodine scanning in patients with thyroid carcinoma. N Engl J Med 1997; 337:888. 162. Haugen B, Pacini F, Reiners C, et al. A comparison of recombinant human thyrotropin and thyroid hormone withdrawal for the detection of thyroid remnant or cancer. J Clin Endocrinol Metab 1999; 84:3877. 163. Rudavsky AZ, Freeman LM. Treatment of scan-negative, thyroglobulin-positive metastatic thyroid cancer using radioiodine 131 I and recombinant human thyroid stimulating hormone. J Clin Endocrinol Metab 1997; 82:11. 164. Spencer CA, Wang CC. Thyroglobulin measurement: techniques, clinical benefits, and pitfalls. Endocrinol Metab Clin North Am 1995; 24:841. 165. Spencer CA. Recoveries cannot be used to authenticate thyroglobulin (Tg) measurements when sera contain Tg autoantibodies. Clin Chem 1996; 42:661. 166. Ringel M, Ladenson P, Levine MA. Molecular diagnosis of residual and recurrent thyroid cancer by amplification of thyroglobulin messenger ribonucleic acid in peripheral blood. J Clin Endocrinol Metab 1998; 83:4435. 167. Nakada K, Katoh C, Kanegae K, et al. Thallium-201 scintigraphy to predict therapeutic outcome of iodine-131 therapy of metastatic thyroid carcinoma. J Nucl Med 1998; 39:807. 168. Carril JM, Quirce R, Serrano J, et al. Total-body scintigraphy with thallium-201 and iodine-131 in the follow-up of differentiated thyroid cancer. J Nucl Med 1997; 38:686. 169. nal S, Menda Y, Adalet I, et al. Thallium-201, technetium-99m-tetrofosmin and iodine-131 in detecting differentiated thyroid carcinoma metastases. J Nucl Med 1998; 39:1897. 170. Lorberboym M, Murthy S, Mechanick JI, et al. Thallium-201 and iodine-131 scintigraphy in differentiated thyroid carcinoma. J Nucl Med 1996; 37:1487. 171. Lin JD, Kao PF, Weng HF, et al. Relative value of thallium-201 and iodine-131 scans in the detection of recurrence or distant metastasis of well differentiated thyroid carcinoma. Eur J Nucl Med 1998; 25:695. 172. Miyamoto S, Kasagi K, Misaki T, et al. Evaluation of technetium-99m-MIBI scintigraphy in metastatic differentiated thyroid carcinoma. J Nucl Med 1997; 38:352. 173. Huang TS, Chieng PU, Chang CC, Yen RF. Positron emission tomography for detecting iodine-131 nonvisualized metastasis of well-differentiated thyroid carcinoma: two case reports. J Endocrinol Invest 1998; 21:392.

174. Bakheet SMB, Powe J. Fluorine-18-fluorodeoxyglucose uptake in rheumatoid arthritis-associated lung disease in a patient with thyroid cancer. J Nucl Med 1998; 39:234. 174a.Wang W, Larson SM, Fazzari M, et al. Prognostic value of [18F]fluorodeox-yglucose positron emission tomographic scanning in patients with thyroid cancer. J Clin Endocrinol Metab 2000; 85:1107. 175. 176. 177. 178. 179. 180. 181. 182. 183. 184. 185. 186. 187. 188. 189. 190. 191. 192. 193. 194. 195. 196. Greenler DP, Klein HA. The scope of false-positive iodine-131 images for thyroid carcinoma. Clin Nucl Med 1989; 14:111. Bakheet SMB, Hammami MM, Hemidan A, et al. Radioiodine secretion in tears. J Nucl Med 1998; 39:1452. Chung JK, Lee YJ, Jeong JM, et al. Clinical significance of hepatic visualization on iodine-131 whole-body scan in patients with thyroid carcinoma. J Nucl Med 1997; 38:1191. Mazzaferri EL. Treating high thyroglobulins with radioiodine: a magic bullet or a shot in the dark? J Clin Endocrinol Metab 1995; 80:1485. Schlumberger M, Challeton C, De Vathaire F, et al. Radioactive iodine treatment and external radiotherapy for lung and bone metastases from thyroid carcinoma. J Nucl Med 1996; 37:598. Young RL, Mazzaferri EL, Rahe AJ, Dorfman SG. Pure follicular thyroid carcinoma: impact of therapy in 214 patients. J Nucl Med 1980; 21:733. Lin JD, Chao TC, Ho J, et al. Poor prognosis of 56 follicular thyroid carcinomas with distant metastases at the time of diagnosis. Cancer J 1998; 11:190. Gilliland FD, Hunt WC, Morris DM, Key CR. Prognostic factors for thyroid carcinoma: a population-based study of 15,698 cases from the surveillance, epidemiology and end results (SEER) program 1973. Cancer 1997; 79:564. Samaan NA, Schultz PN, Hickey RC, et al. Well-differentiated thyroid carcinoma and the results of various modalities of treatment. A retrospective review of 1599 patients. J Clin Endocrinol Metab 1992; 75:714. Brennan MD, Bergstralh EJ, van Heerden JA, McConahey WM. Follicular thyroid cancer treated at the Mayo Clinic, 1946 through 1970: initial manifestations, pathologic findings, therapy, and outcome. Mayo Clin Proc 1991; 66:11. Tennvall J, Biorklund A, Moller T, et al. Prognostic factors of papillary, follicular and medullary carcinomas of the thyroid gland: retrospective multivariate analysis of 216 patients with a median follow-up of 11 years. Acta Radiologica Oncol 1985; 24:17. Mulligan LM, Kwok JBJ, Healey CS, et al. Germ-line mutations of the RET protooncogene in multiple endocrine neoplasia type 2A. Nature 1993; 363:458. Gagel RF, Robinson MF, Donovan DT, Alford BR. Medullary thyroid carcinoma: recent progress. J Clin Endocrinol Metab 1993; 76:809. Howe JR, Norton JA, Wells SA Jr. Prevalence of pheochromocytoma and hyperparathyroidism in multiple endocrine neoplasia type 2A: results of long-term follow-up. Surgery 1993; 114:1070. Raue F, Zink A. Clinical features of multiple endocrine neoplasia type 1 and type 2. Horm Res 1992; 38 (Suppl2):31. Sizemore GW. Medullary carcinoma of the thyroid gland. Semin Oncol 1987; 14:306. Donis Keller H, Dou S, Chi D, et al. Mutations in the RET protooncogene are associated with MEN 2A and FMTC. Hum Mol Genet 1993; 2:851. Eng C, Clayton D, Schufenecker I, et al. The relationship between specific RET proto-oncogene mutations and disease phenotype in multiple endocrine neoplasia type 2: international RET mutation consortium analysis. JAMA 1996; 276:1575. Mendelsohn G. Markers as prognostic indicators in medullary thyroid carcinoma. Am J Clin Pathol 1991; 95:297. Cohen R, Buchsenschutz B, Estrade P, et al. Causes of death in patients suffering from medullary thyroid carcinoma: report of 119 cases. Presse Med 1996; 25:1819. Modigliani E, Cohen R, Campos JM, et al. Prognostic factors for survival and for biochemical cure in medullary thyroid carcinoma: results in 899 patients. Clin Endocrinol (Oxf) 1998; 48:265. Dottorini ME, Assi A, Sironi M, et al. Multivariate analysis of patients with medullary thyroid carcinoma: prognostic significance and impact on treatment of clinical and pathologic variables. Cancer 1996; 77:1556.

196a.Kebebew E, Ituarte PH, Siperstein AE, et al. Medullary thyroid carcinoma: clinical characteristics, treatment, prognostic factors, and a comparison of staging systems. Cancer 2000; 88:1139. 197. Wells SA, Baylin SB, Leight G, et al. The importance of early diagnosis in patients with hereditary medullary thyroid carcinoma. Ann Surg 1982; 195:204. 198. Wells SA Jr, Skinner MA. Prophylactic thyroidectomy, based on direct genetic testing, in patients at risk for the multiple endocrine neoplasia type 2 syndromes. Exp Clin Endocrinol Diabetes 1998; 106:29. 199. Decker RA, Peacock ML. Update on the profile of multiple endocrine neoplasia type 2a RET mutations: practical issues and implications for genetic testing. Cancer 1997; 80:557. 200. Moley JF, DeBenedetti MK, Dilley WG, et al. Surgical management of patients with persistent or recurrent medullary thyroid cancer. J Intern Med 1998; 243:521. 201. Chen HB, Roberts JR, Ball DW, et al. Effective long-term palliation of symptomatic, incurable metastatic medullary thyroid cancer by operative resection. Ann Surg 1998; 227:887. 202. Frank-Raue K, Ziegler R, Raue F. The use of octreotide in the treatment of medullary thyroid carcinoma. Horm Metab Res 1993; 27(Suppl):44. 202a.Juweid ME, Hajjar G, Stein R, et al. Initial experience with high-dose radio-immunotherapy of metastatic medullary thyroid cancer using AHSCR [see comments]. J Nucl Med 2000; 41:93.
131

I-MN-14 F(ab)2 anti-carcinoembryonic antigen MAb and

203. Kurtaran A, Scheuba C, Kaserer K, et al. Indium-111-DTPA-D-Phe-1-octreotide and technetium-99m-(V)-dimercaptosuccinic acid scanning in the preoperative staging of medullary thyroid carcinoma. J Nucl Med 1998; 39:1907. 204. Skinner MA, DeBenedetti MK, Moley JF, et al. Medullary thyroid carcinoma in children with multiple endocrine neoplasia types 2A and 2B. J Pediatr Surg 1996; 31:177. 205. Ledger GA, Khosla S, Lindor NM, et al. Genetic testing in the diagnosis and management of multiple endocrine neoplasia type II. Ann Intern Med 1995; 122:118. 206. O'Riordain DS, O'Brien T, Weaver AL, et al. Medullary thyroid carcinoma in multiple endocrine neoplasia types 2A and 2B. Surgery 1994; 116:1017. 207. Coustan DR, Carpenter MW. The diagnosis of gestational diabetes. Diabetes Care 1998; 21:B5. 208. Decker RA, Geiger JD, Cox CE, et al. Prophylactic surgery for multiple endocrine neoplasia type IIa after genetic diagnosis: is parathyroid transplantation indicated? World J Surg 1996; 20:814. 209. Kraimps JL, Denizot A, Carnaille B, et al. Primary hyperparathyroidism in multiple endocrine neoplasia type IIa: retrospective French multicentric study. World J Surg 1996; 20:808.

CHAPTER 41 UNUSUAL THYROID CANCERS Principles and Practice of Endocrinology and Metabolism

CHAPTER 41 UNUSUAL THYROID CANCERS

MATTHEW D. RINGEL Anaplastic Thyroid Carcinoma Demographics and Pathogenesis Clinical Presentation and Diagnosis Treatment Control of Local Cervical Tumor Treatment of Metastatic Disease Prognosis Summary Poorly Differentiated (Insular) Carcinoma Demographics and Pathogenesis Clinical Presentation and Diagnosis Treatment and Prognosis Summary Tall Cell Variant of Papillary Carcinoma Demographics and Pathogenesis Clinical Presentation and Diagnosis Treatment and Prognosis Summary Columnar Cell Variant of Papillary Carcinoma Clinical Presentation and Diagnosis Treatment and Prognosis Summary Diffuse Sclerosing Papillary Thyroid Cancer Squamous Cell Carcinoma of the Thyroid Demographics and Pathogenesis Clinical Presentation and Diagnosis Treatment and Prognosis Mucoepidermoid Carcinoma Mixed Medullary-Follicular and Medullary-Papillary Carcinoma Thyroid Sarcomas Thyroid Teratoma Thyroid Lymphoma Demographics and Pathogenesis Clinical Presentation and Diagnosis Treatment and Prognosis Summary Metastatic Cancers in the Thyroid Clinical Presentation and Diagnosis Treatment and Prognosis Conclusion Chapter References

Approximately 10% of thyroid cancers are either poorly differentiated epithelial tumors or nonepithelial malignancies that are included in the World Health Organization (WHO) classification of thyroid tumors.1 The variants of epithelial tumors, including anaplastic, squamous cell, tall cell, insular, and other poorly differentiated types of thyroid cancer, are characterized by more aggressive courses than are well-differentiated histologic subtypes. In addition, nonepithelial malignancies (e.g., lymphomas, sarcomas, and others) and tumors that metastasize to the thyroid occur and can now be recognized by fine-needle aspiration (FNA). The clinician must recognize these uncommon tumors because they are treated differently from the typical epithelial thyroid cancers. The clinical presentation, treatment, and prognosis of these unusual thyroid malignancies are discussed in this chapter. More common thyroid malignancies are discussed in Chapter 40.

ANAPLASTIC THYROID CARCINOMA

Anaplastic cancer is a rare, aggressive malignancy with mean survival rates of <1 year. These tumors are comprised of large, pleomorphic cells with mitoses and bizarre nuclei. The cytologic pattern can range from large cells to spindle cells that conform to a variety of different histologic patterns.2,3 All of these tumors share similar clinical characteristics; therefore, clinical presentation, management, and prognosis are considered together. In the past, small cell carcinomas of the thyroid were also considered a variant of anaplastic carcinoma; however, these have been reclassified as thyroid lymphomas, typically respond to lymphoma therapy, and are considered in a separate section. DEMOGRAPHICS AND PATHOGENESIS The annual incidence of anaplastic carcinoma is estimated to be two per million in the United States.4,5 Anaplastic carcinomas are more frequently identified in regions of iodine deficiency,6 and the incidence decreases with introduction of iodine prophylaxis. In all populations studied, the incidence rises with increasing age. Demographically, anaplastic thyroid cancer shows a predominance among females, with a median age of onset in the sixth and seventh decades.2,3 and 4,7 The pathogenesis of anaplastic thyroid cancer is not entirely clear. A stepwise multi-hit hypothesis for thyroid tumor dedifferentiation has been proposed,8 based on the high frequency of concurrent well-differentiated and anaplastic carcinoma, and epidemiologic data showing that as many as 30% of patients with anaplastic cancer have a history of well-differentiated carcinoma.2,3 and 4,7 In support of this model, poorly differentiated thyroid tumors have a higher incidence of mutations of the genes encoding p53, other tumor suppressors, and oncogenes associated with aggressive cancers. Thus, the general belief is that anaplastic thyroid cancers rarely occur de novo. Other factors proposed to be involved in the development of anaplastic thyroid cancer include a history of neck/mantle external radiation and prior iodine-131 (131I) therapy. Although it is associated with development of nodules and hypothyroidism, no convincing data other than several case reports exist linking external irradiation to anaplastic thyroid cancer. Similarly, 131I therapy does not appear to cause anaplastic thyroid cancer.9 CLINICAL PRESENTATION AND DIAGNOSIS Patients with anaplastic cancer classically tend to be older, with a prior history of multinodular goiter that manifests as a rapid enlargement of a nodule or appearance of a lymph node and is characterized by severe pain, dysphagia or odynophagia, and airway compression. The rapid enlargement is often due to hemorrhage into a preexistent tumor.2,3 and 4,7 Superior vena cava syndrome from extensive cervical disease and/or symptoms from distant metastases are sometimes present at initial evaluation. Dense fibrosis and calcifications are also well described.10 Anaplastic carcinoma can also be found in more typical nodules that are not rapidly enlarging and within large compressive multinodular goiters that do not have the rapid enlargement noted above. In addition, paraneoplastic syndromes also occur with anaplastic thyroid cancer and primary squamous cell thyroid cancer (discussed later). Patients with humoral hypercalcemia of malignancy due to production of parathyroid hormonerelated protein (PTHrP) and leukocytosis caused by production of granulocyte colony stimulating factor (G-CSF) have been described. Moreover, cell lines derived from these tumors have been shown to produce these two factors.11,12 Evaluation of thyroid masses typically includes initial FNA. The accuracy of FNA in diagnosing anaplastic thyroid cancer is reported to range from 89% to 100%.2,3 Despite these studies, and the use of flow cytometry to help exclude lymphoma, many centers perform open biopsy to confirm the diagnosis of anaplastic carcinoma, because the distinction from lymphoma is critical to management and prognosis. Frequently, the clinician is confronted with a patient with a large tumor mostly comprised of well-differentiated cancer that contains a small focus of anaplastic cancer. The prognosis of the patient seems to be related to the presence of any focus of anaplastic carcinoma; thus, in the thyroid node metastasis (TNM) and other staging

systems, the presence of anaplastic carcinoma defines a stage 4 lesion, regardless of patient age, tumor size, or lymph node spread. TREATMENT Therapy for anaplastic carcinoma differs from that for other epithelial thyroid malignancies due to the absence of thyroid-stimulating hormone (TSH) receptor expression and iodine uptake. Therefore, radioiodine and TSH suppression are usually ineffective. TSH should be maintained in the lower part of the normal range but should not be suppressed, because nutritional status and weight are critical in chemotherapy. In addition, one of the frequently used chemotherapy medications, doxorubicin hydrochloride (Adriamycin), has cardiotoxic side effects, further limiting a role for TSH-suppression therapy. Thus, treatment paradigms are more similar to those for non-thyroidal poorly differentiated cancers. CONTROL OF LOCAL CERVICAL TUMOR Preservation of vital neck structures to prevent airway and esophageal disasters is the initial focus of anaplastic thyroid cancer management. Preoperative panendoscopy is often performed, as in other head/neck cancers. Initial surgery should be performed with as complete a resection as possible. This may include neck dissection. In cases of smaller, intrathyroidal tumors, or ones with lymph node metastases without invasion into local structures, the surgeon may be able to avoid tracheostomy at this time. In cases in which tumor is invading local structures and gross disease remains, placing a tracheostomy at the time of primary surgery is reasonable. In addition, if periesophageal disease or loss of swallowing reflexes is identified, a percutaneous gastric feeding tube may be placed for either immediate or later use. After surgery, external beam radiation is frequently performed. As with other head and neck cancers, treatment with 5000 to 6000 rads is typical and is associated with significant side effects including esophagitis, malnutrition, and stomatitis. Most patients with anaplastic carcinoma receive postoperative radiation therapy to control gross or microscopic residual disease. Several studies have evaluated the usefulness of radiation-sensitizing chemotherapy agents to reduce the dose of radiation needed. In some of these trials, patients with inoperable disease were treated with combined radiation and chemotherapy before surgery and treatment then continued with the combination therapy after thyroidectomy. 2,3,4,7,13,14,15 and 16 Although local control was frequently obtained, overall survival was not improved. Long-term survival was seen only in patents with no measurable disease after surgery. In general, aggressive surgical resection of the primary tumor is performed, usually without preoperative radiation therapy, although other approaches may be appropriate in selected cases. TREATMENT OF METASTATIC DISEASE Treatment of metastatic anaplastic thyroid cancer has been largely ineffective despite the use of a wide variety of chemotherapy regimens.8,13,15,16 and 17 Partial response rates as high as 20% to 30% have been described, but median and mean survival times still range between 4 and 18 months, consistent with the natural history of the disease.18 The most active agents appear to be doxorubicin (alone or in combination with cisplatin) and perhaps paclitaxel (Taxol) and other taxanes, although the results of phase 1 and 2 clinical trials are not yet complete. Combination chemotherapy with bleomycin, cyclophosphamide, and 5-fluorouracil has also been used, but with similar results and greater toxicity. Several newer agents are under evaluation in phase 1 trials, including antiangiogenic factors and immunomodulators, but results have not yet been reported. 2,19 Autologous or allogeneic transplantation has not been used, probably because of the older age of most of the patients and the poor response to chemotherapy. Because of the absence of data supporting an improvement in length of survival, a decision to use chemotherapy must be made jointly with the patient, with a clear understanding of the quality-of-life issues. After initial surgery and radiation (with or without chemotherapy), patients with measurable metastatic disease who also have a reasonable performance status are appropriate candidates for chemotherapy. PROGNOSIS The overall prognosis for patients with anaplastic thyroid cancer remains dismal and has been largely unaffected by chemotherapy or radiation therapy. Mean and median survival is still measured in months despite aggressive therapy. Most patients can be protected from death due to local airway compression and hemorrhage with aggressive surgery, radiation therapy, and chemotherapy. In two cohorts, more complete surgical resections provided a significant survival advantage for younger individuals presenting with smaller tumors confined to the thyroid.4,7 The 14% of patients in the second cohort 7 who were alive 24 months after diagnosis were younger, had smaller tumors, and were treated aggressively with both surgery and radiation therapy. SUMMARY Anaplastic thyroid cancer is one of the most virulent tumors known to humans. Aggressive therapy to control local disease, particularly in individuals with no measurable metastases at the time of diagnosis, is appropriate to attain control of local disease and to create an opportunity for long-term survival. The side effects and potential effects on quality of life need to be openly discussed with the patient and the family preoperatively when possible. Systemic chemotherapy for metastatic disease is more controversial but is often attempted in younger patients with doxorubicin- or taxane-based regimens. Further clinical and basic research is needed to better understand the pathogenesis of anaplastic thyroid cancer and to develop more effective therapies.

POORLY DIFFERENTIATED (INSULAR) CARCINOMA

Insular thyroid carcinoma was originally described as Wuchernde Struma20 and later became known as insular carcinoma because of the tendency of the polygonal cancer cells to grow in nests or insulae.21 The growth pattern of insular carcinoma is distinct from that of the usual forms of thyroid carcinoma. The cells lack the nuclear features of papillary cancer; thus, these tumors are generally considered variants of follicular carcinoma. However, small foci of more typical well-differentiated cancers, including papillary cancer, are occasionally identified within or adjacent to the poorly differentiated component. The histologic patterns of poorly differentiated follicular cancers can vary from solid, to insular, to trabecular. Frequently, all three histologic patterns are seen within the same tumor. The exact percentage of tumor in the poorly differentiated pattern required for diagnosis is not clear because similar prognoses have been identified for patients with a predominantly poorly differentiated pattern and for those with a predominantly well-differentiated pattern.22 DEMOGRAPHICS AND PATHOGENESIS Patients with poorly differentiated carcinomas tend to be older (mean of 54 years) than those with well-differentiated tumors, and their tumors tend to be larger.23 Insular carcinomas are uncommon, representing 1% to 2% of all thyroid cancers. The absence of a clear pathologic definition for this tumor (e.g., the amount of tumor required to be categorized as this pattern) make incidence studies difficult to interpret. Poorly differentiated carcinomas are thought to represent an intermediate step in the progression of well-differentiated cancer to anaplastic cancer. Indeed, mutations in genes associated with aggressive cancer are reported to be more common in these tumors than in well-differentiated tumors.10 Insular cancer is frequently diagnosed along with well-differentiated cancer in the same nodule; in addition, regions of poorly differentiated histology are often identified in the transition areas from well-differentiated to anaplastic cancer. Most insular carcinomas concentrate iodine and express thyroglobulin but have a prognosis that is intermediate between those of well-differentiated and anaplastic carcinoma.24,25 These factors further support the notion that they represent tumors of moderate dedifferentiation. CLINICAL PRESENTATION AND DIAGNOSIS As previously noted, patients with insular carcinomas tend to be older, have larger tumors, and are more likely to present with metastases as compared to patients with well-differentiated cancers. The male to female ratio is also preserved.23 Most commonly, these tumors present as enlarging masses that may be slow growing, often in the setting of a multinodular goiter. Occasionally they may rapidly enlarge from hemorrhage or anaplastic transformation. FNA results often are suspicious for poorly differentiated carcinoma on the basis of the enlarged, polygonal cells.26 Children with insular carcinoma tend to present with early nodal or distant metastases. TREATMENT AND PROGNOSIS Treatment of poorly differentiated thyroid cancers is similar to that for well-differentiated tumors because they usually concentrate radioiodine. Near-total thyroidectomy should be performed in all patients with these tumors to maximize the effectiveness of postoperative therapy and monitoring. Treatment with radioiodine and thyroxine suppression should be performed, with a goal of undetectable circulating levels of TSH and the complete ablation of iodine-avid tissue on diagnostic scan. This approach may require multiple treatments with radioiodine at 10- to 12-month intervals. Magnetic resonance imaging (MRI), computed tomography (CT), and ultrasonography may also be helpful in defining metastases, particularly in the neck, that may be amenable to surgical excision. In occasional cases with extensive

direct local invasion, external beam radiation may also be useful to control neck disease locally. No proven chemotherapy regimens exist for these tumors if they progress despite radioiodine therapy. Successful redifferentiation therapy with retinoids has been reported, but only in a few cases.27 Thus, the use of retinoids is considered experimental. The prognosis for patients with poorly differentiated thyroid cancer appears to relate mostly to the stage of disease at the time of diagnosis. In one review, 26% of patients with insular cancer presented with distant metastases, and 36% had cervical metastases.23 During the course of follow-up (means of 3.5 to 7 years), 25% of patients died of poorly differentiated cancer and an additional 25% had residual or recurrent diseasefigures much higher than those typically reported for well-differentiated cancer. Most of these patients had metastases at the time of diagnosis and were treated with the aggressive paradigm previously suggested. SUMMARY Insular and other poorly differentiated thyroid cancers appear to be more aggressive tumors than well-differentiated thyroid malignancies; they present with later-stage disease and larger tumors, and have a worse prognosis. The majority are iodine avid; therefore, aggressive surgical therapy followed by radio-iodine therapy and TSH suppression is recommended. Careful monitoring with measurement of serum levels of thyroglobulin, yearly radioiodine scanning, and neck and chest imaging are appropriate.

TALL CELL VARIANT OF PAPILLARY CARCINOMA

The tall cell variant of papillary cancer (TCV) was first identified in a subgroup of papillary cancers of which at least 30% were comprised of large papillary structureswith cells characterized by a height at least twice the width, an oxyphilic cytoplasm, and hyperchromic basal nuclei.28 These characteristic patterns were present in nearly 10% of all papillary cancers. As with poorly differentiated carcinomas, these tumors are believed to represent dedifferentiated cancer cells; however, the prognostic implications of TCV have been controversial. DEMOGRAPHICS AND PATHOGENESIS Most studies report that TCV accounts for 5% to 10% of papillary carcinomas, although the percentage of TCV in the tumor required for a diagnosis has not been entirely consistent. In addition, the TCV must be differentiated from the similar columnar cell variant characterized by similar tall cells that have stratified rather than basal nuclei. The reported cooccurrence of TCV and columnar cell variants suggests a similar pathogenesis for both tumors. TCV has also been associated with anaplastic carcinomas in a manner similar to that of insular carcinoma. In addition, a high incidence of mutations in the p53 tumor-suppressor gene has been reported in TCV, further supporting its transitional nature. Patients with TCV, like those with other dedifferentiated tumors, tend to be older and present with larger tumors. They are also more likely to present with local invasion or metastases.23 The typical female predominance of thyroid cancer is seen for TCV. CLINICAL PRESENTATION AND DIAGNOSIS Patients with TCV tend to present in a similar manner to those with well-differentiated papillary cancer, albeit at a later age. The frequency of local metastases is relatively high, and distant metastases are more common than for papillary carcinoma (19%).23,29,30 and 31,31a Occasionally, young individuals with small intrathyroidal TCV cancers are identified. Much controversy exists regarding the natural history of these tumors. Typically, patients with TCV present with a solitary thyroid nodule or a dominant nodule in a multinodular goiter and then are evaluated by FNA. The cytopathology of TCV generally reflects the papillary origin of the tumor with nuclear grooves, cytoplasmic inclusions, papillary fronds, and occasional psammoma bodies.32 Cells with basilar nuclei and abundant oxyphilic cytoplasm may suggest TCV, but usually the FNA is diagnostic for papillary cancer and the diagnosis is made on final surgical histology. TREATMENT AND PROGNOSIS The management of patients with TCV is similar to that for patients with other poorly differentiated tumors. These tumors are iodine avid and express thyroglobulin. Therefore, near-total thyroidectomy, aggressive radioiodine therapy, and TSH suppression are appropriate. Occasionally, external beam radiation is used for control of local disease. Patients are monitored with serum thyroglobulin measurement and radioiodine scanning. The use of neck CT, MRI, or ultrasonography is reasonable for patients with locally recurrent or invasive disease or those individuals with elevated levels of thyroglobulin to determine if their disease can be managed with surgical removal of the recurrence. Drawing conclusions regarding the prognosis of TCV is difficult due to the absence of large, long-term prospective or retrospective studies. Most published studies are small and retrospective with discordant outcomes. The author and colleagues have attempted to define an overall sense of the prognosis of TCV based on the published data.23 The prognosis of patients with TCV appears to be related to the stage of disease and the age at diagnosis. Patients older than 50 years of age tend to have larger tumors and present with more aggressive disease, and are more likely to have recurrence and to die of their disease than are similarly aged patients with well-differentiated cancer.23,29,30 and 31 Conversely, patients younger than 50 years of age tend to present with early-stage disease and have a prognosis similar to that for patients with well-differentiated cancer, although the follow-up time has been much shorter and the treatments have not been consistent. However, a review has described young patients with small primary tumors and agressive disease.31a Thus, although it seems relatively clear that older patients with TCV have a worse prognosis, in younger patients, the biologic behavior of TCV may be more similar to that of well-differentiated thyroid cancer. However, short lengths of follow-up as well as inconsistencies in treatment regimens may cause underestimation of the true aggressiveness of TCV in young patients. Therefore, the author recommends treating young patients who have TCV with the aggressive protocol previously mentioned unless they have small intrathyroidal primary tumors, in which case they can be treated like other young patients with well-differentiated papillary carcinoma. SUMMARY TCV probably represents a poorly differentiated subtype of papillary cancer that tends to present with later-stage disease in older patients. The biologic behavior in younger patients with smaller tumors appears to be less aggressive. TCV tumors contain molecular characteristics suggesting its role as a transitional tumor in the dedifferentiation of well-differentiated to anaplastic thyroid cancer. Treatment usually includes radioiodine therapy and TSH suppression because most of the tumors concentrate iodine. Additional studies will help clarify the biologic behavior of TCV, particularly in younger patients.

COLUMNAR CELL VARIANT OF PAPILLARY CARCINOMA

Columnar cell variant is similar to TCV in that it appears to be a dedifferentiated subtype characterized by tall cells. The distinction between the two is generally based on the nuclear stratification of the columnar cell versus the basal nuclei typical of TCV and the presence of clear rather than pink cytoplasm. Columnar cell carcinoma is much less common than TCV, with only 24 reported cases. It has been associated with anaplastic carcinoma, well-differentiated thyroid cancer, and TCV.33,34 CLINICAL PRESENTATION AND DIAGNOSIS The mean age of the reported cases of columnar cell carcinoma is 45 years; the female-male ratio is 1.4:1.0. The majority of the reported cases occurred in older individuals with locally invasive tumors that rapidly enlarged. Several young patients with smaller, less aggressive tumors have been described. Distant metastases, including pulmonary, bone, and brain metastases, were present in 29% of the reported patients at the time of diagnosis. As with TCV, several clues may suggest a columnar cell carcinoma on FNA, but usually the aspirate is diagnostic for papillary cancer. Histologic analysis is usually required to make the diagnosis. TREATMENT AND PROGNOSIS In the reported patients, most of the tumors were iodine avid, and patients were treated with thyroidectomy, radioiodine, and TSH suppression. Several patients also

received external beam radiation to the neck to control local spread.34 In the short periods of follow-up included in the reports, 38% of patients died of the cancer and 58% were free of disease at last follow-up, although the monitoring methods used are not clear. SUMMARY Columnar cell carcinoma is a rare variant of papillary cancer that likely represents a dedifferentiated subtype with an aggressive clinical course distinct from TCV. Cure seems possible for patients with small, intrathyroidal tumors, although the reported follow-up times have been short. Treatment with surgery, radioiodine, and TSH-suppression with external radiation for local disease is appropriate.

DIFFUSE SCLEROSING PAPILLARY THYROID CANCER

Although initially described in 1953,35 the diffuse sclerosing variant of papillary cancer was first formally defined in 1985 when it was identified as a subset of papillary cancers characterized by typical papillary cells, dense sclerosis, abundant psammoma bodies, squamous metaplasia, and significant lymphocytic infiltration.36 Subsequent reviews of papillary carcinomas have identified these tumors in 1% to 3% of cases.35,36 and 37 Based on the small number of reported cases, these tumors are typically large and often diffusely involve both lobes. They tend to occur in younger patients (mean age in the 30s) and show predominance among females. Cervical adenopathy is nearly uniformly present at diagnosis, and distant metastases and tumor recurrence are more common than in typical papillary cancer. Despite the apparent aggressive nature of the tumor at presentation, tumor-related mortality has not been reported, and patients surviving decades after diagnosis are well described. This may be related to the lymphocytic response to the tumor. Treatment with near-total thyroidectomy, lymph node removal, and radioiodine all are appropriate. Longer follow-up will be required to better define the clinical course of this disease.

SQUAMOUS CELL CARCINOMA OF THE THYROID

Primary squamous cell carcinoma of the thyroid is defined by the WHO classification as a tumor comprised entirely of cells showing so-called intracellular bridges and/or forming keratin.1 The squamous cells are of uncertain origin.38 When this definition of pure squamous cell carcinoma is used, rather than that of tumors with an adenosquamous appearance or tumors with squamous metaplasia, the incidence is <1% of all thyroid malignancies.39 Before therapy, the clinician must be certain that the tumor does not represent metastases from head and neck or lung cancers. DEMOGRAPHICS AND PATHOGENESIS Forty-five individuals have been identified by MEDLINE review who clearly had tumors meeting the WHO definition of pure squamous cell carcinoma. The median age of 45 years at presentation is younger than that for anaplastic carcinoma but older than that for well-differentiated cancer. Female-male predominance is 1.8:1. As noted previously, the origin of the squamous cells remains quite controversial. The relatively high frequency of squamous cell carcinoma of thyroglossal duct remnants raises the possibility that small nests of squamous cells exist in the thyroid glands. Indeed, a small subset of thyroid squamous cells with high basal cell proliferation has been identified.40 The relatively common finding of squamous metaplasia does not typically occur in association with squamous cell carcinoma, and the pattern of evolution of squamous metaplasia to carcinoma is not characteristic of other squamous cell cancers. Thus, squamous metaplasia is unlikely to represent a premalignant lesion. CLINICAL PRESENTATION AND DIAGNOSIS Most patients with squamous cell thyroid carcinoma present with a rapidly enlarging mass arising within a multinodular goiter, as in anaplastic cancer. Tumor production of PTHrP and G-CSF has been described.12 Rarely, this cancer has been described to occur in conjunction with other thyroid malignancies. Squamous cell carcinomas are difficult to diagnose on FNA because of the difficulty in differentiating metaplasia from carcinoma. Some patients present with distant metastases, but more typically, these tumors are locally invasive in the neck. Squamous carcinomas may be more common in thyroglossal duct remnants and account for 7% of all primary thyroglossal duct carcinomas.23 TREATMENT AND PROGNOSIS The management of squamous cell carcinoma of the thyroid is similar to that of other squamous cell tumors of the head and neck and anaplastic thyroid carcinoma. No role exists for radio-iodine therapy because these tumors do not concentrate iodine, and TSH suppression is not known to be beneficial. A complete surgical resection, often including bilateral neck dissections, is performed. After surgery, local control is paramount, and patients are typically treated with external beam radiation to control local recurrence. Distant metastases have been treated with a variety of chemotherapy regimens including bleomycin, cisplatin, and doxorubicin, with disappointing results. Overall, survival of patients with squamous cell cancer is low, with nearly all patients dying within 16 months of diagnosis. Long-term survival appears to be possible for patients with smaller tumors that are amenable to complete surgical resection.

MUCOEPIDERMOID CARCINOMA
Mucoepidermoid cancer is a rare tumor (~40 cases) that is associated with epithelial thyroid cancers and lymphocytic thyroiditis.41 These tumors may be cystic or mucin filled and present as a typical thyroid nodule. FNA is often nondiagnostic or suggestive of carcinoma, but not of mucoepidermoid cancer; therefore, the diagnosis is made on the basis of surgical histology. Many of the reported cases had metastases (55%), but these were generally confined to the neck. Thyroidectomy, lymph node dissection, and external beam radiation therapy appear to be effective, but the duration of follow-up has been short. The role of radio-iodine therapy and TSH suppression has not been explored.

MIXED MEDULLARY-FOLLICULAR AND MEDULLARY-PAPILLARY CARCINOMA

Rarely, thyroid cancers may display features of both medullary and follicular carcinoma.42 These tumors are distinct from medullary cancer with normal trapped follicles and are characterized by distinct regions of medullary carcinomas adjacent to follicular or papillary cancer. The pathogenesis of these tumors is not clear. Their occurrence suggests a similar response of follicular and parafollicular cells to an oncogenic stimulus rather than the presence of a common progenitor cell. However, cells expressing both thyroglobulin and calcitonin have been identified.42 The presence of mutations in the ret oncogene, typical of medullary cancer, have not been studied, but familial occurrence of these tumors has been described.43,44

THYROID SARCOMAS
True sarcomas of the thyroid are quite rare and must be differentiated from anaplastic cancers with sarcomatoid features. Careful analysis has disclosed several cases of thyroid sarcoma, including leiomyosarcomas, osteosarcomas, chondrosarcomas, fibrosarcomas, and angiosarcomas.45,46 Of these subtypes, angiosarcomas are most common and are found most frequently in regions of iodine deficiency (another term is heman-gioepitheliomas). These tumors, particularly the angiosarcomas, tend to occur in older patients with long-standing goiters. Careful immuno-histochemical analysis must be performed to separate these tumors from carcinomas. The markers should confirm endothelial lineage (e.g., CD34, CD31, and factor VIIrelated antigen) rather than epithelial lineage. The prognosis for thyroid sarcomas, regardless of subtype, is poor. Most patients have large tumors with local invasion and metastases on diagnosis. Rare cures in cases of small intrathyroidal tumors have been described. Patients are treated with surgery, external irradiation, and chemotherapy regimens used for sarcomas arising from other organs. In addition to primary thyroid sarcomas, Kaposi sarcoma infiltrating the thyroid has been described in the setting of acquired immunodeficiency syndrome (AIDS)47 and should be considered in the AIDS patient group.

THYROID TERATOMA
Thyroid teratomas are rare, usually benign, and most frequently diagnosed in childhood. Teratomas of the thyroid, like teratomas at any location, are characterized by the presence of cells arising from all three germ cell layers. Thyroidal origin is determined by identification of the blood supply leading to the tumor.

Malignant teratomas are extremely rare and, when considered as a percentage of all teratomas, are more common in adults. A review has determined that mean age at diagnosis is 31.2 years, size is large (up to 17 cm), and prognosis is poor.48 All patients had metastases, all were treated surgically and with external beam radiation, and all died within 22 months of diagnosis with the exception of one patient for whom only a 7-month follow-up was reported.

THYROID LYMPHOMA
Classically, primary thyroid lymphomas occur in patients with Hashimoto thyroiditis and are usually non-Hodgkin B-cell lymphomas. Lymphomas include the small cell anaplastic carcinoma that has been determined to represent a small B-cell lymphoma. DEMOGRAPHICS AND PATHOGENESIS As previously noted, thyroid lymphoma generally occurs in patients with Hashimoto thyroiditis. Therefore, this malignancy is more common in women. It typically occurs in the fifth and sixth decades of life, but patients younger than 40 years have been reported.23,49,50 Nearly all patients have histologic evidence of Hashimoto thyroiditis and approximately two-thirds have detectable serum levels of antithyroid peroxidase or antithyroglobulin antibodies. Most patients are euthyroid at the time of diagnosis, although hypothyroidism is common. Primary thyroid lymphomas account for only a small percentage of lymphomas or thyroid cancers, but the tumors tend to respond well to appropriate therapy; thus, early and accurate diagnosis is critical. The pathogenesis of thyroid lymphoma is unclear. The association with thyroiditis has given rise to the hypothesis that thyroid lymphoma may be a variety of MALT (mucosa-associated lymphoid tissue) lymphomas. The cells of MALT lymphomas typically resemble cells that surround lymphoid follicles (centrocyte-like); they can undergo plasmacytic changeas has been described in thyroid lymphomas51 and can progress to high-grade lymphoma. Thyroid lymphomas are comprised of a population of monotonous lymphoid cells that replace the thyroid parenchyma, with additional evidence of prior Hashimoto thyroiditis. CLINICAL PRESENTATION AND DIAGNOSIS Most patients with thyroid lymphoma have hypothyroidism or a preexisting goiter that begins to grow, sometimes rapidly and with pain.49 Local symptoms such as dysphonia and dysphagia occur but are uncommon. The classic B symptoms of lymphoma are also usually absent. Presurgical diagnosis is often possible by FNA, provided the clinician's index of suspicion is high before performing the procedure. Flow cytometric analysis of the aspirated cells has been shown to be quite accurate, with a high specificity and sensitivity.49,50 In addition, typical ultrasonographic findings of a pseudocystic pattern50 have been identified in thyroid lymphoma; this pattern is present in only 10% of patients with Hashimoto thyroiditis without lymphoma. These methods are dependent on the expertise of the pathologists and radiologists at a given institution. Therefore, even with these advances, open surgical biopsy, tumor debulking, or thyroidectomy is frequently required to confirm the diagnosis before therapy. Gallium scanning may also be helpful in diagnosing lymphoma. TREATMENT AND PROGNOSIS The treatment of patients with thyroid lymphoma has been controversial, specifically with regard to the need for near total thyroidectomy in all patients. One large study reported 8-year disease-free survivals of 100% in 16 individuals treated with cyclophosphamide, doxorubicin, vincristine, and prednisone (CHOP) for one cycle, followed by 40 to 60 Gy of external radiation and then an additional five cycles of CHOP without a surgical thyroidectomy.50 These results are better than the previously reported cure rates, which approach 80% in the studies that used combined-modality therapy.49 The role for thyroidectomy is not clear, and its routine use in patients with thyroid lymphoma is debated. The physician must have complete confidence in the diagnosis of lymphoma before initiating therapy because the treatment differs dramatically from that for any other thyroid malignancy. If a core sample is required in addition to the FNA, performing a lobectomy or near-total thyroidectomy may be reasonable if the surgical risk is acceptably low. Alternatively, a core biopsy could be performed, and if the results are consistent with lymphoma, treatment could be initiated without waiting several weeks for the patient to recover fully from a thyroidectomy. SUMMARY Thyroid lymphoma is an uncommon malignancy that generally occurs in individuals with preexisting Hashimoto thyroiditis; it presents as an enlarging, tender gland, often with local symptoms. Diagnosis is based on a high level of clinical suspicion and often requires histologic confirmation with a core biopsy of partial or total thyroidectomy. Treatment is based on combined chemotherapy and radiation therapy. Reports for aggressive high-dose therapy suggest a good long-term prognosis.

METASTATIC CANCERS IN THE THYROID

Clinically evident metastases to the thyroid are rare occurrences. They generally present as typical solitary thyroid nodules and are most commonly due to renal cell carcinoma, breast cancer, lung cancer, and malignant melanoma. Usually, widespread disease is evident, but occasionally, thyroid metastases may be the sentinels for metastatic disease. CLINICAL PRESENTATION AND DIAGNOSIS In one study, 40 of 43 patients with cancer metastatic to the thyroid gland presented with nodular disease (solitary or multinodular), and 3 presented with compression of local structures.52 FNA is often adequate to confirm the diagnosis if the patient has widespread metastatic disease at the time of the procedure and the cells resemble those of the original tumor. Rarely, a patient may develop a nodule many years after therapy for the primary tumor. This occurs most commonly with renal cell carcinoma. In this situation, surgical lobectomy may be required to confirm the diagnosis. The tissue of origin for the tumors that metastasize is most commonly renal cell carcinoma, followed by lung cancer, and breast cancer. Spread from adjacent head and neck cancers as well as parathyroid carcinomas is also well described. Hematologic malignancies (i.e., lymphoma, leukemia, and myeloma) can metastasize or infiltrate the thyroid. Kaposi sarcoma metastatic to the thyroid has also been reported.47 TREATMENT AND PROGNOSIS The treatment and prognosis of metastatic cancer to the thyroid gland relates to the primary tumor and not to the thyroid metastases. In selected cases, thyroidectomy may be beneficial, but generally, metastases to the thyroid occur in the setting of other metastatic disease. A discussion of the prognosis of the metastatic cancer should be held with patients, their families, and the physicians (including a medical oncologist), because the outcome is generally poor.

CONCLUSION
This chapter has reviewed a variety of unusual thyroid malignancies. For many of these tumors, too few cases exist for definitive recommendations to be made; however, many, if not all, of these tumors are seen in clinical practice. Therefore, these patients should be diagnosed and treated in a rational manner based on the biology and prognosis of the individual cancer. Clinical trials, which are ongoing in evaluating the use of new treatments for many of these tumors, hopefully will improve clinical outcomes in the next several years. CHAPTER REFERENCES
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. Hedinger C, Williams ED, Sobin LH. Histological typing of thyroid tumors. Berlin: Springer-Verlag, 1988. Ain KB. Anaplastic thyroid carcinoma: behavior, biology, and therapeutic approaches. Thyroid 1997; 8:715. Samaan NA, Ordoez NG. Uncommon types of thyroid cancer. Endocrinol Metab Clin North Am 1990; 19:637. Tan RK, Finley RK, Driscoll D, et al. Anaplastic carcinoma of the thyroid: a 24 year experience. Head Neck 1995; 17:41. dos Santos Silva I, Swerdlow AJ. Thyroid cancer epidemiology in England and Wales: time trends and geographical distribution. Br J Cancer 1993; 67:330. Harach HR, Escalante DA, Onativa A, et al. Thyroid carcinoma and thyroiditis in an endemic goitre region before and after iodine prophylaxis. Acta Endocrinol 1985; 108:55. Vanketesh YSS, Ordoez NG, Schultz PN, et al. Anaplastic carcinoma of the thyroid: a clinicopathologic study of 121 cases. Cancer 1990; 66:321. Fagin JA. Genetic basis of endocrine disease 3: molecular defects in thyroid gland neoplasia. J Clin Endocrinol Metab 1992; 75:1398. Ron E, Doody MM, Becker DV, et al. Cancer mortality following treatment for adult hyperthyroidism. JAMA 1998; 280:347. LiVolsi VA, Brooks JJ, Arendash Durand B. Anaplastic thyroid tumors. Immunohistology. Am J Clin Pathol 1987; 87:434. Yazawa S, Toshimori H, Nakatsuru K, et al. Thyroid anaplastic carcinoma producing granulocyte-colony-stimulating factor and parathyroid hormone-related protein. Int Med 1995; 34:584. Saito K, Kuratomi Y, Yamamoto K, et al. Primary squamous cell carcinoma of the thyroid associated with marked leukocytosis and hypercalcemia. Cancer 1981; 48:2080.

13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31.

Kim JH, Leeper RD. Treatment of locally advanced thyroid carcinoma with combination doxorubicin and radiation therapy. Cancer 1987; 60:2372. Junor EJ, Paul J, Reed NS. Anaplastic thyroid carcinoma: 91 patients treated by surgery and radiotherapy. Eur J Surg Oncol 1992; 18:83. Tennvall J, Lundell G, Hallquist A, et al. Combined doxorubicin, hyperfractionated radiotherapy, and surgery in anaplastic thyroid carcinoma: report on two protocols. Cancer 1994; 74:1348. Schlumberger M, Parmentier C, Delisle M-J, et al. Combination therapy for anaplastic giant cell thyroid carcinoma. Cancer 1991; 67:564. Shimaoka K, Schoenfeld DA, DeWys WD, et al. A randomized trial of doxorubicin versus doxorubicin plus cisplatin in patients with advanced thyroid carcinoma. Cancer 1985; 56:2155. Woolner LB, Beahrs OH, Black BM, et al. Classification and prognosis of thyroid carcinoma: a study of 885 cases observed in a thirty year period. Am J Surg 1961; 102:354. Ain KB, Tofiq S, Taylor KD. Antineoplastic activity of Taxol against human anaplastic thyroid carcinoma cell lines in vitro and in vivo. J Clin Endocrinol Metab 1996; 81:3650. Langhans T. ber die epithelialen formen der malignen struma. Virchows Arch (Pathol Anat) 1907; 385:125. Carcangiu ML, Zampi G, Rosai J. Poorly differentiated (insular) thyroid carcinoma. Am J Surg Pathol 1984; 8:655. Papotti M, Botto Mica F, Favero A, et al. Poorly differentiated thyroid carcinomas with primordial cell component: a group af aggressive lesions sharing insular, trabecular, and solid patterns. Am J Surg Pathol 1993; 17:291. Burman KD, Ringel MD, Wartofsky L. Unusual types of thyroid neoplasms. Endocrinol Metab Clin North Am 1996; 25:49. Justin EP, Seabold JE, Robinson RA, et al. Insular carcinoma: a distinct thyroid carcinoma associated with iodine-131 localization. J Nucl Med 1991; 32:1358. Ashfaq R, Vuitch F, Delgado R, Albores-Saavedra J. Papillary and follicular carcinomas with an insular component. Cancer 1994; 73:416. Pietribiasi F, Sapino A, Papotti M, Bussolati G. Cytologic features of poorly differentiated insular carcinoma of the thyroid, as revealed by fine-needle aspiration biopsy. Am J Clin Pathol 1990; 94:687. Grunwald F, Menzel C, Bender H, et al. Redifferentiation therapy-induced radioiodine uptake in thyroid cancer. J Nucl Med 1998; 39:1903. Hawk WA, Hazard JB. The many appearances of papillary carcinoma of the thyroid. Cleve Clin Q 1976; 45:207. Terry JH, St John SA, Karkowski FJ, et al. Tall cell papillary thyroid carcinoma: incidence and prognosis. Am J Surg 1994; 168:459. Johnson TL, Lloyd RV, Thompson NW, et al. Prognostic implications of the tall cell variant of papillary thyroid cancer. Am J Surg Pathol 1988; 12:22. Ain KB. Papillary thyroid carcinoma: etiology, assessment, and therapy. Endocrinol Metab Clin North Am 1995; 24:711.

31a. Prendiville S, Burman KD, Ringel MD, et al. Tall cell variant: an agressive form of papillary thyroid cancer. Otolaryngol Head Neck Surg 2000; 122:352. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. Harach HR, Zusman SB. Cytopathology of the tall cell variant of thyroid papillary carcinoma. Acta Cytol 1992; 36:895. Asklen LA, Verhaug JE. Thyroid carcinoma with mixed tall-cell and columnar-cell features. Am J Clin Pathol 1990; 94:442. Mizukami Y, Nonomura A, Michigishi T, et al. Columnar cell carcinoma of the thyroid gland: a case report and review of the literature. Hum Pathol 1994; 25:1098. Crile G Jr, Fisher ER. Simultaneous occurrence of thyroiditis and papillary carcinoma. Report of two cases. Cancer 1953; 6:57. Vickery AL, Carcangiu ML, Johannessen JV, Sobrinho-Simoes M. Papillary carcinoma. Session I of thyroid tumor pathology. Semin Diagn Pathol 1985; 2:909. Caplan RH, Wester S, Kisken WA. Diffuse sclerosing variant of papillary thyroid carcinoma: case report and review of the literature. Endocr Pract 1997; 3: 287. LiVolsi VA, Merino MJ. Squamous cells in the thyroid gland. Am J Surg Pathol 1978; 2:133. Misonou J, Aizawa M, Kanda M, et al. Pure squamous cell carcinoma of the thyroid glandreport of an autopsy case and review of the literature. Jpn J Surg 1988; 18:469. Bond JA, Wyllie FS, Ivan M, et al. A variant epithelial sub-population in normal thyroid with high proliferative capacity in vitro. Mol Cell Endocrinol 1993; 93:175. Wenig BM, Adair CF, Heffess CS. Primary mucoepidermoid carcinoma of the thyroid gland: a report of six cases and a review of the literature of a follicular epithelial-derived tumor. Hum Pathol 1995; 26:1099. Papotti M, Negro F, Carney JA, et al. Mixed medullary-follicular carcinoma of the thyroid. A morphological, immunohistochemical and in situ hybridization analysis of 11 cases. Virchows Arch (Pathol Anat) 1997; 430:397. Mizukami Y, Michigishi T, Nonomura A, et al. Mixed medullary-follicular carcinoma of the thyroid occurring in the familial form. Histopathology 1993; 22:284. Lamberg BA, Reissel P, Stenman S, et al. Concurrent medullary and papillary thyroid carcinoma in the same thyroid lobe and in siblings. Acta Med Scand 1981; 209:421. Thompson LDR, Wenig BM, Adair CF, et al. Primary smooth muscle tumors of the thyroid gland. Cancer 1997; 79:579. Syrjnen KJ. An osteogenic sarcoma of the thyroid gland (report of a case and survey of the literature). Neoplasma 1979; 26:623. Mollison LC, Mijch A, McBride G, Dwyer B. Hypothyroidism due to destruction of the thyroid by Kaposi's sarcoma. Rev Infect Dis 1991; 13:826. Bowker CM, Whittaker RS. Malignant teratoma of the thyroid: case report and literature review of thyroid teratoma in adults. Histopathology 1992; 21:81. Weinstein LJ, Ain KB. Primary thyroid lymphoma: a comprehensive assessment and clinical approach. Endocrinologist 1999; 9:45. Matsuzuka F, Miyauchi A, Katayama S, et al. Clinical aspects of primary thyroid lymphoma: diagnosis and treatment based on our experience of 119 cases. Thyroid 1993; 3:93. Isaacson PG. Lymphomas of mucosa-associated lymphoid tissue (MALT). Histopathology 1990; 16: 617. Nakhjavani MK, Gharib H, Goellner JR, van Heerden JA. Metastasis to the thyroid gland. A report of 43 cases. Cancer 1997; 79: 574.

CHAPTER 42 HYPERTHYROIDISM Principles and Practice of Endocrinology and Metabolism

CHAPTER 42 HYPERTHYROIDISM
KENNETH D. BURMAN Pathogenesis of Graves Disease Clinical Manifestations of Graves Disease Ocular Manifestations Laboratory Assessment of Graves Disease Thyroid Function Tests TSH Receptor Antibody Measurements Other Laboratory Tests in Graves Disease Differential Diagnosis of Hyperthyroidism Graves Disease Surreptitious Ingestion of Thyroxine or Triiodothyronine Toxic Multinodular Goiter Solitary Autonomous Nodule Exogenous Iodine Excess Subacute (Granulomatous) Thyroiditis Spontaneously Resolving Thyroiditis Postpartum Thyroiditis Hashitoxicosis Trophoblastic Tumors Struma Ovarii Thyroid-Stimulating-Hormonesecreting Pituitary Tumor Metastatic Thyroid Cancer Hamburger Thyrotoxicosis Natural History of Graves Disease Treatment of Graves Disease Antithyroid Drugs b-Blockers and Other Agents Radioactive Iodine Therapy Surgery Graves Disease in Special Situations Pregnancy and Graves Disease Thyroid Storm Euthyroid Graves Disease Hashimoto Thyroiditis Versus Graves Disease Graves Disease in Children Graves Disease in the Aged
L-Thyroxine Administration in Treated Graves Disease

Suppressed Serum Thyroid-Stimulating Hormone Chapter References

Of the several varieties of hyperthyroidism, one of the most common is Graves disease. Graves disease is an autoimmune disease in which thyroid-stimulating hormone (TSH) receptor antibodies bind to and stimulate the thyroid gland, causing the excessive secretion of thyroxine (T4) or triiodothyronine (T3) or both, resulting in the clinical manifestations of thyrotoxicosis.1,2,3,4,5 and 6 Graves disease has fascinated clinicians because of its possible relationship to stress, its unusual and varied manifestations (e.g., eye involvement, pretibial myxedema), and its unpredictable course, characterized by relapses and exacerbations.

PATHOGENESIS OF GRAVES DISEASE

Graves disease is associated with TSH receptor antibodies that stimulate the thyroid gland.1,2,3,4,5 and 6 In most patients, the population of TSH receptor antibodies is heterogeneous, with some antibodies that stimulate and others that inhibit TSH receptor-mediated action. Patients have been described with even more complex antibodies in their serum, as determined by binding or stimulating actions on the TSH receptors in thyroid membrane or guinea pig fat membrane preparations.7 There are four major hypotheses as to the etiology of Graves disease (Fig. 42-1):

FIGURE 42-1. A through D, Theoretical explanations for the initiation or propagation of Graves disease (see text). T3 receptor complex refers to the T-cell antigen receptor, of which the CD3 molecule is a component. T3 and T4 refer to triiodothyronine and thyroxine derived from thyroidal secretion. (cAMP, cyclic adenosine monophosphate; TSH, thyroid-stimulating hormone.) (From Burman KD, Baker JR Jr. Immune mechanisms in Graves disease. Endocr Rev 1985; 6:183. Reproduced by permission of The Endocrine Society.)

1. A basic defect in antigen-specific suppressor T cells allows an imbalance in helper cell action versus suppressor cell function, resulting in the excessive generation or unregulated synthesis of TSH receptor antibodies.8,9 and 10 Abnormal macrophage migration inhibition and monocyte procoagulant activities in patients with Graves disease represent relatively specific defects observed only with peripheral mononuclear cells of these patients in the presence of thyroidal antigens.11 These same peripheral mononuclear cells are capable of correcting the pancreatic islet cell specific antigen defect found in diabetic mononuclear cells. Although this theory is appealing, it does not explain the etiologic mechanism of the antigen-specific suppressor cell defect,12,13 nor does it explain the polyclonal B- and T-cell patterns observed with molecular probes.14 2. A defect may exist in the mechanism by which thyrocytes and T cells initiate helper T-cell activation. Thyrocytes themselves can express human leukocyte antigen (HLA) class II antigens, and it is speculated that HLA antigens, as well as thyroid-specific antigen, interact with helper cells, presumably through the T-cell T3 receptor complex.15 Although the thyroid gland's ability to present HLA antigens is thought by some to be a primary event, others believe that the expression of D-related (DR) antigens is secondary to the presence of a small number of intrathyroidal lymphocytes that are activated and thus secrete interleukin-2, which is known to cause HLA-DR expression.16 This hypothesis does not explain the mechanism involved in T-cell activation. 3. According to the idiotypic-antiidiotypic network theory, antibodies themselves initiate the development of secondary antibodies against the idiotypic combining sites of the primary antibody.17 This sequence of secondarily developing antibodies results in a cascade by which the body can modulate or regulate antibody formation. As applied to Graves disease, TSH receptor antibodies would provoke secondary antibodies directed against the combining site of the TSH receptor antibodies. These secondary antiidiotypic antibodies would look like a mirror image of the TSH receptor antibody-combining site and thus would bind both TSH and TSH receptor antibodies. The observation that antiidiotypic antibodies are present in the serum of patients with Graves disease is consistent with an

alteration in the network cascade, at least in some patients.18 The fourth hypothesis relates to the presence of TSH receptor-like antigens in various bacteria or parasites. Yersinia enterocolitica possesses a specific TSH-binding site that has binding characteristics similar to those of the thyroidal TSH receptor.19 Iodine-125labeled TSH bound to the receptor could be displaced by unlabeled TSH but not by other glyco-protein hormones. The identification of this TSH receptor in other bacteria (e.g., Mycoplasma) or parasites has led to the hypothesis that infection with agents possessing these TSH receptors could result in the formation of antibodies directed against the thyroidal TSH receptor, causing stimulation (or blockade).19,20 Alternatively, these bacteria or parasites could be engulfed by macrophages and could then be processed and presented on the macrophage surface to initiate T-cell activation in genetically susceptible patients. A potentially very important finding is that the peripheral lymphocytes from patients with Graves disease may contain human spumaretro-virus-related sequences.21 In other work, the TSH receptor has been found to have mutations, especially if ophthalmopathy is present.22 Further work is required to determine the importance of these and other findings and their relevance to the initiation or propagation of autoimmune thyroid disease.

CLINICAL MANIFESTATIONS OF GRAVES DISEASE

The clinical manifestations of Graves disease are myriad and can involve almost any organ system (Table 42-1; Fig. 42-2, Fig. 42-3 and Fig. 42-4).23,24,25,26,27,28,29,30,31,32 and 33 Usually, there is a palpable, diffusely enlarged, smooth goiter that initially may be soft but becomes progressively firmer. Because of the increased vascularity of the gland, there may be a systolic bruit heard with the stethoscope. Patients commonly report nervousness, malaise, irritability, inability to concentrate, hand tremor, weight loss, and burning or itching eyes. Unilateral or bilateral proptosis is frequent, and vitiligo, pruritus, osteoporosis, and gynecomastia can occur. Choreoathetosis and ataxia are rare presenting signs or symptoms.27 A severe, incapacitating, elephantiasis-like syndrome with pretibial myxedema and acropachy is also seen rarely and may be exacerbated by local trauma (see Fig. 42-3 and Fig. 42-4).28

TABLE 42-1. Effects of Thyrotoxicosis

FIGURE 42-2. A, Severe weight loss, nervousness, and sweating in a thyrotoxic man. B, This 57-year-old man with Graves disease reported diplopia and tachycardia.

FIGURE 42-3. A, Graves disease in a 33-year-old man manifested by bilateral exophthalmos, acropachy (clubbing), extensive pretibial myxedema, and insulin-dependent diabetes mellitus. When this photograph was taken, he already had been treated with 131I, had become hypothyroid, and was receiving replacement therapy with exogenous L -thyroxine. B, Radiograph in same patient showing phalangeal periosteal reaction (arrow).

FIGURE 42-4. A and B, Different degrees of involvement of the tissues in pretibial myxedema. Patient in B also shows a patch of vitiligo.

Patients with Graves disease are more likely to have or to acquire several other disorders, such as Addison disease, diabetes mellitus, idiopathic thrombocytopenic purpura, myasthenia gravis, pernicious anemia, rheumatoid arthritis, scleroderma, systemic lupus erythematosus, vitiligo, dermatitis herpetiformis, thymic hyperplasia, and Albright syndrome (cutaneous pigmentation, precocious puberty, and polyostotic fibrous dysplasia of bones).23,24,25,26,27,28,29,30 and 31 Myopathies that are mild or moderate occur frequently in Graves disease, affecting skeletal or ocular muscles. Muscle weakness involving the shoulders, limbs, and

particularly the quadriceps muscles is quite common and occasionally profound. Thyrotoxicosis with periodic paralysis occurs rarely; it is characterized by sporadic attacks, most commonly involving flaccidity and paralysis of the legs, arms, and trunk, although any muscle (e.g., facial) can be involved (see Chap. 210). Respiratory muscles can be affected, but the heart is usually not involved. These attacks may last from minutes to many hours and can occur spontaneously or after the ingestion of carbohydrates or after exercise. Occasionally the serum potassium concentration is decreased. Treatment of thyrotoxicosis with periodic paralysis entails restoration of euthyroidism with antithyroid drugs. Administration of b-adrenergic blockers sometimes rapidly stops the attack, and potassium should be administered when serum levels are low. Thyrotoxicosis with periodic paralysis occurs most frequently in Asians. Changes in bone parameters can also occur.31a Patients with Graves disease may also have myxedema in the pretibial or preradial area. This accumulation of fluid and mucopolysaccharides can be mild and clinically inconsequential or severe and incapacitating. The causes of these connective tissue effects are unknown but may be related to abnormal T- or B-cell stimulation or to cross-reacting epitopes in these tissues and serum antithyroid antibodies. The onset and progression of these problems may not correlate with the severity or progression of the thyrotoxicosis. OCULAR MANIFESTATIONS The onset of Graves ophthalmopathy may precede, coincide with, or follow the thyrotoxicosis,29,30 and sometimes its clinical course appears unrelated to the thyrotoxicosis. Nevertheless, it is a clinical dictum that rapid restoration of the euthyroid state is an important component of the management of ophthalmopathy (see Chap. 43). The pathogenesis of Graves ophthalmopathy and its relationship to the thyroidal manifestations are unclear, but it appears that retroocular fibroblasts contain TSH-receptor transcripts and, perhaps, receptor protein. It is speculated that TSH-receptor antibodies may bind to the TSH receptor within the fibroblasts, mediating or propagating the immune response. Extraocular muscle antigens may also be involved in the pathogenetic process.

LABORATORY ASSESSMENT OF GRAVES DISEASE

The initial laboratory assessment of a patient with thyrotoxic Graves disease should include measurement of hemoglobin, hematocrit, white blood cell count, liver profile (serum alkaline phosphatase, transaminases, and bilirubin), calcium, blood urea nitrogen, creatinine, and electrolytes, as well as thyroid function tests.32,33 THYROID FUNCTION TESTS Serum TSH should be undetectable in patients with active hyperthyroidism; serum total and free T4 and T3 may be measured and are usually elevated. Although there may be different approaches based on background and clinical practice, most patients with suspected Graves disease should have free T4, total triiodothyronine (TT3), and TSH determined initially to help confirm the diagnosis. When patients with known Graves disease are observed, usually free T4 and TT3 are determined, and occasionally TSH may be useful. Evaluation of TT3 levels should be routine, but may be especially useful when the clinical signs or symptoms are marked and seemingly disproportionate to the free T4 elevation.34,35 T 3 toxicosis, a condition of clinical thyrotoxicosis with a normal serum T4 level but an elevated TT3 level, is relatively unusual; it is more likely to occur in patients with an autonomous nodule, early Graves disease, or recurrent disease. To develop tests that would not be influenced by binding proteins, investigators have developed free T4 and free T3 measurements that are sensitive and applicable to clinical use.35,36,37 and 38 Free T4 should replace the more traditional total T4 and resin T3 uptake tests given that it is not influenced by alterations in binding proteins and because it is now widely available. However, free T3 is not yet readily available commercially with a rapid turnaround time; therefore, it is still appropriate to obtain a TT3 level, although it must be remembered that TT3 levels are increased when there are elevations of T4-binding globulin (e.g., pregnancy, estrogen). Serum thyroglobulin levels and antithyroglobulin or antimicrosomal antibodies need not be measured routinely in a patient with Graves disease. In most patients with Graves disease (except pregnant or lactating women), the 24-hour (or 6-hour) radioactive iodine uptake should be determined to differentiate Graves disease from thyroiditis (subacute, painless, postpartum). In painless subacute thyroiditis, the radioactive iodine uptake is <5% when the patient is hyperthyroid with elevated free T4 and TT3 in conjunction with a decreased TSH level. However, the physician also should consider postpartum thyroiditis, subacute thyroiditis, recent exposure to excessive iodide in drugs, and amiodarone use or radiopaque dyes (as used in computed tomography scans or angiograms) that may artifactually lower the uptake. A radioisotopic thyroid scan in a thyrotoxic patient helps confirm that a patient has Graves disease when the scan shows diffuse homogeneous uptake of tracer, although such a scan is not always necessary. However, a palpable thyroid nodule or other unusual finding (e.g., lymphadenopathy) or worrisome signs or symptoms (e.g., pain, rapid growth of thyroid gland) are indications for a radioisotopic thyroid scan to help rule out concomitant neoplasm. Of course, fine-needle aspiration of suspicious or abnormal nodules (intrathyroidally or extrathyroidally) or nodes should be performed as appropriate to help determine the nature of these abnormalities. Occasionally it is difficult to determine clinically that a patient is thyrotoxic, especially when the thyroid gland is only minimally enlarged and the serum T4 or free T4 value is normal or marginally elevated. A thyrotropin-releasing hormone (TRH) stimulation testin which serum TSH is measured 0, 15, and 30 minutes after the administration of 200 to 500 g of TRH by intravenous bolusmay then be indicated. An increase in the TSH level of 2 to 3 U/mL or more after TRH administration excludes hyperthyroidism. With the exception of hypopituitarism, a blunted response almost always indicates hyperthyroidism, although psychiatric patients, patients with hyperemesis gravidarum, or elderly patients may have a blunted or minimal rise in TSH after TRH stimulation, even when they are euthyroid (see Chap. 15). Use of a highly sensitive TSH assay has obviated the utility of most TRH tests. Third-generation TSH assays have a sensitivity of ~0.01 U/mL, and these assays can discriminate subnormal from undetectable TSH concentrations. Thus, thyrotoxic patients have an undetectable TSH level, except in rare instances of a TSH-secreting pituitary tumor.37,38 Serum T4 levels may be elevated in other conditions that can be classified as euthyroid hyperthyroxinemia (see Chap. 36).39 Several other laboratory tests not commonly required in the evaluation of a thyrotoxic patient are occasionally helpful. A 24-hour urinary iodide excretion test may help identify hyperthyroid patients who have a low radioactive iodine uptake secondary to exogenous iodide ingestion. A patient ingesting a normal diet may excrete 500 to 1000 g of iodide in the urine daily. The minimal daily requirement of iodine is ~150 g per day. A patient with a low (<5%) radioactive iodine uptake secondary to exogenous iodine intake usually excretes >1000 g of iodine a day. Serum T4-binding globulin levels may be helpful in selected patients but are unnecessary if there is an obvious cause of the elevated T4-binding globulin (e.g., use of birth control pills) associated with elevated T4 levels and a decreased resin T3 uptake. TSH RECEPTOR ANTIBODY MEASUREMENTS Occasionally it may be important to measure TSH receptor antibodies. TSH receptor antibodies in the sera may be measured by bioassays that detect cyclic adenosine monophosphate production by immunoglobulin G (IgG) isolated from patients with Graves disease (thyroid-stimulating immunoglobulin [TSI] assay), by the ability of Graves IgG to inhibit 125I-TSH binding to thyroid-binding inhibitory immunoglobulins (TBII) of thyroid membranes (TBII assay), or by the ability of sera to bind to TSH receptors in enzyme-linked immunosorbent assay (ELISA) (Table 42-2).1,40,41 and 42 TSH receptor antibodies are usually of the IgG class. In general, assays that detect stimulatory antibodies are preferable. Assays that measure binding to TSH receptors (TBII or ELISA) are easier to perform but may detect antibodies that are not stimulatory and, thus, may have less clinical relevance to Graves thyrotoxicosis. TSI assays have been improved by the use of recombinant TSH receptor. Inhibitory (blocking) TSH receptor antibodies may be found in ~20% of patients with primary hypothyroidism. Some infants may become hypothyroid in the first 3 months of life because of the transplacental passage of these inhibitory antibodies. TSH receptor antibody assays can be modified to detect inhibitory immunoglobulins by assessing the ability of the patient's IgG to inhibit TSH-mediated rises in cyclic adenosine monophosphate. Clinical circumstances in which measurement of TSH receptor antibody may be helpful include differentiation of autoimmune thyroid disease from thyrotoxicosis that is due to other causes; diagnosis of euthyroid Graves disease; prediction of the likelihood that a pregnant woman who has autoimmune thyroid disease will deliver a neonate with thyrotoxicosis; helping to predict the likelihood of relapse or remission; and helping to predict if the thyrotoxicosis in the postpartum period will be transient or persistent.

TABLE 42-2. Thyroid-Stimulating Hormone Receptor Antibody Measurements

Other than the TSH measurement, which reflects thyroid action at the level of the hypothalamic-pituitary axis, there are presently no clinically available laboratory tests that directly and reliably reflect the effect of T4 or T3 on tissues (see Chap. 33). OTHER LABORATORY TESTS IN GRAVES DISEASE In long-standing Graves disease, anemia, which can be normochromic or mildly hypochromic, may be noted. Occasionally there is a mild neutropenia and lymphocytosis. Rarely, thrombocytopenia may occur. Commonly, the serum cholesterol level is decreased and there may be a minimal increase of the serum aspartate aminotransferase value. Perhaps 5% to 10% of patients with thyrotoxicosis manifest elevations in total serum calcium value.32 These elevations, which are due to increased bone resorption, are usually mild. The serum phosphate value is normal; the urinary hydroxyproline level is elevated. The hypercalcemia usually resolves as the serum T4 level decreases. Initially, b-adrenergic blockers may hasten the return to normocalcemia. Rarely, the serum calcium level remains elevated and is found to be associated with coexisting hyperparathyroidism. Similarly, the decreased serum magnesium, elevated serum alkaline phosphatase, and abnormal liver function tests should resolve as the thyrotoxicosis is controlled.

DIFFERENTIAL DIAGNOSIS OF HYPERTHYROIDISM

GRAVES DISEASE Graves disease occurs most frequently in young women (~5:1 female to male ratio) but is not uncommon in any age group or in any geographic population.1,2 and 3,5 No single clinical sign or symptom differentiates Graves disease from other causes of thyrotoxicosis, although the presence of ophthalmopathy, a diffusely enlarged and firm thyroid gland, and pretibial myxedema or acropachy strongly suggests this illness. The cooccurrence of another autoimmune disease in the patient (e.g., myasthenia gravis) or a family history of Hashimoto or Graves disease or of any other autoimmune disease is a frequent accompaniment of this condition. Graves disease is found more frequently in patients with HLA-DR3 antigens. The laboratory diagnosis is supported by the presence of stimulatory TSH receptor antibodies in a thyrotoxic patient.1,2 and 3,43 However, the measurement of TSH receptor antibodies is not indicated in most patients, as usually the diagnosis can be made on the basis of the previously listed clinical characteristics. Although rarely indicated, the thyroid gland may be imaged in such patients by a radioisotopic thyroid scan, CT, magnetic resonance imaging (MRI), or fluorescent iodide scan (Fig. 42-5, Fig. 42-6 and Fig. 42-7) (see Chap. 34, Chap. 35 and Chap. 39). The differential diagnosis of hyperthyroidism includes more than a dozen other conditions (Table 42-3).

FIGURE 42-5. Technetium thyroid scans (anterior views). A, Scan of a normal subject. Note the bilateral, relatively homogeneous tracer distribution (imaging time, 134 seconds). B, Scan illustrating a dominant hyperfunctioning right lower pole thyroid nodule (imaging time, 235 seconds). The nodule was not hyper-functioning sufficiently to completely suppress the remainder of the thyroid gland, but there was no serum thyroid-stimulating hormone rise after thyrotropin-releasing hormone stimulation, and clinically the patient was mildly thyrotoxic. C, Scan of a 50-year-old thyrotoxic patient with subacute thyroiditis. The 24-hour radioactive iodine uptake was <1%. The thyroid scan barely shows a faint outline of thyroid tissue (imaging time, 563 seconds). This patient had painful thyroiditis, a serum thyroxine level of 15.9 g/dL, a resin triiodothyronine uptake of 46%, an erythrocyte sedimentation rate of 32 mm/h, and no serum thyroid-stimulating hormone response to thyrotropin-releasing hormone stimulation.

FIGURE 42-6. A, Computed tomography scan of the neck in a patient with Graves disease. Note mildly enlarged thyroid gland that does not impinge on the subglottic trachea. Asymmetric jugular veins often are observed in normal subjects and are considered a nonpathologic variant. (C, carotid artery; J, external jugular vein; S, subglottic trachea; T, thyroid gland.) B, Technetium thyroid scan of the same patient showing bilateral, relatively homogeneous uptake. The 24-hour radioactive iodine uptake was 31% (imaging time, 152 seconds). C, Fluorescent thyroid scan in the same patient. The thyroid gland had a content of 20.8 mg iodine (normal, 515 mg).

FIGURE 42-7. A, Computed tomography scan showing a suprasellar pituitary tumor (arrow) in a thyrotoxic patient with a thyroid-stimulating hormonesecreting pituitary tumor. These tumors frequently secrete a subunit disproportionately to thyroid stimulating hormone. B, Magnetic resonance image (T1-weighted) in the same patient showing a 2 2-cm pituitary tumor (arrow). T1-weighted images are favorable for demonstrating anatomical detail. (Adapted from Smallridge RC. Thyrotropin-secreting

pituitary tumors. Endocrinol Metab Clin North Am 1987; 16:3.)

TABLE 42-3. Disorders That Can Cause Thyrotoxicosis*

SURREPTITIOUS INGESTION OF THYROXINE OR TRIIODOTHYRONINE Hyperthyroidism that is due to the administration of thyroid hormones is not uncommon, as exogenous thyroid hormone is easily obtained. This medication is often taken by persons who wish to lose weight. The surreptitious ingestion of T4 or T3 (thyrotoxicosis factitia) can be inferred from the measurement of inappropriately low levels of serum thyroglobulin. Because thyroglobulin is normally secreted by the thyroid gland in concert with T4 and T3, its levels should be elevated in cases of thyroidal hyperfunction, whereas its levels are suppressed with ingestion of exogenous thyroid hormones.44 In addition, the thyroid radioiodine uptake is depressed. If the patient is taking T4, serum values of this hormone are typically increased more than those of T3. If the patient is taking T3, the serum T3 level may be preferentially increased, with the T4 level being decreased. The extent of serum T3 changes in these patients depends on the dose ingested and how long after ingestion the blood is obtained. Thyroid extract ingestion may be associated with elevations of both T4 and/or T3, depending on the preparation used and the time interval since ingestion. TOXIC MULTINODULAR GOITER Toxic multinodular goiters that cause thyrotoxicosis are usually very large. Patients with thyrotoxicosis that is due to a multinodular gland do not undergo spontaneous exacerbations and remissions, and definitive therapy (131I or surgery) is usually required. Thyroid scans show diffuse inhomogeneous tracer uptake, reflecting areas of hyperfunction and hypofunction within the gland. Occasionally, CT or MRI of the neck may be helpful in delineating the size of a thyroid gland and in assessing impingement on surrounding structures (see Fig. 42-6). As stated in Chapter 37, the iodine contained in radiocontrast agents used with CT scans may cause thyroid dysfunction. On physical examination, the thyroid glands are multinodular and enlarged. The nodules usually are benign follicular adenomas. Toxic nodular goiters occur equally in men and women and can appear at any age, although they most frequently occur in patients older than 40 years. Usually, there is no exophthalmos, acropachy, onycholysis, or pretibial myxedema. Toxic nodular goiters are not believed to have an autoimmune etiology, because serum TSH receptor antibodies are absent. Antibodies that stimulate thyroid growth may be important in these disorders,45 but this finding is controversial. Low-titer antithyroglobulin or antimicrosomal antibodies may be present. Occasionally, eye findings may also be present. This presentation of stigmata of Graves disease with a nodular rather than diffuse goiter has been termed the Marine-Lenhart syndrome. SOLITARY AUTONOMOUS NODULE Patients who have a solitary autonomous nodule (toxic adenoma, Plummer disease) (see Fig. 42-5 and Chap. 39) as the cause of their thyrotoxicosis generally have a palpable thyroid nodule that, on thyroid scan, is seen to be suppressing the function of the extranodular tissue. However, solitary autonomous nodules usually do not secrete sufficient hormone to cause clinical thyrotoxicosis, although an undetectable T3 may be seen. Most autonomous nodules continue to function at the same rate over time, but 10% to 20% gradually enlarge, and their secretion is increased sufficiently to cause clinical thyrotoxicosis. Nodules more than 3 cm in diameter evolve to cause clinical thyrotoxicosis more frequently than do smaller nodules. Thyroid function tests often are only mildly abnormal. When hyperthyroidism occurs, therapy with 131 I or surgery is indicated. As in toxic multinodular goiter, surgery seems preferable in young adults and children, whereas radioactive iodine therapy is preferable in older patients. In general, 10 to 15 mCi of 131I is given to patients with autonomous nodules and 15 to 25 mCi of 131I to those with multinodular goiters. Because it is mainly the hyperactive tissue that traps the radioiodine and not the suppressed tissue, radioiodine therapy may not result in hypothyroidism, although these patients have to be observed indefinitely for the development of thyroid dysfunction. The differential diagnosis between toxic nodular goiter and Graves disease usually poses no difficulty, with the distinction resting on the anatomy of the thyroid by physical and scan examination and the presence or absence of other signs of Graves disease. In general, definitive therapy usually should not be administered while the patient is clinically thyrotoxic, particularly in older patients. Rather, antithyroid drugs should be given initially to render the patient euthyroid; these latter drugs do not cure the condition. EXOGENOUS IODINE EXCESS Exogenous iodine excess (e.g., iodide supplementation with radiocontrast agents or amiodarone) may induce hyperthyroidism in some individuals with autonomously functioning tissue; hence, this may occur in both Graves disease and nodular goiter (see Chap. 37). A history of iodide excess strongly suggests iodide-induced disease; and the radioactive iodine uptake may be suppressed because of dilution. A low radioactive iodine uptake may also be seen with subacute, silent, or postpartum thyroiditis, although in these circumstances it is thought to be due to impaired ability to trap iodine.5,46,47 and 48 A fluorescent iodine scan (see Chap. 39) may be useful in assessing intrathyroidal iodine content, which is usually increased in Graves disease and iodide-induced hyperthyroidism, and decreased in the hyperthyroidism of subacute thyroiditis (see Fig. 42-6C). SUBACUTE (GRANULOMATOUS) THYROIDITIS Subacute (granulomatous) thyroiditis is a spontaneously resolving inflammation of the thyroid gland that can have a varied course and clinical presentation (see Fig. 42-5C and Chap. 46). Viral infections of the thyroid gland are thought to cause or predispose to this disease (see Chap. 213).46 Although it may present at any age, women and men between the ages of 20 and 50 are most frequently affected. Often, the disorder evolves through a characteristic pattern, from thyrotoxicosis to hypothyroidism to the euthyroid state. Each phase may last ~2 to 10 weeks, and any or all of the clinical phases may be seen in individual patients. The distinction between this disorder and other causes of hyperthyroidism is usually not difficult (Table 42-4). The hyperthyroid phase may be accompanied by fever, pain in the thyroid area that radiates to the neck or jaw, myalgias, tachycardia, nervousness, and palpitations. The thyroid gland is usually enlarged, firm, and tender, and the erythrocyte sedimentation rate is usually elevated. Despite elevated serum T4 and T3 concentrations, the radioactive iodine uptake is very low (<5%). Because the thyrotoxic phase gradually resolves spontaneously, only temporary or palliative treatment is usually required. Some patients may require aspirin or corticosteroids to control the thyroidal pain, and the clinical hyperthyroidism can be treated with b-adrenergic blockers. Subacute thyroiditis may occur in otherwise normal thyroid glands or in patients with preexisting multinodular goiter or Hashimoto thyroiditis.

TABLE 42-4. Clinical and Laboratory Findings in Various Types of Thyroiditis

SPONTANEOUSLY RESOLVING THYROIDITIS In spontaneously resolving thyroiditis,46 the thyroid gland is usually painless and there is little evidence of acute inflammation. The thyrotoxicosis is mild; the 24-hour radioactive iodine uptake is low, and serum iodothyronine concentrations are increased. Thyroid glands demonstrate focal or diffuse lymphocytic infiltrations with varying fibrosis. There are no giant cells (which would be indicative of subacute granulomatous thyroiditis). The clinical course usually evolves from thyrotoxicosis to euthyroidism, with or without transient hypothyroidism. As with granulomatous thyroiditis, treatment is usually supportive, although control of the thyrotoxic signs or symptoms with b-adrenergic blockers is occasionally required. POSTPARTUM THYROIDITIS Postpartum thyroiditis occurs in 5% to 10% of women within the first year after parturition.47,49,50 This disease may evolve through a thyrotoxic, hypothyroid, and euthyroid recovery phase or may simply manifest with either hypothyroidism or thyrotoxicosis that spontaneously and gradually returns to euthyroidism. Approximately 67% of women with positive antimicrosomal thyroid antibodies develop thyroid dysfunction.49 Patients without detectable thyroid antibodies at delivery can also develop postpartum thyroiditis.50 Present knowledge suggests that pregnancy decreases immunologic responsiveness and that there is a rebound effect during the postpartum period; the thyroidal injury observed may be related to increased natural killer cell activity. Treatment of the thyrotoxic symptoms may not be necessary because of their mild elevation and evanescent pattern. Although the thyroid disease resolves spontaneously in many patients, the hypothyroidism or thyrotoxicosis may persist for more than a year after parturition, thus suggesting an unmasking of preexisting thyroid disease. When spontaneous remission does occur, the patient tends to have similar disease progression during each subsequent pregnancy. Moreover, evidence indicates that women with thyroid antibodies during pregnancy may develop gestational hypothyroidism; therefore, periodic monitoring of these women may be indicated.51 HASHITOXICOSIS Hashitoxicosis is a rare cause of hyperthyroidism in which the radioactive iodine uptake is typically increased or within the normal range, serum iodothyronine levels are elevated with an undetectable TSH, and the patient is clinically thyrotoxic. The pathologic findings in the thyroid resemble those of Hashimoto disease, with lymphocytic infiltration, fibrosis, and few or no enlarged follicles. Antimicrosomal and antithyroglobulin antibodies are present in high titer. It is believed that in some patients the condition evolves into hypothyroidism as the thyroid gland undergoes fibrosis and as destruction proceeds. TROPHOBLASTIC TUMORS Thyrotoxicosis that is usually mild, but can be severe, is seen in 10% to 20% of patients with trophoblastic tumors such as benign hydatidiform mole, malignant choriocarcinoma, or embryonal carcinoma of the testis. These tumors cause thyroidal hyperfunction by secretion of human chorionic gonadotropin (hCG), which may be capable of direct thyroid gland stimulation.48,52 There may be no goiter, or only mild thyroidal enlargement may be present. In some cases of choriocarcinoma, the original tumor may be small and difficult to find, but hCG levels in blood or urine are elevated. More patients have abnormal thyroid function tests than have clinical thyrotoxicosis. Treatment of the hyperthyroidism includes b-blockers, iodides, and, on occasion, antithyroidal agents. Therapy for the trophoblastic lesion is discussed in Chapter 111 and Chapter 112. STRUMA OVARII Struma ovarii denotes thyroid hormone secretion by embryonic rests of thyroidal tissue in the ovary (see Chap. 102). This extrathyroidal source of thyroid hormone is associated with relatively suppressed thyroid gland activity, manifested by a low 24-hour radioactive iodine uptake over the cervical area.53 Struma ovarii is extremely rare but should be considered in a thyrotoxic woman without a goiter and with a low radioactive iodine uptake in the neck. The diagnosis can be confirmed by pelvic scan with radioiodine. The primary treatment is surgical with restoration of euthyroidism, if possible, before surgery. THYROID-STIMULATING-HORMONESECRETING PITUITARY TUMOR TSH-secreting pituitary tumors are a very unusual cause of thyrotoxicosis (see Fig. 42-7).54,55 and 56 These tumors occur equally in men and women. The diagnosis should be considered if the thyroid gland regrows after an ablative procedure or in patients with thyrotoxicosis and headaches or visual disturbances. Unilateral proptosis, galactorrhea, amenorrhea, or associated acromegaly may also occur. In contrast to primary thyroidal hyperthyroidism, serum TSH levels are detectable or even elevated (see Chap. 15). A CT scan or MRI of the pituitary is often abnormal. Serum studies show a disproportionately elevated a subunit of TSH. TSH is poorly responsive to TRH stimulation or to T4 suppression. Although rare, the diagnosis should be considered in the previously mentioned circumstances, because the mass effect of the pituitary tumor can be damaging. The treatment is pituitary surgery, irradiation, or both. External radiation alone may be helpful but is not considered curative. Studies suggest that somatostatin analogs have a role in therapy.56 Pituitary tumors that secrete TSH may also secrete other pituitary hormones (growth hormone in particular); therefore, such patients should be screened appropriately. In a report summarizing the experience of the National Institutes of Health,55 in 25 patients with a TSH-secreting pituitary tumor, thyrotoxic symptoms were present in 22 subjects and absent in 3. In 14 patients previously untreated for their pituitary tumor, an elevated a subunit to TSH molar ratio was present in 12 cases, giving 83% sensitivity. Considering all cases, apparent cure was noted in 35% of patients with surgery alone and in an additional 22% with a combined approach. Three deaths were noted, including one with metastatic pituitary cancer. Five of six patients with residual tumor had their hyperthyroidism controlled with octreotide therapy. METASTATIC THYROID CANCER Metastatic thyroid cancer must be widespread before there is sufficient mass to secrete enough iodothyronine to cause hyperthyroidism.57 Follicular carcinoma is the tumor type most likely to produce this syndrome. The diagnosis of this entity derives from finding extensive radioactive iodine uptake outside the thyroid bed on thyroid scan and elevated serum thyroglobulin levels. Usually, there is a history of known thyroid cancer. Treatment consists of 131I administration and the temporary administration of antithyroid drugs and b-blockers. Of interest, tumors (e.g., lymphoma, pancreas, breast, malignant thymoma) that are metastatic to the thyroid gland can also cause hyperthyroidism, probably by local injury and release of cytokines. HAMBURGER THYROTOXICOSIS Hamburger thyrotoxicosis is a fascinating syndrome that was described by astute investigators who noted a regional outbreak of hyperthyroidism. Further study revealed that local abattoirs were including thyroid glands with the other meats obtained from cattle. The iodothyronines in the meat were ingested by individuals, who then exhibited elevated serum T4 and T3 levels and clinical thyrotoxicosis. The limitation of this entity to one geographic area was related to the meat distribution pattern from one local abattoir.58 The clinical manifestations included thyrotoxicosis and elevated serum T4 and T3 levels but low radioactive iodine uptake, similar to what is seen with surreptitious ingestion of thyroid hormone. Although inclusion of the thyroid gland in meat products is now banned in several states, this entity should be considered when patients present with elevated serum iodothyronine levels and a low radioactive iodine uptake. Furthermore, it is possible that similar geographic outbreaks have occurred before and may have accounted for some previously reported cases of spontaneously resolving thyroiditis.

NATURAL HISTORY OF GRAVES DISEASE

Although our understanding of the pathogenesis of Graves disease is progressing rapidly, we still cannot accurately predict the natural history in a given patient. Virtually all patients with Graves disease can be treated effectively with antithyroid drugs, 131I, or surgery. Most patients treated with 131I or surgery eventually become hypothyroid and require L -T4 therapy. Therefore, if a patient does not become endogenously hypothyroid within the first several months after radioactive iodine therapy, he or she will need clinical and laboratory examination once or twice yearly thereafter. Although most patients who do manifest radioiodine-induced hypothyroidism have no recurrences of thyrotoxicosis, an occasional patient with documented hypothyroidism may again develop thyrotoxic Graves disease and require definitive antithyroid therapy. These patients probably have heterogeneous anti-TSH receptor antibodies in their serum, with earlier predominance of the blocking antibodies,40 after which the character or affinity of the antibodies changes to predominantly stimulating antibodies. Patients with stimulating TSH receptor antibodies in high titer or affinity may have particularly aggressive disease, which may include resistance of thyrotoxicosis to two or three standard doses of 131I therapy or recurrence after a near-total thyroidectomy. Some insight into the natural history of Graves disease came from a study of 15 patients who had undergone a course of antithyroid medication with restoration of the euthyroid status 20 to 27 years earlier and who had no subsequent therapy with antithyroid or thyroid medications.59 Four patients had become hypothyroid, and one developed recurrent thyrotoxicosis 25 years after the initial diagnosis; 5 of 10 patients had an abnormal iodide perchlorate discharge test, and 12 of 15 had elevated antimicrosomal antibody titers. These results suggest that patients with Graves disease require medical follow-up throughout their lives.

TREATMENT OF GRAVES DISEASE

The three major modalities of treatment for a thyrotoxic patient with Graves disease are antithyroid medications, 131I, and thyroidectomy. ANTITHYROID DRUGS Either propylthiouracil (PTU) or methimazole (methylmercaptoimidazole, MMI) is considered a first-line agent in the treatment of thyrotoxic Graves disease (Table 42-5).60,61,62,63 and 64 PTU and MMI inhibit iodination of thyroglobulin, iodotyrosine coupling, thyroglobulin synthesis, and lymphocyte in vitro responsiveness. PTU, but not MMI, inhibits conversion of T4 to T3.62 Other possible actions of MMI, which have been suggested for PTU, include inhibition of synthesis of anti-TSH receptor antibodies and inhibition of thyroglobulin-primed antigen-presenting cells.62

TABLE 42-5. Pharmacology of Agents Used to Treat Hyperthyroidism*

Before PTU or MMI therapy is begun, the radioactive iodine uptake should be measured; this can also be accomplished during the first week of therapy if the patient's clinical status requires immediate treatment. PTU and MMI are well absorbed by the gastrointestinal tract. Both drugs are immediately concentrated by the thyroid gland and have a more prolonged biologic effect than indicated by their serum half-lives (PTU, 1 hour; MMI, 5 hours). PTU should be administered 2 or 3 times daily, whereas MMI can be given once daily. The customary starting dose of PTU in a slightly thyrotoxic subject may be 100 to 300 mg per day in divided doses; in a moderately thyrotoxic patient, 300 to 800 mg per day; and in a very thyrotoxic patient, 800 to as high as 1200 mg per day. Patients should be examined periodically to determine the effect of the medications. Serum T4 or free T4 should also be measured, and the dose of PTU or MMI is tapered gradually as the serum hormone levels fall. The aim is to restore the euthyroid state within 1 to 2 months. Because a thyrotoxic patient may have rich hormonal stores in the thyroid gland that continue to be secreted during antithyroid therapy, and because the T4 half-life in these patients is ~5 days, even maximal doses of MMI or PTU may not restore euthyroidism quickly, so 4 to 8 weeks may be required for the serum T4 and T3 levels to normalize. The pathophysiology of Graves disease and treatment options are discussed with each patient. It is stressed that antithyroid agents help treat the thyrotoxic state but usually are not curative and that surgery or 131I may be required.64 A complete blood cell count and chemistry profile (particularly calcium and liver function tests) may be obtained initially and perhaps every 1 to 2 months during treatment, although agranulocytosis or drug-related hepatitis cannot be accurately predicted by these tests. All patients are informed of the potentially adverse effects of PTU or MMI (Table 42-6)62 and are cautioned that a fever, rash, urticaria, arthralgia, or sore throat should be reported. Agranulocytosis occurs in ~1 of every 200 patients treated with PTU or MMI; most frequently it develops in the first several months of drug administration in patients older than age 40 and in patients taking larger doses of PTU or MMI (e.g., >400 mg per day or 40 mg per day, respectively). If there is any question that PTU or MMI may be causing an adverse effect, the individual drug should be discontinued and the patient either observed without medication for a short time or the alternative agent substituted if needed. If there was a serious adverse reaction to one agent, the alternative agent should probably not be used. There is an estimated 20% to 50% likelihood that a patient with a reaction to one agent (e.g., PTU) will have an adverse reaction to the other (i.e., MMI). There has been interest in the use of recombinant human granulocyte colony-stimulating factor in the treatment of antithyroid agent-induced agranulocytosis.64,64a In a review of generalized reactions to PTU or MMI, fewer than 100 cases had been reported since 1945.65 The most common, generalized reactions included vasculitis, lupus erythematosus, and polyarthritis. Reactions were more common with PTU than with MMI, but occasionally, even with discontinuation of the medication, fatality can occur. Cross-reactions to PTU and MMI may occur.

TABLE 42-6. Potential Side Effects of Antithyroid Medications

Most patients tolerate PTU or MMI, and therapy is continued as biochemical and clinical euthyroidism returns. At this time a decision is made whether to continue the antithyroid medications (Fig. 42-8). Some physicians maintain patients on PTU or MMI for ~1 year in an effort to induce a remission (Table 42-7 and Table 42-8). In the United States, it is likely that only 20% to 40% of patients will have a remission; a small goiter size, response to low doses of PTU or MMI, mild thyrotoxicosis at presentation, and perhaps low-titer TSH receptor antibodies and lack of HLA-Dw3 and B8 antigens all increase the likelihood of remission (Table 42-9). Different

geographic areas with variable iodine intake may be associated with different remission rates. In one multicenter study, the efficiency of 40 mg versus 10 mg of MMI was compared in the treatment of Graves disease.60 In this study, 196 of 309 (63%) patients achieved a remission, with the differences in the two groups not being significantly different. Adverse reactions occurred in 16% of the group taking 10 mg and in 26% of the group taking 40 mg, with ~0.8% having a serious hematologic abnormality.

FIGURE 42-8. Upper panel: Patient with Graves disease treated with antithyroid drug. Left: Appearance before therapy. Right: Four months after commencement of therapy. Note the markedly decreased stare and the weight gain. Lower panel: Patient with Graves disease treated with radioiodine. Left: Before therapy. Right: Six months later.

TABLE 42-7. Indications for the Different Therapies of Graves Disease

TABLE 42-8. Complications of Therapies for Graves Disease

TABLE 42-9. Long-Term Antithyroid Medication as a Treatment for Graves Disease

b-BLOCKERS AND OTHER AGENTS When a thyrotoxic patient is symptomatic, a b-adrenergic blocker is often prescribed at the same time PTU or MMI therapy is initiated (see Table 42-5).66 b-Blockers improve many of the symptoms of thyrotoxicosis, improve myocardial efficiency, reduce myocardial oxygen demand, and may decrease nitrogen loss. It may be advisable to recommend that a hyperthyroid patient receiving a b-blocker record his or her pulse several times daily. The time-honored b-blocker is propranolol, and the oral dose range is 20 to 40 mg every 8 hours for mild to moderate thyrotoxicosis and as high as 60 to 80 mg four times a day for more severe cases. The therapeutic objective is subjective improvement and the restoration of the pulse rate to ~80 beats per minute. Newer b-blocking agents have the advantages of cardioselectivity and longer duration of action. A thyrotoxic patient responds very rapidly to the administration of b-blockers. If no clinical or subjective response is noted in 2 to 3 days, the dose should be increased. b-Blockers have no direct effect on radioactive iodine uptake or the synthesis or secretion of T4 or T3. These agents should be given with caution, if at all, to patients with a history of asthma. High-output cardiac failure may be improved by b-blockers, but their use in such patients should be judicious. It is usually not prudent to use long-term b-blockers as the sole therapy in the treatment of Graves disease. Because b-blockers do not alter T4 or T3 synthesis or secretion, the underlying thyrotoxicosis may worsen while a patient is taking b-blockers alone. If a patient is allergic to both PTU and MMI, if the thyrotoxicosis has not responded appropriately, or if the thyrotoxicosis requires more rapid restoration of the euthyroid state, alternative agents such as ipodate, other iodides, or lithium carbonate should be considered (see Table 42-5).64,67,68 and 69 Patients taking one of these latter antithyroid agents usually need closer medical supervision and possibly hospitalization. IODIDES Iodides may be very effective antithyroid agents but should be used in thyrotoxicosis only when specific indications exist.70 Iodides should usually not be used as customary or sole therapy because there can be an escape from their inhibitory effects, and the resultant exacerbation in thyrotoxicosis may be extremely difficult to

control. For this reason, iodides are typically used in combination with thioureas, and therapy is not instituted until a patient has been started on PTU or MMI. Iodides can decrease serum T4 and T3 levels by 30% to 50% in 10 days; iodide and lithium may have an additive effect on decreasing serum T4.68,69 A thorough medical history, with specific inquiry as to allergies to iodides, should be taken before initiating therapy with ipodate or other iodides. The doses of iodides given with these agents are enormous when compared with the minimal daily requirement, and the consequent suppression of radioactive iodine uptake compromises the ability to use radioiodine for either diagnostic studies or therapy. Data suggest that 131I can probably be used effectively again as early as 4 to 6 weeks after ipodate is discontinued,67 but this area requires further investigation. Iodide (or ipodate) does not interfere technically with the measurement of serum iodothyronines or TSH. Iodides, ipodate, and lithium are especially useful in lowering the serum T4 and T3 levels, and thus decrease the chance of thyroid storm when a thyrotoxic patient is being prepared for surgery and there is insufficient time to administer PTU or MMI alone for several weeks. Of the three agents noted, ipodate is a good choice in nonpregnant individuals; lithium therapy requires extremely close observation with frequent serum monitoring of electrolytes, liver function tests, and lithium levels. Ipodate is an iodide-containing radiocontrast agent that is very effective in the treatment of thyrotoxicosis.67 The ipodate molecule inhibits extrathyroidal T4 to T3 conversion, and a dose of 0.5 to 3 g decreases the serum T3 level by 50% within several days. The iodide released from ipodate may help decrease thyroidal secretion when used for short periods (i.e., <1 month). Ipodate has been shown to be safe, with few adverse effects; it may be effective for as long as 6 months; however, its prolonged use in this manner requires a careful assessment. Escape from the antithyroid effects of both ipodate and other iodides may occur, especially after chronic administration. Although the use of ipodate in conjunction with PTU or MMI and b-blockers is gaining acceptance, the drug has not yet been approved by the U.S. Food and Drug Administration for this indication. In a typical nonpregnant patient before thyroid surgery, 0.5 to 1.0 g of ipodate on preoperative days 10, 7, 4, and 1 can be recommended. Serum complete blood count, liver function tests, and thyroid function tests should be measured frequently in a patient receiving ipodate. LITHIUM CARBONATE Lithium carbonate is another second-line agent that can be effective in the treatment of thyrotoxicosis.68,69 Commonly, the drug is first used in hospitalized patients to closely monitor serum lithium levels (therapeutic range, 0.5 to 1.2 mEq/L). Lithium acts on the thyroid gland to decrease the secretion of iodothyronines, without a major effect on iodide uptake. The frequency of adverse effects, such as ataxia and drowsiness, is much reduced when serum levels are maintained within the therapeutic range. The customary antithyroid effect of lithium is seen with doses of 600 to 900 mg per day. Lithium carbonate lowers serum T4 levels by 30% to 50% in ~10 days. Paradoxically, there have been cases of lithium-associated thyrotoxicosis.71 This potential untoward effect of lithium may be related to increases in thyroidal iodine and appears to be less likely when lithium is administered for <1 month. RADIOACTIVE IODINE THERAPY Iodine-131 has been used to treat Graves disease since the 1940s, and its use has enjoyed increasing popularity because of its efficacy and few side effects (see Table 42-7 and Fig. 42-8).72 Follow-up studies have not implicated 131I therapy in a higher risk of carcinoma in general, leukemia, or lymphoma.73 Therapy with 131I does increase the gonadal radiation exposure, but a higher risk of fetal malformation in subsequent pregnancies has not been demonstrated. Nevertheless, 131I therapy should be administered with caution to any female patient of child-bearing age. A retrospective collaborative study analyzed ~35,000 hyperthyroid patients treated with radioiodine between 1946 and 1964.74 Radioiodine therapy and hyperthyroidism itself were not associated with increased cancer mortality. However, thyroid cancer mortality was slightly increased in patients receiving radioiodine. The absolute number of excess deaths was slight, but was statistically significant, throughout the duration of the study. This study can be criticized, however, for several reasons: It is retrospective; patients were not randomly allocated to a given treatment group; initial and follow-up evaluation did not include sonograms, aspirations, or reports of T4 and TSH; many patients required two doses of radioiodine; it cannot be discerned which patients became hypothyroid and required L -T4 therapy; and the histologic diagnosis of thyroid cancer was determined from death certificates rather than from examination of the original slides. It also is conceivable that the risk of thyroid cancer occurring would be decreased if the patients had become hypothyroid within several months after radioiodine therapy. Lastly, the age at which patients received radioiodine and subsequently had thyroid cancer cannot be determined. Another study5 retrospectively analyzed 7209 thyrotoxic patients and noted that the standardized mortality rate was increased slightly, but significantly, for deaths from thyroid disease, cerebrovascular disease, and fracture of the femur. The increased mortality was observed mainly within the initial year after radioiodine therapy. It is possible that these early deaths were related to lack of adequate control of the hyperthyroidism and follow-up issues rather than the radioiodine itself; furthermore, the control group to which the thyrotoxic patients were compared can be questioned. When these two studies are considered in context with earlier studies,5,73,74 it still seems that radioiodine therapy is a relatively safe and effective therapy for Graves disease, and these two studies5,74 do not directly influence the choices of therapy. It is important, however, to discuss these issues and review these reports with patients and to emphasize that further studies in these areas would be useful. A thorough menstrual history and a serum b-hCG level should be obtained to rule out pregnancy just before a 131I dose. 131I crosses the placenta and may damage fetal thyroid gland development. Potential harmful effects of 131I depend on the dose and the time during pregnancy when it is given. Because the fetal thyroid develops during weeks 10 to 15, there is little 131I trapping before that time. Rarely, the issue of how to treat a pregnant woman who inadvertently received radioactive iodine therapy during pregnancy arises. The intraamniotic administration of L -T4 to treat the potentially hypothyroid fetus has been suggested;75 however, this is not approved by the Food and Drug Administration. In my clinic, it is recommended that a woman avoid becoming pregnant for 6 to 12 months after a therapeutic dose of radioactive iodine. In the treatment of Graves disease, the dose of 131I can be calculated by a variety of techniques. In one method, the product of the thyroid weight and the customary dose rate (80 to 120 Ci/g) is divided by the radioactive iodine uptake. For example, a patient with a 60-g goiter and a 60% radioactive iodine uptake would receive 8 to 12 mCi. For calculation of the dose, the author prefers to measure the radioactive iodine uptake 1 to 2 weeks before therapy; on an empirical basis, a higher dose is used if it is thought necessary to ensure a cure of the hyperthyroidism. In the author's hands, a typical patient with Graves disease would perhaps receive 10 to 15 mCi radioactive iodine. PTU or MMI should be discontinued for several days before and after 131I therapy.76 Probably, prior antithyroid agent administration confers relative resistance to a given dose of radioactive iodine; therefore, a relatively higher dose of radioactive iodine is used in these circumstances. Given that Graves hyperthyroidism is mediated by TSH receptor antibodies that are largely unaltered by radioactive iodine administration, the author believes that the optimal and desired goal of 131I therapy is the development of hypothyroidism as rapidly as possible. Thereafter, the patient is given L -T4 therapy and observed on a lifelong basis. The decision to treat a hyperthyroid patient with radioactive iodine early in the disease course (i.e., within several weeks of diagnosis) or to treat the patient initially with several weeks or months of antithyroid agents before radioactive iodine therapy to render the patient euthyroid is a difficult decision. The clinical context, severity of hyperthyroidism, extent of iodo-thyronine elevation, and presence of associated medical and endocrine conditions (e.g., cardiac arrhythmias, ophthalmopathy) are relevant to this decision. In the author's practice, ~70% of patients are treated with radioactive iodine early in their course before the administration of antithyroid agents. The patientand frequently, the familyshould participate in this decision. It should be remembered that hyperthyroid patients may not make decisions in as reasonable a manner as they would if they were euthyroid. therapy may aggravate uncontrolled thyrotoxicosis, can precipitate thyroid storm, and can exacerbate ophthalmopathy (especially if moderate or severe).77,78 The serum T4 level is rarely increased significantly by 131I therapy, but one cannot predict which patients will have a marked T4 rise.76 Radioiodine may make ophthalmopathy worse, particularly if the serum T3 level is increased.78 A study of 443 patients with Graves hyperthyroidism (slight or no ophthalmopathy) randomized them to receive radioiodine, to receive radioiodine followed by a 90-day course of prednisone, or to receive MMI for 18 months.78 After 6 months of radioiodine treatment, 15% of 150 patients developed or had exacerbated ophthalmopathy (transient in 15/23), and none of 55 improved. Of 145 patients treated with radioiodine and prednisone, 75% of those with baseline ophthalmopathy (67%) had improvement; none had progression. Of 148 patients treated with MMI, 141 had no change in their ophthalmopathy, whereas 3 patients (2%) had improvement in baseline ophthalmopathy and 4 patients (3%) had exacerbation. In summary, this prospective study suggests that radioiodine is associated with the presence or exacerbation of ophthalmopathy more frequently than is MMI therapy. Nevertheless, the frequency of the worsening of ophthalmopathy with radioiodine therapy is ~15%, only being persistent ~33% of the time. Prednisone essentially prevented the radioiodine-induced worsening of ophthalmopathy. A patient with moderate or severe thyrotoxic Graves disease is typically treated with PTU or MMI and b-blockers for 1 to 2 months before treatment with 131I is initiated. During this period, the patient is seen frequently, and the other potential therapeutic modalities are discussed, with special emphasis on the potential risks and benefits of 131I treatment. Once treatment with 131I has been agreed to, the PTU or MMI (but not the b-blocker) is discontinued 2 to 3 days before therapy. A negative b-hCG value is required in women of childbearing potential. After informed consent has been obtained and a 131I uptake study has been performed, the patient receives 131I therapy and is cautioned to return if an exacerbation of symptoms is perceived. Patients with persistent mild symptoms are continued on a b-blocker until they become euthyroid. In more symptomatic patients, in patients with potentially serious concomitant medical disorders, and in those with a labile condition in whom hyperthyroxinemia is likely to return before maximal radioiodine effect is achieved, PTU or MMI is restarted 2 to 3 days after 131I therapy. After 131I therapy, patients are examined, and serum T 4 and T3 levels are determined periodically. Based on the progressive improvement, indicated by the results of the physical examination and thyroid function tests, the PTU or MMI dosage is gradually tapered and discontinued, usually within 2 months after the 131I dose. It is controversial whether patients with Graves disease with moderate or severe ophthalmopathy should receive corticosteroids to minimize the possible adverse effect of 131 I on ophthalmopathy.29,30,78 The author discusses the issue with the patient, considers the potential risks of corticosteroid therapy, and analyzes the severity of the ophthalmopathy. Patients are frequently referred to an ophthalmologist who is familiar with thyroid-related eye disease, and medical and/or surgical considerations and decisions are discussed. When appropriate (generally including patients with moderate or severe ophthalmopathy), the author considers treating individual patients with 40 mg prednisone for 7 days before 131I therapy, continuing this dose of prednisone for 14 to 28 days after therapy, and then tapering the dose with the purpose of
131I

discontinuing the medication within 4 to 8 weeks after radioiodine administration. The patients are followed up closely after the discontinuation of PTU or MMI at least 3 to 4 months after the radioiodine administration, depending on the laboratory and clinical evaluation. An increase in serum T4 and T3 levels to above normal (after PTU or MMI discontinuation) indicates that another dose of radioiodine may be required, although, generally, the second dose is not administered until at least 6 months have elapsed since the initial dose. In this circumstance, the TSH value must be interpreted in conjunction with the serum iodothyronine levels and the clinical context. Occasionally, the hypothalamic pituitary axis, as manifest by serum TSH, may not respond appropriately to the serum iodothyronine levels in patients who have had Graves hyperthyroidism within the recent past. If the serum T4 level decreases to below normal and the serum TSH is elevated, L -T4 therapy is initiated. Approximately 80% of patients with Graves disease become euthyroid or hypothyroid after one dose of 131I; 15% require two doses of 131I, and 5% require three or more doses. Patients who remain biochemically and clinically euthyroid after 131I therapy are observed for life with thyroid function studies and are advised that hypothyroidism or recurrent hyperthyroidism could occur at any time. 131I therapy is generally considered for persons older than 20 years of age. However, many factors enter into the decision for a choice of therapy, and it is most important that each patient be counseled about the options and risks so that he or she can participate in the treatment decisions. Patients should be encouraged to obtain information regarding Graves disease from a variety of sources, especially Internet sites sponsored by medical organizations. Although hypothyroidism is an expected and even desirable result of 131I treatment, some physicians prefer to administer several small doses of 131I to achieve euthyroidism and avoid hypothyroidism. The author does not subscribe to this approach. 131I is the isotope most commonly used to treat Graves disease. After radiotherapy, patients are counseled not to have close contact with family members for ~7 days and, because of salivary radioiodine, should not share eating or drinking utensils. Particular care should be exercised if there is a pregnant woman or a child in the household. For purposes of comparison, a typical dose of 131I for therapy for Graves disease is 5 to 15 mCi, but the dosage for therapy for thyroid cancer is often ~100 to 150 mCi. The Nuclear Regulatory Commission guidelines have indicated that a patient given a dose of 131I >30 mCi usually should be hospitalized, whereas lesser doses may be administered to outpatients. SURGERY Although surgery is considered primary definitive therapy by fewer physicians now than in the past, it still is important in the treatment of some patients with Graves disease (see Chap. 44).79 Surgery, rather than 131I therapy, should be considered in a patient of any age if any aspect of the case suggests possible malignancy, such as a palpable thyroid nodule, cervical lymphadenopathy, hoarseness (recurrent laryngeal nerve involvement), pain or tenderness, or rapid growth of the goiter. There is a question as to whether thyroid cancer that does occur in conjunction with Graves disease has a more aggressive course (e.g., increased frequency of local invasion and metastases to lymph nodes and distant sites) as compared with thyroid cancer in a patient without Graves disease.80 Severe or advancing ophthalmopathy may also favor surgery, because some studies suggest that 131I (especially in high doses) may aggravate eye findings. Also, surgery lowers serum T4 and T3 values to normal more quickly than does 131I therapy. Surgery may also be the best choice in a pregnant patient with hyperthyroidism that is difficult to control and in whom 131I therapy is contraindicated. Specific thyroid nodules, detected by physical examination or sonography, or when suggested by radioisotope scan, should usually be aspirated before a decision regarding therapy for the hyperthyroidism, and the results must, obviously, be considered an integral part of the therapeutic decision. Patients with Graves disease who are being considered for surgery should usually be rendered euthyroid with PTU or MMI before surgery is undertaken. Iodides (e.g., saturated solution of potassium iodide or Lugol's solution or ipodate) are generally not required except in moderate to severe cases, although some authorities administer them in a routine manner before surgery (assuming no contraindications) to decrease thyroidal iodothyronine stores and secretion and to decrease gland vascularity (see Chap. 44). Only an experienced surgeon and anesthesiologist should perform or participate in the surgery. A near-total thyroidectomy, leaving 5 to 10 g of thyroid tissue, is usually performed, with care taken to preserve parathyroid and recurrent laryngeal nerve integrity. There is high likelihood that a patient will become hypothyroid after this procedure. Other complications are listed in Table 42-8. Transient laryngeal nerve palsy may occur in ~5% of patients, but <1% of patients should sustain permanent damage. The usually transient hypocalcemia after surgery may not necessarily be due to parathyroid compromise. The syndrome of bone hunger, or increased calcium removal from serum with increased deposition in bone,81 is often short-lived but may persist for 6 to 12 months after surgery; calcium supplements may be needed during this time. A prospective analysis comparing the treatment modalities for 174 Graves disease patients82 showed that patients were generally pleased with whichever therapy they received, and, of interest, the cost estimate for medical therapy (including costs of relapse) was $2284. Surgery and radioactive iodine therapy were each ~1.5 times as expensive as medical therapy.

GRAVES DISEASE IN SPECIAL SITUATIONS

PREGNANCY AND GRAVES DISEASE The diagnosis of thyrotoxic Graves disease in a pregnant patient is especially difficult, because euthyroid pregnant women may share many signs and symptoms considered indicative of hyperthyroidism, including nervousness, irritability, warm skin, hand tremor, palmar erythema, diaphoresis, goiter, widened pulse pressure, and fatigue.83 No signs or symptoms are absolutely discriminatory of the thyrotoxic state, but clinical emphasis should be placed on (a) an inability to gain weight normally while pregnant; (b) a firm thyroid gland weighing more than 40 g and which may demonstrate a bruit; (c) a prior history of Graves disease; and (d) ophthalmopathy with conjunctivitis, chemosis, or proptosis. The total serum T4 level rarely increases to higher than 20 g/dL in euthyroid pregnant women, and the serum free T4 level should not be elevated. TSH receptor antibodies should not be detectable in a euthyroid pregnant woman. The proper interpretation depends on whether binding inhibition or stimulatory assays are performed. Stimulatory TSH receptor antibodies are more supportive of the diagnosis of thyrotoxic Graves disease, whereas positive binding inhibition assays may indicate Graves disease or autoimmune thyroid disease but not necessarily hyperthyroidism. The laboratory evaluation of a pregnant woman should not entail exposure to isotopes or radiation. The use of third-generation TSH assays preclude the necessity of performing a TRH assay in most patients. Moderate to severe thyrotoxic Graves disease in a pregnant woman requires treatment. If it is difficult to distinguish between euthyroidism or mild thyrotoxicosis, the decision to treat with antithyroid medications is more difficult, because mild thyrotoxicosis in some instances may pose less of a risk to the mother and fetus than the administration of antithyroid medications.84,85 Thyrotoxicosis often improves during pregnancy, and antithyroid medication dosages may be less than those required during the nonpregnant state. Occasionally, however, thyrotoxicosis may be severe and difficult to control. It is important to render these women euthyroid, but it is also important not to administer very large PTU or MMI doses, because the antithyroid medications cross the placenta and can cause fetal goiter or hypothyroidism. The author believes there is greater confidence and experience with the use of PTU than with MMI during pregnancy and so consider it the agent of choice. Furthermore, MMI has been reported to cause aplasia cutis, a scalp condition in neonates. The infant's thyroid function should be checked immediately after birth and probably again in 1 to 2 months. Although some complications occur, in general, antithyroid agents do not appear to be highly teratogenic.85,86 and 87 PTU doses of <300 mg per day and MMI doses of <30 mg per day rarely seem to have adverse effects on the fetus. In one study, the neonatal malformation rate was higher in thyrotoxic mothers than in euthyroid mothers, and the restoration of maternal euthyroidism with the use of antithyroid medication had a beneficial effect in decreasing the frequency of the malformations.85 There was no correlation between MMI dose and the frequency or type of malformation. Thus, uncontrolled hyperthyroidism in the mother may be a more important factor in causing malformations than the antithyroid drugs themselves. It is believed the pregnancy itself may decrease the immunologic responses involved in causing Graves disease, and, commonly, improvement or even remission occurs during pregnancy. Unfortunately, however, after delivery the Graves disease often relapses. The therapeutic goal is clinical euthyroidism, a serum T4 level within the normal range for pregnancy (i.e., 10 to 16 g/dL),84 and a normal free T4. It is preferable to monitor free T4 and T3 when observing a hyperthyroid pregnant patient. On the other hand, if larger doses of PTU or MMI are necessary to restore the euthyroid state, or if the condition cannot be well controlled with antithyroid medications, surgical thyroidectomy then becomes a treatment option and is optimally performed in the second trimester.79,83,88 A patient with uncontrolled or poorly controlled thyrotoxicosis should usually be monitored closely for 1 to 2 weeks before thyroidectomy to ensure compliance with medications. Because of the brief period involved, daily doses of PTU of as much as 600 mg (or 60 mg of MMI) may need to be given for a very short time preoperatively. The addition of stable iodine for 7 days or less may be necessary if the thyrotoxicosis is extremely poorly controlled. Extreme caution must be exercised, because iodides, especially when given for longer periods, may cause massive enlargement of the fetal thyroid gland, resulting rarely in asphyxiation. Management of a pregnant Graves disease patient who is severely hyperthyroid should be performed by physicians familiar with this condition. Similarly, thyroidectomy should not generally be undertaken unless an experienced surgeon and anesthesiologist are available and the patient is in nearly a euthyroid condition (see also Chap. 110). When the use of a b-adrenergic blocker is deemed necessary in a pregnant patient with Graves disease, propranolol is preferred because of the extensive experience with this agent.83,84,86,87,88 and 89 Although propranolol is generally safe, intrauterine growth retardation and fetal bradycardia have been associated with its use. Hence,

propranolol should be reserved for severely thyrotoxic pregnant subjects and used judiciously, with careful monitoring of the patient and fetus. The measurement of serum TSH receptor antibody is indicated in most pregnant patients with Graves disease, with or without active thyrotoxicosis. A maternal level of stimulatory TSH receptor antibody more than five times higher than control in the third trimester accurately predicts neonatal thyrotoxicosis.88 If the titer is very high, the fetus is monitored closely, and the neonate is carefully examined at birth, with determination of serum iodothyronines and TSH levels. An apparently euthyroid infant should nevertheless be monitored closely for the development of thyrotoxicosis during the first 3 months of life (see Chap. 47). There is no safe and effective method of treating a thyrotoxic fetus in utero, although this has been done successfully in one reported case.89 It is controversial whether postpartum women with Graves disease who are receiving antithyroid drugs should breast-feed their infants, because the potential harm to the neonates' thyroid gland of such therapy is uncertain. PTU may be preferred over MMI in breast-feeding women.62 Because of its known effect on the neonatal thyroid gland, administration of 131I, whether therapeutic or diagnostic, should be avoided in the nursing woman. If radioiodine therapy must be administered, breast-feeding should be avoided for 3 or 4 months. If a diagnostic radioisotope study must be performed, interruption of the nursing may be done for the following periods: 7 to 10 Ci 131I, 3 to 4 weeks; 50 Ci 131I, 7 to 8 weeks; 10 mCi technetium-99m, 24 hours; 100 Ci 123I, 3 to 4 weeks.90 Elevated hCG levels may increase thyroid hormone secretion during pregnancy and, as a result, hyperemesis gravidarum may be associated with elevated serum free T4 and free T3 with a suppressed TSH.91 This entity is difficult to diagnose, and some authorities recommend short-term antithyroid therapy if the signs, symptoms, and thyroid test results are markedly abnormal. Antithyroid agents for this condition of transient hyperthyroxinemia are seldom indicated. As the hCG levels decline, this condition also tends to resolve. THYROID STORM Thyroid storm is a life-threatening condition in which the customary signs and symptoms of thyrotoxic Graves disease are exaggerated; many patients also have vomiting, diarrhea, dehydration, fever, disorientation or impaired mentation, and, rarely, jaundice.92 Frequently, they have severe tachycardia and may manifest heart failure. The extreme restlessness and adrenergic hyperactivity may proceed to mania. Coma may occur. Thyroid storm is not associated with disproportionately elevated serum total iodothyronine levels compared with other thyrotoxic patients. Thus, it is believed that other factorssuch as increased receptor occupancy or increased free and tissue iodothyronine levelsmust contribute to this condition. Thyroid storm should not be viewed as an entity distinct from thyrotoxicosis but as one end of a continuum of severity of hyperthyroidism. Indeed, authorities may disagree as to what the specific definition or clinical requirements for thyroid storm may be. Although thyroid storm may occur without a known precipitating event, it most frequently occurs during or after thyroidal or nonthyroidal surgery; in the peripartum or postpartum period; after 131I therapy; after administration of iodine-containing materials; or after infections, such as pharyngitis or pneumonitis. The best treatment is prophylactic. Thyroid storm is rare if a patient has received adequate therapy with PTU or MMI before known precipitating events, such as surgery or delivery of a neonate. Because the outcome may be so serious, patients thought to have thyroid storm should usually be treated as if they do have the condition. The diagnosis is particularly difficult to establish in a thyrotoxic patient with a superimposed infectious illness. The customary therapy consists of administration of b-blockers, corticosteroids, iodide or ipodate, and PTU or MMI (Table 42-10). No parenteral preparation of PTU or MMI is available, but MMI or PTU may be administered intrarectally by enema.93 Alternatively, and probably more reliably, the PTU or MMI tablets may be crushed and administered through a nasogastric tube. Antithyroid drugs should be given several hours before the administration of iodides. Hyperactive patients require sedation, and severely pyrexic patients should be treated with acetaminophen and a cooling blanket. If this customary therapy is not effective, peritoneal dialysis, resin or charcoal hemoperfusion, or plasma-pheresis can be attempted if the patient is seriously ill.94

TABLE 42-10. Treatment of Thyroid Storm*

EUTHYROID GRAVES DISEASE Euthyroid Graves disease refers to clinically euthyroid patients who may have detectable serum TSH receptor antibodies and also may have other signs of autoimmune thyroid disease, such as ophthalmopathy.95,96 These euthyroid patients may present with unilateral proptosis, and the clinician must determine whether autoimmune Graves disease is present (see Chap. 43). A thorough medical history and physical examination may uncover data suggesting an earlier occurrence of Graves disease. Serum total, free T4 and T3, and a third-generation TSH may be normal (see Chap. 15). Measurement of TSH receptor antibodies may also be helpful. Because a rare patient may have euthyroid Graves disease and a coexisting retroorbital or brain tumor causing proptosis, orbital and head CT (noncontrast) or MRI may be indicated and could disclose thickening of the orbital muscles or other findings consistent with Graves disease, thereby confirming the presence of autoimmune thyroid eye disease. It should be kept in mind that a CT study with radiopaque dye influences the thyroidal iodine transport system and the ability to perform a radioactive iodine uptake study or treatment with 131I. The diagnosis of euthyroid Graves disease is based on inference and exclusion. The natural history of euthyroid Graves disease is unpredictable, with some patients remaining euthyroid and others becoming thyrotoxic or even hypothyroid. Some may have euthyroid Graves disease without marked clinical eye findings. The distinction between euthyroid Graves disease and Graves disease in remission is difficult. Latent euthyroid Graves disease may be associated with an abnormal TRH test and/or abnormal T3-suppression test. The author rarely performs these tests, especially because a third-generation TSH assay abrogates the necessity for TRH stimulation tests, and a T3-suppression test may be dangerous, without being particularly helpful if TSH receptor antibodies have been measured. HASHIMOTO THYROIDITIS VERSUS GRAVES DISEASE Although earlier studies suggested that Hashimoto and Graves disease were separate entities, there is evidence of some overlap. Thus, antithyroglobulin and peroxidase antibodies may be seen in Graves disease, and some patients with Hashimoto disease may have TSH receptor antibodies.97,98 It has been suggested that autoantibodies against thyroid peroxidase may react to different epitopes, depending on whether Graves disease or Hashimoto thyroiditis is present. Furthermore, patients with Hashimoto thyroiditis may present with thyrotoxicosis, and patients with Graves disease may present with hypothyroidism. Patients with either disease may have blocking antibodies directed against the TSH receptor, and there is preliminary evidence that serum in either disease may contain antibodies that recognize identical or similar antigens in thyroid membranes. It is not known whether the causes of Graves disease and Hashimoto disease are similar and whether these two diseases reflect a varied spectrum of the same pathologic process. Nevertheless, one of the relevant clinical findings is that a condition diagnosed as hypothyroid Hashimoto disease rarely may evolve into thyrotoxicosis, necessitating an evaluation on an annual basis of hypothyroid patients receiving L -T4 replacement therapy. One should not presume that a thyrotoxic patient with high-titer microsomal antibodies has a self-limited variant of hyperthyroidism such as hashitoxicosis, because it is possible that serum T4 levels will remain elevated and that the patient will ultimately need definitive therapy. Patients with Graves disease and high-titer antithyroid (peroxidase or thyroglobulin) antibodies are probably more likely to have thyroid hypofunction after either medical or ablative therapy. The presence of very-high-titer anti-TSH receptor antibodies in a pregnant woman presumed to have Hashimoto disease is particularly worrisome, because these antibodies can cross the placenta to cause either thyrotoxicosis or hypothyroidism in the neonate.97 Finally, postpartum thyroiditis (with thyrotoxicosis) occurs mainly in patients with significant antimicrosomal (anti-peroxidase) or antithyroglobulin antibody titers.97

GRAVES DISEASE IN CHILDREN Graves disease rarely occurs before age 18. Many of the clinical signs and symptoms, laboratory assessments, and treatment protocols in adults and children are similar (see Chap. 47).98 Because it is likely to be more successful, and because of the reluctance to recommend ablative therapy in a child, most patients with Graves disease younger than age 18 are given PTU or MMI for a year in an effort to induce remission. In some subjects with persistent disease who respond to small doses of PTU or MMI, these medications may be continued until the patients are judged to be old enough (or sufficiently mature) to understand the therapeutic options of receiving 131I or undergoing thyroidectomy. The major controversy in the treatment of childhood Graves disease that does not remit with antithyroid drugs concerns the choice between surgery and 131I treatment. Thyroidectomy can be performed in children with the same comparably low complication rates as in adults. Authorities disagree, however, as to whether children and adolescents should be treated with radioactive 131I. In addition to the arguments already noted, young persons are at risk for the development of 131I-related effects for a longer interval than adults so treated, and there is some evidence that children treated with 131I dare more likely to develop benign thyroid nodules. In general, the same guidelines for treatment apply as in adults. Given the advantages and disadvantages of each form of therapy, these guidelines should be individualized to the wishes of the child and his or her parents. GRAVES DISEASE IN THE AGED The elderly patient with Graves disease often does not have hyperkinetic symptomatology despite the occasional presence of severe disease (see Chap. 199). Exophthalmos is often absent, and the goiter may be minimal or absent. Instead of an increased appetite and diarrhea, anorexia and constipation may be present. It is not uncommon for these patients to present with manifestations that may be unusual in younger subjects, such as myopathy, arrhythmia, or heart failure.24,25 and 26 The issue of the risk of thromboembolism occurring in hyperthyroid patients with atrial fibrillation is important. In one retrospective study, 30 of 142 thyrotoxic patients (21%) had atrial fibrillation, and 12 of these patients had thromboemboli.99 None of 112 thyrotoxic patients without atrial fibrillation had emboli. Although these data need to be confirmed, many clinicians prescribe anticoagulants for thyrotoxic patients with atrial fibrillation, assuming there are no contraindications; in an individual patient, the benefits are thought to outweigh the potential disadvantages. This decision also depends on factors such as the presence or absence of other underlying cardiac disease, the results of the echocardiogram, and the entire clinical situation. Close monitoring is essential.100 Patients with a placid facies and withdrawn behavior, limited attention span, or depression have been described as having apathetic hyperthyroidism. The term masked or monosymptomatic hyperthyroidism has been used for those patients who have only one predominant symptom, such as severe weight loss mimicking a malignancy or severe hypomagnesemia (Fig. 42-9). The elderly hyperthyroid patient requires close follow-up and usually nonsurgical therapy.

FIGURE 42-9. Masked hyperthyroidism in a 65-year-old man whose principal symptom was weight loss. No goiter was palpable. Note the gynecomastia.

L-THYROXINE

ADMINISTRATION IN TREATED GRAVES DISEASE

Reports have suggested that it might be useful to administer L -T4 supplementation to Graves disease patients after they have been rendered euthyroid by MMI.101,102 In a controlled study from Japan, patients administered MMI plus L -T4 had lower levels of TSH receptor antibodies and a lower incidence of disease recurrence when observed for as long as 3 years after discontinuation of MMI, as compared with the group treated with MMI alone. The mechanism for this potential effect is unknown. The same investigators also reported that T4 administration during pregnancy and in the postpartum period prevented a recurrence of postpartum hyperthyroidism.101 These studies are of interest, but most authorities believe this therapy should not yet be used in the routine treatment of Graves disease. Indeed, studies have not been able to confirm the potential benefit of L -T4 supplementation in this circumstance.102 SUPPRESSED SERUM THYROID-STIMULATING HORMONE Because of the development of highly sensitive TSH assays, many individuals have been found with normal total T4 and a TSH of <0.1 U/mL in the context of clinical euthyroidism. The precise pathophysiology of this occurrence is unknown, and the natural history is undefined. On occasion, these individuals develop overt hyperthyroidism, whereas in others, transient decrements in TSH return to normal. Many subjects have persistent suppression of TSH with normal total and free T4 and T3. Further studies are needed to determine the proper approach to these patients. However, it appears reasonable to perform a thorough history and physical examination, to obtain a total T3 and free T4 measurement, and, in some instances, to obtain radioactive iodine uptake and a free T3. These patients should be evaluated periodicallyespecially elderly individualsfor the development of atrial fibrillation.103,104 and 105 One approach is to measure free T 4 and T3 and TSH (third generation) monthly for 3 months to ensure that the perturbations are not transient. If the repeat laboratory studies show persistent TSH levels that are undetectable in conjunction with a normal free T4 and free T3, a radioactive iodine uptake and scan are performed and serum TSI levels are obtained to assess the likelihood of overt hyperthyroidism developing. Occasionally, a thyroid sonogram is also performed to determine whether the thyroid gland contains non-palpable nodules. Assuming these tests are normal or only mildly abnormal, one may consider therapy for these patients to decrease the risk of cardiac arrhythmias and possible bone loss over an extended time. The consideration of therapy depends on the clinical context and the patient's desires, but, in general, antithyroid treatment is the primary choice, rather than more definitive therapy such as radioactive iodine or surgery. For example, 5 mg MMI may be given as a therapeutic trial for 6 to 12 months, with monitoring of complete blood count, liver function test, and thyroid function tests perhaps every 4 to 6 weeks. It is unknown if this condition represents a singular entity or is a manifestation of a different disorder. It is also unknown if such patients are at higher risk for complications if they require nonthyroidal surgery. Future studies addressing these issues are required. CHAPTER REFERENCES
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. Burman KD, Baker JR Jr. Immune mechanisms in Graves disease. Endocr Rev 1985; 6:183. Weetman AP, McGregor AM. Autoimmune thyroid disease: developments in our understanding. Endocr Rev 1984; 5:309. Baker JR Jr. Dissecting the immune response to the thyrotropin receptor in autoimmune thyroid disease. J Clin Endocrinol Metab 1993; 77:16. Ajjan RA, Findlay C, Metcalfe RA, et al. The modulation of the human sodium iodide symporter activity by Graves' disease sera. J Clin Endocrinol Metab 1998; 83:1217. Franklyn JA, Maisonneuve P, Sheppard MC, et al. Mortality after the treatment of hyperthyroidism with radioactive iodine. N Engl J Med 1998; 338:712. Jaspan JB, Luo H, Ahmed B, et al. Evidence for a retroviral trigger in Graves' disease. Autoimmunity 1995; 20:135. Zakarija M, McKenzie JM, Munro DS. Immunoglobulin G inhibitor of thyroid-stimulating antibody is a cause of delay in the onset of neonatal Graves disease. J Clin Invest 1983; 72:1352. Ahmann AJ, Burman KD. The role of T lymphocytes in autoimmune thyroid disease. Endocrinol Metab Clin North Am 1987; 16:287. Iitaka M, Aguayo JF, Iwatani Y, et al. Studies of the effect of suppressor T lymphocytes on the induction of antithyroid microsomal antibody-secreting cells in autoimmune thyroid disease. J Clin Endocrinol Metab 1988; 66:708. Dayan CM, Londei M, Corcoran AE, et al. Autoantigen recognition by thyroid-infiltrating T cells in Graves disease. Proc Natl Acad Sci U S A 1991; 88:7415. Topliss D, How J, Lewis M, et al. Evidence for cell-mediated immunity and specific suppressor T lymphocyte dysfunction in Graves disease and diabetes mellitus. J Clin Endocrinol Metab 1983; 57:700. Vento S, Hegarty JE, Bottazzo GF, et al. Antigen-specific suppressor cell function in autoimmune chronic active hepatitis. Lancet 1984; 1:1200. Ludgate ME, Ratanachaiyavong S, Weetman AP, et al. Failure to demonstrate cell-mediated immune responses to thyroid antigens in Graves disease using in vitro assays of lymphokine-mediated migration inhibition. J Clin Endocrinol Metab 1985; 60:98. Kaulfersch W, Baker JR, Burman KD, et al. Immunoglobulin and T cell gene rearrangements demonstrate that the immune response in autoimmune thyroid disease is polyclonal in nature. J Clin

15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31.

Endocrinol Metab 1988; 66:958. Melver B, Morris JC. The pathogenesis of Graves' disease. Endocrinol Metab Clin North Am 1998; 27:73. Weetman AP, Volkman DJ, Burman KD, et al. The in vitro regulation of human thyrocyte HLA-DR antigen expression. J Clin Endocrinol Metab 1985; 61:817. Jerne NK. Towards a network theory of the immune system. Ann Immunol (Paris) 1974; 125:373. Raines KB, Baker JR Jr, Lukes YG, et al. Antithyrotropin antibodies in the sera of Graves disease patients. J Clin Endocrinol Metab 1985; 61:217. Weiss M, Ingbar SH, Winblad S, Kasper DL. Demonstration of a saturable binding site for thyrotropin in Yersinia enterocolitica. Science 1983; 219:1331. Sack J, Zilberstein D, Barile MF, et al. Binding of thyrotropin to selected Mycoplasma species: detection of serum antibodies against a specific Mycoplasma membrane antigen in patients with autoimmune thyroid disease. J Endocrinol Invest 1989; 12:77. Lagaye S, Vexiau P, Morozov V, et al. Human spumaretrovirusrelated sequences in the DNA of leukocytes from patients with Graves disease. Proc Natl Acad Sci U S A 1992; 89:10070. Bahn RS, Dutton CM, Heufelder AE, Sarkar G. A genomic point mutation in the extracellular domain of the thyrotropin receptor in patients with Graves ophthalmopathy. J Clin Endocrinol Metab 1994; 78:256. Gorman CA. Unusual manifestation of Graves disease. Mayo Clin Proc 1972; 47:926. Davis PJ, Davis FB. Hyperthyroidism in patients over the age of 60 years: clinical features in 85 patients. Medicine (Baltimore) 1974; 53:161. Parker JLW, Lawson DH. Death from thyrotoxicosis. Lancet 1973; 2:894. Thomas FB, Mazzaferri EL, Skillman TG. Apathetic thyrotoxicosis: a distinctive clinical and laboratory entity. Ann Intern Med 1970; 72:679. Heffron W, Eaton RP. Thyrotoxicosis presenting as choreoathetosis. Ann Intern Med 1970; 73:425. Singh SP, Ellyin F, Singh SK, Yoon B. Elephantiasis-like appearance of upper and lower extremities in Graves dermopathy. Am J Med Sci 1985; 290:73. Burch HB, Wartofsky L. Graves ophthalmopathy: current concepts regarding pathogenesis and management. Endocr Rev 1993; 14:747. Bahn RS, Heufelder AE. Pathogenesis of Graves ophthalmopathy. N Engl J Med 1993; 329:1468. Cunningham M, Zone JJ. Thyroid abnormalities in dermatitis herpetiformis. Ann Intern Med 1985; 102:194.

31a. Pantazi H, Papapetrou PD. Changes in parameters of bone and mineral metabolism during therapy for hyperthyroidism. J Clin Endocrinol Metab 2000; 85:1099. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. Burman KD, Monchik J, Earll JM, Wartofsky L. Total and ionized calcium and parathyroid hormone in hyperthyroidism. Ann Intern Med 1976; 84:668. Thompson P Jr, Strum D, Boehm T, Wartofsky L. Abnormalities of liver function tests in thyrotoxicosis. Mil Med 1978; 143:548. Sterling K, Refetoff S, Selenkow HA. T3 thyrotoxicosis: thyrotoxicosis due to elevated serum triiodothyronine levels. JAMA 1970; 213:571. Hay ID, Bayer MF, Kaplan MM, et al, for the Committee on Nomenclature of the American Thyroid Association. American Thyroid Association assessment of current free thyroid hormone and thyrotropin measurements and guidelines for future clinical assays. Clin Chem 1991; 37:2002. Sturgess ML, Weeks I, Evans PJ, et al. An immunochemiluminometric assay for serum free thyroxine. Clinical Endocrinol (Oxf) 1987; 27:383. Symposium on Sensitive TSH Assays. Mayo Clin Proc 1988; 63:1026, 1123, 1214. Cooper DS. Subclinical thyroid disease. Ann Intern Med 1998; 129:135. Wartofsky L, Burman KD. Alterations in thyroid function in patients with systemic illness: the euthyroid sick syndrome. Endocr Rev 1982; 3:164. Patibandla SA, Dallas JS, Seetharamaiah GS, et al. Flow cytometric analysis of antibody binding to Chinese hamster ovary cells expressing human thyrotropin receptor. J Clin Endocrinol Metab 1997; 82:1885. Morris JC, Hay ID, Nelson RE, Jiang N-S. Clinical utility of thyrotropin-receptor assays: comparison of radioreceptor and bioassay methods. Mayo Clin Proc 1988; 63:707. Burman KD, Pandian R. Clinical utility of assays for TSH receptor antibodies. The Endocrinologist 1998; 8:284. Kim WB, Chung HK, Lee HK, et al. Changes in epitopes for thyroid-stimulating antibodies in Graves' disease sera during treatment of hyperthyroidism: therapeutic implications. J Clin Endocrinol Metab 1997; 82:1953. Mariotti S, Martino E, Cupini C, et al. Low serum thyroglobulin as a clue to the diagnosis of thyrotoxicosis factitia. N Engl J Med 1982; 307:410. Kraiem Z, Glaser B, Yigla M, et al. Toxic multinodular goiter: a variant of autoimmune hyperthyroidism. J Clin Endocrinol Metab 1987; 65:659. Hay ID, Thyroiditis: a clinical update. Mayo Clin Proc 1985; 60:836. Amino N, Mori H, Iwatani Y, et al. High prevalence of transient postpartum thyrotoxicosis and hypothyroidism. N Engl J Med 1982; 306:849. Stanbury JB, Eermans AE, Bourdoux P, et al. Iodine-induced hyperthyroidism: occurrence and epidemiology. Thyroid 1998; 8:83. Hayslip CC, Fein HG, O'Donnell VM, et al. The value of serum antimicrosomal antibody testing in screening for symptomatic postpartum thyroid dysfunction. Am J Obstet Gynecol 1988; 159:203. Waterman EA, Watson PF, Lazarus JH, et al. A study of the association between a polymorphism in the CTLA-4 gene and postpartum thyroiditis. Clin Endocrinol 1998; 49:251. Glinoer D, Riahi M, Grn J-P, Kinthaert J. Risk of subclinical hypothyroidism in pregnant women with asymptomatic autoimmune thyroid disorders. J Clin Endocrinol Metab 1994; 79:197. Rajanatavin R, Chailurkit L, Srisupandit S, et al. Trophoblastic hyperthyroidism: clinical and biochemical features of five cases. Am J Med 1988; 85:237. Kempers RD, Dockerty MB, Hoffman DL, Bartholomew LG. Struma ovarii, ascitic hyperthyroid, and asymptomatic syndromes. Ann Intern Med 1970; 72:883. Smallridge RC. Thyrotropin-secreting pituitary tumors. Endocrinol Metab Clin North Am 1987; 16:765. Brucker-Davis F, Oldfield EH, Skarulis MC, et al. Thyrotropin-secreting pituitary tumors: diagnostic criteria, thyroid hormone sensitivity, and treatment outcome in 25 patients followed at the National Institutes of Health. J Clin Endocrinol Metab 1999; 84:476. Chanson P, Weintraub BD, Harris AG. Octreotide therapy for thyroid-stimulating hormone-secreting pituitary adenoma: a follow-up of 52 patients. Ann Intern Med 1993; 119:236. Kasagi K, Takeuchi R, Miyamoto S, et al. Metastatic thyroid cancer presenting as thyrotoxicosis. Clin Endocrinol (Oxf) 1994; 40:429. Hedberg CW, Fishbein DB, Janssen RS, et al. An outbreak of thyrotoxicosis caused by the consumption of bovine thyroid gland in ground beef. N Engl J Med 1987; 316:993. Wood LC, Ingbar SH. Hypothyroidism as a late sequela in patients with Graves disease treated with antithyroid agents. J Clin Invest 1979; 64:1429. Reinwein D, Benker G, Lazarus JH, Alexander WD, and the European Multicenter Study Group on Antithyroid Drug Treatment. A prospective randomized trial of antithyroid drug dose in Graves disease therapy. J Clin Endocrinol Metab 1993; 76:1516. Singer PA, Cooper DS, Levy EG, et al. Treatment guidelines for patients with hyperthyroidism and hypothyroidism: Standards of care committee, American Thyroid Association. JAMA 1995; 273:808. Cooper DS. Antithyroid drugs. N Engl J Med 1984; 311:1353. Solomon BL, Evaul JE, Burman KD, Wartofsky L. Remission rates with antithyroid drug therapy: continuing influence of iodine intake? Ann Intern Med 1987; 107:510. Mezquita P, Luna V, Munoz-Torres M, et al. Methimazole-induced aplastic anemia in third exposure: successful treatment with recombinant human granulocyte colony-stimulating factor. Thyroid 1998; 8:791.

64a. Mylonakis E, Akhtar MS, Lopez F, et al. Resolution of drug-induced agran-ulocytosis. Geriatrics 2000; 55:89. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. 103. 104. 105. Mathieu E, Fain O, Sitbon M, Thomas M. Systemic adverse effect of antithyroid drugs. Clin Rheumatol 1999; 18:66. Burman KD. Thyroid hormones. In: Chernow B, ed. The pharmacologic approach to the critically ill patient. Baltimore: Williams & Wilkins, 1988:586. Shen D-C, Wu S-Y, Chopra IJ, et al. Long-term treatment of Graves hyperthyroidism with sodium ipodate. J Clin Endocrinol Metab 1985; 61:723. Boehm TM, Burman KD, Barnes S, Wartofsky L. Synergism of lithium and iodine in the treatment of thyrotoxicosis. Acta Endocrinol (Copenh) 1980; 94:174. Kushner JP, Wartofsky L. Lithium thyroid interactions. In: Johnson FN, ed. Lithium therapy monographs, vol 2. Basel: Karger, 1988:74. Wartofsky L, Ransil BJ, Ingbar SH. Inhibition by iodine of the release of thyroxine from the thyroid glands of patients with thyrotoxicosis. J Clin Invest 1970; 49:78. McDermott MT, Burman KD, Hofeldt FD, Kidd GS. Lithium-associated thyrotoxicosis. Am J Med 1986; 80:1245. Graham DG, Burman KD. Radioiodine treatment of Graves disease. Ann Intern Med 1986; 105:900. Saenger EL, Thoma GE, Tompkins EA. Incidence of leukemia following treatment of hyperthyroidism. JAMA 1968; 205:147. Ron E, Doody MM, Becker DV, et al. Cancer-mortality following treatment for adult hyperthyroidism: cooperative thyrotoxicosis therapy follow-up study group. JAMA 1998; 280:375. Klein AH, Hobel CJ, Sack J, Fisher DA. Effect of intra-amniotic fluid thyroxine injection on fetal serum and amniotic fluid iodothyronine concentrations. J Clin Endocrinol Metab 1978; 47:1034. Burch HB, Solomon BL, Wartofsky L, Burman KD. Discontinuing antithyroid drug therapy before ablation with radioiodine in Graves disease. Ann Intern Med 1994; 121:553. McDermott MT, Kidd GS, Dodson LE, Hofeldt FD. Radioiodine-induced thyroid storm. Arch Intern Med 1983; 75:353. Bartalena L, Marcocci C, Bogazzi F, et al. Relation between therapy for hyperthyroidism and the course of Graves' ophthalmopathy. N Engl J Med 1998; 338:73. Hershman JM. The treatment of hyperthyroidism. Ann Intern Med 1966; 64:1306. Pellegriti G, Belfiiore A, Giuffrida D, et al. Outcome of differentiated thyroid cancer in Graves' disease. J Clin Endocrinol Metab 1998; 83:2805. Dent CE, Harper CM. Hypoparathyroid tetany (following thyroidectomy) apparently resistant to vitamin D. Proc R Soc Med 1958; 51:489. Ljunggren JG, Torring O, Wallin G, et al. Quality of life aspects and costs in treatment of Graves' hyperthyroidism with antithyroid drugs, surgery, or radioiodine: results from a prospective, randomized study. Thyroid 1998; 8:653. Glinoer D. Thyroid hyperfunction during pregnancy. Thyroid 1998; 8:859. Momotani N, Noh J, Oyanagi H, et al. Antithyroid drug therapy for Graves disease during pregnancy: optional regimen for fetal thyroid status. N Engl J Med 1986; 315:24. Momotani N, Ito K, Hamada N, et al. Maternal hyperthyroidism and congenital malformation in the offspring. Clin Endocrinol (Oxf) 1984; 20:695. Rubin PC. Beta-blockers in pregnancy. N Engl J Med 1981; 305:1323. Van Dijke CP, Heydendael RJ, DeKleine MJ. Methimazole, carbimazole, and congenital skin defects. Ann Intern Med 1987; 106:60. Zakarija M, McKenzie JM. Pregnancy associated changes in the thyroid-stimulating antibody of Graves disease and the relationship to neonatal hyperthyroidism. J Clin Endocrinol Metab 1982; 57:1036. Volp R, Ehrlich R, Steiner G, Row VV. Graves disease in pregnancy years after hypothyroidism with recurrent passive-transfer neonatal Graves disease in offspring: therapeutic considerations. Am J Med 1984; 77:572. Dydek GJ, Blue PW. Human breast milk excretion of iodine-131 following diagnostic and therapeutic administration to a lactating patient with Graves disease. J Nucl Med 1988; 29:407. Davis LE, Lucas MJ, Hankins DV, et al. Thyrotoxicosis complicating pregnancy. Am J Obstet Gynecol 1989; 160:63. Burch HB, Wartofsky L. Life-threatening thyrotoxicosis. Endocrinol Metab Clin North Am 1993; 22:263. Nabil N, Miner DJ, Amatruda JM. Methimazole: an alternative route of administration. J Clin Endocrinol Metab 1982; 54:180. Tajiri J, Katsuya H, Kiyokawa T, et al. Successful treatment of thyrotoxic crisis with plasma exchange. Crit Care Med 1984; 12:536. Franco PS, Hershman JM, Haigler ED Jr, Pittman JA Jr. Response to thyrotropin-releasing hormone compared with thyroid suppression tests in euthyroid Graves disease. Metabolism 1973; 22:1357. Solomon DH, Chopra IJ, Chopra U, Smith FJ. Identification of subgroups of euthyroid Graves ophthalmopathy. N Engl J Med 1977; 296:181. Ueta Y, Fukui H, Murakami H, et al. Development of primary hypothyroidism with the appearance of blocking-type antibody to thyrotropin receptor in Graves' disease in late pregnancy. Thyroid 1999; 9:179. Rivkees SA, Sklar C, Freemark M. Clinical review: the management of Graves' disease in children, with special emphasis on radioiodine treatment. J Clin Endocrinol Metab 1998; 83:3767. BarSela S, Ehrenfeld M, Eliakim M. Arterial embolism in thyrotoxicosis with atrial fibrillation. Arch Intern Med 1981; 141:1191. Woeber KA. Thyrotoxicosis and the heart. N Engl J Med 1992; 327:94. Hashizume K, Ichikawa K, Nishii Y, et al. Effect of administration of thyroxine on the risk of postpartum recurrence of hyperthyroid Graves disease. J Clin Endocrinol Metab 1992; 75:6 Rittmaster RS, Zwicker H, Abbott EC, et al. Effect of methimazole with or without L-thyroxine on serum concentrations of thyrotropin (TSH) receptor antibodies on patients with Graves' disease. J Clin Endocrinol Metab 1996; 81:3283. Cooper, DS. Subclinical thyroid disease: a clinician's perspective. Ann Intern Med 1998; 129(2):135. Marqusee E, Haden ST, Utiger RD. Subclinical thyrotoxicosis. Endocrinol Metab Clin North Am 1998; 27:37. Samuels MH. Subclinical thyroid disease in the elderly. Thyroid 1998;

CHAPTER 43 ENDOCRINE OPHTHALMOPATHY Principles and Practice of Endocrinology and Metabolism

CHAPTER 43 ENDOCRINE OPHTHALMOPATHY

MELVIN G. ALPER AND LEONARD WARTOFSKY Definition Etiology Pathology Clinical Course Complications Physical Examination and Mensuration Differential Diagnosis Treatment Immunomodulatory Therapy Supervoltage Therapy Surgical Decompression of Orbit Chapter References

DEFINITION
Endocrine ophthalmopathy is a complex orbital disease of unknown cause characterized by round-cell infiltration, edema, and proliferation of connective tissue. These changes affect predominantly the extraocular muscles and to a lesser degree the lacrimal glands and retrobulbar fat. Usually, exophthalmos, or protrusion of the eyeballs, is present. Endocrine ophthalmopathy may occur alone, in association with diffuse thyrotoxic goiter or pretibial myxedema, or with both conditions. These findings form part of the symptom complex that is termed Graves disease (see Chap. 42). Endocrine ophthalmopathy, or endocrine exophthalmos, the latter term coined by Brain1 in 1959, is a useful but somewhat misleading appellation because the orbital changes may occur without endocrine abnormalities. McKenzie2 defined the association of ophthalmopathy with endocrinopathy as a multisystem disorder of unknown etiology, characterized by one or more of three clinical entities: (a) hyperthyroidism associated with diffuse hyperplasia of the thyroid gland, (b) infiltrative ophthalmopathy, and (c) infiltrative dermopathy (localized pretibial myxedema). The condition is more common in women, but often the most severe cases are encountered in men. The genetics of ophthalmopathy appear to be linked to those of Graves disease, with severity of disease distinguished by certain human leukocyte antigen (HLA) types.3 Reported estimates of the prevalence of ophthalmopathy in Graves disease vary widely from 4% to 60%, probably owing to variation in both definition and the sensitivity of detection techniques. Excluding eyelid signs, only 5% to 20% of Graves patients have ophthalmopathy, but the rate rises to 60% or more when computed tomography (CT) or magnetic resonance imaging (MRI) techniques are used to uncover otherwise unapparent muscle involvement. Werner4 devised a classification of the eye changes of Graves disease: Class 0: No signs or symptoms Class 1: Only signs, no symptoms (signs limited to upper eyelid; retraction and stare, with or without eyelid lag and proptosis) Class 2: Soft-tissue involvement (symptoms and signs) Class 3: Proptosis Class 4: Extraocular muscle involvement Class 5: Corneal involvement Class 6: Sight loss (optic nerve involvement) The progression of eye changes need not be sequential through each of the classes. The principal usefulness of the system relates to its ability to describe and transmit information to other observers about the clinical details of the ocular changes of Graves disease, but it has limitations as a guide for quantitating such changes. As a result, a number of other workers have devised, modified, and revised classifications that provide increased utility.5 Aspects of a typical clinical presentation, characteristics of the natural history with time, and long-term follow-up have been described.6,7,8,9,10,11 and 12 Diagnostic criteria have been described for Graves patients with either new, evolving, or worsening ophthalmopathy.8,11,12

ETIOLOGY
Proptosis, soft-tissue involvement, and ophthalmoplegia occurring in a patient with concurrent thyrotoxicosis or a past history of hyperthyroidism suggests dysfunction of the thyroid. Occasionally, Hashimoto thyroiditis, with or without clinical hypothyroidism, may be coexistent. When the eye changes occur in apparently euthyroid patients, not only Graves ophthalmopathy but also other primary orbital diseases must be considered. Although the fundamental cause of Graves disease is unknown, the hyperthyroidism is generally accepted to be caused by a group of thyroid-stimulating immunoglobulins directed toward thyroid-stimulating hormone (TSH) receptors.13,14,15,16 and 17 Whether or not these same antibodies also are responsible for the ophthalmopathy and associated immune syndromes remains to be determined,18,19 and 20 although B lymphocytes represent a minor fraction of the cells infiltrating orbital connective tissue and muscle. Although no studies have demonstrated in vivo binding of TSH-receptor antibodies to TSH receptor in retroocular tissues, findings of TSH-receptor transcripts and TSH receptor like proteins in orbital fibroblasts suggest an important link to the pathogenesis,21,22 albeit possibly to only a portion of the TSH receptor.23 Some in vitro evidence of cytotoxic responses against eye muscle antigens exists, but no histologic evidence of cytotoxicity against eye muscle has been demonstrated. A 64-kDa protein has been identified that may be a shared eye muscle and thyroid antigen, but its specificity to Graves ophthalmopathy patients has not been established. One hypothesis is that autoreactive CD4+ T lymphocytes are activated by thyroid antigen in Graves disease, leading to the immune response against the TSH receptor and T-cell infiltration in the orbit. T lymphocytes are the predominant cells infiltrating eye muscle and are drawn to the eye by a complex interaction involving expression by retroocular fibroblasts of intercellular adhesion molecule-1 along with vascular and endothelial adhesion molecules.24 A variety of other cytokines (interferon-g, transforming growth factor-b, interleukin-1, tumor necrosis factor, and other fibroblast growth and activating factors) are also involved.25 The cytokines stimulate the fibroblasts to release glycosaminoglycans, which, because of their hydrophilic nature, lead to interstitial edema.26 Progression of the autoimmune response with more intense lymphocyte infiltration, fibroblast proliferation, and edema leads to increased orbital tissue volume, protrusion of the globe, exposure keratopathy, and venous compression, which in turn results in more edema. What initiates the entire process in genetically susceptible people is under intense study. The possible role played by the release of thyroid antigen, such as occurs after radioiodine therapy,27,28,29,30,31 and 32 and by eye muscle as antigen,33 as well as by other exogenous factors, has been reviewed. Superoxide (oxygen free) radicals have been implicated as well, with the proliferative response from orbital fibroblasts being caused by conditions of oxidative stress.34 One such stress may be tobacco smoking, which has been associated with worsening ophthalmopathy.13,14,35,36 and 37 The possibility that patients receiving radioiodine may experience exacerbated ophthalmopathy has been evaluated by frequency comparisons with the outcomes for patients treated with thyroidectomy or antithyroid drugs;27,28,29,30,31,32,38,39 and 40 some of the evidence is compelling. The mechanism appears to be antigenic release of TSH receptor from the thyroid, with activation of T and B lymphocytes leading to increases in serum TSH-receptor antibodies. Until the issue is resolved with a large, well-controlled prospective study, physicians should counsel patients who appear to be at highest risk of worsening ophthalmopathy after radioiodine (e.g., those with more severe thyrotoxicosis, smokers, and those with preexistent ophthalmopathy). If antithyroid drugs or thyroidectomy are not viable alternatives, concomitant therapy with corticosteroids should be considered.39,41 In addition to varying in presentation, the ophthalmopathy of Graves disease may vary in severity. The degree of exophthalmos often is unrelated to the severity of the thyrotoxicosis and, paradoxically, may occasionally appear after the hyperthyroidism is controlled.9,12,29,42,43 The severity of eye changes may vary (Fig. 43-1), but the basic pathogenetic process is the same.

FIGURE 43-1. A, A 38-year-old woman with Graves hyperthyroidism. Note widened palpebral fissures, left greater than right, giving appearance of left unilateral exophthalmos. Measurements showed no proptosis. Both upper eyelids are retracted, baring sclera above the limbus in each eye, in left greater than in right (class 1 of Werner's abridged classification of ocular changes of Graves disease). B, A 56-year-old man with recent onset of class 3 ocular changes 2 years after treatment with radioactive iodine for Graves hyperthyroidism. The patient developed hypothyroidism after therapy and was receiving replacement thyroxine. Proptosis was present; Hertel exophthalmometry measured 33 mm in each eye. Note the soft-tissue involvement with fullness and edema of eyelids, chemosis, edema of caruncles, and conjunctival injection. Both upper lids are retracted. Herniation of retrobulbar fat through the orbital septal spaces of Charpey is present.

PATHOLOGY
The major pathologic changes that occur in Graves disease are inflammatory infiltration by lymphocytes and chronic inflammatory cells in all orbital tissues with proliferation of the connective tissue (Table 43-1). Because autopsy findings are rare in this condition, precise clinicopathologic correlation has not been made for each of the physical signs that are described. Most of the material examined by pathologists is obtained at the time of surgery for orbital decompression or correction of extraocular muscle difficulties. Thus, the earliest changes in the orbital tissues are rarely studied, whereas abundant reports exist of some of the pathologic findings in the more severe forms of ophthalmopathy.

TABLE 43-1. Pathogenesis of Graves Ophthalmopathy

All orbital structures, including lacrimal glands, are affected. Massive swelling of the extraocular muscles is commonly seen; occasionally, they may become 8 to 10 times their normal volume (Fig. 43-2). Histologically, marked interstitial inflammatory edema and focal lymphocytic infiltration of the muscles is noted. The myofibrillar structure remains normal, but loss of striations and disorganization may occur later. With thyrotoxicosis, a diffuse increase in orbital fat content may be present, and a deposition of adipose tissue cells lying in strands between the muscle fibers may be noted. In the absence of thyrotoxicosis, the fat content is decreased and is replaced by connective tissue. The mucopolysaccharide and water content of the tissues increases, and large numbers of mast cells are seen in the perivascular areas.

FIGURE 43-2. A, Orbital contents obtained at autopsy from a 47-year-old man with an 8-month history of Graves hyperthyroidism. Bilateral vision loss had been present for 2 months. While hospitalized for high-dose prednisone and orbital radiotherapy, he experienced acute myocardial infarction. Note massive enlargement of extraocular muscles surrounding the optic nerve. Orbital fat has been removed. B, Photomicrograph of a muscle biopsy from inferior rectus muscle of a 58-year-old woman with endocrine ophthalmopathy. Note focal collection of lymphocytes. Muscle fibers appear normal. 60 (A from Hufnagel TH, Hickey WF, Cobbs WH, et al. Immunohistochemical and ultrastructure studies on the exenterated orbital tissue of a patient with Graves' disease. Ophthalmology 1984; 91:1411. Published courtesy of Ophthalmology.)

As the volume of the orbital tissues increases, proptosis occurs. The swollen muscles become increasingly immobile, and the globe may deviate, often in a downward direction. Fibrosis of the muscles may ensue, causing permanent limitation of motion. As proptosis progresses, corneal exposure with subsequent ulceration may appear (Fig. 43-3). The pressure in the posterior orbit may be so great that venous drainage is impaired, leading to conjunctival injection initially, edema (chemosis), and eventually optic disc swelling (Fig. 43-4). Occlusion of the central retinal vein or artery may occur if the pressure is sufficient. Optic neuropathy with loss of central vision may appear at any time during the course of the ophthalmopathy. Some patients without marked proptosis have normal-appearing optic nerve discs but have vision loss. The cause of the loss of vision in these patients is probably mechanical compression from the massively swollen extraocular muscles pressing on the optic nerve at the apex of the orbit.44 CT scanning confirms this occurrence.45 Prompt recovery of vision after surgical decompression of the orbit and altitudinal visual field defects suggest that vascular compression also may be a factor.46

FIGURE 43-3. Eye of a 60-year-old man with severe endocrine ophthalmopathy causing marked proptosis, extraocular muscle tethering, and corneal exposure. Marked visual loss was present secondary to corneal ulceration in the right eye (class 5 Werner's abridged classification of ocular changes of Graves disease). Note the opacification (arrow) and vascularization of cornea (arrowheads).

FIGURE 43-4. Funduscopic view of the right eye of a 58-year-old woman with severe Graves ophthalmopathy. Marked loss of vision occurred because of optic nerve neuropathy. The funduscopic examination shows a swollen optic disc (arrow) with hemorrhages (arrowheads). The patient was treated with megadose corticosteroids; the process resolved, and vision was restored.

CLINICAL COURSE
Retraction of the upper eyelids is the most important clinical clue and produces a triad that is nearly pathognomonic of endocrine ophthalmopathy: widened palpebral fissures, lid lag on downward gaze, and infrequent blinking (see Fig. 43-1A). Normally, the upper lid covers the upper limbus so that the free border rests at the upper edge of the pupil. Many factors change this relation, and care must be taken in assuming that the upper lid retraction is abnormal. Lid retraction exists if, when the eyes are directed horizontally forward without staring, a band of white sclera is seen above the iris. Because of the cosmetic changes created by this phenomenon, concern over appearance usually is the presenting complaint. In some instances, only one eyelid is retracted, and the patient may complain that the other upper lid is drooped. Lid lag accompanies the retraction of the upper eyelid and is manifested as a failure of the lid to accompany the descent of the globe on downward gaze, with the appearance of a white band of sclera above the upper limbus. These phenomena (lid retraction and lid lag) give the impression of exophthalmos, but in the early stages, usually little or no measurable proptosis is present. In the mild form of orbital involvement, as the process continues, actual exophthalmos may occur. The forward position of the globe is appreciated best by direct measurements but can be suggested by retraction of the lower lid. If the soft tissues of the orbit become involved, symptoms of lacrimation, foreign-body sensation, and tenseness around the eyes appear. The eyelids become edematous and full, and the conjunctiva develop chemosis in the lower cul-de-sac. Conjunctival injection and dilated epibulbar blood vessels overlying the lateral rectus muscles appear, indicating an increase in orbital pressure and venostasis (see Fig. 43-1B). In some patients, the lacrimal gland may become enlarged, reaching enormous size. It may be prominent on physical examination and also appears as a large mass on CT scanning. Involvement of the extraocular muscles with limitation of ocular motility may occur in one or both eyes and, because of the resultant diplopia, may represent the most common major disability of this condition. (Fig. 43-5 is a diagrammatic representation of extraocular movements, the muscles involved, and their controlling nerves.) Typically, limitation of upward gaze is the most common finding and gives the false impression of paralysis of elevation. It is caused, however, by tethering of the inferior rectus muscle by inflammatory infiltration and subsequent fibrosis. The patient often tilts the chin upward to gain fusion in the lower fields of vision. Eventually, as fibrosis progresses, a unilaterally proptosed eye may become fixed in a downward and inward position owing to contracture of the damaged inferior rectus muscle. Then, enophthalmos may replace the former exophthalmos.

FIGURE 43-5. Extraocular movements, the muscles involved, and their controlling nerves. (From Judge RD, et al. Clinical diagnosis, a physiologic approach. Boston: Little, Brown, 1982:107.)

Less commonly, involvement of the medial rectus muscle causes limitation of lateral gaze, simulating a lateral rectus muscle paralysis (Fig. 43-6). Uncommonly, involvement of the superior rectus muscle causes restriction of motility on downward gaze.

FIGURE 43-6. Patient is looking to the right. Note the limitation of abduction of the right eye.

Myasthenia gravis must be considered as a possible cause of extraocular muscle limitation, because 0.2% of patients with Graves disease develop myasthenia and 5% of myasthenic patients have associated Graves disease. In addition to the edrophonium chloride (Tensilon) test, the forced-duction test is helpful in ruling out this condition. It can be conducted with topical anesthesia; one applies either forceps or a suction cup to the globe and attempts to move the eye in a direction against the field of action of the involved muscle. A positive test result (indicative of a restrictive mechanical process, not a neurologic disease) is one in which marked resistance is

encountered during this maneuver, such that the globe cannot be moved. An elevation of intraocular pressure to glaucomatous levels on attempted upward gaze, when compared with readings taken with the globe in the straight-ahead position, frequently accompanies a slight tethering of the inferior rectus muscle. This may be the first sign of ophthalmopathy and a means for early clinical detection of Graves disease.47 Upward gaze-evoked elevation of intraocular pressure does not require treatment if the pressure is normal in downward gaze.

COMPLICATIONS
With increasing exophthalmos, the globe may be proptosed through the eyelids so that exposure of the cornea occurs. Ulceration, with scarring or spontaneous perforation, may lead to vision loss or even loss of the eye. Occasionally, the optic nerve becomes involved; an unusual neuropathy may develop that threatens vision, causing central scotomas and other visual field losses. This neuropathy may present either as a normal or a swollen optic disc and eventually may progress to optic atrophy. Disc swelling may occur without vision loss as a result of increased posterior orbital pressure. The optic neuropathy with vision loss can occur in the absence of clinical signs of markedly increased intraorbital pressure, but in these patients, CT scanning usually reveals enlarged extraocular muscles at the orbital apex. Asymptomatic optic nerve involvement may be detected by measurement of visual evoked cortical potentials.48

PHYSICAL EXAMINATION AND MENSURATION

The purpose of physical examination and mensuration in orbital disease initially is to establish the position of the eyes relative to each other and to the bony landmarks of the orbit. Orbital disease processes usually affect only the soft tissue of the orbit so that, with very few exceptions, the bony margins remain constant. Thus, measuring devices relate the globe to the fixed bony points of reference. Physical examination of the exophthalmic patient begins with inspection, but direct viewing of the subject can be misleading. The best method of determining if proptosis exists is to stand behind the seated patient and evaluate the position of the eyes by looking down over the brows. The patient is then placed in a supine position, and the position of the eyes is reassessed while the physician is standing above and behind the patient. In normal people, the globe sinks back when the patient is recumbent. In such circumstances, exophthalmometric measurements are reduced from the erect position by 1 to 3 mm.49 This retraction does not occur in the seemingly unaffected eye of patients with Graves disease even though the exophthalmos may appear to be unilateral. However, in cases of exophthalmos resulting from a unilateral orbital tumor, the normal postural difference of the unaffected eye is noted; therefore, this may be used as a method of differential diagnosis between seemingly unilateral endocrine exophthalmos and a unilateral orbital tumor. These techniques merely provide the observer with an impression of exophthalmos, but more accurate measurements are needed to establish a baseline and to follow the progress of a patient with the ophthalmopathy of Graves disease. A measuring device, therefore, is necessary. When a measuring device is used, the most commonly accepted reference points are the cornea and the lateral rim of the orbit. With the eye viewed from the side, a transparent ruler may be held against the lateral bony rim of the orbit; the distance to the cornea may be seen and directly measured. The Luedde exophthalmometer is one such device, which can give approximate results; the measurements with this device are less accurate than those obtained with the instrument devised by Hertel (Fig. 43-7). The latter consists of a system of mirrors or prisms that projects a lateral view of the eye forward and superimposes a millimeter rule measuring from the anterior rim of the lateral orbital wall onto the viewing mirror. Certain errors are inherent in the use of this instrument because of factors such as thickness of tissue over the lateral orbital margins, parallax, and facial asymmetry. In practice, the instrument is placed beneath the lateral canthal tendon, pressed firmly against the lateral bony rim, and stabilized in a vertical plane held parallel to the ground. Readings of the position of the anterior aspect of each eye are recorded with the interorbital distance, which is noted on the base scale and varies from patient to patient. Normal readings range from 15 to 20 mm, and the difference between the eyes usually does not exceed 1 mm. A difference of up to 2 mm may occur with extraocular muscle paralysis, but differences of more than 2.5 mm usually indicate a pathologic process in the orbit. Frequently, an isolated reading is inconclusive, but repeated examinations may reveal a progressive exophthalmic process.

FIGURE 43-7. Measuring devices to quantitate exophthalmos. A, Luedde exophthalmometer. B, Hertel exophthalmometer.

Other methods of demonstrating exophthalmos are more complex and expensive. One radiographic method places a radiopaque contact lens on the eye and measures its distance from the anterior clinoid process on a lateral view. CT scanning is a useful adaptation of radiographic techniques. Care must be taken with radiographic techniques to maintain the head in the same position each time a measurement is taken. The Hertel exophthalmometer remains the most practical and accurate method of measuring exophthalmos. Palpation of the orbit may indicate a moderate to severe degree of retrobulbar resistance. Orbitonometers, however, invented to calibrate the force of orbital resistance, have proved to be of little clinical value.

DIFFERENTIAL DIAGNOSIS
When exophthalmos occurs bilaterally and is accompanied by retraction of the upper lid, with lid lag and limitation of upward gaze, little difficulty is encountered in establishing the diagnosis of Graves ophthalmopathy, even if thyrotoxicosis is absent.6,11,12,50 If proptosis is unilateral, this entity is still the most common cause, but other conditions must be ruled out. A logical, sequential evaluation of the orbit should include studies to separate orbital from periorbital and intracranial lesions and should proceed in a rational fashion without risk to the patient.51 A team effort is necessary that may include the ophthalmologist, endocrinologist, otolaryngologist, nuclear medicine specialist, and neuroradiologist. Table 43-2 shows a useful scheme for the diagnostic evaluation of unilateral exophthalmos. The evaluation of thyroid function is discussed in Chapter 33,Chapter 34,Chapter 35 and Chapter 36.

TABLE 43-2. Investigative Studies for Unilateral Exophthalmos

Conventional radiographic films disclose a great variety of diseases in the orbit, paranasal sinuses, and intracranial cavity, but they reveal little information about the

soft-tissue changes of Graves disease; such changes can be visualized with techniques for imaging the orbit, such as ultrasonography, CT, and MRI.52 Ultrasonography measures the reflectivity of sound waves as they pass through a given slice of tissue. B-scan ultrasonography demonstrates the inflammatory changes that occur in the ophthalmopathy of Graves disease and may differentiate it from normal (Fig. 43-8A), from inflammatory pseudotumor, or from expanding orbital masses (see Fig. 43-8B).

FIGURE 43-8. A, Normal B-scan ultrasonogram demonstrates superior (single arrow) and inferior (double arrows) rectus muscles. Optic nerve is denoted by vector. A-scan ultrasonogram is superimposed at bottom of picture. B, Abnormal B-scan ultrasonogram of patient with Graves ophthalmopathy. Because of technicalities in positioning the ultrasonic probe, the superior rectus muscle (arrow) in this scan appears below the vector through the optic nerve. Compare size of this enlarged muscle to the normal size muscle in A.

A CT scan measures the absorption of x-rays as they pass through a given section of tissue. By use of high-resolution instruments, the CT characteristics of pathologic features of Graves ophthalmopathy have been defined (Fig. 43-9). A classic CT scan of Graves ophthalmopathy demonstrates exophthalmos, herniation of retrobulbar fat through the orbital septum, enlargement of extraocular muscles with normal tendons, bilateral involvement, and, frequently, enlargement of the lacrimal glands.

FIGURE 43-9. A, Transaxial computed tomographic (CT) scan of a 50-year-old woman with severe endocrine ophthalmopathy. Optic nerve neuropathy was present in each eye. Note bilateral involvement with exophthalmos, enlarged medial and lateral rectus muscles (M), normal tendinous insertions, moulding of the medial rectus muscle into the ethmoid bones and sinuses bilaterally, and compression of the optic nerves (N) at the orbital apex bilaterally. B, Coronal CT scan through midorbit demonstrates enlarged extraocular muscles. Note that optic nerves (arrows) can be distinguished as separate structures and are not compressed in this plane. (Zy, zygomatic arches; Ch, cheeks; Br, brain.)

MRI is a nonionizing technique that measures the resonance of hydrogen ions in a given section of tissue when subjected to a magnetic field.51,52 Excellent contrast between pathologic and normal anatomic structures has been demonstrated in Graves ophthalmopathy (Fig. 43-10). MRI scans are as useful as those obtained from high-resolution thin-section CT of the orbit. However, because of the greater availability and cheaper cost, CT is the medical imaging technique of choice.

FIGURE 43-10. A, Coronal magnetic resonance imaging (MRI) scan of a 59-year-old man with severe Graves ophthalmopathy. Bilateral optic nerve neuropathy was present. Note the markedly enlarged superior and inferior rectus muscles (double arrows) encroaching on the optic nerve (single arrow). Orbital fat gives a white signal in this T1 sequence. B, Sagittal MRI scan shows markedly enlarged superior and inferior rectus muscles (double arrows) converging on and compressing the optic nerve (single arrow) at the orbital apex. In this T1 sequence, retrobulbar fat is seen as a white signal.

Other conditions that may be confused with Graves ophthalmopathy and that may be diagnosed by these visualizing techniques include orbital tumors, axial myopia, inflammatory granulomas, cysts, arteriovenous aneurysms, lymphomas, and aberrant third-nerve regeneration (in patients who demonstrate lid retraction on downward gaze). A common error in clinical diagnosis is failure to recognize meningioma en plaque because of the exophthalmos and brawny edema of the eyelids that may occur in this condition. In meningioma, however, the edema of the lower lid has a characteristic unilateral baggy appearance, and no lid retraction is seen. A frequent source of confusion in the differential diagnosis is axial myopia, a common cause of unilateral exophthalmos. Retinoscopy and A-scan ultrasonography reveal this condition. Inflammatory pseudotumor is a condition simulating orbital neoplasm that usually affects the orbital floor structures; it presents with sudden proptosis, lid edema, pain, ophthalmoplegia, and vision loss at the onset of the disease. The exquisite sensitivity to corticosteroids that generally accompanies this condition has been used as an aid in the differential diagnosis. Differentiation of Graves ophthalmopathy from inflammatory orbital pseudotumor often presents a diagnostic challenge clinically, but it is readily accomplished by CT scanning. In pseudotumor, a classic CT scan shows (a) exophthalmos; (b) enhancement of the sclerouveal rim by contrast media (rim sign); (c) enlarged extraocular muscles with abnormal tendons; (d) feathery infiltration of the retrobulbar fat, especially in the anterior third of the orbit; and (e) unilateral involvement (Fig. 43-11).

FIGURE 43-11. Transaxial computed tomographic (CT) scan of a 52-year-old woman with classic inflammatory orbital pseudotumor of left orbit. Note unilateral involvement with left exophthalmos, enlargement of left medial and lateral rectus muscles, thickened abnormal tendons, enhancement of sclerouveal rim (arrow) by contrast media (rim sign), and feathery infiltration of fat in the retrobulbar and perineural areas. Prompt resolution was obtained with corticosteroid therapy.

A carotid cavernous aneurysm may produce exophthalmos similar to that seen in patients with unilateral proptosis of Graves ophthalmopathy. In an aneurysm, however, the episcleral veins are distended, a bruit is heard overlying the globe, and the patient hears noises that are synchronous with the pulse. Furthermore, medical imaging demonstrates an enlarged or distended superior ophthalmic vein due to arterial blood shunted by the aneurysm. Lymphomas may cause either bilateral or unilateral exophthalmos. A CT scan demonstrates retrobulbar masses that mold to the posterior aspect of the globe and to the muscle cone. Fine-needle aspiration biopsy under ultrasonographic guidance may yield the diagnosis.

TREATMENT
Treatment of the eye changes of Graves disease is primarily palliative. The patient should be reassured that the ophthalmopathy is usually self-limited. Above all, if thyroid dysfunction is present, the patient must be rapidly treated and maintained in a euthyroid state. Subsequent hypothyroidism should be avoided (see Chap. 45). Initially, during the acute phase of the disease, elevation of the head of the bed and diuretic use may reduce the periorbital edema. A controversy has existed for some time regarding whether any of the three treatments for Graves disease (antithyroid drugs, radioiodine therapy, or thyroidectomy) was associated with greater benefit in preventing onset or exacerbation of ophthalmopathy.13,27,28,29,30,31 and 32,53 Most of these earlier studies were poorly controlled. Three randomized clinical trials have suggested that ophthalmopathy may be more common after treatment with radioactive iodine than after medical or surgical treatment.29,39,41 A clear understanding by the patient that a protracted period of therapy is necessary helps to foster a good patient-doctor relationship. Encouragement that the therapy usually prevents loss of vision and maintains function is helpful for the patient's peace of mind. Many patients are concerned about cosmetic disfigurement caused by the disease. Reassurance that this also can be managed to a considerable degree builds confidence. Treatment of the ophthalmopathy is indicated for functional and cosmetic reasons, which frequently overlap.54 These disabilities relate to lid retraction, extraocular muscle myopathy, exophthalmos, protrusion of retrobulbar fat through the orbital septum, and optic nerve neuropathy.55 Functional disability may result from extraocular muscle malalignment causing diplopia, from vision loss because of corneal exposure owing to exophthalmos or lid retraction, or from optic nerve neuropathy. Also, soft-tissue involvement can cause severe discomfort. Cosmetic deficit may arise from upper eyelid retraction, eyeball malposition, exophthalmos, herniation of retrobulbar fat through the spaces of Charpey in the orbital septum, or soft-tissue involvement. Lid retraction may cause both a cosmetic and functional problem. Drying of the eye from exposure, especially during sleep, leads to keratitis with constant redness, foreign-body sensation, and photophobia. In the early stages, eyedrops of 5% guanethidine (Ismelin), a sympatholytic agent, may relieve lid retraction. This drug has not been approved by the U.S. Food and Drug Administration but is available for use in Europe. Some symptomatic relief may be obtained by increasing the humidity in the home environment, wearing sunglasses when outdoors, using artificial tears (1% methylcellulose) during the day, and applying emollients to lubricate the eyes at bedtime. If these measures fail to relieve the symptoms, or if the disfigurement is too great, one may turn to surgery. However, surgery should not be performed until the patient is stabilized in the euthyroid state. Myectomy of the Mller muscle is effective in relieving lid retraction in many cases (Fig. 43-12). The operation is not indicated, however, in cases of lid retraction secondary to tethering of the inferior rectus muscle (Collier sign) or when organic changes have occurred in the superior levator muscle. When this latter muscle becomes involved by morphologic changes, it becomes tethered so that the upper lid does not close, even under general anesthesia. A technique for testing for this is to grasp the upper eyelashes and pull inferiorly in a direction opposite the field of action of the levator muscle. When the result of this forced duction test is positive, myectomy of the Mller muscle will not alleviate the lid retraction. In this event, one must perform recession of the superior levator muscle with or without the use of some kind of donor material (preserved sclera, Tenon fascia, or preserved dura) as a spacer between the tarsus and muscle. In rare cases, the injection of a soluble corticosteroid (triamcinolone) adjacent to the levator, under ultrasonographic control, has been successful.

FIGURE 43-12. A, Right upper lid retraction in a 42-year-old woman with Graves hyperthyroidism. B, Postoperative appearance after myectomy of Mller muscle of right upper lid.

Tethering of the inferior rectus muscle causes not only vertical diplopia, which probably is the most common functional disability, but also malposition of the eyeball, which can be cosmetically disfiguring. Recession of the muscle and lysis of adhesions between it and the lower lid retractors often restores single binocular vision. Just as often, the lower lid sags postoperatively, which necessitates another operation to elevate the lower lid to prevent exposure of the lower one-third of the globe. Tethering of the medial rectus is not as common as inferior rectus tethering but is equally disabling. It simulates lateral rectus muscle palsy and causes horizontal diplopia and cosmetic disfigurement by eyeball malalignment (see Fig. 43-6). Surgical recession of the medial rectus muscles restores single binocular vision and corrects the disfigurement. Progressive soft-tissue involvement may cause marked functional and cosmetic deficits. These patients may be discomforted by pain, epiphora, and photophobia, becoming functionally disabled, although no vision loss may occur. They may be markedly disfigured, with exophthalmos, lid edema, chemosis edema of the caruncle, and conjunctival injection (see Fig. 43-1B). Ultimately, compressive neuropathy of the optic nerve at the orbital apex resulting from the enlarged extraocular muscles may ensue, and blindness becomes a distinct danger. Therapeutic options for progressive ophthalmopathy include (a) corticosteroid therapy, (b) supervoltage orbital radiotherapy, (c) surgical decompression of the orbit, and (d) combinations of these methods. IMMUNOMODULATORY THERAPY

CORTICOSTEROIDS A large number of immunomodulatory agents have been used in the treatment of Graves ophthalmopathy, but no single agent seems to have a clear advantage over corticosteroids.56 Corticosteroid use may produce dramatic relief of the signs and symptoms of progressive inflammatory changes. The dosage is tailored to the severity of involvement and may range from 40 to 120 mg per day of prednisone, which then is gradually tapered as improvement occurs. When the optic nerve becomes involved, larger doses of corticosteroids are needed to restore function and save vision. The patient may need to be hospitalized if large doses of prednisone are administered (e.g., 120140 mg per day). If adverse reactions occur, the dosage sometimes must be reduced and the treatment augmented with subtenon or retrobulbar injections of triamcinolone or methylprednisolone. As soon as the vision is restored (usually after 7 to 10 days of this therapy), the prednisone dosage is reduced. Sometimes, exacerbations occur when the dose level reaches 40 mg per day. In this event, larger doses once again must be administered. Pulse methylprednisolone treatment has been reported to be effective, with documented clinical improvement and reduced muscle size on CT scan.13 A prospective study indicates that corticosteroids may prevent worsening of ophthalmopathy in patients treated with radioiodine, but this remains controversial.39,41 CYCLOSPORINE Cyclosporine therapy was applied to Graves ophthalmopathy in view of the drug's immunomodulatory effects directed at several points in the presumed pathogenesis of the disease (see Table 43-1). The drug inhibits early induction of T helper cell proliferation, permits activation of T-suppressor cells, blocks activation of cytotoxic T cells and production of cytokines, and may also inhibit synthesis of TSH-receptor antibodies by B lymphocytes.57 Several early reports showed mixed but encouraging results, which led to two prospective randomized studies of note. A comparison of cyclosporine plus prednisone to prednisone alone in 40 patients showed that patients on combined treatment had a faster and more long-lasting response with fewer recurrences. Objective improvement attributed to cyclosporine included reduced eye muscle thickness.58 In a study of 36 patients randomized to receive either prednisone or cyclosporine, 61% of patients responded to prednisone alone, compared with only 22% who showed good responses to cyclosporine. However, 59% of the treatment failures responded to a second course of both drugs in combination.59 Thus, it appears that cyclosporine may provide an added benefit to corticosteroid therapy and shorten the course of the disease. The obvious drawbacks to both corticosteroids and cyclosporine are their potential adverse side effects. Renal function and liver function must be monitored during cyclosporine treatment, with the daily dose kept between 2.5 and 5 mg/kg of body weight. SOMATOSTATIN Trials of somatostatin therapy for ophthalmopathy were undertaken because of the detection of somatostatin receptors in orbital fibroblasts and lymphocytes. The mechanism of action is presumed to be based on reduction of edema mediated by insulin-like growth factor-I and reduction of autoreactivity of T cells via inhibition by somatostatin of proliferation of activated T cells. Octreotide has also been shown to reduce serum levels of intercellular adhesion molecule-1 (sICAM-1),60 thought to be a marker for orbital fibroblast activation that correlates with inflammatory changes.61 Various analogs of somatostatin, including indium-labeled octreotide and pentetreotide, have been used effectively with single photon emission computed tomography (SPECT) to scintigraphically image active ophthalmopathy.62,63,64,65 and 66 A drawback of somatostatin therapy had been the need to give multiple daily subcutaneous injections due to its brief half-life, but longer lasting analogs are now available. One such analog, lanreotide, is given only once every 2 weeks and has been associated with comparable clinical benefit in preliminary trials.67 In one comparison between somatostatin therapy and corticosteroid therapy, the former was as effective in relieving symptoms and soft-tissue inflammation, but not in reducing muscle size.68 On the other hand, a correlation may exist between intensity of octreotide uptake on scintiscan and clinical responsiveness to corticosteroids.69 OTHER IMMUNOMODULATORY AGENTS Plasma exchange or plasmapheresis with or without administration of immunosuppressive agents has been used without striking success. Other antiimmune therapies that have been used with mixed success include cyclophosphamide, ciamexone, and azathioprine as well as infusions of high-dose intravenous immunoglobulin. The results have not been promising.13 One advantage of intravenous immunoglobulin over cortico-steroids is the much lower incidence of adverse side effects.70,71 Strategies are currently under way for the development of cytokine antagonists to be applied to the pathophysiology underlying Graves ophthalmopathy.72 One limited clinical study describes a salutary effect achieved by pentoxifylline, a drug that in vitro has been shown to inhibit glycosaminoglycan synthesis and HLA-DR expression in orbital fibroblasts.73 SUPERVOLTAGE THERAPY Because of adverse side effects of corticosteroids and the inadvisability of long-term therapy, the physician may need to turn to supervoltage orbital radiotherapy or to surgical decompression. Indications for supervoltage therapy are (a) progression of exophthalmos under medical observation, (b) functional incapacitation because of progressive soft-tissue signs, (c) failure of corticosteroid therapy, (d) optic nerve neuropathy, and (e) as an adjunct to surgical decompression. The radiotherapy technique uses a 4-MeV linear accelerator to deliver 1800 to 2000 rad over 2 weeks.74 In an extensive series of 311 patients (one-third of whom also received corticosteroids), 80% had soft-tissue improvement, 51% experienced a reduction in proptosis, 67% had improved vision, and 56% showed some improvement in eye muscle function.75 A combination of prednisone, 40 to 80 mg, and supervoltage radiotherapy may be used as the initial treatment. Soft-tissue signs subside in an average of 6 weeks, and maximal improvement occurs by 3 months (Fig. 43-13). Patients who have short-term disease have the best response, whereas patients who have more chronic disease with significant proptosis and muscle dysfunction due to fibrosis are unlikely to receive any benefit. A diminution in exophthalmos and improvement in muscle balance may be seen, but approximately one-third of patients still require muscle surgery. Lid signs and herniation of retrobulbar fat through the orbital septum do not improve. Also, orbital radiotherapy is ineffective for proptosis without muscle enlargement or for long-standing ophthalmoplegia. Nevertheless, radiation is a safe, effective therapy that may reduce the time required to attain sufficient stability to allow surgical intervention. In one study76 of the effects of orbital radiation plus corticosteroid therapy compared to corticosteroid use alone, the combination therapy was far more effective, particularly in those patients with disease of relatively short duration. An experienced radiotherapist should supervise this treatment to minimize the hazards of radiation retinitis and optic nerve necrosis.

FIGURE 43-13. A, This 48-year-old man with Graves ophthalmopathy was vocationally incapacitated because of involvement of soft tissue and extraocular muscle. Note marked hyperemia and widened palpebral fissures in this view with eyes fixed in primary gaze. On attempted upward gaze, a marked restriction of elevation occurred because of tethered inferior rectus muscles in each eye. B, Transaxial computed tomographic (CT) scan of the patient. Note characteristic enlargement of extraocular muscles noted in endocrine ophthalmopathy. C, Appearance in primary gaze 3 months after orbital radiotherapy. The patient received 2000 rad in 10 fractional doses delivered over a period of 2 weeks by a 6-MeV linear accelerator using the technique of Kriss et al.74 Note diminution of hyperemia and soft-tissue signs. Almost full extraocular motility was restored, and the patient was vocationally rehabilitated. D, The transaxial CT scan after orbital radiotherapy demonstrates marked diminution in size of extraocular muscles.

The efficacy of orbital radiation therapy is putatively due to the radiosensitivity of lymphocytes infiltrating the orbit as well as to reduced proliferation and glycosaminoglycans production by orbital fibroblasts. The most remarkable improvement occurs within 1 year of onset or aggravation of the ophthalmopathy; it is primarily manifest in soft-tissue changes and minimally reduced proptosis, with little beneficial effect on the function of extraocular muscles. Such improvement is associated with the ability to discontinue use of corticosteroids within a few months in as many as 75% of patients. However, approximately one-third of patients still require posttreatment corrective eye surgery. The excellent results observed when systemic corticosteroids are administered with retrobulbar irradiation76 have led to some popularization of this combined modality of therapy in the past decade. In a randomized double-blind study of orbital irradiation versus corticosteroid use,77 after 6 months, significant improvement was noted in

approximately one-half of both treatment groups, results that lead to the conclusion that the therapies were equally effective. Follow-up for longer than 6 months would be of interest because one group has reported that the acute improvement in soft-tissue signs after irradiation may not translate into any apparent long-term benefit.78 Proposed indications for retrobulbar irradiation and guidelines for management13 will have to remain acceptable until availability of more definitive results dictates otherwise. SURGICAL DECOMPRESSION OF ORBIT Relatively conservative orbital decompression can be accomplished with minimal invasiveness by removal of orbital fat.79,80 More aggressive surgical decompression of the orbit may need to be the next step in therapy when fat removal and medical therapy with corticosteroids fail. This operation should be reserved for (a) severe proptosis with corneal exposure or ulceration, (b) optic neuropathy unresponsive to corticosteroid or orbital radiotherapy, (c) marked optic neuropathy in patients who cannot tolerate corticosteroids, and (d) cosmesis. Historically, several decompressive operations of the orbit have been used, in which different regions of the bony orbit have been removed (Fig. 43-14). Currently, many ophthalmologists perform surgical decompression by a transorbital approach (Fig. 43-15) in which three walls of the orbit are removed; the lateral wall, the floor, and the medial wall of the orbit may be removed in one stage. The otolaryngologist performs decompression by the transantral approach, removing the floor and medial wall.81 Both of these approaches decompress the orbital contents into the paranasal sinuses. The transantral route allows better decompression of the orbital apex than does the transorbital procedure. Some neurosurgeons decompress the orbit by removal of the roof through a transcranial or coronal approach.82 Both orbits may be decompressed in one stage by either the transcranial or transantral technique, if indicated. In the transorbital route, the surgeon usually decompresses one orbit, followed by decompression of the other, if needed, after 1 week (see Chap. 215).

FIGURE 43-14. Schematic representation of region of bony orbit removed in several historical decompressive procedures.

FIGURE 43-15. Commonly used orbital decompression procedures. A, Transorbital approach, followed by B, removal of the lateral wall, medial wall, and floor. C, Transantral approach, involving a buccogingival incision, followed by D, removal of the floor and medial wall of the orbit through the antrum and ethmoids. A, copyrighted by Emory, 1983.

A comparable degree of decompression has been reported with a transnasal endoscopic technique.83 Chronic sinusitis is a contraindication to decompression into the paranasal sinuses. Restoration of visual acuity is comparable after either transantral or transcranial decompression, whereas the transorbital technique may require adjunctive orbital radiotherapy. All are equally effective in diminishing proptosis. The transcranial approach has the highest morbidity. Transantral decompression is followed by increased diplopia caused by entrapment of the medial rectus muscles into the ethmoid sinuses in approximately two-thirds of cases, which requires further muscle surgery. Cerebrospinal fluid leaks are reported in a small percentage of patients undergoing transantral decompression resulting from penetration of the cribriform plate. Damage to the infraorbital sensory nerve may occur with either the transantral or orbital approach, resulting in permanent anesthesia of the upper lip and gum. Pain in the jaw when eating or chewing frequently occurs if the temporalis muscle is severed when performing a lateral wall resection during the transorbital approach. Pulsation of the orbital structures may be noted after transcranial decompression. Pneumoproptosis may occur with either transorbital or transantral decompression if an ethmoid air cell ruptures into the orbital tissues after removal of the medial wall. To avoid this frightening experience, patients undergoing this procedure should be warned not to blow their noses forcefully. However, the air rapidly absorbs, and the proptosis subsides. A history of sudden proptosis and crepitation with confirmation by CT scan confirms this diagnosis. CHAPTER REFERENCES
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. Brain R. Pathogenesis and treatment of endocrine exophthalmos. Lancet 1959; 1:109. McKenzie JM. Humoral factors in the pathogenesis of Graves' disease. Physiol Rev 1968; 48:252. Farid NR, Balazs C. The genetics of thyroid associated ophthalmopathy. Thyroid 1998; 8:407. Werner SC. Modification of the classification of the eye changes of Graves' disease: recommendations of the Ad Hoc Committee of the American Thyroid Association. J Clin Endocrinol Metab 1977; 44:203. Bartley GB. Evolution of classification systems for Graves' ophthalmopathy. Ophthal Plas Reconstr Surg 1995; 11:229. Kendall-Taylor P, Perros P. Clinical presentation of thyroid associated orbitopathy. Thyroid 1998; 8:427. Perros P, Kendall-Taylor P. Natural history of thyroid eye disease. Thyroid 1998; 8:423. Bartley GB, Fatourechi V, Kkadrmas EF, et al. Clinical features of Graves' ophthalmopathy in an incidence cohort. Am J Ophthal 1996; 121:284. Bartley GB, Fatourechi V, Kkadrmas EF, et al. The chronology of Graves' ophthalmopathy in an incidence cohort. Am J Ophthal 1996; 121:426. Bartley GB, Fatourechi V, Kkadrmas EF, et al. Long-term follow-up of Graves' ophthalmopathy in an incidence cohort. Ophthal 1996; 103:958. Bartley GB, Gorman CA. Diagnostic criteria for Graves' ophthalmopathy. Am J Ophthal 1995; 119:792. Gorman CA. The measurement of change in Graves' ophthalmopathy. Thyroid 1998; 8:539. Burch HB, Wartofsky L. Graves' ophthalmopathy: current concepts regarding pathogenesis and management. Endocr Rev 1993; 14:747. Bahn RS, Heufelder AE. Pathogenesis of Graves' ophthalmopathy. N Engl J Med 1993; 329:1468. Gorman CA. Pathogenesis of Graves' ophthalmopathy. Thyroid 1994; 4:379. Ludgate M, Crisp M, Lane C, et al. The thyrotropin receptor in thyroid eye disease. Thyroid 1998; 8:411. Bahn RS, Dutton CM, Heufelder AE, Sarkar G. A genomic point mutation in the extracellular domain of the thyrotropin receptor in patients with Graves' ophthalmopathy. J Clin Endocrinol Metab 1994; 78:256. Burch HB, Sellitti D, Barnes SG, et al. Thyrotropin receptor antisera for the detection of immunoreactive protein species in retroocular fibroblasts obtained from patients with Graves' ophthalmopathy. J Clin Endocrinol Metab 1994; 78:1384. Baker JR Jr. Dissecting the immune response to the thyrotropin receptor in autoimmune thyroid disease. (Editorial). J Clin Endocrinol Metab 1993; 77:16. Davies TF. The thyrotropin receptors spread themselves around. (Editorial). J Clin Endocrinol Metab 1994; 79:1232. Spitzweg C, Joba W, Hunt N, Heufelder AE. Analysis of human thyrotropin receptor gene expression and immunoreactivity in human orbital tissue. Eur J Endocrinol 1997; 136:599. Burman KD. Graves' ophthalmopathy: what is the initial abnormality? Eur J Endocrinol 1997; 136:583. Paschke R, Metcalfe A, Alcalde L, et al. Presence of nonfunctional thyrotropin receptor variant transcripts in retroocular and other tissues. J Clin Endocrinol Metab 1994; 79:1234. Heufelder AE, Bahn RS. Soluble intercellular adhesion molecule-1 (sICAM-1) in sera of patients with Graves' ophthalmopathy and thyroid diseases. Clin Exp Immunol 1993; 92:296. Bahn RS. Cytokines in thyroid eye disease: potential for anticytokine therapy. Thyroid 1998; 8:415. Kahaly G, Forster G, Hansen C. Glycosaminoglycans in thyroid eye disease. Thyroid 1998; 8:429. Marcocci C, Bartalena L, Bogazzi F, et al. Relationship between Graves' ophthalmopathy and type of treatment of Graves' hyperthyroidism. Thyroid 1992; 2:171. Manso PG, Furlanetto RP, Wolosker AMB, et al. Prospective and controlled study of ophthalmopathy after radioiodine therapy for Graves' hyperthyroidism. Thyroid 1998; 8:49. Tallstedt L, Lundell G, Torring O, et al. Occurrence of ophthalmopathy after treatment for Graves' hyperthyroidism. N Engl J Med 1992; 326:1733.

30. Kung AWC, Yau CC, Cheng A. The incidence of ophthalmopathy after radioiodine therapy for Graves' disease: prognostic factors and the role of methimazole. J Clin Endocrinol Metab 1994; 79:542. 31. Burmeister LA, Beatty RL, Wall JR. Malignant ophthalmopathy presenting one week after radioiodine treatment of hyperthyroidism. Thyroid 1999; 9:189. 32. Soliman M, Kaplan E, Abdel-Latif A, et al. Does thyroidectomy, radioactive iodine therapy, or antithyroid drug treatment alter reactivity of patients' T cells to epitopes of thyrotropin receptor in autoimmune thyroid diseases? J Clin Endocrinol Metab 1995; 80:2312. 33. Gunji K, Kubota S, Swanson JIL, et al. Role of the eye muscles in thyroid eye disease: identification of the principal autoantigens. Thyroid 1998; 8:553. 34. Burch HB, Lahiri S, Bahn RS, Barnes S. Superoxide radical production stimulates retroocular fibroblast proliferation in Graves' ophthalmopathy. Exp Eye Res 1997; 65:311. 35. Bertelsen JB, Hegedus L. Cigarette smoking and the thyroid. Thyroid 1994; 4:327. 36. Bartalena L, Marcocci C, Tanda ML, et al. Cigarette smoking and treatment outcomes in Graves' ophthalmopathy. Ann Intern Med 1998; 129:632. 37. Hofbauer LC, Muhlberg T, Konig A, et al. Soluble IL-1 receptor antagonist serum levels in smokers and nonsmokers with Graves' ophthalmopathy undergoing orbital radiotherapy. J Clin Endocrinol Metab 1997; 82:2244. 38. Fernandez-Sanchez JR, Pradas JR, Martinez OC, et al. Graves' ophthalmopathy after subtotal thyroidectomy and radioiodine therapy. Br J Surg 1993; 80:1134. 39. Bartalena L, Marcocci C, Bogazzi F, et al. Relation between therapy for hyperthyroidism and the course of Graves' ophthalmopathy. N Engl J Med 1998; 338:73. 40. Sridama V, DeGroot LJ. Treatment of Graves' disease and the course of ophthalmopathy. Am J Med 1989; 87:70. 41. Bartalena L, Marcocci C, Bogazzi F, et al. Use of corticosteroids to prevent progression of Graves' ophthalmopathy after radioiodine therapy for hyperthyroidism. N Engl J Med 1989; 321:349. 42. Gorman CA. Temporal relationship between onset of Graves' ophthalmopathy and diagnosis of thyrotoxicosis. Mayo Clin Proc 1983; 58:515. 43. Wiersinga WM, Smit T, VanderGaag R, Koorneef L. Temporal relationship between onset of Graves' ophthalmopathy and onset of thyroidal Graves' disease. J Endocrinol Invest 1988; 11:615. 44. Hufnagel TH, Hickey WF, Cobbs WH, et al. Immunohistochemical and ultrastructural studies on the exenterated orbital tissues of a patient with Graves' disease. Ophthalmology 1984; 91:1411. 45. Trokel SL, Jakobiec FA. Correlation of CT scanning and pathologic features of ophthalmic Graves' disease. Ophthalmology 1981; 88:553. 46. Trobe JD. Optic nerve involvement in dysthyroidism. Ophthalmology 1981; 88:488. 47. Gamblin GT, Harper DG, Galentine P, et al. Screening for elevated intraocular pressure in Graves' disease: evidence of frequent subclinical ophthalmology. N Engl J Med 1983; 208:420. 48. Salvi M, Spaggiari E, Neri F, et al. The study of visual evoked potentials in patients with thyroid-associated ophthalmopathy identifies asymptomatic optic nerve involvement. J Clin Endocrinol Metab 1997; 82:1027. 49. Hauer J. Additional clinical signs of unilateral endocrine exophthalmos. Br J Ophthalmol 1969; 53:210. 50. Bartley GB. The differential diagnosis and classification of eyelid retraction. Ophthalmol 1996; 103:168. 51. Bailey CC, Kabala J, Laitt R, et al. Magnetic resonance imaging in thyroid eye disease. Eye 1996; 10:617. 52. Nianiaris N, Hurwitz JJ, Chen JC, Wortzman G. Correlation between computed tomography and magnetic resonance imaging in Graves' orbitopathy. Can J Ophthalmol 1994; 29:9. 53. DeGroot JJ, Gorman CA, Pinchera A, et al. Therapeutic controversies: radiation and Graves' ophthalmopathy. J Clin Endocrinol Metab 1995; 80:339. 54. Bahn RS, Gorman CA. Choice of therapy and criteria for assessing treatment outcome in thyroid-associated ophthalmopathy. Endocrinol Metab Clin North Am 1987; 16:391. 55. Tallstedt L. Surgical treatment of thyroid eye disease. Thyroid 1998; 8:447. 56. Prummel MF, Wiersinga WM. Immunomodulatory treatment of Graves' ophthalmopathy. Thyroid 1998; 8:545. 57. Wiersinga WM. Novel drugs for the therapy of Graves' ophthalmopathy. In: Wall JR, How J, eds. Graves' ophthalmopathy. Cambridge, MA: Black-well Scientific, 1990:111. 58. Kahaly G, Schrezenmeir J, Krause U, et al. Cyclosporine and prednisone in treatment of Graves' ophthalmopathy: a controlled, randomized and prospective study. Eur J Clin Invest 1986; 16:415. 59. Prummel MF, Mourits MP, Berghout A, et al. Prednisone and cyclosporine in the treatment of severe Graves' ophthalmopathy. N Engl J Med 1989; 321:1353. 60. Ozata M, Bolu E, Sengul A, et al. Effects of octreotide treatment on Graves' ophthalmopathy and circulating sICAM-1 levels. Thyroid 1996; 6:283. 61. DeBellis A, SiMartino S, Fiordelisi F, et al. Soluble intercellular adhesion molecule-1 (sICAM-1) concentrations in Graves' disease patients followed up for development of ophthalmopathy. J Clin Endocrinol Metab 1998; 83:1222. 62. Kahaly GJ, Gorges R, Diaz M, et al. Indium-111-pentreotide in Graves' disease. J Nucl Med 1998; 39:533. 63. Wiersinga WM, Gerding MN, Prummel MF, Krenning EP. Octreotide scintigraphy in thyroidal and orbital Graves' disease. Thyroid 1998; 8:433. 64. Postema PTE, Krenning EP, Wijngaarde R, et al. [111 In-DTPA-D-phe1]-Oct-reotide scintigraphy in thyroidal and orbital Graves' disease: a parameter for disease activity. J Clin Endocrinol Metab 1994; 79:1845. 65. Kahaly GJ, Forster GJ. Somatostatin receptor scintigraphy in thyroid eye disease. Thyroid 1998; 8:549. 66. Krassas GE. Somatostatin analogues in the treatment of thyroid eye disease. Thyroid 1998; 8:443. 67. Krassas GE, Kaltsas T, Dumas A, et al. Lanreotide in the treatment of patients with thyroid eye disease. Eur J Endocrinol 1997; 136:416. 68. Kung AWC, Michon J, Tai KS, Chan FL. The effect of somatostatin versus corticosteroid in the treatment of Graves' ophthalmopathy. Thyroid 1996; 6:381, 489. 69. Colao A, Lastoria S, Ferone D, et al. Orbital scintigraphy with [111 In-diethylenetriamine pentaacetic acid-D-phe 1]-Octreotide predicts the clinical response to corticosteroid therapy in patients with Graves' ophthalmopathy. J Clin Endocrinol Metab 1998; 83:3790. 70. Kahaly GJ, Pitz S, Muller-Forell W, Hommel G. Randomized trial of intravenous immunoglobulins versus prednisolone in Graves' ophthalmopathy. Clin Exp Immunol 1996; 106:197. 71. Baschieri L, Antonelli A, Nardi S, et al. Intravenous immunoglobulin versus corticosteroid in treatment of Graves' ophthalmopathy. Thyroid 1997; 7:579. 72. Bartalena L, Marcocci C, Pinchera A. Cytokine antagonists: new ideas for the management of Graves' ophthalmopathy. (Editorial). J Clin Endocrinol Metab 1996; 81:446. 73. Balazs C, Kiss E, Vamos A, et al. Beneficial effect of pentoxifylline on thyroid associated ophthalmopathy: a pilot study. J Clin Endocrinol Metab 1997; 82:1999. 74. Kriss JP, McDougall IR, Donaldson SS. Graves ophthalmopathy. In: Krieger DT, Bardin W, eds. Current therapy in endocrinology 1983-84. New York: Decker, 1984:104. 75. Peterson IA, Kriss JP, McDougall IR, Donaldson SS. Prognostic factors in the radiotherapy of Graves' ophthalmopathy. Int J Radiat Oncol Biol Phys 1990; 19:259. 76. Bartalena L, Marcocci C, Chiovato L, et al. Orbital cobalt irradiation combined with systemic corticosteroids for Graves' ophthalmopathy: comparison with systemic corticosteroids alone. J Clin Endocrinol Metab 1983; 56:1139. 77. Prummel MF, Mourits MP, Blank L, et al. Randomized double-blind trial of prednisone versus radiotherapy in Graves' ophthalmopathy. Lancet 1993; 342:949. 78. Kao SCS, Kendler DL, Nugent RA, et al. Radiotherapy in the management of thyroid orbitopathy. Arch Ophthalmol 1993; 111:819. 79. Trokel S, Kazim M, Moore S. Orbital fat removal: decompression for Graves' ophthalmopathy. Ophthalmology 1993; 100:674. 80. Adenis JP, Robert PY, Lasudry JG, Dalloul Z. Treatment of proptosis with fat removal orbital decompression in Graves' ophthalmopathy. Eur J Ophthal 1998; 8:246. 81. Garrity JA, Fatourechi V, Bergstralh EJ, et al. Results of transantral orbital decompression in 428 patients with severe Graves' ophthalmopathy. Am J Ophthal 1993; 116:533. 82. Kalmann R, Mourits MP, van der Pol JP, Koornneef L. Coronal approach for rehabilitative orbital decompression in Graves' ophthalmopathy. Br J Ophthalmol 1997; 81:41. 83. Lund VJ, Larkin G, Fells P, Adams G. Orbital decompression for thyroid eye disease: a comparison of external and endoscopic techniques. J Laryngol Otol 1997; 111:1051.

CHAPTER 44 SURGERY OF THE THYROID GLAND Principles and Practice of Endocrinology and Metabolism

CHAPTER 44 SURGERY OF THE THYROID GLAND

EDWIN L. KAPLAN Preparation for Surgery Hypothyroidism Hyperthyroidism Surgical Approach to Thyroid Nodules Non-Irradiated Patients Irradiated Patients Surgical Approach to Thyroid Cancer Papillary Carcinoma Follicular Carcinoma Anaplastic Carcinoma Medullary Thyroid Carcinoma Operative Technique for Thyroidectomy Postoperative Complications Thyroid Storm Wound Hemorrhage Injury to the Recurrent Laryngeal Nerve Hypoparathyroidism Chapter References

The modern era of thyroid surgery began in the 1860s with the work of Billroth and colleagues.1 Operative techniques and results were so greatly advanced by Kocher2 and others that, by 1920, Halsted3 referred to thyroidectomy as a feat which today can be accomplished by any competent operator without danger of mishap. Decades later, some complications still occur. Nevertheless, with advances in diagnosis and surgical technique, thyroidectomy is one of the safest and most effective major operations performed today.

PREPARATION FOR SURGERY

Most patients undergoing thyroid surgery are euthyroid and require no specific preoperative preparation other than standard preoperative examination and a few tests. A serum calcium value should be determined, and fiberoptic laryngoscopy should be performed, especially in those patients who have had previous thyroid or parathyroid operations, to detect an unrecognized recurrent laryngeal nerve injury.

HYPOTHYROIDISM
Myxedema creates great potential for morbidity and even mortality from the effects of both the anesthesia and the operation.4 Severely hypothyroid patients have a higher incidence of perioperative hypotension, gastrointestinal hypomotility, prolonged anesthetic recovery, and neuropsychiatric disturbances. In addition, they are very sensitive to preoperative medication. Thus, when severe myxedema is present, deferring elective surgery until a euthyroid state is achieved is preferable. Urgent surgery need not be postponed simply for repletion of thyroid hormone, however. Guidelines for thyroxine therapy appear in Chapter 45.

HYPERTHYROIDISM
A number of different regimens are available for the preoperative preparation of patients with thyrotoxicosis; for example, iodine alone; propylthiouracil (PTU) plus iodine; PTU or methimazole (Tapazole) plus thyroxine plus iodine; PTU plus propranolol plus iodine; propranolol plus iodine; and propranolol alone. The author's patients are typically prepared for operation by the use of PTU or methimazole, plus iodine, with or without the addition of propranolol. Each treatment has advantages and disadvantages; however, the preoperative restoration of euthyroidism is recommended.5 Most patients are treated initially with an antithyroid drug, PTU or methimazole, until they approach a normal thyroid state. Many surgeons then prefer to administer several drops of iodine (saturated solution of potassium iodide, or SSKI), two to three times a day, for 8 to 10 days or longer before surgery. The iodine decreases the vascuarity and increases the firmness of the gland. The b-adrenergic receptor blockers, such as propranolol, are useful in the control of tremor, tachycardia, and cardiac arrhythmias. Many surgeons have had excellent results with the use of propranolol, alone6 or with iodine, as preoperative treatment of Graves disease, citing ease and speed of preparation as the major advantages.7 However, reliance on this drug alone may be hazardous,8 and reserving the use of propranolol alone (or with iodine) for selected patients who are allergic to antithyroid medications may be wise. When patients are treated with propranolol alone preoperatively, this drug must be continued into the postoperative period as well, because such patients are still thyrotoxic at that time. The medical management of Graves disease has been discussed in detail in Chapter 42. However, the advantages and disadvantages of the two ablative approaches to treatment (i.e., surgery or radioiodine therapy) are listed in Table 44-1. The author uses subtotal thyroidectomy in young patients with Graves disease, primarily because it rapidly corrects the thyrotoxic state with decreased risk of hypothyroidism, while avoiding the potential risks of radioiodine therapy.5 Operations include bilateral subtotal lobectomies or unilateral lobectomy with contralateral subtotal lobectomy. To minimize the recurrence of disease, only several grams of thyroid tissue are left. When the surgery is to treat ophthalmopathy, however, a near total or total thyroidectomy should be performed.

TABLE 44-1. Ablative Treatment of Graves Disease with Thyrotoxicosis

SURGICAL APPROACH TO THYROID NODULES

NON-IRRADIATED PATIENTS Fine-needle aspiration with cytologic examination is heavily relied on when choosing patients for surgery. Generally, any nodule suspected of being a carcinoma should be completely removed, along with surrounding tissue; thus, a total lobectomy (or lobectomy with isthmusectomy) is the initial operation of choice in most patients. A frozen section should be obtained intraoperatively. If the diagnosis of a colloid nodule is made, the surgery is terminated. If a diagnosis of an adenoma is made, then difficulty arises of differentiating, on frozen section, a follicular adenoma from a follicular carcinoma, or a benign Hrthle cell tumor from a Hrthle cell carcinoma.9 These diagnoses require the careful assessment of cellular morphology as well as capsular and vascular invasion, which are often difficult to evaluate on frozen section analyses. To aid in the diagnosis, enlarged lymph nodes of the central compartment are sampled and a biopsy of the jugular nodes is also performed. If the results are negative, two options exist: (a) stopping the surgery after lobectomy, with the understanding that a second operation may be necessary to complete the thyroidectomy if a carcinoma is ultimately found, or (b) performing a subtotal resection of the contralateral lobe. This latter approach is often used if the patient consents, particularly

when a preoperative needle aspiration suggests that a follicular lesion will be encountered intra operatively (see Chap. 39).10 IRRADIATED PATIENTS In patients who have been exposed to low-dose, external radiation to the head and neck during infancy, childhood, or adolescence, because of the frequency of bilateral disease, the known coincidence of benign and malignant nodules in the same gland, and the prevalence of papillary carcinoma in 35% to 40% of patients, a near total resection of the thyroid gland with a biopsy of the jugular nodes is usually performed, even if a frozen section of the dominant nodule is benign.11 This therapy is thought to be advantageous because the remaining thyroid remnant of these patients can usually be ablated with one 30- mCi dose of radioiodine given on an outpatient basis if a carcinoma is found later.12 Of course, if a carcinoma is diagnosed on frozen section, a total thyroidectomy is attempted. In any event, these patients require therapy with thyroid hormone. Patients who have received high-dose radiation to the thyroid bed, such as those treated with mantle radiation for Hodgkin disease, are also at greater risk for developing thyroid carcinomas years later; they should be followed carefully and their disease should be treated aggressively.

SURGICAL APPROACH TO THYROID CANCER

PAPILLARY CARCINOMA The surgical treatment of papillary carcinoma (see Chap. 40) is best divided into two categories, based on the clinical characteristics and virulence of these lesions. MINIMAL PAPILLARY CARCINOMA The term minimal papillary carcinoma refers to a small papillary tumor, usually <5 mm in diameter, that demonstrates no local invasiveness through the thyroid capsule, is not associated with lymph node metastases, and is often found in a young person as an occult lesion when thyroidectomy has been performed for another benign condition.13 In such instances, lobectomy is sufficient and repeated surgeries are unnecessary; thyroid hormone is given to suppress serum thyroid-stimulating hormone (TSH) levels, and the patient is followed at regular intervals. STANDARD TREATMENT FOR MOST PAPILLARY CARCINOMAS Most papillary carcinomas are neither minimal nor occult. Papillary cancers are known to be microscopically multicentric in up to 80% of cases, occasionally to invade locally into the trahea or esophagus,14 to metastasize commonly to lymph nodes and later to the lungs and other tissues, and to recur clinically in the other thyroid lobe in 7% to 18% of patients if treated by only thyroid lobectomy.15,16 In the United States, ~1200 deaths occur each year from thyroid cancer and half of these deaths occur in patients with anaplastic thyroid carcinoma. Importantly, most patients with papillary thyroid cancer have a very good prognosis, particularly those who are young. Although the best treatment for most papillary cancers is near total or total thyroidectomy, with appropriate central and lateral neck dissections when nodes are involved, the surgeon with limited experience should not perform total thyroidectomy unless capable of achieving a low incidence of recurrent laryngeal nerve injuries and permanent hypoparathyroidism. Otherwise, referral of these patients to a major medical center where such expertise is available may be advisable. Often, patients are treated with radioactive iodine postoperatively17 (see Chap. 40). External irradiation and chemotherapy17 are sometimes used for those special cases of severe, extensive disease that do not accumulate radioiodine.18 In some studies,19 patients receiving total or near-total thyroidectomy with radioiodine therapy for papillary cancers of 1 cm or larger had decreased mortality and recurrence of disease compared with those receiving lesser operations and no radioactive iodine therapy. Studies have attempted to predict the potential aggressiveness of a given papillary carcinoma by determination of nuclear DNA measurement20 as well as by formulations of a set of risk factors, including the age of the patient, the size of the tumor, presence or absence of local invasion or distant metastases, the tumor grade, and whether or not all tumor could be resected.16,21,22 Clearly, young patients with small tumors that can be totally resected and that exhibit no local invasion or distant metastases have an excellent prognosis (approximately a 2% mortality from their papillary cancer). Older age, larger tumor size, local tumor invasiveness, presence of distant metastases, and nonresectable tumor are factors that portend a much greater cancer mortality. Approximately 80% of all papillary cancers fall in the low-risk category by these criteria, but 20% are much more aggressive. FOLLICULAR CARCINOMA The prevalence of true follicular cancers has greatly diminished in recent years, perhaps due to an increase in iodine in the diet. Multicentricity of follicular carcinoma within the thyroid and lymph node metastases is less common than with papillary cancer. Metastatic spread occurs to the lungs, bones, and other peripheral tissues by hematogenous dissemination. The ideal operation for follicular carcinoma is similar to that for papillary cancer, but the rationale differs. Small lesions with only minimal capsular invasion and no vascular invasion have a very good prognosis and require only lobectomy. Near-total or total thyroidectomy should be performed not for multicentricity but to make possible a later total body scan with radio iodine. If peripheral metastases are detected, they should be treated with high-dose radioiodine therapy.17 In patients with either papillary or follicular cancer, if lymph node metastases are present in the lateral neck, a modified radical neck dissection should be performed (see Chap. 40). Prophylactic modified radical neck dissections, in the absence of clinical disease, should not be performed for differentiated thyroid cancers. ANAPLASTIC CARCINOMA Anaplastic thyroid carcinoma remains one of the most virulent of all malignancies. Survival for most of these patients is measured in months. The type previously termed small cell is now considered to be a lymphoma by most pathologists and is treated by a combination of surgery, external radiation, and chemotherapy.23 In treating lymphoma of the thyroid, major surgical resection has not been shown to be of therapeutic benefit. Therefore, operative management is usually limited to obtaining an adequate biopsy. The large cell type may present as a solitary thyroid nodule early in its clinical course and, if surgery is undertaken at that time, a near-total or total thyroidectomy should be performed, with appropriate central and lateral neck dissections. However, most commonly, the carcinoma is advanced when the patient is first evaluated. In these patients, surgical cure is unlikely, no matter how aggressive; in particularly advanced cases, diagnosis by needle biopsy or by small open biopsy may be all that is appropriate. Sometimes, the isthmus can be divided to relieve tracheal compression, or a tracheostomy might be beneficial. Most treatment, however, is by external radiotherapy or chemotherapy or both.24 Hyperfractionation external radiation therapy that uses several treatments per day has some enthusiasts. Radioiodine treatment is ineffective because tumor uptake is absent. Although some success has been observed with doxorubicin, prolonged remissions are rarely achieved and multidrug regimens and combinations of chemotherapy with radiotherapy are being tried.18 Currently, the author's protocol involves preoperative treatment with a chemotherapy regimen of cisplatin, hydroxyurea, and Taxol with external radiation therapy of 7500 rads. Total thyroidectomy follows if no distant metastases are present. 25 (Also see Chap. 41.) MEDULLARY THYROID CARCINOMA Medullary thyroid carcinoma is a C-cell, calcitonin-producing tumor that contains amyloid or an amyloid-like substance, may secrete other peptides and amines, and may be transmitted in a familial pattern (see Chap. 40 and Chap. 188). This tumor, or its precursor, C-cell hyperplasia, occurs as a part of the multiple endocrine neoplasia type 2 syndromes (MEN2A and MEN2B), as well as the less common familial medullary thyroid cancer syndrome.26,26a,26b Hence, patients with medullary thyroid carcinoma should be screened for hyperparathyroidism and pheochromocytoma. If a pheochromocytoma or adrenal medullary hyperplasia is present, this should be operated on first, because it represents the greatest immediate risk. Medullary thyroid carcinoma is unilateral in sporadic cases and bilateral (often with C-cell hyperplasia when familial). It spreads to lymph nodes of the neck and mediastinum and later disseminates to the lungs, bone, liver, and elsewhere. The tumor is relatively radioresistant, it does not take up radioiodine, and it is not responsive to thyroid hormone suppression. Hence, an aggressive surgical approach is mandatory. The operation of choice for medullary thyroid carcinoma is total thyroidectomy with an extensive central compartment dissection down into the mediastinum. Lateral nodes are sampled. If central or lateral lymph nodes contain tumor, then appropriate unilateral or bilateral modified radical neck dissections are required. Repeated surgeries for meta-static tumor based on catheterization studies with blood sampling and calcitonin determinations alone have been proposed by some groups.27 Others believe that surgery should be repeated only after localization of metastatic disease, usually by magnetic resonance imaging or computed tomographic scanning.28 Octreotide or thallium scanning and positron emission tomography are useful in localizing metastatic diseases.29,30 Extensive reoperations of the central and lateral neck compartments have rendered 25% to 30% of patients eucalcitonemic.27,31 Some surgeons recommend laparoscopic evaluation of the liver to rule out metastatic disease before repeat operations in the neck are performed.31

Cure has been shown to be most frequent in young patients with familial disease whose disease was found by calcitonin screening26 or by screening of the RET oncogene.32,33 Prophylactic thyroidectomy is now being practiced at age 5 in patients with MEN2A who exhibit a mutated RET oncogene. Children of families with MEN2B who exhibit a mutated RET oncogene are given prophylactic surgery at an earlier age. Patients with MEN2A have a better prognosis than those with sporadic tumor, but patients with MEN2B have very aggressive tumors and often do not survive past age 40 years. In some patients, postoperative radiotherapy has been used, but most clinicians reserve this treatment for disease that cannot be resected. In MEN2 syndromes, bilateral adrenalectomy is thought to be appropriate by some investigators if a pheochromocytoma or adrenal medullary hyperplasia is diagnosed, because each is likely to be a bilateral condition (see Chap. 188). Others believe that, especially in young children with an apparent unilateral pheochromocytoma, only the affected gland should be removed first, with the knowledge that the other gland may have to be removed in the future. This method has the advantage of permitting the child to grow with normal adrenal cortical function for a period of time. One report34 supports unilateral adrenalectomy, because 48% of patients did not develop a pheochromocytoma in the other gland at 5.2 years of follow-up and no hypertensive crises occurred in these individuals. Furthermore, 23% of patients had an addisonian crisis after bilateral adrenalectomy with the occurrence of one death. Subtotal parathyroidectomy is necessary if primary hyperparathyroidism is diagnosed in MEN2A patients, because parathyroid hyperplasia is almost always present (see Chap. 58 and Chap. 62).

OPERATIVE TECHNIQUE FOR THYROIDECTOMY35,36 and 37

While under general endotracheal anesthesia, the patient is placed in a supine position with the neck extended. A low collar incision is made and is carried down through the subcutaneous tissue and platysma muscle. Superior and inferior flaps are developed, and the strap muscles are divided vertically in the midline and retracted laterally. The thyroid lobe is bluntly dissected free from its investing fascia and rotated medially. The thyroid isthmus is often divided early in the dissection. The middle thyroid vein is ligated. The superior pole of the thyroid is dissected free, and care is taken to identify and preserve the external branch of the superior laryngeal nerve (Fig. 44-1). The superior pole vessels are ligated adjacent to the thyroid lobe, rather than cephalad to it, to prevent damage to this nerve. The inferior thyroid artery and recurrent laryngeal nerve are identified. To preserve the blood supply to the parathyroid glands, the inferior thyroid artery should not be ligated laterally; rather, its branches should be ligated individually on the capsule of the lobe after they have supplied the parathyroid glands (Fig. 44-2). The parathyroid glands are identified and an attempt is made to leave each with an adequate blood supply. Any parathyroid gland that appears to be devascularized can be minced and implanted into the sternocleidomastoid muscle after its identification by frozen section. Care is taken to identify the recurrent laryngeal nerve along its entire course if a total lobectomy is to be done. The nerve is gently unroofed from surrounding tissue so that trauma to it is avoided. The nerve is in greatest danger near the junction of the trachea with the larynx, because here it is adjacent to the thyroid gland. Once the nerve and parathyroid glands have been identified and preserved, the thyroid lobe can be removed from its tracheal attachments (ligament of Berry). The contralateral thyroid lobe is removed in a similar manner when a total thyroidectomy is performed. A near-total thyroidectomy means that a small amount of thyroid tissue is left on the contralateral side to protect the parathyroid glands and recurrent nerve.38,39 Careful hemostasis and visualization of all important anatomic structures are mandatory for success. The strap muscles are approximated in the midline with only one suture to decrease the possibility of a hematoma in this deep space, which could result in tracheal compression. Several deep dermal sutures of 4-0 absorbable sutures are used. The dermis is approximated by a subcuticular stitch of interrupted 5-0 synthetic absorbable sutures, and the skin edges are approximated with sterile paper tapes. A small suction catheter is often inserted through a stab wound and is usually removed by the following morning.

FIGURE 44-1. Proximity of the external branch of the superior laryngeal nerve to the superior thyroid vessels is clearly shown. (Inf., inferior; m., muscle; br., branch; n., nerve; ext., external; Sup., superior; a., artery.) (From Moosman DA, DeWeese MS. The external laryngeal nerve as related to thyroidectomy. Surg Gynecol Obstet 1968; 127:1011. By permission of Surgery, Gynecology, & Obstetrics, now known as the Journal of the American College of Surgeons.)

FIGURE 44-2. The thyroid lobe is retracted medially and, by careful blunt dissection, the recurrent laryngeal nerve, the inferior thyroid artery, and the parathyroid glands are identified. The inferior thyroid artery is not ligated laterally as a single trunk. Rather, each small branch is ligated and divided at a point distal to the parathyroid glands (arrows) to preserve their blood supply. Then, the thyroid lobe can be removed from its tracheal attachments if a lobectomy is to be performed. (From Kaplan EL. Surgery of the thyroid gland. In: De Groot LS, Larsen PR, Refetoff S, Stanbury JB, eds. The thyroid and its diseases. New York: John Wiley and Sons, 1984:851.)

A lateral neck dissection is not done prophylactically for papillary or follicular cancer but, rather, only when metastatic disease is identified in the lateral triangle. Cherry-picking operations are not appropriate, but true modified radical lymph node dissections (Fig. 44-3), which leave the spinal accessory nerve, sternocleidomastoid muscle, and, in most cases, the jugular vein intact, are effective and provide a very satisfactory functional and cosmetic result.

FIGURE 44-3. Modified radical neck dissection. The sternocleidomastoid muscle and internal jugular vein have been divided inferiorly and the dissection performed along the carotid sheath from below upward, exposing underlying structures. The thyroid gland is dissected free from the trachea, sparing the recurrent laryngeal nerve

and leaving the parathyroid glands with a good blood supply. The vagus, phrenic, and spinal accessory nerves as well as the brachial plexus are left intact. At the end of the procedure, the divided sternocleidomastoid muscle can be reattached. In most instances, this muscle does not need to be divided, and the jugular vein can be left in situ if nodes are not fixed to it. (Int. jug. v., internal jugular vein; m., muscle; Spinal access n., spinal accessory nerve; Recurr. laryng. n., recurrent laryngeal nerve; Com. car. a., common carotid artery.) (From Cady B, Rossi R. In: Surgery of the thyroid and parathyroid glands. Philadelphia: WB Saunders, 1991:204.)

Subtotal thyroidectomy is used for the surgical treatment of Graves disease. Two to 4 g of thyroid tissue should be left to attain satisfactory postoperative thyroid function without a high incidence of recurrent thyrotoxicosis. Less thyroid is left in young patients. When operations are performed for ophthalmopathy, a near-total or total thyroidectomy should be performed. Once more, the parathyroid glands and recurrent nerves should be identified and preserved in an intact state. After a thyroidectomy, even with a modified radical neck dissection, the patient is almost always discharged on the first postoperative morning. Some surgeons currently discharge patients several hours after thyroidectomy, but the author thinks this is unsafe because late bleeding may occur. The skin tapes are removed at the time of the first outpatient visit, 8 to 10 days postoperatively. Video-assisted thyroidectomy or endoscopic resection of thyroid tumors is being pioneered in several centers.40,41 and 42 Its aim is to minimize the length of incisions in the neck or to hide the incisions by placing them below the clavicle or far lateral in the neck. In the hands of very skilled operators, the recurrent laryngeal nerve and parathyroid glands can be seen and a lobectomy performed. Other groups have performed mainly nodulectomy. Although this method may have merit for small lesions, the results and complications must be carefully assessed. Is the proper thyroid resection still being performed? Will there be a learning curve (as occurs with most new procedures) with increased rates of bleeding, nerve injuries, or hypoparathyroidism? Finally, do the improved cosmetic results warrant the possibility of greater morbidity? A careful assessment of the new procedures seems warranted.

POSTOPERATIVE COMPLICATIONS
Many authors have reported large series of thyroidectomies with no deaths. In other reports, mortality does not differ greatly from that due to anesthesia alone. Four major complications are associated with thyroid surgery: thyroid storm, wound hemorrhage, recurrent laryngeal nerve injury, and hypoparathyroidism. THYROID STORM Thyroid storm reflects an exacerbation of a thyrotoxic state and is seen most often in Graves disease, but it can occur rarely in patients with toxic adenoma or toxic multinodular goiter. Clinical manifestations and management of thyroid storm are discussed in Chapter 42. WOUND HEMORRHAGE Wound hemorrhage with hematoma is an uncommon complication, reported in 0.3% to 1.0% of patients in most large series. However, it is a well recognized and potentially lethal complication. A small amount of hematoma deep to the strap muscles can compress the trachea. A suction closed-drainage system placed in the wound is not adequate for decompression if bleeding is from a major vessel. Swelling of the neck and bulging of the wound can be followed shortly thereafter by respiratory impairment. Treatment consists of immediately opening the wound and evacuating the clot, even at the bedside. Later, the bleeding vessel can be ligated in a careful and leisurely manner in the operating room. The urgency of treating this condition when it occurs cannot be overemphasized. INJURY TO THE RECURRENT LARYNGEAL NERVE Recurrent laryngeal nerve injuries 43 occur in 1% to 2% or more of thyroid operations, especially when performed for malignant disease. The injuries can be unilateral or bilateral, temporary or permanent. Loss of function can be caused by transection, ligation, or handling of the nerve. In unilateral recurrent nerve injuries, the voice becomes husky because the vocal cords do not approximate one another. Usually vocal function returns within 6 to 9 months. If improvement is insufficient, injection of the paralyzed vocal cord with collagen may help.44 One surgery involves insertion of a plug of silicone rubber, fashioned to the correct size to push the paralyzed vocal cord to the midline and thus improve vocal function. Bilateral recurrent laryngeal nerve damage is much more serious because both vocal cords may assume a median or paramedian position, which results in airway obstruction and difficulty with respiratory toilet. Often, tracheostomy is required. Recurrent nerve injury is best avoided by always attempting to identify each nerve along its course. Accidental transection occurs most often at the level of the upper two tracheal rings, where the nerve is closely approximated to the thyroid lobe in the area of the ligament of Berry. If the transection is recognized, the transected nerve should be reapproximated by microsurgical techniques. A number of procedures to reinnervate the laryngeal muscles have been attempted with limited success.45 Injury to the external branch of the superior laryngeal nerve occurs when the upper pole vessels are divided (see Fig. 44-1) and results in an inability to forcefully project one's voice or to sing high notes. Often, this disability improves during the first 3 months after surgery. HYPOPARATHYROIDISM The incidence of permanent hypoparathyroidism has been reported to be as high as 20% when a total thyroidectomy and radical neck dissection are performed and as low as 0.9% for subtotal thyroidectomy.46 However, many surgeons can perform total or near-total thyroidectomy with a very low incidence of this complication. Postoperative hypoparathyroidism is rarely the result of inadvertent removal of all parathyroid glands but, more commonly, is due to disruption of their blood supply. Devascularization can be minimized by carefully ligating the branches of the inferior thyroid artery on the thyroid capsule distal to their supply of the parathyroid glands and by treating the parathyroids with great care (see Fig. 44-2). If a parathyroid gland is recognized to be nonviable during surery, after identification by frozen section, it can be minced into 1- to 2-mm cubes and placed into pockets in the sternocleidomastoid muscle or elsewhere.47 Postoperative hypoparathyroidism results in hypocalcemia and hyperphosphatemia and is manifested by circumoral numbness, tingling, and intense anxiety during the first few days after operation. Chvostek sign appears early, and carpope the serum calcium level is <8 mg/dL.48 Routinely, the author measures the serum calcium level on the postoperative evening, on the following morning, and every 12 hours thereafter if the patient remains in the hospital. Oral calcium as calcium carbonate is used liberally, especially because most patients now leave on the first postoperative morning. Symptomatic patients should not be discharged and should be given 1 g (10 mL) of 10% calcium gluconate intravenously over 10 minutes. Then 2 to 5 g of this calcium solution should be mixed in a 500-mL bottle to run intravenously during each subsequent 6- to 8-hour period. On this treatment regimen of oral and intravenous calcium, most patients become asymptomatic, and intravenous therapy usually can be stopped. The patient can be discharged if asymptomatic on oral calcium therapy. If more severe hypocalcemia is present, more calcium gluconate may be needed intravenously, and 1,25-dihydroxy-vitamin D should be given orally to promote the absorption of oral calcium. The management of more persistent hypocalcemia is discussed in Chapter 60. CHAPTER REFERENCES
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. Billroth T. Chirurgische Klinik Zurich, 18601867. Berlin, 1867. Kocher T. Zur pathologie und therapie des propfes. Deutsch Z Chir 1874; 4. Halsted WS. The operative story of goitre. Johns Hopkins Hosp Rep 1920; 19:71. Becker C. Hypothyroidism and atherosclerotic heart disease: pathogenesis, medical management, and the role of coronary artery bypass surgery. Endocr Rev 1985; 6:432. Sridama V, Reilly M, Kaplan EL, et al. Long term follow-up study of compensated low dose 131 I therapy for Graves' disease. N Engl J Med 1984; 311:426. Mitchie W, Hamer-Hodges DW, Pegg CAS, et al. Beta-blockade and partial thyroidectomy for thyrotoxicosis. Lancet 1974; 1:1009. Lennquist S, Jortso E, Anderberg B, Smeds S. Beta-blockers compared with antithyroid drugs as preoperative treatment of hyperthyroidism: drug tolerance, complications and postoperative thyroid function. Surgery 1985; 98:1141. Eriksson M, Rubenfeld S, Garber AJ, Kohler PO. Propranolol does not preent thyroid storm. N Engl J Med 1977; 296:263. Heppe H, Armin A, Calandra DB, et al. Hrthle cell tumors of the thyroid gland. Surgery 1985; 98:1162. Brooks JR, Starnes HF, Brooks DC, Pelkey JN. Surgical therapy for thyroid carcinoma: a review of 1249 solitary thyroid nodules. Surgery 1988; 104:940. Kaplan EL. An operative approach to the irradiated thyroid gland with possible carcinoma: criteria technique and results. In: DeGroot LJ, Frohman LA, Kaplan EL, Refetoff S, eds. Radiation associated carcinoma of the thyroid. New York: Grune & Stratton, 1977:371. DeGroot LJ. Clinical features and management of radiation-associated thyroid carcinoma. In: Kaplan EL, ed. Surgery of the thyroid and parathyroid glands. Edinburgh: Churchill-Livingstone, 1983:40. Sampson RJ. Prevalence and significance of occult thyroid cancer. In: DeGroot LJ, Frohman LA, Kaplan EL, Refetoff S, eds. Radiation associated thyroid carcinoma. New York: Grune & Stratton, 1977:137. McCaffrey TV, Bergstralh EJ, Hay ID. Locally invasive papillary thyroid carcinoma: 19401990. Head Neck 1994; 16:165.

15. Clark OH. Total thyroidectomy: the treatment of choice for patients with differentiated thyroid cancer. Ann Surg 1982; 196:361. 16. Grant CS, Hay ID, Gough IR, Bergstralh EJ. Local recurrence in papillary thyroid carcinoma: is extent of surgical resection important? Surgery 1988; 104:954. 17. Beierwaltes WH. Treatment of metastatic thyroid cancer with radioiodine and external radiation therapy. In: Kaplan EL, ed. Surgery of the thyroid and parathyroid glands. Clinical surgery international, vol 4. Edinburgh: Churchill-Livingstone, 1983:103. 18. Hill SC Jr. Chemotherapy of thyroid cancer. In: Kaplan EL, ed. Surgery of the thyroid and parathyroid glands. Clinical surgery international, vol 4. Edinburgh: Churchill-Livingstone, 1983:120. 19. DeGroot LJ, Kaplan EL, Straus FH Jr, Shukla MS. Does the method of management of papillary thyroid carcinoma make a difference in outcome? World J Surg 1994; 18:123. 20. Bckdahl M, Wallin G, Lwhagen T, et al. Fine needle biopsy cytology and DNA analysis. Surg Clin North Am 1987; 67:19 21. Hay ID, Grant CS, Taylor WF, et al. Ipsilateral lobectomy versus bilateral lobar resection in papillary thyroid carcinoma: a retrospective analysis of surgical outcome using a novel prognostic scoring system. Surgery 1987; 102:1088. 22. Cady B, Rossi R. An expanded view of risk-group definition in differentiated thyroid carcinoma. Surgery 1988; 104:947. 23. Rasbach DA, Mondschein MS, Harris NL, et al. Malignant lymphoma of the thyroid gland: a clinical and pathological study of twenty cases. Surgery 1985; 98:1166. 24. Kim JH, Leeper RD. Treatment of anaplastic and spindle cell carcinoma of the thyroid with combination Adriamycin and radiation therapy. Cancer 1983; 52:954. 25. Sweeney PJ, Haraf DJ, Recant W, et al. Clinical case: anaplastic carcinoma of the thyroid. Ann Oncol 1996; 7:739. 26. Sizemore GW, van Heerden JA, Carney JA. Medullary carcinoma of the thyroid gland and the multiple endocrine neoplasia type 2 syndrome. In: Kaplan EL, ed. Surgery of the thyroid and parathyroid glands. Clinical surgery international, vol 4. Edinburgh: Churchill-Livingstone, 1983:75. 26a. Kebrew E, Ituarte PH, Siperstein AE, et al. Medullary thyroid carcinoma: clinical characteristics, treatment, prognostic factors, and a comparison of staging systems. Cancer 2000; 88:1139. 26b. Cohen R, Campos JM, Salaun C, et al. Preoperative calcitonin levels are predictive of tumor size and postoperative calcitonin normalization in medullary thyroid carcinoma. Group d'Etude d'Tumeurs a Calcitonine (GETC). J Clin Endocrinol Metab 2000; 85:919. 27. Tisell LE, Hansson G, Jansson S, Salander H. Reoperation in the treatment of asymptomatic metastasizing medullary thyroid carcinoma. Surgery 1986; 99:60. 28. van Heerden JA, Grant CS, Gharib H, et al. Long-term course of patients with persistent hypercalcitoninemia after apparent curative primary surgery for medullary thyroid carcinoma. Ann Surg 1990; 212:395. 29. Dorr U, Wurstlin S, Frank-Raue K, et al. Somatostatin receptor scintigraphy and magnetic resonance imaging in recurrent medullary thyroid carcinoma: a comparative study. Horm Metab Res Suppl 1993; 27:48. 30. Waddington WA, Kettle AG, Heddle RM, Coakley AJ. Intraoperative localization of recurrent medullary carcinoma of the thyroid using indium-111 pentetreotide and a nuclear surgical probe. Eur J Nucl Med 1994; 2:363. 31. Moley JF, Wells SA, Dilley WG, Tisell LE. Reoperation for recurrent or persistent medullary thyroid cancer. Surgery 1993; 114:1090. 32. Donis-Keller H, Dou S, Chi D, et al. Mutations in the RET proto-oncogene are associated with MEN 2A and FMTC. Hum Mol Genet 1993; 2:851. 33. Wells SA Jr, Donis-Keller H. Current perspectives on the diagnosis and management of patients with multiple endocrine neoplasia type 2 syndromes. Endocrinol Metab Clin North Am 1994; 23:215. 34. Lairmore TC, Ball DW, Baylin SB, Wells SA Jr. Management of pheochromocytomas in patients with multiple endocrine neoplasia type 2 syndromes. Ann Surg 1993; 217:595. 35. Moosman DA, DeWeese JS. The external laryngeal nerve as related to thyroidectomy. Surg Gynecol Obstet 1968; 127:1011. 36. Kaplan EL. Surgery of the thyroid gland. In: DeGroot LJ, Larsen PR, Refet off S, Stanbury JB, eds. The thyroid and its diseases. New York: John Wiley and Sons, 1984:851. 37. Sedgwick CE, Cady B. Surgery of the thyroid and parathyroid glands. Philadelphia: WB Saunders, 1980:180. 38. Sarda AK, Bai S, Kapur MM. Near-total thyroidectomy for carcinoma of the thyroid. Br J Surg 1989; 76:90. 39. Lumsden AB, McGarity WC. A technique for subtotal thyroidectomy. Surg Gynecol Obstet 1989; 168:177. 40. Iacconi P, Bendinelli C, Miccoli P. Endoscopic thyroid and parathyroid surgery. Surg Endosc 1999; 13(3):314. 41. Shimizu K, Akira S, Tanaka S. Video-assisted neck surgery: endoscopic resection of benign thyroid tumor aiming at scarless surgery of the neck. J Surg Oncol 1998; 69(3):178. 42. Yeung GH. Endoscopic surgery of the neck: a new frontier. Surg Laparosc Endosc 1998; 8(3):227. 43. Lennquist S, Cahlin C, Smeds S. The superior laryngeal nerve in thyroid surgery. Surgery 1987; 102:999. 44. Ford CN, Martin DW, Warner TF. Injectable collagen in laryngeal rehabilitation. Laryngoscope 1984; 94:513. 45. May M, Beery P. Muscle-nerve pedicle laryngeal reinnervation. Laryngo scope 1986; 96:1196. 46. Mazzaferri EL. Papillary and follicular cancer: a selective approach to diagnosis and treatment. Annu Rev Med 1981; 32:73. 47. Wells SA, Farndon JR, Dale JK, et al. Long-term evaluation of patients with primary parathyroid hyperplasia managed by total parathyroidectomy and heterotopic autotransplantation. Ann Surg 1980; 192:451. 48. Kaplan EL, Sugimoto J, Yang H, Fredland A. Postoperative hypoparathyroidism: diagnosis and management. In: Kaplan EL, ed. Surgery of the thyroid and parathyroid glands. Clinical surgery international, vol 4. Edinburgh: Churchill-Livingstone, 1983:262.

CHAPTER 45 HYPOTHYROIDISM Principles and Practice of Endocrinology and Metabolism

CHAPTER 45 HYPOTHYROIDISM
LAWRENCE E. SHAPIRO AND MARTIN I. SURKS History and Epidemiology Etiology Systemic Effects of Thyroid Hormone Deficiency Skin and Cutaneous Appendages Gastrointestinal System Cardiovascular System Respiratory System Blood Neuromuscular System Renal Function Bone Adrenocortical System Reproductive System Pregnancy Intermediary Metabolism Drug Metabolism Clinical Diagnosis History Physical Examination Laboratory Diagnosis Thyroid Hormone Measurements Thyroid-Stimulating Hormone Measurement and Thyrotropin-Releasing Hormone Tests Other Tests Therapy Guidelines for Standard Therapy Special Clinical Problems Diagnosis and Management of Patients with Nonthyroidal Disease Effects of Nonthyroidal Disease on Patients with Hypothyroidism Treatment of Asymptomatic or Chemical Hypothyroidism Myxedema Coma Treatment of Thyroid Cancer The Noncompliant Patient Chapter References

Hypothyroidism is the appellation for any degree of thyroid hormone deficiency. The term myxedema is often used to indicate the result of a thyroid hormone deficiency that is of sufficient severity and chronicity to cause characteristic changes in the physical examination.

HISTORY AND EPIDEMIOLOGY

By the end of the nineteenth century, the syndrome of acquired thyroid hormone deficiency in adults was described as a cretinoid state by Gull.1 Ord proposed the term myxoedema to be applied to this condition, which he observed in middle-aged women.2 Kocher3 described its occurrence after thyroidectomy, and by 1894, Raven4 had described the use of oral thyroid medication in the treatment of this disease. Although most physicians appreciate that thyroid disorders are relatively common in the general population, the true incidence is less certain. Undoubtedly, there is a significant population of patients with undiagnosed thyroid disease who still function relatively normally. The Whickham survey, performed in a rural area of England, specifically sought to determine the incidence of thyroid disease in the general population.5 Hypothyroidism, diagnosed by history and blood analysis, was found in more than 2% of 2800 persons tested. The mean age of diagnosis was 57 years, and the disease was ten-fold more common in women than in men. The disease is particularly prevalent in women older than 40 years of age. The disease is prevalent in debilitated geriatric patients of both sexes.6

ETIOLOGY
Hypothyroidism may develop from any one of a number of acquired thyroid diseases or their treatments, from hereditary abnormalities, or secondarily from hypothalamic-pituitary disease (Table 45-1). Usually, hypothyroidism develops as a consequence of Hashimoto disease or of prior radioactive iodine or surgical management of Graves disease. In 20% of patients treated for Graves disease, the hypothyroidism that develops after radioactive iodine therapy occurs within the first 12 months of treatment. The initial incidence is locally influenced by the dosage of radiation used. A slower but relentlessly progressive increase in incidence of hypothyroidism continues at a rate of 2% to 4% of the population at risk each year.7

TABLE 45-1. Causes of Transient or Permanent Hypothyroidism

Hypothyroidism may also develop as a late consequence of Graves disease in patients who previously were not treated with radioactive iodine or surgery.8 Studies suggest that thyrotropin-binding inhibitor immunoglobulins circulate in some of these patients.9 The hyperthyroidism of Graves disease may be supplanted by hypothyroidism, because the autoimmune production of thyrotropin receptor-blocking antibodies has resulted in inhibition of thyroid-stimulating hormone (TSH) receptor function. Idiopathic hypothyroidism implies that an appropriate clinical and laboratory evaluation has failed to reveal a specific etiologic factor for this disorder. This is a common occurrence in patients with acquired hypothyroidism. Because many of these patients have low titers of antithyroid antibodies, and because some may have associated autoimmune disorders, such as pernicious anemia, most investigators believe that the hypothyroidism represents the end stage of autoimmune thyroid disease. The progressive development of hypothyroidism is readily apparent in many patients with Hashimoto thyroiditis. In this disease, there is continual replacement of normal thyroid tissue with lymphocytic and fibrous tissue, until insufficient thyroid tissue remains to maintain normal hormone production. Temporary hypothyroidism may also develop during other forms of thyroiditis, such as viral (i.e., subacute) thyroiditis, silent thyroiditis, or postpartum thyroiditis (see Chap. 46). Although the causes of these forms of thyroiditis probably are distinct, some aspects of their clinical presentation and course can be similar. These patients often present with hyperthyroidism that is associated with decreased thyroid gland function, as determined by a very low uptake of radioactive iodine. The hyperthyroidism probably results from disruption of follicular structure, with the release of hormone systemically. This stage may last from a few weeks to several

months. It is generally followed by a euthyroid stage, then often by a hypothyroid stage that may persist for several months. The hypothyroid stage of the disease is characterized by a healing and partially functioning thyroid gland that is not yet producing sufficient thyroid hormone to maintain the euthyroid state. Hypothyroidism in these settings usually is transient and does not require lifelong thyroid hormone replacement. Several pharmacologic agents may influence thyroid function, but only thionamides and iodides are commonly associated with hypothyroidism. Good clinical management dictates that hypothyroidism should not occur during management of Graves disease with thionamides. As thyrotoxicosis is controlled, the physician should decrease the dosage of thionamides to a level that allows sufficient hormone production to maintain the euthyroid state or, alternatively, supplement thionamides with a daily replacement dose of thyroid hormone. If these modifications are not made, hypothyroidism may occur, often associated with an enlarging thyroid gland. Iodide-induced hypothyroidism may be more difficult to document because the source of excessive iodide ingestion may be obscure. Common sources are iodine-containing pharmaceuticals, such as amiodarone, radiopaque contrast agents, expectorants, and topical antiseptics. Although failure to adapt to the acute Wolff-Chaikoff effect (see Chap. 37) is unusual in individuals without thyroid disease, iodide-induced hypothyroidism by this mechanism occurs readily in patients with pre-existing thyroid diseases, such as Hashimoto disease or Graves disease, previously treated with radioactive iodine.10 Iodide-induced hypothyroidism in these settings may be treated by reducing the excessive iodide intake or by the administration of thyroid hormone. Hypothyroidism may also be caused by several hereditary disorders, particularly of the enzymatic pathways that are responsible for iodothyronine synthesis. Although an occasional adult may present with hypothyroidism and goiter because of hormone dysgenesis, most of these patients are diagnosed and treated during childhood. Defects in the genes coding for the b form of the thyroid hormone nuclear receptor have been reported to occur in familial forms of generalized thyroid hormone resistance.11 Patients with thyroid hormone resistance manifest a spectrum of syndromes, ranging from cases of mental retardation and abnormal somatic development to clinically euthyroid patients diagnosed only because of elevated levels of TSH and thyroid hormones (see Chap. 32). Familial syndromes of thyroid gland agenesis have been shown to be caused by an inactivating mutation of the TSH receptor.12 In a small fraction of hypothyroid patients, the disorder results from decreased stimulation of the thyroid by TSH. The hypothalamic or pituitary disorders that are responsible for decreased TSH production are discussed in Chapter 15. Clinically, attention is focused on secondary or hypothalamic-pituitaryinduced hypothyroidism by the finding of a normal or decreased serum TSH concentration concomitant with hypothyroidism.

SYSTEMIC EFFECTS OF THYROID HORMONE DEFICIENCY

The clinical spectrum of thyroid hormone deficiency ranges from the asymptomatic person without abnormal physical findings to the classic myxedematous patient in whom the diagnosis can readily be made on physical examination. Physical findings are most striking in young, otherwise healthy patients. Advanced age or nonthyroidal disease can commonly produce symptoms or physical findings that suggest thyroid disease and require thyroid function tests to aid in a diagnosis. Thyroid hormone deficiency causes generally decreased oxygen consumption and defects specific to individual organs. The common symptoms of thyroid hormone deficiency include lethargy and decreased physical ability. The decrease in tissue calorigenesis results in cold intolerance. SKIN AND CUTANEOUS APPENDAGES In myxedema, the epidermis demonstrates an atrophied cellular layer with hyperkeratosis of traumatized areas. The dermis is infiltrated with mucopolysaccharides, hyaluronic acid, and chondroitin sulfate, resulting in increased fluid retention and nonpitting edema. Acinar gland secretion is decreased. On physical examination, decreases in surface temperature and pallor, resulting from cutaneous vasoconstriction, are observed. These changes produce the cool, dry roughness of the skin of the myx-edematous patient (see Chap. 218). The hair and nails are brittle. Hair loss is common, but it usually is reversible with treatment. GASTROINTESTINAL SYSTEM The usually modest weight gain in hypothyroidism probably is the result of a decreased metabolic rate. Marked obesity does not result from thyroid hormone deficiency. The sense of taste is decreased. Gastric atrophy and decreased gastric acid production are common. Antiparietal cell antibodies are present in approximately one-third of patients, and pernicious anemia has been reported in 10%.13 Constipation is a common symptom, but ileus is seen only in patients with severe myxedema. Ascites is rarely evident, although it may exist in the markedly myx-edematous patient who may also have pericardial and pleural effusions (see Chap. 204). CARDIOVASCULAR SYSTEM Cardiac function is affected by hypothyroidism; however, this is significant only in patients with coexistent cardiac disease. There is decreased contractility, decreased cardiac output, and diminished cardiac oxygen consumption in hypothyroid patients.14 In hypothyroid patients without coexisting heart disease, low exercise tolerance and shortness of breath on exertion may be partially caused by decreased cardiac function, but symptoms and signs specific for congestive heart failure usually are absent. In the patient with an already compromised heart, hypothyroidism may cause significant further deterioration. However, the lower tissue demands for cardiac output and tissue oxygenation may compensate for poor cardiac function (see Chap. 203). Unless ventricular function is compromised by a large myxedematous pericardial effusion, the authors believe that clinically significant cardiac dysfunction very rarely results from hypothyroidism. Occasionally, severe congestive heart failure coexists in patients with hypothyroidism. Such patients should be given thyroxine (T4) with caution. The potential benefit of T4 would be to increase contractility and cardiac output. The potential harm would be that increased myocardial oxygen requirements might further compromise a failing heart. The diagnosis of autoimmune thyroid disease per se did not increase the risk of ischemic heart disease after a 20-year follow-up.15 However, elevations of serum lipids in hypothyroid patients are likely risk factors for coronary artery disease. The component of the hyperlipidemia that is due to thyroid hormone deficiency is variable, and it resolves after normalization of thyroid hormone levels. Because myocardial oxygen consumption is augmented by thyroid hormone, a hypothyroid patient with a deficient coronary artery circulation may not tolerate full replacement doses. RESPIRATORY SYSTEM In the absence of primary respiratory disease, the diminution of respiratory function in the hypothyroid patient is not significant. These patients have measurably decreased responses to hypoxia and hypercapnea, but usually this lack of response is of little clinical importance. However, respiratory muscle function may be seriously affected in the severely myxedematous patient, in whom hypoventilation may occur. Hypothyroidism may be a reversible component in rare patients with clinically significant sleep apnea syndrome.16 Myxedematous patients experiencing sleep apnea syndrome often have pharyngeal obstruction due to macroglossia. In the patients requiring prolonged mechanical ventilation, aggressive T4 replacement may facilitate weaning from the respirator. However, coexisting heart disease may limit the dosage of T4 that may be given safely (see also Chap. 202). BLOOD Patients with myxedema commonly have a mild normochromic, normocytic anemia coexisting with a decreased red cell mass (see Chap. 212). In the childbearing years, hypothyroid women are often iron deficient because of an associated menorrhagia. Macrocytosis can occur in hypothyroidism because of13 Pernicious anemia occurs folic acid or vitamin B12 deficiency. in 10% of hypothyroid patients. In general, hypothyroid patients do not report abnormal bleeding. However, a clinically evident bleeding tendency may be caused by a defect in clotting factors VIII and IX, increased capillary fragility, or abnormal platelet function. NEUROMUSCULAR SYSTEM Unlike many other tissues, oxygen consumption in the adult cerebrum does not depend on thyroid hormone. However, the hypothyroid patient may have diminished cerebral blood flow that is due to decreased cardiac output. Although this situation is usually clinically insignificant, it may play a role in the psychiatric17 and neurologic disorders occasionally seen in severe hypothyroidism18 (see Chap. 200). Patients with severe myxedema may have general depression of central nervous system function, characterized by sluggish responses to questions and sleepiness. Rarely, in extreme cases, agitated psychosis may be seen, and seizures can occur. Myxedema coma often occurs in the complicated setting of severe trauma or systemic illness, and the diagnosis should be suspected in cases of depression, hypothermia, hypoventilation, hyponatremia, and an exaggerated response to sedatives or narcotics.

Nerve compression by myxedematous fluid may cause the carpal tunnel syndrome. Diminished hearing and poor night vision have been described. Cerebellar dysfunction may occur in severe cases of hypothyroidism; posterior column symptomatology probably indicates coexisting pernicious anemia. Deep tendon reflexes have a characteristic delayed relaxation phase that is relatively specific for hypothyroidism. Muscle cramps are common, but muscle weakness is not prominent. Persistent elevation of serum creatine phosphokinase is commonly seen and is not indicative of active muscle inflammation (see Chap. 210). RENAL FUNCTION The decrease in cardiac output diminishes renal blood flow, which rarely has clinical implications (see Chap. 208). Decreased free water clearance may be encountered in severe hypothyroidism, causing hyponatremia that is similar to the syndrome of inappropriate secretion of antidiuretic hormone (see Chap. 27). However, abnormalities in antidiuretic hormone regulation have not been documented.19 BONE The effects of hypothyroidism on bone and calcium metabolism in adults are rarely significant. Serum calcium is usually normal, and there is no characteristic bone disease. ADRENOCORTICAL SYSTEM Patients with hypothyroidism exhibit a decrease in the rates of cortisol secretion and cortisol metabolism.20 Nevertheless, patients with primary hypothyroidism, in the absence of coexisting adrenal or pituitary disease, have normal plasma cortisol levels and a normal cortisol response to corticotropin. However, hypothyroidism can coexist with primary adrenal deficiency resulting from autoimmune disease21 or can be associated with a secondary adrenal insufficiency caused by an underlying hypothalamic-pituitary disease. If deficiencies of thyroid and adrenocortical hormones do occur, symptoms of the latter disease may be masked by the decreased requirement for glucocorticoids in the unstressed hypothyroid patient. In such a case, manifestations of adrenal failure may occur if thyroid hormone alone is prescribed. REPRODUCTIVE SYSTEM In adult women of childbearing age, excessive menstrual bleeding is common. Fertility decreases, and there is an increased incidence of early abortions.22 Libido is diminished in both sexes. Hypothyroidism alters the metabolism of sex hormones and the concentrations of sex hormonebinding proteins. To avoid confusing results, it is prudent to postpone a nonemergency evaluation of sex hormone function until the hypothyroidism is corrected. Serum prolactin levels can be elevated in women with hypothyroidism, and it resolves with treatment.23 PREGNANCY High levels of estrogens, as occur in pregnancy, result in the elevation of serum T4binding globulin and an ~30% to 50% increase in serum total T4. Consequently, determination of serum total T4 yields an inaccurate assessment of thyroid function in pregnancy, and measurements of serum free T4 (FT4) and TSH should be obtained (see Chap. 110). The thyroid status of the mother affects both fetal and childhood health.23b If the pregnant patient is newly diagnosed as hypothyroid and has no coexisting medical disease, the rapid replacement of thyroid hormone should be prescribed.23b Treatment should begin with a dose of T4 estimated to result in full replacement. The adequacy of this dose can be assessed by determination of TSH and FT4 after 4 weeks.24 A patient who is euthyroid while taking T4 therapy for hypothyroidism should contact her endocrinologist when pregnancy occurs to determine whether the current dose of T4 still results in normal levels of serum TSH. An increase in T4 dose may be required in some pregnant patients.25 All T4-treated hypothyroid patients require reassessment of FT4 and TSH by the end of the first trimester of pregnancy. Any adjustment of T 4 dosage requires reevaluation by assessing FT4 and TSH after 4 weeks. If the pregnant patient has hypothyroidism as a result of previous treatment for Graves disease, further attention is required. Although the continued presence of thyroid-stimulating immunoglobulin may be clinically occult in the mother, the passive transfer of maternal immunoglobulins place the fetus and newborn at risk for Graves disease. The presence of high levels of thyroid-stimulating immunoglobulin in late pregnancy may predict the presence of fetal Graves disease. Postpartum thyroid dysfunction may occur 2 to 8 months after delivery in a patient not previously known to have thyroid disease. The reported incidence of postpartum dysfunction is 2% to 20%.26 The usual course is reminiscent of silent thyroiditis. Patients first develop hyperthyroidism, with a non-tender, firm goiter that has decreased ability to concentrate radioactive iodine. Subsequently, thyroid hormone levels fall, and hypothyroidism occurs. The full-time course of the illness is usually several months before recovery occurs. However, patients with postpartum thyroiditis have serum antithyroid antibodies, often have recurrences after each pregnancy, and may eventually develop permanent hypothyroidism. INTERMEDIARY METABOLISM Some evidence of the general slowing of metabolism may be clinically apparent. An associated diabetes mellitus often is ameliorated in the presence of hypothyroidism with a decreased requirement for insulin or oral hypoglycemic drugs. Hyperlipidemia, caused by reduced clearance, can increase levels of low-density lipoprotein, total cholesterol, and triglycerides. Another risk factor for coronary heart disease, plasma total homocysteine levels, which are elevated in hypothyroidism, is corrected by thyroxine treatment.27,27b DRUG METABOLISM Thyroid hormone regulates the metabolism of drugs by the hepatic cytochrome p450 enzymes. Examples of drugs with a metabolism that decreases in hypothyroidism are antiseizure medications, digoxin, and narcotics. In addition, the effect of warfarin therapy may change because of alteration of the turnover of serum clotting factors. Thus, the dose requirements of these medications may change as a result of T4 treatment.

CLINICAL DIAGNOSIS
HISTORY Adult-onset hypothyroidism usually causes an insidious progression of multiple symptoms and signs, the exact onset of which often cannot be specified by the patient, close family members, or even the personal physician who is seeing the patient on a regular basis (Fig. 45-1 and Table 45-2). This usually is the result of the gradual decline of thyroid hormone levels that occurs in most cases. The physician should seek any history of prior thyroid disease or of therapy directed at the thyroid gland. A history of external irradiation of the head or neck could result in pituitary or thyroid failure. Any history of the use of compounds containing iodine or lithium should be elicited. Iodine deficiency is not a tenable cause of hypothyroidism in the United States because of the commercial practice of supplementing salt and bread with iodine. There may be a family history of thyroid disorders such as Graves or Hashimoto diseases or, more rarely, congenital thyroid disease or dyshormonogenetic goiter. If secondary hypothyroidism is suspected, symptoms pertinent to other pituitary hormonal deficits should be elicited. A personal and family history of autoimmune processes should be explored in cases of idiopathic thyroid failure. Thoracic obstructive symptomssuch as dysphagia, dyspnea, and lightheadedness on raising the armsmay be present because of a large retroclavicular goiter.

FIGURE 45-1. Two hypothyroid patients before (A, C) and after (B, D) full replacement with thyroid hormone. A, Notice the loss of eyelashes, the areas of vitiligo around the mouth, the region around the eyes, and the scalp. C, Notice the lethargic expression and the periorbital puffiness that disappear with therapy (D).

TABLE 45-2. Clinical Presentation of Thyroid Hormone Deficiency

The unusual occurrence of rapidly progressive symptoms of hypothyroidism should suggest one of a few relatively rare diseases. Some forms of thyroiditis are associated with a hyperthyroid and a subsequent hypothyroid phase. These include subacute thyroiditis, silent thyroiditis, and postpartum thyroid disorders (see Chap. 42 and Chap. 46). Occasionally, the physician encounters a demented patient without the ability to give an accurate history who was previously euthyroid on thyroid replacement before the unintentional cessation of treatment. The exposure to iodine or lithium in a patient with an occult thyroid disorder may precipitate an abrupt onset of hypothyroidism. The most common symptoms of hypothyroidism are listed in Table 45-2. There may be a slight weight gain or an inability to lose weight by dieting, but morbid obesity does not occur. Respiratory and cardiovascular symptoms are usually limited to shortness of breath on exertion, with no symptoms at rest or when lying flat in bed. The skin, hair, nail, and voice changes are most obvious in young patients; in the older population, they may erroneously be attributed to aging. PHYSICAL EXAMINATION The features typical of the physical examination are listed in Table 45-2. In myxedema, there may be a typical facies that is puffy, with loss of eyebrows and coarse skin (see Fig. 45-1). The voice may have a unique, deep croaking quality. The tongue may be enlarged. If thyroid failure has resulted from treatment of Graves disease, orbitopathy may be present. Examination of the neck may reveal a scar from prior thyroid surgery. A goiter may be present, although often there is no palpable thyroid tissue. In hypothyroidism, the areolae are normally pigmented, unless there is coexistent adrenal disease, for which there may be decreased pigment (i.e., in pituitary failure) or increased pigment (i.e., in primary adrenal failure). Axillary and pubic hair are present, unless other endocrine failure coexists. Percussion of the chest may reveal basal dullness that is due to the bilateral pleural effusions seen occasionally in advanced cases of myxedema. The breath sounds should be normal. Examination of the heart may reveal remarkable enlargement if myxedematous pericardial effusion exists. In this case, the cardiac impulse is not palpable, and heart sounds are distant. Signs of pericardial constriction should be sought. The abdomen often is mildly obese, without organomegaly. In severe myxedema, the examiner may encounter an ileus. Examination of the extremities may demonstrate a thickened, nonpitting puffiness. The neurologic examination reveals a delay in the relaxation phase of deep tendon reflexes.

LABORATORY DIAGNOSIS
THYROID HORMONE MEASUREMENTS In patients with typical symptoms and signs, the diagnosis of hypothyroidism generally can be confirmed by measurement of the FT4 level. Hypothyroidism is characterized by a decrease of T4 secretion and a decrease of the serum FT4 levels; measurement of total T4 levels in serum usually correlate with levels of FT4. However, alterations of plasma thyroid hormone binding proteins can result in decreased serum total T4 values in the absence of hypothyroidism or in normal total T4 values in the presence of hypothyroidism. A decrease in serum-binding proteins and serum total T4 levels occurs most commonly in chronic nonthyroidal illness, but it also occurs during androgen treatment and in patients with a hereditary decrease in serum T4binding globulin. Alternatively, an increase in serum T4binding may occur because of estrogen-containing medications, pregnancy, acute liver dysfunction, human immunodeficiency virus infection, or several familial syndromes. Because thyroid regulation is not affected by abnormalities of thyroid hormone binding, the levels of serum FT4 and TSH are in the normal ranges. In nonthyroidal disease, serum FT4 may be in the normal range or increased. This contrasts with the findings in a patient with hypothyroidism in whom the serum FT4 is decreased.28 Thus, the measurement of serum FT4 may be helpful. Although triiodothyronine (T3) is the most biologically active of the thyroid hormones, a decrease in serum T3 concentration is not a sufficiently specific marker of hypothyroidism to be useful diagnostically. As with T4, the serum T3 value may be affected by alterations in serum iodothyroninebinding proteins; it also may be low in elderly patients. Nonetheless, in the absence of significant nonthyroidal disease, FT4 levels are normal in patients as old as 110 years of age.29 Serum T3 levels frequently are decreased in euthyroid patients with nonthyroidal disease and during fasting or food restriction because of a diminished extrathyroidal conversion of T4 to T3. Because of the common occurrence of low serum T3 values in euthyroid patients, serum levels of this hormone should not be used to confirm hypothyroidism.30 THYROID-STIMULATING HORMONE MEASUREMENT AND THYROTROPIN-RELEASING HORMONE TESTS In patients with hypothyroidism caused by disease of the thyroid gland, decreased iodothyronine production is always associated with an increase in serum TSH. A low serum FT4 concomitant with an increased serum TSH confirms the diagnosis of hypothyroidism and signifies that the hypothyroidism is caused by failure of the thyroid gland. In the few patients who have hypothyroidism because of hypothalamic or pituitary disease, the serum TSH level is decreased or within normal limits. In the past, inadequate sensitivity of TSH measurements in many laboratories generally precluded differentiating normal from low values. Ultrasensitive immunoas-says for TSH have been used to determine normal or decreased serum TSH in patients with secondary hypothyroidism.31 The surprising detection of normal serum levels of TSH in some patients with secondary hypothyroidism may be the result of the presence of biologically inactive TSH in these patients.32 The use of thyrotropic-releasing hormone (TRH) is not recommended to diagnose or to determine the cause of hypothyroidism. In patients with hypothalamic or pituitary disease, the TSH response to TRH may be depressed or normal. Moreover, the TSH response to TRH is not sufficiently specific to determine whether the lesion is in the pituitary or hypothalamus.32 However, the TSH response to TRH is always increased in patients with primary hypothyroidism. Because an elevation in the baseline serum TSH value confirms the diagnosis of primary hypothyroidism, testing with TRH does not play a role in the diagnosis of this disorder.

The negative-feedback regulation of TSH secretion by thyroid hormones is sensitive to changes of 10% to 20% in serum T4 and T3.33 Even a small decrease in iodothyronine production that results in a decrease in serum T4 to a value that may still be within the normal range may result in a TSH elevation (Fig. 45-2, time A). The finding of a normal serum T4 value associated with an elevation in the serum TSH level is probably the first indication of failure of the thyroid gland and subclinical hypothyroidism. Because an elevation of the serum TSH concentration is the most sensitive indication of early primary hypothyroidism, TSH levels should be monitored, particularly in patients who are at high risk for developing this condition, including those who have had thyroid surgery or radioiodine treatment, patients with Hashimoto disease or Graves disease in remission, patients with malignancies of the head or neck who have undergone radiation therapy, and patients with multinodular goiter or with a strong family history of thyroid dysfunction (see Chap. 15).

FIGURE 45-2. Progressive changes in serum thyroxine (T4), triiodothyronine (T3), and thyroid-stimulating hormone (TSH) concentrations during the development of hypothyroidism. The shaded areas represent the normal ranges for the three hormones. The time points may be months or years apart. At time A, the minimal decrease in serum T4 to a value that is still within the normal range results in a rise in serum TSH. At time B, serum T4 has decreased to a very low value, and serum TSH is markedly elevated. The patient may have few symptoms, because serum T3 is still relatively normal. At time C, all patients have symptomatic hypothyroidism.

OTHER TESTS Before radioimmunoassay measurements of serum FT4, T4, and TSH were available, tests using radioactive iodine were a mainstay in the diagnosis of hypothyroidism. A decrease in thyroidal uptake of radioiodine does occur in patients with hypothyroidism, but an increasing iodide intake in the United States has resulted in a lower normal range for thyroidal radioiodine uptake, which now significantly overlaps with the range for hypothyroid patients.34 Although radioiodine testing may occasionally be useful to elucidate the abnormal iodine kinetics of certain forms of hypothyroidism, the inconvenience, expense, and radiation exposure of this test have markedly reduced its use to substantiate hypothyroidism. Other tests that reflect, in part, the biologic activity of thyroid hormones were extensively used. These procedures included determination of oxygen consumption, serum cholesterol, and Achilles tendon reflex relaxation time. Later, tests such as cardiac systolic ejection intervals and serum levels of sex hormonebinding globulin were used (see Chap. 33). All these tests suffered from deficient sensitivity and have been supplanted by the specific and sensitive measurements of serum FT4 and TSH concentrations. Several additional manifestations of the hypothyroid state include an increase in the activity of serum creatinine phosphokinase, serum oxaloacetic transaminase, and lactate dehydrogenase, but these changes are not sufficiently specific to be useful in routine diagnosis.

THERAPY
GUIDELINES FOR STANDARD THERAPY Several pharmaceutical preparations are available for the treatment of hypothyroidism. The careful use of any of these preparations can reverse the hypothyroidism and restore the euthyroid state. Before commencing therapy, it is important to define whether hypothyroidism is caused by thyroid failure or by the much less common failure of TSH secretion because of hypothalamic or pituitary disease. In the latter case, coexisting impairment of adrenocorticotropic hormone secretion may precipitate an adrenal failure that had been satisfactorily tolerated during hypothyroidism. Even when TSH elevation demonstrates hypothyroidism that is due to thyroid failure, it is important to consider coexisting autoimmune disease of the adrenal gland. Treatment of the hypothyroidism alone raises the metabolic rate and the rate of cortisol metabolism and may then unmask adrenocortical hormone deficiency. In these settings, cortisol should be administered with the thyroid hormone replacement. Animal products, such as desiccated thyroid and thyroglobulin, were standard replacement therapy for several decades. These products are being replaced with synthetic preparations of L -thyroxine (L -T4), L -triiodothyronine (L -T3), or medications that combine synthetic T4 and T3. The use of synthetic preparations is recommended because of their stability and because their standardization is by hormone content rather than by iodine content. Among the synthetic products, only L -T4 produces relatively unchanged serum T4 and T3 levels throughout the 24-hour day.35 Because of its small distribution pool in blood, T3 as well as synthetic combinations of T4 and T3 result in a postabsorptive increase in serum T3 to hyperthyroid values. The consequent instability of serum T3 levels in patients treated with T3 does not allow assessment of the thyroidal state by measurement of serum levels of iodothyronine. Thyroid hormone receptors in some tissues (e.g., brain) are occupied by the T3 that is produced within the target tissue from deiodination of T4.36 Because oral T3 would not deliver thyroid hormone to such tissues in a physiologic manner, it is recommended that synthetic L -T4 be used for replacement therapy. One of the well-standardized brands (e.g., Synthroid, Levothroid, Levoxyl) is recommended, because variations in potency have been reported for some generic preparations.37 The initial dosage of L -T4 depends on the severity of the hypothyroid disorder, the age of the patient, and the presence of associated or underlying medical conditions. The half-life of L -T4, which is ~6 days in euthyroid individuals, may be significantly prolonged in hypothyroid patients. A full replacement dose has been considered to be between 1.5 and 2.0 g T4/kg of body weight.33 Most patients' deficiencies are fully replaced with a daily dosage of 0.075 to 0.15 mg per day. In young, otherwise healthy patients with mild hypothyroidism, a full daily replacement dose of L -T4 (usually 0.075 to 0.10 mg) may be given at the beginning of therapy. The initial dose of L -T4 in patients with severe hypothyroidism or in patients with clinically apparent or probable underlying atherosclerotic heart disease or in elderly patients is generally 0.0125 to 0.025 mg per day. This low dose is recommended because an abrupt increase in metabolic rate and demand for increased cardiac output may precipitate angina pectoris, myocardial infarction, congestive heart failure, or arrhythmias.38 After this low initial daily dosage of L -T4 is fully equilibrated (4 to 6 weeks), and if the patient has no symptoms or signs of cardiac decompensation, the daily dosage may be cautiously increased. Doses are assessed at monthly intervals until the clinical syndrome is relieved or until laboratory tests demonstrate that full replacement has been achieved. The therapeutic goal for lifelong replacement therapy should be alleviation of the clinical syndrome and the normalization of serum TSH. It is essential to use a TSH assay of sufficient sensitivity to detect serum TSH levels that fall below the normal range in response to excess T4. Appropriate practice requires that T4 dosage be adjusted to normalize serum TSH to prevent possible adverse effects of chronic excess T4 therapy on bone mineral density39 or cardiac function.40 These therapeutic goals may be modified in elderly patients with cardiovascular disease. In such individuals, optimal therapy may be a daily dosage of L -T4 that relieves most of the symptoms of hypothyroidism without causing myocardial decompensation. Such a dose may not fully normalize serum T4 and TSH. After the desired dose of T4 is established, patients are reevaluated annually to assess changing requirements or altered absorption of L -T4 and their reliability in ingestion of medication. The dosage of L -T 4 should be evaluated when the patient's status changes. Pregnancy may increase the dosage needed to normalize serum TSH. The prescription of some medications may alter the absorption (bile acid binders, iron supplements, or aluminum-containing antacids), or metabolism (antiseizure medications or rifampin), of L -T4. If the patient develops symptoms of thyroid dysfunction or heart disease, a dosage adjustment may be required.

SPECIAL CLINICAL PROBLEMS

DIAGNOSIS AND MANAGEMENT OF PATIENTS WITH NONTHYROIDAL DISEASE Nonthyroidal disease complicates the evaluation of thyroid function. Differentiating hypothyroidism from the effects of nonthyroidal disease is a common reason to request endocrine consultation in hospitalized patients. Although patients who have nonthyroidal disease are not clinically hypothyroid, their clinical state may have some features in common with hypothyroidism, such as lethargy, hypothermia, and constipation. Moreover, the effects of nonthyroidal disease on binding proteins and iodothyronine metabolism may result in changes that suggest associated hypothyroidism.30 Most clinical investigators think that the decreased conversion of T4 to T3 that occurs in illness facilitates the conservation of body protein stores and does not indicate hypothyroidism (see Chap. 30, Chap. 36 and Chap. 232). Although patients with primary hypothyroidism have decreased serum T4 and increased serum TSH levels, patients with nonthyroidal disease may have decreased levels of serum T3 and normal or reduced levels of serum T4.28 Decreased serum T3 levels are encountered so frequently in these sick patients that this determination is of little value.41 When the serum T4 level remains normal in patients with nonthyroidal disease, there is no difficulty in excluding hypothyroidism.28 However, when patients with severe nonthyroidal disease have decreased serum T4 levels, the serum FT4 concentration often remains normal, a finding that also would exclude hypothyroidism. Nevertheless, because of the presence of circulating inhibitors of thyroid hormone binding to plasma-binding proteins, and depending on the technique used to determine FT4, even this value may be decreased in very sick patients.42 A diagnosis of primary hypothyroidism usually can be excluded in such patients by finding a normal level of serum TSH, but studies have shown that the magnitude of the TSH secretory response to TRH after an experimentally produced decrease in serum T4 and T3 levels may be reduced in approximately one-half of older individuals and in patients with nonthyroidal disease.43 These findings suggest that TSH may fail to rise in some in whom nonthyroidal disease coexists with primary hypothyroidism. Experience and clinical judgment are essential to complement laboratory values in evaluating thyroid status in some sick patients with low serum T4 and TSH concentrations. The finding of TSH elevation in an intensive care unit (ICU) patient usually indicates primary hypothyroidism. In a study of 86 patients who were hospitalized in an ICU, only two patients had hypothyroidism, and both had elevated serum levels of TSH.44 There is a caveat to the interpretation of an elevated TSH in an ICU patient recovering from a recent, severe illness: TSH levels are elevated during the recovery of patients without endocrine disease whose serum T4 levels were diminished during severe nonthyroidal illness. In such patients, the TSH elevation is transient and resolves after normalization of the serum T4 level.45 Because the concentration of reverse T3 is normal or increased in nonthyroidal disease and is decreased in hypothyroidism, the measurement of reverse T3 may be helpful in differentiating patients with low TSH and secondary hypothyroidism caused by hypopituitarism from patients with nonthyroidal illness.42 Serum T4 and TSH concentrations may be decreased in both clinical settings. The finding of low serum levels of T4 without elevations of serum TSH in a severely ill ICU patient most often occurs in the absence of endocrine disease. However, these laboratory data are also consistent with TSH deficiency. Because pituitary failure may coexist in an ICU patient, assessment of this life-threatening possibility is an important part of the endocrine evaluation of these patients. EFFECTS OF NONTHYROIDAL DISEASE ON PATIENTS WITH HYPOTHYROIDISM Patients who have hypothyroidism and nonthyroidal disease represent special management problems, because the treatment of either disorder may adversely affect the other. For example, the use of iodine-rich radiopaque contrast agents for diagnostic procedures may acutely worsen subclinical or established hypothyroidism. The treatment of hypothyroidism may precipitate angina pectoris, myocardial infarction, arrhythmia, or congestive heart failure in patients with underlying heart disease. The coexistence of hypothyroidism with a severe nonthyroidal disease represents a challenge to the clinician. Hypothyroid patients frequently have substantial anemia, a decrease in free water clearance, and a decrease in the rate of drug metabolism. These factors must be carefully considered in the management of the associated nonthyroidal disease. These issues are heightened when surgical management of the nonthyroidal disease is required. Hypothyroidism significantly increases the risk of anesthesia and surgery because of delayed metabolism of anesthetic agents and drugs, abnormal fluid balance, alveolar hypoventilation, reduced cardiac output, and other factors. Because of these potentially serious problems, a decision for surgery should be made only after extensive discussions among the surgeon, anesthesiologist, and the clinicians responsible for medical management. Evidence suggests that optimal combined management results in an excellent outcome for hypothyroid patients who undergo coronary bypass surgery. In this instance, thyroid hormone replacement usually is delayed until after surgery has been performed.38 TREATMENT OF ASYMPTOMATIC OR CHEMICAL HYPOTHYROIDISM Because hypothyroidism may develop insidiously, over many years in some patients, abnormal serum T4 and TSH concentrations may be found before the typical symptoms and signs of hypothyroidism become apparent. The increase in serum TSH, which accompanies a decreasing serum T4 level as thyroid function fails, stimulates the thyroid to produce more T3 relative to T4. Patients in the early stages of thyroid failure may have relatively normal serum T3 levels, despite a marked decrease in serum T4 and increase in serum TSH concentration (see Fig. 45-2, time B). These patients may be clinically euthyroid.46 Because of the frequency of thyroid hormone tests as part of multiphasic screening, several individuals who are identified as having chemical hypothyroidism because of abnormal results do not appear to have a compatible clinical syndrome. Some clinicians do not think that treatment with thyroid hormone replacement should be started in such patients, because the rate of progression of the thyroid failure is uncertain. It is possible that clinically relevant hypothyroidism may never develop in some of these persons. However, the presence of elevated serum TSH marks the patient with exhausted thyroid reserve who is at the edge of worsening hypothyroidism. The presence of antithyroid antibodies increases the risk of progressive thyroid failure, prompting L -T4 treatment.47,48 If the patient is not treated, the physician must plan to reevaluate thyroid function on a regular basis. Although there are no prospective studies to reference, the authors suggest that there are several reasons why most patients with subclinical hypothyroidism should be treated: 1. Symptoms of hypothyroidism may be subtle and not appreciated by the patient until full replacement therapy has resulted in beneficial changes. 2. Patients with underlying thyroid disease and subclinical hypothyroidism are at risk for rapid development of severe hypothyroidism after exposure to iodides, because in many instances, they fail to adapt to the acute Wolff-Chaikoff effect. In a similar manner, treatment with lithium can precipitate worsening thyroid hormone deficiency. 3. The T3 content of some tissues in the body is derived from intracellular T4 rather than the plasma T3. The failure to restore T4 levels to normal in patients with chemical hypothyroidism may theoretically result in low T3 levels within these tissues. 4. The effects of potent drugs, such as warfarin and digoxin, and antiseizure medications may be inconstant. 5. T4 therapy is safe, simple, and relatively inexpensive. The assays for TSH allow the physician to avoid subtle T4 overdosage. 6. The need for routine follow-up is as great (or greater) for the patient who is observed without treatment than it is for the treated patient. 7. Treatment of subclinical hypothyroidism decreases serum cholesterol.49 MYXEDEMA COMA Coma, as an end stage of myxedema, was described by Ord in 1880.2 Thereafter, it was accepted that severe myxedema with coma was usually lethal. Only within the last three decades has there been success in treatment. Two factors seem important for improved survivability: The first is successful specific treatment of the intercurrent nonthyroidal disease that often precipitates the comatose state, and the second is that newer methods of intensive care for general medical management have increased survival for critically ill patients in general. Hypothyroidism is compatible with life in the otherwise healthy individual. The patient with a history of years or even decades of hypothyroidism is not rare. However, such a person has dysfunction of many organs and is less able to tolerate life-threatening illness than the euthyroid person. Survival of the critically ill patient with severe hypothyroidism depends on vigorous treatment of nonthyroidal life-threatening processes and prompt therapy with thyroid hormones. The presence of coma may be a marker of the patient's clinical deterioration more than a primary effect of hypothyroidism. Often, there is a critical insult in the form of infection, surgery, convulsive seizure, congestive heart failure, cerebral vascular accident, gastrointestinal bleeding, cold exposure, or the administration of sedatives that precipitates myxedema coma.50 Generally, the patient has clinical features of hypothyroidism on physical examination, is hypothermic, and has laboratory evidence of primary hypothyroidism. Other frequent findings are hypoventilation and hyponatremia.

The rare occurrence of myxedema coma and the heterogeneity of coexisting nonthyroidal diseases make it difficult to evaluate modes of treatment. Therapy is based on the successful treatment of seven clinically myxedematous patients with various associated and severe nonthyroidal diseases.51 These investigators administered 400 to 500 g of L -T4 intravenously. This dose is sufficient to restore normal T4 levels. They also used pharmacologic doses of glucocorticoids in some of these patients, a practice commonly recommended to treat potential adrenal insufficiency. Vigorous therapeutic attention should be paid to coexisting problems, such as hypothermia, shock, and CO2 retention. Some clinicians use high doses of intravenous T4, although others prefer to use smaller doses of T 4 or to use T3 because of its more rapid onset of effect. 52,53 and 54 However, in some reports in which relatively high doses of T3 have been used, an increased incidence of cardiovascular complications was observed.55 The authors' approach is to consider each case individually, with a careful search for coexisting diseases and assessment of the integrity of the patient's coronary arteries. The patient should be in an ICU. Vigorous therapeutic attention should be paid to coexisting diseases and life-threatening manifestations, such as hypothermia, shock, and CO2 retention. T4 should be given intravenously. At least 50 g is given dailythis dosage is sufficient to initiate therapy and minimizes adverse effects on a potentially diseased cardiovascular system. Glucocorticoids are used if the setting suggests coexisting adrenal insufficiency. In the authors' experience, the principal determinant of the patient's outcome is the degree of success achieved in diagnosing and treating coexisting diseases while the appropriate hormone therapy is being applied. TREATMENT OF THYROID CANCER Patients treated for thyroid epithelial cell cancer are often athyrotic after surgery and radioiodine treatment. In these patients, the goal of L -T4 therapy is to just suppress serum TSH, thus creating minimal biochemical thyroid hormone excess. Current studies suggest that this goal can be accomplished without the occurrence of cardiac dysfunction56 or diminished bone mineral density.57 THE NONCOMPLIANT PATIENT A common cause of failure to normalize serum TSH is patient noncompliance. Because of its long serum half-life, L -T4 can be given at greater than daily intervals. The patient should double the daily dosage on the day after a missed dose. In the extreme, the single ingestion of a week's dosage of T4 every 7 days has been reported to achieve a reasonable therapeutic result.58 CHAPTER REFERENCES
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. Gull WW. On a cretinoid state supervening in adult life in women. Trans Clin Soc Lond 1874; 7:180. Ord WM. Cases of myxoedema. Trans Clin Soc Lond 1879; 13:15. Kocher TH. Uber kropfextirpation und ihre folgen. Langenbecks Arch Klin Chir 1883; 29:254. Raven TF. Myxoedema treated with thyroid tablets. BMJ 1894; 1:12. Tunbridge WM, Evered DC, Hall R, et al. Lipid profiles and cardiovascular disease in the Whickham area with particular reference to thyroid failure. Clin Endocrinol (Oxf) 1977; 7:481. Helfand M, Crapo LM. Screening for thyroid disease. Ann Intern Med 1990; 112:840. Cevallos JL, Hagen GA, Maloof F, et al. Low-dosage 131I therapy of thyrotoxicosis (diffuse goiters). A five-year follow-up study. N Engl J Med 1974; 290:141. Tamai H, Kasagi K, Takaichi Y, et al. Development of spontaneous hypothyroidism in patients with Graves' disease treated with antithyroid drugs: clinical, immunological and histological findings in 26 patients. J Clin Endocrinol Metab 1989; 69:49. Konishi J, Iida Y, Kasagi K, et al. Primary myxedema with thyrotrophin-binding inhibitor immunoglobulins. Clinical and laboratory findings in 15 patients. Ann Intern Med 1985; 103:26. Wolff J. Iodide goiter and the pharmacologic effects of excess iodide. Am J Med 1969; 47:101. Takeda K, Balzano S, Sakurai A, et al. Screening of nineteen unrelated families with generalized resistance to thyroid hormone for known point mutations in the thyroid hormone receptor gene and detection of a new mutation. J Clin Invest 1991; 87:496. Gagne N, Parma J, Deal C, et al. Apparent congenital athyreosis contrasting with normal thyroglobulin levels and associated with inactivating mutations in the thyrotropin gene. J Clin Endocrinol Metab 1998; 83:1771. Tudhope GR, Wilson GM. Anaemia in hypothyroidism. Q J Med 1960; 29:513. Woeber KA. Thyrotoxicosis and the heart. N Engl J Med 1992; 232:94. Vanderpump MP, Tunbridge WM, French JM et al. The development of ischemic heart disease in relation to autoimmune thyroid disease in a 20-year follow up study of an English community. Thyroid 1996; 6:155. Skjodt NM, Atkar R, Easton PA. Screening for hypothyroidism in sleep apnea. Am J Respir Crit Care Med 1999; 160:732. Jackson IM. The thyroid axis and depression. Thyroid 1998; 8:951. Hall RCW. Psychiatric effects of thyroid hormone disturbance. Psychosomatics 1983; 24:7. Iwasaki Y, Oiso Y, Yamauchi K, et al. Osmoregulation of plasma vasopressin in myxedema. J Clin Endocrinol Metab 1990; 70:534. Gordon GG, Southren AL. Thyroid hormone effects on steroid hormone metabolism. Bull NY Acad Med 1977; 53:241. Edmonds M, Lamki L, Killinger DW, Volpe R. Autoimmune thyroiditis, adrenalitis and oophoritis. Am J Med 1973; 54:782. Stephenson MD. Frequency of factors associated with habitual abortion in 197 couples. Fertil Steril 1996; 66:24. Honbo KS, Van Herle AJ, Kellett KA. Serum prolactin levels in untreated primary hypothyroidism. Am J Med 1978; 64:782.

23b. Haddow JE, Palomaki GE, Allan WC, et al. Maternal thyroid deficiency during pregnancy and subsequent neuropsychological development of the child. N Engl J Med 1999; 341:541. 24. 25. 26. 27. Montoro, MN. Management of hypothyroidism during pregnancy. Clin Obstet Gynecol 1997; 40:65. Mandel PJ, Larsen PR, Seely GW, et al. Increased need for thyroxine during pregnancy in women with primary hypothyroidism. N Engl J Med 1990; 323:126. Gerstein HC. How common is postpartum thyroiditis? Arch Intern Med 1990; 150:1397. Nedrebo BG, Ericsson U-B, Nygard O, et al. Plasma total homocysteine levels in hyperthyroid and hypothyroid patients. Metabolism 1998; 47:89.

27b. Hussein WI, Green R, Jacobsen DW, et al. Normalization of hyperhomocysteinemia with L-thyroxine in hypothyroidism. Ann Intern Med 1999; 131:348. 28. Surks MI, Hupart KH, Pan C, et al. Normal free thyroxine in critical non-thyroidal illnesses measured by ultrafiltration at undiluted serum and equilibrium dialysis. J Clin Endocrinol Metab 1988; 67:1031. 29. Mariotti S, Barbesino G, Caturegli P. Complex alteration of thyroid function in healthy centenarians. J Clin Endocrinol Metab 1993; 77:1130. 30. Tibaldi JM, Surks MI. Effects of nonthyroidal illness on thyroid function. Med Clin North Am 1985; 69:899. 31. Spencer CA. Clinical utility and cost effectiveness of sensitive thyrotropin assays in ambulatory and hospitalized patients. Mayo Clin Proc 1988; 63:1214. 32. Faglia G, Bitensky L, Pinchera A, et al. Thyrotropin secretion in patients with central hypothyroidism: evidence for reduced biological activity of immunoreactive thyrotropin. J Clin Endocrinol Metab 1979; 48:989. 33. Ordene KW, Pan C, Barzel US, et al. Variable thyrotropin response to thyrotropin releasing hormone after small decreases in plasma thyroid hormone concentrations in patients of advanced age. Metabolism 1983; 32:881. 34. Pittman JA, Dailey GE, Beschi RJ. Changing normal values for thyroidal radioiodine uptake. N Engl J Med 1969; 280:1431. 35. Surks MI, Schadlow AR, Oppenheimer JH. A new radioimmunoassay for plasma L-triiodothyronine: measurements in thyroid disease and in patients maintained on hormonal replacement. J Clin Invest 1972; 51:3104. 36. Larsen PR, Silva JE, Kaplan MM. Relationships between circulating and intracellular thyroid hormones: physiological and clinical implications. Endocr Rev 1981; 2:87. 36b. Toft AD (Editorial). Thyroid hormone replacementone hormone or two? N Engl J Med 1999; 340:469. 37. Fish LH, Schwartz HL, Cavanaugh J, et al. Replacement dose metabolism and bioavailability of levothyroxine in treatment of hypothyroidism. N Engl J Med 1987; 37:764. 38. Becker C. Hypothyroidism and atherosclerotic heart disease: pathogenesis, medical management and the role of coronary bypass surgery. Endocr Rev 1985; 6:432. 39. Taelman P, Kaufman JM, Janssens X, et al. Reduced forearm bone mineral content and biochemical evidence of increased bone turnover in women with euthyroid goitre treated with thyroid hormone. Clin Endocrinol (Oxf) 1990; 33:107. 40. Biondi B, Fazio S, Carella C, et al. Cardiac effects of long term thyrotropin-suppressive therapy with levothyroxine. J Clin Endocrinol Metab 1993; 77:334. 41. Bermudez F, Surks MI, Oppenheimer JH. High incidence of decreased serum triiodothyronine concentration in patients with nonthyroidal disease. J Clin Endocrinol Metab 1975; 41:27. 42. Chopra IJ, Solomon DH, Hepner GW, et al. Misleading low free thyroxine index and usefulness of reverse triiodothyronine measurement in non-thyroidal disease. Ann Intern Med 1979; 90:905. 43. Maturlo SJ, Rosenbaum RL, Pan C, et al. Variable thyrotropin response to thyrotropin-releasing hormone following small decreases in plasma free thyroid hormone concentrations in patients with nonthyroidal diseases. J Clin Invest 1980; 66:451. 44. Slag MF, Morley JE, Elson MK, et al. Hypothyroxinemia in critically ill patients as a predictor of high mortality. JAMA 1981; 245:43. 45. Hamblin PS, Dyer SA, Mohr VS, et al. Relationship between thyrotropin and thyroxine changes during recovery from severe hypothyroxinemia of critical illness. J Clin Endocrinol Metab 1986; 62:717. 46. Zulewski H, Muller B, Exer P, et al. Estimation of tissue hypothyroidism by a new clinical score. J Clin Endocrinol Metab 1997; 82:771. 47. Rosenthal MJ, Hunt WC, Garry PJ. Thyroidal failure in the elderly: microsomal antibodies as a discriminant for therapy. JAMA 1987; 258:209. 48. Vanderpump MP, Turnbridge WM, French JM et al. The incidence of thyroid disorders in a community hospital: a twenty year follow up of the Whickham Survey. Clin Endocrinol (Oxf) 1995;43:55. 49. Tanis BC, Westendorp GJ, Smelt HM. Effect of thyroid substitution on hypercholesterolaemia in patients with subclinical hypothyroidism. Clin Endocrinol (Oxf) 1996; 44:643. 50. Wartofsky L. Myxedema coma. In: Braverman LE, Utiger RD, eds. The thyroid, 6th ed. Philadelphia: JB Lippincott, 1991:1084. 51. Holvey DN, Goodner CJ, Nicoloff JT, et al. Treatment of myxedema coma with intravenous thyroxine. Arch Intern Med 1964; 113:139.

52. 53. 54. 55. 56.

McConahey WM. Diagnosing and treating myxedema and myxedema coma. Geriatrics 1978; 33:61. Rosenberg IN. Hypothyroidism and coma. Surg Clin North Am 1968; 48:353. Capiferri R, Evered D. Investigation and treatment of hypothyroidism. Clin Endocrinol Metab 1979; 8:39. Hylander B, Rosenquist U. Treatment of myxoedema coma-induced factors associated with fatal outcome. Acta Endocrinol (Copenh) 1985; 108:65. Shapiro LE, Sievert R, Ong L, et al. Minimal cardiac effects in asymptomatic athyreotic patients chronically treated with thyrotropin-suppressive doses of L-thyroxine. J Clin Endocrinol Metab 1997; 82:2592. 57. Rosen HN, Moses AC, Garber J, et al. Randomized trial of pamidronate in patients with thyroid cancer: bone density is not reduced by suppressive dose of thyroxine, but is increased by cyclic intravenous pamidronate. J Clin Endocrinol Metab 1998; 83:2324. 58. Grebe SK, Cooke RR, Ford HC, et al. Treatment of hypothyroidism with once weekly thyroxine. J Clin Endocrinol Metab 1997; 82:870.

CHAPTER 46 THYROIDITIS Principles and Practice of Endocrinology and Metabolism

CHAPTER 46 THYROIDITIS
IVOR M. D. JACKSON AND JAMES V. HENNESSEY Suppurative Thyroiditis Etiology Clinical Features and Laboratory Evaluation Treatment Subacute Thyroiditis Etiology Clinical Features Laboratory Evaluation Treatment Hashimoto Thyroiditis (Autoimmune Thyroiditis) Etiology Pathology Clinical Features Other Associated Clinical Disorders Laboratory Evaluation Treatment Silent Thyroiditis Etiology Clinical Features Laboratory Evaluation and Treatment Postpartum Thyroiditis Clinical Features and Laboratory Evaluation Treatment Riedel Thyroiditis Radiation Thyroiditis Radioisotope Therapy-Induced Injury External Radiation-Induced Injury Other Forms of Thyroiditis Perineoplastic Thyroiditis Palpation Thyroiditis Granulomatous Disorders Chapter References

Thyroiditis is a term used for a wide variety of thyroid disorders or different etiologies. Patients may present with vastly different clinical features. The evaluation and therapy of thyroiditis and the relevant pathology and etiology are herein reviewed.

SUPPURATIVE THYROIDITIS
ETIOLOGY Suppurative thyroiditis, an uncommon but potentially serious disorder, is an acute, subacute, or chronic infection of the thyroid caused by either a bacterial or fungal infection. The organisms most commonly involved in acute suppurative thyroiditis include Staphylococcus aureus, Streptococcus hemolyticus, Escherichia coli, pneumococcus, and Salmonella sp.1 Bacteroides sp and other anaerobic bacteria may occasionally initiate this disorder.2 Acute inflammatory changes in the thyroid usually occur by direct hematogenous spread, either from a nearby infected tissue or from a distant site. Conversely, tuberculous or syphilitic involvement and fungal infection typically lead to a more chronic indolent process (see Chap. 213). Pneumocystis carinii has been documented as a cause of thyroid infection in patients with acquired immunodeficiency syndrome. In addition, cases of Coccidioides immitis thyroiditis have been reported in immune-suppressed patients in the context of disseminated coccidioidal disease.3 These findings suggest that patients infected with human immunodeficiency virus and other immune-compromised individuals may be predisposed to a variety of thyroid infections with more unusual, opportunistic organisms. CLINICAL FEATURES AND LABORATORY EVALUATION Patients with acute bacterial suppurative thyroiditis usually present with severe pain, tenderness, and swelling of the thyroid. Systemic signs and symptoms such as fever, rigor, and malaise frequently occur. Although pain usually is localized to the thyroid, it may be referred to the occiput, ear, or mandible. Examination characteristically reveals a tender thyroid gland, with warmth and erythema over the site of a nodule, and cervical lymphadenopathy. Patients with tuberculous thyroiditis may have thyroid swelling but absence of pain and tenderness. Serum thyroid hormone levels in all types of suppurative thyroiditis are usually normal, but rare cases with thyrotoxicosis are reported that are due to tissue necrosis with release of a high concentration of thyroid hormone into the systemic circulation.4 A specific diagnosis, with demonstration of the offending organism, usually can be made after aspiration of the infected area. TREATMENT Treatment of suppurative thyroiditis necessitates surgical drainage of any abscess and administration of the appropriate antibiotics. Although some patients may respond to antibiotics alone, a failure to respond indicates the need to consider surgical measures.5 Patients with a persistent thyroglossal duct or an internal fistula may require surgery for prevention of recurrence.6 The prognosis in properly treated patients is good; normal residual thyroid function is the rule.

SUBACUTE THYROIDITIS
ETIOLOGY Subacute thyroiditis is an inflammatory disorder of the thyroid thought to be viral in origin. Previously, this condition was termed granulomatous thyroiditis, De Quervain thyroiditis, struma granulomatosa, and pseudotuberculous thyroiditis. The pathologic hallmarks are granulomas and pseudo giant cells (Fig. 46-1). Evidence that suggests a viral infection includes the following: 1. 2. 3. 4. 5. Subacute thyroiditis often occurs after an upper respiratory tract infection or during a viral epidemic. There is no associated polymorphonuclear leukocytosis. Coxsackie virus antibody titers have been noted to rise.7 The mumps virus has been cultured from thyroid tissue of patients with subacute thyroiditis.8 The process is self-limited.

FIGURE 46-1. Subacute (granulomatous) thyroiditis. Note the numerous giant cells. 125 (From LiVolsi VA, LoGerfo P, eds. Thyroiditis. Boca Raton, FL: CRC Press,

1981.)

All current evidence suggests that subacute thyroiditis is not an autoimmune disease. Although low levels of antithyroid antibodies may appear transiently during the course of the disorder, the levels are not of the same magnitude seen in patients with other autoimmune thyroid disorders. Subjects with subacute thyroiditis, but not those with Hashimoto disease, have a higher frequency of a certain human leukocyte antigen (HLA) haplotype (HLA-B35)7; it is possible that this HLA association may reflect a genetic susceptibility to subacute thyroiditis after various viral infections.9 CLINICAL FEATURES These patients characteristically present with a painful, tender goiter that is firm, with pain radiating to the ears, mandible, or occiput. In some patients, however, these features may be minimal or even absent.7 Cervical lymphadenopathy usually is not present, but many patients have prodromal symptoms of an upper respiratory tract infection, fever, myalgia, or arthralgia. Additionally, patients with subacute thyroiditis often experience a series of changes in thyroid function related to the inflammatory effects in the gland. Early in the course, symptomatic thyrotoxicosis may arise from the leakage of thyroid hormones from damaged follicular cells (see Chap. 42). Subsequently, patients may develop transient hypothyroidism after passing through a euthyroid phase. The hypothyroidism results from a depletion of thyroxine (T4) and triiodothyronine (T3) from the thyroid gland after the destructive, inflammatory process. However, patients usually experience a spontaneous recovery of normal thyroid function. Not all patients with subacute thyroiditis experience all phases of these alterations in thyroid function. The duration of the illness often is 6 weeks but may be 3 to 6 months. The inflammatory process may migrate from one area of the thyroid to another and often crosses between the two lobes (creeping thyroiditis). LABORATORY EVALUATION Laboratory evaluation usually reveals the following: elevated erythrocyte sedimentation rate (often >55 mm per hour); normal or near-normal leukocyte count; transient elevations of antithyroid antibodies in low titers; and serum T4 and T3 values that may be high, normal, or low, depending on the phase of the disease. In the thyrotoxic phase, the serum thyroid-stimulating hormone (TSH) value is suppressed but may be detectable in third-generation assays in nearly 80% of those evaluated.10 TSH is normal during the euthyroid phase and elevated during the hypothyroid phase. A thyroid scan during an episode of subacute thyroiditis shows a cold region in the involved section of the thyroid. Additionally, during the thyrotoxic and euthyroid phases, the radioactive iodine (RAI) uptake is often <1%. This finding is of diagnostic importance in the evaluation of patients with suspected thyrotoxicosis, as a low uptake reflects the underlying etiology of destructive thyroiditisnot increased thyroid hormone production and secretion, as in Graves disease or toxic nodular goiter (in which the uptake is normal or high). Further, the serum thyroglobulin is frequently markedly elevated, a finding that is of value in the differential diagnosis of thyrotoxicosis that is due to surreptitious ingestion of thyroid hormone (thyrotoxicosis factitia). In the latter disorder, the RAI uptake and the serum thyroglobulin are low. Although the diagnosis of subacute thyroiditis usually is apparent in patients with typical clinical features, occasionally there is difficulty in establishing the diagnosissuch as with sudden enlargement of the thyroid glandraising the possibility of anaplastic carcinoma. In this situation, a thyroid biopsy may be helpful. The usual microscopic features include degeneration of the follicular epithelium; infiltration of the tissue with leukocytes, lymphocytes, and histiocytes; and formation of granulomas composed of giant cells surrounding colloid. In approximately one-half of patients, microabscesses may be found.11 In most patients with subacute thyroiditis, the diagnosis can be made clinically. TREATMENT The management of patients with subacute thyroiditis is based on relief of symptoms, because no specific therapy is available. Two aspects of this therapy must be addressed in each patient: the local symptoms and the effects of thyroid dysfunction. Frequently, pain and tenderness in the neck respond to treatment with salicylates, 0.65 g every 4 hours. If no benefit ensues within 48 hours, nonsteroidal antiinflammatory agents may be administered, although it is unclear whether these agents are superior to the salicylates. Approximately 5% of patients require cortico-steroids for the relief of symptoms. Prednisone in doses of 30 to 40 mg per day usually results in prompt improvement; however, it may be difficult to reduce the dose without return of pain, so that treatment for 4 to 8 weeks may be necessary before a successful tapering is accomplished.7 If the patient does not respond to prednisone within 24 to 48 hours, the diagnosis of subacute thyroiditis should be reevaluated. Thyroidectomy occasionally is used in patients with severe or recurrent pain that is unresponsive to other therapy, but this approach is required infrequently. If thyrotoxic symptoms are foremost, the patient should be managed with b-blockade (e.g., propranolol). The thyrotoxic phase in these patients is transient, and definitive therapy with RAI is neither indicated nor of value because the thyroid uptake is low. Likewise, the antithyroid drugs propyl-thiouracil or methimazole (Tapazole) should not be given because these medications are also ineffective. These drugs impair thyroid hormone synthesis, but the excess serum thyroid hormone levels in subacute thyroiditis result from leakage of T3 and T4 from the damaged follicles and not from enhanced synthesis and secretion. The hypothyroid phase of subacute thyroiditis is usually transient and is often asymptomatic; therefore, thyroid hormone replacement frequently is not necessary. Additionally, the increased TSH secretion at this time may be important in the recovery of thyroid gland function; consequently, there may be a relative contraindication to thyroid hormone treatment unless the patient is clearly symptomatic. The long-term prognosis for patients with subacute thyroiditis is excellent. Recurrent episodes, although uncommon, do occur, in as many as 2% of affected patients yearly,12 but eventual normal thyroid function is the rule. Discomfort in the thyroid area can, however, continue for many months. Rarely, patients with subacute thyroiditis may develop permanent hypothyroidism. Patients who have had prior thyroid surgery or who have coexistent autoimmune thyroiditis13 may be especially prone to such an outcome.

HASHIMOTO THYROIDITIS (AUTOIMMUNE THYROIDITIS)

Hashimoto thyroiditis, an autoimmune disorder, is the most common cause of hypothyroidism. It is also referred to as chronic lymphocytic thyroiditis and autoimmune thyroiditis. It occurs in all age groups, but it is most common in middle age. ETIOLOGY Numerous features of Hashimoto thyroiditis indicate an autoimmune cause in which multiple genetic factors likely play an important role.14 Because the disorder is much more common in women than in men, the possibility of a gene related to sex-steroid action has been considered.14 Patients with Hashimoto thyroiditis are characterized by the presence of circulating antibodies to microsomal and thyroglobulin antigens, often present in high titer. The disorder occurs frequently in certain families, particularly among the female members.15 Patients with Hashimoto thyroiditis and their relatives have an increased prevalence of other autoimmune disorders, such as Addison disease, diabetes mellitus, pernicious anemia, and myasthenia gravisone or more of these may develop over time. The disease is characterized pathologically by infiltration of the thyroid by plasma cells and lymphocytes. Autoimmune thyroiditis resulting in hypothyroidism may represent a spectrum from Hashimoto disease (with a goiter) to atrophic lymphocytic thyroiditis (without a goiter, often called primary myxedema). The variability in these related disorders may reflect the production of different types of antithyroid antibodies. For example, patients with agoitrous autoimmune thyroiditis may have a blocking antibody that inhibits the growth of thyroid tissue induced by TSH.16 This antibody most likely is absent in patients with Hashimoto thyroiditis and a goiter. The antimicrosomal antibodies found in patients with Hashimoto thyroiditis are primarily directed against the thyroid peroxidase (TPO) enzyme. Inhibition of this enzyme by the TPO antibody may be partly responsible for the hypothyroidism seen in these patients.17 In addition, patients with Hashimoto thyroiditis may have antibodies that block TSH-dependent production of cyclic adenosine monophosphate, thus inhibiting thyroid hormone production. In one study,18 it was shown that patients with overt hypothyroidism that is due to Hashimoto thyroiditis have a higher prevalence of TSH-blocking antibodies than patients with subclinical hypothyroidism. The autoimmune thyroid diseases represent a wide spectrum of immune abnormalities, from Graves thyrotoxicosis, caused by antibodies that stimulate the TSH receptor, to Hashimoto thyroiditis, which results from a complex process of immunologic events leading to impaired thyroid function. Both these autoimmune thyroid diseases share a common pathophysiologic bond: the production of an abnormal thyroid state resulting from an immune process directed against the thyroid gland. Volp15 suggested that Hashimoto thyroiditis is predominantly a disorder of cell-mediated immunity that is manifested by a genetic defect in the suppressor T-cell function. This theory postulates that as a result of defective suppressor T cells, helper (CD4) T cells are not suppressed appropriately and, therefore, are able to activate and cooperate with certain B lymphocytes. Additionally, the helper T cells produce various cytokines, including interferon-g, which induce thyrocytes to express major histocompatibility complex class II surface HLA-DR antigens and render them susceptible to immunologic attack. The activated B lymphocytes then produce antibodies that react with thyroid antigens. However, the antimicrosomal (anti-TPO) and antithyroglobulin antibodies used as serologic markers to identify patients with Hashimoto thyroiditis are likely not the direct cause of the thyroid destruction. There is evidence that cytotoxic T lymphocytes show variants in the promoter of antigen 4 (CTLA-4) and represent a risk factor for Hashimoto disease.19 These cells act in concert with complement and killer (or K) lymphocytes to cause lymphocytic thyroiditis. In addition to autoimmune mechanisms of thyroid cell damage, various environmental factors may be involved in the pathogenesis of Hashimoto thyroiditis.19a For example, iodide ingestion may help trigger the development of autoimmune thyroiditis.20 Further, the administration of granulocyte macrophage colony-stimulating factor or interleukin-2 may induce the temporary development of thyroid autoantibodies and reversible hypothyroidism in treated patients.21, 22 This finding suggests that patients receiving hematopoietic growth factors or cytokines should be monitored for the development of autoimmune thyroiditis and hypothyroidism. Apoptosis

(programmed cell death) may play a major role in the massive thyrocyte destruction in Hashimoto thyroiditis.23 Cross-linking of the Fas receptor with its ligand (Fas L) induces apoptosis, whereas this process is inhibited by the protooncogene Bc1-2. Immunostaining of Hashimoto tissue shows enhanced expression of Fas and Fas L but diminished Bc1-2 protein in thyroid follicles, effects activated by the cytokine interleukin-1b(IL-1b), which is abundantly produced in a Hashimoto gland.24 Certain HLA tissue types, such as HLA-DR5, are more common in patients with Hashimoto thyroiditis than in the general population (see Chap. 194). Patients with primary myxedema (agoitrous hypothyroidism) have an increased prevalence of DR3 but not the DR5 antigen, suggesting that these autoimmune thyroid disorders are separate genetic entities.25 Because Graves disease and Hashimoto thyroiditis may coexist in the same patient, it is interesting that in both conditions there is an increased prevalence of HLA-AW30.26 Graves disease, however, is associated with an excess of DR3 (as in primary myxedema) but not the DR5 antigen.27 HLA-DR antigens are not normally expressed on thyroid cells. However, in Hashimoto disease the thyroid follicular cells present these antigens on their surface, which may trigger the autoimmune process.28 There remains, however, considerable debate on the role that the HLA system plays in the pathogenesis of Hashimoto thyroiditis. It has been suggested that the tendency to develop thyroid autoantibodies is inherited as a mendelian dominant trait and not primarily through the HLA-related gene systems.20 In addition, age, sex, and environmental factors (e.g., iodide or viral exposures) may play modifying roles in defining which genetically predisposed patients develop autoimmune thyroiditis. PATHOLOGY Pathologically, Hashimoto thyroiditis is characterized by the infiltration of the gland with lymphocytes and plasma cells, follicular destruction and fibrosis, colloid depletion, and, at times, the presence of germinal centers (Fig. 46-2 and Fig. 46-3). In some patients, only isolated areas of lymphocytic infiltration are noted (termed focal thyroiditis). This may represent an early phase of Hashimoto thyroiditis, because there is a correlation between the degree of lymphocytic infiltration and the presence of antimicrosomal and antithyroglobulin antibodies in the serum of affected patients.29 Also characteristically found in the glands of adults (but not children) with Hashimoto thyroiditis are eosinophilic epithelial cells called Hrthle or Askanazy cells.16 In the juvenile form of Hashimoto thyroiditis, lymphoid follicles usually are not present.

FIGURE 46-2. Histology of Hashimoto thyroiditis. 125 (From LiVolsi VA, LoGerfo P, eds. Thyroiditis. Boca Raton, FL: CRC Press, 1981.)

FIGURE 46-3. Histology of Hashimoto thyroiditis. Note the presence of lymphocytic and plasma cell infiltration, germinal centers, and eosinophilic Askanazy cells. 200 (From LiVolsi VA, LoGerfo P, eds. Thyroiditis. Boca Raton, FL: CRC Press, 1981.)

CLINICAL FEATURES Patients may present in one of or a combination of the following three ways: with a goiter or thyroid nodule; with hypothyroidism; or with thyrotoxicosis, in which the condition is often referred to as hashitoxicosis. It appears likely that in most of the latter patients, the thyrotoxicosis represents the coexistence of Graves disease or silent thyroiditis with Hashimoto thyroiditis. The clinical features of the hyperthyroidism in patients with an apparent combination of Graves and Hashimoto disorders are essentially identical to the features of thyrotoxicosis in patients with only Graves disease. It is probable, however, that patients with Graves disease and coexistent Hashimoto thyroiditis have an increased chance of spontaneous remission of the thyrotoxicosis and may be more prone to develop spontaneous hypothyroidism. Goiter is a common presentation for patients with Hashimoto thyroiditis. In these patients, the gland is characteristically firm and bosselated, often with a palpable pyramidal lobe. Usually, patients have a diffusely enlarged gland; however, some patients may have a multinodular goiter or, more rarely, a single nodule. Although the goiter most often is asymptomatic, occasional patients develop pain and tenderness in the thyroid. In these patients, goiter growth may have been rapid, and the antithyroid antibody titers usually are high. The goiter also may cause local symptoms, such as pressure in the neck or dysphagia, but these symptoms are unusual. Hashimoto thyroiditis is not associated with cervical lymphadenopathy, and most patients (7580%) presenting with a goiter are euthyroid when initially evaluated. Other patients have hypothyroidism that may be apparent clinically, or it may be subclinical and documented only by thyroid function testing. Commonly, patients present with hypothyroid symptoms, and a careful physical examination reveals a small goiter, the presence of which the patient may have been unaware. The diagnosis of Hashimoto thyroiditis should be strongly suspected in a euthyroid or hypothyroid patient who presents with a thyroid gland that has the feel of a Hashimoto gland. This determination is more easily made with extensive experience so that the typical firm and somewhat rubbery consistency of the thyroid tissue can be recognized. This disorder, which can occur in children, may be present as an acute confusional state with a gradual impairment of cognitive function, even mimicking stroke.31a OTHER ASSOCIATED CLINICAL DISORDERS Patients with depression may have an increased prevalence of antithyroid antibodies, suggesting that autoimmune thyroiditis may predispose to this disorder.30 Additionally, a rapidly progressive dementia may occur in patients with Hashimoto disease (Hashimoto encephalopathy), which often responds to treatment with high dose glucocorticoid therapy.31 LABORATORY EVALUATION A full laboratory evaluation of any patient suspected of having Hashimoto thyroiditis must include a determination of thyroid function. Rarely, patients may be thyrotoxic, in which case elevated serum T4 and/or T3 levels usually are found, along with a suppressed TSH. However, most patients are euthyroid or hypothyroid when first evaluated. In the hypothyroid patients, one may variably see reduced serum T4 and T3 levels and T3 uptake with an elevated serum TSH. In patients with milder degrees of hypothyroidism, however, one may see only an elevated serum TSH concentration (or even a normal basal TSH value with an exaggerated TSH response to thyrotropin-releasing hormone). In these patients, the hypothyroidism probably is at an earlier stage, as manifested by the normal levels of serum T4and T3. Without thyroid hormone replacement, many of these patients would eventually develop frank clinical hypothyroidism with low serum T4 and T3 levels. An immunologic diagnosis of Hashimoto thyroiditis can be made by the determination of antimicrosomal and antithyroglobulin antibodies in the serum using various methods.32 The availability of recombinant TPO has led to the development of assays for anti-TPO, which have enhanced the sensitivity of the serologic diagnosis for Hashimoto

thyroiditis. As a consequence, the antimicrosomal antibody determination is becoming obsolete. In biopsy-proven Hashimoto thyroiditis, the antimicrosomal levels are elevated more commonly than are the antithyroglobulin levels. Low titers of antithyroid antibodies may be found in patients with a wide variety of thyroid disorders and in apparently asymptomatic patients who have no evidence of thyroid disease. It is possible that the presence of low levels of antithyroid antibodies represents an early stage of Hashimoto thyroiditis, before the development of a goiter or thyroid function abnormalities. Furthermore, seronegative Hashimoto thyroiditis, presumably related to an organ-restricted form of this autoimmune disorder, has been reported.33 Low titers of the antibodies also may represent a nonspecific response to thyroid tissue damage, as may be seen in patients with thyroid carcinoma. In many patients who present with a diffuse goiter or hypothyroidism, a thyroid RAI uptake and scan are not crucial to the management. When a patient presents with a solitary nodule, a fine-needle aspiration biopsy is indicated to evaluate the possibility of malignancy (see Chap. 34). Of note is evidence that the presence of a cold nodule on nuclear scan in Hashimoto thyroiditis may reflect an even higher risk for malignancy than is normally found in such nodules.34 If a thyroidal RAI uptake is done, one finds normal or low values in most patients, with rare patients showing an elevated uptake. Additionally, patients with Hashimoto thyroiditis may have a positive perchlorate discharge test (see Chap. 34), but this finding is also seen in patients with other thyroid disorders, such as in Graves disease after iodine-131 (131I) therapy. The scan in patients with Hashimoto thyroiditis often is symmetric but with a patchy uptake evident.35 In summary, to establish a diagnosis of Hashimoto thyroiditis in patients presenting with a goiter or thyroid function abnormalities, the determination of the titer of the anti-TPO antibody is the most helpful laboratory test available. A thyroid fine needle aspiration biopsy is not routinely necessary, but can be helpful in establishing the diagnosis in antibody negative subjects (~10%). In patients with a solitary nodule, it should be undertaken to rule out a thyroid malignancy. Ultrasonography demonstrates reduced thyroid echogenicity in both Hashimoto and Graves diseases.35a This finding on ultrasound may be a useful predictor of these autoimmune diseases.35a TREATMENT Treatment usually involves the administration of physiologic doses of thyroid hormone, for example 0.1 to 0.15 mg per day of L -T4.(In elderly hypothyroid patients or those with coronary disease, the initial dosage is lower, 0.01250.025 mg per day.) In patients with clinical or even subclinical hypothyroidism and Hashimoto thyroiditis, the therapy is directed at the hypothyroidism. In patients with a goiter but no hypothyroidism, the authors usually still consider the use of thyroid hormone replacement, for two reasons. First, the thyroid hormone administration may shrink the goiter, or at least limit growth. In patients with local symptoms (such as dysphagia), the thyroid hormone may lead to some improvement. The second reason relates to the natural history of the disease. In a number of patients with Hashimoto thyroiditis, hypothyroidism develops eventually.36 Therefore, it seems reasonable to institute replacement in these patients early in the course of the disease rather than waiting until symptoms develop. This decision, however, is optional, and it would not be unreasonable to observe such patients carefully and to institute T4 replacement when hypothyroidism actually occurs. Previously, it was thought that hypothyroidism resulting from Hashimoto thyroiditis was likely permanent. However, more recent evidence has shown that some patients with hypothyroidism that is due to Hashimoto thyroiditis may have transient hypothyroidism. Approximately 20% of such patients have a spontaneous recovery of thyroid function while taking thyroid hormone replacement.37 Another study38 demonstrated that the mechanism of such recovery likely involves a disappearance of the TSH-blocking antibodies that previously were the apparent cause of the hypothyroidism. In the atypical, uncommon patient with rapid thyroid growth and pain, Hashimoto thyroiditis can be treated with corticosteroids, primarily to alleviate the local symptoms. An initial dosage of 60 to 80 mg per day of oral prednisone has been recommended, with a gradual tapering over a period of 3 to 4 weeks. Surgery is rarely indicated in patients with Hashimoto thyroiditis; it usually is undertaken to relieve severe local effects (e.g., obstructive symptoms unresponsive to corticosteroids), to exclude the possible presence of a thyroid malignancy in a patient with a solitary thyroid nodule, or to treat a patient whose goiter continues to grow despite thyroid hormone administration. The long-term prognosis of most patients with Hashimoto thyroiditis is good, because the hypothyroidism can be corrected by L -T4, and the goiter usually is not large enough to cause serious problems. However, patients with Hashimoto thyroiditis may be at a significantly higher risk of developing a thyroid lymphoma39 (see Chap. 40). Although this is uncommon, any increase in thyroid gland size during L -T4 administration necessitates the exclusion of a malignant change.

SILENT THYROIDITIS ETIOLOGY

During the past two decades, physicians increasingly have observed patients presenting with signs and symptoms of thyrotoxicosis; a small, painless, nontender goiter; and a low RAI uptake. Furthermore, the thyrotoxicosis has been transient. This syndrome, usually termed silent thyroiditis (absence of thyroid pain)40 also has been termed hyperthyroiditis,41 painless subacute thyroiditis, atypical subacute thyroiditis, and lymphocytic thyroiditis with spontaneously resolving hyperthyroidism. It is apparent from the range of the terms used to describe this disorder that the etiology and relation to several other forms of thyroiditis are somewhat unclear. Additionally, the reported frequency with which silent thyroiditis causes thyrotoxicosis ranges widely, from 4% to 23%.42, 43 The clinical course resembles that seen in subacute thyroiditis in regard to thyroid function. At biopsy, however, the thyroid tissue from patients with silent thyroiditis reveals findings more suggestive of Hashimoto disease (chronic lymphocytic thyroiditis), because extensive lymphocytic infiltration is observed. Often, severe disruption of the follicles is seen, Hrthle cell transformation is absent, and occasionally lymphoid follicles are present.44 Despite these histologic changes that are compatible with Hashimoto thyroiditis, as many as 40% of patients with silent thyroiditis do not have significantly elevated titers of antimicrosomal or antithyroglobulin antibodies, the serologic hallmark of Hashimoto disease.7 Silent thyroiditis is associated with an excess of HLA-DR3.45 This association links silent thyroiditis to autoimmune thyroiditis rather than to subacute thyroiditis.46 Other observations that implicate an autoimmune etiology for this entity are the onset of silent thyroiditis (which has been observed after the resection of a thymoma)47 and the appearance of a mixture of TSH-receptor antibodies that may mediate the various phases of thyroid function that are observed.48 Silent thyroiditis has also been reported in conjunction with amiodarone7 and lithium, and with the use of interferon-a in the treatment of hepatitis C patients.49, 50 The prevalence of antibodies before interferon treatment generally is low,51 but treatment may induce antibodies in as many as 15% of patients, most of whom may experience thyrotoxicosis and/or hypothyroidism.52 A report53 was made of silent thyroiditis in a husband whose wife previously had subacute thyroiditis. Similar to subacute thyroiditis, the thyrotoxicosis54 in silent thyroiditis results from damage to follicular cells by the inflammatory process, with the subsequent release of excessive amounts of thyroid hormone. This process also impairs the capacity of the thyroid to trap iodinehence, the observed low RAI uptake. Because the damaged follicular cells temporarily cannot make additional thyroid hormone, patients experience a euthyroid phase and then become hypothyroid when the stored thyroid hormone is depleted; in most cases, they ultimately are restored to euthyroidism. The presence of high titers of anti-TPO (especially in the postpartum variety) may predispose to permanent hypothyroidism. CLINICAL FEATURES Patients with silent thyroiditis present with the signs and symptoms of thyrotoxicosis, such as nervousness, weight loss, and often palpitations.7 Thyroid examination, although variable, often reveals a slightly enlarged, firm, nontender gland. The thyroid, however, may be normal in size, but if enlarged, it usually is symmetric without nodules. These patients do not have the infiltrative eye signs or skin changes (pretibial myxedema) seen in Graves hyperthyroidism, a distinction that is helpful clinically. A history of a recent viral illness is not usually obtained. The differentiation from Graves hyperthyroidism is important as the prevalence of silent thyroiditis ranges from 1% of cases of thyrotoxicosis in Wales to 23% in some parts of the United States.7 As in subacute thyroiditis, patients characteristically progress from a thyrotoxic state to euthyroid and hypothyroid phases before recovering normal thyroid function. The thyrotoxic phase of silent thyroiditis usually is transient, lasting ~2 to 4 months (rare patients may have thyrotoxicosis lasting 12 months44). Occasional patients, however, may develop permanent, rather than transient, hypothyroidism, especially if the antimicrosomal (or anti-TPO) antibody titers are elevated. This, along with the HLA findings, suggests that silent thyroiditis in these patients represents a form of Hashimoto disease. LABORATORY EVALUATION AND TREATMENT The diagnosis of silent thyroiditis is based on the finding of hyperthyroidism in the context of a low RAI uptake and a painless thyroid gland. These patients usually present in the thyrotoxic phase, and laboratory evaluation shows an elevated serum, T4, T3 and T3 uptake, and a suppressed TSH. TSH suppression is below the lower limit of detectability with a third-generation TSH assay in 90% of subjects with silent thyroiditis,10 similar to Graves disease.10 Anti-TPO or antithyroglobulin antibodies may only be present in relatively low titers (although some patients have raised titers), and the white blood cell count typically is normal. The erythrocyte sedimentation rate is normal or slightly elevated. An RAI uptake is low (often 1% to 2%) at 6 and 24 hours,55 differentiating this form of thyrotoxicosis from Graves hyperthyroidism or Plummer disease (see Chap. 34). This is important because the management of thyrotoxic patients with silent thyroiditis is different from that for Graves hyperthyroidism or toxic nodular goiter. Because the hyperthyroxinemia in silent thyroiditis, as in subacute thyroiditis, results from the excessive release of thyroid hormone from damaged thyroid follicular cells and not from excess synthesis, antithyroid medication (propylthiouracil, methimazole) or 131I therapy are inappropriate. The thyrotoxicosis should normally be treated symptomatically with b-adrenergic blockade, such as propranolol, 10 to 40 mg orally three or four times daily, if there are

no contraindications to this form of therapy (i.e., congestive heart failure or asthma). Decreased T4 to T3 conversion may also be accomplished by administering sodium ipodate in doses of 500 mg per day. Ipodate has been demonstrated to be56 It may effective in rapidly decreasing T3 levels and symptoms.It may be continued for 2 to 6 weeks as the clinical syndrome resolves in individuals with severe destructive thyrotoxicosis. In occasional patients (10% to 20%) who do not respond to b-blockade, prednisone may be tried. Prednisone, started at 50 mg per day orally, then tapered over 4 weeks, has been used in patients with silent thyroiditis. Rarely, surgery may be indicated in patients with recurrent episodes of severe thyrotoxic silent thyroiditis. Patients in the hypothyroid phase of silent thyroiditis often do not need thyroid hormone replacement because the hypothyroidism is usually transient and often symptomatically mild. Furthermore, as in subacute thyroiditis, the elevated serum TSH at this time may enhance the spontaneous return of normal thyroid function. In patients who are symptomatic and require treatment, it has been suggested that a lower-than-usual dose of thyroid hormone replacement be given to allow the TSH to remain slightly elevated. This may enhance the subsequent recovery of thyroid function.57 Similarly, such patients may be treated with L -T4 for ~9 months and then receive a tapering dose of thyroid hormone to determine if recovery of thyroid function has occurred.58 In patients who appear to have developed permanent hypothyroidism (persistence of low serum T4, elevated serum TSH levels, and hypothyroid symptoms for longer than 4 to 9 months), long-term thyroid hormone replacement is indicated. Thyroid biopsy usually is not necessary to make the diagnosis. In many patients with this disorder, the thyroid gland is only barely enlarged; therefore, this procedure can be technically difficult. However, all nonpregnant patients with suspected hyperthyroidism, whose diagnosis appears to be other than Graves disease or a toxic nodular goiter by examination, should have an RAI uptake to establish the presence of this disorder. If the gland is painless and the uptake is low, a diagnosis of silent thyroiditis can usually be made. These patients can be reassured that the thyrotoxic phase is generally transient, and b-blockade is the treatment of choice if relief of thyrotoxic symptoms is necessary. Studies of the natural history of silent thyroiditis suggest that a number of patients have recurrences (in one study, 11%59); patients with chronically elevated titers of antithyroid antibodies tend to develop permanent hypothyroidism. Because this disorder was recognized only relatively recently, however, additional studies are needed to document more clearly the long-term features of the disorder. At least five conditions exist in which symptoms of thyrotoxicosis may be associated with a low RAI uptake: subacute thyroiditis; silent thyroiditis; recent iodide ingestion by patients with Graves disease, toxic multinodular goiter, or toxic adenoma; thyrotoxicosis factitia; and ectopic hyperthyroidism (struma ovarii or functioning metastatic thyroid carcinoma).

POSTPARTUM THYROIDITIS
Like spontaneous lymphocytic thyroiditis, postpartum thyroiditis is associated with HLA-DR3, but an association with DR5 also is observed.45 This postpartum entity occurs in 5% to 9% of all pregnancies.58, 60, 61 and 62 These observations are of particular interest for the probable autoimmune nature of this and the classic form of silent thyroiditis, as it is well established that many autoimmune diseases flare during the postpartum period after resolution of the immune suppression normally seen to occur during pregnancy. Higher ratios of helper (CD4+) to suppressor (CD8+) T cells seen in the postpregnancy state may play a critical role in the development of this postpartum dysfunction.63 Patients with type 1 diabetes mellitus are especially prone to the development of postpartum thyroiditis; as many as 25% experience this disorder.59, 64 CLINICAL FEATURES AND LABORATORY EVALUATION Fewer than 33% of affected women have a classic thyrotoxic, euthyroid, hypothyroid pattern followed by euthyroid recovery.7 Many patients may present only as hypothyroidism when the thyrotoxic phase is mild or brief.64 For those presenting with clear symptoms or laboratory evidence of thyrotoxicosis or hyperthyroxinemia and thyrotropin suppression, differentiation of a postpartum flare of Graves hyperthyroidism with an RAI uptake is essential to appropriately intervene. This diagnostic intervention may be problematic, because it interrupts the course of breast-feeding at a critical stage of bonding and necessitates discarding RAI-contaminated milk for 1 week after the ingestion of the diagnostic iodine dose. It is not enough to assume that thyrotoxicosis in the postpartum period is due to thyroiditis or even preexisting Graves disease; one study has demonstrated that 26 of 96 women with prior Graves disease who subsequently developed postpartum thyrotoxicosis actually had low uptake thyroiditis.65 Postpartum thyroiditis typically occurs with subsequent pregnancies among those who are HLA-DR5 positive45 and may be partially predicted by the presence of thyroid autoantibodies which are seen in ~10% of affected women at 16 weeks of gestation.58,66 Titers of antithyroid antibodies reach peak levels 3 to 7 months postpartum, corresponding to the thyrotoxic (~14 weeks) and hypothyroid (19 weeks) phases of the process.58 Approximately 25% of patients with prior episodes of postpartum thyroiditis subsequently develop permanent hypothyroidism.62 The majority, therefore, should be expected to recover from their postpartum episode and maintain euthyroidism.7,58 TREATMENT Treatment of the thyrotoxic phase with a b-blocker may be necessary for symptomatic individuals. L -T4 may be indicated for symptoms during the hypothyroid phase. As most patients do recover normal thyroid function, T4 should be withdrawn after 6 to 9 months of replacement.

RIEDEL THYROIDITIS
Riedel thyroiditis is a most unusual and rare form of thyroid disease, which is characterized by extensive fibrosis of the thyroid gland. In these patients, the gland is hard, woody in consistency, and often fixed to the adjacent tissues. Occasionally, the gland may be asymmetric, presenting as a unilateral mass. In many patients with Riedel thyroiditis, local symptoms, including hoarseness (recurrent laryngeal nerve involvement) or dysphagia, are present. Pathologically, the gland shows an inflammatory fibrous process consisting of activated T lymphocytes, macrophages, aggregates of B lymphocytes, eosinophilia, and deposition of eosinophilic products, which may stimulate fibrosis67 both within and outside the thyroid capsule (Fig. 46-4). Because invasion of local tissues typically occurs, the pathologic appearance initially may resemble invasive anaplastic carcinoma.68 The etiology is unclear, but it may represent a variant of Hashimoto thyroiditis, as evidenced by the mononuclear infiltrate, the detection of thyroid-specific autoantibodies in many patients, and the response to glucocorticoid treatment.67 Riedel thyroiditis may represent the extension into the thyroid of a more generalized, invasive fibrosis syndrome, such as retroperitoneal fibrosis. Indeed, patients have been described with both Riedel thyroiditis and retroperitoneal fibrosis. This diagnosis should be considered in patients who present with a woody, hard thyroid gland that is fixed in position and associated with local symptoms. Laboratory evaluationincluding thyroid function tests, erythrocyte sedimentation rate, and thyroid antibodiesmay be normal, but a thyroid scan may reveal a nonspecific reduction in uptake in the affected areas. The diagnosis can be best established by a biopsy; however, specimens from patients with other thyroid disorders (e.g., Hashimoto thyroiditis) also may contain notable fibrotic tissue. Surgical treatment may be required if there are significant local symptoms or evidence of tracheal compression. Corticosteroids occasionally have been helpful in reducing the goiter size and alleviating local symptoms.69

FIGURE 46-4. Riedel thyroiditis. Note fibrous tissue proliferation with cellular exudate consisting of lymphocytes, plasma cells, and polymor phonuclear leukocytes. 650 (From Thomson JA, Jackson IMD, Duguid WP. The effect of steroid therapy on Riedel's thyroiditis. Scott Med J 1968; 13:13.)

RADIATION THYROIDITIS
Radiation injury to the thyroid gland may affect it in several ways, depending on the dose and type of radiation. Pathologically, the thyroid changes seen after acute

radiation injury (radiation thyroiditis) include cellular necrosis, follicular disruption, infiltration by neutrophils, and local edema.70 Soon thereafter, vascular damage is evident in the form of hemorrhage and thrombosis. The more long-term changes of radiation thyroiditis (years after the injury) include fibrosis, atrophy of the follicles, nuclear abnormalities, lymphocytic infiltration, oxyphilic changes, and arteriolar hyalinization. RADIOISOTOPE THERAPY-INDUCED INJURY Internal irradiation (e.g., 131I or 125I) for the treatment of hyper-thyroidism or thyroid cancer can induce a radiation thyroiditis that may abruptly present clinically as pain and tenderness in the thyroid, swelling in the neck, and transient thyrotoxicosis. Indeed, thyroid storm occasionally has occurred after 131I treatment of thyrotoxicosis.71 These acute changes usually occur within 1 week of the RAI treatment and subside within 3 to 4 weeks.70 As a consequence of 131I or 125I therapy for hyperthyroidism, the most common long-term clinical abnormality relates to the development of hypothyroidism. Therefore, after RAI therapy, patients need to be monitored indefinitely for the occurrence of hypothyroidism. However, there does not appear to be any increased incidence of thyroid cancer in patients receiving RAI therapeutically. This absence of an increased incidence of thyroid cancer after 131I or 125I therapy or high-dose external irradiation for head and neck cancer may relate to the fact that very high levels of intrathyroidal radiation are achieved, thereby destroying thyroid cells and obviating the potential for the later development of thyroid cell carcinoma. EXTERNAL RADIATION-INDUCED INJURY Three effects may be seen on the thyroid gland after external irradiation of the thyroid: goitrous formation caused by thyroiditis, the induction of thyroid tumors, and the development of hypothyroidism. An increased incidence of goiter and of thyroid carcinoma has been observed in patients who previously received external irradiation to the head and neck in infancy or childhood for the treatment of such conditions as enlarged thymus or tonsils or for acne vulgaris (radiation dose usually <10 Gy [<1000 rads]; see Chap. 40). The development of thyroid cancer 10 to 40 years later after this type of low-dose external irradiation may occur. The effects are dose related from 0.5 to 10 Gy (501000 rads). Hypothyroidism does not occur at this dose range. The radiation doses used for treatment of head or neck cancers or lymphomas (4050 Gy) may lead to the development of thyrotoxic thyroiditis with subsequent hypothyroidism months to years after the radiation exposure.72 Experience indicates that thyroid cancer is not a risk from this treatment.

OTHER FORMS OF THYROIDITIS PERINEOPLASTIC THYROIDITIS

Perineoplastic thyroiditis is a term used to describe thyroiditis in a gland containing a thyroid neoplasm. Most commonly, this refers to lymphocytic thyroiditis in the tissue surrounding a thyroid neoplasm, most often a papillary thyroid carcinoma (PTC), but this finding has also been reported in medullary thyroid cancer. The frequency with which perineoplastic lymphocytic infiltration in PTC occurs varies with race and sex. It has been suggested that the lymphocytic thyroiditis represents an autoimmune reaction to the thyroid neoplasm and that it may be beneficial in reducing the spread of the tumor. Lymphocytes derived from tissue containing PTC are unusual, because they express both CD4 and CD8 markers and may display intense cytolytic activity in vitro. Significantly improved outcomes occur in PTC and medullary thyroid cancer patients who have lymphocytic infiltration when compared with patients with these tumors and no lymphocytic tissue response.73 This finding may, if confirmed, be useful in treatment planning for patients with thyroid cancer.74 Despite the increased incidence of thyroid lymphoma in patients with Hashimoto thyroiditis, clinical lymphocytic thyroiditis generally is not regarded as a significant premalignant lesion and was found in only ~1% of PTC cases in one series.73 Therefore, the removal of thyroid glands containing chronic lymphocytic thyroiditis, which might be aimed at preventing the development of a thyroid neoplasm, does not seem justified. The coexistence of thyroiditis and a neoplasm in the same gland indicates that caution must be exercised in the interpretation of a thyroid biopsy that demonstrates thyroiditis. Because the lymphocytic changes are often noted near or mixed with the tumor, a biopsy of a tumor may show only thyroiditis. Multiple biopsy samples should allow sampling of actual tumor tissue and prevent the cancer from being missed. PALPATION THYROIDITIS Palpation of the thyroid gland also can lead to a type of thyroiditis referred to as palpation thyroiditis. Most thyroid glands removed from patients who had recent thyroid physical examinations may reveal this form of inflammation.75 Histologic examination demonstrates infiltration of lymphocytes, plasma cells, and occasionally giant cells. Hemorrhage also may be present. The clinical significance of palpation thyroiditis is unclear; however, few adverse effects are expected in most patients. Transient thyroiditis causing thyrotoxicosis may occur after throat trauma, thyroid biopsy, parathyroid surgery, and after surgical resection of hypopharyngeal cancer. These patients most likely develop a form of palpation thyroiditis because of trauma to the thyroid at the time of surgery. The possibility of such thyroiditis should be considered in postoperative patients who develop signs or symptoms suggestive of hyperthyroidism. GRANULOMATOUS DISORDERS Infiltrative disorders, such as sarcoidosis or eosinophilic granuloma, may involve the thyroid gland and result in a lymphocytic thyroiditis, probably occurring as a reaction to the induced damage in the gland. Noncaseating granulomas have been noted occasionally in the thyroid glands of patients with sarcoidosis and usually are found between the thyroid follicles. If such granulomas are noted in a biopsy or surgical specimen from the thyroid, the patient should be evaluated for the systemic presence of sarcoidosis. CHAPTER REFERENCES
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. Volp R. Acute and subacute thyroiditis. Pharmacol Ther 1976; 1:171. Abe K, Taguchi T, Ohumo A, et al. Recurrent acute suppurative thyroiditis. Am J Dis Child 1978; 132:990. Smilack JD, Rodolfo A. Coccidioidal infection of the thyroid. Arch Intern Med 1998; 158:89. Golshan MM, McHenry CR, deVente J, et al. Acute suppurative thyroiditis and necrosis of the thyroid gland: a rare endocrine manifestation of acquired immunodeficiency syndrome. Surgery 1997; 233:593. Dolgin C, LoGerfo P. Surgical management of thyroiditis. In: LiVolsi VA, LoGerfo P, eds. Thyroiditis. Boca Raton, FL: CRC Press, 1981:174. Takai S, Matsuzaka F, Miyauchi A, et al. Internal fistula as a route of infection in acute suppurative thyroiditis. Lancet 1979; 1:751. Ross DS. Syndromes of thyrotoxicosis with low radioactive iodine uptake. Endocrinol Metab Clin North Am 1998; 27:169. Eylan E, Zmucky R, Sheba C. Mumps virus and subacute thyroiditis: evidence of a causal association. Lancet 1957; 1:1062. Bech K, Nerup J, Thomsen M, et al. Subacute thyroiditis de Quervain: a disease associated with a HLA-B antigen. Acta Endocrinol (Copenh) 1977; 86:504. Ito M, Takamatsu J, Yoshida S, et al. Incomplete thyrotroph suppression determined by third generation thyrotropin assay in subacute thyroiditis compared to silent thyroiditis or hyperthyroid Graves' disease. J Clin Endocrinol Metab 1997; 82:616. Hannibal E, LiVolsi V. Subacute thyroiditis. In: LiVolsi VA, LoGerfo P, eds. Thyroiditis. Boca Raton, FL: CRC Press, 1981:31. Iitaka M, Momotani N, Ishii J. Incidence of subacute thyroiditis recurrences after a prolonged latency: 24-year survey. J Clin Endocrinol Metab 1996; 81:466. Tikkanen M, Lamberg BA. Hypothyroidism following subacute thyroiditis. Acta Endocrinol (Copenh) 1982; 101:348. Barbesino G, Tomer Y, Concepcion ES, et al. International Consortium for the Genetics of Autoimmune Thyroid Disease. Linkage analysis of candidate genes in autoimmune thyroid disease II. Selected gender-related genes and the X-chromosome. J Clin Endocrinol Metab 1998; 83:3290. Volp R. Autoimmune thyroiditis. In: Braverman LE, Utiger RD, eds. Werner and Ingbar's the thyroid, 5th ed. Philadelphia: JB Lippincott, 1991:921. Drexhage H, Bottazzo G, Bitensky L, et al. Thyroid growth-blocking antibodies in primary myxedema. Nature 1981; 289:594. Okamoto Y, Hamada N, Saito H, et al. Thyroid peroxidase activity-inhibiting immunoglobulins in patients with autoimmune thyroid disease. J Clin Endocrinol Metab 1989; 68:730. Chiovato L, Vitti P, Santini F, et al. Incidence of antibodies blocking thyrotropin effect in vitro in patients with euthyroid or hypothyroid autoimmune thyroiditis. J Clin Endocrinol Metab 1990; 71:40. Braun J, Donner H, Siegmund T, et al. CTLA-4 promoter variants in patients with Graves' disease and Hashimoto's thyroiditis. Tissue Antigens 1998; 51:563.

19a. Tomer Y, Barberino G, Greenberg DA, et al. Mapping the major susceptibility loci for familial Graves' and Hashimoto's diseases: evidence for genetic heterogeneity and gene interaction. J Clin Endocrinol Metab 1999; 84:4656. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. Weetman A, McGregor AM. Autoimmune thyroid disease: further developments in our understanding. Endocr Rev 1994; 15:788. Hoekmank K, von Blombergvan Der Flier BME, Wagstaff J, et al. Reversible thyroid dysfunction during treatment with GM-CSF. Lancet 1991; 338:541. Atkins MB, Mier JW, Parkinson DR, et al. Hypothyroidism after treatment with interleukin-2 and lymphokine-activated killer cells. N Engl J Med 1988; 318:1557. Mitsiades N, Poulaki V, Kotoula V, et al. Fas/Fas ligand up-regulation and Bcl-2 down-regulation may be significant in the pathogenesis of Hashimoto's thyroiditis. J Clin Endocrinol Metab 1998; 83:199. Giordano C, Stassi G, De Maria R, et al. Potential involvement of FAS and its ligand in the pathogenesis of Hashimoto's thyroiditis. Science 1997; 275:960. Doniach D. Hashimoto's thyroiditis and primary myxedema viewed as separate entities. Eur J Clin Invest 1981; 11:245. Brown J, Solomon DH, Beall GN, et al. Autoimmume thyroid disease: Graves' and Hashimoto's. Ann Intern Med 1978; 88:379. Strakosch C, Wenzel B, Row V, Volp R. Immunology of autoimmune thyroid diseases. N Engl J Med 1982; 307:1499. Botazzo G, Pujol-Borrell R, Hanafusa T. Role of HLA-DR expression and antigen presentation in induction of endocrine autoimmunity. Lancet 1983; 2:1115. Yoshida H, Amino N, Yagawa K, et al. Association of serum antithyroid antibodies and lymphocytic infiltration of the thyroid gland: studies of 70 autopsied cases. J Clin Endocrinol Metab 1978; 46:859. Nemeroff CB, Simon JS, Haggerty JJ, Evans DL. Antithyroid antibodies in depressed patients. Am J Psychiatry 1983; 14:840. Wilhelm-Gossling C, Weckbecker K, Brabant EG, Dengler R. Autoimmune encephalopathy in Hashimoto's thyroiditis. A differential diagnosis in progressive dementia syndrome. Deutsche Medizinische Wochenschrift 1998; 123:279.

31a. Watemberg N, Willis D, Pollock JM. Encephalopathy as the presenting symptom of Hashimoto's thyroiditis. J Child Neurol 2000; 15:66.

32. Engler H, Staub J-J, Althaus B, et al. Assessment of antithyroglobulin and antimicrosomal autoantibodies in patients with autoimmune thyroid disease: comparison of haemagglutination assay, enzyme-linked immunoassay and radio-ligand assay. Clin Chim Acta 1989; 179:251. 33. Baker JR Jr, Saunders NB, Wartofsky L, et al. Seronegative Hashimoto thyroiditis with thyroid autoantibody production localized to the thyroid.Ann Intern Med 1988; 108:26. 34. Ott RA, Calandra DB, McCall A, et al. The incidence of thyroid carcinoma in patients with Hashimoto's thyroiditis and solitary cold nodules. Surgery 1985; 98:1203. 35. Yarman S, Mudun A, Alagol F, et al. Scintigraphic varieties in Hashimoto's thyroiditis and comparison with ultrasonography. Nucl Med Commun 1997; 18:951. 35a. Pederson OM, Aordal NP, Larssen TB, et al. The value of ultrasonography in predicting autoimmune thyroid disease. Thyroid 2000; 10:251. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. Gordin A, Lamberg BA. Natural course of symptomless autoimmune thyroiditis. Lancet 1975; 2:1234. Takasu N, Komiya I, Asawa T, et al. Test for recovery from hypothyroidism during thyroxine therapy in Hashimoto's thyroiditis. Lancet 1990; 336:1084. Takasu N, Yamada T, Takasu M, et al. Disappearance of thyrotropin-blocking antibodies and spontaneous recovery from hypothyroidism in autoimmune thyroiditis. N Engl J Med 1992; 326:513. Holm L, Blomgren H, L';whagen T. Cancer risks in patients with chronic lymphocytic thyroiditis. N Engl J Med 1985; 312:601. Papapetrou P, Jackson IMD. Thyrotoxicosis due to silent thyroiditis. Lancet 1975; 1:36. Jackson IMD. Hyper-thyroiditis: a diagnostic pitfall. N Engl J Med 1975; 293:661. Vitug A, Goldman J. Silent (painless) thyroiditis: evidence of a geographic variation in frequency. Arch Intern Med 1985; 145:473. Nikolai TF, Brosseau J, Kettrick MA, et al. Lymphocytic thyroiditis with spontaneously resolving hyperthyroidism (silent thyroiditis). Arch Intern Med 1980; 140:478. Woolf P. Transient painless thyroiditis with hyperthyroidism: a variant of lymphocytic thyroiditis? Endocr Rev 1980; 1:411. Farid NR, Hawe B, Walfish PG. Association of HLA-DR3 and HLA-DR5 to antigens in the painless thyroiditis with transient thyrotoxicosis syndrome. Clin Endocrinol Metab 1983; 19:699. Volp R. Is silent thyroiditis an autoimmune disease? [Editorial]. Arch Intern Med 1988; 148:1907. Murao S, Yoshinouchi T, Sato M, et al. Silent thyroiditis after excision of a thymoma. Intern Med 1998; 37:604. Sarlis NJ, Brucker-Davis F, Swift JP, et al. Graves' disease following thyrotoxic painless thyroiditis: analysis of antibody activities against the thyrotropin receptor in two cases. Thyroid 1997; 7:829. Wada M, Shimoyama T, Hamada A, Fukuchi M. Antithyroid peroxidase antibody and development of silent thyroiditis during interferon-2a treatment of chronic hepatitis C. Am J Gastroenterol 1995; 90:1366. Roti E, Minelli R, Giuberti T, et al. Multiple changes in thyroid function in patients with chronic HCV hepatitis treated with recombinant interferon-a. Am J Med 1996; 101:482. Boardas J, Rodriguez-Espinosa J, Enriquez J, et al. Prevalence of thyroid autoantibodies is not increased in blood donors with hepatitis C virus infection. J Hepatol 1995; 22:611. Custro N, Montalto G, Scafidi V, et al. Prospective study on thyroid autoimmunity and dysfunction related to chronic hepatitis C and interferon therapy. J Endocrinol Invest 1997; 20:374. Morrison J, Caplan RH. Typical and atypical (silent) subacute thyroiditis in a wife and husband. Arch Intern Med 1978; 138:45. Barclay ML, Brownlie BE, Turner JG, Wells JE. Lithium associated thyrotoxicosis: a report of 14 cases with statistical analysis of incidence. Clin Endocrinol (Oxf) 1994; 40:759. Hennessey JV, Berg LA, Ibrahim M, Markert RJ. Evaluation of the early (5-6 hour) iodine-123 uptake for diagnosis and treatment planning in Graves' disease. Arch Intern Med 1995; 155:621. Aren R, Munipalli B. Ipodate therapy in patients with severe destruction induced thyrotoxicosis. Arch Intern Med 1996; 156:1752. Nikolai TF, Coombs GJ, McKenzie AK, et al. Treatment of lymphocytic thyroiditis with spontaneously resolving hyperthyroidism (silent thyroiditis). Arch Intern Med 1982; 142:2281. Lazarus JH. Prediction of postpartum thyroiditis. Eur J Endocrinol 1998; 139:12. Alvarez-Marfany M, Roman SH, Drexler AJ, et al. Long-term prospective study of postpartum thyroid dysfunction in women with insulin-dependent diabetes mellitus. J Clin Endocrinol Metab 1994; 79:10. Davies TF. The thyroid immunology of the postpartum period. Thyroid 1999; 9:675. Hayslip CC, Fein HG, O'Donnell VM, et al. The value of serum antimicrosomal antibody testing in screening for symptomatic postpartum thyroid dysfunction. Am J Obstet Gynecol 1988; 159(1):203. Amino N, Miyai K, Kuro R, et al. Transient post-partum hypothyroidism: fourteen cases with autoimmune thyroiditis. Ann Intern Med 1977; 87:155. Stagnaro-Green A, Roman S, Cobin R, et al. A prospective study of lymphocyte initiated immunosuppression in normal pregnancy: evidence of a T-cell etiology for postpartum thyroid dysfunction. J Clin Endocrinol Metab 1992; 74:645. Stagnaro-Green A. Postpartum thyroiditis: prevalence, etiology and clinical implications. Thyroid Today 1993; 16:1. Momotani N, Noh J, Ishikawa N, et al. Relationship between silent thyroiditis and recurrent Graves' disease in the postpartum period. J Clin Endocrinol Metab 1994; 79:285. Roti E, Emerson C. Postpartum thyroiditis. J Clin Endocrinol Metab 1992; 74:3. Heufelder AE, Goellmer JR, Bahn RS, et al. Tissue eosinophilia and eosinophil degranulation in Riedel's invasive fibrous thyroiditis. J Clin Endocrinol Metab 1996; 81:977. Wan SK, Chan JK, Tang SK. Paucicellular variant of anaplastic thyroid carcinoma: a mimic of Riedel's thyroiditis. Am J Clin Pathol 1996; 105:388. Bagnasco M, Passalacqua G, Pronzato C, et al. Fibrous invasive (Riedel's) thyroiditis with critical response to steroid treatment. J Endocrinol Invest 1995; 18:305. LiVolsi VA. Radiation-associated thyroiditis. In: LiVolsi VA, LoGerfo P, eds. Thyroiditis. Boca Raton, FL: CRC Press, 1981:113. McDermott MT, Kidd GS, Dodson LE, Hofeldt FD. Radioiodine-induced thyroid storm: case report and literature review. Am J Med 1983; 75:353. Aizawa T, Watanabe T, Suzuki N, et al. Radiation-induced painless thyrotoxic thyroiditis followed by hypothyroidism: a case report and review of the literature. Thyroid 1998; 8:273. Matsubayashi S, Kawai K, Matsumoto Y, et al. The correlation between papillary thyroid carcinoma and lymphocytic infiltration in the thyroid gland. J Clin Endocrinol Metab 1995; 80:3421. Baker JR. The immune response to papillary thyroid carcinoma. J Clin Endocrinol Metab 1995; 80:3419. LiVolsi VA. Palpation thyroiditis. In: LiVolsi VA, LoGerfo P, eds. Thyroiditis. Boca Raton, FL: CRC Press, 1981:43.

CHAPTER 47 THYROID DISORDERS OF INFANCY AND CHILDHOOD Principles and Practice of Endocrinology and Metabolism

CHAPTER 47 THYROID DISORDERS OF INFANCY AND CHILDHOOD

WELLINGTON HUNG HypothalamicPituitaryThyroid Interrelationships Physiologic Changes in Thyroid Function Tests Hypothyroidism in the Neonate and Child Etiology Transient Disorders of Thyroid Function Clinical Findings in Hypothyroidism Laboratory Findings Therapy Neonatal Screening for Hypothyroidism Human Breast Milk and Neonatal Screening Hyperthyroidism in the Child Clinical Findings Laboratory Findings Treatment Neonatal Graves Disease and Fetal Thyrotoxicosis Clinical Findings Treatment Goiters Chronic Lymphocytic Thyroiditis Colloid or Simple Goiter Evaluation of the Thyroid Nodule Clinical Findings Laboratory Findings Treatment Thyroid Cancer Etiology Pathology Clinical Findings Diagnosis Treatment Follow-up Chapter References

Thyroid hormones are important for the general growth and development of the infant and child, particularly in the differentiation and function of the nervous system. They are essential elements in the maturational events involved in the transition of the neonate to the adult.

HYPOTHALAMICPITUITARYTHYROID INTERRELATIONSHIPS
In the human fetus, the period of maturation of thyroid function extends throughout gestation and into the neonatal period.1,2 Thyrotropin-releasing hormone has been detected in the human fetal hypothalamus at 10 to 12 weeks of gestation, and levels increase to term. Pituitary thyroid stimulating hormone (TSH) is detectable at 8 to 10 weeks of gestation, remains low until 16 to 18 weeks, and then increases to a plateau level by 28 weeks of gestation. Human fetal serum TSH is detectable by 10 weeks of gestation and increases to a mean level of ~15 U/mL between 20 and 30 weeks. Fetal serum total thyroxine (T4) levels are low before 16 to 18 weeks of gestation and, thereafter, progressively increase to levels of 10 to 12 g/dL at 34 to 36 weeks of gestation. By 40 weeks, the mean serum fetal TSH decreases to a mean value of ~10 U/mL, indicating maturation of the feedback mechanism. Normal children and adults show a diurnal variation of TSH secretion with lower values at 11 a.m. and peak values around 11 p.m. No diurnal rhythm is detectable in preterm and full-term infants younger than 4 weeks of age.3 Development of TSH circadian rhythm begins after the first month of life.

PHYSIOLOGIC CHANGES IN THYROID FUNCTION TESTS

The proper interpretation of serum TSH, T4, triiodothyronine (T3), reverse T3, T3 uptake, and thyroglobulin (Tg) concentrations in pediatric patients requires a knowledge of the age dependency of these hormones. Normal values relative to age are shown in Table 47-1. Importantly, preterm and small-for-gestational-age neonates have serum thyroid hormone values different from those of normal-term infants. In preterm infants, the cord serum T4 is decreased in proportion to gestational age and birth weight. Newborn screening serum T4 concentrations obtained from filter paper blood specimens in very-low-birth-weight infants (<1500 g) and in low-birth-weight infants (15002499 g) are presented in Table 47-2.4

TABLE 47-1. Age-Related Levels for Serum Parameters of Thyroid Function

TABLE 47-2. Newborn Screening Serum T4 Values by Birth Weight

HYPOTHYROIDISM IN THE NEONATE AND CHILD

ETIOLOGY The causes of pediatric hypothyroidism are listed in Table 47-3. Because most of the congenital and acquired causes are similar, they are discussed together. The terms cretinism and congenital hypothyroidism are used for hypothyroidism present before or at birth.

TABLE 47-3. Causes of Hypothyroidism

Endemic iodine deficiency may cause congenital hypothyroidism (cretinism) (Fig. 47-1). Thyroid dysgenesis is the most common cause of congenital hypothyroidism in nonendemic areas. The term includes aplasia, ectopy, and hypoplasia of the thyroid gland. It is twice as frequent among women as men. Interestingly, circulating antithyroid antibodies are found in a higher percentage of sera from mothers of nongoitrous cretins than in control mothers. Although these antibodies can cross the placenta, they probably do not destroy the fetal thyroid gland. Thus, the possible role, if any, of circulating or cellular antithyroid antibodies in the causation of congenital athyrosis is not clear. In mothers with Graves disease with or without thyrotoxicosis, TSH receptor-blocking antibodies may cross the placenta and cause transient neonatal hypothyroidism (see Chap. 42).5

FIGURE 47-1. A, Endemic cretin of Zaire. These children, born in areas of severe iodine deficiency, have marked mental retardation, short stature, muscle weakness, and motor incoordination. They may or may not have a goiter. Note the obesity, the protuberant abdomen, and the dry skin of the hands. This condition is due to prenatal iodine deficiency, probably combined with the maternal ingestion of cassava; the thiocyanate content of this plant crosses the placenta and further injures the developing fetal thyroid gland. Myxedematous endemic cretinism may be prevented by an injection of iodized oil to the pregnant mother. B, Other members of the community are clinically euthyroid, but, as in this youth, some of them have very large, multinodular goiters.

Ectopy of the thyroid gland is an important cause of hypothyroidism. Approximately 75% of patients with lingual thyroid glands do not have any normally located thyroid tissue (Fig. 47-2 and Fig. 47-3). Usually, the amount of ectopic thyroid tissue is insufficient to prevent hypothyroidism, although it can respond to TSH stimulation. Ectopic thyroid glands must be included in the differential diagnosis of midline lingual and sublingual masses. All children with ectopic thyroid glands should have a trial of full replacement thyroid hormone therapy before surgical excision is contemplated. Thyroid hormone therapy prevents further hypertrophy and hyperplasia. There is no evidence that lingual or sublingual thyroid glands are malignant in the pediatric patient. In infants, the combination of elevated serum TSH and normal or low-normal T4 levels suggests the presence of thyroid ectopy.6 Infants with aplasia or hypoplasia of the thyroid gland also have high serum TSH and low T4 levels.

FIGURE 47-2. Eight-year-old child with lingual thyroid gland (arrow).

FIGURE 47-3. Diagrammatic representation of development and migration of thyroid, indicating locations of thyroid-related anomalies and of ectopic thyroid tissue. The thyroid gland develops from a midline endodermal thickening in the pharyngeal floor. The descending thyroid is connected to the tongue by the thyroglossal duct; its lingual opening is termed the foramen cecum. By the seventh week, the thyroglossal duct usually has disappeared. A and B show possible locations of thyroglossal duct cysts, as well as a thyroglossal duct sinus (opening at the skin surface). The broken line illustrates the normal course of the descending thyroglossal duct and thyroid gland. C shows the most common locations of ectopic thyroid tissue. (The commonly present pyramidal lobe of the thyroid is formed from a persistent remnant of the lower end of the thyroglossal duct.) Although accessory thyroid tissue is often functional, it frequently fails to maintain the euthyroid state if the thyroid gland is removed. (From Moore KL. The developing human, 5th ed. Philadelphia: WB Saunders, 1993:202.)

Hypothyroidism may occur because of a hereditary enzymatic deficiency that prevents synthesis of T4 and T3. Hereditary enzymatic defects of thyroid hormone synthesis are the second most common cause of congenital hypothyroidism (Table 47-4).7 Patients with enzymatic defects may present with hypothyroidism with or without goiter or euthyroidism with goiter. Because the thyroid gland is not always palpable in infants, infants with congenital hypothyroidism caused by an enzymatic defect cannot always be differentiated by physical examination from those with athyrotic hypothyroidism. The diagnosis depends on specific tests of the various steps in thyroid hormone synthesis and release. Pendred syndrome is an autosomal recessive disorder characterized by a goiter resulting from a peroxidase defect, with euthyroidism or mild hypothyroidism and congenital sensorineural hearing loss. The Pendred syndrome gene is mapped to chromosome 7q31 and has been found to encode a putative sulfate transporter.8,9 Mutations have been found in the Pendred syndrome gene.8,9 Congenital hypothyroidism caused by mutations in the thyrotropin-receptor gene has been reported.10 The inherited disorders of thyroid hormone transport, T4-binding globulin deficiency must also be considered in the differential diagnosis of hypothyroidism (Table 47-5). However, in this condition, the patient is clinically euthyroid.

TABLE 47-4. Hereditary Defects in Thyroid Hormone Synthesis or Action Causing Hypothyroidism

TABLE 47-5. Inherited Defects in Thyroid Hormone Transport

Chronic lymphocytic thyroiditis is the most common cause of acquired hypothyroidism in pediatric practice. TRANSIENT DISORDERS OF THYROID FUNCTION Several transient disorders of thyroid function may occur in pediatric patients. These include transient hypothyroxinemia, transient hyperthyrotropinemia, and the euthyroid sick syndrome. CLINICAL FINDINGS IN HYPOTHYROIDISM Patients with hypothyroidism differ in appearance depending on the age at which the deficiency occurs and its duration and severity before therapy is instituted. In complete athyrosis, symptoms may be present at birth but, more commonly, they occur during the first 2 months of life. Features in the neonate that should suggest the possible presence of hypothyroidism include prolonged gestation with large size at birth, large posterior fontanelle, respiratory distress, hypothermia, peripheral cyanosis, hypoactivity, feeding difficulties, constipation, abdominal distention with vomiting, prolonged jaundice, and dry skin. The skin may be dry and cool, and circulatory mottling may be present. As the neonate ages, the facial features become coarse and puffy, and the tongue becomes broad and thick (Fig. 47-4). Linear growth may decline during the first month of life. Mental retardation occurs if the diagnosis is not made and therapy is not instituted. Diminished physical activity is a prominent finding. Any infant who must be awakened for feedings and who rarely cries must be suspected of having hypothyroidism.

FIGURE 47-4. A, Hypothyroid neonate. Note typical cretinoid facies with dull appearance and myxedema of the face, eyelids, lips, and tongue. B, The same patient after therapy and disappearance of myxedema.

Patients who develop hypothyroidism in early childhood have clinical manifestations different from neonates. Symptoms may appear gradually over several years. Delayed dentition and decelerated linear growth occurs. Mental sluggishness develops, but mental retardation usually does not occur if hypothyroidism develops after 2 years of age. If hypothyroidism has been present since infancy, infantile body proportions persist. Cold intolerance, dry skin, constipation, and muscle weakness are common findings. An infrequently recognized syndrome in hypothyroid children is the Kocher-Debr-Smelaigne syndrome (see Chap. 210), which consists of generalized muscular hypertrophy involving particularly the muscles of the extremities. The pathogenesis of the muscle hypertrophy is not known; it disappears with thyroid hormone therapy. In the older child and adolescent, growth retardation is a common finding. Hypothyroidism affects the circulatory and neuromuscular systems, as it does in the younger patient. Significant delay in sexual maturation is characteristic. Multicystic ovaries occur frequently in girls with hypothyroidism. A rare syndrome in children consists of severe hypothyroidism and isosexual precocity. The etiology is not entirely resolved. Studies suggest that there is some binding of the elevated circulating TSH to the follicle-stimulating hormone receptor, with follicle-stimulating hormone activity producing sexual precocity.11 The signs of

hypothyroidism and sexual precocity regress with therapy, and complete sexual maturation occurs at a normal age. LABORATORY FINDINGS In primary hypothyroidism, serum TSH levels are elevated, whereas in secondary and tertiary hypothyroidism, serum TSH levels are low or undetectable. Serum T4, T3, and T3 uptake values all are low. Plasma Tg levels are very low or undetectable in congenital hypothyroidism that is due to athyrosis. However, it is possible that low levels may be present in a patient with a defect in Tg synthesis or secretion. Thyroid scintiscan or ultrasonography may provide important information concerning the etiology of hypothyroidism in the neonate. Plasma somatomedin C concentrations are low in hypothyroidism, as is the serum alkaline phosphatase. The bone age is retarded. Epiphyseal dysgenesis or stippling of the epiphyses can be present (Fig. 47-5). Enlargement of the sella turcica occasionally can be seen in primary hypothyroidism.6 With thyroid hormone therapy, the bone age returns to normal, the epiphyseal dysgenesis regresses, and the radiographic appearance normalizes.

FIGURE 47-5. A, Radiograph of 13-year-old hypothyroid child with epiphyseal dysgenesis. There is fragmentation of the femoral heads and widening of the epiphyseal lines (arrows). B, At 6 months after commencement of thyroid hormone treatment, the femoral heads are now well formed and the epiphyseal lines appear relatively normal.

THERAPY The treatment goal is to restore euthyroidism as rapidly as is safe for the patient. In the neonate and very young infant, therapy is essential to protect the brain from damage. However, caution is necessary in the therapy for markedly hypothyroid neonates and infants because of the possible presence of a myxedematous myocardium. Vigorous therapy may cause cardiac failure or serious arrhythmias. The initial single daily oral dose of L -T4 in full-term neonates is 10 to 15 g/kg per day.12 In preterm neonates, the starting dosage is 10 g/kg per day; usually the dose can be increased to 15 g/kg per day in 4 to 6 weeks. In infants, children, and adolescents, the initial oral dose of L -T4is ~100 g/m2 per day.13 In children, suppression of serum TSH to normal levels is the best index of adequate therapy in primary hypothyroidism. However, neonates and infants with congenital hypothyroidism may have an abnormal threshold for the inhibition of TSH secretion; the feedback set-point seems to be increased, so that excessive serum levels of T4 are required to suppress the TSH.14 Serum TSH values may remain elevated for as long as 2 years, despite normal T4 and T3 levels produced by replacement therapy. Therefore, normal serum TSH levels must not be used as the only criterion of adequacy of therapy. Other indicators of adequate therapy include normal growth and development and normal skeletal maturation; however, children with long-standing acquired hypothyroidism appear to sustain a permanent height deficit at maturity, despite treatment. 15 Pseudotumor cerebri may occur after the initiation of L -T4 therapy. Excessive thyroid hormone therapy may delay neurologic development.16 NEONATAL SCREENING FOR HYPOTHYROIDISM The clinical diagnosis of congenital hypothyroidism during the neonatal period can be difficult and is frequently missed. Thus, neonatal screening programs for congenital hypothyroidism have been established. Most screening programs measure T4 and TSH in blood collected on filter paper before the neonate is discharged from the hospital. A supplemental TSH determination is performed on the lowest 3% to 5% of T4 results. This provides the most comprehensive screening, identifying infants with primary hypothyroidism and those at risk for secondary hypothyroidism and Tg deficiency. Positive results should be confirmed by serum determinations. Screening programs throughout the world have shown an incidence of 1:3500 to 1:4500 births.17 The preliminary results of follow-up psychometric studies of neonates detected by screening programs have been encouraging. Studies indicate that the earlier the diagnosis of congenital hypothyroidism is made and therapy started, the better the prognosis for intelligence.18 The prognosis also appears to be related to the etiology of the congenital hypothyroidism. HUMAN BREAST MILK AND NEONATAL SCREENING Both T4 and T3 are present in human milk, but levels of T4 and T3 in various reports differ widely, probably because sampling methods were different. Also, in some studies, extracted milk was analyzed, whereas in other studies, nonextracted milk was used. It has been suggested that there is sufficient thyroid hormone in human milk to mitigate the effect of congenital hypothyroidism. Nevertheless, breast-feeding does not impair the detection of congenital hypothyroidism in neonatal screening programs.19

HYPERTHYROIDISM IN THE CHILD

The various types of hyperthyroidism seen in pediatric patients are congenital (neonatal Graves disease and transitory Graves disease caused by transplacental passage of thyroid-stimulating immunoglobulins [TSIs]); and acquired (autoimmune thyroid disease, Graves disease, chronic lymphocytic thyroiditis, nodular toxic goiter [Plummer disease], thyroid carcinoma, exogenous iodide-induced hyperthyroidism [jod-basedow phenomenon], TSH-producing pituitary tumor, inappropriate secretion of TSH, and factitious hyperthyroidism). Graves disease is the most common type of hyperthyroidism. TSIs are responsible for causing Graves disease (see Chap. 42). CLINICAL FINDINGS Graves disease in the child and adolescent is much like the disease in adults. Its course is characterized by variable severity and a tendency for spontaneous remissions and exacerbations. The disease may appear at any age and occurs approximately five times as frequently among girls as among boys. The frequency increases with age after the first year of life, peaking during adolescence. Nervousness, irritability, emotional lability, excessive sweating, increased bowel movements, and increased appetite with or without weight loss are the most commonly noted symptoms. Occasionally, obesity may occur. Mild exophthalmos, lid retraction and stare, goiter, tachycardia, increased systolic blood pressure with wide pulse pressure, and tremors almost always are present on physical examination. Usually, the onset of symptoms and signs is gradual. Deterioration in school performance is not an uncommon symptom. Mild ophthalmopathy, including proptosis, stare, and lid retraction, is present. However, severe ophthalmopathy, such as might be seen in adults, is very rare in pediatric patients. Accelerated growth may occur, but this is not followed by a significant increase in maximal attained adult height. Thyroid storm is very rare in children and adolescents. LABORATORY FINDINGS

Serum T4, T3, free T4, free T3, and T3 uptake values are elevated in most patients; T3toxicosis and T4 toxicosis can occur in pediatric patients. TREATMENT There is controversy over the therapy of Graves disease in children and adolescents because no available therapy is ideal; each has distinct advantages and disadvantages.20,21 Antithyroid drugs, surgery, and radioactive iodine have been used. The selection of treatment depends on many factors, including the age and sex of the patient, severity and duration of the disease, size of the thyroid gland, presence of other complicating medical conditions, availability of experienced surgeons, ability of the patient and family to cooperate, and fear of ionizing irradiation and its potential genetic effects. Antithyroid drugs probably are the most common therapy for Graves disease in pediatric patients. The usual initial daily dosage of propylthiouracil is 300 mg and that of methimazole is 30 mg. The daily dosage of each drug should be divided into three equal doses and given approximately every 8 hours. Some physicians prescribe methimazole once daily. On the other hand, propylthiouracil may be less effective when given as a single daily dose. The recommended duration of therapy ranges from 1 to 3 years. Children and adolescents experience the same potential adverse effects of antithyroid drugs as do adults. The contraindications to medical therapy are drug toxicity and patient noncompliance. The main disadvantages are the relatively long period of therapy required and the difficulty of obtaining the continued cooperation of the patient. Surgery usually offers the advantage of rapid and permanent control of hyperthyroidism. Potential disadvantages include mortality; hypoparathyroidism; hypothyroidism; injury to the recurrent laryngeal nerves; and laryngotracheal edema, sometimes requiring tracheotomy. The indications for surgery include (a) toxicity to antithyroid drugs, (b) failure of cure after an adequate course of antithyroid drug therapy, (c) lack of patient compliance, and (d) failure of antithyroid drug therapy to shrink a large, conspicuous goiter. The use of 131I in the treatment of Graves disease in pediatric patients has been controversial.20,21 The advantages of such therapy are that it is easily administered and is effective. The disadvantages include a high incidence of hypothyroidism, the fact that complete control requires weeks to months, the unknown long-term effects on the induction of neoplasia, and the unknown genetic effects.

NEONATAL GRAVES DISEASE AND FETAL THYROTOXICOSIS

Neonatal Graves disease is a rare condition in neonates born to mothers with active or previously active Graves disease. Neonatal Graves disease occurs in ~1 in 25,000 pregnancies, and almost all cases are caused by transplacental passage of TSI from mother to fetus.22 In some neonates, the disease lasts for months or years, much longer than could be explained by persistence of passively transferred TSI in utero. The disease can be severe. There have been reports of neonatal Graves disease caused by a mutation in the thyrotopin-receptor gene.23 Fetal thyrotoxicosis is rarely diagnosed but is important because of the potential for significant morbidity and mortality.24,25 It is caused by transplacental passage of TSI. CLINICAL FINDINGS The clinical findings of neonatal Graves disease include prematurity, goiter, exophthalmos, tachycardia, hypertension, hyperirritability, congestive heart failure, jaundice, hepatosplenomegaly, and thrombocytopenia (Fig. 47-6). If the neonate has been exposed to antithyroid medication in utero, the illness may not manifest clinically for 5 to 10 days after birth. The mortality ranges between 15% and 20%, and sequelae include craniosynostosis and intellectual impairment.

FIGURE 47-6. Neonate with Graves disease. Note prominent eyes, stare, and wet hair, a result of increased sweating. Pacifier had to be used to decrease constant movement of hands and arms.

Clinical findings in fetal thyrotoxicosis include tachycardia (>160 beats per minute), goiter, congestive heart failure, and intrauterine growth retardation. The goiter can be detected by ultrasound examination. If there is evidence of fetal thyrotoxicosis, cordocentesis should be performed to monitor fetal thyroid function and the effects of maternal antithyroid therapy. However, the need for accurate assessment of fetal thyroid function must be weighed against the risk of cordocentesis. TREATMENT The diagnosis of neonatal Graves disease should be followed immediately by therapy with propylthiouracil, 5 to 10 mg/kg per day, or methimazole, 0.5 to 1.0 mg/kg per day, in divided. doses every 8 hours by gavage. This, then, should be followed by the administration of iodides, either orally (1 drop of super-saturated potassium iodide daily) or intravenously (as sodium iodide). Propranolol may be necessary if marked tachycardia and other catecholamine-mediated signs and symptoms are present. The dose of propranolol is 2 mg/kg per day in divided doses. Propranolol should never be used alone because of the associated complications and hazards of this drug in the neonate. In life-threatening situations, vigorous supportive therapiesincluding the use of glucocorticoids (prednisone, 2 mg/kg per day), intravenous fluids, sedation, oxygen, and a cooling blanketmay be necessary. Assuming that the Graves disease is self-limited, therapy should be gradually discontinued at 2 months. If there is a recurrence, antithyroid drugs should be restarted.

GOITERS
Goiter in the pediatric patient is not rare. The incidence in school-age children and adolescents is between 3.9% and 6%.26 The most common cause is chronic lymphocytic thyroiditis (autoimmune or Hashimoto thyroiditis).27 Chronic lymphocytic thyroiditis (see Chap. 46) is the single most frequent thyroid disorder seen in pediatric patients and is the most common cause of acquired hypothyroidism. Colloid or simple goiter is the second most common cause of euthyroid goiter. CHRONIC LYMPHOCYTIC THYROIDITIS CLINICAL FINDINGS Usually, the goiter is completely asymptomatic and the patient is clinically euthyroid. However, symptoms and signs of hypothyroidism or hyperthyroidism may be present. The goiter may be firm and smooth, lobulated, or nodular and is non-tender. If asymmetry of the thyroid gland is present, the right lobe is usually larger. LABORATORY FINDINGS The presence of antibodies (anti-Tg and antithyroid peroxidase) against thyroid gland components is consistent with the diagnosis of chronic lymphocytic thyroiditis. However, antithyroid antibodies may not be detectable even in histologically proven cases of this disease. Serum levels of T4 and T3 may be normal, low, or elevated. Serum TSH levels may be normal, but they are frequently elevated despite normal serum T4 concentrations. Thyroid scintiscan may show thyromegaly, with asymmetric or patchy areas of radioisotope uptake. Defective binding of inorganic iodide may occur in chronic lymphocytic thyroiditis, with substantial accumulation of

iodide; significant discharge follows the inhibition of iodide transport. One test of the integrity of the organic-binding mechanism is the iodide-perchlorate discharge test (see Chap. 34). A definite diagnosis of chronic lymphocytic thyroiditis may require a biopsy. However, the diagnosis may be made with reasonable certainty by the clinical picture and appropriate laboratory studies. TREATMENT There is no general agreement about the treatment of chronic lymphocytic thyroiditis in pediatric patients, except for those with hypothyroidism for whom thyroid hormone therapy is essential. Some authors believe that the condition ultimately leads to hypothyroidism if untreated; therefore, they place all patients on full replacement doses of L -T4 indefinitely to prevent either continued enlargement of the thyroid gland or hypothyroidism. Other physicians wait until there is biochemical evidence of hypothyroidism before instituting thyroid hormone therapy. Other autoimmune disorders occur in increased frequency in pediatric patients with chronic lymphocytic thyroiditis and include Graves disease, type 1 diabetes mellitus, Addison disease, idiopathic hypoparathyroidism, and pernicious anemia. Therefore, periodic evaluation for these disorders is recommended. COLLOID OR SIMPLE GOITER Colloid or simple goiter constitutes the second most common cause of euthyroid goiter in pediatric patients.28 Almost all patients have no symptoms suggestive of hypothyroidism or hyperthyroidism. The physical examination and laboratory studies frequently do not differentiate between a colloid goiter and chronic lymphocytic thyroiditis. The course of a colloid goiter is unpredictable. It has been postulated that simple diffuse enlargement of the thyroid gland in a young patient can lead to a large multinodular goiter in the older individual. This is the rationale for prescribing thyroid hormone therapy in a young patient with a large colloid goiter.

EVALUATION OF THE THYROID NODULE

Single or multiple nodules of the thyroid gland are an uncommon finding in pediatric patients. Patients with nontoxic goiters may present with either solitary or multiple discrete nodules. Multinodular goiter may appear in childhood, particularly in areas in which there is chronic dietary iodine deficiency. The differential diagnosis of solitary nodules of the thyroid gland include chronic lymphocytic thyroiditis, cyst, adenoma, carcinoma, hyperplasia, toxic nodule, and embryonic defects such as intrathyroidal thyroglossal duct cysts.29 The causes of multiple thyroid nodules include toxic goiter, toxic goiter after131I therapy, chronic lymphocytic thyroiditis, and adenomatous goiter. Nodules of the thyroid gland in children and adolescents usually are asymptomatic and are noted incidentally by the patient or parents or detected during a routine physical examination. The primary challenge is to rule out the presence of a malignancy. The appropriate evaluation and therapy for solitary or multiple nodules in pediatric patients, particularly for solitary thyroid nodules, are controversial because the diagnostic techniques used to distinguish a benign from a malignant lesion vary in reliability. CLINICAL FINDINGS In evaluation, the history is important and should include questions about previous irradiation of the head, neck, or upper thorax. External therapeutic irradiation of the head and neck is associated with the subsequent development of neoplastic changes in the thyroid gland.30 Also, extensive diagnostic irradiation of the head and neck of children may be a causative factor in inducing subsequent thyroid neoplasia.31 The family history should be explored for evidence of thyroid disease, pheochromocytoma, or hyperparathyroidism. Medullary thyroid carcinoma can occur sporadically or be transmitted as an autosomal dominant trait and may be associated with pheochromocytomas and hyperparathyroidism (multiple endocrine neoplasia type 2A) or can occur in association with pheochromocytoma and mucosal neuromas (multiple endocrine neoplasia type 2B) (see Chap. 188). Information on the rate of growth of the mass, the presence of local or systemic symptoms, and hoarseness or dysphagia should be obtained. The painless, rapid growth of a nodule suggests anaplastic thyroid cancer. Pain in the thyroid gland is an uncommon symptom with malignancy. Pain is associated more commonly with a rapidly enlarging nodule caused by hemorrhage into a cyst or a benign degenerating nodule. Careful palpation of a thyroid nodule may help define its nature. A soft, compressible, circumscribed nodule is probably not malignant; it is more likely to be a colloid or adenomatous cyst. Tenderness in a nodule suggests hemorrhage or an inflammatory process. Thyroid malignancy should be suspected if the nodule is hard or firm, if there is fixation to surrounding structures, or if there is vocal cord paralysis. Enlarged cervical lymph nodes, particularly low in the neck, increase the likelihood that a thyroid nodule is malignant. In most patients with malignant nodules, the surrounding thyroid tissue feels normal and the gland is of normal size. Medullary thyroid carcinoma is implicated in a patient who presents with multiple mucosal neuromas, which appear as whitish nodules on the tongue, lips, and the palpebral conjunctiva; such patients may also manifest a marfanoid habitus and various skeletal defects (multiple endocrine neoplasia type 2B). LABORATORY FINDINGS Serum T4, T3, and TSH determinations are helpful in determining the functional status of the thyroid gland. Importantly, the degree of hyperthyroidism resulting from a toxic nodule may not be sufficiently severe to allow diagnosis on clinical grounds alone. Serum antithyroid antibody studies may be helpful if one suspects chronic lymphocytic thyroiditis. Serum Tg lacks specificity as a marker of thyroid cancer when used preoperatively but has been useful in detecting postoperative recurrences of differentiated thyroid carcinomas. The measurement of serum calcitonin is essential if one suspects medullary thyroid carcinoma. Patients with this tumor may have elevated basal levels of serum calcitonin that increase further after the infusion of calcium or pentagastrin or both (see Chap. 40 and Chap. 188). Thyroid scintiscan is not particularly helpful in distinguishing between benign and malignant lesions (see Chap. 34 and Chap. 40). Nevertheless, it should be performed preoperatively to preclude a nodule being an ectopic thyroid gland and perhaps being the sole functioning thyroid tissue. Furthermore, the scan might identify other nodules that are not detected by physical examination as well as an autonomous nodule. Ultrasonography is useful in defining accurately the size of a nodule, the anatomic relationship to other structures, and whether a cold nodule is cystic, solid, or mixed. Roentgenography of the neck may be helpful because medullary thyroid carcinoma may be associated with calcification in the region of the thyroid gland or in the cervical lymph nodes. Pulmonary metastases of thyroid cancer may be detected on a chest radiograph. Fine-needle aspiration biopsy is used routinely in the evaluation of thyroid nodules in adults (see Chap. 39) and is being used more frequently in pediatric patients. TREATMENT The primary challenge in the management of a multinodular thyroid gland or a solitary nodule of the thyroid is to exclude malignancy. The risk of malignancy is lower for a thyroid with multiple nodules than for a thyroid with a solitary nodule. Some clinicians recommend excision of all thyroid nodules in pediatric patients.29 Others recommend thyroid suppression for a 4- to 6-month period in all patients who have no findings that suggest that their solitary nodule might be malignant.32 If the nodule grows during this period, or if the nodule does not decrease in size by 50% over a 1-year period, surgery is recommended. More widespread application of needle aspiration biopsy (see Chap. 39) in the future should direct management choices more specifically. A hyperfunctional or hot nodule may represent a follicular adenoma, focal hyperplasia, chronic lymphocytic thyroiditis, or, very rarely, a carcinoma. A hot nodule can be mimicked in patients with agenesis of the contralateral lobe of the thyroid. This particular anomaly must be differentiated from a hot nodule because excision is contraindicated. Repeat thyroid scanning after TSH stimulation excludes the functioning tissue as being the only thyroid tissue present. Solitary hot nodules may be autonomous and may be toxic or nontoxic. A toxic, hot nodule should be removed surgically after proper preparation with antithyroid drugs. Hot nodules that do not produce hyperthyroidism initially may occasionally increase their secretion of thyroid hormones gradually and insidiously until the patient becomes hyperthyroid; therefore, prolonged observation of the pediatric patient is necessary. The surgical excision of nontoxic, autonomous, solitary thyroid nodules is recommended to avoid lifelong observation. A hypofunctional or cold nodule presents the greatest clinical problem in differentiating malignant from benign lesions. Cold nodules may be found in chronic

lymphocytic thyroiditis or may represent cysts, follicular adenomas, abscesses, carcinomas, or embryonic defects such as intrathyroidal thyroglossal duct cysts.29 In one series of surgically removed solitary nodules of the thyroid, 5 of 27 patients (18.5%) with a cold nodule had carcinoma, whereas no patients with either a warm or hot nodule had a malignancy.29 In the few reports concerning the histopathology of solitary thyroid nodules in children and adolescents, the incidence of carcinomas has been from 16% to 24%.29

THYROID CANCER
Carcinoma of the thyroid gland in pediatric patients is rare.33 Its incidence has increased since 1951, when Winship was able to find 93 cases in the world's literature, to which he added 99 cases in 1970.34 Since 1970, the incidence may be decreasing. ETIOLOGY There is a significant association between thyroid carcinoma in children and adolescents and irradiation to the head and neck (see Chap. 39 and Chap. 40). Also, there is some evidence that an increased serum TSH level may be implicated in thyroid cancer. Several retrospective studies have indicated that the recurrence of cancer can be minimized and survival improved with thyroid hormone therapy. Differentiated thyroid cancers have TSH receptors, and in vitro these tumors respond metabolically to TSH stimulation. PATHOLOGY All of the histologic types of thyroid cancer that occur in adults are found in children and adolescents but, fortunately, with a slightly higher proportion of differentiated carcinomas. Papillary adenocarcinoma is the most common histologic type. Papillary cancers often contain follicular elements and may be designated mixed papillary and follicular carcinomas. Papillary carcinomas spread to the normal surrounding thyroid parenchyma and regional lymph nodes as they invade lymphatics. These carcinomas characteristically grow slowly. Follicular carcinomas have a marked tendency for vascular invasion and spread to bone and lung. Medullary thyroid cancer arises from the C cells of the thyroid and secretes calcitonin and other substances. Early spread is typically through the lymphatics. Undifferentiated carcinoma (anaplastic) is uncommon in children and causes early death because of extensive local and disseminated disease (see Chap. 40). CLINICAL FINDINGS Thyroid cancer is approximately twice as common in women as in men. In children it usually presents as one or more firm, painless nodules in the neck. These nodules may be a metastatic lesion to the cervical lymph nodes without any detectable thyroid nodule, the primary thyroid tumor being occult. Usually, these lymph nodes are movable, smooth, nontender, and discrete. Carcinoma localized only to the thyroid gland is unusual. DIAGNOSIS The diagnostic investigation for suspected thyroid cancer is identical with that discussed for the evaluation of thyroid nodules. Almost all pediatric patients with thyroid cancer are euthyroid. A definitive diagnosis of malignancy can be made only by histopathologic examination. TREATMENT The therapy for thyroid cancer in children and adolescents is primarily surgical. However, there is controversy because of differences in criteria for diagnosis, a variable and usually prolonged course of the disease, and the lack of controlled studies of surgical and adjuvant types of therapy. The main controversy relates to the extent of surgery; recommendations vary from none, if the cancer has extended beyond the capsule or to the regional lymph nodes, to total thyroidectomy and prophylactic node dissection in all cases.35,36 There is seldom an indication for radical surgery in children. Most forms of undifferentiated thyroid cancer are unresectable at the time of surgery; therefore, the goal is to obtain tissue for histologic examination so that radiation therapy can be initiated. A total-body 131I scan should be obtained after surgery to determine if there is any 131I-concentrating tissue anywhere in the body. Remnants of normal thyroid tissue are usually present after surgery is performed for thyroid cancer. Most differentiated thyroid cancers do not concentrate 131I in the presence of significant amounts of normal thyroid tissue. This necessitates ablation of the normal thyroid remnants with low doses of 131I before the administration of therapeutic doses of radioactive iodine. Thyroid hormone should be prescribed after the surgery and radioactive iodine therapy to suppress TSH release. Thyroid hormone therapy seems to be more effective in papillary and mixed papillary-follicular carcinomas than in follicular carcinomas. The serum Tg concentration is a useful marker for metastatic differentiated thyroid cancer. However, for reliability, residual normal thyroid tissue must have been ablated. Chemotherapy is indicated for a patient with thyroid cancer that is unresponsive to surgery, thyroid hormone therapy, or radiation therapy. The effectiveness of available chemotherapeutic agents has not, however, been evaluated extensively in pediatric patients. Patients with medullary thyroid cancer who have undergone thyroidectomy should be evaluated for residual tumor or metastases by basal and provocative tests using calcium or pentagastrin to stimulate the release of calcitonin. If residual medullary thyroid cancer is present, serum and urine calcitonin levels increase in response to stimulation. External-beam irradiation should be considered in locally advanced, incompletely resected, recurrent, or metastatic disease. FOLLOW-UP Children or adolescents with differentiated thyroid carcinoma should be observed for the rest of their lives. The course of these tumors is extremely slow and unpredictable.37 The tumor may remain quiescent for a long period and then produce metastases. Follow-up evaluation should include a clinical examination for evidence of lung metastases, measurement of serum thyroid hormone and TSH levels to ensure an adequate suppressive dose of exogenous thyroid hormone, determination of serum Tg in patients with differentiated thyroid cancer, and measurement of serum or urine calcitonin in patients with medullary thyroid cancer. CHAPTER REFERENCES
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. Burrow GN, Fisher DA, Larsen PR. Maternal and fetal thyroid function. N Engl J Med 1994; 331:1072. Fisher DA, Nelson JC, Carlton EI, Wilcox RB. Maturation of human hypothalamic-pituitary-thyroid function and control. Thyroid 2000; 10:229. Mantagos S, Koulouris A, Makai M, Vegenakis AG. Development of thyrotropin circadian rhythm in infancy. J Clin Endocrinol Metab 1992; 74:71. Frank FE, Faix JE, Hermos RJ, et al. Thyroid function in very low birth weight infants: effects on neonatal hypothyroidism screening. J Pediatr 1996; 128:548. Usala AL, Wexler I, Poech A, Gupta MK. Elimination kinetics of maternally derived thyrotropin receptor-blocking antibodies in a newborn with significant thyrotropin elevation. Am J Dis Child 1992; 146:1074. Hung W, Fitz CR, Lee EDH. Pituitary enlargement due to lingual thyroid gland and primary hypothyroidism. Pediatr Neurol 1990; 6:60. Medeiros-Neto G. Clinical and molecular advances in inherited disorders of the thyroid system. Thyroid Today 1996; 19:1. Everett LA, Glaser B, Beck JC, et al. Pendred syndrome is caused by mutations in a putative sulphate transporter gene (PDS). Nat Genet 1997; 17:411. Coyle B, Reardon W, Herbrick JA, et al. Molecular analysis of the PDS gene in Pendred syndrome. Hum Mol Genet 1998; 7:1105. Biebermann H, Schoeneberg T, Gudermann T. Congenital hypothyroidism caused by mutations in the thyrotropin-receptor gene. N Engl J Med 1997; 336:1390. Anasti JN, Flack MR, Froehlich J, et al. A potential novel mechanism for precocious puberty in juvenile hypothyroidism. J Clin Endocrinol Metab 1995; 80:276. Fisher DA. Management of congenital hypothyroidism. J Clin Endocrinol Metab 1991; 72:523. Hodges S, O'Malley BP, Northover BN, et al. Reappraisal of thyroxine treatment in primary hypothyroidism. Arch Dis Child 1990; 65:1129. Redmond GP, Soyka LF. Abnormal TSH secretory dynamics in congenital hypothyroidism. J Pediatr 1981; 98:83. Rivkees SA, Bode HH, Crawford JD. Long-term growth in juvenile acquired hypothyroidism: the failure to achieve normal adult stature. N Engl J Med 1988; 318:599. Weichsel ME. Thyroid hormone replacement therapy in the perinatal period: neurologic considerations. J Pediatr 1978; 92:1035. Fisher DA. Hypothyroidism. Pediatr Rev 1994; 15:227. Dubuis JM, Glorieux J, Richer F, et al. Outcome of severe congenital hypothyroidism: closing the developmental gap with early high dose levothyroxine treatment. J Clin Endocrinol Metab 1996; 81:222. Hahn HB, Spiekerman AM, Otto WR, Hossalla DE. Thyroid function tests in neonates fed human milk. Am J Dis Child 1983; 137:220. Zimmerman D, Lteif AN. Thyrotoxicosis in children. Endocrinol Metab Clin North Am 1998; 27:109. Rivkees SA, Sklar C, Freemark M. The management of Graves' disease in children, with special emphasis on radioiodine treatment. J Clin Endocrinol Metab 1998; 83:3767. Zimmerman D. Fetal and neonatal hyperthyroidism. Thyroid 1999; 9:727. de Roux N, Polak M, Couet J, et al. A neomutation of the thyroid-stimulating hormone receptor in a severe neonatal hyperthyroidism. J Clin Endocrinol Metab 1996; 81:2023.

24. Wilace C, Couch R, Ginsberg J. Fetal thyrotoxicosis: a case report and recommendations for prediction, diagnosis and treatment. Thyroid 1995; 5:125. 25. Bowman ML, Bergmann M, Smith JF. Intrapartum labetalol for the treatment of maternal and fetal thyrotoxicosis. Thyroid 1998; 8:795. 26. Rallison ML, Dobyns BM, Meikle AW, et al. Natural history of thyroid abnormalities: prevalence, incidence, and regression of thyroid disease in adolescents and young adults. Am J Med 1991; 91:363. 27. Moore DC. Natural course of subclinical hypothyroidism in childhood and adolescence. Arch Pediatr Adolesc Med 1996; 150:293. 28. Hung W, Chandra R, August GP, Altman RP. Clinical, laboratory and histologic observations in euthyroid children and adolescents with goiter. J Pediatr 1973: 82:10. 29. Hung W. Solitary thyroid nodules in 93 children and adolescents. Horm Res 1999; 52:15. 30. Hempelman LH, Hall WL, Phillips M, et al. Neoplasms in persons treated with x-ray in infancy: fourth survey in 20 years. J Natl Cancer Inst 1975; 55:519. 31. Pillay R, Graham-Pole J, Miraldi F, et al. Diagnostic irradiation as a possible etiologic agent in thyroid neoplasms of childhood. J Pediatr 1982; 101:566. 32. Fisher DA. Thyroid nodules in childhood and their management. J Pediatr 1976; 89:866. 33. Hung W. Well-differentiated thyroid carcinoma in children and adolescents: a review. Endocrinologist 1994; 4:117. 34. Winship T, Rosvoll RF. Thyroid carcinoma in children: final report on a 20-year study. Clin Proc Child Hosp DC 1970; 26:327. 35. Robie DK, Dinauer CW, Tuttle RM, et al. The impact of initial surgical management on outcome in young patients with differentiated thyroid cancer. J Pediatr Surg 1998; 33:1134. 36. Millman B, Pellitteri PK. Thyroid carcinoma in children and adolescents. Arch Otolaryngol Head Neck Surg 1995; 121:1261. 37. Farahati J, Bucsky P, Parlowsky T, et al. Characteristics of differentiated thyroid carcinoma in children and adolescents with respect to age, gender, and histology. Cancer 1997; 80:2156.

CHAPTER 48 MORPHOLOGY OF THE PARATHYROID GLANDS Principles and Practice of Endocrinology and Metabolism

CHAPTER 48 MORPHOLOGY OF THE PARATHYROID GLANDS

VIRGINIA A. LIVOLSI Embryology Anatomy Gross Anatomy Histology Ultrastructure Pathology Parathyroid Adenoma Primary Parathyroid Hyperplasia Chief Cell Hyperplasia Clear Cell (Water Clear Cell) Hyperplasia Parathyroid Carcinoma Intraoperative Assessment of Parathyroid Pathology Fat Stains Density Gradients Other Types of Parathyroid Pathology Secondary Hyperparathyroidism Familial Hypocalciuric Hypercalcemia Humoral Hypercalcemia of Malignancy Chapter References

EMBRYOLOGY
The parathyroid glands develop from the third and fourth pharyngeal (branchial) pouches (see Chap. 53, Fig. 53-4). The third pharyngeal pouch, containing the thymus and parathyroid tissue, migrates downward until the embryo is 18 mm in size, at which time the two inferior parathyroids attain their normal location at the lower poles of the thyroid.1 The fourth pharyngeal pouch is the source of the two upper parathyroid glands, which remain attached to the upper poles of the thyroid.1

ANATOMY
Anatomic studies have attempted to enumerate and localize the parathyroid glands in normal human cadavers.1,2 and 3 Although 80% of normal adults have four parathyroid glands, from 1 to 12 may be observed.1,2 and 3 The location of the four parathyroid glands also varies. The superior glands may be found embedded in the capsule of the thyroid or in the pharynx, behind the esophagus, or lateral to the larynx.1,3 The inferior glands, which normally rest near the lower pole of the thyroid, may be found near the bifurcation of the common carotid arteries, behind any part of the thyroid, paratracheally, in the thorax, or within the thymus. Despite this variable location, the parathyroid glands tend to be bilaterally symmetric.1

GROSS ANATOMY
In normal adults, the total weight (mean standard deviation) of all parathyroid tissue is 120 3.5 mg in men and 142 5.2 mg in women.4 (If both the parenchyma and fat are included in total gland weight, the weights are more variable.) Densitometry measurements show that parenchymal cell mass accounts for 74% of parathyroid gland weight.1,5 Variations in parathyroid gland weights are found between males and females and between blacks and whites.6 The size of the parathyroid glands ranges from 2 to 7 mm in length, 2 to 4 mm in width, and 0.5 to 2.0 mm in thickness. The glands are kidney bean shaped, soft, and brown to rust coloredalthough color varies with fat content, degree of vascular congestion, and the number of oxyphil cells present.1,4

HISTOLOGY
The parathyroid glands are each surrounded by a thin connective tissue capsule that extends into the parenchyma, where fibrous septa divide the gland into lobules.1,4 The parenchymal cells of the parathyroid glands are arranged in cords, nests, or sheets around capillaries.1,4 Small clusters or spheres of cells are interspersed with foci of adipose tissue. The adult parathyroid is composed of chief and oxyphil cells, a fibrous stroma, and variable amounts of fat (Fig. 48-1). The ratio of parenchymal to fat cells varies and probably is age dependent. Studies indicate that the formerly designated fat/cells ratio (50:50) may not reflect the normal state; in truth, normal parathyroid glands have only 17% fat.1,6,7,8 and 9

FIGURE 48-1. Normal parathyroid gland from a 40-year-old woman who had surgery for a thyroid nodule. Note 70:30 cell/fat ratio. Lymphoid tissue is at right. 20

Normal parathyroid glands contain chief cells, oxyphil cells, and clear cells. These distinguishable cells probably represent different morphologic expressions of the same basic parenchymal element. The chief cell is polyhedral and measures 6 to 8 nm in diameter.4 It has amphophilic or slightly eosinophilic cytoplasm, a sharply outlined nuclear membrane, and abundant nuclear chromatin. Intracellular fat is found in most normal chief cells. After puberty, oxyphils appear,4 and these increase in number with age (Fig. 48-2). The oxyphil measures 8 to 12 nm in diameter and has a well-demarcated cell membrane, abundant eosinophilic granular cytoplasm, and a pyknotic nucleus. Clear cells represent chief cells with an excessive amount of cytoplasmic glycogen.4 The presence of follicles is a normal finding, but these may mimic thyroid histology and cause concern during intraoperative frozen section analysis.10

FIGURE 48-2. Oxyphil nodule, which abuts fat on right, and smaller chief cells on left (arrow). Cells are larger than chief cells with more abundant cytoplasm; the latter is distinctly eosinophilic and granular. Nodules such as this are common in adult glands. 100.

ULTRASTRUCTURE
Electron microscopic studies of parathyroid cells have shown that the chief cells undergo morphologic change caused by the synthesis of parathyroid hormone. The Golgi apparatus increases in size; vesicles are prominent. In active parathyroid cells, little lipid is present.11 In the involuting and resting phase, the cytoplasm of the chief cells contains glycogen and lipid. In the euparathyroid normal adult, only ~20% of the chief cells appears to be producing hormone at any one time.4,11 Ultrastructurally, oxyphil cells contain numerous tightly packed mitochondria, some glycogen, and few free ribosomes. Evidence at the ultrastructural level for parathyroid hormone secretion is rarely observed in oxyphils.4,11

PATHOLOGY
Most patients with primary hyperparathyroidism harbor a parathyroid adenoma (see Chap. 58); usually, a single adenoma occurs in ~80% of patients.1,4,12,13 Occasionally, multiple adenomas are found either at surgery or after the single adenoma is removed. These cases most likely represent asynchronous parathyroid hyperplasia.

PARATHYROID ADENOMA
The parathyroid adenoma is more frequently located among the lower glands than among the upper ones. Grossly, a parathyroid adenoma is oval, reddish-brown, and smooth, circumscribed, or encapsulated (Fig. 48-3). It may show areas of hemorrhage and, if large, cystic degeneration. In small adenomas, a grossly visible rim of yellow-brown normal parathyroid tissue occasionally may be seen. Weights vary from 300 mg to several grams. The overall size ranges from smaller than 1 cm to larger than 3 cm.4

FIGURE 48-3. Transected parathyroid adenoma is shown. It weighed 1.1 g and had a brown, homogeneous surface. No rim is visible.

Histologically, parathyroid adenomas are composed of sheets of parathyroid chief cells interspersed with a delicate capillary network. If not very large, most adenomasconsonant with their gross appearancereveal a rim of normal or atrophic parathyroid tissue beyond the adenoma capsule1,4,12,13 (Fig. 48-4). The extraadenomatous, normal cells tend to be smaller and more uniform than those in the adenoma. Stromal and cytoplasmic fat is frequently abundant here but absent in the adenoma.13,14 The absence of a rim of normal tissue does not preclude the diagnosis of an adenoma, because large tumors may have overgrown the preexisting normal gland; alternatively, the rim may be lost when the tissue is sectioned.15 Zones of fibrosis with hemorrhage, cholesterol clefts, and calcification may be found in larger tumors.3,5

FIGURE 48-4. Portion of parathyroid adenoma (left) abutting normal rim. In this example, the adenoma cells have clear to slightly foamy cytoplasm. Cords of chief cells also are seen. On the right is fat and a group of normal chief cells. 100

Adenoma cells range in configuration from regular to severely atypical (Fig. 48-5 and Fig. 48-6). Focally, bizarre multinucleated cells with dark crinkled nuclei are occasionally seen; only rarely do these cells occur in large areas of the tumor. Nuclear DNA measurements in such multinucleate lesions show polyploid, not aneuploid,16 values, reflecting degenerative changes. Mitoses are only rarely found in a parathyroid adenoma. The belief was long held that mitotic figures are virtually never found in a benign parathyroid adenoma and that their presence should raise the suspicion of malignancy. This has been called into question, and parathyroid tumors with mitotic activity may, in fact, be benign.17,18 and 19 However, long-term follow-up in the reported series is brief, and parathyroid carcinoma has a long natural history, so the issue remains controversial. Not infrequently, in a predominantly chief cell adenoma, scattered small foci of oxyphil cells may be found.1,4 Ultrastructurally, adenoma cells have features of secretory activity, with prominent Golgi, rough endoplasmic reticulum, and secretory granules.11

FIGURE 48-5. Center of chief cell adenoma. Note vascularity of the lesion; in this area, cells are nearly uniform. 250.

FIGURE 48-6. Another adenoma is shown. Bizarre nuclear shapes and large size of nuclei characterize this area of the lesion. No mitoses were found. Bizarre cells are considered a degeneration phenomenon; the cells are polyploid, not aneuploid, by microspectrophotometric measurements. 250

When a single adenoma is present, the three nonadenomatous glands often show normal to increased fat content, and examination of biopsy tissue from the normal glands shows hypercellularity. Whether this observation reflects a real increase in cell number, difficulty in defining normal appearance, or sampling error is uncertain. Furthermore, some glands that contain a normal cell/fat ratio may have parenchyma that is unevenly dispersed throughout the gland, which makes a biopsy nonrepresentative. Rarely, oxyphil or oncocytic adenomas occur as functional lesions.4 Such tumors tend to be large and are associated clinically with mild degrees of hypercalcemia. Ultrastructurally, large numbers of mitochondria are found. In light of the embryology of the parathyroids, the finding of parathyroid tissue in nonclassic locations is not unexpected. Hence, adenomas are found in the superior mediastinum, behind the esophagus, or in an intrathyroidal location.2,3 and 4

PRIMARY PARATHYROID HYPERPLASIA

CHIEF CELL HYPERPLASIA Chief cell hyperplasia accounts for ~15% of patients with primary hyperparathyroidism.1,4 Significantly, 25% to 35% of individuals with chief cell hyperplasia have a history of familial hyperparathyroidism or multiple endocrine neoplasia1,4 (see Chap. 188). In chief cell hyperplasia, all four glands are enlarged.4 If the glands are of unequal size, it is usually the lower glands that are larger.5 Occasionally, the finding of one gland that is much larger than the others may erroneously suggest an adenoma. The combined weight of all four hyperplastic glands may range from 150 mg to >20 g, but usually it is 1 to 3 g.4 Lobulation, a red-brown color, and a homogeneous appearance are characteristic. Microscopically, two major patterns are found. In diffuse chief cell hyperplasia, solid masses of cells are present with minimal or absent stromal fat. Usually, these sheets comprise chief cells with rare oxyphil elements. Nodular (adenomatous or pseudoadenomatous) hyperplasia consists of circumscribed nodules of chief or oxyphil cells. Each nodule is devoid of fat, and little fat is found in the intervening stroma.1,4 Usually, no rim of normal parathyroid can be found; however, in nodular glands, relatively parvicellular zones may encompass totally cellular nodules, mimicking an adenoma.4 Bizarre nuclei are rare in primary hyperplasia; mitoses are found occasionally.19 Electron microscopy reveals active-appearing chief cells, with large Golgi complexes, secretory vacuoles, and prominent endoplasmic reticulum virtually identical with chief cell adenomas.11 Usually, intracellular fat is decreased or absent at the ultrastructural level and by fat stain.13,15,20 CLEAR CELL (WATER CLEAR CELL) HYPERPLASIA Clear cell (water clear cell) hyperplasia,21 a rare cause of primary hyperparathyroidism, occurs as a sporadic, nonfamilial disorder. Grossly, unlike other hyperplasias, the superior glands are larger than the lower. Total weight of such parathyroids always exceeds 1 g, and weights of 5 to 10 g may occur.1,4 The glands are irregular and exhibit cysts. No nodules are seen. The color is distinctly mahogany. Histologically, the glands are composed of diffuse sheets of water clear cells without the admixture of other cell types. No rim of normal tissue is present. The clear cell is large, 10 to 15 mm in diameter, with sharp cell borders that occasionally fuse. The nuclei are located basally. Nuclear pleomorphism is absent, although multinucleation may be observed. Acini, papillae, or follicles may form and may be interspersed with solid sheets of clear cells.1,4 Ultrastructurally, the cytoplasm is not empty but contains numerous membrane-bound vacuoles, secretion granules, Golgi apparatus, and endoplasmic reticulum.11

PARATHYROID CARCINOMA
Parathyroid carcinoma is responsible for <1% of cases of primary hyperparathyroidism.1,4,17,22,23 Often, the tumors are larger than adenomas, with an average weight of 12 g. Microscopically, parathyroid carcinoma is characterized by a trabecular arrangement of tumor cells (the latter divided by thick, fibrous bands), capsular and blood vessel invasion, and the presence of mitotic figures.17,22,23 The mitotic figures must be found in tumor cells, not in stromal or endothelial elements22 (Fig. 48-7). Cellular atypia, frequently found in benign parathyroid tumors, is uncommon in parathyroid carcinoma.17,22 The mere presence of capsular invasion cannot be equated with malignancy in a parathyroid tumor because large parathyroid adenomas may have undergone prior hemorrhage, with subsequent fibrosis and trapping of tumor cells within the capsule. Similarly, vascular invasion is difficult to determine, except if seen outside the vicinity of the neoplasm. Grossly, the surgeon often notes that the gland is adherent to or invasive into neighboring tissues (e.g., nerve, esophagus). Metastases at the time of presentation are unusual, but they may be found in local or regional lymph nodes. Indeed, in some patients, the diagnosis of parathyroid carcinoma may not be made until metastases appear. The loss of the retinoblastoma (Rb) tumor-suppressor gene formerly was considered to be a marker for parathyroid carcinoma.24 However, its loss in some adenomas and its retention in some carcinomas have diminished the diagnostic usefulness of testing for loss of Rb in parathyroid tumors.23

FIGURE 48-7. Parathyroid carcinoma from a 42-year-old patient with hyperparathyroidism. At surgery, the gland stuck to the esophagus and thyroid. The arrow points to mitotic figure in the parathyroid tumor. Such a finding is common in parathyroid malignancy. 250.

A group of solitary parathyroid tumors demonstrate some of the features of parathyroid carcinoma. These lesions appear to represent a pathologic spectrum of tumors that deserve careful clinicopathologic study and long-term follow-up. Such atypical lesions include (a) large parathyroid tumors with fibrous bands and uniform cytology but no mitoses (atypical adenoma), (b) parathyroid tumors of uniform pattern with easily identifiable mitoses but no other features of malignancy (atypical adenoma), and (c) circumscribed parathyroid adenomas within which are found one or more nodules of uniform histologic pattern with easily discernible mitoses (possibly, carcinoma in situ). In rare instances of hyperparathyroidism due to primary hyperplasia, nests of hyperplastic parathyroid cells can be found in the neck outside of hyperplastic glands. This pathology is referred to as parathyromatosis. In some individuals, these nests of parathyroid cells are discovered at the first neck exploration. During embryologic development, nests of pharyngeal tissue containing parathyroid cells are assumed to be scattered throughout the adipose tissue of the neck and mediastinum. Normally, these nests are inconspicuous, but in the process of diffuse hyperplasia of the parathyroids, all functioning tissue becomes hyperplastic and appears as separate fragments on histologic evaluation. More commonly, a similar lesion may occur after surgery due to spillage and implantation of hyperplastic parathyroid tissue in the neck25,26 and 27; this may occur in the setting of either primary or secondary hyperparathyroidism.

INTRAOPERATIVE ASSESSMENT OF PARATHYROID PATHOLOGY

FAT STAINS Because prominent, large fat droplets are found in the cytoplasm of normal chief cells and are absent in abnormal ones, the use of a sudan IV fat stain intraoperatively, at the time of frozen section, may conveniently and rapidly distinguish pathologic parathyroid tissue from normal tissue. In normal parathyroid glands, 80% of the cells are in a nonsecretory phase and contain cytoplasmic fat.10 The potential usefulness of the fat stain is illustrated as follows. A biopsy specimen of an enlarged parathyroid gland is sent for frozen section, and, by routine hematoxylin and eosin stain, is shown to be hypercellular with little or no stromal fat. Thus, it appears to represent an adenoma or hyperplastic gland. A biopsy specimen of the second parathyroid gland is normocellular or minimally hypercellular. The fat stain shows abundant cytoplasmic fat; hence, this second gland is normal. The enlarged gland, therefore, represents an adenoma. However, these clear-cut distinctions are not supported by some authors, who indicate that the use of the fat stain is helpful only in distinguishing normal from abnormal in ~80% of cases. Complicating this issue is the fact that normal glands may show decreased fat.8,13,20 Thus, fat stains should serve as an adjunctive technique and should be considered in the light of gross findings, gland weight, and size.1,13,14,20,28 DENSITY GRADIENTS Another rapid technique that may prove useful for the intraoperative assessment of parathyroid pathology is density gradient measurements of tissue. The gland is weighed, small pieces are removed from the center and from the rim, and their densities are determined in a 25% mannitol solution.1,5 Abnormal parathyroid tissue sinks because of decreased fat and high parenchymal mass. One study reported >90% accuracy in distinguishing abnormal from normal parathyroid glands with this technique performed intraoperatively.29

OTHER TYPES OF PARATHYROID PATHOLOGY

SECONDARY HYPERPARATHYROIDISM Grossly, in secondary hyperparathyroidism, all four glands are enlarged. Without clinical data, however, primary hyperplasia cannot be distinguished from secondary parathyroid hyperplasia microscopically.4 The anatomic variations of the parathyroid gland locations makes the surgical treatment of secondary hyperparathyroidism a challenge.30 FAMILIAL HYPOCALCIURIC HYPERCALCEMIA In familial hypocalciuric hypercalcemia (see Chap. 58) the parathyroid glands are described as normal or as showing minimal hyperplasia.31 HUMORAL HYPERCALCEMIA OF MALIGNANCY In patients with hypercalcemia caused by a nonparathyroid malignancy that is secreting a humoral substance, the parathyroids would be expected to demonstrate a normal histologic appearance32 or atrophy, because they would be suppressed by the systemic hypercalcemia. Although uncommon, hyperplastic parathyroid tissue in such patients has been described by some authors.5 CHAPTER REFERENCES
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. Grimelius L, Akerstrm G, Johansson H, Bergstrm R. Anatomy and histopathology of human parathyroid glands. Pathol Annu 1981; 16(Part II):1. Akerstrm G, Malmaeus J, Bergstrm S. Surgical anatomy of human parathyroid glands. Surgery 1984; 95:14. Wang CA. The anatomic basis of parathyroid surgery. Ann Surg 1976; 183:271. Castleman B, Roth SI. Tumors of the parathyroid glands, series II, fasc 14. Washington: Armed Forces Institute of Pathology, 1978. Akerstrm G, Grimelius L, Johansson H, et al. Estimation of parathyroid parenchymal cell mass by density gradients. Am J Pathol 1980; 99:685. Dufour DR, Wilkerson SY. Factors related to parathyroid weight in normal persons. Arch Pathol Lab Med 1983; 107:167. Dufour DR, Wilkerson SY. The normal parathyroid revisited: percent of stromal fat. Hum Pathol 1982; 13:717. Dekker A, Dunsford HA, Geyer SJ. The normal parathyroid gland at autopsy: the significance of stromal fat in adult patients. J Pathol 1979; 128:127. Saffos RO, Rhatigan RM, Urgulu S. The normal parathyroid and the borderline with early hyperplasia: a light microscopic study. Histopathology 1984; 8:407. Cinti S, Balercia G, Zingaretti MC, et al. The normal human parathyroid gland. Submicrosc Cytol 1983; 15:661. Johannessen JV. Parathyroid glands. In: Johannessen JV, ed. Electron microscopy in human medicine, vol 10. New York: McGraw-Hill, 1981:111. LiVolsi VA. Pathology of the parathyroid glands. In: Barnes L, ed. Surgical pathology of the head and neck. New York: Marcel Dekker, 1985:1487. Grimelius L, Johansson H. Pathology of parathyroid tumors. Semin Surg Oncol 1997; 13:142. Dekker A, Watson CG, Barnes EL. The pathologic assessment of primary hyperparathyroidism and its major impact on therapy: a prospective evaluation of 50 cases with oil-red-O stain. Ann Surg 1979; 190:671. Ghandur-Mnaymneh L, Kimura N. The parathyroid adenoma. Am J Pathol 1984; 115:70. Bowiby LS, DeBault LE, Abraham SR. Flow cytometric DNA analysis of parathyroid glands. Am J Pathol 1987; 128:338. Bondeson L, Sandelin K, Grimelius L. Histopathological variables and DNA cytometry in parathyroid carcinoma. Am J Surg Pathol 1993; 17:820. Chaitin BA, Goldman RL. Mitotic activity in benign parathyroid disease. (Letter). Am J Clin Pathol 1981; 76:363. Snover DC, Foucar K. Mitotic activity in benign parathyroid disease. Am J Clin Pathol 1981; 75:345. Dufour DR, Durkowski C. Sudan IV staining: its limitations in evaluating parathyroid functional status. Arch Pathol Lab Med 1982; 106:224. Hedback G, Oden A. Parathyroid water clear cell hyperplasia, an O-allele associated condition. Hum Genet 1994; 94:195. Schantz A, Castleman B. Parathyroid carcinoma: a study of 70 cases. Cancer 1973; 31:600. Favia G, Lumachi F, Polistina F, D'Amico DF. Parathyroid carcinoma: sixteen new cases and suggestions for correct management. World J Surg 1998; 22:1225. Cryns VL, Thor A, Hu HJ, et al. Loss of the retinoblastoma tumor suppressor gene in parathyroid carcinoma. N Engl J Med 1994; 330:757.

25. 26. 27. 28. 29. 30. 31. 32.

Reddick RL, Costa JC, Marx SJ. Parathyroid hyperplasia and parathyromatosis. Lancet 1977; 1:549. Fitko R, Roth SI, Hines JR, et al. Parathyromatosis in hyperparathyroidism. Hum Pathol 1990; 21:234. Sokol MS, Kavolius J, Schaaf M, et al. Recurrent hyperparathyroidism from benign neoplastic seeding: a review with recommendations for management. Surgery 1993; 113:456. Bondeson AG, Bondeson L, Ljungberg O, Tibblin S. Fat staining in parathyroid disease: diagnostic value and impact on surgical strategy. Clinicopathologic analysis of 191 cases. Hum Pathol 1985; 16:1255. Wang CA, Rieder SV. A density test for the intraoperative differentiation of parathyroid hyperplasia from neoplasia. Ann Surg 1978; 187:63. Butterworth PC, Nicholson MI. Surgical anatomy of the parathyroid glands in secondary hyperparathyroidism. J R Coll Surg Edinb 1998; 43:271. Thorgiersson U, Costa J, Marx SJ. The parathyroid glands in familial hypocalciuric hypercalcemia. Hum Pathol 1981; 12:229. Dufour DR, Marx SJ, Spiegel AM. Parathyroid gland morphology in nonparathyroid hormone-mediated hypercalcemia. Am J Surg Pathol 1985; 9:43.

CHAPTER 49 PHYSIOLOGY OF CALCIUM METABOLISM Principles and Practice of Endocrinology and Metabolism

CHAPTER 49 PHYSIOLOGY OF CALCIUM METABOLISM

EDWARD M. BROWN Role of Calcium Forms of Calcium in Blood Overall Calcium Balance Hormonal Control of Calcium Homeostasis Regulation of Parathyroid Hormone Secretion by Calcium and Other Factors Intracellular Mechanisms Regulating Parathyroid Hormone Release Effects of Parathyroid Hormone on Calcium-Regulating Tissues Role of the Skeleton in Calcium Homeostasis Regulation of Circulating Calcium Concentration Clinical Assessment of Calcium Homeostasis Direct Assays of Calcium-Regulating Hormones in Blood Assessment of the Function of Target Organs for Calcium-Regulating Hormones Chapter References

ROLE OF CALCIUM
Calcium ions (Ca2+) play numerous critical roles in both intra-cellular and extracellular physiology (Table 49-1). Intracellular Ca2+ is an important regulator of a variety of cellular functions, including processes as diverse as muscle contraction, hormonal secretion, glycogen metabolism, and cell division.1,2 and 3 Many of these functions are accomplished through the interaction of Ca2+ with intracellular binding proteins, such as calmodulin, which then activate enzymes and other intracellular effectors.2,3 The cytosolic free calcium concentration in resting cells is ~100 nmol/L. It is regulated by channels, pumps, and other transport mechanisms that control the movements of Ca2+ into and out of cells and between various intracellular compartments.1,2 and 3 Consonant with its role as a key intracellular second messenger, the cytosolic free Ca2+ concentration (Cai) can rise by as much as 100-fold (i.e., to 110 mol/L) during cellular activation. Such increases in Cai are the result of uptake of extracellular Ca2+ through Ca2+-permeable channels in the plasma membrane, release of Ca2+ from its intra-cellular stores, such as the endoplasmic reticulum, or both factors. Despite the importance of intracellular Ca2+ in cellular metabolism, this compartment comprises only 1% of total body calcium.4

TABLE 49-1. Some Properties of Calcium in Humans

In contrast to intracellular Ca2+ concentration, the extracellular free Ca2+ concentration (Ca2+o) is ~1 mmol/L. It is closely regulated by a complex homeostatic system involving the parathyroid hormone (PTH)secreting parathyroid glands and calcitonin-secreting thyroidal C cells as well as specialized Ca2+-transporting cells in the intestine, skeleton, and kidney.4,5,6 and 7 This system regulates the flow of Ca2+ into and out of the body as well as between various bodily compartments, particularly between the skeleton and extracellular fluid. The rigid control of Ca2+o ensures a steadysupply of Ca2+ for vital intracellular functions. Ca2+o has otherimportant roles, such as maintaining intercellular adhesion, promoting the integrity of the plasma membrane, and ensuring the clotting of blood, which further emphasizes the importance of maintaining near constancy of Ca2+o. The total amount of soluble extracellular calcium, however, like intracellular Ca2+, constitutes only a minute fraction of total bodily Ca2+ (~0.1%; see Table 49-1). Most of the Ca2+ within the body (>99%) resides as calcium phosphate salts within the skeleton, where it serves two important functions. First, it protects vital internal organs and acts as a rigid framework that facilitates locomotion and other bodily movements. Second, it provides a nearly inexhaustible reservoir of calcium and phosphate ions for times of need when intestinal absorption and renal conservation are insufficient to maintain adequate levels of these ions within the extracellular fluid. Thus, although Ca2+ within all bodily compartments plays essential roles, the fraction that is most closely regulated by the homeostatic system and, in turn, affects all other compartments is Ca2+o. An understanding of overall extracellular calcium homeostasis requires knowledge of the various forms of Ca2+ in the circulation and extracellular fluid as well as the movements of Ca2+ between the organism and the environment and among various bodily compartments (i.e., overall Ca2+ balance). Finally, Ca2+ homeostasis depends critically on a recognition system for Ca2+o and the effector systems by which changes in the circulating levels of calciotropic hormones (i.e., PTH; 1,25-dihydroxyvitamin D3[1,25(OH)2D3]; and calcitonin) restore Ca2+ homeostasis. The form of Vitamin D produced by the body is vitamin D3, which is then metabolized to 25-hydroxyvitamin D3 and 1,25-dihydroxyvitamin D3 as described later in this chapter. Thus, vitamin D3 and its respective metabolites are referred to throughout this chapter. The reader should recognize, however, that some dietary forms of vitamin D and medications contain vitamin D2. Vitamin D2 is metabolized to 25-hydroxyvitamin D2 and 1,25-dihydroxyvitamin D2, which in humans are equivalent in their potencies to the same metabolites of vitamin D3. As is described in more detail later, mineral ions themselves (i.e., extracellular calcium and phosphate ions) can also function in a calciotropic hormone-like manner, directly regulating the functions of many, if not all, of the tissues involved in maintaining Ca2+ homeostasis.7 FORMS OF CALCIUM IN BLOOD Although the extracellular ionized Ca2+ concentration within the interstitial fluid that bathes the various tissues of the body is perhaps most relevant to the Ca2+ homeostatic system, it is not readily measurable. Generally, the total or ionized serum Ca2+concentration is determined. Of the serum total Ca2+ concentration, some 47% is ionized or free Ca2+, an equivalent amount (~46%) is protein-bound, and the remainder (510%) is complexed to small anions, including phosphate, citrate, bicarbonate, and others.8 Ultrafilterable Ca2+ comprises both free and complexed Ca2+, but only the former is metabolically active (i.e., available for binding to extracellular Ca2+ binding sites or for uptake into cells). Albumin is the principal protein that binds Ca2+ in the circulation, and ~75% of bound serum calcium resides on albumin. The remainder is bound to various globulins. Several factors alter the amount of Ca2+ bound to proteins. Hypoalbuminemia and hyperalbuminemia reduce and increase, respectively, the amount of Ca2+ in the blood that is bound to albumin (and, therefore, the total Ca2+ concentration) without altering the level of ionized Ca2+ (see Chap. 60). A simple way to correct the total Ca2+ concentration for changes in serum albumin is to add or subtract 0.8 mg/dL of total Ca2+ for each 1 g/dL decrease or increase, respectively, in serum albumin concentration. Rarely, patients with multiple myeloma present with an increase in total Ca2+ concentration despite a normal serum ionized Ca2+ concentration because of a Ca2+-binding paraprotein in the circulation.9 The binding of Ca2+ to serum proteins is also pH dependent; acidosis reduces binding and alkalosis increases it, thereby producing reciprocal changes in serum ionized Ca2+ concentration. The carpal spasm encountered in some patients during hyperventilation is, in part, due to a concomitant reduction in serum ionized Ca2+ concentration caused by alkalosis. For each increase in pH of 0.1 unit, the serum ionized Ca2+ concentration decreases by ~0.1 mEq/L, and vice versa.

OVERALL CALCIUM BALANCE During skeletal and somatic growth in childhood, more Ca2+enters the body through the gastrointestinal (GI) tract than leaves it through the kidneys, GI tract, and perspiration (generally, loss of Ca2+ in sweat is insignificant). That is, the organism is in positive Ca2+ balance. Conversely, in the elderly (see Chap. 64), total body Ca2+ decreases, principally because of loss from the skeleton (i.e., Ca2+ balance is negative). During several decades of adult life, however, the Ca2+ homeostatic system precisely balances intake and output of Ca2+ to maintain a constant total body level (Fig. 49-1) while at the same time maintaining near constancy of Ca2+o. Of the 1000 mg of elemental Ca2+ ingested daily by this hypothetical individual, ~30% or 300 mg is actually absorbed by the intestine. Some 100 mg of Ca2+ is lost into the gut lumen by intestinal secretion daily, so that net absorption is 200 mg. Approximately 500 mg of Ca2+ enters and leaves the skeleton daily through the processes of bone formation and resorption, respectively. To maintain Ca2+ balance, 200 mg of Ca2+ are lost in the urine. The latter comprises only 2% of the daily filtered load of Ca2+ (10 g), illustrating the remarkable efficiency of the kidney in reabsorbing Ca2+ and its potential for modifying Ca2+ balance via changes in net Ca2+ reabsorption.

FIGURE 49-1. Overall Ca2+ balance in a normal individual. Of the 1 g of elemental Ca2+ ingested, net absorption is 200 mg (300 mg true absorption, 100 mg endogenous fecal secretion). Balance is achieved by renal excretion of 200 mg of Ca2+ because equivalent amounts of Ca2+ are laid down and removed from the skeleton on a daily basis in this individual. (ECF, extracellular fluid.) (Adapted from Brown EM, LeBoff MS. Pathophysiology of hyperparathyroidism. In: Rothmund M, Wells SA Jr, eds. Progress in surgery, vol 18. Parathyroid surgery. Basel: Karger, 1986:13.)

Because of the increasing recognition of osteoporosis as a major public health problem for individuals in later life, a great deal of interest has been shown in Ca2+ nutrition as a function of age, hormonal status, and other factors (see Chap. 64). The recommended daily allowance (RDA) in the United States had been 800 mg. However, although this level of intake may be sufficient to maintain some young adults in zero or positive balance, it may be insufficient for growing children and can lead to a negative balance in older people. In particular, a mean calcium intake of ~1000 mg prevents negative balance in perimeno-pausal, estrogen-replete women, whereas nearly 1500 mg is necessary to achieve zero balance in estrogen-deficient women of this age.10 Indeed, the RDAs for various groups of individuals have been modified to the following values: 800 mg until age 10; 1200 mg in adolescence; and 1000 mg thereafter, increasing to 1200 mg during pregnancy and lactation and to 1500 mg if at increased risk of osteoporosis or >65 years of age.11 Intake of calcium in these quantities should not be considered as a supplementation but rather as a nutritional requirement for skeletal health. Interestingly, the RDA for humans is nearly five-fold lower than that for other species of animals ranging in size from 1 to 800 kg.10 The form of calcium ingested has some impact on the efficiency with which it is absorbed. A high phosphate/calcium ratio may have deleterious skeletal consequences, because increases in serum phosphate can lower Ca2+o concentration and promote secondary hyperparathyroidism. Nevertheless, the relatively high content of phosphorus in dairy products does not appear to reduce substantially the absorption of Ca2+from this source. Calcium carbonate is a commonly used dietary Ca2+ supplement. A pH below ~5 is required for solubilization of calcium carbonate in the stomach.12 Therefore, in the presence of achlorhydriaa common condition in the aging populationabsorption of the carbonate salt of calcium can be reduced. In this situation, more soluble forms of Ca2+, such as calcium citrate, are preferable.13 Certain dietary sugars, such as lactose, enhance the absorption of ingested Ca2+ by a mechanism that is independent of vitamin D. This potential beneficial effect of lactose is often offset by lactose intolerance in older patients. The treatment of milk with lactase largely overcomes this problem in the lactose-intolerant patient. Conversely, phytate, which is present in some cereal grains, can bind Ca2+ in the intestinal lumen and impair absorption.

HORMONAL CONTROL OF CALCIUM HOMEOSTASIS

A complex homeostatic system has evolved that is designed to maintain near constancy of Ca2+o through calciotropic hormone, and direct, Ca2+o-induced alterations in the GI, renal, and/or skeletal handling of Ca2+ (Fig. 49-2).4,5,6 and 7 Usually, this is accomplished in such a way that extracellular (including skeletal) and cellular Ca2+ homeostasis and balance are maintained simultaneously. Conditions of gross Ca2+ excess, however, may induce positive Ca2+ balance (e.g., deposition of Ca2+ in bone and soft tissues) to maintain a normal or near-normal level of Ca2+o.Conversely, severe Ca2+ deficiency leads to mobilization of skeletal Ca2+ to preserve normocalcemia. Overall Ca2+o homeostasismay be understood in terms of the following general principles: (a) The first priority of the homeostatic system is to maintain a normal level of Ca2+o; (b) with moderate stresses on the system, intestinal and renal adaptations are usually sufficient to sus- tain Ca2+o homeostasis without changes in net bone mass; and (c) with severe hypocalcemic stresses, skeletal stores of Ca2+ are mobilized to maintain Ca2+o homeostasis, potentially compromising the structural integrity of the skeleton.

FIGURE 49-2. The hormonal control of calcium and phosphate homeostasis by parathyroid hormone (PTH) and vitamin D (solid lines and arrows). Ca2+ regulates PTH secretion in an inverse fashion. In turn, PTH affects renal handling of calcium and phosphate as well as renal synthesis of 1,25(OH)2D3. PTH and 1,25(OH)2D3 synergistically mobilize calcium and phosphate from bone, whereas the latter increases intestinal absorption of both ions. Also shown are direct actions of calcium and phosphate ions on tissues involved in maintaining mineral ion homeostasis (lines and arrows), illustrating the role of these ions as extracellular, first messengers. For example, elevated Ca2+o directly suppresses renal synthesis of 1,25(OH)2D3, tubular reabsorption of Ca2+ in the thick ascending limb, and the resorptive activities of osteoclasts. Thus, calcium ions modulate their own homeostasis not only through direct actions on the secretion of calciotropic hormones but also by exerting direct effects on target tissues that tend to lower Ca2+o. In effect, calcium and phosphate ions act as Ca2+o-regulating factors or hormones because they transmit information about the state of mineral ion metabolism from one part of the homeostatic system to other parts. For additional details, see text. (Reproduced from Brown EM, Pollak M, Hebert SC. Cloning and characterization of extracellular Ca2+-sensing receptors from parathyroid and kidney. Molecular physiology and pathophysiology of Ca2+-sensing. The Endocrinologist 1994; 4:419.)

The homeostatic system comprises two essential components. The first is several cell types that sense changes in Ca2+o and respond with appropriate alterations in

their output of Ca2+o-regulating hormones (PTH, calcitonin, and 1,25[OH]2D3). The parathyroid glands are key sensors of variations in Ca2+o, responding withalterations in PTH secretion that are inversely related to ambient levels of Ca2+o.7 Parathyroid cells recognize changes in Ca2+o through a cell-surface Ca2+o-sensing receptor14described in more detail laterthat is a member of the superfamily of G protein coupled receptors that includes the receptors for PTH and calcitonin. The C cells of the thyroid gland also respond to changes in Ca2+o with alterations in calcitonin secretionwith high Ca2+o stimulating calcitonin secretionthrough the same Ca2+o-sensing receptor. The physiologic relevance of changes in circulating levels of calcitonin in adult humans, however, remains uncertain (see Chap. 53). Finally, the production of 1,25(OH)2D3 is likewisedirectly regulated by Ca2+o, with hypocalcemia stimulating its synthesis in the renal proximal tubule and hypercalcemia inhibiting it. 15 The second key component of the homeostatic system is the effector cells that control the renal, intestinal, and skeletal handling of Ca2+. The translocation of calcium and phosphate ions into or out of the extracellular fluid by these tissues is regulated by PTH, calcitonin, and 1,25(OH)2D3 as well as by mineral ions themselves. The overall operation of the homeostatic system is as follows (see Fig. 49-2). In response to slight decrements in the Ca 2+o, forexample, a prompt increase occurs in the secretory rate for PTH. This hormone has several important effects on the kidney, which include inducing phosphaturia, enhancing distal tubular reabsorption of Ca2+, and increasing generation of 1,25(OH)2D3 from 25-hydroxyvitamin D3, or 25(OH)D3.16 The increased circulating levels of this potent metabolite of vitamin D3 directly stimulate absorption of calcium and phosphate by the intestine through independent transport systems.4,5 and 6,17 PTH and 1,25(OH)2D3 also synergistically enhance the net release of calcium and phosphate from bone. The increased movement of Ca2+ into the extracellular fluid from intestine and bone, coupled with PTH-induced renal retention of this ion, normalize circulating levels of Ca2+o, thereby inhibiting PTH release and closing the negative-feedback loop. In addition, 1,25(OH)2D3, formed in response to the action of PTH on the kidney, directly inhibits the synthesis and secretion of PTH,7 contributing to the negative feedback control of parathyroid function by the homeostatic system. Excess phosphate mobilized from bone and intestine is excreted in the urine via the phosphaturic action of PTH. Both calcium and phosphate ions can exert direct effects on various cells and tissues involved in mineral ion metabolism (see Fig. 49-2). For instance, calcium ions not only reduce PTH secretion but also directly inhibit proximal tubular synthesis of15 stimulate the function of osteoblasts,18 and inhibit 1,25(OH)2D3, the function of osteoclasts.19 Phosphate ions reduce 1a-hydroxylation of vitamin D, stimulate bone formation, inhibit bone resorption,4,5,6 and 7 and stimulate various aspects of parathyroid function,20 as described in more detail in the next section. These actions play important roles in mineral ion homeostasis by enabling both the hormone-secreting and effector elements of the homeostatic system to sense changes in the ambient concentrations of mineral ions and to respond in a physiologically appropriate manner. Indeed, by virtue of their direct actions on cells involved in mineral ion metabolism, calcium and phosphate ions can be viewed as serving a hormone-like role as extracellular first messengers.7 Although the cloning of the Ca2+o-sensing receptorthat mediates direct actions of Ca2+o on parathyroid and renalfunction has clarified substantially how cells sense Ca2+o,21 the mechanism underlying phosphate sensing remains obscure. REGULATION OF PARATHYROID HORMONE SECRETION BY CALCIUM AND OTHER FACTORS A steep, inverse sigmoidal relationship exists between PTH secretion and Ca2+o (Fig. 49-3)hence, the exquisite sensitivity ofthe parathyroid gland to Ca2+o that is critical to maintaining stable normocalcemia.7 This sigmoidal function can be described in terms of maximal and minimal secretory rates, midpoint or set-point, and slope at the midpoint (see Fig. 49-3). The maximal secretory rate provides a measure of the acute secretory reserve of the gland in response to a maximal hypocalcemic stress. Maximal secretory capacity increases with parathyroid cellular hyper-plasia, as in primary or secondary hyperparathyroidism.4,5,6 and 7,21 The secretion of PTH from the parathyroid cell apparently is incapable of being totally suppressed, even at very high levels of Ca2+o(see Fig. 49-3). Importantly, this implies that a sufficient mass of even normal parathyroid tissue could cause hypersecretion of PTH with resultant hypercalcemia. This has, in fact, been documented in rats by transplanting sufficient numbers of normal rat parathyroid glands to cause hypercalcemia in the recipient animals.22 This may, in part, contribute to the hypersecretion of PTH in hypercalcemic hyperparathyroidism The setpoint of normal human parathyroid cells in vitro is ~1.0 mmol/L of ionized Ca2+and is close to the ambient level of Ca2+o in humans (1.11.3mmol/L).7 This parameter is important in determining the level at which the Ca2+o is set by the homeostatic system (the setpointfor Ca2+o is slightly higher than that for the parathyroid cell per se, also being a function of the tissues that are targets for the actions of PTH). In primary hyperparathyroidism, the setpoint of pathologic parathyroid cells is increased to 1.2 to 1.4 mmol/L Ca2+, or sometimes higher, and the serum Ca 2+ concentration is correspondingly reset to an elevated level7,21 (see Chap. 58). The steep slope of the relationship between PTH release and Ca2+oensures a large change in secretory rate for a small change in Ca2+o and is important in maintaining the serum Ca2+ concentra- tion within a narrow range. In addition to its sensitivity to Ca2+o,the parathyroid cell alters its secretory rate within seconds to a perturbation in Ca2+o and may also be responsive to other variables, such as the rate and direction of the change.23,24

FIGURE 49-3. A, Relationship between parathyroid hormone (PTH) secretion and extracellular Ca2+ concentration in normal human parathyroid cells. Dispersed parathyroid cells were prepared and incubated with the indicated ionized Ca2+ concentrations. The PTH released was determined using a radioimmunoassay that recognizes the intact hormone and the COOH-terminal fragments of the molecule. B, The four parameters that can be used to describe the relationship between PTH secretion and the extracellular ionized Ca2+ concentration: maximal and minimal secretory rates at low and high Ca2+ concentrations, respectively; slope of the curve at its midpoint; and setpoint (the level of Ca2+ producing half of the maximal inhibition of PTH release). (A from Brown EM. Regulation of the synthesis, metabolism and actions of parathyroid hormones. In: Brenner BM, Stein H, eds. Contemporary issues in nephrology, vol 2. Divalent ion homeostasis. New York: Churchill-Livingstone, 1983.)

The technique of expression cloning in Xenopus laevis oocytes enabled isolation of a phosphoinositide-coupled, Ca2+o-sensing receptor from bovine parathyroid gland through which parathyroid, kidney, and other cells sense Ca2+o.14 This receptor has a large amino-terminal extracellular domain that is involved in binding Ca2+o, followed by seven membrane-spanning segments characteristic of the superfamily of G proteincoupled receptors, and a cytoplasmic carboxy terminus (Fig. 49-4). In addition to recognizing Ca2+o, the receptor responds to other polyvalent cations as well, including Mg2+, trivalent cations, and neomycin.14 The possibility exists, therefore, that the Ca 2+o-sensing receptor can function as a Mg2+o-sensing receptor, and that some of the directeffects of Ca2+o as well as the toxic actions of aminoglycosides on the kidney are mediated through it.21 The receptor plays a key role in Ca2+o sensing by the parathyroid and kidney because the human homolog of the receptor harbors inactivating or activating mutations in several human genetic diseases that manifest abnormal Ca2+o sensing.21 In familial hypocalciuric hypercalcemia (FHH),25 individuals with heterozygous inactivating mutations of the Ca2+o-sensing receptor have mild to moderate hypercalcemia with inappropriately normal (e.g., nonsuppressed) circulating levels of PTH and normal or even frankly low levels of urinary calcium excretion. In contrast, persons with homozygous FHH, which presents clinically as neonatal severe hyperparathyroidism (NSHPT), have severe hypercalcemia with marked hyperparathyroidism.25 (NSHPT can also be caused by compound heterozygous inactivating mutations of the Ca2+o-sensing receptor [i.e., patients harbor a different mutation in each allele of the receptor26] or by the presence of heterozygous inactivating mutations that exert a dominant negative effect on the wild-type receptor.27) Finally, some families with an autosomal dominant or, occasionally, a sporadic form of hypocalcemia accompanied by relative hypercalciuria (e.g., inappropriately high for the serum calcium concentration) harbor activating mutations in this receptor, a finding that provides further evidence of its importance in Ca2+o sensing in the parathyroid and kidney.28

FIGURE 49-4. Proposed structural model of the predicted bovine parathyroid Ca2+-sensing receptor protein. The large amino-terminal domain is located extracellularly and contains nine consensus N-glycosylation sites that are shown as branched chains. Amino acids of the deduced protein are shown at intervals of 50 in the amino-terminal domain. The amino-acid segments that comprise each of the seven predicted membrane-spanning helices are numbered (M1M7). Potential protein kinase C (PKC) phosphorylation sites are also shown. Each symbol (circle or triangle) represents an individual amino acid; those that are conserved with the metabotropic glutamate receptors are shown as solid symbols, of which acidic residues are indicated by solid triangles and all others by solid circles. Also indicated are multiple cysteine residues conserved with the metabotropic glutamate receptors (which may be involved in stabilizing the large extracellular domain) as well as clusters of acidic residues in the extracellular domain (open and solid triangles) that could be involved in binding of Ca2+ and other polyvalent cations. (SP, signal peptide; HS, hydrophobic segment.) (From Brown EM, Gamba G, Riccardi D, et al. Cloning and characterization of an extracellular Ca2+-sensing receptor from bovine parathyroid. Nature 1993; 366:575.)

In addition to being present in parathyroid, the Ca2+o-sensing receptor is also present along most of the renal tubule29; it is present at the highest levels in the cortical thick ascending limb of the nephron (a segment that is likely responsible for the abnormal renal handling of Ca2+ in FHH).25 Receptor transcripts are also found in thyroid, brain, and several other tissues, including intestine and bone, as described later.30 The presence of the receptor in kidney probably accounts for several of the long-recognized but poorly understood direct effects of Ca2+o on renal function (such as the reduced urinary concentrating ability encountered in some hypercalcemic patients).31 In the thyroid, the Ca2+o-sensing receptor resides almost exclusively in the C cells and is thought to mediate the stimulatory effect of elevated Ca2+o on calcitonin secretion. Ca2+o-sensing receptors in the brain might mediate as yet unknown actions of Ca2+o on neuronal function.30 Interestingly, among the G proteincoupled receptors, the Ca2+o-sensing receptor exhibits only limited homology with three subclasses of receptors sharing its overall topology, particularly the presence of a very large amino-terminal extracellular domain. These are the metabotropic glutamate receptors (mGluRs), targets for glutamate, the major excitatory neurotransmitter in the brain; the GABAB receptor, which is activated by g-aminobutyric acid (GABA), the major inhibitory neurotransmitter in the brain; and a family of putative pheromone receptors.30 All three classes of receptor presumably arose from a common ancestral gene. Further understanding of their structural similarities may clarify previously unknown relationships between Ca2+o homeostasis or other functions of Ca2+ in the nervous system and in systemic Ca2+o homeostasis. As with Ca2+, high concentrations of Mg2+o inhibit PTH release via the Ca2+o-sensing receptor, although, on a molar basis, Mg2+o is two to three times less potent than Ca2+o.32 Levels of Mg2+o in the blood sufficiently high to modulate PTH secretion are generally achieved only during pharmacologic interventions, such as the intravenous infusion of Mg2+ for the treatment of preeclampsia. Because of the reabsorption of more NaCl, water, and Ca2+than of Mg2+ in the proximal tubule, however, the level of Mg2+ in the thick ascending limb can rise to approximately 1.6-fold higher than that initially filtered at the glomerulus.33 Therefore, locally higher Mg2+ concentrations could interact with Ca2+o-sensing receptors in the thick ascending limb, a major site of hormonally regulated (i.e., PTH-regulated) Mg2+ and Ca2+ reabsorption. Abnormally low Mg2+o is also associated with a reduced rate of PTH release due to an unknown mechanism, which can cause hypocalcemia.34 This biphasic secretory response to Mg2+o differs from that to Ca2+o because near-maximal rates of PTH release can be maintained at very low levels of Ca2+o. Various other factors, including metabolites of vitamin D, a variety of catecholamines and other biogenic amines, prosta-glandins and peptide hormones, phosphate, and monovalent cations such as potassium and lithium also modify PTH release.7,35 Of these, the most physiologically relevant are the active vitamin D metabolite 1,25(OH)2D3 and changes in the serum phosphate concentration. Studies have documented that 1,25(OH)2D3 plays an important role in the longer-term (over days or longer) regulation of parathyroid function, tonically inhibiting PTH secretion,36,37 PTH gene expression,38,39 and probably parathyroid cellular proliferation.40 Therefore, the actions of PTH on its target tissues produce negative feedback regulation of parathyroid cellular function not only by increasing the serum Ca2+ concentration but also by stimulating the formation of 1,25(OH)2D3, which exerts direct negative feedback actions on the parathyroid. Elevations and reductions in serum phosphate concentrations increase and decrease, respectively, PTH gene expression, PTH secretion (perhaps indirectly through changes in the expression of its gene), and parathyroid cellular proliferation.20 The use of lithium in the treatment of manic-depressive illness may induce hypercalcemia and hypersecretion of PTH.41 These findings can mimic the biochemical abnormalities of primary hyperparathyroidism or FHH (e.g., lithium treatment can also be accompanied by hypocalciuria and a tendency toward hypermagnesemia). Indeed, of patients being treated long term with lithium, some 10% to 15% develop overt hyperparathyroidism, although whether this results from a direct action of lithium to induce the disorder or from unmasking of preexistent mild primary hyperparathyroidism is not entirely clear. In vitro, lithium produces a rightward shift in the setpoint of bovine parathyroid cells for Ca2+o42 (e.g., a change similar to that seen in parathyroid adenomas), possibly due to interference with the process of Ca2+o sensing. INTRACELLULAR MECHANISMS REGULATING PARATHYROID HORMONE RELEASE After its biosynthesis as preproPTH, its subsequent metabolism to proPTH, and then its metabolism to PTH, the hormone is stored in secretory granules before its release into the circulation43 (see Chap. 51). The regulation of hormonal secretion involves the transduction of information about changes in Ca2+o by the Ca2+o-sensing receptor into alterations in the levels of one or more intracellular mediators that ultimately modify the secretory process.7,14,43 The definitive identification of the relevant mediators, however, remains elusive. Although cyclic adenosine monophosphate (cAMP) mediates the actions of potent secretagogues, such as b-adrenergic catecholamines, on PTH release, it cannot account quantitatively for the effects of Ca2+o on hormonal secretion.7 Increases in Ca2+o produce an initial transient rise in Cai,44 most likely because of Ca2+-receptor mediated hydrolysis of polyphosphoinositides (PIs). The resultant production of inositol 1,4,5-trisphosphate14,45 stimulates the release of Ca2+ from intracellular stores. A more sustained increase in Cai then ensues,44 owing to uptake of extracellular Ca2+ through plasma membrane channels for which the properties and mode of regulation remain to be fully defined. Activation of phospholipase C and the resultant hydrolysis of polyphosphoinositides in response to elevations in Ca2+o likelyplays a key role in the concomitant inhibition of PTH secretion and of parathyroid cellular proliferation. However, the identification of the most relevant downstream mediator(s) of these inhibitory effects on parathyroid function remain elusive. Patients with homozygous FHH have dramatic hypersecretion of PTH due to severe primary hyperparathyroidism with a substantial shift to the right in setpoint as well as marked glandular enlargement and parathyroid cellular hyperplasia.46 This experiment in nature shows, therefore, that the Ca2+o-sensing receptor contributes to the tonic inhibition of parathyroid cellular proliferation under normal circumstances. EFFECTS OF PARATHYROID HORMONE ON CALCIUM-REGULATING TISSUES CONTROL OF GASTROINTESTINAL CALCIUM ABSORPTION The net amount of Ca2+ absorbed from the GI tract is the difference between the total mass of Ca 2+ moving from lumen to plasma (absorption) and plasma to lumen.47 The latter, termed endogenous fecal calcium, results from the secretion by the intestinal mucosa of ~100 mg per day of Ca2+; it varies little in different states of Ca 2+ balance. Ca2+ absorption results from both passive diffusion across the intestinal mucosa and active transport. The passive diffusion of Ca2+ is a concentration-dependent, nonsaturable process that accounts for absorption of 10% to 15% of the Ca2+ in the normal diet (i.e., 100150 mg per day of ingested Ca2+). The active component of Ca2+ absorption is a saturable, carrier-mediated mechanism46a,46b regulated by 1,25(OH)2D3. The highest density of sites of active Ca2+ absorption is in the duodenum.47,48 Clearly, however, vitamin D-responsive Ca2+ absorption occurs in more distal portions of the bowel as well, including the small intestine (ileum more than jejunum) and portions of the colon. Because these segments of the GI tract are anatomically much longer than the duodenum, they may

contribute significantly to overall Ca2+ absorption. After 1,25(OH)2D3 is administered to vitamin Ddeficient animals, an increase occurs in the GI absorption of Ca2+ over several hours, which generally is paralleled by the induction in intestinal mucosal cells of several vitamin Ddependent proteins, including a Ca2+-binding protein (calbindin D9K), alkaline phosphatase, and Ca2+-Mg2+-adenosine triphosphatase.47,48 These proteins may be involved in the mechanism by which vitamin D enhances Ca2+ absorption. Probably, Ca2+ ions in the intestinal lumen move down their electrochemical gradient across the brush border of the epithelium via an uptake channel46a,46b and are pumped out of the basolateral aspect of the cell on their way to the extracellular fluid. The metabolite 1,25(OH)2D3 appears to stimulate both the influx and the egress of Ca2+ from the intestinal epithelial cells.49,50 Most phosphate absorption from the intestine occurs in the small bowel through a vitamin Dresponsive mechanism distinct from the Ca2+mechanism.4,5 and 6,47 Even in vitamin D deficiency, however, approximately half of dietary phosphorus is absorbed. The less stringent regulation of phosphate absorption in the gut is consonant with the ubiquity of this ion in the diet and the looser control of the serum phosphate concentration. An important aspect of the Ca2+o homeostatic system is itscapacity to adapt the efficiency of Ca2+ absorption to dietary intake. When patients are placed on a low Ca2+ diet, serum levels of 1,25(OH)2D3 increase by 50% within 24 to 48 hours, whereas when they are exposed to a high Ca2+ diet this metabo-lite decreases by 50% over this period.51 In experimental animals, the increase in 1,25(OH)2D3 levels on a low Ca2+ diet is largely abolished by prior parathyroidectomy,52 a finding which suggests that dietary Ca2+-induced changes in 1,25(OH)2D3 concentration arise from changes in serum Ca2+ concentration that alter vitamin D metabolism through alterations in the rate of PTH secretion. Nevertheless, lowering and raising the level of Ca2+o also directly stimulate or inhibit, respectively, the 1-hydroxylation of 25(OH)D3.15 The latter effects of Ca2+o could potentially be mediated by the Ca2+o-sensing receptor presentin the renal proximal tubular cells (see later).29 The Ca2+o-sensing receptor is also expressed along the entire GI tract, but whether it directly regulates the absorption of mineral ions is not known.30 Because of the Ca2+o- and PTH-evoked, 1,25(OH)2D3 mediated adaptation in the efficiency of the absorption of Ca2+by the intestine, the absorption of this ion varies less than its content in the diet. Absorption of supplemental Ca2+ may be predominantly through the vitamin Dindependent route.53 Phosphate intake also modulates the production of 1,25(OH)2D3, with hypophosphatemia stimulating and hyperphosphatemia inhibiting its renal synthesis.4,5 and 6 CONTROL OF RENAL CALCIUM EXCRETION PTH- and direct Ca2+o-induced changes in renal Ca2+ handling play an important role in overall fine-tuning of Ca2+balance54,55,55b (see Chap. 206). Conversely, vitamin D and its metabolites have only minor direct effects. Of the approximately 10 g of Ca2+ filtered daily by the kidney, 65% is reabsorbed in the proximal tubule.55 Ca2+ reabsorption in this site is closely linked to bulk solute and water transport. PTH has little effect on Ca2+ transport in this nephron segment. In fact, in some studies, PTH inhibits the proximal tubular reabsorp tion of Ca2+, perhaps because the hormone reduces sodium reabsorption.55 Of the more distal portions of the renal tubule, the descending and ascending thin limbs of Henle loop transport little Ca2+.55 Conversely, the thick ascending limb (TAL) of the loop and the distal convoluted tubule (DCT) reabsorb ~20% and 10%, respectively, of the filtered load of Ca2+. In experimental animals, PTH rapidly increases the reabsorption of Ca2+ in both segments of the nephron by increasing transport of the ion from lumen to plasma.5,6,55 It exerts this effect, like its other actions in the kidney and in other tissues, by interacting with a G protein coupled cell-surface receptor linked to activation of both adenylate cyclase and phospholipase C.56 Several lines of evidence support a primary role for cAMP in mediating PTH-induced changes in renal Ca2+ handling. In microdissected portions of the mammalian renal tubule, PTH-sensitive adenylate cyclase is present in the proximal tubule, cortical portion of the TAL, and portions of the distal tubule (e.g., DCT).57 The location of the enzyme in the proximal tubule is thought to be linked to the well-known PTH-induced phosphaturia. The PTH-activated adenylate cyclase activity in the latter two locations correlates well with the sites of action of the hormone in promoting Ca2+reabsorption. Moreover, the exposure of renal tubules to analogs of cAMP mimics the effects of PTH on Ca2+ transport, further supporting the mediatory role of cAMP. In the cortical thick ascending limb (CTAL), PTH is thought to increase the overall activity of the Na/K/2Cl cotransporter that drives transcellular reabsorption of NaCl in this nephron segment (Fig. 49-5).31,33,55 This increase in transcellular transport is associated with an increase in the lumen-positive, transepithelial potential difference that is responsible for driving ~50% of the reabsorption of NaCl and most of the reabsorption of Ca2+ and Mg2+ in the CTAL. In contrast, raising Ca2+o by activating the Ca2+osensing receptor present in the same epithelial cells of the CTAL reduces overall cotransporter activity, probably by inhibiting both the cotransporter itself as well as the apical potassium channel that recycles potassium ions back into the tubular lumen.31 The transepithelial potential gradient is consequently reduced (see Fig. 49-5), leading to diminished paracellular reabsorption of both Ca2+ and Mg2+. In effect, Ca2+o, acting in a fashion analogous to that of loop diuretics (e.g., furosemide), regulates its own reabsorption (and that of Mg2+) by a direct renal action that antagonizes the action of PTH.31

FIGURE 49-5. Diagram showing how the Ca2+o-sensing receptor (CaR) may regulate intracellular second messengers and, in turn, transport of NaCl, K+, Mg2+, and Ca2+ in the cortical thick ascending limb. Hormones stimulating cyclic adenosine monophosphate (cAMP) accumulation, such as parathyroid hormone, increase the reabsorption of Ca2+and Mg2+ through the paracellular pathway by elevating the lumen-positive transepithelial potential, Vte, via stimulation of the activity of the Na/K/2Cl cotransporter and an apical K+ channel. The CaR, also present on the basolateral membrane, increases arachidonic acid (AA) formation by activating phospholipase A2 (PLA2) (2). AA is metabolized via the P450 pathway to produce an active metabolite, probably 20-hydroxyeicosatetraenoic acid, which inhibits the apical K+ channel (4) and, perhaps, the Na/K/2Cl cotransporter (3). Both actions reduce overall cotransporter activity, thereby diminishing Vte and, in turn, paracellular transport of Ca2+ and Mg2+. The CaR probably also inhibits adenylate cyclase (1), thereby decreasing hormoneand cAMP-stimulated divalent cation transport. (Reproduced with permission from Brown EM, Hebert SC. Calcium-receptor regulated parathyroid and renal function. Bone 1997; 20:303.)

Although the detailed cellular mechanism by which PTH regulates Ca2+ transport in the DCT remains incompletely understood, it likely involves a PTH-stimulated increase in the apical uptake of Ca2+ through a channel.46a,55,55a The ensuing transcellular transport of Ca2+ is facilitated by a vitamin Ddependent Ca2+-binding protein, calbindin D28K, which is expressed in this nephron segment and is distinct from that participating in vitamin Dmediated GI absorption of Ca2+.58 Another small amount of Ca2+ (~5% of the filtered load) is reabsorbed in the collecting ducts, but transport at this site is not PTH regulated. Although the Ca2+o -sensing receptor is present in the DCT,29 its role, if any, in regulating tubular reabsorption of Ca2+ is not known. The net action of PTH on renal Ca2+ handling is to reduce the amount of Ca2+ excreted at any given level of serum Ca2+.54 This has been demonstrated by examining renal Ca2+ excretion as a function of serum Ca2+ concentration in subjects with underactive, normal, or overactive parathyroid function (Fig. 49-6).54 In patients with primary hyperparathyroidism, although the total amount of Ca2+ excreted over 24 hours in the urine may be elevated, less Ca2+ is excreted than in a normal person whose serum Ca2+ concentration is equally elevated. Conversely, hypoparathyroid patients have a renal Ca2+ leak, excreting more calcium than normal at a given serum Ca2+ concentration. Therefore, during therapy with vitamin D, the total serum Ca2+concentration of patients with hypoparathyroidism should be maintained in the range of 8 to 9 mg/dL to avoid hypercalciuria (see Chap. 60). The data in Figure 49-6 also illustrate the steep positive relationship between serum and urine Ca2+, which is likely mediated by the Ca2+o-sensing receptor.31 This relationship is reset as a function of the prevailing state of parathyroid function, shifting rightward and leftward with chronic increases and decreases, respectively, in the circulating levels of PTH.54

FIGURE 49-6. Urinary excretion of Ca2+ expressed as a function of serum Ca2+ in normal individuals (area enclosed by the dotted lines, which shows mean 2 standard deviations) as well as in hypoparathyroid subjects (open and solid triangles; solid triangles represent basal values) and hyperparathyroid subjects (solid circles). The shaded area is the normal physiologic situation. (GF, glomerular filtrate.) (From Nordin BEC, Peacock M. Role of the kidney in regulation of plasma calcium. Lancet 1969; 2:1280.)

Along with its effects on renal Ca2+ handling in the distal nephron (i.e., CTAL and DCT), PTH also inhibits phosphate reabsorption in both proximal and distal sites and enhances the synthesis of 1,25(OH)2D3 in the proximal tubule.5,6 The first of these effects, like the actions of the hormone on Ca2+ reabsorption, appears to be mediated by cAMP. PTH also activates PI turnover,56 and evidence has implicated this pathway in the PTH mediated stimulation of the synthesis of 1,25(OH)2D359. In addition to regulating renal handling of Ca2+ and Mg2+, the Ca2+o-sensing receptor probably mediates the known action of hypercalcemia to reduce urinary concentrating ability.31 It is thought to do so by two actions: First, by inhibiting NaCl reabsorption in the TAL, it reduces the medullary hypertonicity needed for passive, vasopressin-stimulated reabsorption of water in the collecting ducts. Second, in the inner medullary collecting duct, raising Ca2+o directly reduces vasopressinstimulated water flow, probably by an action mediated by the Ca2+o-sensing receptor on the apical membrane of these cells60. The resultant excretion of Ca2+ in a more dilute urine may reduce the risk of renal stone formation when a Ca2+ load requires disposal during times of antidiuresis.31 In this way, the Ca2+o-sensing receptor may serve to coordinate the homeostatic mechanisms governing Ca2+o and water metabolism. ROLE OF THE SKELETON IN CALCIUM HOMEOSTASIS During the constant remodeling of the skeleton, osteoclastic bone breakdown is closely coupled to osteoblastic bone formation.4,5 and 6,61 The formation of osteoblasts is coupled to the generation of osteoclasts from their precursors. Osteoclastic resorption of bone, in turn, is tightly linked to the ensuing replacement of the resorbed bone by osteoblasts. This constant turnover and renewal of bone (see Chap. 50) plays an important role in maintaining the structural integrity of this tissue. The precision of the coupling between resorption and formation is dramatically illustrated in Paget disease of bone, in which up to 10-fold increases in the rate of skeletal turnover are often unassociated with any alterations in the serum calcium concentration or overall calcium balance (see Chap. 65). Key mechanisms that link the differentiation and function of osteoblasts and osteoclasts are described briefly below and are discussed in more detail in Chap. 50. PTH and other agents stimulating bone resorption (e.g., interleukin-11, prostaglandin E2, and 1,25[OH]2D3) activate osteoclast maturation and function only indirectly by enhancing the expression of osteoclast differentiating factor (ODF) (also sometimes called TRANCE, RANKL,or osteoprotegerin-ligand).62 ODF is expressed on the cell surface of osteoblasts and stromal cells. It activates osteoclast development and increases the activity of mature osteoclasts by interacting with the ODF receptor (a protein called RANK or a closely related protein) on preosteoclasts, which then differentiate into mature osteoclasts if macrophage colony-stimulating factor (M-CSF) is also present.62 Osteoclastic bone resorption, in turn, is coupled to subsequent osteoblastic formation of bone, at least in part through the release of skeletal growth factors, such as transforming growth factor-b and insulin-like growth factor-I, which stimulate osteoblastic differentiation and activity.63 The skeleton also serves as a nearly inexhaustible reservoir for calcium and phosphate ions.5,6 and 7 Because the skeletal content of Ca2+ is 1000-fold higher than that of the extracellular fluid, this function can be subserved by the net movement of relatively small amounts of Ca2+ into or out of bone. After administration of PTH to animals, the structure of osteoclasts, osteoblasts, and osteocytes (those bone cells trapped within the calcified matrix) is altered within minutes.5 Those morphologic changes are accompanied by increased activity of osteoclasts and inhibition of osteoblastic function, leading to an increase in net skeletal calcium release within 2 to 3 hours. The PTH-induced increase in the size of the periosteocytic lacunae also has been considered presumptive evidence for a role for this cell type in skeletal calcium release. Continued exposure to PTH causes an increase in the number and activity of osteoclasts that is ultimately accompanied by an increase in osteoblastic activity through the coupling of bone resorption to formation, as noted previously. The mechanisms by which PTH modulates bone cell function remain to be fully elucidated. The hormone raises skeletal levels of cAMP64 but may also act through other second messenger systems, including the activation of phospholipase C.56 In addition to modulating bone turnover indirectly, by altering the rate of PTH secretion and/or 1,25(OH)2D3 production, changes in Ca2+o also directly regulate bone cell function in vitro in ways that probably contribute to the control of bone turnover in vivo. Raising Ca2+o stimulates several aspects of the functions of osteoblasts and preosteoblasts, such as enhancing their proliferation and chemotaxis.18 These actions could contribute to the coupling of osteoclastic bone resorption to the subsequent replacement of the missing bone by osteoblasts. Conversely, elevated levels of Ca2+o directly inhibit osteoclastic function,19 possibly providing a way for osteoclasts to autoregulate their activity as a function of the amount of calcium that they have resorbed. Pharmacologic evidence has implicated distinct mechanisms for Ca2+o sensing in osteoblasts and osteoclasts,18,19 which differ in turn from the Ca2+o-sensing receptor that regulates various aspects of parathyroid and renal function.7,14 The latter receptor, however, can be expressed by both osteoblastic65 and osteoclastic cells,66 and further studies are needed to establish with certainty the molecular mechanisms through which bone cells sense Ca2+o. Changes in the level of extracellular phosphate also modulate bone turnover as noted previously. Elevations in phosphate stimulate bone formation and inhibit bone resorption, whereas hypophosphatemia produces the converse effects.7 As with the known direct actions of phosphate on parathyroid function, the mechanisms underlying the sensing of extracellular phosphate ions by bone cells remain obscure. REGULATION OF CIRCULATING CALCIUM CONCENTRATION The information in Figure 49-2 does not delineate the temporal sequence of how the homeostatic system reacts to perturbations in Ca2+o. Actually, there is a hierarchy of responses by both the parathyroid gland and the effector systems that regulate calcium transport in the skeleton, kidney, and intestine.7 The initial alteration in the secretory rate of PTH (release of preformed stores of hormone) occurs within seconds of the change in Ca2+o. Within 15 to 30 minutes, an increase occurs in the net synthesis of PTH, without any change in the levels of messenger RNA (mRNA), because of reduced intracellular degradation of PTH. If the hypocalcemic stimulus persists, the levels of PTH mRNA increase over the ensuing 12 to 24 hours or more.38,39 Eventually, chronic hypocalcemia is associated with enhanced parathyroid cellular proliferation over days to weeks,67 which further increases the secretory rate of the hormone. Reduced levels of 1,25(OH)2D3 are also a likely stimulus for parathyroid cellular proliferation and contribute to the severe hyperparathyroidism that can be encountered in chronic renal insufficiency. The most rapid changes in the handling of calcium by the effector organs of the homeostatic system occur in the skeleton and kidney. Alterations in distal tubular calcium reabsorption take place within minutes in vitro,55 whereas the skeletal release of calcium occurs within 2 to 3 hours.68 If these two functional changes are insufficient to restore normocalcemia, the continued hypersecretion of PTH stimulates increased synthesis of 1,25(OH)2D3 within 1 to 2 days,51 enhances the activity of existing osteoclasts, and promotes the appearance of new osteoclasts within days to weeks. Only rarely (e.g., in severe vitamin D deficiency) is the full complement of homeostatic responses insufficient to restore normocalcemia. Exposure of the homeostatic system to a calcium load produces responses that are largely the opposite of those seen with hypocalcemia. The suppression of parathyroid function induces a renal leak of Ca2+, reduces the release of skeletal Ca2+, and ultimately suppresses GI absorption of Ca2+ by inhibiting the synthesis of 1,25(OH)2D3. The remarkable sensitivity of the system is illustrated by its response to the ingestion of the Ca2+ present in a glass of milk. A minute rise in serum Ca2+ causes approximately a 30% reduction in PTH secretion, which leads to prompt excretion of much of the extra Ca2+ in the urine.5 In response to severe loads of Ca2+, the skeleton can buffer substantial quantities of Ca2+, and the kidney can excrete up to 1 g of Ca2+ over 24 hours; however, hypercalcemia may ensue, particularly if

renal impairment develops. One of the most elegant features of the homeostatic system for Ca2+ is that it simultaneously contributes to the regulation of the serum phosphate concentration. With Ca2+ deficiency sufficient to produce secondary hyperparathyroidism, the excess phosphate mobilized into the extracellular fluid from intestine and bone is excreted in the urine (mild hypophosphatemia may ensue). Conversely, with oral Ca2+ loading, reduced GI and skeletal availability of phosphate are managed by decreased renal phosphate clearance because of reduced PTH secretion. Primary abnormalities in phosphate metabolism also elicit changes in the Ca2+o homeostatic system that tend to correct both serum phosphate and calcium concentrations. Hypophosphatemia, for example, enhances the synthesis of 1,25(OH)2D3, which then increases intestinal absorption and skeletal release of phosphate and calcium52 (Fig. 49-7). The increased movement of Ca2+ into the extracellular fluid suppresses PTH release, thereby enhancing the excretion of the excess Ca2+ as well as retaining the phosphate mobilized from intestine and bone. The net result of these adaptations is some normalization of serum phosphate without any change in the serum Ca2+ concentration. Through poorly defined mechanisms, the kidney also retains phosphate more avidly in states of phosphate depletion, independent of alterations in PTH secretion. Clearly, as with the actions of phosphate on parathyroid and bone cells, some mechanism must exist through which these alterations in extracellular phosphate availability and/or levels are recognized by the kidney and transduced into alterations in 1,25(OH)2D3 synthesis and tubular phosphate reabsorption. Moreover, additional hormones regulating phosphate homeostasis likely will be discovered that are involved in the pathogenesis of inherited and acquired disorders of the regulation of the serum phosphate concentration.69

FIGURE 49-7. Response of the homeostatic system to hypophosphatemia. Hypophosphatemia per se stimulates synthesis of 1,25(OH)2D3. Excess Ca2+ mobilized with phosphate from bone and intestine suppresses parathyroid hormone (PTH) release, facilitating calciuresis. Decreased renal phosphate clearance resulting from reduced circulating levels of PTH retains phosphate available from intestine and bone. Low phosphate also may directly inhibit PTH release, increase bone resorption, and enhance renal phosphate reabsorption. (ECF, extracellular fluid.)

CLINICAL ASSESSMENT OF CALCIUM HOMEOSTASIS

The clinical evaluation of normal and abnormal calcium homeostasis requires an accurate assessment of serum calcium and phosphate concentrations as well as the functions of the parathyroid glands and the effector systems regulating calcium homeostasis. Total serum calcium and phosphate measurements are usually reliable. Both dip-type and flow-through electrodes of improved quality are widely available for measuring serum ionized calcium and may be useful when changes in serum protein levels make the accurate estimation of ionized from total calcium levels difficult. Measurement of the serum ionized calcium concentration may also be helpful in cases of mild primary hyperparathyroidism, in which an elevation of serum ionized calcium is sometimes more readily detectable than one of serum total calcium concentration. DIRECT ASSAYS OF CALCIUM-REGULATING HORMONES IN BLOOD PARATHYROID HORMONE Most of the immunoreactive PTH in the circulation represents inactive fragments of the hormone comprising the mid-molecule and carboxy-terminal portions of the molecule70 (see Chap. 51). The amount of biologically active PTH (1-84) in the blood represents <10% of the total immunoreactivity. The advent of specific and sensitive double-antibody (e.g., immunoradiometric or immunochemiluminescent) assays specific for intact PTH has largely supplanted the previously used mid-molecule and carboxy-terminal assays.71 Although these double-antibody assays were originally thought to recognize only the intact form of PTH, later studies have shown that they also detect variable amounts (1025% or more) of additional forms of immunoreactive PTH.72 Nevertheless, these intact PTH assays permit reliable diagnosis of primary and secondary hyperparathyroidism when interpreted in the context of the simultaneously measured serum calcium concentration. Furthermore, in hypercalcemia due to nonparathyroid causes, intact PTH levels are generally frankly suppressed, greatly facilitating the differential diagnosis of hypercalcemia. The need seldom arises, therefore, to measure PTH by the older immunoassays or more cumbersome bioassays. Because high-quality assays for parathyroid hormonerelated protein are also routinely available,73 measurement of urinary cAMP excretion is usually not required to diagnose parathyroid hormonerelated, protein-mediated hypercalcemia of malignancy. An assay that appears to be specific for PTH (1-84) per se is under development. CALCITONIN The determination of the immunoreactive calcitonin level is most useful in screening patients suspected of harboring medullary thyroid carcinoma (MTC).74 In this setting, the calcitonin level is determined both before and after the administration of a secretagogue, such as calcium or pentagastrin. An abnormally large increase in the circulating level of calcitonin after penta-gastrin administration generally indicates that the patient harbors a C-cell neoplasm. The identification of the gene (the retoncogene) causing familial MTC and multiple endocrine neoplasia type 2 (MEN2)75 now permits the clinician to identify obligate gene carriers by genetic means rather than relying solely on pentagastrin testing. How or to what extent serum calcitonin influences calcium homeostasis in normal humans is uncertain (see Chap. 53). VITAMIN D METABOLITES Assays are available for measuring 25(OH)D3 and 1,25(OH)2D3.4,5 and 6 The former is measured to assess circulating stores of the vitamin. Serum is extracted and purified chromatographically; 25(OH)D3 is then measured either directly by its absorption of ultraviolet light, by a radioligand-binding assay that uses endogenous binding proteins, or by radioimmunoassay. Because of the low levels of 1,25(OH)2D3 in blood (~30 pg/mL), this metabolite generally must be extensively purified before assay and then measured using the naturally occurring cellular receptor in radioligand assays or by radioimmunoassay. The clinical settings in which the measurement of this metabolite is useful are discussed in Chapter 63, Chapter 70 and Chapter 219. ASSESSMENT OF THE FUNCTION OF TARGET ORGANS FOR CALCIUM-REGULATING HORMONES KIDNEY Along with measurement of urinary cAMP excretion, determination of the tubular reabsorption of phosphate (TRP) is another indirect assessment of PTH action on the kidney that is seldom required now that high-quality assays for intact PTH are widely available (see previous section on PTH assays). This parameter is calculated from the equation TRP = 1 [ Up Scr/Ucr Sp], where Up = urinary phosphate excretion, Ucr = urinary creatinine excretion, and Sp and Scr = serum phosphate and creatinine excretion, respectively. A particularly useful expression of renal phosphate handling is the tubular maximum for phosphate corrected for glomerular filtration rate, which can be calculated using an appropriate nomogram.76 This parameter is particularly useful in assessing the contribution of renal leak of phosphate to hypophosphatemic, osteomalacic disorders (see Chap. 63 and Chap. 70). The measurement of the relationship between serum calcium and urine calcium excretion provides some indirect information about the PTH-calciferol axis (see Fig. 49-6). Calcium excretion is often expressed in terms of the calcium/creatinine excretion ratio. A more rigorous parameter of renal calcium handling is the ratio of calcium to creatinine clearance, which is characteristically lower in patients with FHH (usually <0.01) than in those with primary hyperparathyroidism.25

INTESTINE The direct measurement of calcium absorption provides an indication of the intestinal response to circulating levels of 1,25(OH)2D3 and, indirectly, PTH. Calcium absorption may be assessed by determining the difference between oral calcium intake and fecal calcium excretion or by using isotopic techniques.77 The former is laborious and requires equilibration of mineral homeostasis to a constant diet over weeks. The latter is accomplished by administering a tracer dose of calcium-47 (47Ca) or other isotopes of calcium with a fixed amount of stable calcium (40Ca, usually 100 mg) and measuring the appearance of the isotope in blood samples during the following 0.5 to 6 hours. Unfortunately, neither technique is widely available. BONE Alkaline phosphatase in serum originates from many sources, of which bone and liver are the two principal ones in adult humans.78 Circulating alkaline phosphatase of skeletal origin derives primarily from osteoblasts; measurement of its activity in serum can be a useful, albeit indirect, indicator of osteoblastic activity. A crude way of distinguishing the activity of skeletal alkaline phosphatase from that arising from liver is the greater lability of the former to heat. Immunoassays specific for the skeletal isoenzyme of alkaline phosphatase are available; their utility will become clearer as greater experience with them accumulates.78 A bone-derived serum protein, osteocalcin (sometimes called bone Gla protein), can be a marker of osteoblast function. This protein contains g-carboxyglutamic acid (Gla) synthesized through a vitamin Kdependent carboxylation of glutamate, and it may be measured by radioimmunoassay.79 When bone resorption and formation are coupled, it is a marker of bone turnover. If they are not well coupled (i.e., during therapy with glucocorticoids), it is a marker of bone formation.78 Assays of bone formation are available that measure the serum levels of the amino-terminal and carboxy-terminal propeptides of type I collagen,78 which are cleaved off during the maturation of bone collagen. Additional experience is needed to evaluate further their clinical utility, but as with other bone markers, they are a rather indirect measure of bone cell activity. Several clinically useful markers of bone resorption arise from products of the degradation of bone collagen.78 Because ~60% of total body collagen resides in bone, the determination of urine, or in some case serum, levels of these markers provides information about the turnover of bone matrix. The assay of urinary hydroxyproline as a measure of bone turnover has been supplanted by the developments of methods for measuring various breakdown products of the amino acids that cross-link collagen molecules in bone (hydroxylysyl pyridinolines = pyridinolines [Pyr]; or lysyl pyridinolines = deoxypyridino-lines [d-Pyr]).78 Approximately two-thirds of the Pyr and d-Pyr in urine are in the form of small peptides, and the remainder are the free amino acids. Pyr and d-Pyr are generally measured after acid hydrolysis to convert Pyrand d-Pyrcontaining peptides to the free amino acids. Two additional markers of bone resorption that are currently being evaluated are pyridinoline-containing peptides arising from the aminoand carboxy-termini of bone collagen (cross-linked N-telopeptides [NTx] and C-telopeptides [ICTP], respectively). As clinical experience accumulates, the indication is that assays based on cross-links of bone collagen will be useful, relatively specific urine (and perhaps serum) markers of bone resorption (see Chap. 56). CHAPTER REFERENCES
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. Rasmussen H. Calcium messenger system. N Engl J Med 1986; 314:1089. Meldolesi J, Pozzan T. The endoplasmic reticulum Ca 2+ store: a view from the lumen. Trends Biochem Sci 1998; 23:10. Nathanson MH. Cellular and subcellular calcium signaling in gastrointestinal epithelium. Gastroenterology 1994; 106:1349. Parfitt AM, Kleerkoper M. The divalent ion homeostatic system: physiology and metabolism of calcium, phosphorous, magnesium, and bone. In: Maxwell MH, Kleeman CR, eds. Clinical disorders of fluid and electrolyte metabolism, 3rd ed. New York: McGraw-Hill, 1980:269. Stewart AF, Broadus AE. Mineral metabolism. In: Felig P, Baxter JD, Broadus AE, Frohman LA, eds. Endocrinology and metabolism. New York: McGraw-Hill, 1987:1317. Bringhurst FR. Calcium and phosphate distribution, turnover and metabolic actions. In: DeGroot LJ, ed. Endocrinology, 3rd ed. Philadelphia, WB Saunders, 1995:1015. Brown EM. Extracellular Ca2+-sensing, regulation of parathyroid cell function, and role of calcium and other ions as extracellular (first) messengers. Physiol Rev 1991; 71:371. Walser M. Ion association. VI. Interactions between calcium, magnesium, inorganic phosphate, citrate, and protein in normal human plasma. J Clin Invest 1961; 40:723. Lindgarde F, Zettervall O. Hypercalcemia and normal ionized serum calcium in a case of myelomatosis. Ann Intern Med 1973; 78:396. Heaney RP, Gallagher JC, Johnston CC, et al. Calcium nutrition and bone health in the elderly. Am J Clin Nutr 1982; 36:986. Miller GD, Groziak SM, DiRienzo D. Age considerations in nutrient needs for bone health. J Am Coll Nutr 1996; 15:553. Bo-Linn GW, Davis GR, Buddru DJ, et al. An evaluation of the importance of gastric acid secretion in the absorption of dietary calcium. J Clin Invest 1984; 73:640. Nicar MJ, Pak CY. Calcium bioavailability from calcium carbonate and calcium citrate. J Clin Endocrinol Metab 1985; 61:391. Brown EM, Gamba G, Riccardi D, et al. Cloning and characterization of an extra-cellular Ca 2+-sensing receptor from bovine parathyroid. Nature 1993; 366:575. Weisinger JR, Favus MJ, Langman CB, Bushinsky DA. Regulation of 1,25- dihydroxyvitamin D 3 by calcium in the parathyroidectomized, parathyroid hormone-replete rat. J Bone Miner Res 1989; 4:929. Fraser DR, Kodicek E. Regulation of 25-hydroxy-cholecalciferol-hydroxylase activity in kidney by parathyroid hormone. Nature New Biol 1973; 241:163. Holick MF. Vitamin D: photobiology, metabolism, and clinical applications. In: DeGroot L, Besser, H, Burger HG, et al., eds. Endocrinology, 3rd ed. Philadelphia, WB Saunders, 1995:990. Quarles LD. Cation-sensing receptors in bone: a novel paradigm for regulating bone remodeling? J Bone Miner Res 1997; 12:1971. Zaidi M, Adebanjo OA, Moonga BS, Sun L, Huang CL. Emerging insights into the role of calcium ions in osteoclast regulation. J Bone Miner Res 1999; 14:669. Slatopolsky E, Finch J, Denda M, et al. Phosphorus restriction prevents parathyroid gland growth. High phosphorus directly stimulates PTH secretion in vitro. J Clin Invest 1996; 97:2534. Brown EM, Vassilev PM, Quinn S, Hebert SC. G protein-coupled, extracellular Ca2+-sensing receptor: a versatile regulator of diverse cellular functions. Vitamins and Hormones 1999; 55:1. Gittes RF, Radde JC. Experimental model for hyperparathyroidism: effect of excessive numbers of transplanted isologous parathyroid glands. J Urol 1966; 95:595. Adami S, Muirhead N, Manning RM, et al. Control of secretion of parathyroid hormone in secondary hyperparathyroidism. Clin Endocrinol 1982; 16:463. Grant FD, Conlin PR, Brown EM. Rate and concentration dependence of parathyroid hormone dynamics during stepwise changes in serum ionized calcium in normal humans. J Clin Endocrinol Metab 1990; 71:370. Brown EM, Kifor O, Bai M. Decreased Responsiveness to Extracellular Ca 2+due to abnormalities in the Ca2+o-sensing receptor. In: Jameson U, ed. Contemporary Endocrinology: hormone resistance syndromes. Totowa, NJ: Humana Press 1999:87. Kobayashi M, Tanaka H, Tsuzuki K, et al. Two novel missense mutations in calcium-sensing receptor gene associated with neonatal severe hyperparathyroidism. J Clin Endocrinol Metab 1997; 82:2716. Bai M, Pearce SH, Kifor O, et al. In vivo and in vitro characterization of neonatal hyperparathyroidism resulting from a de novo, heterozygous mutation in the Ca2+-sensing receptor gene: normal maternal calcium homeostasis as a cause of secondary hyperparathyroidism in familial benign hypocalciuric hypercalcemia. J Clin Invest 1997; 99:88. Pollak MR, Brown EM, Estep HL, et al. Autosomal dominant hypocalcemia caused by a Ca 2+-sensing receptor gene mutation. Nat Genet 1994; 8:303. Riccardi D, Hall AE, Chattopadhyay N, et al. Localization of the extracellular Ca2+/polyvalent cation-sensing protein in rat kidney. Am J Physiol 1998; 274:F611. Chattopadhyay N, Brown EM. Calcium-sensing receptor: roles in and beyond systemic calcium homeostasis. Biol Chem 1997; 378:759. Hebert SC, Brown EM, Harris HW. Role of the Ca 2+-sensing receptor in divalent mineral ion homeostasis. J Exp Biol 1997; 200:295. Habener JT, Potts JT Jr. Relative effectiveness of calcium and magnesium on the secretion and biosynthesis of parathyroid hormone in vitro. Endocrinology 1976; 98:197. De Rouffignac C, Quamme GA. Renal magnesium handling and its hormonal control. Physiol Rev 1994; 74:305. Anast CS, Mohs JM, Kaplan SL, Burns TW. Evidence for parathyroid failure in magnesium deficiency. Science 1972; 177:606. Brown EM. Parathyroid secretion in vivo and in vitro: regulation by calcium and other secretagogues. Miner Electrolyte Metab 1982; 8:130. Russell J, Silver J, Sherwood LM. The effects of calcium and vitamin D metabolites on cytoplasmic mRNA coding for preproparathyroid hormone in isolated parathyroid cells. Trans Assoc Am Physicians 1984; 97:269. Chan YK, McKay C, Dye E, Slatopolsky E. The effect of 1,25 dihydroxy-cholecalciferol on parathyroid hormone secretion by monolayer cultures of bovine parathyroid cells. Calcif Tissue Int 1986; 38:27. Yamamoto M, Igarishi T, Muramatsu M, et al. Hypocalcemia increases and hypercalcemia decreases the steady-state level of parathyroid hormone messenger RNA in the rat. J Clin Invest 1989; 83:1053. Silver J, Naveh-Many T, Mayer J, et al. Regulation by vitamin D metabolites of parathyroid hormone gene transcription in vivo in the rat. J Clin Invest 1986; 78:1296. Kremer R, Bolivar I, Goltzman D, Hendy GN. Influence of calcium and 1,25-dihydroxycholecalciferol on proliferation and proto-oncogene expression in primary cultures of bovine parathyroid cells. Endocrinology 1989; 125:935. Mallette LE, Khouri K, Zengolita H, et al. Lithium treatment increases midregion parathyroid hormone and parathyroid volume. J Clin Endocrinol Metab 1989; 68:654. Brown EM. Lithium induces abnormal calcium-regulated PTH release in dispersed bovine parathyroid cells. J Clin Endocrinol Metab 1981; 52:1046. Silver J, Moallem E, Kilav R, et al. New insights into the regulation of parathyroid hormone synthesis and secretion in chronic renal failure. Nephrol Dial Transplant 1996; 11:2. Nemeth EF, Scarpa A. Cytosolic Ca++ and the regulation of secretion in parathyroid cells. FEBS Lett 1986; 213:15. Kifor O, Kifor I, Brown EM. Effects of high extracellular calcium concentrations on phosphoinositide turnover and inositol phosphate in dispersed bovine parathyroid cells. J Bone Miner Metab 1992; 7:1327. Marx SJ, Lasker R, Brown E, et al. Secretory dysfunction in parathyroid cells from a neonate with severe primary hyperparathyroidism. J Clin Endocrinol Metab 1986; 62:445.

46a. Hoenderop JGJ, Van de Graf AWCM, Hartog A, et al. Molecular identification of the apical Ca 2+ channel in 1,25-dihydroxyvitamin D 3-responsive epithelia. J Biol Chem 1999; 274:22739 46b. Peng JB, Chen XZ, Berger UV, et al. Molecular cloning and characterization of a channel-like transporter mediating intestinal absorption. J Biol Chem 1999; 274:22739. 47. Favus MJ. Intestinal absorption of calcium, magnesium and phosphorus. In: Coe FL, Favus MJ, eds. Disorders of bone and mineral metabolism. New York: Raven Press, 1992:57. 48. DeLuca HF. The Vitamin D story: a collaborative effort of basic science and clinical medicine. Fed Proc Am Soc Exper Biol 1988; 2:2124. 49. Rasmussen H, Fontaine O, Max EE, Goodman DBP. The effect of 1,alpha- hydroxyvitamin D3 administration on calcium transport in chick intestine brush border membrane vesicles. J Biol Chem 1979; 25:2993. 50. Bikle DD. Regulation of intestinal calcium transport by vitamin D: role of membrane structure. In: Aloia RC, Curtain KC, Gordon LM, eds. Membrane transport and information storage. New York: Wiley Ross, 1990:191. 51. Adams ND, Gray RW, Lemann J. The effect of oral CaCO3 loading and dietary calcium deprivation on plasma 1,25-dihydroxyvitamin D concentrations in healthy adults. J Clin Endocrinol Metab 1979; 48:1008. 52. Hughes MR, Brumbaugh PF, Haussler MR, et al. Regulation of serum 1,alpha-25-dihydroxyvitamin D 3 by calcium and phosphate in the rat. Science 1975; 190:578. 53. Sheikh MS, Ramirez A, Emmett M, et al. Role of vitamin D-independent mechanisms in absorption of food calcium. J Clin Invest 1988; 81:126. 54. Nordin BEC, Peacock M. Role of the kidney in regulation of plasma calcium. Lancet 1969; 2:1280.

55. Friedman PA, Gesek FA. Cellular calcium transport in renal epithelia: measurement, mechanisms. and regulation. Physiol Rev 1995; 75:429. 55a. Hoenderop JG, Willems PH, Bindels RJ. Toward a comprehensive molecular model of active calcium reabsorption. Am J Physiol 2000; 278:F352. 56. Abou-Samra A, Juppner H, Force T, et al. Expression cloning of a common receptor for parathyroid hormone and parathyroid hormone-related peptide from rat osteoblast-like cells: a single receptor stimulates intracellular accumulation of both cAMP and inositol phosphates and increases intra-cellular free calcium. Proc Natl Acad Sci U S A 1992; 89:2732. 57. Morel F, Chabardes D, Imbert-Teboul M, et al. Multiple hormonal control of adenylate cyclase in distal segments of the rat kidney. Kidney Int 1982; 11(Suppl):555. 58. Sonnenberg J, Pansini AR, Christakos S. Vitamin D-dependent rat renal calcium-binding proteins: development of a radioimmunoassay, tissue distribution, and immunologic identification. Endocrinology 1984; 115:640. 59. Ro HK, Tembe V, Favus MS. Evidence that activation of protein kinase C can stimulate 1,25-dihydroxyvitamin D 3 secretion by rat proximal tubules. Endocrinology 1992; 131:1424. 60. Sands JM, Naruse M, Baum M, et al. Apical extracellular calcium/polyvalent cation-sensing receptor regulates vasopressin-elicited water permeability in rat kidney inner medullary collecting duct. J Clin Invest 1997; 99:1399. 61. Rodan GA, Martin TJ. Role of osteoblasts in hormonal control of bone resorption: a hypothesis. Calcif Tissue Int 1981; 33:349. 62. Yasuda H, Shima N, Nakagawa N, et al. Osteoclast differentiation factor is a ligand for osteoprotegerin/osteoclastogenesis-inhibitory factor and is identical to TRANCE/RANKL. Proc Natl Acad Sci U S A 1998; 95:3597. 63. Canalis E, Pash J, Varghese S. Skeletal growth factors. Crit Rev Eukaryot Gene Expr 1993; 3:155. 64. Chase LR, Fedak SA, Aurbach GD. Activation of skeletal adenyl cyclase by parathyroid hormone in vitro. Endocrinology 1969; 84:761. 65. Yamaguchi T, Chattopadhyay N, Kifor O, et al. Mouse osteoblastic cell line (MC3T3-E1) expresses extracellular calcium (Ca 2+o)-sensing receptor and its agonists stimulate chemotaxis and proliferation of MC3T3-E1 cells. J Bone Miner Res 1998; 13:1530. 66. Kameda T, Mano H, Yamada Y, et al. Calcium-sensing receptor in mature osteoclasts, which are bone resorbing cells. Biochem Biophys Res Commun 1998; 245:419. 67. Lee MJ, Roth SI. Effect of calcium and magnesium on deoxyribonucleic acid synthesis in rat parathyroid glands in vitro. Lab Invest 1975; 33:72. 68. Robertson WG, Peacock M, Alkins D. The effect of parathyroid hormone on the uptake and release of calcium by bone in tissue culture. Clin Sci 1972; 43:715. 69. Rowe PS, Oudet CL, Francis F, et al. Distribution of mutations in the PEX gene in families with X-linked hypophosphataemic rickets (HYP). Hum Mol Genet 1997; 6:539. 70. Martin KJ, Hruska KA, Freitag JJ, Slatopolsky E. The peripheral metabolism of parathyroid hormone. N Engl J Med 1979; 301:1092. 71. Nussbaum SR, Zahradnik RJ, Lavigne JR. Highly sensitive two-site immunoradiometric assay of parathyrin, and its clinical utility in evaluating patients with hypercalcemia. Clin Chem 1987; 33:1364. 72. Lepage R, Roy L, Brossard JH, et al. A non-(1-84) circulating parathyroid hormone (PTH) fragment interferes significantly with intact PTH commercial assay measurements in uremic samples. Clin Chem 1998; 44:805. 73. Burtis WJ, Brady TG, Orloff JJ, et al. Immunochemical characterization of hypercalcemia of malignancy. N Engl J Med 1990; 322:1106. 73a. John MR, Goodman WG, Gao P, et al. A novel immunoradiometric assay detects full-length human PTH but not amino-terminally truncated fragments: implications for PTH measurements in renal failure. J Clin Endocrinol Metab 1999; 84:4287. 74. 75. 76. 77. 78. 79. Deftos LJ. Calcitonin and medullary thyroid carcinoma. In: Bennett JC, Plum F, eds. Cecil textbook of medicine, 20th ed. Philadelphia: WB Saunders, 1996:1372. Gagel RF. Multiple endocrine neoplasia type II and familial medullary thyroid carcinoma. Impact of genetic screening on management. Cancer Treat Res 1997; 89:421. Walton RJ, Bijvoet OLM. Nomogram for determination of renal threshold phosphate concentration. Lancet 1975; 2:309. Heaney RP, Recker RR, Hinders SM. Variability of calcium absorption. Am J Clin Nutr 1988; 47:262. Garnero P, Delmas PD. Biochemical markers of bone turnover. Applications for osteoporosis. Endocrinol Metab Clin North Am 1998; 27:303. Price PA, Baukol SA. 1,25-Dihydroxyvitamin D3 increases synthesis of the vitamin K-dependent bone protein by osteosarcoma cells. J Biol Chem 1980; 255:11660.

CHAPTER 50 PHYSIOLOGY OF BONE Principles and Practice of Endocrinology and Metabolism

CHAPTER 50 PHYSIOLOGY OF BONE

LAWRENCE G. RAISZ Embryology and Anatomy of Bone Modeling and Remodeling Natural History of the Skeleton Bone Chemistry and Mineralization Cell Biology of Bone Osteoblasts Osteoclasts Other Cell Types in Bone Effects of Hormones on Bone Cells Parathyroid Hormone 1,25-Dihydroxyvitamin D3 Calcitonin Systemic Hormones that Affect Bone Metabolism Glucocorticoids Growth Hormone and Insulin-Like Growth Factors Insulin Thyroid Hormones Sex Hormones Other Systemic Hormones Local Regulators Chapter References

Understanding of the physiology of skeletal tissue has advanced remarkably during the past few decades. Studies of the various cell types in bone and their interactions have led to the concept of a system involving not only systemic hormones but also local factors that regulate bone turnover. Also, molecular biology techniques are now being applied to bone cells and have provided rapid advances. Structurally, bone must provide a framework for locomotion, must protect internal organs and marrow, and must be able to adapt to changing physical stress. Metabolically, the skeleton functions as a storehouse and as a homeostatic buffer system. Presumably, bone evolved to fulfill both its structural and metabolic roles as our ancestors moved from the calcium-rich, buoyant ocean to fresh water, and then to dry land. Although the major reservoir function of bone is to supply calcium and phosphorus, the skeleton also serves as a reservoir and source of other ions, such as magnesium and sodium, and as a buffer to deal with hydrogen ion excess. Also, the ability of the skeleton to take up a variety of trace elements may serve as an important safeguard against their toxicity. To achieve its mechanical functions, the skeleton needs to be selectively responsive to different kinds of strain, light, of high tensile strength, and rigid but not brittle.1,2 This is achieved by an orderly, slightly deformable, mineralized collagen structure distributed as a combination of dense cortical bone and spongy trabecular bone.

EMBRYOLOGY AND ANATOMY OF BONE

The formation of the skeleton begins with condensation and differentiation of mesenchymal cells into cartilage. These provide the template for subsequent bone formation in two ways. Bone may begin to form by differentiation of osteoblasts adjacent to the cartilage rudiments. This occurs in the membranous bones, such as the skull and the periosteum of long bones. Here, the mesenchyme condenses, and osteoblasts differentiate on the surface of the cartilaginous template, which then degenerates. Endochondral bone formation occurs at the cartilage growth plate (Fig. 50-1). Here, the osteoblasts differentiate directly on calcified cartilage and form spicules of bone with a cartilaginous core. In either case, the first bone formed is of a loose, woven structure, and it is then replaced by lamellar bone.3

FIGURE 50-1. Bone remodeling and bone growth at the epiphyseal plate. (R, resorption; CC, calcified cartilage; Rev, reversal; F, formation; WB, woven bone; LB, lamellar bone.) (From Baron R. Anatomy and ultrastructure of bone. In: Primer on the metabolic bone diseases and disorders of bone metabolism, 3rd ed. New York: Raven Press, 1996.)

Long bones lengthen by the proliferation of cartilage cells. The cartilage undergoes an orderly change, in which the columns of cells in the proliferative zone become hypertrophied and the matrix between these columns becomes mineralized. This probably involves both breakdown of the highly hydrated high-molecular-weight proteoglycans of cartilage matrix and the release of matrix vesicles from hypertrophic chondrocytes. After the cartilage is mineralized, new bone formation by osteoblasts begins on the surface of the calcified cartilage spicules; this is called the primary spongiosa. Subsequently, the spicules are resorbed and replaced by bone, termed the secondary spongiosa. Early in fetal bone formation, collagen is laid down in a woven and irregular pattern, but soon the osteoblasts deposit an orderly lamellar arrangement of collagen. The combination of a smooth, dense outer layer of cortical (compact) bone and spongy trabecular (cancellous) bone provides both the necessary strength without excessive weight and an extended surface on which rapid changes in formation or resorption can respond to changing metabolic needs. The metabolic responses of the skeleton occur mainly on the trabecular bone surfaces and the endosteal (inner) surface of the cortex. However, even the periosteal (outer) surface of the cortex can be affected by calcium-regulating hormones, as evidenced by the development of subperiosteal bone resorption in severe hyperparathyroidism.

MODELING AND REMODELING

The term modeling refers to the process by which bone grows and alters its shape through resorption and formation at different sites. For example, the long bones enlarge by periosteal formation and endosteal resorption. As they lengthen, the large amount of bone formed at the growth plate is resorbed to maintain a hollow cylindrical structure (Fig. 50-2). The flat bones of the skull and pelvis grow and change their shape by this process of formation at one site and resorption at another. Modeling can be influenced by mechanical stress. For example, bone mass increases in the most used long bones of athletes by both periosteal and endosteal apposition. Bone formation on one surface and resorption on the other permit teeth to be moved by mechanical forces. Even pressure from soft tissues can produce modeling changes. For example, increased endosteal resorption and periosteal apposition can occur in response to bone marrow hyperplasia.

FIGURE 50-2. Modeling during the longitudinal growth of long bones. As the growth plate moves upward (see Fig. 50-1), the wider parts of the bone must be reshaped into a tubular diaphysis. (From Baron R. Anatomy and ultrastructure of bone. In: Primer on the metabolic bone diseases and disorders of bone metabolism, 3rd ed. New York: Raven Press, 1996.)

The term remodeling refers to the process in which resorption is followed by formation at the same site; hence, the two processes are coupled.4 This process is important for the overall growth and functional integrity of skeletal tissues as well as for the metabolic responses of the bone mineral reservoir. In large mammals, the cortical bone is remodeled by the development of a haversian system of osteons (Fig. 50-3). These structures are formed by osteoclastic removal of a cylinder of bone. Behind these osteoclasts are a vascular loop and mesenchymal cells that differentiate into osteoblasts and form concentric lamellae of new bone around the central vascular canal. The major importance of this system of osteons may be that it enables the cortex to participate in the metabolic functions of the skeleton without excessive loss of strength. Haversian remodeling may also be important in the repair of fatigue damage in bone.5 Remodeling also occurs on the trabecular bone surface. This begins with the formation of the primary spongiosa. Osteoclasts excavate scalloped areas (called Howship lacunae) on mature trabecular bone, which are then replaced by packets of new lamellar bone laid down by osteoblasts. Although much is known about the anatomic sequence and time course of cycles of cortical and trabecular remodeling (Fig. 50-4), the cellular mechanisms are poorly understood. The initial activation involves a change in the lining cells or resting osteoblasts on the surface of bone, which may normally protect it from attack by osteoclasts. These may not only change shape but also release factors that stimulate osteoclastic activity. The development of osteoclasts requires an interaction between precursor cells of the osteoblastic and osteoclastic lineages; some of the proteins involved in this interaction are known (see later). Once osteoclasts have removed the bulk of bone mineral and matrix, a reversal phase occurs that involves the removal of additional elements of bone matrix by macrophages and the preparation of the bone surface for new osteoblastic formation. In the formation phase, successive generations of osteoblasts synthesize lamellar bone and replace the resorbed bone with a packet of new bone, called a bone structural unit (BSU).

FIGURE 50-3. Haversian remodeling. This is a longitudinal section through a cortical bone remodeling unit (400). The haversian canal is formed by an osteoclast-cutting cone. After the canal reaches maximal diameter, mesenchymal cells differentiate into osteoblasts and begin to form an osteon made of concentric lamellae of new matrix. (Courtesy of Dr. Robert Schenk.)

FIGURE 50-4. Bone remodeling cycle. Activation of this cycle begins with a change in the resting osteoblasts (also called lining cells) on the bone surface, which permits osteoclastic bone resorption, and a signal from cells of the osteoblastic lineage, which make contact with osteoclast precursors. Osteoclasts and macrophages are probably derived from different immediate precursors, but both originate from hematopoietic stem cells. During the reversal phase in trabecular bone remodeling, macrophages are seen on the bone surface, but their role is not established. Osteoblasts come from a separate mesenchymal stem cell population, probably related to stromal stem cells in the marrow, which differentiate into mature osteoblasts that replace the bone lost during the resorption phase. (From Raisz LG. Local and systemic factors in the pathogenesis of osteoporosis. N Engl J Med 1988; 318:818.)

NATURAL HISTORY OF THE SKELETON

In humans, the skeleton continues to grow until 25 to 35 years of age. Before puberty, the bones grow by periosteal apposition and endosteal resorption. During puberty and adolescence, a period of endosteal apposition and thickening of the trabeculae occurs, so that bone mass increases by 10% to 20% even after linear growth has ceased.6 During the years from 20 to 50, bone mass is stable, because rates of formation and resorption are equal or coupled. The formation of new BSUs both in the cortex and on the surface of trabecular bone continues throughout life. Later in life, bone mass begins to decrease due to uncoupled remodeling. Age-related bone loss is greater for trabecular than for cortical bone, and more rapid in women than in men, largely because of the acceleration of remodeling at menopause. Bone mass can decrease simply because more BSUs are formed (increased turnover), because a time gap exists between resorption and formation even when the processes are coupled. This deficit is theoretically reversible if the rate of turnover decreases and the formation phase is allowed to proceed to complete replacement. Uncoupled remodeling can occur because so much endosteal or trabecular bone is removed that a template for osteoblastic replacement no longer exists. Thus, areas may appear in which trabecular plates develop holes or are converted to attenuated rods. Finally, the ability of successive populations of osteoblasts to complete the formation process is impaired with age, so that haversian canals remain enlarged and trabecular surfaces show only partial replacement of resorbed bone with small BSUs, that is, thinner packets of new bone.

BONE CHEMISTRY AND MINERALIZATION

Bone mineral consists largely of hydroxyapatite [Ca10 (PO4) 6 (OH)2] together with transition forms and other minerals absorbed on the surface. The hydroxyapatite crystals are small and often have lattice defects, although these crystals become more complete as bone matures. The major absorbed minerals are carbonate, magnesium, and sodium. Ninety-nine percent of body calcium, 90% of phosphorus, 80% of carbonate, 80% of citrate, 60% of magnesium, and 35% of sodium are in the skeleton. All these ions may be accessed when deficits exist and stored when excesses occur. Hydrogen ion is generated when hydroxyapatite is formed from circulating calcium and hydrogen phosphate. When a hydrogen ion excess occurs in the extracellular fluid, it can be buffered by demineralization, in which carbonate and

phosphate are released from bone. Many bone-seeking elements exist, such as aluminum, fluoride, lead, and strontium. Deposition of these elements in the skeleton can prevent soft-tissue damage but is likely to alter bone cell function. The organic matrix of bone is made up largely of type I collagen, which consists of tightly coiled, long, triple helical molecules containing two a1 chains and one a2 chain. The collagen molecules are strengthened by covalent intramolecular and intermolecular cross-links and are assembled in ordered fibers. The other collagen types, such as types III, IV, and V, are not deposited in the matrix but are present in interstitial and vascular structures. Type II collagen predominates in cartilage. Although 95% of the matrix is collagen, the noncollagenous components that constitute the remaining 5% are important in providing bone with some of its physical and chemical properties.7 The proteoglycan of bone is of lower molecular weight and is more compact than that in cartilage. Small amounts of proteolipid, which can form complexes with calcium phosphate, are also found. Noncollagen proteins include a calcium-binding, g-carboxyglutamic acidcontaining protein (BGP or osteocalcin) and osteonectin, a highly phosphorylated glycoprotein that binds to collagen and calcium. Osteopontin and bone sialoprotein are highly acidic, have a high affinity for calcium, and have binding sites for integrin receptors. All these proteins are probably important in regulating mineralization, and their distribution may account for the delay between matrix deposition and mineralization. These proteins may also be involved in bone resorption. Osteocalcin is chemotactic for osteoclasts and their precursors; and osteopontin, as well as other proteins, may be involved in the adhesion of osteoclasts to mineralized matrix. Mineralization may also be controlled by cellular elements.8 Matrix vesicles have been identified in calcifying cartilage and fetal bone. They contain cell membrane elements, are rich in alkaline phosphatase, and may initiate mineralization by increasing the local phosphate concentration or providing membrane proteolipids. The concept that phosphatase activity is essential for mineralization has been with us for >50 years, but the precise role of phosphatase is uncertain. Such enzymes may increase the local inorganic phosphate concentration by acting on organic phosphates, or they may cleave pyrophosphate, a potent inhibitor of calcification. The delay between the deposition of matrix by osteoblasts and its mineralization is probably essential for extracellular modifications of matrix, such as collagen cross-linking, formation of large collagen fibers, and deposition of noncollagen proteins, all of which may increase the strength of mineralized tissues. Little evidence exists for direct hormonal control of these steps, but adequate supplies of calcium and phosphate are essential. Thus, animals lacking a vitamin D receptor, given enough calcium and phosphate, show normal mineralization.9 Moreover, little evidence exists that parathyroid hormone (PTH) or calcitonin is essential for mineralization. The more important direct effects of the calcium-regulating hormones appear to be on matrix formation and bone resorption.

CELL BIOLOGY OF BONE

The lineages of bone-forming cells (osteoblasts) and bone-resorbing cells (osteoclasts) probably become separate early in development. Their function is controlled by a complex system of intercellular signals that involve not only systemic calcium-regulating and growth-regulating hormones but also local factors. Bone disease occurs when an imbalance exists between the functions of forming and resorbing cells. Thus, accelerated resorption and diminished formation exacerbate decreases in bone mass in osteoporosis. Excessive bone mass can occur because bone resorption is impaired, as in congenital osteopetrosis, or because formation is excessive and disorderly, as in virally induced avian osteopetrosis and Paget disease. OSTEOBLASTS The osteoblast, a highly specialized bone matrixsynthesizing cell, is derived from precursor cells in the periosteum or the stroma of the bone marrow called determined osteoprogenitor cells.10 Bone can also form in ectopic sites from undifferentiated mesenchymal cells in response to certain inducing agents, particularly demineralized bone matrix. This may be due to the effects of specific proteins produced by bone cells and deposited in matrix.11 Under these conditions, the sequence of endochondral bone formation is recapitulated; that is, cartilage is formed, mineralized, and then replaced by bone. The mature osteoblast is a plump polygonal cell that has an eccentric nucleus, a prominent Golgi apparatus, and abundant rough endoplasmic reticulum (Fig. 50-5). These cells deposit a collagenous matrix and extend cytoplasmic processes into this matrix. As they complete their synthetic activity, they become buried in their own matrix and are called osteocytes. Osteoblasts may also stop producing matrix but remain on the bone surface; these lining cells or resting osteoblasts can be the sites for new remodeling cycles of activation, resorption, and formation. Lining cells may also be reactivated to functional osteoblasts in response to mechanical loading.

FIGURE 50-5. Osteoblasts shown on electron micrograph. A layer of plump cells rich in rough endoplasmic reticulum (RER) and with a large Golgi apparatus (G) is seen forming an osteoid seam consisting of a collagenous matrix that subsequently mineralizes. ~8000 (Courtesy of Dr. Marijke E. Holtrop.)

Osteoblasts produce most of the constituents of bone matrix; however, many proteins are taken up by matrix from the circulation, including a2-HS-glycoprotein and albumin.12 Osteoblasts produce alkaline phosphatase and release it systemically. Hence, serum alkaline phosphatase activity correlates with osteoblastic activity. BGP is also released into the circulation from osteoblasts and provides another marker of their activity.13 When procollagen is converted to collagen, large N- and C-terminal peptides are released. These can be measured in the circulation and reflect the overall rate of collagen synthesis in the body. Osteoblasts have receptors for PTH and 1,25-dihydroxyvitamin D3 [1,25(OH)2 D3] and for systemic growth regulators. 14 They apparently are the major source of the autocrine factors that regulate local bone turnover, including cytokines, prostaglandins, and bone-derived growth factors. Osteoblasts may trigger bone resorption in the remodeling process. This activation step may involve shape changes or secretion of collagenase and related metalloproteinases as well as other proteolytic enzymes, such as plasminogen activator.15 These changes may enhance the access of osteoclasts to the mineralized bone surface. The concept that cells of the osteoblast lineage are critical in activating the resorption phase of the bone remodeling cycle has long been recognized, and some of the specific proteins involved in this interaction are known.16,17 and 18 The ligands that regulate the osteoblast-osteoclast interaction are not specific for these cells but are also involved in lymphoid activation and may be produced by many other cells types. The activator protein, which is probably expressed in stromal precursor cells in the marrow, has been termed osteoclast differentiating factor (ODF) or osteoprotegerin ligand (OPGL) and is identical to the protein in the lymphoid series termed TRANCE. Its receptor on preosteoclasts is identical with another membrane protein on lymphoid cells called RANK. This interaction can be blocked by osteoprotegerin (OPG), which binds to OPGL. The importance of this regulatory system has been demonstrated experimentally. The addition of synthetic ODF to cultures of preosteoclasts can replace cells of the osteoblast lineage in promoting differentiation and osteoclast formation. Animals in whom OPG has been knocked out show severe osteoporosis, due to increased bone resorption.19 Once the osteoblasts have completed their allotted portion of matrix synthesis, they can undergo one of three fates. Some cells become buried as osteocytes. These cells remain connected to the overlying osteoblast and to each other by cell processes enclosed in canaliculi within the mineralized bone. This syncytium of osteocytes and osteoblasts is probably critical for the signaling of the response to mechanical forces. One mechanism is the fluid shear-stress produced by small strains on the bone in the canaliculi and osteocyte lacunae, which can generate cellular responses, such as increased nitric oxide and prostaglandin production. A few osteoblasts remain on the inactive cell surface and spread out as thin lining cells. Finally, osteoblasts can undergo programmed cell death or apoptosis.20 This process may be regulated by systemic hormones and local factors. A fourth possibility exists, namely, that osteoblasts can dedifferentiate and return to the marrow stoma as precursor cells, but this has not been demonstrated experimentally. OSTEOCLASTS Osteoclasts resorb bone and calcified cartilage. These large multinucleated cells are formed by the fusion of mononuclear precursors. Osteoclasts presumably are derived from a hematopoietic stem cell rather than from the mesenchymal precursor of the osteoblast. The osteoclast cell line is related to the monocyte-macrophage

lineage, but many of the surface markers for macrophages are missing from osteoclasts, and the osteoclast and monocyte precursor cell lines probably separate early in differentiation.21 The stimulation of osteoclast precursor replication may be an important mechanism for increasing bone resorption. Once fusion has occurred, however, the nuclei in an osteoclast do not undergo further cell division. The osteoclast cytoplasm contains abundant mitochondria, many lysosomes, and relatively little rough endoplasmic reticulum. The unique feature of the osteoclast is the ruffled border, which is the site of active resorption (Fig. 50-6). This is surrounded by a clear or sealing zone, which functions to attach the osteoclast to bone and isolate the ruffled border from extracellular fluid so that a high local concentration of hydrogen ions and lysosomal enzymes can be maintained. Osteoclasts are rich in acid phosphatase as well as other lysosomal enzymes, and in carbonic anhydrase, which facilitates hydrogen ion secretion. Osteoclasts also produce large amounts of cathepsin K, a relatively selective lysosomal enzyme that can degrade all components of bone matrix, including collagen, at a low pH.22 Thus, the ruffled border area simulates a giant exteriorized phagolysosome in which hydrogen ions are important not only in mobilizing mineral but also in activating lysosomal enzymes.

FIGURE 50-6. Resorbing apparatus of osteoclasts shown on electron micrograph. A section of an osteoclast with a highly infolded ruffled border is seen between two areas relatively devoid of subcellular particles, termed the clear or sealing zone. Large vacuoles are present in the osteoclast cytoplasm, and mitochondria are abundant. ~8000 (Courtesy of Dr. Marijke E. Holtrop.)

The proton pump that transports hydrogen ions into the ruffled border area is similar but not identical to the vacuolar proton pump found in lysosomes and kidney cells. Other important features of the osteoclast are the cell attachment apparatus, which involves vitronectin receptors that can bind a wide variety of proteins containing Arg-Gly-Asp sequences.23 The osteoclast may also have calcium-sensing receptors that mediate the inhibition of osteoclastic activity by high calcium concentrations. Osteoclast precursors express RANK; fully differentiated osteoclasts have not been shown to express RANK, but do respond to OPG, perhaps through some other binding protein.24 Another critical factor for the formation of osteoclasts is colony-stimulating factor-1 (CSF-1) or macrophage colony-stimulating factor (M-CSF), which may play a role in the replication and differentiation of precursors.25 Animals lacking CSF-1 have osteopetrosis. OTHER CELL TYPES IN BONE Macrophages may be found at resorption sites after the initial removal of bone osteoclasts. Their function is uncertain, but they may remove residual matrix that has not been completely digested, because macrophages can secrete collagenase as well as lysosomal enzymes. Macrophages may also be a source of interleukin-1 (IL-1) and prostaglandin E2 (PGE2), which not only are potent stimulators of bone resorption but also can stimulate the replication of osteoblast precursors and may be involved in initiating the formation phase of the remodeling cycle. Lymphocytes may also play a role by secreting bone-resorbing factors.26 Moreover, calcium-regulating hormones can act on these hematopoietic cells. For example, 1,25(OH)2 D3 increases the differentiation of monocyte precursors into macrophages. Fibroblastic cells may also play a role by secreting local regulators. Insulin-like growth factor-I (IGF-I), which can be produced by fibroblasts, is a potent stimulator of bone growth. Mast cells are found adjacent to resorbing bone and can produce heparin, which enhances bone resorption in some culture systems27 (see Chap. 181). Finally, endothelial cells and nerve endings may play a role by producing neuropeptides and growth factors.28 Endothelial cells are also a major source of prostaglandins and nitric oxide, which can influence bone resorption and formation as well as affect bone blood flow.29

EFFECTS OF HORMONES ON BONE CELLS

Despite many studies of the direct effects of hormones on bone, the complex regulation of bone resorption and formation still is inadequately explained. The effects of systemic agents are probably modulated by interactions with local intercellular mediators. For example, PTH is clearly a potent stimulator of bone resorption both in vivo and in vitro but has not been shown to act on isolated osteoclasts in the absence of other bone cells. Prostaglandins are also potent stimulators of bone resorption, but they decrease motility and resorptive activity of isolated osteoclasts, as does calcitonin. The major factors that influence bone metabolism and their most important direct effects are listed in Table 50-1.

TABLE 50-1. Systemic and Local Regulation of Bone Metabolism

PARATHYROID HORMONE Although PTH was first shown to act directly on bone as a stimulator of resorption, much more is now known about its effects on osteoblasts at the cellular and molecular levels.14,30,31,32,33,34 and 35 Much of this information has been derived from investigations of isolated bone cells or cloned osteosarcoma cells that have an osteo-blastic phenotype, as well as from in vivo studies. The first effect of PTH on osteoblast-like cells is an activation of adenylate cyclase and protein kinase A. PTH can also accelerate phosphatidylinositol turnover, which leads to increased intracellular calcium and activation of protein kinase C. PTH causes rapid changes in osteoblast cell shape associated with polymerization of actin. A subsequent decrease in collagen synthesis is associated with a diminution of procollagen messenger RNA (mRNA) levels in the cell. Release of metalloproteinases from osteoblasts is also increased in response to PTH. Plasminogen activator activity is increased, in part by decreased production of an inhibitor, and this may result in the activation of latent collagenase. Alkaline phosphatase levels are usually decreased. PTH may also decrease BGP synthesis. PTH can increase prostaglandin synthesis and cell replication in bone cell and organ cultures.36 Most of these effects of PTH on bone can be considered catabolic; but prolonged, intermittent administration of low doses of PTH can elicit an anabolic effect, with increased bone mass.37 This is the basis for the use of intermittent PTH administration to treat osteoporosis. The anabolic response may be due to stimulation of precursor cell replication by PTH or release of growth factors from bone cells or matrix. Although the ability of PTH to stimulate bone resorption has been recognized for many years, the mechanism is poorly understood. In vivo, PTH rapidly increases the number and activity of osteoclasts as measured by the extent of ruffled border area. Increased release of calcium and matrix constituents of bone can be measured within a few hours, but increases in lysosomal enzyme release can be detected within minutes. PTH stimulation of bone resorption presumably is receptor mediated. The role of cyclic adenosine monophosphate (cAMP) in mediating this response is

controversial.32,34 Stimulators of adenylate cyclase activity and phosphodiesterase inhibitors can enhance bone resorption, but their effect is generally smaller than that of PTH. Moreover, under certain conditions, these agents inhibit bone resorption, mimicking the action of calcitonin. Possibly, the stimulation of bone resorption involves multiple pathways, including activation of phospholipase C to release phosphatidylinositol and diacylglycerol. This results in increased cell calcium and activation of protein kinase C. The major PTH receptor has been cloned. It is present in osteoblasts and their precursors but has not been demonstrated in mammalian osteoclasts. This single receptor can mediate increases in cAMP, phosphatidylinositol breakdown, and intracellular calcium. The receptor can be activated either by 1-34 PTH or by 1-34 PTHrelated protein (PTHrP).31 PTH and PTHrP, like other stimulators of bone resorption, can increase the expression of ODF and may also decrease the expression of OPG in osteoblastic cells.38 Another PTH receptor has been identified with greater affinity for PTH than for PTHrP,39,40 but its physiologic role is not known. 1,25-DIHYDROXYVITAMIN D3 The major physiologic role of the hormonal form of vitamin D, 1,25(OH)2 D3 or calcitriol, is to promote intestinal absorption of calcium and phosphorus. It is also one of the most potent hormones acting on bone. Although the effects of 1,25(OH)2 D3 on the intestine promote skeletal growth and mineralization, the effects of high concentrations on bone appear to be catabolic, stimulating resorption and inhibiting formation. As with PTH, receptors for 1,25(OH)2 D3 have been demonstrated in osteoblastic cells. The hormone resembles PTH in inhibiting collagen synthesis but differs from PTH in its selective ability to increase osteocalcin production. Moreover, 1,25 (OH)2 D3 does not show the same marked anabolic effect as PTH, although it may prevent bone loss after ovariectomy in rats.41 Presumably 1,25(OH)2 D3 acts largely by a classic steroid hormone pathway, through a nuclear receptor that binds to chromatin.42 This receptor has been demonstrated in osteoblasts and cells and can mediate such transcriptional effects as the increase in mRNA for BGP and the decrease in mRNA for collagen. A nongenomic effect of 1,25(OH)2 D3 may also be present, based on the finding of a binding protein for the hormone in plasma membrane and matrix vesicles, which appears to activate protein kinase C in intestinal cells and chondrocytes.43 Although 1,25(OH)2 D3 stimulates bone resorption, receptors have not been demonstrated in isolated osteoclasts, and the effect is likely to be mediated through osteoblasts as it is for other bone resorbers. The 1,25(OH)2 D3 also stimulates the production of multinucleated cells with an osteoclastic phenotype in bone marrow and spleen cell cultures when cells of the osteoblastic lineage are present. Resorption occurs at such low concentrations in organ culture that this effect can be used as a serum bioassay.44 However, this assay requires extraction of the serum and, therefore, measures 1,25(OH)2 D3 that would normally circulate bound to vitamin Dbinding protein. Hence, the concentrations that stimulate bone resorption in organ culture are probably at least 10-fold higher than the normal concentration of free hormone in the blood and extracellular fluid. When calcium and phosphate deprivation is severe, such high concentrations of 1,25(OH)2 D3 may be achieved that vitamin D is unable to maintain calcium and phosphorus levels by increasing intestinal absorption but must draw on the skeletal reservoir. Although substantial evidence exists that 1,25(OH)2 D3 is the bioactive form of vitamin D, some evidence is found for effects of other metabolites on skeletal tissue. The metabolite 25-hydroxy-vitamin D can stimulate bone resorption and inhibit bone formation at high concentrations, which may occur in vivo when toxic doses are given. The metabolite 24,25-dihydroxyvitamin D is formed by an alternative hydroxylation pathway in the kidney. This probably represents an inactivation process, but evidence exists that 24,25-dihydroxyvitamin D has anabolic effects, particularly on cartilage. Not only does 1,25(OH)2 D3 have important direct effects on bone cells but it also may act on bone indirectly through its function as an immunomodulator.45 The physiologic importance of these actions remains uncertain (see Chap. 195). CALCITONIN Calcitonin is a potent direct inhibitor of osteoclastic activity (see Chap. 53). In contrast to PTH and 1,25(OH) 2 D3, receptors for calcitonin have been identified on osteoclasts.39 Moreover, isolated osteoclasts show a decrease in motility and resorptive activity when treated with this hormone. Calcitonin rapidly decreases the amount of active ruffled border of osteoclasts in organ cultures, although the cells may remain attached to the bone surfaces by their clear zones. In vivo, the number of osteoclasts decreases as the cells appear to migrate away from the bone surface. Calcitonin increases cAMP content in cell populations enriched with osteoclasts. Cyclic AMP is the mediator of inhibition of bone resorption, because agents that increase the cAMP concentration can mimic the action of calcitonin both in organ culture and in isolated osteoclasts. However, evidence also exists for an inhibitory pathway mediated by protein kinase C.46 The direct inhibition of osteoclastic activity by calcitonin is transient in isolated cell systems, in organ cultures, and in patients with hyperparathyroidism or hypercalcemia of malignancy. This escape phenomenon may be due to down-regulation of the calcitonin receptors.39 Although calcitonin has been used clinically as an inhibitor of bone resorption, its physiologic role is probably limited. Patients with medullary carcinoma of the thyroid who have extremely high blood levels of calcitonin have normal bone turnover. A number of peptides with homology to calcitonin may also affect bone metabolism. These include calcitonin gene-related peptide, amylin, and adrenomedullin.47 The importance of these peptides in bone physiology is unknown; however, they have been shown to affect both bone resorption and bone formation in animal and in vitro models.48

SYSTEMIC HORMONES THAT AFFECT BONE METABOLISM

Many hormones that regulate somatic growth act directly or indirectly on the skeleton. These hormones not only modulate physiologic skeletal growth and development but also are important in the pathogenesis of metabolic bone disease. GLUCOCORTICOIDS Glucocorticoids have complex direct and indirect effects on skeletal tissue. The major indirect effect is the inhibition of calcium absorption in the intestine. This may lead to secondary hyperparathyroidism, which could explain the clinical observation that bone resorption is increased in some patients treated with glucocorticoids. Nevertheless, the major direct effect on bone is a dose-related decrease in formation.49 This is probably mediated by a decrease in the replication and differentiation of osteoblast precursors. However, glucocorticoids can also have a positive effect on osteoblast function. In some cell and organ cultures, physiologic concentrations of glucocorticoids can increase collagen synthesis, increase alkaline phosphatase levels, and enhance the response to other hormones. These effects may be due to an increase in osteoblast differentiation.14 Glucocorticoids can have both stimulatory and inhibitory effects on bone resorption. Stimulation may be due to enhanced osteoclast differentiation as well as secondary hyperparathyroidism. Inhibition of resorption may be due to decreased replication of osteoclast precursors or decreased production of bone-resorbing factors by osteoblasts, such as prostaglandins and interleukins. Whatever the mechanisms, hypercalcemia has been observed in adrenal insufficiency, and decreased bone mass is an important adverse effect of glucocorticoid excess (see Chap. 59 and Chap. 64). GROWTH HORMONE AND INSULIN-LIKE GROWTH FACTORS Excesses and deficiencies of growth hormone are associated with increases and decreases in skeletal growth. This effect is probably mediated by IGF-I (see Chap. 12 and Chap. 173).50 A major source of IGF is the liver, but many other tissues, including skeletal tissue, can produce IGF-I and also IGF-II, which is not under growth hormone control. Moreover, growth hormone can stimulate IGF-I production by bone cells. IGF-I and IGF-II have pleiotropic effects, stimulating bone cell replication as well as collagen and noncollagen protein synthesis. The relative importance of IGF-I and IGF-II in regulating bone growth and in metabolic bone disease is uncertain. Possibly, the decrease in bone mass seen in malnutrition and gastrointestinal disorders is related to decreased production of these factors. An age-related decrease in growth hormone and IGF-I secretion may play a role in bone loss. Bone tissue produces not only IGF-I and IGF-II but also a number of IGF-binding proteins that can be both inhibitory and stimulatory. The regulation of these binding proteins may be as important as the regulation of IGFs themselves in the local control of bone formation.51 INSULIN Insulin is an important regulator of somatic growth. At physiologic concentrations, insulin appears to stimulate osteoblast function, selectively increasing collagen synthesis without affecting cell replication or protein synthesis in the periosteum.14 At higher concentrations, insulin can produce a pleiotropic effect, perhaps by acting

on the IGF-I receptor. Insulin may be important in skeletal development. Bone mass may be increased by hyperinsulinism in the infants of diabetic mothers and decreased in diabetic children with insulin deficiency. THYROID HORMONES Thyroid hormones (thyroxine and triiodothyronine) are essential for maintaining skeletal growth and remodeling. They not only increase cartilage growth directly but also probably have a positive interaction with IGF-I. Thyroid hormones can increase bone turnover, an effect apparently due to direct stimulation of both bone resorption and formation.52,53a Decreased bone mass has been observed in hyperthyroid patients, including those who received large doses of exogenous thyroid hormones for long periods.53 SEX HORMONES Perhaps the greatest limitation in our understanding of skeletal physiology is that we do not know precisely how sex hormones act on bone. Androgens and estrogens are involved in the pubertal growth spurt and the maintenance of bone mass. Some of this may be due to changes in muscle mass, which then influence bone. The accelerated bone loss that occurs with estrogen withdrawal at menopause can be attributed to an increase in bone resorption. This has led to the concept that estrogens oppose the resorptive activity of PTH or other stimulators of osteoclastic activity. Both estrogen and androgen receptors have been identified in bone cells, particularly those of the osteoblastic lineage. The sex hormones change the production of cytokines, prostaglandins, and growth factors by bone cells.36,54,55 In rodent models, inhibition of IL-1 or tumor necrosis factor-a (TNF-a) can reverse the bone loss that follows ovariectomy. Moreover, ovariectomy does not cause bone loss in animals in whom the IL-1 receptor has been knocked out by homologous recombination. However, the relative importance of these effects in mediating the changes seen with sex hormone deficiency in humans remains to be determined. Experiments of nature have helped us to understand the relative importance of estrogen and androgen.56,56a An abnormal skeletal phenotype has been defined in a man with loss of function due to a mutation in the estrogen receptor and in two men with aromatase deficiency who cannot convert testosterone to estrogen. These individuals showed failure of epiphyseal closure, high bone turnover, and decreased bone mass. In the aromatase-deficient men, treatment with estrogen could reverse these abnormalities. Because androgens are a source for estrogens, defining their specific role in bone is difficult. In women with androgen insensitivity, skeletal development is normal, although bone mineral density may be low.57 However, estrogen may also play a role here, either because replacement has been inadequate or because a critical role is played by estrogen produced locally in bone by aromatase, which is expressed in human bone cells. OTHER SYSTEMIC HORMONES In view of the many hormonal influences on the skeleton identified in the past few decades, additional effects of systemic hormones on skeletal function are likely to be discovered. Among the hormones that have been considered as possible skeletal regulators are progesterone,58 prolactin,59 and neuropeptides.60,61 LOCAL REGULATORS The ability of the skeleton to respond to local forces must depend on the existence of local regulators; moreover, these regulators are probably important in mediating the responses of systemic hormones as previously noted. PROSTAGLANDINS Prostaglandins are ubiquitous local modulators of cell function (see Chap. 172). Their role in skeletal physiology was first suggested by the observations that PGE2 can raise the cAMP concentration and stimulate bone resorption in vitro. Locally, prostaglandin production in bone can be stimulated by mechanical stress, by cytokines associated with inflammation and injury, and by growth factors. Systemically, PTH can stimulate and glucocorticoids can inhibit prostaglandin production in bone. The effects of PGE 2 on bone metabolism are complex and biphasic.36 PGE2 can stimulate bone cell replication and differentiation and increase bone formation in vivo. At high concentrations in vitro, inhibition of osteoblastic collagen synthesis occurs. The ability of bone to respond to mechanical forces may be prostaglandin dependent.62 Stimulation of bone resorption by PGE2 is important in mediating bone loss in inflammation and immobilization, and might play a role in osteoporosis. BONE-DERIVED GROWTH FACTORS Two major families of growth factors are produced in bone and deposited in bone matrix: (a) the IGFs, including the IGF-binding proteins,50 and (b) the transforming growth factor-b and bone morphogenetic protein family.63,64 The latter group contains at least 10 different proteins that are related to other regulatory peptides, including the activin-inhibins and invertebrate growth factors. OTHER GROWTH FACTORS Epidermal, fibroblast, and platelet-derived growth factors can all act on skeletal tissue (see Chap. 173). These factors can be derived from bone cells or from adjacent hematopoietic or vascular tissue. They can stimulate bone resorption by either prostaglandin-dependent or prostaglandin-independent mechanisms.65 Their mitogenic effects may be associated with decreased collagen synthesis in acute experiments in vitro, but prolonged treatment may increase bone formation in vivo. CYTOKINES Both bone cells and adjacent hematopoietic cells can produce cytokines, which profoundly influence bone metabolism54 (see Chap. 173 and Chap. 212). Cytokine-mediated bone resorption was first identified as being due to osteoclast-activating factor, produced by mitogen- or antigen-stimulated human leukocyte cultures. Subsequently, IL-1 was found to be the major bone-resorbing factor from macrophages. TNF is also produced by leukocytes and is an active bone resorber. These cytokines can stimulate bone resorption by both prostaglandin-dependent and prostaglandin-independent mechanisms and can inhibit collagen synthesis in osteoblasts. IL-6 is less potent as a direct stimulator of resorption but can synergize with IL-1 to increase bone resorption and prostaglandin production.66 Some cytokines, such as IL-4, may inhibit bone resorption.67 Both IL-1 and IL-6 have been implicated as mediators of the increased bone resorption that occurs at menopause based largely on findings from rodent models.54,68 Cytokines may mediate inflammatory bone loss and the hypercalcemia of multiple myeloma and other lymphoproliferative disorders.69,70 IONS AS REGULATORS Calcium, phosphate, and other ions important in bone metabolism not only are involved in feedback control of calcium-regulating hormone secretion but also can act as direct regulators. In addition to the critical role played by the supply of calcium and phosphorus in mineralization, organ culture studies suggest that both calcium and phosphate can influence the rate of matrix formation.14 The effect of calcium is nonspecific in that this ion is required for cell growth generally. High serum phosphate concentrations are associated with rapid rates of bone growth. For example, in humans, phosphate concentrations are higher in the first year of life and at puberty, when the relative rates of skeletal growth are fastest. PTH secretion is inhibited and calcitonin secretion stimulated by calcium, whereas phosphate loading can lower ionized calcium and, thus, stimulate PTH secretion. Both calcium and phosphate have direct inhibitory effects on 1a-hydroxylase in the kidney and decrease 1,25(OH)2 D3 levels. High local concentrations of calcium, which develop during resorption in the ruffled border area, can cause loss of activity and detachment of osteoclasts. Phosphate may inhibit bone resorption directly by a physicochemical effect on mineral dissolution. Magnesium has complex effects on skeletal metabolism (see Chap. 68). High concentrations can inhibit mineralization and PTH secretion, whereas severe magnesium depletion also is associated with decreased PTH secretion and decreased hormone responsiveness. Changes in hydrogen ion concentration can affect bone mineralization and demineralization. The increased hydrogen ion concentration in the ruffled border area is necessary not only for removal of mineral but also for maximal activity of the lysosomal enzymes that resorb matrix. Decreasing the hydrogen ion supply by inhibiting the proton pump, blocking carbonic anhydrase, or interfering with chloride-bicarbonate exchange all can inhibit bone resorption. CHAPTER REFERENCES
1. Turner CH. Three rules for bone adaptation to mechanical stimuli. Bone 1998; 23:339.

2. McLeod KJ, Rubin CT, Otter MW, Qin YX. Skeletal cell stresses and bone adaptation. Am J Med Sci 1998; 316:176. 3. Gorski JP. Is all bone the same? Distinctive distributions and properties of non-collagenous matrix proteins in lamellar vs. woven bone imply the existence of different underlying osteogenic mechanisms. Crit Rev Oral Biol Med 1998; 9:201. 4. Eriksen EF. Normal and pathological remodeling of human trabecular bone: three dimensional reconstruction of the remodeling sequence in normals and in metabolic bone disease. Endocrinol Rev 1986; 7:379. 5. Burr DB, Turner CH, Niack P, et al. Does microdamage accumulation affect the mechanical properties of bone? J Biomech 1998; 31:337. 6. Hui SL, Johnston CC Jr, Mazess RB. Bone mass in normal children and young adults. Growth 1985; 49:34. 7. Robey PG. Vertebrate mineralized matrix proteins: structure and function. Connect Tissue Res 1996; 35:131. 8. Hsu HHT, Anderson HC. Evidence of the presence of a specific ATPase responsible for ATP-initiated calcification by matrix vesicles isolated from cartilage and bone. J Biol Chem 1996; 271:26383. 9. Li YC, Amling M, Pirro AE, et al. Normalization of mineral ion homeostasis by dietary means prevents hyperparathyroidism, rickets, and osteomalacia, but not alopecia in vitamin D receptor-ablated mice. Endocrinology 1998; 139:4391. 10. Friedenstein AJ, Latzinik NV, Gorskaya YF, et al. Bone marrow stromal colony formation requires stimulation by haemopoietic cells. Bone Miner 1992; 18:199. 11. Raval P, Hsu HH, Schneider DJ, et al. Expression of bone morphogenetic proteins by osteoinductive and non-osteoinductive human osteosarcoma cells. J Dent Res 1996; 75:1518. 12. Delmas PD, Tracy RP, et al. Identification of the noncollagenous proteins of bovine bone by two-dimensional gel electrophoresis. Calcif Tissue Int 1984; 36:308. 13. Garnero P, Delmas PD. Biochemical markers of bone turnoverapplications for osteoporosis. Endocrinol Metab Clin North Am 1998; 27:303. 14. Raisz LG, Kream BE. Medical progress: regulation of bone formation. N Engl J Med 1983; 309:29. 15. Allan EF, Martin TJ. The plasminogen activator inhibitor system in bone cell function. Clin Orthop 1995; 313:54. 16. Yasuda H, Shima N, Nakagawa N, et al. Osteoclast differentiation factor is a ligand for osteoprotegerin/osteoclastogenesis-inhibitory factor and is identical to TRANCE/RANKL. Proc Natl Acad Sci U S A 1998; 95:3597. 17. Quinn JMW, Elliott J, Gillespie MT, Martin TJ. A combination of osteoclast differentiation factor and macrophage-colony stimulating factor is sufficient for both human and mouse osteoclast formation in vitro. Endocrinology 1998; 139:4424. 18. Fuller K, Wong B, Fox S, et al. TRANCE is necessary and sufficient for osteoblast-mediated activation of bone resorption in osteoclasts. J Exp Med 1998; 188:997. 19. Mizuno A, Amizuka N, Irie K, et al. Severe osteoporosis in mice lacking osteoclastogenesis inhibitory factor/osteoprotegerin. Biochem Biophys Res Commun 1998; 147:610. 20. Jilka RL, Weinstein RS, Bellido T, et al. Osteoblast programmed cell death (apoptosis): modulation by growth factors and cytokines. J Bone Miner Res 1998; 13:793. 21. Teitelbaum SL, Tondravi MM, Ross FP. Osteoclasts, macrophages, and the molecular mechanisms of bone resorption. J Leukoc Biol 1997; 61:381. 22. Garnero P, Borel O, Byrjalsen I, et al. The collagenolytic activity of cathepsin K is unique among mammalian proteinases. J Biol Chem 1998; 48:32347. 23. Engleman VW, Nickols GA, Ross FP, et al. A peptidomimetic antagonist of the alpha (v) beta3 integrin inhibits bone resorption in vitro and prevents osteoporosis in vivo. J Clin Invest 1997; 99:2284. 24. Hakeda Y, Kobayashi Y, Yamaguchi K, et al. Osteoclastogenesis inhibitory factor (OCIF) directly inhibits bone-resorbing activity of isolated mature osteoclasts. Biochem Biophys Res Commun 1998; 251:796. 25. Flanagan AM, Lader CS. Update on the biologic effects of macrophage colony-stimulating factor. Curr Opin Hematol 1998; 5:181. 26. Miyaura C, Onoe Y, Inada M, et al. Increased B-lymphopoiesis by interleukin 7 induces bone loss in mice with intact ovarian function: similarity to estrogen deficiency. Proc Natl Acad Sci U S A 1997; 94:9360. 27. Nakamura M, Kuroda H, Narita K, Endo Y. Parathyroid hormone induces a rapid increase in the number of active osteoclasts by releasing histamine from mast cells. Life Sci 1996; 58:1861. 28. Konttinen Y, Imai S, Suda A. Neuropeptides and the puzzle of bone remodeling. State of the art. Acta Orthop Scand 1996; 67:632. 29. Chow JWM, Fox SW, Lean JM, Chambers TJ. Role of nitric oxide and prostaglandins in mechanically induced bone formation. J Bone Miner Res 1998; 13:1039. 30. Tetradis S, Nervina JM, Nemoto K, Kream BE. Parathyroid hormone induces expression of the inducible cAMP early repressor in osteoblastic MC3T3-E1 cells and mouse calvariae. J Bone Miner Res 1998; 13:1846. 31. Lanske B, Divieti P, Kovacs CS, et al. The parathyroid hormone (PTH)/PTH-related peptide receptor mediates actions of both ligands in murine bone. Endocrinology 1998; 139:5194. 32. Takasu H, Bringhurst FR. Type-1 parathyroid hormone (PTH) PTH-related peptide (PTHrP) receptors activate phospholipase C in response to carboxyl-truncated analogs of PTH (1-34). Endocrinology 1998; 139:4293. 33. Huang YF, Harrison JR, Lorenzo JA, Kream BE. Parathyroid hormone induces interleukin-6 heterogeneous nuclear and messenger RNA expression in murine calvarial organ cultures. Bone 1998; 23:327. 34. Schwindinger WF, Fredericks J, Watkins L, et al. Coupling of the PTH/PTHrP receptor to multiple G-proteinsDirect demonstration of receptor activation of G s , Gq11, and Gi (1) by {alpha-P-32} GTP-gamma-azidoanilide photoaffinity labeling. Endocrine 1998; 201. 35. Torrungruang K, Feister H, Swartz D, et al. Parathyroid hormone regulates the expression of the nuclear mitotic apparatus protein in the osteoblast-like cells, ROS 17/2.8. Bone 1998; 22:317. 36. Kawaguchi H, Pilbeam CC, Harrison JR, Raisz LG. The role of prostaglandins in the regulation of bone metabolism. Clin Orthop 1995; 313:36. 37. Dempster DW, Cosman F, Parisien M, et al. Anabolic actions of parathyroid hormone on bone. Endocr Rev 1993; 14:690. 38. Horwood NJ, Elliott J, Martin TJ, Gillespie MT. Osteotropic agents regulate the expression of osteoclast differentiation factor and osteoprotegerin in osteoblastic stromal cells. Endocrinology 1998; 139:4743. 39. Findlay DM, Martin TJ. Receptors of calciotropic hormones. Horm Metab Res 1997; 29128. 40. Turner PR, Mefford S, Bambino T, Nissenson RA. Transmembrane residues together with the amino terminus limit the response of the parathyroid hormone (PTH) 2 receptor to PTH-related peptide. J Biol Chem 1998; 273:3830. 41. Erben RG, Bromm S, Stangassinger M. Therapeutic efficacy of 1a, 25-dihydroxyvitamin D3 and calcium in osteopenic ovariectomized rats: evidence for direct anabolic effect of 1a, 25-dihydroxyvitamin D3 on bone. Endocrinology 1998; 139:4319. 42. Haussler MR, Whitfield GK, Haussler CA. The nuclear vitamin D receptor: biological and molecular regulatory properties revealed. J Bone Miner Res 1998; 13:325. 43. Nemere I, Schwatz Z, Pedroszo H, et al. Identification of a membrane receptor for 1,25-dihydroxyvitamin D 3 which mediates rapid activation of protein kinase C. J Bone Miner Res 1998; 13:1353. 44. Stern PH, Phillips TE, Mavreas T. Bioassay of 1,25-dihydroxyvitamin D in human plasma purified by partition, alkaline extraction, and high-pressure chromatography. Anal Biochem 1980; 102:22. 45. Morgan JW, Sliney DJ, Morgan DM, Maizel AL. Differential regulation of gene transcription in subpopulations of human B lymphocytes by vitamin D-3. Endocrinology 1999; 140:381. 46. Moonga BS, Dempster DW. Effects of peptide fragments of protein kinase C on isolated rat osteoclasts. Exp Physiol 1998; 83:717. 47. Cooper GJS. Amylin compared with calcitonin gene-related peptidestructure, biology, and relevance to metabolic disease. Endocrin Rev 1994; 15:163. 48. Cornish J, Callon KE, King AR. Systemic administration of amylin increases bone mass, linear growth, and adiposity in adult male mice. Am J Physiol 1998; 38:E694. 49. Advani S, LaFrancis D, Bogdanovic E, et al. Dexamethasone suppresses in vivo levels of bone collagen synthesis in neonatal mice. Bone 1997; 20:41. 50. Rosen CJ, Donahue LR. Insulin-like growth factors and bonethe osteoporosis connection revisited. Proc Soc Exp Biol Med 1998; 219:1. 51. Hakeda Y, Kawaguchi H, Hurley M, et al. Intact insulin-like growth factor binding protein-5 (IGFBP-5) associates with bone matrix and the soluble fragments of IGFBP-5 accumulated in culture medium of neonatal mouse calvariae by parathyroid hormone and prostaglandin E2-treatment. J Cell Physiol 1996; 166:370. 52. Kawaguchi H, Pilbeam CC, Raisz LG. Anabolic effects of 3,3',5-triiodothyronine and triiodothyroacetic acid in cultured neonatal mouse parietal bones. Endocrinology 1994; 135:971. 53. Derosa G, Testa A, Giacomini D, et al. Prospective study of bone loss in pre- and postmenopausal women on L-thyroxine therapy for non-toxic goitre. Clin Endocrinol 1997; 47:529. 53a. Pantazi H, Papapetrou PD. Changes in pararmeters of bone and mineral metabolism during therapy for hyperthyroidism. J Clin Endocrinal Metab 2000; 85:1099. 54. Pacifici R, Cytokines, estrogen, and postmenopausal osteoporosisthe second decade. Endocrinology 1998; 139:2659. 55. Srivastava S, Weitzmann MN, Kimble RB, et al. Estrogen blocks M-CSF gene expression and osteoclast formation by regulating phosphorylation of Egr-1 and its interaction with Sp-1. J Clin Invest 1998; 102:1850. 56. Bilezikian JP, Morishima A, Bell J, Grumbach MM. Increased bone mass as a result of estrogen therapy in a man with aromatase deficiency. N Engl J Med 1998; 339:599. 56a. Takagi M, Miyashita Y, Koga M, et al. Estrogen deficiency is a potential case for osteopenia in adult male patients with Noonan's syndrome. Calcif Tissue Int 2000; 66:200. 57. Mizunuma H, Soda M, Okano H, et al. Changes in bone mineral density after orchidectomy and hormone replacement therapy in individuals with androgen insensitivity syndrome. Hum Reprod 1998; 13:2816. 58. Manzi DL, Pilbeam CC, Raisz LG. The anabolic effects of progesterone on fetal rat calvaria in tissue culture. J Soc Gynecol Invest 1994; 1:302. 59. Clement-Lacroix P, Ormandy C, Lepescheux L, et al. Osteoblasts are a new target for prolactin: analysis of bone formation in prolactin receptor knock-out mice. Endocrinology 1999; 140:96. 60. Konttinsen Y, Imai S, Suda A. Neuropeptides and the puzzle of bone remodeling. State of the art. Acta Orthop Scand 1996; 67:632. 61. Togari A, Arai M, Mizutani S, et al. Expression of MRNAs for neuropeptide receptors and beta-adrenergic receptors in human osteoblasts and human osteogenic sarcoma cells. Neurosci Lett 1997; 233:125. 62. Klein-Nulend J, Berger EH, Semeins CM, et al. Pulsating fluid flow stimulates prostaglandin release and inducible prostaglandin G/H synthase mRNA expression in primary mouse bone cells. J Bone Miner Res 1997; 12:45. 63. Bonewald LF, Dallas SL. Role of active and latent transforming growth factor beta in bone formation. J Cell Biochem 1994; 55:350. 64. Sakou T. Bone morphogenetic proteins: from basic studies to clinical approaches. Bone 1998; 22:591. 65. Zhang ZM, Chen JT, Jin D. Platelet-derived growth factor (PDGF)BB stimulates osteoclastic bone resorption directly: the role of receptor beta. Biochem Biophys Res Commun 1998; 251:190. 66. Tai H, Miyaura C, Pilbeam CC, et al. Transcriptional induction of cyclooxygenase in osteoblasts is involved in interleukin-6-induced osteoclast formation. Endocrinology 1997; 138:2372. 67. Kawaguchi H, Nemoto K, Raisz LG, et al. Interleukin-4 inhibits prostaglandin G/H synthase-2 and cytosolic phospholipase A2 induction in neonatal mouse parietal bone cultures. J Bone Miner Res 1996; 11:358. 68. Papanicolaou DA, Wilder RL, Manolagas SC, Chrousos GP. The pathophysiologic roles of interleukin-6 in human disease. Ann Intern Med 1998; 128:127. 69. Daroszewska D, Bucknall RC, Chu P, Frazier WD. Severe hylpercalcemia in B-cell lymphoma: combined effects of PTH-RP, IL-6 and TNF. Postgrad Med J 1999; 75:672. 70. Trico G. New insights into role of microenvironment in multiple myeloma. Lancet 2000; 355:248.

CHAPTER 51 PARATHYROID HORMONE Principles and Practice of Endocrinology and Metabolism

CHAPTER 51 PARATHYROID HORMONE

DAVID GOLTZMAN AND GEOFFREY N. HENDY Biosynthesis of Parathyroid Hormone Molecular Biology of Parathyroid Hormone Characteristics of the Normal Parathyroid Hormone Gene Alterations in the Parathyroid Hormone Gene Structure Hormone Secretion General Features of Hormonal Secretion Modulators of Parathyroid Gland Secretion Peripheral Metabolism Actions of Parathyroid Hormone Parathyroid Hormone Functions Structure of Parathyroid Hormone Correlated with Function Parathyroid Hormone Receptors Parathyroid HormoneInduced Signal Transduction Adenylate Cyclase Effects of Parathyroid Hormone in Target Tissues Measurement Radioimmunoassays Bioassays Conclusions Chapter References

Parathyroid hormone (PTH) is essential for the physiologic maintenance of calcium homeostasis, and a marked excess or deficiency can cause severe and potentially fatal illness. Because of its essential role in metabolism, skeletal function, and renal function, considerable effort has been expended and substantial advances have been made in understanding the biosynthesis, molecular biology, secretion, metabolism, and action of this hormone. Improved quantitation methods have enhanced our appreciation of its complex physiology and pathophysiology. The molecular biology approach has contributed valuable information about the structure and synthesis of this molecule and has led to the identification and characterization of a parathyroid hormonelike molecule, parathyroid hormonerelated protein (PTHrP); a receptor that mediates the actions of PTH and PTHrP, the PTH/PTHrP receptor; and a parathyroid calcium-sensing receptor. This chapter examines the production, metabolism, function, and measurement of PTH as a foundation for understanding the cause, pathogenesis, and diagnosis of disorders of the parathyroid glands.

BIOSYNTHESIS OF PARATHYROID HORMONE

Parathyroid hormone follows a pattern of biosynthesis and of vectorial transport through organelles of the cell that now is well established for many peptide hormones (Fig. 51-1).1,2 The major glandular form of the hormone, an 84-amino-acid, straight-chain peptide, PTH 184, is biosynthesized on the polyribosomes of the rough endoplasmic reticulum of the parathyroid gland. The gene for PTH encodes a precursor, prepro-PTH, that is extended at the amino-terminus of PTH 184 by 31 residues. The NH2-terminal, 25-residue portion, characterized by its hydrophobicity, is called the signal, leader, or pre sequence, and it facilitates entry of the nascent hormone into the cisternae of the endoplasmic reticulum.

FIGURE 51-1. The biosynthesis of precursor and secretory forms of parathyroid hormone (PTH). Within the nucleus, transcription of the gene encoding PTH is followed by processing of the pre-mRNA through removal of intervening sequences. The mature mRNA leaves the nucleus and attaches to polyribosomes in the cytoplasm. The signal or pre-sequence of the hormone then binds to a signal recognition particle that interacts with a docking protein on the membrane of the endoplasmic reticulum, facilitating entry of the nascent peptide into the cisternae. The signal sequence is removed, leaving the precursor proPTH. The NH2-terminal hexapeptide of this molecule is then removed in the Golgi apparatus, and the mature hormone, PTH 184, is packaged into secretory granules.

As the signal sequence of the synthesized hormone emerges from the ribosome, it binds to a signal recognition particle that stops further synthesis of the nascent protein. The signal recognition particle carrying the ribosome then binds to an integral membrane protein of the endoplasmic reticulum, called the docking protein or signal recognition particle receptor.3 This protein releases the block in protein synthesis, and the nascent peptide is transported across the membrane into the cisternae of the endoplasmic reticulum. The signal sequence is simultaneously removed at the inner surface of the endoplasmic reticulum, enzymatically at a glycyllysyl bond. The resultant precursor molecule, proPTH, is extended at the NH2-terminus of PTH 184 by only six amino acids. The pro sequence is necessary for efficient translocation and cleavage of the signal peptide. Once formed, proPTH is transported to the Golgi apparatus. The prohormone hexapeptide has several basic residues that serve as a recognition sequence to yield the mature hormone. Unlike many other prohormones, proPTH does not contain another sequence at the COOH terminus and has not been detected within the circulation in states of parathyroid gland hyperfunction. ProPTH has little intrinsic biologic activity until cleaved to create the hormonal form.1,4 The translocation of proPTH from the rough endoplasmic reticulum to the Golgi apparatus is a process that requires energy.2 The conversion of proPTH to PTH appears to occur within the Golgi apparatus through the action of endopeptidases with trypsin-like specificity.1,4 The enzymes likely to be involved are furin and PC7, mammalian proprotein convertases that are related to bacterial subtilisins.5,6 Little proPTH is stored within the gland. The resultant mature 84-amino-acid form of the hormone is packaged in secretory granules and transported to the region of the plasma membrane. This appears to occur by a process involving vesicular budding and fusion that is driven by low-molecular-weight guanine nucleotidebinding proteins. The hormone is released by exocytosis in response to the principal stimulus to secretion, hypocalcemia. The calcium ion has not been shown to influence the enzymatic cleavages involved in the processing of preproPTH or proPTH. Little information is available concerning the posttranslational modification of the amino acid sequence of the hormone. Glycosylation does not appear to occur. Phosphorylation of proPTH and PTH occurs in vitro, where 10% to 20% of the hormone is phosphorylated at serine residues within the NH2-terminal region of the molecule.7 However, the influence of this process on intraglandular processing of the molecule or on bioactivity remains undefined.

MOLECULAR BIOLOGY OF PARATHYROID HORMONE

CHARACTERISTICS OF THE NORMAL PARATHYROID HORMONE GENE The structural characterization of messenger RNAs (mRNAs) and genes encoding preproPTH from several mammalian species (i.e., humans, cattle, pigs, dogs, and

both rats and mice) and one avian species (i.e., chicken) has been accomplished using the techniques of molecular biology.8,9,10,11 and 12 In mammals, mature PTH has 84 amino acids, but the chicken form has 88 amino acids. The preproPTH gene is organized into three exons: exon I encoding the 5' untranslated region, exon II encoding the pre-propeptide coding region and part of the prohormone cleavage site recognition sequence, and exon III encoding the Lys 2-Arg 1 of the prohormone cleavage site, the 84 amino acids of the mature hormone, and the 3' untranslated region. Some of these organizational features are shared with the PTHrP gene, in which the same functional domainsthe 5' untranslated region, prepro region of the precursor peptide, and the prohormone cleavage site and most or all of the mature peptideare encoded by single exons13,14,15 and 16 (Fig. 51-2; see Chap. 52). For PTHrP, exons encoding alternative 5' untranslated regions, carboxyl-terminal peptides, and 3' untranslated regions may also exist, depending on the species.

FIGURE 51-2. Comparison of the structural organization of the human parathyroid hormone (PTH) and parathyroid hormonerelated protein (PTHrP) genes. (Nc, noncoding region; CTP, carboxyl-terminal peptide.) Roman numerals denote exons.

The PTH and PTHrP genes are both single-copy genes and have been mapped to the short arms of chromosomes 11 and 12, respectively.17,18 These two human chromosomes are thought to have been derived by an ancient duplication of a single chromosome, and the PTH and PTHrP genes and their respective gene clusters have been maintained as syntenic groups in the human, rat, and mouse genomes.18 Overall, because of the similarity in NH2-terminal sequence of their mature peptides, their gene organization, and chromosomal location, it is likely that the PTH and PTHrP genes evolved from a single ancestral gene and form part of a single gene family. Two common restriction-fragment-length polymorphisms for the restriction enzymes PstI and TaqI have been detected at the human PTH gene locus, and a variable number of tandem repeat polymorphisms that occur with high frequency in the general population have been described within the human PTHrP gene.19 These DNA markers are useful for testing for linkage of inherited disease genes to these loci in family studies. ALTERATIONS IN THE PARATHYROID HORMONE GENE STRUCTURE The human PTH produced by patients with hyperparathyroidism is structurally normal.8,20 In a small number of parathyroid tumors examined, the PTH gene sequence is rearranged, and the 5' flanking region of the PTH gene is placed upstream of the cyclin D gene located on the long arm of chromosome 11.21 This is thought to lead to deregulated expression of the cyclin gene that contributes to tumor development. However, this type of gene rearrangement occurs infrequently in parathyroid tumors. A more common event involves the loss or inactivation of the multiple endocrine neoplasia type 1 gene, also on the long arm of chromosome 11. The protein encoded by the MEN1 gene,22,23 called menin, is a 610-amino-acid nuclear protein;24 its tumor suppressor function is thought to involve its binding to the transcription factor JunD and inhibition of JunD-activated transcription.25 Germline mutations in the MEN1 gene cause familial and sporadic MEN1 and are found in 20% of non-MEN1 parathyroid adenomas. Loss of heterozygosity at 11q13 is found in MEN1 tumors and sporadic parathyroid adenomas, which is consistent with MEN1 being a tumor-suppressor gene.26,27 Mutations have been identified in the PTH gene in some cases of familial isolated hypoparathyroidism. One patient with autosomal-dominant hypoparathyroidism had a mutation within the protein-coding region of the PTH gene in which there was a single base substitution (T > C) in exon II, resulting in the substitution of arginine (CGT) for cysteine (TGT) in the signal peptide.28 This places a charged amino acid in the hydrophobic core of the signal peptide, leading to inefficient processing of the mutant preproPTH to PTH. In a family with autosomal-recessive isolated hypoparathyroidism, a donor splice mutation was identified in the PTH gene of affected individuals at the exon II-intron II boundary that resulted in the loss of exon II, which encodes the initiation codon and signal peptide.29 A search for evidence of ectopic PTH synthesis (i.e., synthesis outside of parathyroid tissue) indicates that it occurs only rarely in malignancies associated with hypercalcemia.30 However, with the advent of highly specific PTH immunoassays and mRNA analysis, a few cases of true ectopic PTH production have been documented. In one case of an ovarian carcinoma, the 5' PTH gene regulatory region that normally silences gene transcription in nonparathyroid cells was replaced by a foreign sequence that allowed inappropriate transcription to take place.31

HORMONE SECRETION
GENERAL FEATURES OF HORMONAL SECRETION Relatively little PTH is stored in secretory granules within the parathyroid glands. In the absence of a stimulus for release, intraglandular metabolism occurs, causing complete degradation of the hormone to its constituent amino acids or partial degradation to fragments through a calcium-regulated enzymatic mechanism.32 In cases of hypercalcemia, the predominant hormonal entities released from parathyroid glands appear to be fragments composed of midregion or COOH-terminal sequences or both, containing little or no bioactivity.33,34 In cases of hypocalcemia, degradation of PTH within the parathyroid cell is minimized, and the major hormonal entity released appears similar to or identical with the bioactive PTH 184 molecule.35 Consequently, in the presence of hypocalcemia, increased amounts of bioactive PTH are secreted, even in the absence of additional synthesis of hormone. With a brief hypocalcemic stimulus, a biphasic secretory response often occurs. Hormone, presumably newly synthesized and derived from immature Golgi vesicles, is initially released in a large burst over a few minutes.2 This is followed by a lower response sustained for a longer period, presumably representing hormone stored in secretory granules. However, hormone stores are insufficient to maintain secretion for more than a few hours in the presence of a sustained severe hypocalcemic stimulus; although other mechanismstranscriptional and posttranscriptionalincrease PTH synthesis to some extent, additional PTH secretion ultimately depends on an increase in the number of parathyroid cells. Such an increase appears to be modulated by the reductions in circulating 1,25-dihydroxyvitamin D that often accompany hypocalcemia. A second protein, identical to chromogranin A from the adrenal medulla, is cosecreted with PTH in most conditions leading to the release of PTH.36,37 This 50-kilodalton (kDa) protein is synthesized within the parathyroid gland and stored with PTH within secretory granules. This molecule, which can be glycosylated and phosphorylated, is a member of the chromogranin-secretogranin (granin) family of proteins that occurs in virtually all neuroendocrine cells.38,39 and 40 The family of proteins plays several roles in the process of regulated secretion, including targeting peptide hormones and neurotransmitters to secretory granules of the regulated secretory pathway. Chromogranin A also functions as a precursor of biologically active peptides that modulate neuroendocrine cell secretion in an autocrine or paracrine fashion (see Chap. 175). MODULATORS OF PARATHYROID GLAND SECRETION The calcium ion is the main regulator of parathyroid gland activity, although several other agents influence the release of PTH from parathyroid glands. These include various ions, agents altering the activity of the parathyroid cell adenylate cyclase system (e.g., b-adrenergic catecholamines, histamine), peptides derived from chromogranin A, and vitamin D metabolites. A circadian rhythm has been reported for PTH secretion, with increased blood levels occurring at night and small amplitude pulses of PTH secretion occurring at much shorter intervals.41,42 These studies may suggest neural or central nervous system influences on PTH secretion, or they could reflect circadian alterations in the levels of extracellular fluid calcium.

IONS Cations. The most potent of the cations modulating PTH release and the secretagogue that is most important in altering PTH release under physiologic and pathophysiologic circumstances is the calcium ion (Fig. 51-3). Although there is an inverse relationship between ambient calcium levels and PTH release, this is a curvilinear rather than proportional relationship.43 From in vivo studies of cattle, maximal rates of PTH secretion of ~16 ng/kg per minute appear to be rapidly achieved at calcium levels below 7.58.0 mg/dL (1.882.00 mmol/L). When calcium levels are reduced from 10.0 to 9.5 mg/dL (2.502.38 mmol/L), a small and gradual increase in PTH secretion occurs that does exhibit a proportional relationship to the ambient calcium level. Half-maximal secretion rates normally occur at calcium levels of ~8.5 mg/dL (2.12 mmol/L), which is the set-point for PTH secretion. Basal secretion rates result after ambient calcium levels have risen above 11 mg/dL (2.75 mmol/L) and appear to persist despite further increases in calcium concentration, even up to 1618 mg/dL (4.004.50 mmol/L). Similar results have been observed in studies with human parathyroid tissue in vitro. The nonsuppressible (more correctly, noncalcium suppressible) component of constitutive PTH secretion appears to comprise mainly bioinactive midregion and COOH-terminal fragments. However, direct determination of the activity of hormonal material released when calcium levels are markedly elevated also demonstrates some bioactivity.

FIGURE 51-3. The sites of calcium regulation for parathyroid hormone (PTH) biosynthesis, intraglandular metabolism, and secretion. Sites known to be influenced by calcium (*) include reversible reduction of preproPTH mRNA levels by elevated extracellular fluid calcium levels acting on preproPTH gene transcription and mRNA stability, increase in the production and release of COOH-terminal fragments by elevated calcium levels, and increase in secretion of the mature form of PTH by reduction in extracellular fluid calcium concentration. Whether calcium acts to alter the splicing of the pre-mRNA is unclear (?).

This inverse relationship between PTH and extracellular calcium contrasts with the influence of the calcium ion as a secre-tagogue in most other secretory systems in which elevations in this ion enhance release of the secretory product. This distinction between the parathyroid cell and other secretory cells is maintained intracellularly, where elevations rather than decreases in cytosol calcium correlate with decreased PTH release.44 Alterations in extracellular fluid calcium levels are transmitted through a parathyroid plasma membrane calcium sensing receptor that couples through a G-protein complex to phospholipase C. Increases in extracellular calcium lead to increases in inositol 1,4,5-trisphosphate (IP3) and mobilization of intracellular calcium stores. The manner in which this inhibits hormone secretion is not understood. Although it would be anticipated that increases in diacylglycerol would accompany IP3 increases caused by hydrolysis of phosphoinositides, activate protein kinase C, and result in reduced PTH secretion, this does not appear to occur in the parathyroid cell.45 Paradoxically, agents that do stimulate protein kinase C, such as phorbol esters, stimulate rather than inhibit hormone secretion. This may be a result of phosphorylation of amino acids in the COOH-terminus of the CaSR, leading to its desensitization by blocking interaction of the receptor with its G protein. The precise steps in the pathway from changes in extracellular calcium levels to hormone release remain to be elucidated. The parathyroid cell Ca2+-sensing receptor cDNA was identified by expression cloning in Xenopus laevis oocytes.46 The mRNA of the human receptor encodes a polypeptide of 1078 amino acids that is predicted to contain a very large extracellular domain of ~600 amino acids and a seven transmembrane-spanning region characteristic of G-proteincoupled cell-surface receptors. Compared with the other known G-proteincoupled receptor family members, the Ca2+-sensing receptor shows some homology with the metabotropic glutamate, g-aminobutyric acid-B, and vomeronasal odorant receptors, sharing conserved cysteine residues and a hydrophobic sequence in the NH2-terminal region. The Ca2+-sensing receptor has a low affinity for Ca2+, and consistent with this, the receptor sequence does not contain any of the Ca2+ binding motifs found in high-affinity calcium-binding proteins. Highly acidic regions in the extracellular NH2-terminal domain and the second extracellular loop may bind calcium as they do in other known low-affinity Ca2+ binding proteins. Besides the parathyroid, the Ca2+-sensing receptor is also expressed in other cells having Ca2+-sensing functions, such as those of the kidney tubule, the calcitonin-secreting thyroid C-cells; and in diverse other organs and tissues such as brain, keratinocytes, gastrointestinal tract, and retina. Neomycin binds the receptor, possibly accounting for the toxic renal effects of aminoglycoside antibiotics. The human Ca2+-sensing receptor gene has been partially characterized and shown to consist of seven exons spanning more than 20 kb of genomic DNA.47 The long extracellular domain is encoded by the first two to six exons, and the remainder of the molecule is encoded by exon seven. Exons 1A and 1B encode two alternative 5' untranslated regions of the CaSR mRNA. Inherited abnormalities of the CaSR gene, located on chromosome 3p13.3-21, can lead either to hypercalcemia or to hypocalcemia depending on whether they are inactivating or activating, respectively. Heterozygous loss-of-function mutations give rise to familial (benign) hypocalciuric hypercalcemia (FHH), in which the lifelong hypercalcemia is asymptomatic.47,48 and 49 The homozygous condition manifests itself as neonatal severe hyperparathyroidism (NSHPT), which is a rare disorder characterized by extreme hypercalcemia and the bony changes of hyperparathyroidism that occur in infancy.50,51 In several cases of NSHPT the parents are normocalcemic, and the cases seem to be sporadic. Autosomal-dominant hypocalcemia (ADH) is caused by gain-of-function mutations in the CaSR gene.52 Autosomal-dominant hypocalcemia may be asymptomatic or present with neonatal or childhood seizures. Because of the overactive CaSR in the nephron, these patients are at greater risk of developing renal complications during vitamin D therapy than are patients with idiopathic hypoparathyroidism. A common polymorphism in the intracellular tail of the CaSR, Ala to Ser at position 986, has a modest effect on the serum calcium concentration in healthy individuals.53 CaSR polymorphisms might also affect urinary calcium excretion and bone mass; therefore, the CaSR is a candidate gene for involvement in disorders such as idiopathic hypercalciuria and osteoporosis. The CaSR is a target for phenylalkylamine compounds so-called calcimimetics that are allosteric stimulators of the affinity for cations by CaSR.54 These orally active compounds have been evaluated for treatment of primary and secondary hyperparathyroidism and parathyroid carcinoma.55 The CaSR expressed in the developing parathyroid glands and in the placenta plays an important role in regulating fetal calcium concentrations. Normally, the fetal blood calcium level is elevated above the maternal level. This depends on the action of PTHrP (released from the fetal parathyroids and placenta) on placental calcium transport. Disruption of the CaSR, as shown by studies in CaSR-deficient mice, causes fetal hyperparathyroidism and hypercalcemia as a result of fetal bone resorption.56 The transfer of calcium across the placenta is reduced, and renal calcium excretion is increased. In vitro reductions in extracellular fluid calcium do not appear capable of enhancing specific PTH biosynthesis, which seems to occur at a maximal or near-maximal rate per cell; however, in vivo the reverse is the case, with biosynthesis of PTH normally being set at a low level. In vivo, PTH mRNA levels are markedly stimulated by decreased circulating calcium concentrations. It is thought that the relative insensitivity of parathyroid cell cultures to extracellular calcium is the result of reduced expression of the CaSR.57,58 In vitro and in vivo studies have reported that elevated calcium levels reversibly and specifically reduce PTH mRNA.59 This occurs, in part, by a direct effect on transcription of the pre-proPTH gene60 and, in part, by a posttranscriptional mechanism, whereby hypocalcemia stabilizes and hypercalcemia destabilizes the PTH mRNA.61 Although the results of in vitro studies are conflicting, prolonged hypocalcemia in vivo may stimulate DNA replication, cell division, and the production of increased numbers of parathyroid cells or parathyroid hyperplasia. This would increase the synthesis of proteins, including PTH, within the hypercellular parathyroid gland and ultimately would increase PTH release. In primary parathyroid gland hyperfunction resulting in hyperparathyroidism, alterations in the calcium-sensing mechanism may manifest as a setpoint error, producing a shift to the right of the curve relating PTH secretion to extracellular calcium levels.62 Consequently, elevated concentrations of extracellular fluid calcium may be required to reduce PTH secretion, resulting in an adenomatous or hyperplastic parathyroid gland that is incompletely suppressed by calcium. Such a mechanism may underlie the observation that an increase in the mass of parathyroid tissue, such as that produced by transplantation, can be associated with hypercalcemia. The parathyroid glands of patients with primary and severe uremia secondary to hyperparathyroidism have reduced CaSR expression as assessed by immunostaining.63 Loss of a functional CaSR, as in humans with NSHPT or in mice in which the CaSR gene has been ablated, leads to severe parathyroid hyperplasia.64 If basal secretion per cell produces a significant amount of bioactive PTH, the cumulative increase in this basal or noncalcium-suppressible secretion arising from an increase in parathyroid cells also could be responsible for the hypercalcemia. The precise mechanistic relationship of extracellular calcium to parathyroid cell growth remains to be determined. In studies in vitro, magnesium appears to parallel the effects of calcium on PTH release, although with reduced efficacy.65,66 This is consistent with the known affinity of the parathyroid calcium-sensing receptor for magnesium that is lower than that for calcium itself. Mild hypomagnesemia or hypermagnesemia stimulates or suppresses,

respectively, PTH secretion.67 A special situation exists in clinical disorders associated with severe hypomagnesemia in which PTH secretion is impaired67 (see Chap. 68). With the discovery of the syndrome of aluminum toxicity in uremia, characterized by osteomalacia, low circulating PTH levels, and a tendency toward hypercalcemia during treatment with vitamin D, the effects of aluminum on PTH secretion were assessed in vitro. Such studies, performed with rather high ambient aluminum concentrations, have shown suppression of PTH release by this cation,68 whereas, PTH mRNA expression is unaffected. A direct effect of aluminum on inhibition of PTH secretion in vivo therefore could contribute to the low circulating PTH levels seen in the presence of aluminum intoxication. It is unlikely that aluminum plays any physiologic role in the normal regulation of PTH secretion (see Chap. 131 and Chap. 61). Anions. The phosphate ion is of most interest clinically as a potential modulator of PTH release. Hyperphosphatemia induced by intravenous or oral administration of phosphate is associated with increased circulating levels of PTH. The effects have been thought to be indirect and a result of the hypocalcemia that accompanies the rise in serum phosphate.69 However, it has been suggested that the anion can have a direct action on PTH secretion although the mechanism is unknown.70,71 A parathyroid phosphate sensor analogous to the CaSR has yet to be identified. The physiologic role of other anions, such as chloride, that have been associated with alterations of PTH levels in vitro remains unclear. MODULATORS OF ADENYLATE CYCLASE ACTIVITY IN THE PARATHYROID CELL A group of adenylate cyclasestimulating agonists, such as catecholamines, prostaglandins of the E series, calcitonin, gut hormones of the secretin family, and histamine, influences PTH release in vitro.72,73 and 74 b-Adrenergic agonists (e.g., epinephrine, isoproterenol) also enhance PTH release in animals, and some decreases in circulating PTH levels in humans have been reported with b-blockers. Despite the ability of the b-adrenergic catecholamines and histamine to induce PTH secretion, it is uncertain what physiologic or pathophysiologic role these biogenic amines may have in humans.74,75 Agents that may lower cyclic adenosine monophosphate (cAMP) levels within the parathyroid gland, such as a2-adrenergic agonists, prostaglandin F2, and somatostatin, have been associated with decreased PTH secretion, mainly in vitro. CHROMOGRANIN ADERIVED PEPTIDES Peptides derived from chromogranin A (CgA), such as b-granin (i.e., CgA 1113), synthetic b CgA 140, pancreastatin (i.e., porcine CgA 240288), and parastatin (i.e., porcine CgA 347419), have been reported to inhibit low-calciumstimulated PTH release from parathyroid cells in culture.76,77 The physiologic significance of these observations remains unclear. Within the parathyroid gland, processing of CgA to peptides occurs to a lesser extent than in some other endocrine tissues, such as the B cells of the pancreas, and the parathyroid cells could respond in a paracrine or endocrine fashion to CgA-derived peptides generated elsewhere. STEROLS AND STEROIDS Various studies have demonstrated a role for vitamin D metabolites in the modulation of PTH release. A feedback loop between PTH-induced increase of vitamin D metabolite levels and vitamin D metaboliteinduced decrease of PTH levels has been postulated. A high-affinity 1,25-dihydroxyvitamin D receptor (a member of the nuclear/steroid hormone receptor superfamily of transcriptional regulators) exists in parathyroid cells, and with the localization of injected 1,25-[3H]-dihydroxy-vitamin D within the parathyroid cell, interest in this metabolite as a potential regulator of PTH synthesis or secretion has arisen. Efforts that focused on the role of 1,25-dihydroxyvitamin D in the biosynthesis of PTH provided evidence, in vitro and in vivo, that the sterol reversibly reduces levels of mRNA responsible for PTH synthesis.78,79 This is effected by the action of the sterol on preproPTH gene transcription. A vitamin D response element has been identified in the human PTH gene promoter, although the binding of the vitamin D receptor (VDR) to this element apparently does not require the retinoid X receptor, the normal heterodimerization partner of the VDR.80 The precise mechanism of the down-regulation of PTH gene transcription by 1,25-dihydroxyvitamin D remains to be elucidated. Although 1,25-dihydroxyvitamin D is of uncertain importance in influencing immediate hormone release, it plays a role in modulating hormone synthesis within the gland, altering the quantities of hormone available for immediate release by secretagogues. Moreover, an early in vivo effect of reduction in 1,25-dihydroxyvitamin D appears to be an increase in maximal releasable PTH as a result of stimulation of glandular proliferation.79 In vitro studies show that this effect of the sterol on cell proliferation is mediated by its action on early immediate response gene expression such as the C-MYC protooncogene.81 The role of other metabolites of vitamin D in modulating PTH biosynthesis or secretion remains unclear. However, nonhyper-calcemic analogs of 1,25-dihydroxyvitamin D have been developed that do diminish PTH secretion in vitro and in vivo and that may serve in the future as therapeutics for hyperparathyroidism. Cortisol, in some studies, has been reported to elevate extracellular levels of PTH in vivo and in vitro.82 It is unclear, however, whether this effect is a direct effect on synthesis, on alteration of hormonal release, or on an alteration of the levels of enzymes metabolizing PTH within the parathyroid cell.82 Although these findings could help explain the mild secondary hyperparathyroidism occasionally reported with glucocorticoid therapy, their overall significance remains unknown. PERIPHERAL METABOLISM Besides metabolism within the parathyroid cell, PTH 184 undergoes extensive peripheral metabolism after release into the general circulation (Fig. 51-4). Although PTH 184 can be cleared from the circulation by the kidney and hormonal fragments may be generated in the kidney, such fragments probably do not reenter the circulation. Studies using sequence-specific radioimmunoassays coupled with gel filtration of serum or plasma have shown that the major circulating forms of PTH are the midregion and COOH-terminal fragments.83,84,85,86,87 and 88 Such fragments may be released from the parathyroid gland in the presence of hypercalcemia.

FIGURE 51-4. Model of parathyroid hormone (PTH) metabolism in the presence of normal renal function (NRF) and chronic renal failure (CRF). In the presence of NRF, PTH 184 released from the parathyroid gland is cleared by the kidney or metabolized in the liver, where amino (N; NH2 ) and carboxyl (C; COOH) fragments are generated. Carboxyl, but generally not NH2 , fragments, are released into the circulation; these COOH fragments enter a pool, which also includes a contribution from the parathyroid gland, and are cleared by the kidney. In chronic renal failure, PTH secretion can be increased by a tendency toward hypocalcemia, and fewer COOH fragments are released from the parathyroid glands. Failure of the nonfunctioning kidneys to clear PTH results in hepatic metabolism of this moiety, with increased production of NH2 - and COOH fragments. Under these conditions, COOH-terminal fragments reach high circulating concentrations because of increased production and reduced renal clearance.

Another major site of origin of PTH fragments appears to be the liver. Studies using 125I-labeled PTH 184 injected intravenously into animals have demonstrated that initial cleavage occurs between residues 33 and 34 and at sites COOH-terminal to this. These and other studies have indicated that the hepatic Kupffer cell is the most likely site of peripheral hormonal cleavage.89 The midregion and COOH-terminal fragments thus generated can enter the general circulation and subsequently be metabolized in the kidney. In view of structure-function studies that have localized the bioactivity of PTH 184 to the NH2-terminal third of the molecule, such midregion and COOH-terminal fragments should be biologically inert, at least with respect to calcium homeostatic actions mediated through the PTH/PTHrP receptor. Because of their long half-life in the circulation relative to intact PTH 184, such inert moieties generally comprise most circulating PTH-related entities.90 Conversely, the corresponding NH2-terminal fragments generated during hepatic cleavage of PTH 184 would be expected to be bioactive, but such fragments seem to be completely degraded in the liver and do

not reenter the circulation. The major circulating bioactive moiety is similar to or identical with the intact molecule PTH 184. Clearance of midregion and COOH-terminal fragments of PTH apparently depends almost solely on renal mechanisms. In cases of renal insufficiency, concentrations of midregion and COOH fragments may increase considerably (see Chap. 58 and Chap. 61). By using a sensitive renal cytochemical bioassay for PTH in association with gel filtration fractionation of plasma, it was possible to confirm that the major circulating bioactive moiety in hyperparathyroidism is similar to or identical with PTH 184.91 These observations have been confirmed with the development of sensitive immunoradiometric assays that simultaneously recognize NH2 and COOH epitopes on the PTH molecule and therefore detect only intact PTH 184. In secondary hyperparathyroidism associated with severe chronic renal failure, additional low-molecular-weight bioactive forms may occasionally be detected.4 The finding of a half-life of disappearance of intact bioactive PTH in end-stage renal disease that cannot be differentiated from normal kinetics indicates that, with the gradual evolution of renal insufficiency, extrarenal sites of degradation assume increasing importance in the clearance of the hormone.4 This conclusion agrees with the finding that the clearance of NH2-terminal immunoreactivity is not decreased in patients with renal failure.90

ACTIONS OF PARATHYROID HORMONE

PARATHYROID HORMONE FUNCTIONS The major function of PTH appears to be the maintenance of a normal level of extracellular fluid calcium. The hormone exerts important effects on bone and kidney and indirectly influences the gastrointestinal tract. In response to a fall in the extracellular fluid ionized calcium concentration, PTH is released from the parathyroid cell and acts directly on the kidney to enhance renal calcium reabsorption and promote the conversion of 25-hydroxy-vitamin D to 1,25-dihydroxyvitamin D.92 This latter metabolite increases gastrointestinal absorption of calcium and, with PTH, induces skeletal resorption, causing the restoration of extracellular fluid calcium and the neutralization of the signal initiating PTH release. The opposite series of homeostatic events occurs in response to a rise in extracellular fluid calcium levels. Although this scheme outlines the overall events that occur after a fall in calcium levels, aspects of the response may vary with the extent of the fall in calcium concentration and the consequent rise in PTH. There is some evidence that the actions of PTH on target tissues may depend on its prevailing concentration.93 Consequently, certain actions of PTH, such as renal calcium retention and even skeletal anabolic actions, may predominate at relatively low circulating concentrations of PTH, but skeletal lysis may become evident only at higher levels of circulating PTH. In addition to regulating calcium homeostasis, PTH elicits various other responses. Whether these other functions evolved independently or to complement the role of the hormone in maintaining calcium homeostasis is unclear. Among these other responses are perturbations of other ions, the most marked of which are those involving phosphate. As a consequence of PTH-enhanced 1,25-dihydroxyvitamin D production, the gastrointestinal absorption of phosphate is facilitated to some extent, and with PTH-induced skeletal lysis, phosphate and calcium are released. These effects increase the extracellular fluid phosphate levels, but the predominant effect of PTH on phosphate homeostasis is to inhibit renal phosphate reabsorption and produce phosphaturia. Consequently, a net decrease in extracellular fluid phosphate concentration occurs, and this may be viewed as adjunctive to the role of PTH in raising calcium levels. In common with other peptide hormones, PTH interacts through a receptor on the outer surface of the plasma membrane of target cells. In these target tissues, the result of this interaction of PTH with the membrane receptor has classically been appreciated to be the stimulation of the enzyme adenylate cyclase on the inner surface of the plasma membrane, although the same receptor can couple to phosphatidylinositol turnover as well. The product of this adenylate cyclase activity, cellular cAMP, and the products of phospholipase activity, IP3, diacyl-glycerol, and intracellular Ca2+, initiate a cascade of events leading to the final cellular response to the hormone. The PTH receptor may activate both these intracellular signaling path-ways in some target tissues and act preferentially by means of one or another pathway in other cells, but these issues are still under study. STRUCTURE OF PARATHYROID HORMONE CORRELATED WITH FUNCTION The amino acid sequences of the major glandular form of mammalian PTH, PTH 184, have now been determined in humans, cattle, pigs, dogs, and rats (Fig. 51-5).11,20 The corresponding amino acid sequence of the chicken peptide has also been determined and contains 88 rather than 84 amino acids. Studies correlating hormonal structure and function have emphasized the importance of the NH2 -terminal region to bioactivity. Considerable deletion of the middle and COOH-terminal region of the intact peptide can be tolerated without apparent loss of biologic activity. A peptide composed of the NH2-terminal 34 residues, PTH 134, appears to contain all of the conventional bioactivity of the intact hormone when tested in various bioassay systems.94 Synthetic NH2-terminal fragments of PTHrP, which share a high degree of homology with PTH within the NH2-terminal 13 residues, have been shown to mimic PTH actions in many bioassay systems.95,96 and 97

FIGURE 51-5. Amino acid sequence of mammalian parathyroid hormone 184. The backbone is that of the human sequence, and substitutions in each of the bovine, porcine, and rat hormones are shown at the specific sites.

Considerable effort has been expended to define the regions of the NH2 -terminal domain of the PTH molecule (and more recently of the PTHrP molecule) that are required to interact with the PTHR and stimulate downstream signaling. These studies examined full agonist, partial agonist, and antagonist activities of a variety of synthetic analogs of PTH (1-34), generally using in vitro adenylate cyclasestimulating activity as an index of bioresponse. Early studies demonstrated that either deletion or extension of the NH2 terminus of PTH markedly inhibited the capacity of this hormone to increase cAMP in target tissues.94,98 Subsequently, the NH2 terminus of PTHrP was shown to manifest similar sensitivity.96 This work suggested that the NH2 -terminal domain of these hormones is a critical site for activation of the adenylate cyclase signaling system. Other work focused on the role of the COOH-terminal domain of PTH (1-34) in receptor binding, attempting to dissect a COOH-terminal binding domain in this entity from the NH2-terminal activation domain. Stimulated by the initial observation that poorly active NH2-terminal deleted analogs of PTH could still function as antagonists,98 this work ultimately suggested that the 14-34 portion of PTH contains the predominant binding domain.99 A number of studies have now been reported in which the structural basis of PTHrP function has also been analyzed, generally in comparison with PTH. For example, although PTH and PTHrP lack significant amino acid sequence homology in the 14-34 domains, these domains bind to osteoblastic cells with similar avidity, suggesting that these regions retain similar molecular topologies.100,101 Despite many additional studies, however, it remains premature to assign exact functional equivalence to most of the different domains and residues in these two hormones. Both PTH and PTHrP assume only a modestly ordered structural configuration in aqueous solution. Several nuclear magnetic resonance spectroscopy studies of PTH (1-34) and of PTHrP (1-34) in nonaqueous solutions (which are intended to mimic the membrane-bound receptor milieu) have disclosed some elements of secondary structure and have assigned two a-helical domains to these molecules, one less stable, extending for 711 amino acids in the NH2-terminal region, and another more stable, of 915 residues in the COOH domain.102,103,104,105,106 and 107 Some, but not all, studies have suggested interaction between the helices via a hydrophobic interface. In addition, the linker between the two helices is believed to be a turn or flexible hinge region. Cyclization of PTH and PTHrP analogs by a specific lactam apparently promotes a-helical conformation, and the stabilization of the a-helix in the putative COOH-terminal receptor-binding domain of PTH and PTHrP108,109 has been correlated with enhanced bioactivity of these hormones. To date, however, determination of the three-dimensional x-ray structure of PTH or PTHrP has not been feasible because of the difficulty in obtaining high-quality crystals. Although most in vitro studies correlating PTH structure and function have examined adenylate cyclase stimulating activity, several studies have been performed in

which other in vitro activities have been assessed. Such studies have indicated that the NH2 residues of PTH 134 may not be essential to induce effects on phosphoinositide metabolism and on the augmentation of intracellular calcium levels; rather that the PTH 2334 sequence mediates this response. In vitro studies have described the mitogenic effects of the 2547 region of PTH in osseous cells (not dependent on cAMP accumulation) and alkaline phosphatasestimulating activity of the 5384 sequence of PTH in osseous cells. The clinical significance of these in vitro observations remains to be determined.

PARATHYROID HORMONE RECEPTORS

The PTH/PTHrP receptor is a member of the G-proteinlinked receptor superfamily. These glycoproteins have hydrophobic sequences that are thought to span the plasma membrane seven times. These receptors couple to intracellular effectors through guanine nucleotide binding regulatory proteins or G proteins, which are heterotrimeric, consisting of alpha, beta, and gamma subunits. In the inactive state, the alpha subunit binds GDP; however, when the receptor is occupied by ligand, the G protein is activated by exchange of GTP for GDP and dissociates into a and b-g subunits. The alpha subunit, with GTP bound, can then stimulate (Gsa) or inhibit (Gia), the activity of effector molecules such as adenylate cyclase or phospholipase C. The a subunit also has an intrinsic GTPase activity that hydrolyzes GTP to GDP, terminating the effector activation. The PTH/PTHrP receptor belongs to a subgroup of the G-protein coupled receptor superfamily (group II GPCR) that, by virtue of their structural similarities, also includes the receptors binding secretin, calcitonin, vasoactive intestinal peptide, glucagon, glucagon-like peptide I, and growth hormonereleasing hormone. These receptors show no significant sequence identity with other known G-proteinlinked receptors. All of these receptors couple to the adenylate cyclase effector by means of Gsa. The activated PTH/PTHrP receptor couples to at least two effectors: adenylate cyclase and phospholipase C. Other members of this subgroupincluding the receptors for calcitonin, glucagon, and glucagon-like peptide Ialso couple to multiple signaling pathways. PTH/PTHrP receptor cDNAs have been cloned and characterized from several species and tissues, including opossum kidney, rat bone, and human kidney and bone.110,111,112 and 113 The amino acid sequences of the receptors are highly homologous, demonstrating a marked conservation across species. The opossum and rat PTH/PTHrP receptors bind NH2-terminal analogs of PTH and PTHrP with similar affinity, while the human receptor appears to bind PTH somewhat preferentially. This is consistent with the study results of relative bioactivities of PTH and PTHrP in vivo in humans, that showed PTHrP 134 to be less potent than PTH 134.114,114a The gene for the PTH/PTHrP receptor contains multiple introns (Fig. 51-6). For example, the mouse gene has at least 15 exons14 of which encode the receptor proteinthat span more than 32 kb of genomic DNA.115 There are eight exons containing predicted membrane-spanning domains. These exons are heterogeneous in length, and three of the exon-intron boundaries fall within putative transmembrane sequences, suggesting that these exons did not arise from duplication events. The exon-intron organization of the PTH/PTHrP gene is similar to that of the growth hormonereleasing hormone gene, especially in the transmembrane regions, suggesting that the two genes evolved from a common precursor. PTH/PTHrP receptor gene transcription in mice is controlled by two promoters; P1, which is selectively active in kidney, and P2, which functions in a variety of tissues. Although P1 and P2 are conserved in the human PTH/PTHrP receptor gene, P1 activity in the kidney is weak, and a third promoter, P3, accounts for the majority of renal PTH/PTHrP receptor transcripts in humans.116 During development, only P2 is active at midgestation in many human tissues, including calvaria and long bone. Thus, factors regulating the well-conserved P2 promoter control PTH/PTHrP receptor gene expression during skeletal development. Later in development, receptor gene expression is up-regulated with the induction of both P1 and P3 promoter activities.116a

FIGURE 51-6. Schematic representation of the parathyroid hormone (PTH) and parathyroid hormonerelated protein receptor. The gene contains at least 15 exons. Exon I encodes the 5' untranslated region of the mRNA. The portions of the receptor encoded by the remaining 14 exons are schematically depicted by the alternate blocks of unfilled and filled circles. Exons are depicted as follows: SS; putative signal sequence; E1E4; extracellular sequences; T1T7; transmembrane sequences; C; cytoplasmic sequence. The mouse PTH/PTHrP receptor of 591 amino acids is shown.

The PTH/PTHrP receptor gene promoters lack consensus TATA elements, and the downstream promoter P2 and P3 promoters are GC-rich, which would be consistent with the widespread expression of the receptor mRNA. Although the receptor mRNA is highly expressed in kidney and bone, which are the primary target tissues of PTH, it is also expressed in nonclassic target tissues such as liver, brain, smooth muscle, spleen, testis, and skin.112,117 In most of these tissues, the receptor probably mediates the local action of PTHrP. Although the predominant transcript is 2.5 kb, some tissues also express smaller or larger transcripts that probably are the result of alternative splicing of the primary transcript. The functional significance of these different forms of the receptor is unknown. A second related receptor, PTHR2, which is the product of a distinct gene and binds PTH but not PTHrP, has been identified.118 Its expression is limited to brain, pancreas, testis, and placenta; its function is unknown. Ligand-binding specificity of the PTH/PTHrP and the PTH2 receptors resides predominantly, but not exclusively, in its NH2-terminal extracellular domain, whereas activation and generation of the cAMP signal is generated in the membrane-associated domain and involves specific amino acids in transmembrane domain (TM) 3 and extracellular loop 2. This suggests a two-step interaction of PTH and receptor, whereby the ligand first complexes with the NH2 -terminal domain of the receptor, after which the complex interacts with the membrane-associated part of the receptor to generate the signal. Extensive analysis of the PTHR has also been performed in an attempt to delineate the structural basis of its function. Such studies using mutant and chimeric PTHR have indicated that the NH2 -terminal extracellular domain of the PTHR in particular is important for ligand binding and appears to interact with the COOH region of PTH or PTHrP.119 Nevertheless, amino acid residues elsewhere in the PTHR, such as in the third extracellular loop, may also contribute to this binding.120 Residues in the membrane-spanning helices of the PTHR may be required for binding residues in the NH2 -terminal domain of the ligand. Amino acids in the extracellular loops and the transmembrane regions are involved in signaling by PTH analogs, and residues in several intracellular loops and the COOH terminal intracellular tail are implicated in linkage to both Gsa and Gqa. Residues in the second intracellular loop appear to be especially critical for interaction with Gqa.121 Other studies have identified sites in the COOH tail of the PTHR that may be important for receptor internalization and phosphorylation.122 Consequently, a great deal of important information has been gleaned and continues to be generated on specific sites in the PTHR that may be necessary for hormone binding, for hormone signaling, and for PTHR regulation. Nevertheless, a comprehensive understanding of the detailed interaction of PTH and of PTHrP with the PTHR will have to await the determination of the three-dimensional structure of the bimolecular complex. High circulating levels of PTH in hyperparathyroid states have been associated with hormonal desensitization in target tissues, apparently caused by diminished receptor capacity and a postreceptor reduction in functional levels of Gs.123,124 PTH receptors, similar to many peptide hormone receptors, appear to be subject to down-regulation. The renal resistance to PTH often seen in hyperparathyroid states may be the partial result of this kind of regulatory mechanism.125 More widespread reductions in Gs are associated with hormone resistance in the disorder pseudohypoparathyroidism type 1a. The human PTH/PTHrP receptor gene has been localized to chromosome 3p21.1-22112 (and the PTHR2 gene to chromosome 2q33). A search for PTH/PTHrP receptor defects in patients with apparent resistance to endogenous PTH, such as in patients with pseudohypoparathyroidism type 1b, has produced negative results.126,127 These patients have end-organ resistance to PTH without typical features of Albright hereditary osteodystrophy and therefore are thought likely to manifest a defect in PTH/PTHrP receptor expression or function. Because no mutations were identified in the receptor gene in such patients, it was indicated that mutations in genes for other proteins involved in the PTH/PTHrP signaling pathway are most likely responsible for the defects. Linkage to chromosome 20q13.3, which includes the GNAS1 locus encoding Gsa has been established in kindreds with PHP-1b.128 In addition, the genetic defect is imprinted paternally and is therefore inherited in the same fashion as the PTH resistance in kindreds with PHP-1a and/or pseudopseudo-hypoparathyroidism and in a mouse model heterozygous for ablation of the Gnas gene.129 Although the precise nature of mutations causing PHP-1b remains to be elucidated, an understanding of how mutations at a single chromosomal locus cause overlapping and/or distinct phenotypes is likely to be related to (a) the appreciation of the complex nature of the GNAS1 gene that because of its bidirectional imprinting encodes maternally, paternally, and biallelically derived proteins,130 and (b) the subtle cell-specific imprinting of the Gsa transcript. Newly

discovered exons upstream of the Gsa exons encode two different proteins, XLas and NESP55, which are expressed in neuroendocrine cells and are probably important for secretory vesicle formation and function. It therefore is possible that a structural or regulatory mutation within the complex GNAS1 locus could account for PHP-1b, although this remains to be demonstrated. Direct evidence that the PTH/PTHrP receptor mediates the calcium homeostatic actions of PTH and the growth plate actions of PTHrP in humans has come from the study of rare genetic disorders. Jansen metaphyseal chondrodysplasia (JMC) is inherited in an autosomal-dominant fashion although most reported cases are sporadic.131 The disorder comprises short-limbed dwarfism secondary to severe growth plate abnormalities, asymptomatic hypercalcemia, and hypophosphatemia. There is increased bone resorption similar to that in primary hyperparathyroidism and urinary cAMP levels are elevated, but circulating PTH and PTHrP levels are low or undetectable. Although the PTH/PTHrP receptor is widely expressed in fetal and adult tissues, it is most abundant in three major organs, the kidney, bone, and metaphyseal or cartilaginous growth plate. It was hypothesized that the changes in mineral ion homeostasis and the growth plate in JMC are caused by constitutively active PTH/PTHrP receptors. Indeed, heterozygous PTH/PTHrP receptor gene mutations have now been identified in JMC patients. Two recurrent mutations have been described: one changes His to Arg at position 223 at the cytoplasmic end of the second transmembrane domain and the other is a Thr-to-Pro conversion at position 410 toward the cytoplasmic end of the sixth transmembrane domain.132 Current theory holds that in the absence of agonist, G protein-coupled receptors are constrained in an inactive conformation by interactions between select amino acids in the transmembrane domains. Both His-223 and Thr-410 are highly conserved in members of the group II GPCRs, attesting to their functional importance for these receptors. Mutation of position 223 to any positively charged amino acid results in constitutive activity. Possibly, in the wild-type receptor, agonist binding leads to protonation of His-223 and results in activation. In the case of Thr-410 substitution with any other amino acid leads to constitutive activity indicating the key role Thr-410 plays in maintaining the receptor in an inactive state. Inactivating or loss-of-function mutations in the PTH/PTHrP receptor have been implicated in the molecular pathogenesis of Blomstrand lethal chondrodysplasia (which was first described as a disease entity in 1985).133 This rare disease is characterized by advanced endochondral bone maturation, short-limbed dwarfism, and fetal death, thus mimicking the phenotype of PTH/PTHrP receptor-less mice. The majority of patients with Blomstrand lethal chondrodysplasia were born to phenotypically normal, consanguineous parents, suggesting an autosomal-recessive mode of inheritance. In one such fetus, a homozygous missense mutation converting Pro-132 to Leu in the extracellular domain of the PTH/PTHrP receptor was identified.134 Although the mutated receptor is expressed on the plasma membrane, it demonstrates reduced ligand binding and adenylate cyclase signaling. In another patient born to noncon-sanguineous parents, a heterozygous deletion of 11 amino acids in the fifth transmembrane domain of the receptor was found.135 This was the result of a G A substitution in exon M5, which created a novel splice acceptor site in the maternal allele. The paternal allele is not expressed, although the reasons for this are not known. As in the other example, whereas the mutant receptor is well expressed in cells, it fails to bind ligand or stimulate cAMP or inositol phosphate production. Evidence has accumulated for the possible existence of other receptors in the PTH/PTHrP system in addition to the two receptors already characterized. A receptor responsive to NH2 -terminal PTH and PTHrP, which signals through changes in intracellular free calcium rather than cAMP, has been found to exist in keratinocytes and other cells but has not yet been characterized. A receptor that signals in a similar manner but is only responsive to PTHrP-(1-34) mediates the release of arginine-vasopressin from the supraoptic nucleus. Unique biologic roles are predicted for the mid-region of PTHrP (transplacental calcium transport), the basic region of PTHrP (nuclear import and inhibition of apoptosis), the carboxyl-terminal region of PTHrP (inhibition of bone resorption), and the mid/carboxyl-terminal part of PTH (binding to osteoblast-like cells and modulation of their activity), each of which may act via discrete receptors. Finally, discovery of the PTHR2 suggested the presence of an additional ligand in the PTH family. Thus, the PTHR2, responsive only to PTH and not PTHrP, is expressed mainly in brain, placenta, testis, and pancreas. A novel PTH-like ligand for the PTHR2 has been isolated from the hypothalamus.135a PARATHYROID HORMONEINDUCED SIGNAL TRANSDUCTION ADENYLATE CYCLASE Adenylate cyclase stimulation and the subsequent generation of cAMP are believed to be important events in the actions of PTH in the kidney and in the skeleton. Cyclic adenosine monophosphate mimics phosphaturic and calcium-retaining effects of PTH in the kidney in vivo and in vitro and mediates PTH-stimulated renal 1a-hydroxylase activity.136 Cyclic adenosine monophosphate has also been implicated in the hypercalcemic action of PTH and may simulate PTH-induced bone resorption in vitro. Furthermore, the bone anabolic effect of PTH appears dependent on the capacity to stimulate cAMP. Cyclic AMP produced in target cells for PTH is believed to stimulate cAMP-dependent protein kinase isoenzymes; types I and II have different biochemical characteristics and may serve different functions. The isoenzymes are tetramers consisting of two regulatory and two catalytic subunits that have similar catalytic but different regulatory components, termed RI and RII, respectively. With binding of cAMP to the regulatory component, the holoenzyme dissociates, releasing the catalytic component that facilitates the transfer of a terminal phosphate group from a nucleotide donor, usually ATP, to an amino acid residue (i.e., serine, threonine, or tyrosine) of the substrate protein. PTH-induced stimulation of the two types of protein kinase has been demonstrated in normal osteoblasts and in a malignant osteoblast line.137 One of the substrates of PKA is the widely expressed cAMP response element binding protein (CREB).40 The phosphorylated CREB bound to the cAMP response element (CRE) in the promoters of responsive genes couples to the basal transcriptional machinery via cointegrators such as CREB-binding protein (CBP). PTH activates several osteoblastic genes, including c-fos,138 by this mechanism. OTHER SECOND MESSENGERS OF PARATHYROID HORMONE ACTION In addition to cAMP, other second messengers, such as the calcium ion, have been implicated as being potentially important in PTH action. In some species, transient hypocalcemia, presumably caused by calcium entry into bone cells, is the earliest event in the action of PTH on the skeleton in vivo. In vitro studies have shown that PTH promotes the uptake of calcium into isolated bone cells, that elevated calcium mimics or potentiates the effects of PTH on the enzymatic activities of isolated bone cells, and that calcium antagonists inhibit and calcium ionophores stimulate bone resorption. PTH stimulates phosphatidylinositol turnover in certain cell types and in renal membranes.139,140 In response to PTH-induced augmentation of phospholipase C action, presumably via Gq stimulation after occupancy of the PTH/PTHrP receptor, increased production of IP3 and diacylglycerol occurs. Increases in cytoplasmic calcium, presumably induced by IP3, have also been demonstrated, as has increased protein kinase C activity. The cellular response to PTH may therefore involve multiple mechanisms of cell signaling, and modulation of one message by another (cross talk) may affect the final response to the hormone. The relative contributions of adenylate cyclase versus phospholipase C stimulation to the pleiotropic effects of PTH in bone and kidney still remains to be clarified. EFFECTS OF PARATHYROID HORMONE IN TARGET TISSUES BONE Consistent with its prime function of raising the extracellular fluid calcium concentration, the most appreciated effect of PTH is a catabolic one in bone.141,142 The end result is the breakdown of mineral constituents and bone matrix, as manifested in vivo by the release of calcium and phosphate, by increases in plasma and urinary hydroxyproline, and by other indices of bone resorption. This process appears to be mediated by osteo-clastic osteolysis, but the mechanism by which PTH causes osteoclastic stimulation is indirect. Unlike calcitonin, PTH, when administered in vivo, does not bind directly to multinucleated osteoclasts. In vivo studies using autoradiography have revealed PTH binding to mature osteoblasts and to skeletal mononuclear cells in the intertrabecular region of the metaphysis (apparently to preosteoblastic stromal cells).143,144 Consequently, PTH-mediated increases in the number and function of osteoclasts (which are of hematogenous origin) appear to occur indirectly, through effects on cells of the osteoblast series (which are of mesenchymal origin).145 The results of these studies have been confirmed in experiments with the cloned PTH/PTHrP receptor; these have confirmed expression of the receptor in osteoblastic cells but not osteoclasts.146 PTH-induced stimulation of multinucleated osteoclasts occurs through the action of PTH-stimulated osteoblastic activity.147 This osteoblastic activity is characterized by release of intermediary factors such as cytokines. One of the most important of these intermediary factors is osteoclast differentiation factor (ODF), also known as osteoprotegerin ligand (OPGL).148 This protein is expressed in bone-lining cells or in osteoblast precursors that support osteoclast recruitment. ODF/OPGL, which is a member of the tumor necrosis factor (TNF) family, is identical to the molecule TRANCE/RANKL, which is expressed in T cells and previously was identified as a dendritic cell survival factor in vitro. Several bone resorbing factors, such as PTH, PTHrP, PGE2, some of the interleukins, and 1,25(OH)2D3, up-regulate ODF/OPGL gene expression in osteoblasts and bone stromal cells (Fig. 51-7). Interaction of ODF/OPGL with its receptor on osteoclast progenitors and osteoclasts then stimulates their recruitment and activation and delays their degradation. The ablation of the ODF/OPGL gene in mice has confirmed it to be a key regulator of osteoclastogenesis as well as of lymphocyte development and lymphnode organogenesis.149 Osteoprotegerin (OPG)/osteoclastogenesis inhibitory factor (OCIF)150 is a soluble receptor for ODF/OPGL that inhibits recruitment,

activation, and survival of osteoclasts and therefore inhibits osteoclastic bone resorption. OPG/OCIF acts as a natural decoy receptor to disrupt the interaction between ODF/OPGL, which is released by osteoblast-related cells, and the ODF/OPGL receptor on osteoclast progenitors.

FIGURE 51-7. Model of the activation of bone resorption and formation by parathyroid hormone (PTH) or parathyroid hormonerelated protein (PTHrP). PTH of PTHrP bind via their NH2 terminus to the PTH/PTHrP receptor (PTHR) of osteoblastic stromal cells or osteoblasts. Osteoblastic stimulation may lead to bone formation. PTH/PTHrP binding to the PTHR, most likely on the osteoblastic stromal cell, can also lead to the binding of ODF/OPGL/RANKL to receptors (RANK) on cells of the osteoclast series. This cytokine can then enhance commitment, proliferation, differentiation, and fusion of osteoclast precursors to form active bone-resorbing osteoclasts. Alternatively, ODF/OPGL/RANKL may be bound by OPG, preventing osteoclastic bone resorption.

Early and late phases of calcium mobilization have been described after in vivo administration of PTH.151 The early hypercalcemic response occurs from 10 minutes to 3 hours after hormone exposure. This response may be a consequence of increased metabolic activity of preexisting osteoclasts or of other cell types that enhance the transfer to bone of calcium in bone that is already in solution. A more sustained hypercalcemic response to PTH administration, occurring over approximately 24 hours, appears to depend on new protein synthesis and to involve a quantitative increase in osteoclasts, a change in the structure of the osteoclasts (i.e., increased ruffled borders, which is a zone of the cell believed to be involved in skeletal resorption); an increase in the secretion of lysosomal enzymes, including collagenase and acid hydrolases such as acid phosphatase; and acidification of the extracellular milieu of the osteoclast. In contrast to sustained administration of PTH, which causes bone resorption and hypercalcemia, low and intermittent doses of PTH-(1-34)and PTHrP-(1-34) and related analogspromote bone formation.152 Furthermore, in clinical studies, daily injections of PTH-(1-34) have been reported to increase hip and spine bone mineral density (BMD). The consequences of the effects of PTH on osteoblast activity are therefore complex. Thus, in addition to using osteoblastic cells to relay signals to the osteoclast lineage to resorb bone, PTH also appears to stimulate osteoblastic cells to enhance new bone formation. This might occur as a result of the ability of PTH to stimulate the production of growth factors, such as IGF-I, by osteoblasts. However, the bone-forming activity of PTH may depend more on promoting differentiation of precursor cells into secretory osteoblasts than on an action on mature osteoblasts.153 The precise mechanisms of the anabolic effect of PTH still remain to be clarified, however. The anabolic and catabolic effects of PTH on osteoblasts may therefore represent a combination of direct and indirect effects; effects of different domains of the PTH molecule that signal differently; discrete functions of morphologically similar but functionally distinct osteoblastic cells; or differences in hormonal effects based on different times of exposure or different hormone concentrations. KIDNEY One of the first-described effects of the administration of PTH intravenously was phosphaturia (see Chap. 206). Renal tubular reabsorption of phosphate is an active process, with a limited transport capacity resulting in a maximum rate of tubular reabsorption (Tm PO4).154 Because the absolute values of Tm PO4 vary considerably among individuals and most of this variation can be explained by differences in glomerular filtration rate (GFR), the Tm PO4/GFR ratio has been suggested as a more accurate index of phosphate reabsorption. This index, which represents the sum of the heterogeneous maximum reabsorption rates of all individual nephrons, is reduced by increased concentrations of PTH and increased in its absence. The molecular basis of the inhibition of Na+-dependent phosphate transport by PTH was clarified by the cloning of the Na+ -phosphate co-transporter, NPT-2.155 Thus, inhibition of phosphate reabsorption by PTH in the renal proximal tubule appears to involve PTH-induced endocytosis of NPT-2 and its subsequent intracellular degradation. Another relatively immediate effect of PTH that contributes to its role in calcium homeostasis is enhanced fractional reabsorption of calcium from the glomerular fluid. However, excessive circulating concentrations of PTH ultimately are associated with a rise in urinary calcium because of increases in extracellular calcium levels and therefore in the filtered load of calcium. The capacity of PTH to produce a mild renal tubular acidosis appears linked to its ability to inhibit the activation of the type 3 Na+-H+ exchanger (NHE3).156 Augmentation of urinary cAMP excretion in response to PTH administration is one of the earliest renal responses and is consistent with its postulated role as a second messenger for many or most renal responses. Nevertheless, evidence for an important role for inositol phosphates, protein kinase C, and intracellular calcium in the renal actions of PTH is also available; the relative in vivo contributions of the two pathways to the multiple effects of PTH in the kidney still requires clarification. The renal actions of PTH to increase urinary phosphate and cAMP excretion form the basis for the modified Ellsworth-Howard test that is used clinically to establish renal PTH responsiveness (see Chap. 60). PTH alters renal function because of its interactions with multiple regions of the nephron. Evidence for PTH binding to the primary foot processes of podocytes of renal corpuscles and for the stimulation of adenylate cyclase activity in rat glomeruli can be correlated with reduction of the GFR in rats by PTH as a result of decreasing the ultrafiltration coefficient.144 Evidence for the expression of the cloned PTH/PTHrP receptor in renal glomeruli confirms the previous findings. PTH appears to reach the luminal surface of polar tubular cells by glomerular filtration and the basolateral surface through the peritubular capillary plexus. High-capacity binding sites have been described on the periluminal portion of the earliest part of the proximal tubule, at which, related to the microvillar surface, PTH degradative events appear to occur. Saturable, specific, low-capacity binding sites for PTH have been localized after injection in rats to the basolateral surface of the proximal convoluted tubule and pars recta, the thick ascending limb of the loop of Henle, and the distal convoluted tubule and pars arcuata of the collecting duct.144 This pattern of PTH binding and activity correlates well with the expression of the cloned PTH/PTHrP receptor and with the localization of PTH-stimulated adenylate cyclase activity in rat nephrons, which was demonstrated in studies using microdissection of tubules.156 These observations emphasize the diversity of PTH actions on the renal tubule, most of which can be mimicked by the infusion of cAMP onto the luminal aspect of tubular cells.157,158 The use of various techniques, including micropuncture and microperfusion, has localized PTH-induced inhibition of phosphate reabsorption to the proximal convoluted tubule and to the pars recta. 158 Inhibition of phosphate reabsorption in the proximal convoluted tubule appears to be accompanied by inhibition of sodium and fluid reabsorption. However, sodium is also reabsorbed more distally. Inhibition of phosphate reabsorption also may occur, although perhaps to a lesser extent, in the distal tubule. The proximal tubule appears to be the major site of action of PTH in stimulating the 1a-hydroxylase and increasing the production of 1,25-dihydroxyvitamin D. The PTH-induced inhibition of bicarbonate transport occurs in the proximal tubule and the pars recta, and the effect of PTH on this nephron segment may explain the rise in urinary bicarbonate produced by PTH infusion.158 The important site of PTH action to increase calcium and probably magnesium transport appears to be in the thick ascending limb of the loop of Henle and in the distal convoluted tubule and earliest portion of the cortical collecting tubule. Despite the demonstration of the dual role of PTH in increasing extracellular fluid calcium concentrations through renal and skeletal routes, the relative importance of each is unclear.159 However, contributions of the kidney and skeleton to calcium homeostasis may depend on the concentration of circulating PTH and the duration of elevated hormonal levels.

OTHER TARGET TISSUES Although the administration of PTH in vivo can enhance intestinal calcium absorption, this effect appears to be indirect and mediated by the increased production of 1,25-dihydroxyvitamin D.92 Among other reported effects of PTH are direct effects on vascular tone, stimulation or inhibition of mitosis of various cells in vitro, increased concentrations of calcium in mammary and in salivary glands, enhanced hepatic gluconeogenesis, and enhanced lipolysis in isolated fat cells. The demonstration of the widespread expression of the PTHrP gene and the equally disseminated expression of the PTH/PTHrP receptor gene suggests that many of the noncalcemic actions of PTH may be carried out by PTHrP acting in an autocrine or paracrine manner.160

MEASUREMENT
RADIOIMMUNOASSAYS Radiommunoassays (RIAs) developed for human PTH generally do not use human PTH (hPTH) 184 as a tracer because of its scarcity and because it lacks a tyrosine residue for convenient radioiodination. Instead, bovine PTH (bPTH) 184 is used, possibly contributing to the reduced sensitivity for hPTH. The antisera used in PTH RIAs are generally polyclonal and have been raised against bPTH 184 or hPTH 184. Nevertheless, populations of antibodies within polyclonal antisera directed against specific epitopes contained in the 184 molecule may predominate. Antibody populations recognizing the COOH-terminal region of PTH 184 usually are readily obtained after immunization with bPTH 184; antibodies recognizing the midregion also occur with high frequency.161 Sufficiently sensitive antisera containing antibodies predominantly interacting with the NH2-terminal region are more unusual but have been successfully raised.162 Attempts to direct antibody specificity have used several strategies. Although the development of monoclonal antibodies to PTH or PTH fragments would seem to be the most direct route to achieve specificity, this approach has not been successful for various technical and biochemical reasons. Instead, specificity has been achieved by using polyclonal antisera with predominant specificity for selected regions, as determined by their reactivity toward synthetic fragments of discrete regions of the molecule or by enhancing the specificity of such antisera by using synthetic fragments of the midregion or COOH-terminal end of PTH as tracers. An important advance in the measurement of PTH is the development of immunoradiometric assays (IRMAs) that use antisera directed against the NH2-terminal and midregion or COOH region of the molecule. These assays (i.e., intact assays) detect mainly intact PTH 184 and appear to be the most sensitive and specific.163 Defining the specificity of an RIA is important because of the complex metabolism of PTH, which results in a multiplicity of circulating molecular forms. The most abundant circulating forms are midregion and COOH fragments because of their longer half-life in the circulation; these become even more predominant in renal failure when clearance is further impaired.89,90 The presumed bioactive forms, intact PTH 184 and the NH2-terminal fragment, are cleared more readily and circulate at lower levels. Midregion and COOH-directed RIAs detect long-lived fragments and intact bioactive hormone. These RIAs generally do not require the high sensitivity of NH2-directed assays to be useful, because they measure higher absolute levels of PTH. Most material measured by these RIAs is thought to be inactive, but this has not greatly restricted their clinical usefulness in the diagnosis of primary or secondary hyperparathyroidism. For primary hyperparathyroidism, the midregion and COOH-directed RIAs may be especially useful in providing an index of integrated secretory function of the parathyroid glands.85 However, first it should be established that a given midregion or COOH-terminal RIA is clinically useful. Amino-terminal RIAs, which determine the level of short-lived PTH forms, can also be used to detect hypersecretion in primary hyperparathyroidism but may be especially useful in assessing acute changes in PTH secretion in physiologic studies. Two-site IRMAs, which also measure bioactive hormone, are generally more sensitive and can substitute for NH2-directed assays. For cases of hyperparathyroidism associated with renal failure, NH2-directed assays have the theoretical advantage of providing a better estimate of circulating bioactive PTH forms and differentiating decreased PTH clearance from hypersecretion. From a practical point of view, two-site IRMAs are preferred to NH2-directed assays. If PTH levels are determined serially during the progression of renal failure, COOH-directed assays also may provide useful estimates of the severity of the hyperparathyroidism, and hormone levels determined by COOH-directed assays in renal failure have correlated well with the extent of skeletal resorption.85 It has been suggested that midregion assays may detect intact PTH more readily than COOH-directed assays and may have a somewhat greater ability to measure a variety of PTH forms than the COOH-directed assays. However, midregion assays are subject to restrictions and advantages similar to those of the COOH-directed assays. Few RIAs, no matter the degree of specificity, completely discriminate between concentrations of immunoreactive PTH found in healthy persons and those detected in hyperparathyroidism, although best results appear to be obtained with two-site IRMAs. It has been suggested that PTH levels should be interpreted in conjunction with prevailing levels of blood calcium. This assists in the assay analysis; if a given PTH level is measured in a normocalcemic individual and that same PTH level is found in a patient with hypercalcemia, the level in the latter case can be considered inappropriately elevated. Another limitation of most RIAs is the inability to measure values in all healthy individuals. A lower limit of normality generally cannot be established, which makes it difficult to ascertain reduced PTH levels. Sensitive two-site IRMAs often approach a lower limit of normality. PTH RIAs are useful in differentiating hypocalcemia resulting from PTH deficiency (i.e., hypoparathyroidism) from hypocalcemia resulting from other causes. In the former situation, PTH is undetectable or is found in very low concentrations, but in the latter situation, the hypocalcemic stimulus is associated with increased PTH secretion and elevated levels of PTH (see Chap. 58 and Chap. 61). BIOASSAYS With the unraveling of the complex metabolism of PTH and the resulting heterogeneity of circulating hormone, the care that must be exercised in interpreting the results of PTH RIAs became apparent, and the possibility of using PTH bioassays to supplement the information obtained from RIAs arose. Some PTH bioassays estimate the biologic effects of the hormone in vivo, and some estimate levels of the hormone based on biologic effects in vitro. It has been estimated that the normal circulating levels of bioactive PTH are ~1 pmol/L. IN VIVO BIOASSAYS Measurements of in vivo phosphaturic effects (e.g., determination of Tm PO4/GFR) and renal calcium retention (e.g., fractional excretion of calcium) provide estimates of PTH effects in vivo and have been useful as adjunctive studies of PTH. Partly because of their lack of specificity, these assays have not achieved widespread popularity as indices of PTH concentrations in most clinical situations. Conversely, estimates of urinary cAMP excretion (measured by RIA) have been fairly widely used.164 Several hormones in addition to PTH (e.g., vasopressin) contribute to urinary cAMP levels, although the PTH-produced component is the major fraction. Greater specificity of urinary cAMP for PTH can be achieved by measuring the fraction of urinary cAMP that is of renal origin. This nephrogenous component can be determined from the clearance of cAMP relative to the GFR, an estimate requiring plasma cAMP measurements, or it can be determined by relating the urinary cAMP to 100 mL of glomerular filtrate. Nephrogenous cAMP is a specific and rather sensitive in vivo bioassay for PTH.164 Nephrogenous cAMP is often elevated in primary hyperparathyroidism and decreased in hypoparathyroidism. There is, however, overlap with the normal range. Most nephrogenous cAMP determinations accurately reflect circulating PTH levels as determined by RIA, but they do not have the sensitivity or specificity of the best two-site IRMAs. IN VITRO BIOASSAYS Several attempts have been made to develop clinically useful in vitro bioassays for the measurement of active PTH. The capacity of PTH to stimulate adenylate cyclase, with the addition of guanyl nucleotides, in purified renal membrane preparations or tumor cell lines and the availability of synthetic antagonistic PTH fragments have imparted useful specificity to such renal membrane bioassay preparations.165,166 Unfortunately, the relative insensitivity of this approach has precluded its use for anything other than experimental purposes. Another approach is the cytochemical PTH bioassay. The most widely used has been a renal cytochemical bioassay (CBA) performed in guinea pig kidney segments maintained in non-proliferative organ culture.4,91,167 The major drawbacks of the method are its technical difficulty and low throughput, which have precluded its use for routine clinical assay purposes.

CONCLUSIONS
Considerable progress has been made in examining the biosynthesis of PTH, the regulation of PTH secretion, and the mechanism of PTH action. These advances include elucidation of the molecular genetics of PTH, discovery of the parathyroid calcium sensor, and determination of the structure of the PTH/PTHrP receptor. Discovery of the ODF/OPGL/RANKL system and the cloning of NPT-2 have greatly contributed to the understanding of the mechanism of PTH action in bone and kidney. The identification of PTHrP has disclosed a new member of the PTH gene family and provided possibilities for understanding the biologic effects of both molecules. This improved comprehension of the biochemistry and physiology of PTH should translate into exciting insights into the pathophysiology of disease states in which PTH is implicated. CHAPTER REFERENCES
1. Kronenberg HM, Igarashi T, Freeman MW, et al. Structure and expression of the human parathyroid hormone gene. Recent Prog Horm Res 1986; 42:641. 2. Cohn DV, Elting J. Biosynthesis, processing and secretion of parathormone and secretory protein-I. Recent Prog Horm Res 1983; 39:181. 3. Walter P, Ibrahimi I, Blobel G. Translocation of proteins across the endoplasmic reticulum. 1. Signal recognition protein (SRP) binds to in vitroassembled polysomes synthesizing secretory protein. J Cell Biol 1981; 91:545. 4. Goltzman D, Bennett HPJ, Koutsilieris M, et al. Studies of the multiple forms of bioactive parathyroid hormone and parathyroid hormone-like substances. Recent Prog Horm Res 1986; 42:665. 5. Hendy GN, Bennett HPJ, Lazure C, et al. Proparathyroid hormone (ProPTH) is preferentially cleaved to parathyroid hormone (PTH) by the prohormone convertase furin: a mass spectrometric study. J Biol Chem 1995; 270:9517. 6. Canaff L, Bennett HPJ, Hou Y, et al. Proparathyroid hormone processing by the proprotein convertase PC7: comparison with furin and assessment of modulation of parathyroid convertase mRNA levels by calcium and 1,25-dihydroxy vitamin D3. Endocrinology 1999; 140:3633. 7. Rabbani SA, Kremer R, Bennett HPJ, Goltzman D. Phosphorylation of parathyroid hormone by human and bovine parathyroid glands. J Biol Chem 1984; 259:2949. 8. Hendy GN, Kronenberg HM, Potts JT Jr, Rich A. Nucleotide sequence of cloned DNAs encoding human preproparathyroid hormone. Proc Natl Acad Sci U S A 1981; 78:7365. 9. Vasicek TJ, McDevitt BE, Freeman MW, et al. Nucleotide sequence of genomic DNA encoding human parathyroid hormone. Proc Natl Acad Sci U S A 1983; 80:2127. 10. Weaver CA, Gordon DF, Kissil MS, et al. Isolation and complete nucleotide sequence of the gene for bovine parathyroid hormone. Gene 1984; 28:319. 11. Heinrich G, Kronenberg HM, Potts JT Jr, Habener JF. Gene encoding parathyroid hormone. Nucleotide sequence of the rat gene and deduced amino acid sequence of rat preproparathyroid hormone. J Biol Chem 1984; 259:3320. 12. Russell J, Sherwood LM. Nucleotide sequence of the DNA complementary to avian (chicken) preproparathyroid hormone mRNA and the deduced sequence of the hormone precursor. Mol Endocrinol 1989; 3:325. 13. Yasuda T, Banville D, Hendy GN, Goltzman D. Characterization of the human parathyroid hormonelike peptide gene: functional and evolutionary aspects. J Biol Chem 1989; 264:7720. 14. Mangin M, Ikeda K, Dreyer BE, Broadus AE. Isolation and characterization of the human parathyroid hormonelike peptide gene. Proc Natl Acad Sci U S A 1989; 86:2408. 15. Suva LJ, Mather KA, Gillespie MT, et al. Structure of the 5' flanking region of the gene encoding human parathyroid hormonerelated protein (PTHrP). Gene 1989; 77:95. 16. Karaplis AC, Yasuda T, Hendy GN, et al. Gene encoding parathyroid hormone like peptide: nucleotide sequence of the rat gene and comparison with the human homologue. Mol Endocrinol 1990; 4:441. 17. Naylor SL, Sakaguchi AY, Shows TB, et al. Human parathyroid hormone gene (PTH) is on the short arm of chromosome 11. Somatic Cell Mol Genet 1983; 9:609. 18. Seldin MF, Mattei M-G, Hendy GN. Localization of mouse parathyroid hormone-like peptide gene (Pthlh) to distal chromosome 6 using interspecific backcross mice and in situ hybridization. Cytogenet Cell Genet 1992; 60:252. 19. Pausova Z, Morgan K, Fujiwara TM, et al. Molecular characterization of an intragenic minisatellite (VNTR) polymorphism in the human parathyroid hormonerelated peptide gene in chromosome region 12p12.1p11.2. Genomics 1993; 17:243. 20. Keutmann HT, Sauer MM, Hendy GN, et al. Complete amino acid sequence of human parathyroid hormone. Biochemistry 1978; 17:5723. 21. Motokura T, Bloom T, Kim HG, et al. A novel cyclin encoded by bcl1-linked candidate oncogene. Nature 1991; 350:512. 22. Chandrasekharappa SC, Guru SC, Manickam P, et al. Positional cloning of the gene for multiple endocrine neoplasia-type 1. Science 1997; 276:404. 23. European Consortium on MEN1. Identification of the multiple endocrine neoplasia type 1 (MEN1) gene. Hum Mol Genet 1997; 6:1177. 24. Guru SC, Goldsmith PK, Burns AL, et al. MENIN, the product of the MEN1 gene, is a nuclear protein. Proc Natl Acad Sci U S A 1998; 95:1630. 25. Agarwal SK, Guru SC, Heppner C, et al. Menin interacts with the AP1 transcription factor JunD and represses JunD-activated transcription. Cell 1999; 96:143. 26. Friedman E, Sakaguchi K, Bale AE, et al. Clonality of parathyroid tumors in familial multiple endocrine neoplasia type 1. N Engl J Med 1989; 321:213. 27. Thakker RV, Bouloux P, Wooding C, et al. Association of parathyroid tumors in multiple endocrine neoplasia type 1 with loss of alleles on chromosome 11. N Engl J Med 1989; 321:218. 28. Arnold A, Horst SA, Gardella TJ, et al. Mutation of the signal peptide-encoding region of the preproparathyroid hormone gene in familial isolated hypoparathyroidism. J Clin Invest 1990; 86:1084. 29. Parkinson DB, Thakker RV. A donor splice site mutation in the parathyroid hormone gene is associated with autosomal recessive hypoparathyroidism. Nature Genet 1992; 1:149. 30. Simpson EL, Mundy GR, D'Souza SM, et al. Absence of parathyroid hormone messenger RNA in nonparathyroid tumors associated with hypercalcemia. N Engl J Med 1983; 309:325. 31. Nussbaum SR, Gaz RD, Arnold A. Hypercalcemia and ectopic secretion of parathyroid hormone by an ovarian carcinoma with rearrangement of the gene for parathyroid hormone. N Engl J Med 1990; 323:1324. 32. Fischer JA, Oldham SB, Sizemore GW, Arnaud CD. Calcium-regulated parathyroid hormone peptidase. Proc Natl Acad Sci U S A 1972; 69:2341. 33. Flueck JA, DiBella FP, Edis AJ, et al. Immunoheterogeneity of parathyroid hormone in venous effluent serum from hyperfunctioning parathyroid glands. J Clin Invest 1977; 60:1367. 34. Hanley DA, Takatsuki K, Sultan JM, et al. Direct release of parathyroid hormone fragments from functioning bovine parathyroid glands in vitro. J Clin Invest 1978; 62:1247. 35. Mayer GP, Keaton JA, Hurst JG, Habener JF. Effect of plasma calcium concentrations on the relative proportion of hormone and carboxyl fragments in parathyroid venous blood. Endocrinology 1979; 104:1778. 36. Kemper B, Habener JF, Rich A, Potts JT Jr. Parathyroid secretion: discovery of a major calcium-dependent protein. Science 1974; 184:167. 37. Cohn DV, Zangerle R, Fischer-Colbrie R, et al. Similarity of secretory protein-I from parathyroid gland to chromogranin A from adrenal medulla. Proc Natl Acad Sci U S A 1982; 79:6056. 38. Bhargava G, Russell J, Sherwood LM. Phosphorylation of parathyroid secretory protein. Proc Natl Acad Sci U S A 1983; 80:878. 39. Mouland AJ, Bevan S, White JH, Hendy GN. Human chromogranin A gene; molecular cloning, structural analysis and neuroendocrine cell-specific expression. J Biol Chem 1994; 269:6918. 40. Canaff L, Bevan S, Wheeler D, et al. Analysis of molecular mechanisms controlling neuroendocrine cell specific transcription of the chromogranin A gene. Endocrinology 1998; 139:1184. 41. el-Hajj Fuleihan G, Klerman EB, Brown EN, et al. The parathyroid hormone circadian rhythm is truly endogenousa general clinical research center study. J. Clin Endocrinol Metab 1997; 82:281. 42. Samuels MH, Veldhais JD, Kramer P, et al. Episodic secretion of parathyroid hormone in postmenopausal women: assessment by deconvolution analysis and approximate entropy. J Bone Miner Res 1977; 12:616. 43. Mayer GP, Hurst JG. Sigmoidal relationship between parathyroid hormone secretion rate and plasma concentration in calves. Endocrinology 1978; 102:1036. 44. Shoback DM, Thatcher J, Leombruno R, Brown EM. Relationship between parathyroid hormone secretion and cytosolic calcium concentration in dispersed bovine parathyroid cells. Proc Natl Acad Sci U S A 1984; 81:3113. 45. Brown EM, Redgrave J, Thatcher J. Effect of the phorbol ester TPA on PTH secretion. Evidence for a role for protein kinase C in the control of PTH release. FEBS Lett 1984; 175:72. 46. Brown EM, Gambda G, Riccardi D, et al. Cloning and characterization of an extracellular Ca 2+-sensing receptor from bovine parathyroid. Nature 1993; 366:575. 47. Pollak MR, Brown EM, Chou Y-HW, et al. Mutations in the human Ca2+-sensing receptor gene cause familial hypocalciuric hypercalcemia and neonatal severe hyperparathyroidism. Cell 1993; 75:1237. 48. Heath H III, Odelberg S, Jackson CE, et al. Clustered inactivating mutations and benign polymorphisms of the calcium receptor gene in familial benign hypocalciuric hypercalcemia suggest receptor functional domains. J Clin Endocrinol Metab 1996; 81:1312. 49. Cole DEC, Janicic N, Salisbury SR, Hendy, GN. Neonatal severe hyperparathyroidism, secondary hyperparathyroidism, and familial hypercalciuric hypercalcemia: multiple different phenotypes associated with an inactivating Alu insertion mutation of the calcium-sensing receptor gene. Am J Med Genet 1997; 71:202. 50. Pollak MR, Chou Y-HW, Marx SJ, et al. Familial hypocalciuric hypercalcemia and neonatal severe hyperparathyroidism. Effects of mutant gene dosage on phenotype. J Clin Invest 1994; 93:1108. 51. Bai M, Pearce SHS, Kifor O, et al. In vivo and in vitro characterization of neonatal hyperparathyroidism resulting from a de novo, heterozygous mutation in the Ca 2+-sensing receptor gene: normal maternal calcium homeostasis as a cause of secondary hyperparathyroidism in familial benign hypocalciuric hypercalcemia. J Clin Invest 1997; 99:88. 52. Pearce SHS, Williamson C, Kifor O, et al. A familial syndrome of hypocalcemia with hypercalciuria due to mutations in the calcium-sensing receptor. N Engl J Med 1996; 335:1115. 53. Cole DEC, Peltekova VD, Rubin LA, et al. A986S polymorphism of the calcium-sensing receptor and circulating calcium concentrations. Lancet 1999; 353:112. 54. Nemeth EF, Steffey ME, Hammerland LG, et al. Calcimimetics with potent and selective activity on the parathyroid calcium receptor. Proc Natl Acad Sci U S A 1998; 95:4040. 55. Silverberg SJ, Bone GH III, Marriott TB, et al. Short-term inhibition of parathyroid hormone secretion by a calcium-receptor agonist in patients with primary hyperparathyroidism. N Engl J Med 1997; 337:1506. 56. Kovacs CS, Ho-Pao CL, Hunzelman JL, et al. Regulation of murine fetal-placental calcium metabolism by the calcium-sensing receptor. J Clin Invest 1998; 101:2812. 57. Brown AJ, Zhong M, Ritter C, et al. Loss of calcium responsiveness in cultured bovine parathyroid cells is associated with decreased calcium receptor expression. Biochem Biophys Res Comm 1995; 212:861. 58. Mithal A, Kifor O, Kifor I, et al. The reduced responsiveness of cultured bovine parathyroid cells to extracellular Ca2+ is associated with marked reduction in the expression of extracellular Ca(2+)-sensing receptor messenger ribonucleic acid and protein. Endocrinology 1995; 136:3087. 59. Russell J, Lettieri D, Sherwood LM. Direct regulation by calcium of cytoplasmic messenger ribonucleic acid coding for pre-proparathyroid hormone in isolated bovine parathyroid cells. J Clin Invest 1983; 72:1851. 60. Russell J, Sherwood LM. The effects of 1,25(OH)2 D3 and high calcium on transcription of the pre-proparathyroid hormone gene are direct. Trans Assoc Am Physicians 1987; 100:256. 61. Moallem E, Kilav R, Silver J, Naveh-Many T. RNA-protein binding and posttranscriptional regulation of parathyroid hormone gene expression by calcium and phosphate. J Biol Chem 1998; 273:5253. 62. Brown EM. Four-parameter model of the sigmoidal relationship between parathyroid hormone release and extracellular calcium concentration in normal and abnormal parathyroid tissue. Endocrinology 1983; 56:572. 63. Kifor O, Moore FD, Wang P, et al. Reduced immunostaining for the extracellular Ca2+-sensing receptor in primary and secondary hyperparathyroidism. J Clin Endocrinol Metab 1996; 81:1598. 64. Ho C, Conner DA, Pollack MR, et al. A mouse model of human familial hypocalciuric hypercalcemia and neonatal severe hyperparathyroidism. Nature Genet 1995; 11:389. 65. Habener JF, Potts JT Jr. Relative effectiveness of magnesium and calcium on the secretion and biosynthesis of parathyroid hormone in vitro. Endocrinology 1976; 98:197. 66. Rude RK, Oldham SB, Sharp CF Jr, Singer FR. Parathyroid hormone secretion in magnesium deficiency. J Clin Endocrinol Metab 1978; 47:800. 67. Cholst IN, Steinberg SF, Tropper PJ, et al. The influence of hypermagnesemia on serum calcium and parathyroid hormone levels in human subjects. N Engl J Med 1984; 310:1221. 68. Morrissey J, Rothstein M, Mayer G, Slatopolsky E. Suppression of parathyroid hormone by aluminum. Kidney Int 1983; 23:699. 69. Sherwood LM, Mayer GP, Ramberg DF, et al. Regulation of parathyroid hormone secretion: proportional control by calcium, lack of effect of phosphate. Endocrinology 1968; 83:1043. 70. Almaden Y, Canelejo A, Hernandez A, et al. Direct effect of phosphorous on PTH secretion from whole rat parathyroid glands in vitro. J Bone Miner Res 1996; 11:970. 71. Slatopolsky E, Finch J, Denda M, et al. Phosphorous restriction prevents parathyroid gland growth: high phosphorous directly stimulates PTH secretion in vitro. J Clin Invest 1996; 97:2534. 72. Abe M, Sherwood LM. Regulation of parathyroid hormone secretion by adenylate cyclase. Biochem Biophys Res Commun 1972; 48:396. 73. Brown EM, Gardner DG, Windek RA, Aurbach GD. Relationship of intracellular 3',5'-adenosine monophosphate accumulation to parathyroid hormone release from dispersed bovine parathyroid cells. Endocrinology 1978; 103:2323. 74. Heath H III. Biogenic amines and the secretion of parathyroid hormone and calcitonin. Endocr Rev 1980; 1:319. 75. Epstein S, Heath H III, Bell NH. Lack of influence of isoproterenol, propranolol and dopamine on immunoreactive parathyroid hormone and calcitonin in normal man. Calcif Tissue Int 1983; 35:32. 76. Fasciotto BH, Gorr S-U, De Franco DJ, et al. Pancreastatin, a presumed product of chromogranin-A (secretory protein-1) processing, inhibits secretion from porcine parathyroid cells in culture. Endocrinology 1989; 125:1617.

77. Drees BM, Rouse J, Johnson J, Hamilton JW. Bovine parathyroid glands secrete a 26-kDa N-terminal fragment of chromogranin-A which inhibits parathyroid cell secretion. Endocrinology 1991; 129:3381. 78. Silver J, Russell J, Sherwood LM. Regulation by vitamin D metabolites of messenger ribonucleic acid for preproparathyroid hormone in isolated bovine parathyroid cells. Proc Natl Acad Sci U S A 1985; 82:4270. 79. Hendy GN, Stotland MS, Grunbaum D, et al. Characteristics of secondary hyperparathyroidism in vitamin D deficient dogs. Am J Physiol 1989; 256:E765. 80. Mackey SL, Heymont JL, Kronenberg HM, Demay MB. Vitamin D receptor binding to the negative human parathyroid hormone vitamin D response element does not require the retinoid x receptor. Molec Endocrinol 1996; 10:298. 81. Kremer R, Bolivar I, Goltzman D, Hendy GN. Influence of calcium and 1,25-dihdroxycholecalciferol on proliferation and protooncogene expression in primary cultures of bovine parathyroid cells. Endocrinology 1989; 125:935. 82. Fucik RF, Krukeja SC, Hargis GK, et al. Effect of glucocorticoids on function of the parathyroid glands in man. J Clin Endocrinol Metab 1975; 40:152. 83. Silverman R, Yalow RS. Heterogeneity of parathyroid hormone: clinical and physiologic implications. J Clin Invest 1973; 52:1958. 84. Segre GV, Niall HD, Habener JF, Potts JT Jr. Metabolism of parathyroid hormone: physiological and clinical significance. Am J Med 1974; 56:774. 85. Arnaud CD, Goldsmith RS, Bordier PJ, et al. Influence of immunoheterogeneity of circulating parathyroid hormone on radioimmunoassays of serum in man. Am J Med 1974; 56:785. 86. Reiss E, Canterbury JM. Emerging concepts of the nature of circulating parathyroid hormones: implications for clinical research. Recent Prog Horm Res 1974; 30:391. 87. Neuman WF, Neuman MW, Lane K, et al. The metabolism of labeled parathyroid hormone V. Collected biological studies. Calcif Tissue Res 1975; 18:271. 88. Martin KJ, Hruska KA, Freitag JJ, et al. The peripheral metabolism of parathyroid hormone. N Engl J Med 1979; 301:1092. 89. Segre GV, D'Amour P, Hultman A, Potts JT Jr. Effects of hepatectomy, nephrectomy, and nephrectomy/uremia on the metabolism of parathyroid hormone in the rat. J Clin Invest 1981; 67:439. 90. Papapoulos SE, Hendy GN, Tomlinson S, et al. Clearance of exogenous parathyroid hormone in normal and uraemic man. Clin Endocrinol 1977; 7:211. 91. Goltzman D, Henderson B, Loveridge N. Cytochemical bioassay of parathyroid hormone: characteristics of the assay and analysis of circulating hormonal forms. J Clin Invest 1980; 65:1309. 92. DeLuca HF. Recent advances in the metabolism of vitamin D. Ann Rev Physiol 1981; 43:199. 93. Parsons JA, Rafferty B, Gray D, et al. Pharmacology of parathyroid hormone and some of its fragments and analogues. In: Talmage RV, Owen M, Parsons JA, eds. Calcium-regulating hormones. Amsterdam: Excerpta Medica, 1975:33. 94. Tregear GW, van Rietschoten J, Greene E, et al. Bovine parathyroid hormone: minimum chain length of synthetic peptide required for biological activity. Endocrinology 1973; 93:1349. 95. Kemp BE, Moseley JM, Rodda CP, et al. Parathyroid hormonerelated protein of malignancy: active synthetic fragments. Science 1987; 23:1568. 96. Horiuchi N, Caulfield MP, Fisher JE, et al. Similarity of synthetic peptide from human tumor to parathyroid hormone in vivo and in vitro. Science 1987; 23:1566. 97. Rabbani SA, Mitchell J, Roy DR, et al. Influence of the amino-terminus on in vitro and in vivo biological activity of synthetic parathyroid hormone like peptides of malignancy. Endocrinology 1988; 123:2709. 98. Goltzman D, Peytremann A, Callahan E, et al. Analysis of the requirements for parathyroid hormone action in renal membranes with the use of inhibiting analogues. J Biol Chem 1975; 250:3199. 99. Nussbaum SR, Rosenblatt M, Potts JT Jr. Parathyroid hormone renal receptor interactions: demonstration of two receptor binding domains. J Biol Chem 1980; 255:10183. 100. Abou-Samra A-B, Uneno S, Jppner H, et al. Non-homologous sequences of parathyroid hormone and the parathyroid hormone related peptide bind to a common receptor on ROS 17/2.8 cells. Endocrinology 1989; 125:2215. 101. Caulfield MP, McKee RL, Goldman ME, et al. The bovine renal parathyroid hormone (PTH) receptor has equal affinity for two different amino acid sequences: The receptor binding domains of PTH and PTH-related protein are located within the 14-34 region. Endocrinology 1990; 127: 83. 102. Klaus W, Dieckmann T, Wray V, et al. Investigation of the solution structure of the human parathyroid hormone fragment (1-34) by 1H NMR spectroscopy, distance geometry, and molecular dynamics calculations. Biochemistry 1991; 30:6936. 103. Cohen FE, Strewler GJ, Bradley MS, et al. Analogues of parathyroid hormone modified at positions 3 and 6: Effects on receptor binding and activation of adenylyl cyclase in kidney and bone. J Biol Chem 1991; 266:1997. 104. Barden JA, Kemp BE. Stabilized NMR structure of the hypercalcemia of malignancy peptide PTHrP [Ala-26](1-34) amide. Biochim Biophys Acta 1994; 1208:256. 105. Wray V, Federau T, Gronwald W, et al. The structure of human parathyroid hormone from a study of fragments in solution using 1H NMR spectroscopy and its biological implications. Biochemistry 1994; 33:1684. 106. Neugebauer W, Barbier J-R, Sung WL, et al. Solution structure and adenylyl cyclase stimulating activities of C-terminal truncated human parathyroid hormone analogues. Biochemistry 1995; 34:8835. 107. Maretto S, Mammi S, Bissacco E, et al. Mono- and bicyclic analogs of parathyroid hormone-related protein. 2. Conformational analysis of antagonists by CD, NMR, and distance geometry calculations. Biochemistry 1997; 36:3300. 108. Barbier J-R, Neugebauer W, Morley P, et al. Bioactivities and secondary stuctures of constrained analogues of human parathyroid hormone: cyclic lactams of the receptor binding region. J Medicinal Chem 1997; 40:1373. 109. Vickery BH, Avnur Z, Cheng Y, et al. RS-66271, a C-terminally substituted analog of human parathyroid hormone-related protein (1-34), increases trabecular and cortical bone in ovariectomized, osteopenic rats. J Bone Miner Res 1996; 11:1943. 110. Juppner H, Abou-Samra AB, Freeman MW, et al. A G proteinlinked receptor for parathyroid hormone and parathyroid hormonerelated peptide. Science 1991; 254:1024. 111. Abou-Samra AB, Jppner H, Force T, et al. Expression cloning of a parathyroid hormone/parathyroid hormone-related peptide receptor from rat osteoblast-like cells: a single receptor stimulates intracellular accumulation of both cAMP and inositol triphosphate and increases intracellular free calcium. Proc Natl Acad Sci U S A 1992; 89:2732. 112. Pausova Z, Bourdon J, Clayton D, et al. Cloning of a parathyroid hormone/parathyroid hormone-related peptide receptor (PTHR) cDNA from a rat osteosarcoma (UMR 106) cell line: chromosomal assignment of the gene in the human, mouse, and rat genomes. Genomics 1994; 20:20. 113. Schipani E, Harga H, Karaplis AC, et al. Identical complementary deoxyribonucleic acids encode a human renal and bone parathyroid hormone (PTH)/PTH-related peptide receptor. Endocrinology 1993; 132:2157. 114. Fraher LJ, Hodsman AB, Jonas K, et al. A comparison of the in vivo biochemical responses to exogenous parathyroid hormone-(134) [PTH-(134)] and PTH-related peptide-(134) in man. J Clin Endocrinol Metab 1992; 75:417. 114a. Fraher LJ, Avram R, Watson PH, et al. Comparison of the biochemical responses to human parathyroid hormone-(1-31)NH 2 and hPTH-(1-34) in healthy humans. J Clin Endocrinol Metab 1999; 84:2739. 115. McCuaig KA, Clarke JC, White JH. Molecular cloning of the gene encoding the mouse parathyroid hormone/parathyroid hormone related peptide receptor. Proc Natl Acad Sci U S A 1994; 91:5051. 116. Bettoun JD, Minagawa M, Hendy GN, et al. Developmental upregulation of human parathyroid hormone (PTH)/PTH-related peptide receptor gene expression from conserved and human-specific promoters. J Clin Invest 1998; 102:958. 116a. Bettoun JD, Kwan MY, Minagawa M, et al. Methylation patterns of human parathyroid hormone (PTH)/PTH-related peptide receptor gene promoters are established several weeks prior to onset of their function. Biochem Biophys Res Commun 2000; 267:482. 117. Urena P, Kong X-F, Abou-Samra A-B, et al. Parathyroid hormone (PTH)/PTH-related peptide receptor messenger ribonucleic acids are widely distributed in rat tissues. Endocrinology 1993; 133:617. 118. Usdin TB, Gruber C, Bonner TI. Identification and functional expression of a receptor selectively recognizing parathyroid hormone, the PTH2 receptor. J Biol Chem 1995; 270:15455. 119. Lee CW, Gardella TJ, Abou-Samra AB, et al. Role of the extracellular regions of the PTH/PTHrP receptor in hormone-binding. Endocrinology 1994; 135:1488. 120. Lee CW, Luck MD, Jppner H, et al. Homolog-scanning mutagenesis of the parathyroid hormone (PTH) receptor reveals PTH-(1-34) binding determinants in the third extracellular loop. Mol Endocrinol 1995; 9:1269. 121. Iida KA, Guo J, Takemura M, et al. Mutations in the second cytoplasmic loop of the rat parathyroid hormone (PTH)/PTH-related protein receptor result in selective loss of PTH-stimulated phospholipase C activity. J Biol Chem 1997; 272:6882. 122. Huang Z, Chen Y, Nissenson RA. The cytoplasmic tail of the G protein-coupled receptor for parathyroid hormone and parathyroid hormone-related protein contains positive and negative signals for endocytosis. J Biol Chem 1996; 270:151. 123. Mahoney CA, Nissenson RA. Canine renal receptors for parathyroid hormone: down-regulation in vivo by exogenous parathyroid hormone. J Clin Invest 1983; 72:411. 124. Forte LR, Langeluttig SG, Poelling RE, Thomas ML. Renal parathyroid hormone receptors in the chick: downregulation in secondary hyperparathyroid animal models. Am J Physiol 1982; 242:E154. 125. Tomlinson S, Hendy GN, Pemberton DM, O'Riordan JLH. Reversible resistance to the renal action of parathyroid hormone in man. Clin Sci Mol Med 1976; 51:59. 126. Schipani E, Weinstein LS, Bergwitz C, et al. Pseudohypoparathyroidism type 1b is not caused by a defect in the coding exons of the human parathyroid hormone (PTH)/PTH-related peptide gene. J Clin Endocrinol Metab 1995; 80:1611. 127. Bettoun JD, Minagawa M, Kwan MY, et al. Cloning and characterization of the promoter regions of the human parathyroid hormone (PTH)/PTH-related peptide receptor gene: analysis of deoxyribonucleic acid from normal subjects and patients with pseudohypoparathyroidism type 1b. J Clin Endocrinol Metab 1997; 82:1031. 128. Jppner H, Schipani E, Bastepe M, et al. The gene responsible for pseudohypoparathyroidism type 1b is paternally imprinted and maps in four unrelated kindreds to chromosome 20q13.3. Proc Natl Acad Sci U S A 1998; 95:11798. 129. Yu S, Yu D, Lee E, et al. Variable and tissue-specific hormone resistance in heterotrimeric Gs protein a-subunit (Gs a) knockout mice is due to tissue-specific imprinting of the G s a gene. Proc Natl Acad Sci U S A 1998; 95:8715. 130. Hayward BE, Moran V, Strain L, Bonthron DT. Bidirectional imprinting of a single gene: GNAS1 encodes maternally, paternally, and biallelically derived proteins. Proc Natl Acad Sci U S A 1998; 95:15475. 131. Jansen M. ber atypische Chondrodystrophie (Achondroplasie) und ber eine noch nicht beschriebene angeborene Wachstumsst;rung des Knochen-systems: Metaphysre Dysostosis. Orthop Chir 1934; 61:253. 132. Schipani E, Langman CB, Parfitt AM, et al. Constitutively activated receptors for parathyroid hormone-related peptide in Jansen's metaphyseal chondrodysplasia. N Engl J Med 1996; 335:708. 133. Blomstrand S, Claesson I, Save-Soderbergh J. A case of lethal congenital dwarfism with accelerated skeletal maturation. Pediatr Radiol 1985; 15:141. 134. Zhang P, Jobert A-S, Couvineau A, Silve C. A homozygous inactivating mutation in the parathyroid hormone/parathyroid hormone-related peptide receptor causing Blomstrand chondrodysplasia. J Clin Endocrinol Metab 1998; 83:3365. 135. Jobert A-S, Zhang P, Couvineau A, et al. Absence of functional receptors for parathyroid hormone and parathyroid hormone-related peptide in Blom-strand chondrodysplasia. J Clin Invest 1998; 102:34. 135a. Usdin TB, Hoare SR, Wang T, et al. TIP39: a new neuropeptide and PTH2-receptor agonist from hypothalamus. Nature Neuroscience 1999; 2:941. 136. Brenza HL, Kimmel-Jehan C, Jehan F, et al. Parathyroid hormone activation of the 25-hydroxyvitamin D3-1 alpha-hydroxylase gene promoter. Proc Natl Acad Sci U S A 1998; 95:1387. 137. Livesey SA, Kemp BE, Re CA, et al. Selective hormonal activation of cyclic AMP-dependent protein-kinase isoenzymes in normal and malignant osteoblasts. J Biol Chem 1982; 257:14983. 138. Pearman AT, Chou WY, Bergman KD, et al. Parathyroid hormone induces c-fos promoter activity in osteoblastic cells through phosphorylated cAMP response element (CRE)-binding protein to the major CRE. J Biol Chem 1996; 271:25715. 139. Stern PH. Cationic agonists and antagonists of bone resorption. In: Cohn DV, Fujita T, Potts JT Jr, Talmage RV, eds. Endocrine control of bone and calcium metabolism. Amsterdam: Excerpta Medica, 1984:109. 140. Hruska KA, Moskowitz D, Esbrit P, et al. Stimulation of inositol triphosphate and diacylglycerol production in renal tubular cells by parathyroid hormone. J Clin Invest 1987; 79:230. 141. Bingham P, Brazell I, Owen M. The effect of parathyroid extract on cellular activity and plasma calcium levels in vivo. J Endocrinol 1969; 45:387. 142. Holtrop ME, Raisz LG, Simmons HA. The effect of parathyroid hormone, colchicine and calcitonin on the ultrastructure and activity of osteoblasts in organ culture. J Cell Biol 1974; 60:346. 143. Rouleau MF, Mitchell J, Goltzman D. In vivo distribution of parathyroid hormone receptors in bone. Evidence that a predominant osseous target cell is not the mature osteoblast. Endocrinology 1988; 123:187. 144. Rouleau MF, Warshawsky H, Goltzman D. Parathyroid hormone binding in vivo to renal, hepatic and skeletal tissues of the rat using a radioautographic approach. Endocrinology 1986; 118:919. 145. Rizzoli RE, Somerman M, Murray TM, Aurbach GD. Binding of radioiodinated parathyroid hormone to cloned bone cells. Endocrinology 1983; 113:1832.

146. 147. 148. 149. 150. 151. 152. 153. 154. 155. 156. 157. 158. 159. 160. 161. 162. 163. 164. 165. 166. 167.

Amizuka N, Karaplis AC, Henderson JE, et al. Haploinsufficiency of parathyroid hormone-related peptide (PTHrP) results in abnormal postnatal bone development. Dev Biol 1996; 175:166. Takahashi N, Akatsu T, Udagawa N, et al. Osteoblastic cells are involved in osteoclast formation. Endocrinology 1988; 123:2600. Lacey DL, Timms E, Tan HL, et al. Osteoprotegerin ligand is a cytokine that regulates osteoclast differentiation and activation. Cell 1998; 93:165. Kong Y-Y, Yoshida H, Sarosi L, et al. OPGL is a key regulator of osteoclastogenesis, lymphocyte development and lymphnode organogenesis. Nature 1999; 397:315. Simonet WS, Lacey DL, Dunstan CR, et al. Osteoprotegerin: A novel secreted protein involved in the regulation of bone density. Cell 1997; 89:309. Parsons JA, Potts JT Jr. Physiology and chemistry of parathyroid hormone. In: MacIntyre I, ed. Clinics in endocrinology and metabolism, vol 1. Calcium metabolism and bone disease. London: WB Saunders, 1972:33. Tam CS, Heersche JNM, Murray TM, Parsons JA. Parathyroid hormone stimulates the bone apposition rate independently of its resorptive action: differential effects of intermittent and continual administration. Endocrinology 1982; 110:506. Corral DA, Amling M, Priemel M, et al. Dissociation between bone resorption and bone formation in osteopenic transgenic mice. Proc Natl Acad Sci U S A 1998; 95:13835. Bijvoet OLM. Kidney function in calcium and phosphate metabolism. In: Avioli LV, Krane SM, eds. Metabolic bone disease, vol 1. New York: Academic Press, 1977:48. Hartmann CM, Hewson AS, Kos CH, et al. Structure of murine and human renal type II Na+-phosphate cotransporter genes (Npt2 and NPT2). Proc Natl Acad Sci U S A 1996; 93:7409. Azarani A, Goltzman D, Orlowski J. Structurally diverse N-terminal peptides of parathyroid hormone (PTH) and PTH-related peptide (PTHRP) inhibit the Na +/H+ exchanger NHE3 isoform by binding to the PTH/PTHRP receptor type I and activating distinct signalling pathways. J Biol Chem 1996; 271:14931. Morel F. Regulation of kidney functions by hormones: a new approach. Recent Prog Horm Res 1983; 39:271. Agus ZS, Wasserstein A, Goldfarb S. PTH, calcitonin, cyclic nucleotides and the kidney. Annu Rev Physiol 1981; 43:583. Nordin BEC, Peacock M. Role of the kidney in regulation of plasma calcium. Lancet 1969; 2:1280. Amizuka N, Warshawsky H, Henderson JE, et al. Parathyroid hormone-related peptidedepleted mice show abnormal epiphyseal cartilage development and altered endochondral bone formation J Cell Biol 1994; 126:1611. Marx SJ, Sharp ME, Krudy A, et al. Radioimmunoassay for the middle region of human parathyroid hormone: studies with a radioiodinated synthetic peptide. J Clin Endocrinol Metab 1981; 53:76. Papapoulos SE, Manning RM, Hendy GN, et al. Studies of circulating parathyroid hormone in man using a homologous amino-terminal specific immunoradiometric assay. Clin Endocrinol 1980; 13:57. Nussbaum SR, Zahnadnik RJ, Labigne JR, et al. A highly sensitive two-site immunoradiometric assay of parathyrin (PTH) and its clinical utility in evaluating patients with hypercalcemia. Clin Chem 1987; 33:1364. Broadus AE. Nephrogenous cyclic AMP. Recent Prog Horm Res 1981; 37:667. Nissenson RA, Abbott SR, Teitelbaum AP, et al. Endogenous biologically active human parathyroid hormone: measurement by a guanyl nucleotide-amplified renal adenylate cyclase assay. J Clin Endocrinol Metab 1981; 52:840. Sato K, Han DC, Ozawa M, et al. A highly sensitive bioassay for PTH using ROS 17/2.8 subclonal cells. Acta Endocrinol (Copenh) 1987; 116:113. Chambers DJ, Dunham J, Zanelli JM, et al. A sensitive bioassay of parathyroid hormone in plasma. Clin Endocrinol 1978; 9:375.

CHAPTER 52 PARATHYROID HORMONE-RELATED PROTEIN Principles and Practice of Endocrinology and Metabolism

CHAPTER 52 PARATHYROID HORMONE-RELATED PROTEIN

GORDON J. STREWLER Structure and Biologic Properties Structure and Processing of Parathyroid Hormone-Related Protein and its Gene Secreted Forms: Peptide Heterogeneity Immunoassays Parathyroid Hormone-Related Protein Receptors Biologic Properties: Action on Bone and Kidney Role of Parathyroid Hormone- Related Protein in Hypercalcemia Tumors Secreting Parathyroid Hormone-Related Protein Pathogenetic Role of Parathyroid Hormone-Related Protein Use of Parathyroid Hormone Related Protein Assays in the Differential Diagnosis of Hypercalcemia Role of Parathyroid Hormone Related Protein in Physiology Cartilage Breast Skin Teeth Smooth Muscle Uterus and Placenta Chapter References

Parathyroid hormonerelated protein (PTHrP), sometimes referred to as parathyroid hormonelike protein, is the sister of PTH. Originally identified as the cause of humoral hypercalcemia in malignancy, PTHrP has a distinct set of physiologic functions that are unrelated to the regulation of systemic calcium homeostasis but rival the actions of PTH in their importance. It has been recognized since Fuller Albright's time that in some ways patients with malignant tumors causing hypercalcemia resemble patients with primary hyperparathyroidism (pHPT). Malignancy-associated hypercalcemia is characterized not only by humorally mediated bone resorption but also by diminished tubule resorption of phosphate with consequent phosphaturia and hypophosphatemia. When it was recognized that the disorder also produced an increase in nephrogenous cyclic adenosine monophosphate (cAMP), which is the component of urinary cAMP secreted into the urine from the renal tubule,1 it became clear that a humoral factor in patients with malignancy-associated hypercalcemia was mimicking PTH at the kidney. Increased nephrogenous cAMP was previously thought to be unique to hyperparathyroidism, reflecting increased secretion of cAMP from the renal tubule, where cAMP is the intracellular second messenger for PTH. Once it was understood that increased cAMP concentrations in kidney or bone cells could be used as a bioassay to detect the humoral factor secreted by malignant tumors, the substance responsible for malignancy-associated hypercalcemia was purified and shown to be a protein that was structurally related to PTH.2,3 and 4

STRUCTURE AND BIOLOGIC PROPERTIES

STRUCTURE AND PROCESSING OF PARATHYROID HORMONE-RELATED PROTEIN AND ITS GENE PTHrP is a protein of 139 to 173 amino acids.5,6 and 7 It resembles PTH in primary sequence only at the aminoterminus, where 8 of the first 13 amino acids in the two peptides are identical (Fig. 52-1). Although this is a limited region of homology, it is a critical region of both peptides; it is not required for binding but it is necessary for receptors occupied by either hormone to activate adenylate cyclase and hence to produce cAMP as a second messenger. Thus, the homology of PTH and PTHrP at the aminoterminal end is responsible for their shared biologic properties (see later in this chapter).

FIGURE 52-1. Primary structures of the aminoterminal part of parathyroid hormonerelated protein and of parathyroid hormone. Compared are the human sequences 134. Identical amino acids are highlighted.

The gene for PTHrP is located on human chromosome 12. It is considerably more complex than the PTH gene, which is located on chromosome 11, with three promoters and four alternatively spliced exons upstream of the coding sequence.8 The complexity of the promoter region allows for considerable flexibility in the regulation of PTHrP gene transcription. There is evidence that the promoters are used differentially in different tissues8 and by the viral transactivating protein tax.9 Most of the prepro sequence of PTHrP is encoded on one exon, with the last 2 amino acids of the prepro sequence and 139 amino acids of the mature peptide encoded on a second exon. The splice junction between these two coding exons is precisely conserved between PTHrP and PTH. Downstream of the main coding sequence are two additional exons, which are alternatively spliced to contribute distinct carboxyterminal ends to the isoforms of PTHrP and distinct 3' noncoding sequences. The protein isoforms encoded by these exons are identical through amino acid 139 but comprise 139, 141, or 173 amino acids in toto. This downstream complexity of the PTHrP gene may also be used for regulation. The three transcript families with different 3' untranslated sequences have markedly different halflives, ranging from 20 minutes to ~7 hours. In addition, their stability is differentially regulated. Exposure of cells to transforming growth factor-b specifically increases the halflife of the most unstable transcript, which encodes the 139 amino acid form of the protein. It is clear from conservation of sequence, gene structure, and chromosomal localization that PTHrP and PTH arose from a common ancestral gene. In addition to the striking resemblance of their aminoterminal sequence and the conservation of the intronexon boundaries between the two major coding exons, each gene is flanked by duplicated genes for lactate dehydrogenase. It is believed that chromosome 12, on which the PTHrP gene is located, and chromosome 11, where the PTH gene resides, arose by an ancient duplication in which the ancestral PTH/PTHrP gene was evidently copied along with its neighbors. SECRETED FORMS: PEPTIDE HETEROGENEITY The three mature PTHrP peptides predicted by cDNA are 139, 141, and 173 amino acids long. It is not clear whether any of them are secreted as intact proteins, although it seems likely they are. AMINOTERMINAL FRAGMENTS The molecular size of PTHrP containing the bioactive aminoterminal part found in plasma and in tumor extracts ranges anywhere from 6 to 18 kDa,2,3 and 4,10 and, apparently, several different aminoterminal fragments of PTHrP are secreted.11 Three different cell types (renal carcinoma cells, PTHrP-transfected RIN-141 cells, and normal keratinocytes) use prohormone convertases to cleave PTHrP after arginine-37 to produce an aminoterminal fragment that is probably processed to PTHrP(136)12 (Fig. 52-2). This aminoterminal PTHrP fragment contains the PTH-like region and could account completely for the PTH-like effects of PTHrP, as

discussed later.

FIGURE 52-2. To p, Structural features of PTHrP. The PTH-homologous domain PTHrP(113) is delineated by hatched lines ( ), and potential cleavage sites are shown as vertical lines or crosshatched regions ( ). Middle, Peptides known or postulated to be derived from PTHrP. Bottom, Proven or postulated sites of action of individual PTHrP peptides are shown beneath each peptide.

MIDREGIONAL FRAGMENTS The intracellular cleavage step at arginine-37 also produces midregional PTHrP fragments of ~50 to 70 amino acids beginning with alanine-3812 (see Fig. 52-2). The carboxytermini of these fragments are located at amino acids 94, 95, and 101 in the highly basic region PTHrP(88106).13 Midregional fragments of PTHrP generated intracellularly can apparently enter the regulated pathway of peptide secretion and become packaged in dense neurosecretory granules, suggesting that they may be under separate secretory control from aminoterminal fragments.12 Midregional PTHrP fragments of similar size are the predominant circulating forms in the plasma of patients withhumoral hypercalcemia of malignancy.14 These midregional PTHrPs play a physiologic role in transplacental calcium transport (see later), and synthetic midregion peptides also appear to have bioactivities in squamous carcinoma cells. CARBOXYTERMINAL FRAGMENTS The size of the midregional fragments suggests the existence of additional carboxyterminal PTHrP fragments resulting from cleavage at the multibasic amino acid clusters at amino acids 88106. Because the most carboxyterminal cleavage site is at amino acid 101, which precedes the lysine-arginine cluster at PTHrP(102106), it is likely that carboxyterminal peptides such as PTHrP(107-139) are produced and secreted. Peptides with PTHrP(109138) but not PTHrP(136) immunoreactivity have been found in sera of patients with humoral hypercalcemia of malignancy, and a peptide with PTHrP(109138) immunoreactivity circulates as a separate peptide that accumulates in patients with chronic renal failure.10,15,16 The high degree of conservation between human, rat, and chicken PTHrP in the portion up to amino acid 111 suggests some biologic relevance of such peptides. The very carboxyterminal part of the conserved region, PTHrP(107111), was found in several studies to be a potent inhibitor of osteoclastic bone resorption in vitro,17,18 and has been given the name osteostatin, although this result has not been confirmed in all bone resorption assays.19 The corresponding synthetic peptides (PTHrP[107111] and PTHrP[107139]) induce calcium transients in hippocampal neurons.20 PTHrP is posttranslationally modified in additional ways21; for example, PTHrP secreted by keratinocytes is glycosylated posttranslationally.22 Whether this is also the case with other cell types and with PTHrP forms circulating in plasma is unknown. The findings that cells process PTHrP to multiple peptide fragments, that some of these fragments may be under separate secretory control, and that one or more (in addition to its PTH-like aminoterminus) has its own bioactivity indicate that PTHrP is, like the adrenocorticotropic hormone-melanocyte-stimulating hormoneendorphin precursor pro-opiomelano-cortin, a polyprotein precursor for multiple bioactive peptides.23 The pattern of posttranslational modification may be tissue-specific, and this would add another dimension of specificity to the secreted forms of PTHrP. IMMUNOASSAYS Because the known PTH-like bioactivity of PTHrP is located within the first 34 amino acids, most immunoassays have been directed against the aminoterminal sequence 134 to 140 of PTHrP in radioimmunoassays (RIAs)24,25 and 26 or against larger fragments containing aminoterminal PTHrP in immunoradiometric assays.16,27 Because the amounts of aminoterminal PTHrP extractable from normal serum are very low (<2.5 pmol/L),26 prior extraction steps are often required to obtain clinically useful results with RIAs for aminoterminal PTHrP.25,26 Typically, aminoterminal PTHrP is elevated in 60% to 80% of patients with hypercalcemia of malignancy and is undetectable in most healthy subjects (normal range, <2.5 pmol/L; Fig. 52-3).

FIGURE 52-3. Plasma concentrations of parathyroid hormonerelated protein (PTHrP) in patients with hyperparathyroidism (HPT), normocalcemic patients with malignancy (Normocalc), and patients with hypercalcemia of malignancy resulting from a solid tumor (Solid) or a hematologic malignancy (Hematol). Radioimmunoassay (RIA) for aminoterminal PTHrP(134) (left panel), an immunoradiometric assay (IRMA) for PTHrP(174) (middle panel), and an RIA for midregion PTHrP(5384) (right panel). The hatched area represents the normal ranges, the dotted line the limits of detection, the numbers attached to each group indicate the number of patients. In the PTHrP(174) assay, the group Solid includes five patients classified as local osteolytic type of hypercalcemia (D) and two patients with lymphoma. Note the different scales of the Y-axes. (Modified from Budayr AA, Nissenson RA, Klein RF, et al. Increased serum levels of a parathyroid hormone-like protein in malignancy-associated hypercalcemia. Ann Intern Med 1989; 111:807; Burtis WJ, Brady TJ, Orloff JJ, et al. Immunochemical characterization of circulating parathyroid hormone-related protein in patients with humoral hypercalcemia of cancer. N Engl J Med 1990; 322:1106; and Blind E, Raue F, Gtzmann J, et al. Circulating levels of midregional parathyroid hormone-related protein in hypercalcaemia of malignancy. Clin Endocrinol [Oxf] 1992; 37:290.)

As previously mentioned, a midregional PTHrP peptide is generated intracellularly and secreted. A RIA directed against a midregional epitope within the 5384 region of PTHrP detects elevated PTHrP concentrations in the sera of ~80% of patients with hypercalcemia of malignancy.28 The mean concentration of midregional PTHrP in these patients is ~50 pmol/L (see Fig. 52-3); this is ~10-fold higher than aminoterminal PTHrP concentrations.25,26 Results with RIAs against this portion of PTHrP appear to be variable.29,30 A RIA directed against a PTHrP epitope for the carboxyterminal (109138) detected elevated levels in humoral hypercalcemia of malignancy but also revealed highly elevated levels in patients without malignancy who had chronic renal failure.16 The 109138 immunoreactive species appears to circulate as a separate carboxyterminal peptide.10,15,16 Assays with increased sensitivity and specificity have been developed by applying the two-site immunoradiometric assay technique to detect larger fragments containing aminoterminal and midregional PTHrP epitopes using PTHrP(174) to (186) standards.16,27 Such two-site assays are now the preferred method for determining levels of circulating PTHrP. In patients with humoral hypercalcemia of malignancy, two-site assays usually detect elevated PTHrP serum concentrations in at least 80% of cases, with an average PTHrP concentration of 21 pmol/L (normal range, <5 pmol/L) in a typical assay (see Fig. 52-3), which is a range of concentrations similar to that detected by aminoterminal RIAs. Although such assays would fail to detect some bioactive aminoterminal fragments of PTHrP (e.g., PTHrP[136]), their performance in comparison to aminoterminal RIAs in detecting PTHrP in patients with humoral hypercalcemia suggests that such fragments rarely

predominate in tumor patients. Even the most sensitive assays of this type (reporting a lower limit of detection of 0.2 pmol/L) fail to detect PTHrP in at least half of normal people,27,31 in contrast to the much higher normal range found for intact PTH in serum (16 pmol/L). Direct comparison of PTHrP levels in different immunoassays shows heterogeneity of circulating forms secreted by individual tumors, which is consistent with the peptide's heterogeneity. Whether there are secretion patterns typical for certain tumors remains to be seen. Assay samples require the addition of proteinase inhibitors and rapid handling to avoid proteolysis after collection.24,31 These assays are helpful in diagnosing humoral hypercalcemia of malignancy but not yet sufficiently sensitive to allow firm conclusions to be drawn about the presence or absence of PTHrP in the circulation of people under physiologic conditions. PARATHYROID HORMONE-RELATED PROTEIN RECEPTORS Humoral hypercalcemia of malignancy shares several biochemical features with primary hyperparathyroidism (pHPT), closely reflecting the similarities of PTH and PTHrP action on bone and kidney. PTH and PTHrP bind with similar affinities to a shared receptor on osteoblast-like cells,32,33 and 34 canine renal membranes,33,35 and other tissues, owing to their homology in the aminoterminal portion (see Fig. 52-1). This adenylate cyclase-coupled receptor is a member of the family of G-protein coupled receptors36 and is required for the developmentally crucial actions of PTHrP on the development of cartilage and bone, because ablation of the PTH/PTHrP receptor by homologous recombination produces a similar phenotype to ablation of the PTHrP gene itself,37 and expression of a constitutively active receptor produces a phenotype similar to overexpression of PTHrP.38,39 In the brain,40 in insulinoma cells,41 and in skin,42 there is evidence for receptors that are specific for aminoterminal PTHrP and do not recognize PTH. A receptor that is specific for PTH and does not recognize native PTHrP has been identified in brain by molecular cloning,43 and the basis for its ligand specificity has been partially clarified.44,45 The observations that midregion PTHrP peptides have their own unique bioeffects (e.g., in the placenta,46,47 as discussed later) clearly indicate that a separate receptor for midregion PTHrP must also exist. There is also evidence of a carboxyterminal PTHrP receptor in brain20 and putatively in bone, where the carboxyterminal fragment of PTHrP (sometimes called osteostatin) appears to inhibit bone resorption.18,48 Finally, as discussed later, some effects of PTHrP may result from direct intracrine actions in the nucleus of cells that secrete PTHrP,49,50 and thus are possibly not mediated by a cell-surface receptor at all. BIOLOGIC PROPERTIES: ACTION ON BONE AND KIDNEY Aminoterminal PTHrP is able to elicit the classic effects of PTH on target cells of calcium homeostasis. A number of studies have shown nearly identical effects with similar doses of aminoterminal fragments of PTH and PTHrP(134 to 140 region) on bone cells, such as stimulation of adenylate cyclase,51,52 and 53 inhibition of alkaline phosphatase and growth, and activation of the intracellular calcium second messenger system.54,55 In animals and in humans, aminoterminal PTHrP peptides reproduce the phosphaturic, hypocalciuric effects of PTH and activate the synthesis of 1,25-dihydroxyvitamin D.51,53,56,57

ROLE OF PARATHYROID HORMONE- RELATED PROTEIN IN HYPERCALCEMIA

TUMORS SECRETING PARATHYROID HORMONE-RELATED PROTEIN The clinical manifestations of hypercalcemia in malignancy are discussed in Chapter 59. Here, evidence is presented on the involvement and the pathogenetic role of PTHrP. SOLID TUMORS Elevated serum levels of PTHrP are found in ~80% of patients with solid tumors and hypercalcemia.16,24,25,26,27 and 28,30,30a PTHrP levels are almost always elevated in patients with hypercalcemia in the absence of bone metastasis (humoral hypercalcemia) and in solid tumors with squamous histologic features.16,58 PTHrP secretion by tumor tissue should precede the development of hypercalcemia. Indeed, PTHrP elevations can be detected in a small percentage of cancer patients when they are still normocalcemic.16,59 HUMORAL COMPONENT OF HYPERCALCEMIA It was long assumed that in patients with radiologic evidence of extensive skeletal destruction by malignancy (especially in breast cancer), local osteolytic hypercalcemia was the mechanism responsible for hypercalcemia. A humoral form of the syndrome (humoral hypercalcemia of malignancy) was defined as hypercalcemia without bone metastasis. However, PTHrP levels are elevated with similar frequency in hypercalcemic patients with and without bone metastasis. In addition, the propensity of a given tumor type (e.g., lung cancer) to metastasize to bone is poorly correlated with the frequency of hypercalcemia; for example, tumors with small-cell histologic characteristics have the highest frequency of bone metastasis but rarely cause hypercalcemia. Thus, PTHrP can produce humoral hypercalcemia even in patients with bone metastases, and the diagnosis of humoral hypercalcemia in malignancy rests largely on the demonstration of elevated plasma PTHrP levels. BREAST CANCER Hypercalcemia in breast cancer is most often seen in patients with extensive bone metastasis, but in 50% to 90%16,24,26 of these patients, the serum PTHrP level is also elevated.59 The finding that in breast cancer, hypercalcemia can occur in patients with or without increased circulating PTHrP levels suggests that hypercalcemia in breast carcinoma can have either a humoral or a local osteolytic cause. Fifty percent to 60% of breast cancer tissue is immunocytochemically positive for PTHrP,59,60 but metastases in bone have been found to be PTHrP-positive (92%) more often than at other sites (17%).61 This suggests that PTHrP may cause either local osteolytic hypercalcemia, when secreted at low levels from bone metastases, or humoral hypercalcemia, when secreted systemically at higher levels. Moreover, PTHrP-positive tumors may be more likely to metastasize to bone, possibly because of a survival advantage that results from the local osteolytic action of PTHrP. This possibility has been confirmed in an animal model of local osteolysis, in which introduction of PTHrP cDNA into breast cancer cells markedly increases the prevalence of lytic bone metastases.62 This may have therapeutic consequences, because such a subgroup of patients could be treated prophylactically with bisphosphonates.63 In addition, secretion of PTHrP may be activated in bone metastases by the local effects of bone matrix growth factors, such as transforming growth factor b, as shown in an animal model.64 Approximately a century ago, Sir James Paget proposed that both seed and soil are important in bone metastasis, and the recent findings with regard to PTHrP illustrate this adage. HEMATOLOGIC MALIGNANCIES PTHrP is rarely involved in the pathogenesis of hypercalcemia in hematologic malignancies. Multiple myeloma and lymphoma generally cause hypercalcemia by a local osteolytic mechanism, but humoral secretion of PTHrP is evident in a small fraction of hypercalcemic patients.24,26,65 However, in the adult T-cell leukemia/lymphoma syndrome, which is caused by infection with the retrovirus human T-cell leukemia virus type 1 (HTLV-1), PTHrP-related hypercalcemia occurs in more than one-half of patients.66,67,68,69 and 70 The tax protein encoded in the HTLV-1 genome can directly activate the PTHrP promoter,9,71 and PTHrP expression in T lymphocytes may also be activated indirectly by HTLV-induced cytokines.72,73

PATHOGENETIC ROLE OF PARATHYROID HORMONE-RELATED PROTEIN

Patients with humoral hypercalcemia of malignancy as well as normal persons infused with aminoterminal fragments of PTHrP57,74 and animals treated with synthetic PTHrP show many of the hallmark effects of excess PTH, such as hypercalcemia, hypophosphatemia, hyperphosphaturia, elevated nephrogenous cAMP levels, and stimulation of osteoclastic bone resorption. PTHrP levels and calcium in plasma are positively correlated in patients with humoral hypercalcemia of malignancy, as are PTHrP levels and urinary cAMP concentrations.16 Furthermore, the infusion of antibodies against PTHrP is able to reverse hypercalcemia and other manifestations in experimental humoral hypercalcemia of malignancy, including models in which human tumor cells produce hypercalcemia in nude mice.75,76 Therefore, not only is PTHrP capable of producing hypercalcemia, but also secretion of PTHrP is necessary to develop hypercalcemia in such models. To produce hypercalcemia in patients with hypercalcemia of malignancy, PTHrP may act in concert with bone-resorbing cytokines. Evidence from animal studies suggests that bone-resorbing factors such as tumor necrosis factor a may modulate hypercalcemia in humoral hypercalcemia of malignancy,77 although there is little evidence that such factors by themselves cause humoral hypercalcemia of malignancy. Some discrepancies exist between the clinical pictures of humoral hypercalcemia of malignancy and the PTH excess in pHPT with regard to bone and kidney. Although

PTH and PTHrP act similarly on osteoblast-like cells in vitro, bone formation appears to be relatively lower in humoral hypercalcemia of malignancy than in pHPT,78 although bone resorption is increased in both cases. Depressed bone formation in humoral hypercalcemia of malignancy may be caused by differences of action of PTHrP compared with PTH; but it may also result from the release of additional substances from tumors, such as cytokines, or it may be more indirectly caused by less specific effects of severe malignant disease. Another discrepancy is the level of 1,25-dihydroxyvitamin D in the circulation: 1,25-dihydroxyvitamin D levels are often elevated in patients with pHPT but tend to be close to or below the lower limit of normal in hypercalcemia of malignancy.1 Suppressed levels of 1,25-dihydroxyvitamin D in humoral hypercalcemia of malignancy (despite the presence of high PTHrP levels) rebound during calcium-lowering therapy, suggesting a dominance of inhibition of the renal 1a-hydroxylase by high serum calcium levels over the stimulatory effect of PTHrP.79,80 As an underlying pathophysiologic mechanism, the existence of factors inhibiting the renal 1a-hydroxylase released by tumors has been suggested.67 Hypercalcemia of malignancy usually responds well to potent inhibitors of bone resorption, such as bisphosphonates, but some studies have shown that these substances are less effective in patients with increased serum PTHrP levels than in those with local osteolytic hypercalcemia.81,82 Humoral hypercalcemia may be caused not only by bone resorption but also by increased calcium reabsorption by the kidney, owing to the hypocalciuric effect of PTHrP. The weaker response of patients with high PTHrP to bisphosphonates may be attributable to the effect of PTHrP-mediated renal calcium reabsorption on hypercalcemia in these patients, an effect that is not influenced by bisphosphonates. Additionally, high PTHrP levels may result in a more powerful stimulation of bone resorption, requiring higher doses of bisphosphonates.82

USE OF PARATHYROID HORMONE RELATED PROTEIN ASSAYS IN THE DIFFERENTIAL DIAGNOSIS OF HYPERCALCEMIA
PTHrP, as measured in immunoassays available today, is elevated in most patients with hypercalcemia caused by a solid tumor (see Fig. 52-3); an elevated PTHrP level can make a positive diagnosis of hypercalcemia of malignancy in ~80% of these patients. PTHrP levels are normal in almost all other causes of hypercalcemia, with rare exceptions, such as mammary hyper-plasia (see later). However, to discriminate between pHPT and hypercalcemia of malignancy, the measurement of serum intact PTH is the variable of first choice. Although hypercalcemia is usually a complication of advanced malignancies that are obvious, a determination of intact PTH should be obtained on all patients because pHPT, which is a prevalent disorder, can present as an intercurrent illness in patients with malignancies. The demonstration of a suppressed or low-normal intact PTH level (<20 pg/mL) can exclude this possibility.

ROLE OF PARATHYROID HORMONE RELATED PROTEIN IN PHYSIOLOGY

Unlike PTH, whose expression is highly specific to the parathyroid glands, PTHrP is expressed in many tissues of fetuses and adults. These tissues include cartilage, bone, breast, skin, skeletal, heart and smooth muscle, uterus, and placenta. In addition, several endocrine organssuch as the fetal83,84 and adult parathyroid,85 fetal thyroid,83 and pancreatic islets86 contain PTHrP transcripts or protein. In the central nervous system, PTHrP is expressed in neurons of the hippocampus, cerebral cortex, and cerebellar cortex.87 Using the polymerase chain reaction, it has been shown that PTHrP expression is even more widespread.88 This, together with the finding that the circulating level of PTHrP is significantly lower than that of PTH, suggests that PTHrP has protean tissue-specific functions, some of which are described here. CARTILAGE The best understood action of PTHrP in normal physiology is as a regulator of cartilage growth and differentiation during the process of endochondral bone formation. The abolition of PTHrP gene expression by a targeted mutation of both PTHrP alleles produces embryonic lethality as the result of a severe disorder of cartilage development.89 The limb bones are short, and there is a marked reduction in the width of the zone of proliferation of the growth plate, suggesting that the proliferation of growth plate chondrocytes is grossly impaired. In addition, endochondral bones in the skull base are heavily mineralized at birth, and the cranial vault is misshapen. The ribs are also prematurely mineralized, with a resultant decrease in the circumference of the chest cavity. Thus, the absence of PTHrP is associated with defective proliferation of chondrocytes and accelerated maturation to the hypertrophic stage, in which chondrocytes mineralize their matrix. The converse phenotype is obtained when PTHrP is overexpressed in cartilage from a transgene driven by the cartilage-specific type II collagen promoter:90 chondrocyte maturation is impaired, the growth plate is widened, and immature chondrocytes may be found within trabecular bone. To carry out its actions in cartilage, PTHrP uses the shared PTH/PTHrP receptor, as evidenced by the similarity of phenotypes when PTHrP and the PTH/PTHrP receptor are ablated in mice37 or in a rare human chondrodysplasia, the Blomstrand type.91,92 Conversely, mutations that constitutively activate the PTH/PTHrP receptor inhibit chondrocyte maturation38 and also induce life-long hypercalcemia, as reported in the human disorder Jansen-type metaphysial chondrodysplasia.39 As predicted from the sharing of a common receptor, PTH has most of the cartilage effects of PTHrP. PTH is a powerful mitogen for proliferating growth plate chondrocytes,93,94 and it inhibits the mineralization of cartilage.95 Yet PTH cannot substitute for the absence of PTHrP in PTHrP knockout mice. The common receptor in avascular cartilage may be relatively inaccessible to circulating PTH, compared with higher levels of PTHrP produced locally, and there may also be fewer PTH/PTHrP receptors in cartilage than in the primary target tissues of PTH, bone, and kidney. PTHrP functions in endochondral bone formation as part of a feedback loop. Chondrocytes, which have just exited from the proliferative phase and begun to hypertrophy and terminally differentiate, produce the morphogen Indian hedgehog, a member of the hedgehog family of morphogens that regulate many important steps in vertebrate morphogenesis. Via its own receptor in the perichondrial layer surrounding the developing bone, Indian hedgehog signals for the secretion of PTHrP from proliferating chondrocytes, thus up-regulating the proliferation of cartilage cells and keeping them out of the terminal differentiation program.37,96 The effect of the Indian hedgehog-PTHrP system is to relay a signal from mature cells back to the zone of decision. This feedback loop regulates the rate of entry into the differentiation pathway and thereby ensures that proliferation and maturation of chondrocytes is balanced and that cells are synchronized as they enter the last phase of their life, so that the linear growth of bone is orderly. This conclusion is supported by studies of chimeric mice in which clones of PTH/PTHrP receptor (/) cells are present in developing cartilage. These cells differentiate prematurely, activate the PTHrP-Indian hedgehog axis, and slow the differentiation of surrounding normal chondrocytes.97 To make available for study PTHrP knockout mice, they have been rescued from lethality by crossing heterozygotes for the PTHrP null genotype with transgenic mice in which expression either of PTHrP98 or of a constitutively active PTH/PTHrP receptor38 has been directed to cartilage. This produces some offspring, which express PTHrP or its constitutively active receptor in cartilage only, thus rescuing them from the fatal chondrodysplasia, while allowing study of the PTHrP null phenotype in other tissues. BREAST Breast development in the PTHrP knockout mouse does not progress beyond the earliest stages of ingrowth of the mammary epithelial rudiment, which normally expresses PTHrP, into the underlying mesenchyme, which normally expresses its receptor98 (see Fig. 52-2). This suggests that PTHrP is one of the signals in the complex interactions of epithelium and mesenchyme that underlie breast development. Restoration of PTHrP to mammary epithelium can be accomplished by a strategy similar to the rescue of PTHrP knockout mice, by crossing rescued PTHrP knockout mice with mice that express PTHrP constitutively in the mammary epithelium,99 thus resulting in a mouse that produces PTHrP only in cartilage (thus, ensuring its survival) and the ingrowing mammary rudiment. This strategy restores the ingrowth of the breast bud, but the nipple still does not develop normally.98 It should be possible to analyze the role of PTHrP in epithelial-mesenchymal signaling in detail by reconstituting explants of epithelium and mesenchyme from animals with differing genetic backgrounds. The role of PTHrP in the breast is not confined to the developmental period. Glandular epithelial cells of the lactating breast as well as myoepithelial cells produce large amounts of PTHrP. This has led to the proposal that PTHrP may be the long-postulated signal that is responsible for the adaptation of maternal calcium metabolism to the stress of lactation. A nursing mother secretes gram quantities of calcium into milk to mineralize the skeleton of her offspring; in multiparous animals, lactation presents a particular challenge to calcium homeostasis. Yet the adaptation to this challenge does not seem to require any of the hormones of calcium homeostasis: a mother can successfully suckle her offspring in the face of calcium deficiency, hypoparathyroidism, or vitamin D deficiency, sacrificing a large portion of her skeletal mineral to do so.100 Hypoparathyroid mothers can even maintain normocalcemia during lactation, but thereafter must return to vitamin D supplementation.101 Some workers have reported detectable levels of PTHrP in the serum of nursing mothers.102,103 However, even when detected, the levels of PTHrP are very low, and the level of PTH is not suppressed in lactation, as expected in the face of PTHrP-induced bone resorption. It remains to be shown conclusively that PTHrP is the long-sought lactational signal. Nonetheless, mammary PTHrP can probably produce systemic effects, because PTHrP levels are increased in rare syndromes of hypercalcemia associated with lactation104 and massive mammary hypertrophy.105 In all mammalian and marsupial species that have been tested, PTHrP is found in milk at enormous levels~10,000 times higher than the serum level.58,106 The production of PTHrP during lactation is under the control of prolactin,107 and the protein is secreted into milk by the glandular epithelial cells of the lactating mammary gland. This recapitulates for the nursing newborn the environment of its developing gut in utero, because amniotic fluid swallowed by the fetus also contains high levels

of PTHrP.108 It is possible that PTHrP plays a physiologic role in the developing alimentary tract or is absorbed to fulfill other functions. Receptors for PTHrP are present in the intestinal epithelium.109 However, infants raised on soy-based formulas that do not contain PTHrP are healthy. It therefore appears that PTHrP in milk is dispensable. SKIN One prominent site of normal production of PTHrP is the epidermis. The activation of the gene in epidermal keratinocytes probably accounts for the frequency of PTHrP-induced hypercalcemia in squamous carcinomas, which arise from this cell type. Expression of PTHrP seems to control the size of the proliferative pool of basal keratinocytes, from which keratinocytes exit to undergo terminal differentiation, keratinization, and programmed cell death. This pool is increased when PTHrP is overexpressed in the skin, but is diminished in rescued PTHrP knockout mice, with a reciprocal increase in granular cells that have entered the differentiation pathway, suggesting that in postnatal mice PTHrP regulates epidermal cell traffic in much the same way that it does in endochondral ossification prenatally.110 PTHrP is also produced by cells of the inner root sheath of the hair follicle. The PTH/PTHrP receptor is present in the underlying dermis. Treatment with a PTHrP antagonist enhances entry of hair follicles into anagen, the active phase of hair growth.111 This is another example of epithelial-mesenchymal interactions in which PTHrP serves as the messenger. TEETH Rescue of PTHrP knockout mice, as previously described, discloses a failure of tooth eruption, which can be attributed to the absence of another epithelial-mesenchymal interaction. The formation of the teeth appears normal, but PTHrP is absent from the enamel epithelium, which caps the tooth rudiment as it pushes its way through the overlying alveolar bone to erupt. Restoration of PTHrP expression to enamel epithelium, using a strategy similar to those previously described, in which PTHrP is directed to this cell layer by a tissue-specific promoter, restores tooth eruption.112 This implies that PTHrP secreted by the epithelial layer is normally targeted to receptors in the overlying bone, where it activates resorption of alveolar bone by osteoclasts to allow passage of the tooth. The signaling circuit in alveolar bone is distinctive, because there is no general defect in osteoclast function in transgenic mice. If present, impaired osteoclast function would produce generalized osteopetrosis. SMOOTH MUSCLE PTHrP is a vasodilator113,114 and also relaxes bladder, gastric, duodenal, and uterine smooth muscle beds.115,116 and 117 The expression of PTHrP is increased in several types of smooth muscle by mechanical stretch 115 or by treatment with vasoconstrictors, such as angiotensin II.118 Together, these findings raise the possibility that PTHrP released in response to increased smooth muscle tone acts locally to relax smooth muscle in a short-loop negative feedback system.119,120 In keeping with this hypothesis, transgenic mice that overexpress PTHrP in vascular smooth muscle manifest a reduction in blood pressure.121 This circuitry could be operative in the heart, in which PTHrP is released by atrial and ventricular myocytes122 and has a positive chronotropic effect as well as a positive inotropic effect that probably results from coronary vasodilation.123,124 PTHrP is also released from stromal cells of the spleen and other organs in response to endotoxic shock125; neutralization of PTHrP effects with antibodies prolongs survival after administration of lethal doses of endotoxin.126 In the arterial wall, PTHrP is expressed in proliferating vascular smooth muscle cells in culture118 and after balloon angioplasty in vivo.127 The level of PTHrP is increased in atherosclerotic coronary arteries.128 Exposure of rat vascular smooth muscle cells to PTHrP has an antimitotic effect, suggesting that locally released PTHrP might act to throttle the response to a proliferative stimulus.118 In contrast, when transfected into A10 rat vascular smooth muscle cells, PTHrP markedly induces proliferation.49 The proliferative response does not occur with transfection of mutant forms of PTHrP from which polybasic amino acid sequences between residues 88 and 106 have been deleted. These sequences have been shown to function as nuclear localization sequences in other cells50,129; wild-type PTHrP, but not the deletion mutants, is targeted to the nucleus of A10 cells. It thus appears possible that in addition to binding to cell-surface receptors, PTHrP can have direct nuclear actions, termed intra-crine actions. Because secreted fragments of PTHrP and its intracrine actions appear to have opposing effects on proliferation, PTHrP could interplay in a complex fashion with other proliferative factors in determining the response of the vascular wall to injury or atherosclerosis. UTERUS AND PLACENTA In the human uteroplacental unit, PTHrP is expressed at high levels in the amniotic sac (as well as chorion, placenta, and decidua) and is secreted into the amniotic fluid.108 The expression of PTHrP in amnion as well as rat myometrium115 appears to be sensitive to stretch, and PTHrP in turn could modulate smooth muscle tone. Both fetal skin and the fetal gastrointestinal tract are exposed to high concentrations of PTHrP through at least the last half of pregnancy; thus, PTHrP could have a developmental function in these tissues. The placenta maintains a calcium gradient by actively pumping calcium into the fetal circulation.130 This gradient is abolished by fetal parathyroidectomy and in PTHrP knockout mice.131 The transport of calcium can be restored by infusion of midregional fragments of PTHrP (e.g., PTHrP[4668]), but not by aminoterminal PTHrP or by PTH.46,47,131 As discussed in a previous section, this provides strong evidence for a distinct receptor for the midregion domain and establishes that PTHrP is a polyhormone. The fetus uses a midregion peptide from PTHrP for the regulation of systemic calcium economy, rather than the PTH-like aminoterminal domain. In so doing the fetus makes use of midregion PTHrP, perhaps secreted from the parathyroid gland (which expresses PTHrP83,84) in much the same way that PTH is used in the postnatal state. CHAPTER REFERENCES
1. Stewart AF, Horst R, Deftos LJ, et al. Biochemical evaluation of patients with cancer-associated hypercalcemia: evidence for humoral and non-humoral groups. N Engl J Med 1980; 303:1377. 2. Moseley JM, Kubota M, Diefenbach-Jagger H. Parathyroid hormone-related protein purified from a human lung cancer cell line. Proc Natl Acad Sci U S A 1987; 84:5048. 3. Strewler GJ, Stern PH, Jacobs JW. Parathyroid hormone-like protein from human renal carcinoma cells: structural and functional homology with parathyroid hormone. J Clin Invest 1987; 80:1803. 4. Stewart AF, Wu T, Goumas D, et al. N-terminal amino acid sequence of two novel tumor-derived adenylate cyclasestimulating proteins: identification of parathyroid hormone-like and parathyroid hormone-unlike domains. Biochem Biophys Res Commun 1987; 146:672. 5. Thiede MA, Strewler GJ, Nissenson RA, et al. Human renal carcinoma expresses two messages encoding a parathyroid hormone-like peptide: evidence for the alternative splicing of a single-copy gene. Proc Natl Acad Sci U S A 1988; 85:4605. 6. Suva LJ, Winslow GA, Wettenhall RE, et al. A parathyroid hormone-related protein implicated in malignant hypercalcemia: cloning and expression. Science 1987; 237:893. 7. Mangin M, Webb AC, Dreyer BE, et al. Identification of a cDNA encoding a parathyroid hormone-like peptide from a human tumor associated with humoral hypercalcemia of malignancy. Proc Natl Acad Sci U S A 1988; 85:597. 8. Broadus A, Stewart A. Parathyroid hormone-related protein: structure, processing, and physiological actions. In: Bilezikian J, Levine M, Marcus R, eds. The parathyroids. New York: Raven Press, 1994:259. 9. Dittmer J, Gitlin SD, Reid RL, Brady JN. Transactivation of the P2 promoter of parathyroid hormone-related protein by human T-cell lymphotropic virus type I Tax1: evidence for the involvement of transcription factor Ets1. J Virol 1993; 67:6087. 10. Burtis WJ, Fodero JP, Gaich G, et al. Preliminary characterization of circulating amino- and carboxy-terminal fragments of parathyroid hormone-related peptide in humoral hypercalcemia of malignancy. J Clin Endocrinol Metab 1992; 75:1110. 11. Soifer NE, Stewart AF. Measurement of PTH-related protein and the role of PTH-related protein in malignancy-associated hypercalcemia. In: Halloran BP, Nissenson RA, eds. Parathyroid hormone-related protein: normal physiology and its role in cancer. Boca Raton: CRC Press, 1992. 12. Soifer NE, Stewart AF. Measurement of PTH-related protein and the role of PTH-related protein in malignancy-associated hypercalcemia. In: Halloran BP, Nissenson RA, eds. Parathyroid hormone-related protein: normal physiology and its role in cancer. Boca Raton: CRC Press, 1992:93. 13. Wu TL, Vasavada RC, Yang K, et al. Structural and physiologic characterization of the mid-region secretory species of parathyroid hormone-related protein. J Biol Chem 1996; 271:24371. 14. Burtis WJ, Dann P, Gaich GA, Soifer NE. A high abundance midregion species of parathyroid hormone-related protein: immunological and chromatographic characterization in plasma. J Clin Endocrinol Metab 1994; 78:317. 15. Orloff JJ, Soifer NE, Fodero JP, et al. Accumulation of carboxy-terminal fragments of parathyroid hormone-related protein in renal failure. Kidney Int 1993; 43:1371. 16. Burtis WJ, Brady TG, Orloff JJ, et al. Immunochemical characterization of circulating parathyroid hormone-related protein in patients with humoral hypercalcemia of cancer. N Engl J Med 1990; 322:1106. 17. Fenton AJ, Kemp BE, Hammonds RG Jr, et al. A potent inhibitor of osteoclastic bone resorption within a highly conserved pentapeptide region of parathyroid hormone-related protein; PTHrP[107-111]. Endocrinology 1991; 129:3424. 18. Cornish J, Callon KE, Nicholson GC, et al. Parathyroid hormone-related protein-(107-139) inhibits bone resorption in vivo. Endocrinology 1997; 138:1299. 19. Sone T, Kohno H, Kikuchi H, et al. Human parathyroid hormone-related peptide-(107-111) does not inhibit bone resorption in neonatal mouse calvariae. Endocrinology 1992; 131:2742. 20. Fukayama S, Tashjian AH Jr, Davis JN, et al. Signaling by N- and C-terminal sequences of parathyroid hormone-related protein in hippocampal neurons. Proc Natl Acad Sci U S A 1995; 92:10182. 21. Orloff JJ, Reddy D, de Papp AE, et al. Parathyroid hormone-related protein as a prohormone: posttranslational processing and receptor interactions. Endocr Rev 1994; 15:40. 22. Wu TL, Soifer NE, Burtis WJ, et al. Glycosylation of parathyroid hormone-related peptide secreted by human epidermal keratinocytes. J Clin Endocrinol Metab 1991; 73:1002. 23. Wysolmerski JJ, Stewart AF. The physiology of parathyroid hormone-related proteinan emerging role as a developmental factor. Annu Rev Physiol 1998; 60:431. 24. Grill V, Ho P, Body JJ, et al. Parathyroid hormone-related protein: elevated levels in both humoral hypercalcemia of malignancy and hypercalcemia complicating metastatic breast cancer. J Clin Endocrinol Metab 1991; 73:1309. 25. Kao PC, Klee GG, Taylor RL, Heath H III. Parathyroid hormone-related peptide in plasma of patients with hypercalcemia and malignant lesions. Mayo Clin Proc 1990; 65:1399. 26. Budayr AA, Nissenson RA, Klein RF, et al. Increased serum levels of a parathryoid hormone-like protein in malignancy-associated hypercalcemia. Ann Intern Med 1989; 111:807. 27. Ratcliffe WA, Norbury S, Heath DA, Ratcliffe JG. Development and validation of an immunoradiometric assay of parathyrin-related protein in unextracted plasma. Clin Chem 1991; 37:678. 28. Blind E, Raue F, Gtzmann J, et al. Circulating levels of midregional parathyroid hormone-related protein in hypercalcaemia of malignancy. Clin Endocrinol (Oxf) 1992; 37:290.

29. Bucht E, Eklund A, Toss G, et al. Parathyroid hormone-related peptide, measured by a midmolecule radioimmunoassay, in various hypercalcaemic and normocalcaemic conditions. Acta Endocrinol (Copenh) 1992; 127:294. 30. Ratcliffe WA, Norbury S, Stott RA, et al. Immunoreactivity of plasma parathyrin- related peptide: three region-specific radioimmunoassays and a two-site immunoradiometric assay compared. Clin Chem 1991; 37:1781. 30a. Motellon JL, Javort Jiminez JF, deMiguel F, et al. Parathyroid hormone-related protein, parathyroid hormone, and vitamin D in hypercalcemia of malignancy. Clin Chim Acta 2000; 290:189. 31. Pandian MR, Morgan CH, Carlton E, Segre GV. Modified immunoradiometric assay of parathyroid hormone-related protein: clinical application in the differential diagnosis of hypercalcemia. Clin Chem 1992; 38:282. 32. Shigeno C, Yamamoto I, Kitamura N, et al. Interaction of human parathyroid hormone-related peptide with parathyroid hormone receptors in clonal rat osteosarcoma cells. J Biol Chem 1988; 263:18369. 33. Nissenson RA, Diep D, Strewler GJ. Synthetic peptides comprising the aminoterminal sequence of a parathyroid hormone-like protein from human malignancies: binding to parathyroid hormone receptors and activation of adenylate cyclase in bone cells and kidney. J Biol Chem 1988; 263:12866. 34. Jppner H, Abou-Samra A, Uneno S, et al. The parathyroid hormone-like peptide associated with humoral hypercalcemia of malignancy and parathyroid hormone bind to the same receptor on the plasma membrane of ROS 17/2.8 cells. J Biol Chem 1988; 263:8557. 35. Orloff JJ, Wu TL, Heath HW, et al. Characterization of canine renal receptors for the parathyroid hormone-like protein associated with humoral hypercalcemia of malignancy. J Biol Chem 1989; 264:6097. 36. Lanske B, Kronenberg HM. Parathyroid hormone-related peptide (PTHrP) and parathyroid hormone (PTH)/PTHrP receptor. Crit Rev Eukaryot Gene Expr 1998; 8:297. 37. Lanske B, Karaplis AC, Lee K, et al. PTH/PTHrP receptor in early development and Indian hedgehog-regulated bone growth. Science 1996; 273:663. 38. Schipani E, Lanske B, Hunzelman J, et al. Targeted expression of constitutively active receptors for parathyroid hormone and parathyroid hormone-related peptide delays endochondral bone formation and rescues mice that lack parathyroid hormone-related peptide. Proc Natl Acad Sci U S A 1997; 94:13689. 39. Schipani E, Langman CB, Parfitt AM, et al. Constitutively activated receptors for parathyroid hormone and parathyroid hormone-related peptide in Jansen's metaphyseal chondrodysplasia. N Engl J Med 1996; 335:708. 40. Yamamoto S, Morimoto I, Yanagihara N, et al. Parathyroid hormone-related peptide-(1-34) [PTHrP-(1-34)] induces vasopressin release from the rat supraoptic nucleus in vitro through a novel receptor distinct from a type I or type II PTH/PTHrP receptor. Endocrinology 1997; 138:2066. 41. Gaich G, Orloff JJ, Atillasoy EJ, et al. Amino-terminal parathyroid hormone- related protein: Specific binding and cytosolic calcium responses in rat insulinoma cells. Endocrinology 1993; 132:1402. 42. Orloff JJ, Ganz MB, Ribaudo AE, et al. Analysis of PTHrP binding and signal transduction mechanisms in benign and malignant squamous cells. Am J Physiol 1992; 262:E599E607. 43. Usdin TB, Gruber C, Bonner TI. Identification and functional expression of a receptor selectively recognizing parathyroid hormone, the PTH2 receptor. J Biol Chem 1995; 270:15455. 44. Behar V, Nakamoto C, Greenberg Z, et al. Histidine at position 5 is the specificity switch between two parathyroid hormone receptor subtypes. Endocrinology 1996; 137:4217. 45. Gardella TJ, Luck MD, Jensen GS, et al. Converting parathyroid hormone-related peptide (PTHrP) into a potent PTH-2 receptor agonist. J Biol Chem 1996; 271:19888. 46. Abbas SK, Pickard DW, Rodda CP, et al. Stimulation of ovine placental calcium transport by purified natural and recombinant parathyroid hormone-related protein (PTHrP) preparations. Q J Exp Physiol 1989; 74:549. 47. Care AD, Abbas SK, Pickard DW, et al. Stimulation of ovine placental transport of calcium and magnesium by mid-molecule fragments of human parathyroid hormone-related protein. Exp Physiol 1990; 75:605. 48. Fenton AJ, Kemp BE, Kent GN, et al. A carboxyl-terminal peptide from the parathyroid hormone-related protein inhibits bone resorption by osteoclasts. Endocrinology 1991; 129:1762. 49. Massfelder T, Dann P, Wu TL, et al. Opposing mitogenic and anti-mitogenic actions of parathyroid hormone-related protein in vascular smooth muscle cellsa critical role for nuclear targeting. Proc Natl Acad Sci U S A 1997; 94:13630. 50. Henderson JE, Amizuka N, Warshawsky H, et al. Nucleolar localization of parathyroid hormone-related peptide enhances survival of chondrocytes under conditions that promote apoptotic cell death. Mol Cell Biol 1995; 15:4064. 51. Horiochi N, Caulfield MP, Fisher JE, et al. Similarity of synthetic peptide from human tumor to PTH in vivo and in vitro. Science 1987; 238:1566. 52. Kemp BE, Moseley JM, Rodda CP, et al. Parathyroid hormone-related protein of malignancy: active synthetic fragments. Science 1987; 238:1568. 53. Stewart AF, Mangin M, Wu T, et al. A synthetic human parathyroid hormone- like protein stimulates bone resorption and causes hypercalcemia in rats. J Clin Invest 1988; 81:596. 54. Civitelli R, Martin TJ, Fausto A, et al. Parathyroid hormone-related peptide transiently increases cytosolic calcium in osteoblast-like cells: comparison with parathyroid hormone. Endocrinology 1989; 125:1204. 55. Yamada H, Tsutsumi M, Fukase M, et al. Effects of human PTH-related peptide and human PTH on cyclic AMP production and cytosolic free calcium in an osteoblastic cell clone. Bone Miner 1989; 6:45. 56. Yates AJ, Gutierrez GE, Smolens P, et al. Effects of a synthetic peptide of a parathyroid hormone-related protein on calcium homeostasis, renal tubular calcium reabsorption, and bone metabolism in vivo and in vitro in rodents. J Clin Invest 1988; 81:932. 57. Fraher LJ, Hodsman AB, Jonas K, et al. A comparison of the in vivo bio-chemical responses to exogenous parathyroid hormone-(1-34) [PTH-(1- and PTH-related peptide-(1-34) in man. J Clin Endocrinol Metab 1992; 75:417. 58. Budayr AA, Halloran BP, King JC, et al. High levels of a parathyroid hormone- like protein in milk. Proc Natl Acad Sci U S A 1989; 86:7183. 59. Bundred NJ, Ratcliffe WA, Walker RA, et al. Parathyroid hormone related protein and hypercalcemia in breast cancer. BMJ 1991; 303:1506. 60. Southby J, Kissin MW, Danks JA, et al. Immunohistochemical localization of parathyroid hormone-related protein in human breast cancer. Cancer Res 1990; 50:7710. 61. Powell GJ, Southby J, Danks JA, et al. Localization of parathyroid hormone- related protein in breast cancer metastases: increased incidence in bone compared with other sites. Cancer Res 1991; 51:3059. 62. Guise TA, Yin JJ, Taylor SD, et al. Evidence for a causal role of parathyroid hormone-related protein in the pathogenesis of human breast cancer-mediated osteolysis. J Clin Invest 1996; 98:1544. 63. Heath DA. Parathyroid hormone related protein. Clin Endocrinol (Oxf) 1993; 38:135. 64. Yin JJ, Selander K, Chirgwin JM, et al. TGF-b signaling blockade inhibits PTHrP secretion by breast cancer cells and bone metastases development. J Clin Invest 1999; 103:197. 65. Firkin F, Seymour JF, Watson AM, et al. Parathyroid hormone-related protein in hypercalcaemia associated with haematological malignancy. Br J Haematol 1996; 94:486. 66. Fukumoto S, Matsumoto T, Ikeda K, et al. Clinical evaluation of calcium metabolism in adult T-cell leukemia/lymphoma. Arch Intern Med 1988; 148:921. 67. Fukumoto S, Matsumoto T, Yamoto H, et al. Suppression of serum 1,25- dihydroxyvitamin D is caused by elaboration of a factor that inhibits renal 1,25-dihydroxyvitamin D3 production. Endocrinology 1989; 124:2057. 68. Motokura T, Fukumoto S, Takahashi S, et al. Expression of parathyroid hormone-related protein in a human T cell lymphotrophic virus type I-infected T cell line. Biochem Biophys Res Commun 1988; 154:1182. 69. Motokura T, Fukumoto S, Matsumoto T, et al. Parathyroid hormone-related protein in adult T-cell leukemia-lymphoma. Ann Intern Med 1989; 111,484. 70. Ikeda K, Ohno H, Hane M, et al. Development of a sensitive two-site immunoradiometric assay for parathyroid hormone-related peptide: evidence for elevated levels in plasma from patients with adult T-cell leukemia/lymphoma and B-cell lymphoma. J Clin Endocrinol Metab 1994; 79:1322. 71. Watanabe T, Yamaguchi K, Takatsuki K, et al. Constitutive expression of parathyroid hormone-related protein gene in human T cell leukemia virus type 1 (HTLV-1) carriers and adult T cell leukemia patients that can be trans-activated by HTLV-1 tax gene. J Exp Med 1990; 172:759. 72. Ikeda K, Okazaki R, Inoue D, et al. Interleukin-2 increases production and secretion of parathyroid hormone-related peptide by human T cell leukemia virus type I-infected T cells: possible role in hypercalcemia associated with adult T cell leukemia. Endocrinology 1993; 132:2551. 73. Ikeda K, Okazaki R, Inoue D, et al. Transcription of the gene for parathyroid hormone-related peptide from the human is activated through a cAMP-dependent pathway by prostaglandin E1 in HTLV-I-infected T cells. J Biol Chem 1993; 268:1174. 74. Henry JG, Mitnick M, Dann PR, et al. Parathyroid hormone-related protein-(1-36) is biologically active when administered subcutaneously to humans. J Clin Endocrinol Metab 1997; 82:900. 75. Kukreja SC, Shevrin DH, Wimbiscus SA, et al. Antibodies to parathyroid hormone-related protein lower serum calcium in athymic mouse models of malignancy-associated hypercalcemia due to human tumors. J Clin Invest 1988; 82:1798. 76. Kukreja SC, Rosol TJ, Wimbiscus SA, et al. Tumor resection and antibodies to parathyroid hormone-related protein cause similar changes on bone histomorphometry in hypercalcemia of cancer. Endocrinology 1990; 127:305. 77. Guise TA, Yoneda T, Yates AJ, Mundy GR. The combined effect of tumor-produced parathyroid hormone-related protein and transforming growth factor-alpha enhance hypercalcemia in vivo and bone resorption in vitro. J Clin Endocrinol Metab 1993; 77:40. 78. Stewart A, Vignery A, Silverglate A, et al. Quantitative bone histomorphometry in humoral hypercalcemia of malignancy: uncoupling of bone cell activity. J Clin Endocrinol Metab 1982; 55:219. 79. Budayr AA, Zysset E, Jenzer A, et al. Effects of treatment of malignancy-associated hypercalcemia on serum parathyroid hormone-related protein. J Bone Miner Res 1994; 9:521. 80. Schilling T, Pecherstorfer M, Blind E, et al. Parathyroid hormone-related protein (PTHrP) does not regulate 1,25-dihydroxyvitamin D serum levels in hypercalcemia of malignancy. J Clin Endocrinol Metab 1993; 76:801. 81. Blind E, Raue F, Meinel T, et al. Levels of parathyroid hormone-related protein (PTHrP) in hypercalcemia of malignancy are not lowered by treatment with the biphosphonate BM 21.0955. Horm Metab Res 1993; 25:40. 82. Dodwell DJ, Abbas SK, Morton AR, Howell A. Parathyroid hormone-related protein 5069 and response to pamidronate therapy for tumour-induced hypercalcemia. Eur J Cancer 1991; 27:1629. 83. Moseley JM, Hayman JA, Danks JA, et al. Immunohistochemical detection of parathyroid hormone-related protein in human fetal epithelia. J Clin Endocrinol Metab 1991; 73:478. 84. Rodda CP, Kubota M, Heath JA, et al. Evidence for a novel parathyroid hormone-related protein in fetal lamb parathyroid glands and sheep placenta: comparisons with a similar protein implicated in humoral hypercalcaemia of malignancy. J Endocrinol 1988; 117:261. 85. Ikeda K, Arnold A, Mangin M, et al. Expression of transcripts encoding a parathyroid hormone-related peptide in abnormal human parathyroid tissues. J Clin Endocrinol Metab 1989; 69:1240. 86. Drucker DJ, Asa SL, Henderson J, Goltzman D. The parathyroid hormone-like peptide gene is expressed in the normal and neoplastic human endocrine pancreas. Mol Endocrinol 1989; 3:1589. 87. Weir EC, Brines ML, Ikeda K, et al. Parathyroid hormone-related peptide gene is expressed in the mammalian central nervous system. Proc Natl Acad Sci U S A 1990; 87:108. 88. Selvanayagam P, Graves K, Cooper C, Rajaraman S. Expression of the parathyroid hormone-related peptide gene in rat tissues. Lab Invest 1991; 64:713. 89. Karaplis AC, Luz A, Glowacki J, et al. Lethal skeletal dysplasia from targeted disruption of the parathyroid hormone-related peptide gene. Genes Dev 1994; 8:277. 90. Weir EC, Philbrick WM, Amling M, et al. Targeted overexpression of parathyroid hormone-related peptide in chondrocytes causes chondrodysplasia and delayed endochondral bone formation. Proc Natl Acad Sci U S A 1996; 93:10240. 91. Jobert AS, Zhang P, Couvineau A, et al. Absence of functional receptors for parathyroid hormone and parathyroid hormone-related peptide in Blomstrand chondrodysplasia. J Clin Invest 1998; 102:34. 92. Zhang P, Jobert AS, Couvineau A, et al. A homozygous inactivating mutation in the parathyroid hormone/parathyroid hormone-related peptide receptor causing Blomstrand chondrodysplasia. J Clin Endocrinol Metab 1998; 83:3365. 93. Crabb ID, O'Keefe RJ, Puzas JE, Rosier RN. Differential effects of parathyroid hormone on chicken growth plate and articular chondrocytes. Calcif Tissue Int 1992; 50:61. 94. Koike T, Iwamoto M, Shimazu A, et al. Potent mitogenic effects of parathyroid hormone (PTH) on embryonic chick and rabbit chondrocytes: differential effects of age on growth, proteoglycan, and cyclic AMP responses of chondrocytes to PTH. J Clin Invest 1990; 85:626. 95. Kato Y, Shimazu A, Nakashima K, et al. Effects of parathyroid hormone and calcitonin on alkaline phosphatase activity and matrix calcification in rabbit growth-plate chondrocyte cultures. Endocrinology 1990; 127:114. 96. Vortkamp A, Lee K, Lanske B, et al. Regulation of rate of cartilage differentiation by Indian hedgehog and PTH-related protein. Science 1996; 273:613. 97. Chung UI, Lanske B, Lee K, et al. The parathyroid hormone/parathyroid hormone-related peptide receptor coordinates endochondral bone development by directly controlling chondrocyte differentiation. Proc Natl Acad Sci U S A 1998; 95:13030. 98. Wysolmerski JJ, Philbrick WM, Dunbar ME, et al. Rescue of the parathyroid hormone-related protein knockout mouse demonstrates that parathyroid hormone-related protein is essential for mammary gland development. Development 1998; 125:1285. 99. Wysolmerski JJ, McCaughern-Carucci, Daifotis AG, et al. Overexpression of parathyroid hormone-related protein or parathyroid hormone in transgenic mice impairs branching morphogenesis during mammary gland development. Development 1995; 121:3539. 100. Kovacs CS, Kronenberg HM. Maternal-fetal calcium and bone metabolism during pregnancy, puerperium, and lactation. Endocr Rev 1997; 18:832. 101. Mather KJ, Chik CL, Corenblum B. Maintenance of serum calcium by parathyroid hormone-related peptide during lactation in a hypoparathyroid patient. J Clin Endocrinol Metab 1999; 84:424. 102. Grill V, Hillary J, Ho PMW, et al. Parathyroid hormone-related protein; a possible endocrine function in lactation. Clin Endocrinol (Oxf) 1992; 37:405.

103. 104. 105. 106. 107. 108. 109. 110. 111. 112. 113. 114. 115. 116. 117. 118. 119. 120. 121. 122. 123. 124. 125. 126. 127. 128. 129. 130. 131.

Sowers MF, Hollis BW, Shapiro B, et al. Elevated parathyroid hormone-related peptide associated with lactation and bone density loss. JAMA 1996; 276:549. Lepre F, Grill V, Ho PW, et al. Hypercalcemia in pregnancy and lactation associated with parathyroid hormone-related protein. N Engl J Med 1993; 328:666. Braude S, Graham A, Mitchell D. Lymphoedema/hypercalcaemia syndrome mediated by parathyroid-hormone-related protein. Lancet 1991; 337:140. Thurston AW, Cole JA, Hillman LS, et al. Purification and properties of parathyroid hormone-related peptide isolated from milk. Endocrinology 1990; 126:1183. Thiede MA. The mRNA encoding a parathyroid hormone-like peptide is produced in mammary tissue in response to elevations in serum prolactin. Mol Endocrinol 1989; 3:1443. Ferguson JE II, Gorman JV, Bruns DE, et al. Abundant expression of parathyroid hormone-related protein in human amnion and its association with labor. Proc Natl Acad Sci U S A 1992; 89:8384. Li H, Seitz PK, Thomas ML, et al. Widespread expression of the parathyroid hormone-related peptide and PTH/PTHrP receptor genes in intestinal epithelial cells. Lab Invest 1995; 73:864. Foley J, Longely BJ, Wysolmerski JJ, et al. PTHrP regulates epidermal differentiation in adult mice. J Invest Dermatol 1998; 111:1122. Holick MF, Ray S, Chen TC, et al. A parathyroid hormone antagonist stimulates epidermal proliferation and hair growth in mice. Proc Natl Acad Sci U S A 1994; 91:8014. Philbrick WM, Dreyer BE, Nakchbandi IA, et al. Parathyroid hormone-related protein is required for tooth eruption. Proc Natl Acad Sci U S A 1998; 95:11846. Roca-Cusachs A, DiPette DJ, Nickols GA. Regional and systemic hemodynamic effects of parathyroid hormone-related protein: preservation of cardiac function and coronary and renal flow with reduced blood pressure. J Pharmacol Exp Ther 1991; 256:110. Winquist RJ, Baskin EP, Vlasuk GP. Synthetic tumor-derived human hypercalcemic factor exhibits parathyroid hormone-like vasorelaxation in renal arteries. Biochem Biophys Res Commun 1987; 149:227. Thiede MA, Daifotis AG, Weir EC, et al. Intrauterine occupancy controls expression of the parathyroid hormone-related peptide gene in preterm rat myometrium. Proc Natl Acad Sci U S A 1990; 87:6969. Mok LL, Ajiwe E, Martin TJ, et al. Parathyroid hormone-related protein relaxes rat gastric smooth muscle and shows cross-desensitization with parathyroid hormone. J Bone Miner Res 1989; 4:433. Mok LL, Cooper CW, Thompson JC. Parathyroid hormone and parathyroid hormone-related protein inhibit phasic contraction of pig duodenal smooth muscle. Proc Soc Exp Biol Med 1989; 191:337. Pirola CJ, Wang HM, Kamyar A, et al. Angiotensin II regulates parathyroid hormone-related protein expression in cultured rat aortic smooth muscle cells through transcriptional and post-transcriptional mechanisms. J Biol Chem 1993; 268:1987. Massfelder T, Helwig JJ, Stewart AF. Parathyroid hormone-related protein as a cardiovascular regulatory peptide. Endocrinology 1996; 137:3151. Massfelder T, Fiaschitaesch N, Stewart AF, et al. Parathyroid hormone-related peptide-a smooth muscle tone and proliferation regulatory protein. [Review.] Curr Opin Nephrol Hyperten 1998; 7:27. Maeda S, Sutliff RL, Qian J, et al. Targeted overexpression of parathyroid hormone-related protein (PTHrP) to vascular smooth muscle in transgenic mice lowers blood pressure and alters vascular contractility. Endocrinology 1999; 140:1815. Deftos LJ, Burton DW, Brandt DW. Parathyroid hormone-like protein is a secretory product of atrial myocytes. J Clin Invest 1993; 92:727. Ogino K, Burkhoff D, Bilezikian JP. The hemodynamic basis for the cardiac effects of parathyroid hormone (PTH) and PTH-related protein. Endocrinology 1995; 136:3024. Hara M, Liu YM, Zhen LC, et al. Positive chronotropic actions of parathyroid hormone and parathyroid hormone-related peptide are associated with increases in the current, I-F, and the slope of the pacemaker potential. Circulation 1997; 96:3704. Funk JL, Krul EJ, Moser AH, et al. Endotoxin increases parathyroid hormone- related protein mRNA levels in mouse spleen. Mediation by tumor necrosis factor. J Clin Invest 1993; 92:2546. Funk JL, Moser AH, Strewler GJ, et al. Parathyroid hormone-related protein is induced during lethal endotoxemia and contributes to endotoxin-induced mortality in rodents. Molecular Medicine 1996; 2:204. Ozeki S, Ohtsuru A, Seto S, et al. Evidence that implicates the parathyroid hormone-related peptide in vascular stenosis-increased gene expression in the intima of injured rat carotid arteries and human restenotic coronary lesions. Arterioscler ThrombVascul Biol 1996; 16:565. Nakayama T, Ohtsuru A, Enomoto H, et al. Coronary atherosclerotic smooth muscle cells overexpress human parathyroid hormone-related peptides. Biochem Biophys Res Commun 1994; 200:1028. Aarts MM, Levy D, He B, et al. Parathyroid hormone-related protein interacts with RNA. J Biol Chem 1999; 274:4832. Strenler GJ. The philosophy of parathyroid hormone-related protein. N Engl J Med 2000; 342:177. Kovacs CS, Lanske B, Hunzelman JL, et al. Parathyroid hormone-related peptide (PTHrP) regulates fetal-placental calcium transport through a receptor distinct from the PTH/PTHrP receptor. Proc Natl Acad Sci U S A 1996; 93:15233.

CHAPTER 53 CALCITONIN GENE FAMILY OF PEPTIDES Principles and Practice of Endocrinology and Metabolism

CHAPTER 53 CALCITONIN GENE FAMILY OF PEPTIDES

KENNETH L. BECKER, BEAT MLLER, ERIC S. NYLN, RGIS COHEN, OMEGA L. SILVA, JON C. WHITE, AND RICHARD H. SNIDER, JR. Calcitonin Distribution of Calcitonin Control of Calcitonin Secretion Effects of Calcitonin Serum Levels of Immunoreactive Calcitonin Therapeutic Uses of Calcitonin Procalcitonin and its Derivative Peptides Disorders Associated with Aberrant Production and Secretion of Calcitonin Precursors and Derivative Peptides Calcitonin GeneRelated Peptides Alternative Gene Expression Effects Amylin Effects Adrenomedullin Effects Pathophysiology Receptors of the Calcitonin Gene Family of Peptides Calcitonin Receptors Calcitonin ReceptorLike Receptors Receptor-ActivityModifying Protein and the Plasticity of the Receptors for the Hormones of the Calcitonin Gene Family of Peptides Conclusion Chapter References

The calcitonin gene family consists of five genes (CALC-I to CALC-V)that are located on chromosome 11 (CALC-I, CALC-II, CALC-III, and CALC-V) and on chromosome 12 (CALC-IV). In humans, the mRNAs arising from these genes generate multiple peptides, including calcitonin (CT), calcitonin generelated peptide (CGRP), amylin, adrenomedullin (ADM), and various circulating precursor and derivative peptides, some of which may have biologic functions (Fig. 53-1).1,2,3,4 and 5

FIGURE 53-1. The human calcitonin (CT) gene family: organization of genes, mRNAs, and their hormone precursors. Based on their nucleotide sequence homologies, five genes belong to this family: CALC-I (CT/calcitonin generelated peptide-I [CGRP-I]), CALC-II (CGRP-II), CALC-III, CALC-IV (amylin), and the CALC-V (adrenomedullin [ADM]) genes. Two structural features that are essential for full functional activity are conserved between the peptides: They contain two N-terminal cysteines that form a disulfide bridge resulting in an N-terminal loop and a C-terminal amide. A, The CALC-I primary transcript is processed into three different mRNAs: CT, CT-II, and CGRP-I mRNAs. The different products are generated by the inclusion or exclusion of exons by a mechanism termed splicing. Exons IIII are common for all mRNAs. Exon IV codes for CT, and exon V codes for CGRP-I. CT mRNA includes exons I + II + III + IV. CT-II mRNA includes exons I + II + III + IV (partial) + V + VI. CGRP-I mRNA is composed of exons I + II + III + V + VI. Each mRNA codes for a specific precursor. CT mRNA codes mainly for an N-terminal region, mature CT, and a specific C-terminal peptide (i.e., katacalcin, PDN-21, or calcitonin carboxy-terminal peptide-I [CCP-I]) that consists of 21 amino acids (aa). The N-terminal region includes a signal peptide of 25 amino acids and an N-terminal peptide of 57 amino acids (i.e.,aminoprocalcitonin [NProCT] or PAS-57). The CT-II precursor differs from the CT-I precursor by its specific C-terminal peptide, CCP-II. CCP-I differs from CCP-II by its last eight amino acids. CGRP-I mRNA codes for an N-terminal region, mature CGRP-I, and a cryptic peptide. The commitment of primary transcript in the different splicing pathways is determined, in part, by a tissue specificity. Although some overlap is seen, the CGRP-I mRNA is expressed mainly in nervous tissue, and CT mRNA is the major mRNA product in thyroid tissue and other tissues, whereas CT-II was found to be expressed in liver. B, The CALC-II gene codes only for a CGRP-II precursor. Its organization is similar to that of the CALC-I gene, containing 6 exons. Sequence homologies are important. Examination of the exon IVlike region of CALC-II indicates that CT mRNA is unlikely. Splicing at the site equivalent to the exon IIIexon IV junction in human CT mRNA would result in a stop codon within the reading frame of the precursor polypeptide. Although CALC-II appears to be a pseudogene for CT, it is a structural gene for CGRP-II. The CGRP-II hormone differs from CGRP-I by three amino acids. C, The CALC-III gene contains only two exons. Their sequences have homologies with exon II and III of the CALC-I and CALC-II genes. The CALC-III gene does not seem to encode a CT- or CGRP-related peptide hormone and is probably a pseudogene that is not translated into a protein. D, The CALC-IV gene codes for a precursor containing the amylin peptide. This gene contains only three exons. The third exon codes for amylin. This 37-amino-acid peptide has marked homology with the CGRP peptides. The suggestion has been made that the CT and CGRP exons are derived from a primordial gene and that the different CT/CGRP/ADM/amylin genes have arisenby duplication and sequence-divergent events. E, The CALC-V gene is translated into ADM. This gene contains four exons. ADM is coded by the fourth exon. The amino-terminal peptides, encoded by exons II and III, also have some bioactivity.

Several common features characterize the calcitonin gene family of peptides (Fig. 53-2).CT, the CGRPs, amylin, and ADM all contain two N-terminal cysteines that form a disulfide bridge resulting in a ring structure at the amino terminus. In addition, the carboxy-terminal amino acid of all these peptides is amidated. When the sequences of CT, CGRP, and amylin are aligned with a gap introduced in the CT sequence to maximize homology, the 12 identical matches and five conservative amino-acid substitutions suggest gene duplication of a common ancestral gene. Furthermore, the midregions of CGRP, CT, amylin, and ADM form an a-helical structure. Finally, the CT gene family peptides exert their overlapping bioeffects by binding to the same family of receptors.

FIGURE 53-2. Amino-acid sequences of human calcitonin (hCT), salmon CT (sCT), human calcitonin generelated peptide-I (hCGRP-I), human calcitonin generelated peptide-II (hCGRP-II), human amylin (hAmylin), and human adrenomedullin (hADM).

CALCITONIN

For reasons that are elucidated later, the following discussion distinguishes between CT and its precursor peptides (CTpr), which have been found to have important clinical and biologic functions. CT is initially biosynthesized as a larger precursor, procalcitonin (ProCT; Fig. 53-3), which is enzymatically cleaved into aminoprocalcitonin (NProCT) and the conjoined calcitonin:calcitonin carboxypeptide-I (CT:CCP-I), some of which, in turn, is further enzymatically processed to yieldfree immature CT and CCP-I. Immature CT is then amidated to yield the mature CT. (Procalcitonin and its subsequent posttranslational processing are discussed later.)

FIGURE 53-3. Amino-acid sequence of procalcitonin. This prohormone consists of 116 amino acids. At the amino terminus is a 57-amino-acid peptide, aminoprocalcitonin. The midportion consists of the 33-amino-acid immature calcitonin. The final 21 amino acids comprise calcitonin carboxy-terminal peptide-I (CCP-I). (Smaller amounts of another flanking peptide [CCP-II] are found in nervous tissue.)

In this chapter, unless otherwise stated, the term CT, without further qualification, refers to the mature amidated hormone (molecular mass 3.42 kDa). The term immunoreactive calcitonin (iCT) is used to indicate either that the material being detected immunologically is not the mature hormone (i.e., it is incompletely processed CT or prohormone) or that in the particular referenced study, its precise molecular structure has not been determined or clarified. One should keep in mind that nearly all studies have used methodologies that do not distinguish mature CT from immature CT or CTpr and were performed before the distinct biologic functions of the various CT peptides were known. This important limitation should be kept in mind when scientific data of tissue distribution, control of secretion, and bioeffects are interpreted. Immature CT is a 33-amino-acid glycine-terminated peptide contained initially within the procalcitonin molecule. In the peripheral blood of normal persons, however, immature CT is found within the procalcitonin molecule as a separate peptide in which it is conjoined to the calcitonin carboxy-terminal peptide (CCP-I), and also asthe free 33-amino-acid peptide. Mature CT is a free 32-amino-acid peptide terminating in an amidated proline. As with many other peptides, this amidation is important to the bioactivity of the hormone (see Chap. 167). CT, in evolutionary terms, is a very conserved ancestral hormone that is found in primitive protochordate marine animals. Although the midportion of the hormone varies considerably among different species, common characteristics include the disulfide bond between the cysteine residues at the amino-terminal positions 1 and 7 and a carboxy-terminal proline amide. Of the known CTs, hamster and rat CT are the closest to the human form, differing by only two amino acids. Despite species differences, all of the CTs are bioactive in laboratory animals, although their potency and time curve of action vary with the receptor affinity and the rate of degradation. DISTRIBUTION OF CALCITONIN The principal source of the iCT found in mammalian serum is probably the C cells. In submammals (e.g., birds, fish, amphibians, reptiles), the thyroidal C cells occur within a discrete, separate ultimobranchial gland. During early embryogenesis, CT-containing cells migrate from the neural crest to this gland. In humans and other mammals, however, the ultimobranchial gland subsequently fuses with the thyroid gland (Fig. 53-4). In the human, the C cells (also called parafollicular cells) are found in the central portion of each lateral lobe (see Chap. 29). Within the mammalian thyroid gland, these cells constitute an endocrine system that is quite separate from the thyroid follicular cells.

FIGURE 53-4. Embryology of the endocrine structures of the human neck, showing the progressive development and migration of the ultimobranchial body. A, In the fifth week, pharyngeal arches appear (I, II, III, IV). They are separated and demarcated by pharyngeal clefts (1, 2, 3, 4). Simultaneously, along the lateral walls of the primitive pharynx, the pharyngeal pouches form (1, 2, 3, 4, 5). The pharyngeal arches (i.e., brachial arches) participate in the formation of structures of the face, head, and neck. B, The third pharyngeal pouches develop into the inferior parathyroid glands and the thymus; the fourth pouches develop into the superior parathyroid glands. The fifth pouches, which some consider to be part of the fourth, develop into the ultimobranchial bodies. C, Notice the migration of the ultimobranchial bodies into the thyroid gland after the latter organ has descended from the level of the foramen cecum. The caudally migrating thymus pulls down the inferior para-thyroid glands, and even though these glands have originated from the third pouches, their eventual location is below the superior parathyroid glands, which had originated from the fourth pouches. (inf., inferior; sup., superior; Ext., exterior.) (From Sadler TW. Langman's medical embryology. Baltimore: Williams & Wilkins, 1985.)

In addition to the thyroid gland, iCT is found in many other tissues, such as the lungs, adrenal medulla, hypothalamus, pituitary gland, thymus, parathyroid glands, and the gastrointestinal and genitourinary tracts.6 In part, these tissue levels reflect the production of the hormone by their constituent neuroendocrine cells (see Chap. 175). However, nonneuroendocrine cells may contain low levels of iCT. Immunoreactive CT is found in the blood, urine, bile, gastric juice, and seminal fluid, and the concentration is particularly high in milk. The wide distribution of iCT is similar to that of several other peptides of the diffuse neuroendocrine system (see Chap. 175) (e.g., somatostatin, cholecystokinin, mammalian bombesin, neurotensin) that were originally thought to emanate from a single tissue. In mammals, a large contribution to serum iCT is made by the C cells of the thyroid gland, and the responsivity of the mature CT to induced hypercalcemia appearsto reside mostly within these cells. The highest concentration of CT in the body is found within the C-cell region of the thyroid gland. Within the lung, iCT is found in the pulmonary neuroendocrine (PNE) cells, and in humans, more iCT is found within the lungs than in the thyroid gland7; however, the extent to which extrathyroidal neuroendocrine production of CT contributes to circulating levels of the hormone is uncertain. Although the thyroid C cells originate from the neural crest, the iCT-containing cells of the gut and respiratory tract probably arise embryologically from a different source. (For a discussion of iCT in extrathyroidal tissues, see later in section Calcitonin Precursors and Derivative Peptides.) CONTROL OF CALCITONIN SECRETION The secretory control of CT is poorly understood. The physiologic secretagogues probably vary with the location of the hormone. Because of the hypocalcemic effects of pharmacologic doses of the hormone, early experiments searching for a possible feedback mechanism had focused on the stimulatory influence of increased serum calcium on iCT secretion.8 The relationship between serum calcium and iCT secretion is not consistent, however, and small, sudden changes of endogenous circulating

calcium do not influence the serum hormone levels in humans. Hypermagnesemia but not hyperphosphatemia induces iCT release from the thyroid gland.9 Various hormones are known torelease thyroidal iCT, some of them involving cyclic AMP (cAMP); usually, these effects are induced by pharmacologic concentrations, and whether these responses are physiologically relevant in humans is unknown. These substances include cholecystokinin, cerulein, secretin, glucagon, and gastrin. Physiologically, modulation by the sympathetic nervous system may be important because b-adrenergic agonists may induce secretion, and b-adrenergic antagonists and a-adrenergic agonists may diminish iCT secretion. Within the lung, the iCT of the pulmonary neuroendocrine cells is released by a nicotinic-cholinergic mechanism.10,11 Factors that decrease iCT secretion include dopamine, somatostatin, and the H2-receptor blocker, cimetidine. EFFECTS OF CALCITONIN Despite hundreds of studies, the physiologic role of CT remains unclear. Many of the investigations used pharmacologic doses of the hormone. The studies used differentin vivo and in vitro models that employed different species of the hormone and different species of animals.The full bioeffects of CT require the presence of the disulfide bridge and an intact NH2-terminal end, and the majority of pharmacologic studies have used the mature (processed and amidated) forms of CT. Although multiple target tissues have been identified, bones, the gastrointestinaltract, the kidneys, and the central nervous system appear to be particularly important. BONE EFFECTS The osseous effect of CT is related to its inhibition of osteoclast function, which causes diminished bone resorption (Fig. 53-5).12 In addition, some data suggest that CT may also exert stimulatory effects on osteoblasts.13 In vivo, in most animals, the administration of large amounts of CT induces hypocalcemia and hypophosphatemia, an effect that is attributed to the decreased bone resorption. In vitro, CT protects against the bone-resorbing effects of parathyroid hormone, prostaglandins, and vitamin A. In humans, however, CT does not appear to play a role in the fine regulation of serum calcium. Thus, although the hormone may prevent a postprandial increase in serum calcium in some animals, this is not true in the human adult.

FIGURE 53-5. Effect of calcitonin on the osteoclast. Series of video micrographs taken with phase-contrast microscopy shows a single osteoclast. Frame at left was taken immediately before addition of calcitonin (100 ng/mL final concentration). Middle frame shows osteoclast 10 minutes after addition of calcitonin, by which time the retraction of peripheral pseudopods was initiated. Frame at right shows the same osteoclast 40 minutes after treatment with calcitonin, with complete retraction of pseudopods. (Courtesy of S. Jeffrey Dixon and Stephen M. Sims, University of Western Ontario, London, Canada.)

An effect of CT that may indirectly impact bone metabolism is its apparent influence on vitamin D metabolism; in some species, the hormone stimulates formation of 1,25-dihydroxyvitamin D, which increases intestinal absorption of calcium.14 In humans, neither endogenous hypercalcitonemia (e.g., in medullary thyroid cancer [MTC]) nor diminished CT (e.g., in total thyroidectomy) is associated with alterations of serum calcium, nor are any marked effects seen on osseous metabolism. Accordingly, some researchers have postulated that the main function of CT in relation to calcium metabolism is to protect the skeleton during times of increased need, such as during growth, pregnancy, or lactation. The hormone may also combat postprandial hypercalcemia in the newborn.15 GASTROINTESTINAL EFFECTS Pharmacologically, CT decreases the secretion of pancreatic enzymes, gastrin, gastric acid, pancreatic glucagon, motilin, pancreatic polypeptide, gastric inhibitory peptide, and glucose-stimulated insulin; it also decreases the small intestinal secretion of potassium, chloride, sodium, and water. The gastrointestinal secretion of somatostatin is increased. Because some hormones that are found within the gastrointestinal tract stimulate the release of iCT, the existence of a food-related gut hormoneCT axis has been proposed.16,17 RENAL EFFECTS Renal cell membranes have receptors to CT, and this hormone decreases the tubular reabsorption of calcium and phosphate. Often this results in hypercalciuria and phosphaturia (see Chap. 206). In humans, CT increases the excretion of sodium, chloride, potassium, and magnesium. PULMONARY EFFECTS For a discussion of pulmonary effects, see the section Neuroendocrine Tumors Other Than Medullary Thyroid Cancer. CENTRAL NERVOUS SYSTEM EFFECTS The intracerebral administration of CT induces analgesia, decreases the intake of food and water, decreases gastric acid secretion, diminishes intestinal motility, andaugments the glucose-stimulated release of insulin.18 EFFECTS IN THE REPRODUCTIVE SYSTEM Patients with ovarian failure have been reported to have decreased circulating iCT concentrations and a decreased iCT reserve in the thyroidal C cells.19,20 Whether exogenous administration of estrogen increases the production of iCT by C cells is controversial.21,22 Interestingly, iCT is secreted by the pregnant uterus.23 This secretion is tightly regulated by the ovarian hormones estrogen and progesterone, which limit its expression to a brief period during the time of blastocyst implantation. CT is expressed in the preimplantation embryo, as shown in a mouse model.24 The binding of CT to its receptor activates adenylate cyclase and elevates cytosolic calcium levels, accelerating the development of the preimplantation embryo. Thus, CT might play a role in pregnancy during the implantation of the embryo. METABOLIC EFFECTS Experimentally, CT increases plasma lactate and glucose concentrations, inhibits insulin secretion, and, in the rat model, causes peripheral insulin resistance by inhibiting insulin-stimulated incorporation of glucose in glycogen.5 SERUM LEVELS OF IMMUNOREACTIVE CALCITONIN Serum iCT has been measured by bioassay, radioreceptor assay, and immunoassay. Bioassay studies (e.g., induced hypocalcemia in laboratory animals, in vitro generationof adenylate cyclase from renal cell membranes, inhibition of calcium-45 release from prelabeled mouse calvaria) have demonstrated that not all of the

immunologically detectable CT is bioactive. Because of differences in the structure of CT, often little or no immunologic cross-reaction occurs among species, except among some rodents. In humans, after total thyroidectomy some immunoassays still detect serum iCT, but at low levels. Nevertheless, the serum iCT of the thyroidectomized human does not respond to an intravenous calcium challenge. When serum iCT levels are being evaluated, only well-characterized assays should be used, and the assay conditions should not vary. In the past, most physiologic studies involving serum determinations have used techniques and antibodies that detect the immature CT (whether free or as part of precursor molecules), as well as mature CT. Mature CT assays (usually double-antibody techniques) use an antibody that specifically recognizes the carboxy terminus of CT. Previously, many investigators have reported that serum iCT levels are higher in men than in women.25 Some have reported that iCT levels are higher in children than in adults, and that the premature infant has higher levels than the full-term newborn, who has higher levels than the young child.26 Serum iCT levels do not change appreciably after meals. The existence of a circadian variation in serum levels is controversial. Values do not vary significantly during the menstrual cycle. Serum iCT is higher during pregnancy and may be increased during lactation. These findings have not been adequately verified using assays specific for mature CT. The basal levels of CT, when measured using a sensitive and specific immunoassay are <10 pg/mL (<3 fmol/mL). The endogenous secretion rate of iCT is 100 to 200 g per day. In humans, the half-life of intravenously administered CT is ~10 minutes, and the hormone is degraded predominantly by the kidney.27 Immunoreactive CT appears in the urine in a molecular form that is larger than the mature hormone. The specific chemical structure of this iCT is unknown. Values in the urine are also higher in the young than in the adult, and higher in men than in women when carboxy terminalrecognizing antisera are used.28 THERAPEUTIC USES OF CALCITONIN The CT preparation available in the United States for therapeutic purposes is synthetic salmon CT; the human peptide is no longer available for this purpose. Salmon CT differs from human CT by 16 of its 32-amino-acid sites(see Fig. 53-2). It is thought to have greater hypocalcemic potency in humans than does the human hormone (on a weight basis), greater binding affinity to target tissues, and slower degradation and clearance. When injected subcutaneously, local itching and redness at the site of injection are common but progressively diminish. Other side effects may include flushing, nausea, vomiting, and anorexia. Usually, these effects are mild and disappear or diminish with time; therefore, initial therapy should be at bedtime. A slight natriuretic effect may cause mild urinary frequency. In an occasional patient, the nausea is sufficiently bothersome to require cessation of therapy. The nasal preparation causes fewer systemic symptoms. PAGET DISEASE The most widespread and efficacious use of synthetic CT is for treatment of Paget disease (see Chap. 65). Usually, subcutaneous CT therapy (100 IU daily) results in a decrease of serum alkaline phosphatase, osteocalcin, and urinary hydroxyproline. Pain often is improved, the overlying warm skin may become cooler, the progress of the disease may be halted, and neurologic compressive symptoms may regress dramatically. Radiologically, an arrest of the lytic front and some remineralization of lytic lesions may be apparent. Moreover, CT has an analgesic effect in Paget disease that partially, but not entirely, depends on its effect on decreasing bone turnover. The progression of hearing loss, if present, may be halted.29 Another use of CT in Paget disease is to decrease hyperemia of involved bones before surgical procedures, such as hip replacement. To minimize nausea, the patient should start at half-dose levels for the first week. The nasal preparation, which has a markedly diminished effect in Paget disease, is not approved for the treatment of this condition. HYPERCALCEMIA Salmon CT has a role in the treatment of acute hypercalcemia.30,31, The drug is particularly useful for short-term therapy. It lowers the increased serum calcium by 0.5 to 1.5 mg/dL in ~75% of patients regardless of the cause (Fig. 53-6). It can be used in association with any other therapy (e.g., intravenous fluids, intravenous pamidronate). Salmon CT is extremely well tolerated in this setting and can be used safely in patients with congestive heart failure, renal disease, liver disease, or bone marrow deficiency. The principal value of CT is to take the top off the hypercalcemiaan obtunded patient may become more alert, diagnostic studies may proceed, and the physician need have no concern about possible contraindications. One of the most effective drug therapies for acute hypercalcemia, intravenous pamidronate, requires 48 to 72 hours to exert its hypocalcemic effect; consequently, salmon CT has a very useful role before this time. Occasionally a patient with mild hypercalcemia can be treated with long-term use of salmon CT, although resistance (CT escape) usually eventuates fairly rapidly; the cause of this early resistance, which also occurs when using human CT, is not known. Corticosteroid therapy has been alleged to nullify this escape phenomenon, but this has not been adequately documented, and such treatment adds the effects of harmful exogenous hypercortisolism to the clinical condition. Thus, intravenous pamidronate, repeated as needed, is a more appropriate therapy for the long-term control of hypercalcemia. Salmon CT is administered subcutaneously at a dosage of 4 IU/kg every 12 hours; in the authors' experience, a further increase of the dosage or the use of intravenous therapy seldom provides any additional benefit. Whether the nasal route of administration would be effective for hypercalcemia is not known, and it is not approved for this purpose.

FIGURE 53-6. Salmon calcitonin treatment of hypercalcemia of various etiologies. Mean serum calcium ( standard error of the mean) for 24 hypercalcemic patients (0-, 24-, 48-, 72-, and 96-hour determinations are before first morning injection); p value indicates statistical significance of difference from initial pretherapy serum calcium value. (From Wis-neski LA, Croom WP, Silva OL, et al. Salmon calcitonin in hypercalcemia. Clin Pharmacol Ther 1978; 24:219.)

OSTEOPOROSIS The use of CT in osteoporosis seems logical because many forms of this illness are caused by excessive bone resorption. The proposal has been made, based on no direct evidence, that a relative endogenous CT deficiency contributes to the development of osteoporosis. In this regard, many investigators found that serum iCT is lower in normal women than in men and responds less to a calcium infusion.32 Some investigators found that serum iCT is higher in blacks (who have greater bone density than whites). Some found that estrogen therapy increases basal levels of serum iCT in postmenopausal women and may augment the CT responsivity to intravenous calcium challenge. In addition, a decreased iCT reserve in response to stimulation testing has been reported in the elderly of both sexes. Osteoporotic women, in particular, have been reported to have poor iCT reserves, although this latter finding is disputed. (None of these aforementioned studies used assays specific for mature CT.) Nevertheless, in the patient undergoing thyroid hormone replacement, total thyroidectomy does not cause osteoporosis. Furthermore, patients with extremely high CT levels, such as those with MTC, do not have increased bone mass, nor does the uninvolved bone of elderly patients with Paget disease who have received long-term CT therapy appear to be influenced markedly. Several studies have evaluated the use of CT as therapy for established osteoporosis or for the control of its progression. In cases of postmenopausal osteoporosis, some investigators observed a reduction of bone loss, and others reported increased bone mineral calcium and decreased urinary total hydroxyproline.33,34 and 35 CT treatment of postmenopausal osteoporosis appears to be most effective in patients with increased bone turnover.36 The increase of bone mass in patients with post-menopausal osteoporosis appear to be less for CT than for alendronate.37

After several months to 1 to 2 years, the beneficial effects of CT therapy may become limited by a subsequent decline in bone formation. The intermittent use of CT (e.g., 2 or 3 months on, and 2 or 3 months off) may prolong the beneficial effects on bone.38 Both the subcutaneous and nasal preparations have been reported to increase axial and appendicular bone density, although not all studies agree. Evidence is increasing that the fracture rate is diminished,39,40 and 41 perhaps with greater efficacy than the augmentation of bone density might suggest. The dosage of subcutaneous salmon CT is 100 IU daily; the dosage of the nasal spray is 200 IU daily. The administration of either subcutaneous or nasal CT induces antibody formation in many patients after several months of use. Usually, these antibodies do not play a role in resistance to the action of this hormone, unless they are present in very high titer. Patient compliance is greatly increased when the nasal spray is used.42 Few studies are available concerning the efficiency of the nasal route in other conditions, and this route is approved only for osteoporosis. When treating osteoporosis, both the subcutaneous and nasal preparations should be supplemented with daily oral doses of 1000 mg of elemental calcium and 400 IU of vitamin D. Further studies are needed to systematically compare the effects of nasal CT with those of subcutaneous CT. CT has been used to treat the osteoporosis of immobilization as well as that caused by corticosteroid usage with rather disappointing results. In patients with temporal arteritis and poly-myalgia rheumatica who were starting high-dose, long-term corticosteroids, salmon CT with calcium and vitamin D provided no greater bone preservation than that observed with calcium and vitamin D alone.43 Therapy using this hormone has been tried in osteogenesis imperfecta with variable, but not dramatic, results. In treatment of the regional osteoporosis of Sudeck atrophy (reflex sympathetic dystrophy), CT has been alleged to diminish the pain, edema, and osteolytic process. OTHER USES In acute pancreatitis, the diminution of pancreatic enzyme secretion caused by the administration of CT has been reported to produce a quicker clinical and biochemical recovery. CT also has been used as therapy for pain of various types with variable results.44 Receptors for CT have been identified in some human cancer cells, including those of lung, breast, bone, prostate, and MTC, suggesting that an imaging agent for the receptors might be useful in nuclear oncology. Radiolabeled mature CT is not optimal for nuclear imaging, however, because of rapid breakdown and clearance from target tissues.45

PROCALCITONIN AND ITS DERIVATIVE PEPTIDES

The CALC-I gene encodes the information for the structure of the polypeptide precursor of CT, preprocalcitonin (Pre-ProCT). Early in posttranslational processing, the leader sequence is cleaved from this precursor molecule by a signal peptidase. The resultant prohormone, procalcitonin (ProCT, also termed PAN116), consists of 116 amino-acid residues (molecular mass 12.79 kDa) (see Fig. 53-3).The function of ProCT, which is present at low levels in normal human serum (<5 pg/mL; <0.4 fmol/mL),46 is unknown; an immunomodulatory role has been postulated. At the amino-terminus portion of ProCT is found a 57-amino-acid peptide, NProCT (also termed PAS-57), which has a molecular mass of 6.22 kDa. NProCT, which is readily detectable in normal human serum (1070 pg/mL; 1.612 fmol/mL), has no known definite function, although it has been reported to be a mitogen for osteoclasts.46,47 The immature CT is centrally placed within the ProCT molecule. This peptide, which also is present in normal serum, has no known function other than as a precursor of mature CT. The final 21-amino-acid residues comprise CCP-I (also termed PDN-21 and previously called katacalcin). As shown in Figure 53-1, the CALC-I gene can also produce a CCP-II peptide arising from a different form of CT mRNA. CCP-I, which differs from CCP-II in its eight terminal amino acids, occurs predominantly in thyroidal C cells and neuroendocrine malignancies such as MTC tissue, and also in the serum of normal persons, whereas CCP-II, which has been insufficiently studied, is found mostly in the pituitary and in neurologic tissues.48 Although these peptides may have a biologic function, none has yet been proven. Lastly, as part of the processing of the ProCT molecule, a peptide consisting of the conjoined immature CT and CCP-I peptide (CT:CCP-I) occurs in normal serum, where its function, if any, is unknown.46 All of the peptides that contain the CT molecule in its immature form (i.e., ProCT, CT:CCP-I, and immature CT) cross-react with antisera specific for the midportion of the immature CT molecule and, along with NProCT, have been termed CT precursors (CTpr), precalcitonin peptides, or large-molecular-weight CT. Regulated versus Constitutive Secretion. The extent of posttranslational processing of ProCT largely depends on whether the prohormone is secreted mainly by a regulated pathway, as in the thyroidal C cells, or a constitutive pathway.49 In regulated secretion, the Golgi apparatus identifies and selects ProCT to be routed to newly formed secretory vesicles, where the mature CT is produced and progressively concentrated. Subsequently, these electron-dense vesicles serve as storage repositories for later secretion. On the appropriate signal at the plasma membrane, a brief increased concentration of cytosolic free Ca2+ induces secretion (i.e., migration to the cell periphery, fusion with the apical portion of the cell membrane, and discharge of hormone contents in a quantal release). In regulated secretion, in addition to CT, these vesicles also contain and extrude variable concentrations of the other processed products of ProCT (i.e., NProCT and CCP-I). ProCT and CT:CCP-I are largely absent from the storage vesicles. In contrast, under certain conditions, constitutive secretion occurs, in which less-dense vesicles receive CT precursors, bud off from the trans-Golgi, and migrate rapidly to the cell surface. This nonstoring, continuous secretion of newly synthesized peptide(s) is a bulk-flow system by which hormonal material is rapidly secreted. Because of the lack of processing machinery within these different, rapidly transported and rapidly extruded vesicles, the relative proportion of CT in relation to precursor peptides is considerably diminished. In the case of the ProCT molecule, this form of secretion may occur to a small extent in normal persons, whereas in some pathophysiologic conditions, it is greatly increased (see later). It is uncertain why some tumors (e.g., MTC, small cell lung cancer [SCLC], bronchial carcinoid) secrete relatively small amounts of CT in comparison with large amounts of CTpr despite their abundant number of dense-core secretion vesicles, which are characteristic of regulated secretion. DISORDERS ASSOCIATED WITH ABERRANT PRODUCTION AND SECRETION OF CALCITONIN PRECURSORS AND DERIVATIVE PEPTIDES MEDULLARY THYROID CARCINOMA Although the assay for CT has become the standard marker for the clinical detection and follow-up of MTC, CT precursors and other derivative peptides are also increased in the sera of these patients, frequently to a greater extent than is the CT.50 Furthermore, some of these lesions secrete CGRP and precursor CGRP peptides as well. Nevertheless, the time-honored marker remains mature CT, and care should be taken to use the most specific assay.51 Nearly all MTC patients have increased basal serum CT (see Chap. 40 and Chap. 188), and nearly all patients with MTC respond markedly to a calcium and/or pentagastrin challenge, even when in a subclinical stage (i.e., C-cell hyperplasia). A patient with a thyroid mass and increased basal CT probably has MTC, and this usually can be readily confirmed by a fine-needle aspiration biopsy and immunocytochemical staining. In most sporadic cases of MTC, the basal serum CT level usually has not been obtained preoperatively, and the diagnosis most often is made by examination of the surgical specimen. In this respect, some clinicians believe that a screening basal serum CT level should be measured preoperatively for all nodular thyroid disease.52 Furthermore, preoperative basal CT levels are predictive of the tumor size and offer information concerning the probability of postoperative normalization of the hormone level.53 In those patients proven to have MTC, follow-up CT testing after surgery is essential to detect persistent local disease or to document the presence of metastases; later, long-term follow-up is essential to detect a subsequent recurrence or further tumor growth. In the past, the major use of provocative testing had been for the initial screening of the family members of a patient with MTC to help determine whether the patient had the sporadic or the familial variety of the disease, and also for the screening and follow-up of relatives of known familial cases. Currently, genetic studies (i.e., Ret protooncogene) have largely replaced the use of provocative testing of family members of an index case with known disease for the purpose of determining whether any specific relative is affected. In the future, provocative testing will probably become limited to the follow-up examination of individual nonhypercalcitonemic family members previously shown to possess the genetic abnormality and for whom prophylactic thyroidectomy is not selected (see Chap. 188). The authors' preference had been to perform a combined calcium and pentagastrin stimulation, because the effect of the latter hormone is additive with that of induced hypercalcemia. Pentagastrin often causes marked epigastric and chest pain, however, and is not an approved drug in the United States. Intravenous stimulation testing of family members is onerous and, for children, frightening. For family members and for patients with diagnosed C-cell hyperplasia whose serum CT has normalized after surgery, the measurement of iCT in a single urine specimen has proved to be more sensitive than serum measurements. The urine assay readily detects early disease in a family member or recurrent disease in

the index case and often obviates the need for stimulation testing.54 NEUROENDOCRINE HYPERPLASIAS The study of some chronic nonneoplastic pulmonary conditions has demonstrated a slight to moderate increase of CTpr. Usually, levels of serum CT remain within the normal range. These conditions include chronic bronchitis, chronic obstructive pulmonary disease, chronic pulmonary tuberculosis, and cystic fibrosis. PNE cells (see later) are hyperplastic in some chronic lung conditions and may be the cause of this phenomenon.55 A similar occurrence may be encountered in inflammatory bowel disease (e.g., regional enteritis), in which iCT-producing enteric neuroendocrine cells may undergo hyperplasia. In the case of the PNE cells, one of the exogenous stimuli to PNE cell hyperplasia may be cigarette smoke. In hamsters and in humans, cigarette smoke immediately increases serum iCT, probably due to the nicotine content of this smoke. Cholinergic-nicotinic receptors are found on the PNE cell, and nicotine stimulates the growth of these cells. A similar phenomenon occurs after exposure to the carcinogen diethylnitrosamine, which has nicotinic characteristics.56 NEUROENDOCRINE TUMORS OTHER THAN MEDULLARY THYROID CANCER The PNE cells, which have electron-dense vesicles containing iCT as well as other peptides, are situated near the basement membrane of the human trachea, larynx, and bronchial tree.57 They are particularly common within the lungs of the newborn, where they often are strategically located in clusters called neu-roepithelial bodies. In both the newborn and adults, the PNE cells secrete their contents into the bronchial lumen and also communicate with other types of pulmonary cells in a paracrine manner. Functionally, PNE cells appear to play a role in pulmonary maturation, transcellular and intracellular movements of calcium, and chondrogenesis of the bronchial tree; also, perhaps they locally modulate pulmonary blood flow (see Chap. 177). The iCT that is secreted in the short term by these cells is mostly the mature CT hormone. SCLC, the third most common form of lung cancer, is a very virulent tumor. It is nearly always associated with cigarette smoking. Electron microscopy reveals that these cells are strikingly similar to PNE cells; they contain similar secretion granules, and contain and secrete iCT. This cancer as well as the more benign pulmonary carcinoid tumor are believed to arise from PNE cells or their progenitors.58 Although these tumor cells contain mostly CT intracellularly, in contrast to PNE cells, they secrete predominantly CTpr. The carcinoid tumor, which occurs in extrapulmonary locations (e.g., small intestine), and extrapulmonary SCLC behave in a similar fashion. Occasionally, other neuroendocrine tumors may contain and secrete iCT (e.g., pancreatic islet cell cancer, pheochromocytoma, thymoma). Furthermore, pulmonary neoplasms other than SCLC or carcinoid tumor may be associated with high levels of iCT in the serum (e.g., pulmonary epidermoid cancer, adenocarcinoma, or large cell cancer). This may be due to an admixture of SCLC cells within the tumor, hyperplasia of PNE cells in the vicinity of the cancer, or the smoking-related PNE cell hyperplasia that occurs in some of these affected patients.59 In the case of SCLC and pulmonary and extrapulmonary carcinoid tumors, as well as some of the above-mentioned iCT-secreting neoplasms, the sequential measurement of serum iCT may be a very useful marker for the presence of these lesions and their response to therapeutic measures (i.e., surgery, irradiation, and/or chemotherapy).60 In such cases, the measurement of serum CTpr, which are always present in considerably higher concentrations than is CT, is preferable. INFLAMMATION, INFECTION, AND SEPSIS Potent stimuli (e.g., trauma, extensive surgery, burns, pancreatitis, stroke, or bacterial infection) often initiate a marked augmentation of proinflammatory cytokines and other humoral substances that may result in inflammation. Locally, among other phenomena, inflammation is characterized by vasodilation and a concomitantly increased blood supply, attraction of leukocytes and lymphocytes, activation of macrophages, alteration of capillary function, and transudation into the tissues of serum along with its protein, peptide, and electrolyte constituents. Inflammation may be systemic, in which case the patient may note malaise, loss of appetite, or muscle aching. Fever or hypothermia, leukocytosis, or leukopenia, tachycardia, and/or tachypnea may be present. Systemic inflammation may be self-limited or may progress to multiple organ dysfunction, which is characterized by varying degrees of hypoxemia, myocardial insufficiency, hypoperfusion, diminished mental status, and circulatory dysfunction. Further complications include coagulopathy, acute respiratory distress syndrome, cardiac failure, shock, renal failure, and/or coma (see Chap. 227 and Chap. 228). The term sepsis has been used to describe a condition in which a patient's cytokine and humoral response to an infection leads to some of these marked symptoms, signs, and complications. In sepsis, the patient is ill, not primarily because of the initial injury or infection, but because of the harmful patho-physiologic overreaction by the host. Calcitonin Precursors as a Marker for Inflammation, Infection, and Sepsis. Although the serum levels of several humoral substances may be increased in the patient with systemic inflammation, infection, or sepsis, very few are reliable markers for the presence of the conditions, their clinical course, the response to therapy, or the prognosis. Data that have been accumulating since the 1980s indicate that large-molecular-weight CT-containing peptides (i.e., CTpr) as well as ProCT itself are dramatically increased in these conditions, attaining levels ranging from moderate elevations to thousands of times higher than normal.7,61,62 and 63 In all of these illnesses, although the relative proportions of the different peptides may vary, serum levels of CT remain normal or only modestly increased because of incomplete processing and/or more rapid degradation. Several techniques have been used to quantitate the CTpr of patients with inflammation, infection, and sepsis (e.g., gel filtration or high-performance liquid chromatography with associated immunoassay, and single-site assay, or two-site immunoassays). When selecting an assay, measuring more than one of the CTpr may be best, because of the varying degrees of increases in the serum of any individual patient. The two current assays include a single-site NProCT assay, which detects NProCT, ProCT, and procalcitonin generelated peptide, and a two-site assay, which detects ProCT and CT:CCP-I.63,64 When these assays are used, most patients with marked illness are found to have increased serum CTpr, and values tend to correlate with the severity of the condition and with the prognosis. Indeed, serum CTpr levels have now become the best markers, and one can follow the day-to-day course of the patient with considerable reliability.65 Stimulus to Increased Serum Calcitonin Precursor Levels in Systemic Inflammation, Infection, and Sepsis. The injection of endotoxin into normal human volunteers provokes an early increase of tumor necrosis factor-a (TNF-a) and interleukin-1b (IL-1 b) Injection of TNF-a into normal hamsters provokes a marked increase of CTpr, reaching levels comparable to those observed in septic animals.66 Various injuries (e.g., burns or trauma) are known to induce translocation across the bowel wall of endotoxin that normally is contained within the cell walls of the Escherichia coli flora of the gut. Endotoxin arising endogenously from the gut or exogenously from an acquired infection likely is the initial triggering factor that precipitates the cytokine cascade leading to increased CTpr. Bacterial exotoxins, other known precipitants of a cytokine cascade, may constitute additional offenders. Calcitonin Precursors as a Toxic Factor in Sepsis. In experimental systemic sepsis, CTpr have been found to contribute to the severity and mortality of the condition. In hamsters with induced sepsis due to peritonitis, the mortality is doubled by the addition of human ProCT but not by the addition of CT. Furthermore, in these septic animals, immunoneutralization of ProCT using an antiserum to the CT portion of ProCT improves outcome and reduces mortality.67 In a similar model of sepsis in pigs, the preventive immunoneutralization of ProCT using an antiserum to the N-ProCT portion of ProCT markedly mitigates the onset and symptomatology of sepsis and prolongs life.68 Origin of the High Serum Calcitonin Precursor Levels in Systemic Inflammation, Infection, and Sepsis. The levels of CTpr in septic tissues (e.g., kidney, liver, lung) are significantly increased compared with levels in tissues from healthy controls. Reverse transcriptase polymerase chain reaction studies in septic hamsters and pigs reveal detectable CT mRNA, not only in the thyroid gland, but also in all extrathyroidal tissues studied (e.g., kidney, liver, lung, brain, fat, muscle, spleen, testes, etc.). In contrast, in control hamsters, CT mRNA is detectable only in the thyroid and lung. Furthermore, in situ hybridization studies reveal that multiple cell types within these organs and tissues are involved. In this respect, during the marked host response, the whole organism is transformed into an endocrine gland.69 Presumably, the increase of CALC gene transcription is mediated by one or several stimulus-specific sepsisresponse elements in the gene promotor. Cause of Calcitonin Precursor Toxicity in Sepsis. In preliminary and limited experiments in healthy hamsters, the injection of human ProCT did not cause mortality. A mild hyperglycemia and a slight decrease of serum calcium were noted. In vitro, the addition of human ProCT to human endotoxinstimulated mononuclear cells of peripheral blood markedly increased the ambient levels of proinflammatory cytokines (i.e., TNF-a, IL-1b and interleukin 8)70; presumably this would bea harmful occurrence in sepsis. Furthermore, the administration of human ProCT to normal hamsters blocks the marked hypocalcemic effect of administered CT, suggesting that ProCT blocks tissue receptors for CT.71 In this respect, previous studies have shown that ProCT does indeed bind to CT receptors. Such a blockage becomes potentially even more significant when one considers the extremely long half-life of serum ProCT (probably because of prolonged stimulus to secretion and the resistance to degradation) that occurs after endotoxin injection in humans. Thus, one could speculate that the toxicity of ProCT, may, in part, be related to receptor blockade, which prevents the various peptides of the CT gene family (e.g., CT, CGRP, amylin, and ADM) from exerting an otherwise beneficial effect in sepsis.

CALCITONIN GENERELATED PEPTIDES

ALTERNATIVE GENE EXPRESSION The CALC-I gene gives rise to more than one bioactive peptide: CT and CGRP. This occurs because of alternative processing of the primary RNA transcript72 (see Chap. 3; see Fig. 53-1). Two very similar human CGRPs are found: CGRP-I (or a) from the CALC-I gene and CGRP-II (or b) from the CALC-II gene. CGRP-I is a 37-amino-acid hormone; in humans, CGRP-II differs from CGRP-I by only three amino acids (see Fig. 53-2).73 The CGRPs have only a modest homology with CT. As is the case for CT, however, the CGRPs have an NH 2-terminal disulfide bridge and a COOH-terminal amide. The CALC-II gene does not seem to be alternatively expressed. The CALC-I and CALC-II genes, plus a pseudogene, CALC-III, probably arose by local duplication of a common ancestral gene.74 Unlike with CT, the structure of the CGRPs is highly conserved in different animal species. CT and CGRP-I can be coexpressed by a single cell, and several tissues contain both of them.75 However, usually a preference is seen for synthesis of either CT or CGRP-I according to the cellular type and localization. In the C cell, CT is preferentially produced; in the nervous system, CGRP-I predominates. The decision concerning the prevailing mode of synthesis is not immutable; for example, it can be modified by dexamethasone and by the naturally occurring fatty acid butyrate. Whether the expression of CALC-I and CALC-II genes is independently regulated is uncertain. CGRP-I and CGRP-II occur in tissues of the normal human.76 They are also found in some tumors, such as SCLC.77 In the rat, in situ hybridization demonstrates a similar tissue distribution of the CGRP-I and CGRP-II mRNAs. Most immunohistochemical studies of specific tissue localization have used antisera that probably cannot differentiate the very similar CGRP-I and CGRP-II; consequently, when location and action are discussed, the more noncommital term CGRP is used. CGRP is predominantly a neuropeptide, found mostly in various regions of the brain, cranial nerves, ganglia, and spinal cord. In peripheral nerves, CGRP is often costored and released with other neurotransmitters. Within the gastrointestinal tract, it is found in the myenteric plexus and in neuro-endocrine cells. It also is found in the pituitary gland, the adrenal medulla, the pancreatic islets, the pulmonary nerves, and the PNE cells. The peptide circulates in a heterogeneous form in the blood. Serum levels of immunoreactive CGRP are higher in females than in males, and oral contraceptive use further increases these levels. Levels are normal in as many as 50% of patients with MTC, and its measurement does not improve detection of this tumor. EFFECTS Many receptor sites react to CGRP in the nervous system and peripheral tissues; most, but not all, of these are not linked to adenylate cyclase (see the section Receptors of the Calcitonin Gene Family of Peptides). As with CT, induced hypercalcemia can stimulate the release of CGRP, but in the case of this latter peptide, this probably is a nonphysiologic phenomenon. CGRP acts directly on rat osteoclasts to inhibit bone resorption.78 Although administration of this hormone in some laboratory animals has a hypocalcemic effect, however, this does not occur in humans. CGRP functions as a peptidergic messenger and is involved in neurocrine modulation. Capsaicin, the sensory nerve neuro-toxin that is found in hot peppers, depletes CGRP, suggesting that this peptide is involved in sensory function. Intracere-brally, the hormone decreases food intake and gastric acid secretion. Commonly, CGRP is colocalized with substance P and sometimes is found with somatostatin, galanin, vasoactive intestinal peptide, or cholecystokinin.79 This association with other peptides may have functional consequences. For example, the synergistic action of CGRP with substance P and the other neurokinins may induce plasma protein leakage from the vasculature and promote the inflammatory response. Experimentally, the administration of CGRP has many effects, most of which may have no physiologic relevance but some of which may reflect a physiologic neurosensory influence. Using various experimental models, some have reported pharmacologic effects including flushing of skin (related to peripheral arteriolar vasodilatation), hypotension, chronotropic and inotropic cardiac effects, contraction of smooth muscle of the gut, and inhibition of contraction of smooth muscle of the uterus and vas deferens. The extremely potent vasorelaxant actions of CGRP are mediated partly by cAMP-dependent mechanisms. The release of this hormone from vascular nerve endings may be of importance in the physiologic control of blood pressure. Although blood levels of CGRP are only slightly increased in sepsis, an increase in paracrine secretion of CGRP may play an important role in sepsis-induced vasodilation and shock. In vitro studies suggest that CGRP may be implicated in the regulation of B-lymphocyte development and T-lymphocyte function.80 The hormone exerts positive inotropic and chronotropic effects. The CGRPs may have a role in pulmonary pathophysiology. CGRP-containing pulmonary neuroendocrine cells are increased in infants with chronic lung disease associated with broncho-dysplasia. CGRP and other sensory peptides are potentially capable of producing many of the characteristic features of asthma, including bronchoconstriction and edema.81 CGRP also induces a dose-dependent relaxation of the spontaneously contracting myometrium of pregnant women. Infusion of pharmacologic doses of CGRP has metabolic effects similar to those of amylin (i.e., it also induces insulin resistance, decreases peripheral glucose clearance, and increases hepatic glucose output).82 CGRP, similiarly to CT, may also have central effects: These include alteration of food intake and release of growth hormone and prolactin from the pituitary.83 Interestingly, increased serum levels of CGRP are found in patients with migraine, even between attacks, suggesting a possible role in this disorder.84

AMYLIN
Amylin was first isolated from amyloid deposits in an insulinoma, but it is also found in the amyloid of pancreatic islets of patients with type 2 diabetes, as well as in normal islets. Pancreatic amyloid is formed by the aggregation of amylin. Consequently, the hormone is also known as islet amyloid polypeptide. This 37-amino-acid peptide originates from the CALC-IV gene (see Fig. 53-1). It has 43% sequence homology with CGRP-I and 49% homology with CGRP-II, as well as overlapping bioactivity.85. EFFECTS Amylin is packaged within B-cell granules and is cosecreted with insulin in response to hyperglycemia. In the human fetal pancreas, amylin is colocalized within the pancreatic neuroendocrine cells, along with insulin and glucagon. Intraislet somatostatin appears to be an inhibitory regulator of both of these peptide hormones.86 In obese insulin-resistant patients, amylin values increase ten-fold after meals or glucose, whereas in patients with type 1 diabetes, serum values remain low or absent. Various metabolic actions of amylin are related to the regulation of fuel metabolism, but whether they are physiologic or pharmacologic is uncertain. Thus, in muscle, the hormone opposes glycogen synthesis and activates glycogenolysis and glycolysis. The resultant increased lactate output provides substrate for hepatic gluconeogenesis and glycogen synthesis (Cori cycle).Presumably, the increased serum glucose reflects the gluconeogenesis. Intracerebrally, amylin exhibits an anorectic action. In experimental models, amylin causes insulin resistance by inhibiting glycogen synthesis in muscle, but whether serum amylin levels are increased in humans with insulin resistance is controversial. In the kidney, amylin stimulates the secretion of renin. Because of its restraining effect on gastric emptying,87 the insulin agonist pramlintide has been suggested as a potential therapy to mitigate hypoglycemia in insulin-treated patients with type 1 diabetes. After a long period of therapy with this agent, however, a potential risk of inducing systemic amyloidosis may be present. In this respect, pancreatic amyloid may be pathogenic88; amyloid fibrils, which accumulate in the islets of patients with type 2 diabetes, are cytotoxic to pancreatic cell lines and may be directly associated with the development or the worsening of type 2 diabetes. Furthermore, amylin fibrils, found in Alzheimer disease, promote the production of proinflammatory cytokines such as TNF-a,IL-1b, and IL-6, and may contribute to the pathology of this disease.89 In the rat, amylin appears to act centrally to elevate blood pressure, perhaps through activation of the renin-angiotensin system. However, pramlintide does not alter blood pressure in humans. Amylin has weak osteoclast-inhibitory and hypocalcemic actions and, like CT and CGRP, has an antiinflammatory action. Amylin activates CGRP receptors and thereby induces vasodilation. Like CT, amylin has a gastroprotective action in some models of gastric ulcer.

ADRENOMEDULLIN
ADM, originating from the CALC-V gene (see Fig. 53-1 and Fig. 53-2), is a 52-amino-acid peptide originally isolated from human adrenal pheochromocytoma.90,91 As is the case for iCT, a significant fraction of circulating immunoreactive ADM (iADM) consists of one or more precursor peptides with potential bioactivity. ADM is produced by and secreted from vascular endothelium, smooth musclecells, fibroblasts, monocytes/macrophages, and many other cell types.92,93 Immunoreactive ADM has been detected in many tissues throughout the body (e.g., kidney, lung, skin). ADM is cleared by the kidneys. High levels of the hormone are found in the urine, and serum levels increase with renal failure.94 EFFECTS Endothelial cells produce ADM and express specific receptors linked to the cAMP second messenger system.95 ADM is a potent vasodilator, apparently via the generation of nitric oxide. Alternative cleavage products of the ADM prohormone, adrenotensin, have been reported to have an opposite effect on vascular tone. Hence, ADM may function not only as a circulating vasodilating hormone but also as a local regulating agent for vascular tone. It exerts positive inotropic and chronotropic effects. Other effects of ADM include stimulation of insulin secretion, inhibition of adrenocorticotropic hormone (ACTH) secretion, inhibition of angiotensin IIinduced aldosterone secretion, bronchodilation, natriuresis and diuresis, delay of gastric emptying, immunomodulation, stimulation of the growth of osteo-blasts, and an antiinflammatory activity.96,97,98 and 99 PATHOPHYSIOLOGY Lipopolysaccharide and inflammatory cytokines (e.g., IL-1 and TNF) are potent stimulants of ADM production.100 In patients with septic shock, serum ADM concentrations are increased compared with controls, and the levels correlate with hemodynamic disturbances (i.e., high cardiac output and low vascular resistance).101 Serum ADM is elevated in patients with essential hypertension, congestive heart failure, myocardial infarction, and chronic obstructive pulmonary disease. The N-terminal peptide of proadrenomedullin appears to function presynaptically to inhibit adrenergic nerves that innervate blood vessels. This latter peptide is increased in patients with essential hypertension. Its vasodilating effect is considerably weaker than that of ADM.

RECEPTORS OF THE CALCITONIN GENE FAMILY OF PEPTIDES

The peptides of the CT-gene family exert their bioeffects through binding to characteristic receptors with seven transmembrane hydrophobic domains. The receptors belong to the G proteincoupled receptors, in which the guanidine nucleotide guanosine triphosphate mediates receptor function by binding to specific mediator proteins.5 The receptors are categorized into two major subgroups, CT receptors (CRs) and CT receptorlike receptors (CRLRs). The different members of the CT family of peptides bind with different affinity to CRs and to CRLRs.5 Therefore, their bioeffects are in part overlapping. CALCITONIN RECEPTORS Several isoforms of the CR are found. They are derived from a single gene by alternative splicing of mRNA, which results in variable amino-acid sequence, length, and functional properties of the receptor domains.102 The CRs consist of a single glycosy-lated polypeptide of 70 to 90 kDa, equivalent to 450 to 500 amino acids. CRs are related to receptors of parathyroid hormone and parathyroid hormonerelated protein, secretin, vasoactive intestinal peptide, pituitary adenylate cyclaseactivating peptide, growth hormonereleasing hormone, glucagon-like peptide-1, and glucagon. This suggests that these receptors may comprise a distinct subfamily of G proteincoupled receptors.103 The CRs are widely distributed and are particularly numerous in osteoclasts of the bone, in cells of the distal nephron of the kidney, and in the central nervous system.104 CRs are also expressed in preimplantation embryos in mice. As mentioned earlier, CT is thought to be involved in blastocyst differentiation and implantation. Human T lymphocytes have highaffinity CRs on their surfaces. Thus, the markedly increased CTpr levels seen in infection and sepsis might be immunoregulatory.105 The bioactivity of CR is coupled to several signal-transduction pathways. Stimulation of the CR involves binding of a guanine nucleotidecoupled regulatory protein. This binding is coupled with the activation of adenylate cyclase with cAMP formation and activation of protein kinase A.106 The same receptor is also coupled to activation of phosphoinositol-dependent phospholipase C, which results in calcium mobilization, and protein kinase C activation.107 The two signal-transduction pathways require the presence of mediator proteins (namely, the cholera toxinsensitive Gs and the pertussis toxinsensitive Gi). Coupling of the CT-CR complex to these distinct G proteins leads to opposite bioeffects. For example, the Na+ pump is stimulated by protein kinase A and inhibited by protein kinase C. Furthermore, selective activation of the pathways can occur in a cell cycledependent manner because Gs is activated in G2 phase and no longer stimulated in S phase. Conversely, CRs couple to Gi in S phase, but not in G2 phase.108 The expression and activity of the CR is strongly regulated. CRs exhibitconstitutive signaling activity (i.e., they cause elevation of intracellular cAMP even in the absence of CT agonist).This may be a mechanism to adapt their signaling properties in the absence of agonist, so that the cell is more sensitive to the hormone.109 Conversely, treatment of mature osteoclasts with CT causes a rapid and prolonged decrease in CR mRNA levels, which is reflected in a loss of cell-surface CRs. Because these cells regain resorption competence, the data support the concept that the clinical phenomenon of escape from the hypocalcemic action of CT results from a loss of sensitivity of osteoclasts to the hormone.110 CALCITONIN RECEPTORLIKE RECEPTORS The classic ligands for the CRLRs had been considered to be CGRP-I and CGRP-II. At least two isoforms of the receptor were postulated, according to the presence or absence of antagonistic activity of the 8-37 fragment of CGRP-I. The molecular sequence of a CGRP receptor type 1 has been cloned. This receptor has a significant peptide sequence homology with the CR and shares similarities with this receptor in general structure and length.111 The CRLRs are expressed in a variety of tissues. They are most prevalent in various brain areas, where the regional distribution is well conserved among different species. In addition, they are expressed in high densities in the cardiovascular system, especially within the intima and media layers of blood vessels. In the heart, CGRP binds in highest densities in coronary arteries, heart valves, and the right atrium. CRLRs are also expressed in the adrenal and pituitary glands, exocrine pancreas, kidney, and bone. RECEPTOR-ACTIVITYMODIFYING PROTEIN AND THE PLASTICITY OF THE RECEPTORS FOR THE HORMONES OF THE CALCITONIN GENE FAMILY OF PEPTIDES The search for the specific receptors for the peptides of the calcitonin gene family has been arduous and elusive. Identifying a specific and unique receptor for any one of these peptides has been difficult or impossible. Rather, a complex and fascinating overlapping of receptivity by the candidate receptors appears to exist. This phenomenon is related to the multi-potentiality of the two principal receptors (or classes of receptors): CRs and CRLRs. These receptors manifest alternative physiologic profiles, which are conferred on them by the actions of certain accessory proteins; these allow a response to CT, to CGRP, to amylin, and/or to ADM ina dynamic and varying manner. RAMPs, or receptor-activitymodifying proteins, are single-transmembrane-domain proteins that alter the phenotype of both CRs and CRLRs.112 Three RAMPs are known: RAMP-1, RAMP-2, and RAMP-3. RAMPs act on CRs, probably by modifying the receptor gene. They act on the CRLRs, however, by influencing their transport to the plasma membrane. The specific cellular phenotype of the receptors that is ultimately expressed on the cell surface is determined by the presence, concentration, and/or timing of one or more of the three known RAMPs. Action of Receptor-ActivityModifying Proteins on Calcitonin Receptors. RAMP-1 induces a high-affinity amylin receptor, which also responds well to CGRP but poorly to CT (Table 53-1). RAMP-2 induces receptors with a high affinity for CT, a lower affinity for CGRP, and a weak binding to amylin. RAMP-3 induces a high-affinity amylin receptor with poor affinity for CGRP.

TABLE 53-1. RAMP-Induced Heterogeneity of Receptor Affinity to the Calcitonin Gene Family of Peptides and the Postulated Effects

Action of Receptor-ActivityModifying Proteins on Calcitonin ReceptorLike Receptors. RAMP-1 induces a CRLR with a strong affinity for CGRP. RAMP-2 or RAMP-3 induces a receptor with strong affinity for ADM (see Table 53-1). Impact of the Receptor-ActivityModifying Proteins. RAMPs, the expression of which is subject to humoral influences,113 effect an astonishingly elegant regulation and diversification of receptor function. They induce alternative profiles in the receptors for the peptides of the calcitonin gene family and presumably modulate hormonal responsivity according to ambient needs.

CONCLUSION
Although much remains to be elucidated, knowledge concerning CT has increased enormously. The role of the peptide, first thought to be an importantregulator of serum calcium, is still elusive. In spite of its as yet imperfectly defined functions, CT has become an important therapeutic agent. Furthermore, appreciable amounts of CTpr are now known to circulate in normal persons, and their potential hormonal roles require elucidation. Several common and often life-threatening conditions are associated with dramatic increases of these peptides. One of these peptides, ProCT, may contribute markedly to the morbidity and mortality of some of these conditions, and its immunoneutralization offers considerable therapeutic promise. Importantly, other related peptides have been discovered that comprise a calcitonin gene family. This family includes the CGRPs, amylin, and ADM. These latter peptides have some overlapping effects, as well as some very unique functions. Thus, that which conceptually began as a single hormone has broadened to include a great panoply of known or potential hormones. CHAPTER REFERENCES
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. Le Moullec JM, Jullienne A, Chenais J, et al. The complete sequence of human preprocalcitonin. FEBS Lett 1984; 167:93. Steenbergh PH, Hoppener JW, Zandberg J, et al. Structure and expression of the human calcitonin/CGRP genes. FEBS Lett 1986; 209:97. Nishi M, Sanke T, Seino S, et al. Human islet amyloid polypeptide gene: complete nucleotide sequence, chromosomal localization, and evolutionary history. Mol Endocrinol 1989; 3:1775. Yeakley JM, Hedjran F, Morfin JP, et al. Control of calcitonin/calcitonin gene-related peptide pre-mRNA processing by constitutive intron and exon elements. Mol Cell Biol 1993; 13:5999. Wimalawansa SJ. Amylin, calcitonin gene-related peptide, calcitonin, and adrenomedullin: a peptide superfamily. Crit Rev Neurobiol 1997; 11:167. Becker KL, Snider RH, Moore CF, et al. Calcitonin in extrathyroidal tissues of man. Acta Endocrinol (Copenh) 1979; 92:746. Becker KL, Silva OL, Snider RH, et al. The pathophysiology of pulmonary calcitonin. In: Becker KL, Gazdar AF, eds. The endocrine lung in health and disease. Philadelphia: WB Saunders, 1984:chap 16,277. Copp DH, Cameron EC, Cheney BA, et al. Evidence for calcitonina new hormone from the parathyroid that lowers blood calcium. Endocrinology 1962; 70:638. Suzuki K, Kono N, Onishi T, Tarin S. The effect of hypermagnesemia on serum immunoreactive calcitonin levels in normal human subjects. Acta Endocrinol (Copenh) 1987; 116:282. Tabassian AR, Nyln ES, Giron AE, et al. Evidence for cigarette smoke-induced calcitonin secretion from lungs of man and hamster. Life Sci 1988; 42:2323. Nyln ES, Becker KL, Snider RH Jr., et al. Cholinergic-nicotinic control of growth and secretion of cultured pulmonary neuroendocrine cells. Anat Rec 1993; 236:129. Chambers TJ, Chambers JC, Symonds J, Darby JA. The effect of human calcitonin on the cytoplasmic spreading of rat osteoclasts. J Clin Endocrinol Metab 1986; 63:1080. Wergedal J. The bone anabolic effect of antiresorptive drugs. In: Chris-tiansen C, ed. Postmenopausal osteoporosis. State of the art 1994. Rodovre, Denmark: Osteopress ApS, 1994:20. Jaeger P, Jones W, Clemens TL, Hayslett JP. Evidence that calcitonin stimulates 1,25-dihydroxyvitamin D production and intestinal absorption of calcium in vivo. J Clin Invest 1986; 78:456. Stevenson JC, Hillyard CJ, MacIntyre I, et al. A physiological role for calcitonin: protection of the maternal skeleton. Lancet 1979; 2:769. Care AD, Bruce JB, Boelkins J, et al. Role of pancreozymin-cholecystokinin and structurally related compounds as calcitonin secretogogues. Endocrinology 1971; 89:262. Cooper CW, Schwesinger WH, Ontjes DA, et al. Stimulation of secretion of pig thyrocalcitonin by gastrin and related hormonal peptides. Endocrinology 1972; 91:1079. Greeley GH Jr, Cooper CW, Jeng YJ, et al. Intracerebroventricular administration of calcitonin enhances glucose-stimulated release of insulin. Regul Pept 1989; 24:259. Taggart HM, Chesnut CH, Ivey JL, et al. Deficient calcitonin response to calcium stimulation in postmenopausal osteoporosis? Lancet 1982; 1:475. Stevenson JC, Whitehead MI. Calcitonin secretion and postmenopausal osteoporosis. [Letter]. Lancet 1982; 1:804. Hurley DL, Tiegs RD, Barta J, et al. Effects of oral contraceptive and estrogen administration on plasma calcitonin in pre- and postmenopausal women. J Bone Miner Res 1989; 4:89. Dick IM, Prince RL. Transdermal estrogen replacement does not increase calcitonin secretory reserve in postmenopausal women. Acta Endocrinol (Copenh) 1991; 125:241. Kumar S, Zhu LJ, Polihronis M, Armant DR. Progesterone induces calcitonin gene expression in human endometrium within the putative window of implantation. J Clin Endocrinol Metab 1998; 83:4443. Wang J, Rout UK, Bagchi IC, et al. Expression of calcitonin receptors in mouse preimplantation embryos and their function in the regulation of blastocyst differentiation by calcitonin. Development 1998; 125:4293. Heath HD, Sizemore GW. Plasma calcitonin in normal man. Differences between men and women. J Clin Invest 1977; 60:1135. Samaan NA, Anderson GD, Adam-Mayne ME. Immunoreactive calcitonin in the mother, neonate, child and adult. Am J Obstet Gynecol 1975; 121:622. Simmons RE, Hjelle JT, Mahoney C, et al. Renal metabolism of calcitonin. Am J Physiol 1988; 254:F593. Snider RH, Moore CF, Silva OL, Becker KL. Radioimmunoassay of calcitonin in normal human urine. Anal Chem 1978; 50:449. Nyln ES, O'Neill W, Jordan MH, et al. Serum procalcitonin as an index of inhalation injury in burns. Horm Metab Res 1992; 24:439. Silva OL, Becker KL. Salmon calcitonin in the treatment of hypercalcemia.Arch Intern Med 1973; 132:337. Wisneski LA, Croom WP, Silva OL, et al. Salmon calcitonin in hypercalcemia. Clin Pharmacol Ther 1978; 24:219. Garcia-Ameijeiras A, De La Torre W, Rodriguez-Espinosa J, et al. Does testosterone influence the post-stimulatory levels of calcitonin in normal men? Clin Endocrinol (Oxf) 1987; 27:545. Mazzuoli GF, Passeri M, Gennari C, et al. Effects of salmon calcitonin in postmenopausal osteoporosis: a controlled double-blind clinical study.Calcif Tissue Int 1986; 38:3. Gnudi S, Mongiorgi R, Moroni A, Bertocchi G. Densitometric analysis in in vivo evaluation of synthetic salmon calcitonin activity. J Clin Pharmacol Ther 1988; 13:125. Peichl P, Rintelen B, Kumpan W, Broll H. Increase of axial and appendicular trabecular and cortical bone density in established osteoporosis with intermittent nasal salmon calcitonin therapy. Gynecol Endocrinol 1999; 13:7. Civitelli R, Gonnelli S, Zacchei F, et al. Bone turnover in postmenopausal osteoporosis. Effect of calcitonin treatment. J Clin Invest 1988; 82:1268. Downs RW Jr, Bell NH, Ettinger MP, et al. Comparison of alendronate and intranasal calcitonin for treatment of osteoporosis in postmenopausal women. J Clin Endocrinol Metab 2000; 85:1783. Gennari C, Agnusdei D, Camporeale A. Long-term treatment with calcitonin in osteoporosis. Horm Metab Res 1993; 25:484. Chesnut C, Baylink D, Doyle D, et al. Salmon calcitonin nasal spray prevents vertebral fractures in established osteoporosis. Osteoporosis Int 1998; 8:13. Hizmetli S, Elden H, Kaptanoglu E, et al. The effect of different doses of calcitonin on bone mineral density and fracture risk in postmenopausal osteoporosis. Int J Clin Pract 1998; 52:453. Kanis JA, McCloskey EV. Effect of calcitonin on vertebral and other fractures. QJM 1999; 92:143. Reginster JY, Denis D, Deroisy R, et al. Long-term (3 years) prevention of trabecular postmenopausal bone loss with low-dose intermittent nasal salmon calcitonin. J Bone Miner Res 1994; 9:69. Healey JH, Paget SA, Williams-Russo P, et al. A randomized controlled trial of salmon calcitonin to prevent bone loss in corticosteroid-treated temporal arteritis and polymyalgia rheumatica. Calcif Tissue Int 1996; 58:73. Roth A, Kolaric K. Analgesic activity of calcitonin in patients with painful osteolytic metastases of breast cancer. Results of a controlled randomized study. Oncology 1986; 43:283. Blower PJ, Puncher MRB, Kettle AG, et al. Iodine-123 salmon calcitonin, an imaging agent for calcitonin receptors: synthesis, biodistribution, metabolism and dosimetry in humans. Eur J Nucl Med 1998; 25:101. Snider RH Jr, Nyln ES, Becker KL. Procalcitonin and its component peptides in systemic inflammation: immunochemical characterization. J Investig Med 1997; 45:552. Born W, Beglinger C, Fischer JA. Diagnostic relevance of the amino-terminal cleavage peptide of procalcitonin (PAS-57), calcitonin and calcitonin gene-related peptide in medullary thyroid carcinoma patients. Regul Pept 1991; 32:311. Cohen R, Giscard-Darteville S, Bracq S, et al. Calcitonin genes (I and II) expression in human nervous and medullary thyroid carcinoma tissues.Neuropeptides 1994; 26:215. Burgess TL, Kelly RB. Constitutive and regulated secretion of proteins.Annu Rev Cell Biol 1987; 3:243. Snider RH, Silva OL, Moore CF, Becker KL. Immunochemical heterogeneity of calcitonin in man: effect on radioimmunoassay. Clin Chim Acta 1977; 76:1. Engelbach M, Gorges R, Forst T, et al. Improved diagnostic methods in the follow-up of medullary thyroid carcinoma by highly specific calcitonin measurements. J Clin Endocrinol Metab 2000; 85:1890. Pacini F, Fontanelli M, Fugazzola L, et al. Routine measurement of serum calcitonin in nodular thyroid diseases allows the preoperative diagnosis of unsuspected sporadic medullary thyroid carcinoma. J Clin Endocrinol Metab 1994; 78:826. Cohen R, Campos JM, Salaun C, et al. Preoperative calcitonin levels are predictive of tumor size and postoperative calcitonin normalization in medullary thyroid carcinoma. Groupe d'Etudes des Tumeurs calcitonine (GETC). J Clin Endocrinol Metab 2000; 85:919. Silva OL, Snider RH Jr, Moore CF, Becker KL. Urine calcitonin as a test for medullary thyroid cancer: a new screening procedure. Ann Surg 1979; 189:269. Becker KL, Nash D, Silva OL, et al. Increased serum and urinary calcitonin levels in patients with pulmonary disease. Chest 1981; 79:211. Linnoila RI, Becker KL, Snider RH, Moore CF. Calcitonin as a marker for diethylnitrosamine induced pulmonary endocrine cell hyperplasia in hamsters. Lab Invest 1984; 51:39. Becker KL. The coming of age of a bronchial epithelial cell. Am Rev Respir Dis 1993; 148:1166. Becker KL, Gazdar AF. What can the biology of small cell cancer teach us about the endocrine lung? Biochem Pharmacol 1985; 34:155. Kelley MJ, Becker KL, Rushin JM, et al. Calcitonin elevation in small cell lung cancer without ectopic production. Am J Respir Crit Care Med 1994; 149:183. Silva OL, Broder LE, Doppman JL, et al. Calcitonin as a marker for bron-chogenic cancer: a prospective study. Cancer 1979; 44:680. Mallet E, Lanse X, Devaux AM, et al. Hypercalcitoninaemia in fulminant meningococcaemia in children. Lancet 1983; 1:294. Chesney RW, McCarron DM, Haddad JG, et al. Pathogenic mechanisms of the hypocalcemia of the staphylococcal toxic-shock syndrome. J Lab Clin Med 1983; 101:576. Assicot M, Gendrel D, Carsin H, et al. High serum calcitonin concentrations in patients with sepsis and infection. Lancet 1993; 34:515.

64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. 103. 104. 105. 106. 107. 108. 109. 110. 111. 112. 113. 114. 115. 116. 117.

Whang K, Steinwald PM, White JC, et al. Serum calcitonin precursors in sepsis and systemic infection. J Clin Endocrinol Metab 1998; 83:3296. Mller B, Becker KL, Schchinger H, et al. Calcitonin precursors are reliable markers of sepsis in a medical intensive care unit. Crit Care Med 2000; 28:977. Whang KT, Vath SD, Becker KL, et al. Procalcitonin and pro-inflammatory cytokine interactions in sepsis. Shock 2000; 14:73. Nyln ES, Whang KT, Snider RH Jr, et al. Mortality is increased by procalcitonin and decreased by an antiserum reactive to procalcitonin in experimental sepsis. Crit Care Med 1998; 26:1001. Wagner KE, Vath SD, Snider RH, et al. Immunoneutralization of elevated calcitonin precursors markedly attenuates the harmful physiologic response to sepsis. Paper presented at: 40th Interscience Conference on Antimicrobial Agents and Chemotherapy; September 1720, 2000; Toronto, Canada. Mller B, White JC, Nyln ES, et al. Ubiquitous expression of the calcitonin-1 gene in multiple tissues in response to sepsis. J Clin Endocrinol Metab 2001; 86:396. Mller B, Becker KL, Snider RH, et al. Procalcitonin induces the synthesis of inflammatory cytokines by human peripheral blood mononuclear cells.Paper presented at: 39th Interscience Conference on Antimicrobial Agents and Chemotherapy; September 2629, 1999; San Diego, CA. Whang KT, Snider RH, White JC, et al. Impact of precalcitonin peptides on calcium in experimental sepsis. Paper presented at: Endocrine Society; June 24, 1998; New Orleans, LA. Amara SG, Jonas V, Rosenfeld MG, et al. Alternative RNA processing in calcitonin gene expression generates mRNAs encoding different polypeptide products. Nature 1982; 298:240. Steenbergh PH, Hoppener JW, Zandberg J, et al. A second human calcitonin/CGRP gene. FEBS Lett 1985; 183:403. Hoppener JW, Steenbergh PH, Zandberg J, et al. A third human CALC (pseudo)gene on chromosome 11. FEBS Lett 1988; 233:57. Williams ED, Ponder BJ, Craig RK. Immunohistochemical study of calcitonin gene-related peptide in human medullary carcinoma and C cell hyperplasia. Clin Endocrinol (Oxf) 1987; 27:107. Petermann JB, Born W, Chang JY, Fischer JA. Identification in the human central nervous system, pituitary, and thyroid of a novel calcitonin gene-related peptide, and partial amino acid sequence in the spinal cord. J Biol Chem 1987; 262:542. Kelley MJ, Snider RH, Becker KL, Johson BE. Small cell lung carcinoma cell lines express mRNA for calcitonin and a- and b-calcitonin gene related peptides. Cancer Lett 1994; 81:19. Zaidi M, Chambers TJ, Gaines DR, et al. A direct action of human calcitonin gene-related peptide on isolated osteoclasts. J Endocrinol 1987; 115:511. Donnerer J, Schuligoi R, Stein C, Amann R. Upregulation, release and axonal transport of substance P and calcitonin gene-related peptide in adjuvant inflammation and regulatory function of nerve growth factor.Regul Pept 1993; 46:150. Fernandez S, Knopf MA, McGillis JP. Calcitonin-gene related peptide (CGRP) inhibits interleukin-7-induced pre-B cell colony formation. J Leukoc Biol 2000; 67:669. Palmer JB, Cuss FM, Mulderry PK, et al. Calcitonin gene-related peptide is localized to human airway nerves and potently constricts human airway smooth muscle. Br J Pharmacol 1987; 91:95. Molina JM, Cooper GJ, Leighton B, Olefsky JM. Induction of insulin resistance in vivo by amylin and calcitonin gene-related peptide. Diabetes 1990; 39:260. Fahim A, Rettori V, McCann SM. The role of calcitonin gene-related peptide in the control of growth hormone and prolactin release. Neuroendocrinology 1990; 51:688. Ashina M, Bendsten L, Jensen R, et al. Evidence for increased plasma levels of calcitonin gene-related peptide in migraine outside of attacks. Pain 2000; 86:133. Benvenga S, Trimarchi F, Facchiano A. Homology of calcitonin with the amyloid-related proteins. J Endocrinol Invest 1994; 17:119. Kleinman RM, Fagan SP, Ray MK, et al. Differential inhibition of insulin and islet amyloid polypeptide secretion by intraislet somatostatin in the isolated perfused human pancreas. Pancreas 1999; 19:346. Samsom M, Szarka LA, Camilleri M, et al. Pramlintide, an amylin analog selectively delays gastric emptying: potential role of vagal inhibition. Am J Physiol Gastrointest Liver Physiol 2000; 278:G946. Tenidis K, Waldner M, Bernhagen J, et al. Identification of a penta- and hexapeptide of islet amyloid polypeptide (IAPP) with amyloidogenic and cytotoxic properties. J Mol Biol 2000; 295:1055. Yates SL, Burgess LH, Kocsis-Angle J, et al. Amyloid beta and amylin fibrils induce increases in proinflammatory cytokine and chemokine production by THP-1 cells and murine microglia. J Neurochem 2000; 74:1017. Kitamura K, Kangawa K, Kawamoto M, et al. Adrenomedullin: a novel hypotensive peptide isolated from human pheochromocytoma. Biochem Biophys Res Commun 1993; 192:553. Hinson JP, Kapas S, Smith DM. Adrenomedullin, a multifunctional regulatory peptide. Endocr Rev 2000; 21:138. Ishihara T, Kato J, Kitamura K, et al. Production of adrenomedullin in human vascular endothelial cells. Life Sci 1997; 60:1763. Kubo A, Minamino N, Isumi Y, et al. Adrenomedullin production is correlated with differentiation in human leukemia cell lines and peripheral blood monocytes. FEBS Lett 1998; 426:233. Sato K, Hirata Y, Imai T, et al. Characterization of immunoreactive adrenomedullin in human plasma and urine. Life Sci 1995; 57:189. Kato J, Kitamura K, Kangawa K, Eto T. Receptors for adrenomedullin in human vascular endothelial cells. Eur J Pharmacol 1995; 289:383. Kanazawa H, Kurihara N, Hirata K, et al. Adrenomedullin, a newly discovered hypotensive peptide, is a potent bronchodilator. Biochem Biophys Res Commun 1994; 205:251. Petrie MC, Hillier C, Morton JJ, McMurray JJ. Adrenomedullin selectively inhibits angiotensin II-induced aldosterone secretion in humans. J Hypertens 2000; 18:61. Kamoi H, Kanazawa H, Hirata K, et al. Adrenomedullin inhibits the secretion of cytokine-induced neutrophil chemoattractant, a member of the interleukin-8 family, from rat alveolar macrophages. Biochem Biophys Res Commun 1995; 211:1031. Yamaguchi T, Baba K, Doi Y, et al. Inhibition of aldosterone production by adrenomedullin, a hypotensive peptide, in the rat. Hypertension 1996; 28:308. Sugo S, Minamino N, Shoji H, et al. Interleukin-1, tumor necrosis factor and lipopolysaccharide additively stimulate production of adrenomedullin in vascular smooth muscle cells. Biochem Biophys Res Commun 1995; 207:25. Nishio K, Akai Y, Murao Y, et al. Increased plasma concentrations of adrenomedullin correlate with relaxation of vascular tone in patients with septic shock. Crit Care Med 1997; 25:953. Moore EE, Kuestner RE, Stroop SD, et al. Functionally different isoforms of the human calcitonin receptor result from alternative splicing of the gene transcript. Mol Endocrinol 1995; 9:959. Goldring SR. The structure and molecular biology of calcitonin receptor.In: Bilezikian JP, Raisz LG, Rodan GA, eds. Principles of bone biology. New York: Academic Press, 1996:461. Marx SJ, Woodward CJ, Aurbach GD. Calcitonin receptors of kidney and bone. Science 1972; 178:999. Body JJ, Fernandez G, Lacroix M, et al. Regulation of lymphocyte calcitonin receptors by interleukin-1 and interleukin-2. Calcif Tissue Int 1994; 55:109. Barsony J, Marx SJ. Dual effects of calcitonin and calcitonin gene-related peptide on intracellular cyclic 3',5'-monophosphate in a human breast cancer cell line. Endocrinology 1988; 122:1218. Teti A, Paniccia R, Goldring SR. Calcitonin increases cytosolic free calcium concentration via capacitative calcium influx. J Biol Chem 1995; 270:16666. Chakraborty M, Chatterjee D, Kellokumpu S, et al. Cell cycle-dependent coupling of the calcitonin receptor to different G proteins. Science 1991; 251:1078. Cohen DP, Thaw CN, Varma A, et al. Human calcitonin receptors exhibit agonist-independent (constitutive) signaling activity.Endocrinology 1997; 138:1400. Findlay DM, Martin TJ. Receptors of calciotropic hormones. Horm Metab Res 1997; 29:128. Aiyar N, Rand K, Elshourbagy NA, et al. A cDNA encoding the calcitonin gene-related peptide type 1 receptor. J Biol Chem 1996; 271:11325. McLatchie LM, Fraser NJ, Main MJ, et al. RAMPs regulate the transport and ligand specificity of the calcitonin-receptor-like receptor. Nature 1998; 393:333. Frayon S, Cueille C, Gnidehou S, et al. Dexamethasone increases RAMP1 and CRLR mRNA expressions in human vascular smooth muscle cells.Biochem Biophys Res Commun 2000; 270:1063. Muff R, Bhlmann N, Fischer JA, Born W. An amylin receptor is revealed following co-transfection of a calcitonin receptor with receptor activity modifying proteins-1 or -3. Endocrinology 1999; 140:2924. Christopoulos G, Perrk J, Morfis M, et al. Multiple amylin receptors arise from receptor activity-modifying protein interaction with the calcitonin receptor gene product. Mol Pharmacol 1999; 56:235. Sumpe ET, Tilakaratne N, Fraser NJ, et al. Multiple ramp domains are required for generation of amylin receptor phenotype from the calcitonin receptor gene product. Biochem Biophys Res Commun 2000; 267:368. Martnez A, Kapas S, Miller M-J, et al. Coexpression of receptors for adrenomedullin, calcitonin gene-related peptide, and amylin in pancreatic b cells. Endocrinology 2000; 141:406.

CHAPTER 54 VITAMIN D Principles and Practice of Endocrinology and Metabolism

CHAPTER 54 VITAMIN D
THOMAS L. CLEMENS AND JEFFREY L. H. O'RIORDAN Chemistry Production and Metabolism of Vitamin D Photoproduction Dietary Sources Metabolic Activation of Vitamin D 25-Hydroxylation in the Liver 1 a-Hydroxylation in the Kidney 24-Hydroxylation Other Pathways of Vitamin D Metabolism Regulation of Vitamin D Metabolism Parathyroid Hormone Phosphorus Other Regulators Control of 24-Hydroxylation Maternal-Fetal Metabolism Absorption, Transport, and Excretion Vitamin DBinding Protein Excretion Biologic Actions of Vitamin D Intestine Skeleton Actions on Other Tissues Nonhypercalcemic Vitamin D Analogs Biochemical Mechanism of Action Measurement of Vitamin D Metabolites Methods Assessment of Vitamin D Status Chapter References

CHEMISTRY
Vitamin D and its analogs form a group of fat-soluble secosterols with antirachitic properties.1 The two parent forms of vitamin D are ergocalciferol (vitamin D2) and cholecalciferol (vitamin D3). Ergocalciferol derives its name from its immediate precursor, ergosterol, the plant sterol from which it was originally prepared. Cholecalciferol is the natural form of the vitamin and is produced by irradiation of the precursor, 7-dehydrocholesterol. The D vitamins are structurally related to C-21 steroids but differ by having an opened B ring that forms a conjugated triene structure (Fig. 54-1). Vitamin D2 differs structurally from vitamin D3 by having a double bond between the carbon positions 22 and 23 and a methyl group at C-24. Both vitamin D2 and vitamin D3 are metabolized along the same pathways, producing active metabolites with equivalent biologic effects (Fig. 54-2). When written without the subscript, the term vitamin D may refer to either compound.

FIGURE 54-1. The structure of vitamin D3 : A 910 secosterol (right), and a C-21 steroid (left).

FIGURE 54-2. The main metabolic pathways of vitamin D3 production and activation. Vitamin D2 from the diet is metabolized along the same pathways, yielding the corresponding vitamin D2 metabolites (not shown).

PRODUCTION AND METABOLISM OF VITAMIN D

PHOTOPRODUCTION Technically, vitamin D3 is not a true vitamin because it can be produced in the body. Its synthesis in skin can provide the body's entire requirement unless exposure to sunlight is restricted (see Chap. 185). The reaction proceeds by nonenzymatic photolysis of 7-dehydrocholesterol (provitamin D) in the epidermis.2 Near-ultraviolet (UV) light (wavelengths 290315 nm) penetrates to the basal layer and cleaves the bond between carbon atoms 9 and 10 of 7-dehydrocholesterol, forming pre-vitamin D3 (see Fig. 54-2). Previtamin D3 and vitamin D3 are in thermal equilibrium; production of the latter is favored at body temperature. The time-dependent thermal isomerization of pre-vitamin D3 to vitamin D3 allows the continuous release of vitamin D3 into the circulation for several days after exposure to sunlight. Previtamin D3 is subject to further photolysis, yielding the biologically inert photoproducts, lumisterol and tachy-sterol. Studies measuring circulating vitamin D3 in subjects who have been given graded amounts of UV light have estimated that exposure to one minimal erythema dose results in the formation of at least 30 g of vitamin D3 per square meter of body surface. The skin pigment melanin, when present in large amounts, competes with 7-dehydrocholesterol for absorption of UV energy. Thus, heavily pigmented black persons make less vitamin D3 in response to UV exposure than do lightly pigmented white persons.

DIETARY SOURCES Few foods naturally contain appreciable amounts of vitamin D. The livers of fatty fish are a relatively rich source,2a which accounts for the efficacy of cod-liver oil as a cure for rickets. Insignificant amounts of vitamin D exist naturally in dairy products. The limited dietary availability of vitamin D led to its use as a supplement in certain foods. In the United States, federal regulation stipulates that 10 g (400 IU) of vitamin D be added to every quart of milk. (In the United Kingdom, milk is not fortified with vitamin D, which may partly explain the lower vitamin D stores of its residents.) Multivitamin preparations typically contain 10 g, which is twice the recommended dietary allowance of vitamin D as noted by the National Academy of Sciences. It has been estimated that the true vitamin D requirement in the absence of sunlight exposure could be as much as 600 IU or 15 g per day.3 Previously, the exclusive use of vitamin D2 as a food additive made it the source of the circulating metabolites. Today, however, either vitamin D2 or vitamin D3 is added to foods, so that vitamin D3 metabolites in the circulation may have originated from endogenous or from dietary sources. METABOLIC ACTIVATION OF VITAMIN D Vitamin D has little, if any, intrinsic bioactivity. Sequential hydroxylation reactions, which occur first in the liver and then in the kidney, are required for the production of the biologically active form, 1,25-dihydroxyvitamin D (1,25[OH]2D).1 The addition of the two hydroxyl groups enables 1,25(OH)2D to bind to intracellular receptors in target tissues with high affinity. The metabolism and mode of action of vitamin D, therefore, are similar to those of other steroid hormones. 25-HYDROXYLATION IN THE LIVER The first step in the activation of vitamin D is 25-hydroxylation, which occurs in the liver and is obligatory for further metabolism of the vitamin. Although other tissues can metabolize [3 H]vitamin D3 to [3H]25(OH)D3 in vitro, the liver is the principal site of production; total hepatectomy causes the virtual disappearance of 25-hydroxyvitamin D3 (25[OH]D3) from the circulation. Hepatic 25-hydroxylase enzyme activity is associated with both mitochondrial and microsomal fractions. The mitochondrial enzyme, a cytochrome P-450 protein designated CYP27,4,5 has been cloned and also has been shown to hydroxylate cholesterol and sterols involved in bile acid synthesis. Normally, 25(OH)D circulates at 5 to 65 ng/mL and has a circulating half-life of ~15 days. It is not considered bioactive at physiologic concentrations in vivo, but appears to be active at high concentrations when assessed in vitro. Its order of potency in calcium transport bioassay systems is similar to its 1,25(OH)2D intestinal receptor binding activity, suggesting that, at high serum concentrations in vivo, it has agonist activity.6 Accordingly, the hypercalcemia produced by vitamin D intoxication is almost certainly mediated by 25(OH)D. 1 a-HYDROXYLATION IN THE KIDNEY The final step in the activation of vitamin D occurs in the kidney, in which 25(OH)D is converted to 1,25(OH)2D, the biologically active and most potent metabolite.6a The major site of 1 a-hydroxylation is the renal proximal tubule.7 Although other tissues and cultured cells produce 1,25(OH)2D3 in vitro, 1,25-(OH)2D3 is absent from the serum of anephric animals and humans, suggesting that extrarenal sources do not contribute significantly to circulating 1,25(OH)2D concentrations under normal physiologic conditions. However, in pregnancy, the placenta synthesizes 1,25(OH)2D and supplies additional hormone to the circulation.8 Extrarenal synthesis of the metabolite also occurs in patients with sarcoidosis, in whom hypercalcemia and elevated 1,25(OH)2D levels have been demonstrated, even in the anephric state9 (Fig. 54-3). The synthesis of [3H]1,25(OH)2D3 by pulmonary alveolar macrophages from patients with sarcoidosis has been demonstrated in vitro, suggesting that these cells are the source of the ectopic production.10 It is likely that the hypercalcemia associated with other granulomatous diseases, such as tuberculosis, and with certain lymphomas also is attributable to extrarenal production of 1,25(OH)2D (see Chap. 59).

FIGURE 54-3. Extrarenal production of 1,25(OH)2D in an anephric patient with sarcoidosis. The patient had progressive hypercalcemia in the face of renal failure that was associated with elevated serum 1,25(OH)2D levels and suppressed immunoreactive parathyroid hormone (iPTH) concentrations. Bilateral nephrectomy and reduction of the prednisone dosage resulted in marked elevations in serum 1,25(OH)2D levels and worsening of the hypercalcemia, implicating an extrarenal source of the 1,25(OH)2D production. (From Barbour GL, Coburn JW, Slatopolsky E, et al. Hypercalcemia in a anephric patient with sarcoidosis: evidence for extrarenal generation of 1,25-dihydroxyvitamin D. N Engl J Med 1981; 305:440.)

The enzyme responsible for the conversion of 25(OH)D to 1,25(OH)2D, vitamin D-25 1a-hydroxylase (CYP1), has been cloned11 and is a mixed-function cytochrome P450 protein.11a The enzyme requires NADPH generated in the inner mitochondrial membrane of the renal proximal convoluted tubule cell by an energy-dependent trans-hydrogenase. Thus, inhibitors of oxidative phosphorylation or electron transport will inhibit 1 a-hydroxylation in vitro.12 Along with 25(OH)D, 24,25(OH)2D and 25,26(OH)2D also can serve as substrates to yield 1,24,25(OH)3D and 1,25,26(OH)3D, respectively. Under normal physiologic conditions, the circulating concentration of 1,25(OH)2D is 500 to 1000 times lower than that of 25(OH)D. In humans, 1 g of 1,25(OH)2D is produced each day; its circulating half-life is ~4 to 6 hours.13 24-HYDROXYLATION 25-Hydroxyvitamin D also is subject to hydroxylation at C-24, producing 24,25(OH)2D, which is the second most abundant circulating metabolite of vitamin D. Most circulating 24,25(OH)2D is made in the kidney. Other tissues, most notably cartilage and intestine, also can produce 24,25(OH)2D3 in vitro. The renal 24- hydroxylase (CYP24) also has been cloned,14 and like the vitamin D-25 1-hydroxylase is a mixed function P450-dependent enzyme. It is expressed in mitochondria of renal tubular cells, but is distinct from the vitamin D-25 1 a-hydroxylase and demonstrates a broader tissue distribution.14,15 The main substrate is 25(OH)D, but under certain conditions, the enzyme also will hydroxylate 1,25(OH)2D to form 1,24,25(OH)3D. Normally, 24,25(OH)2D circulates at concentrations that are ten-fold lower than those of 25(OH)D. Compared with 1,25(OH)2D3, 24,25(OH)2D3 is far less active when bioassayed in the systems that are responsive to vitamin D. No clear role can be ascribed to 24,25(OH)2D; however, it may simply be a product of the further metabolism and inactivation of vitamin D3, because 24-hydroxylation renders the molecule susceptible to side-chain cleavage and oxidation. OTHER PATHWAYS OF VITAMIN D METABOLISM Other circulating forms of vitamin D have been isolated and identified.16 25,26-Dihydroxyvitamin D is produced in the kidney and probably in the liver. Its serum concentration usually is slightly lower than that of 24,25(OH)2D. The 25(OH)D3 26,23-lactone has been identified in the plasma of animals and humans. Studies of the metabolism of 1,25(OH)2D have identified a major oxidative cleavage product as the C-23 carboxylic acid. Other side-chain metabolites of vitamin D have been isolated from animals given large doses of the vitamin or from tissues incubated with 25(OH)D3 in vitro. The physiologic significance of these metabolites, aside from being intermediates in the vitamin D catabolic pathway, is unknown.

REGULATION OF VITAMIN D METABOLISM

The metabolic activation of vitamin D is subject to both coarse-and fine-control mechanisms operating at the principal enzymatic steps.17 Hepatic 25-hydroxylation is more efficient at low levels of substrate, and the absolute amount of 25(OH)D in the circulation declines when the supply of vitamin D increases. Nevertheless, the degree of regulation of 25(OH)D synthesis is not sufficient to prevent vitamin D intoxication after the ingestion of large amounts of vitamin D.18 In contrast to the relatively coarse control of hepatic 25- hydroxylation, renal 1 a-hydroxylation is regulated strictly and represents the most important regulatory mechanism in the metabolism of vitamin D. In normal adults, serum concentrations of 1,25(OH)2D change little in response to repeated dosing with vitamin D, and remain normal or even decline in vitamin D intoxication. The three major factors that regulate the enzyme's activity are blood parathyroid hormone (PTH), calcium, and phosphorus. PARATHYROID HORMONE Parathyroid hormone is the main hormonal stimulus for 1 a-hydroxylation, increasing systemic 1,25(OH)2D levels after administration to human research subjects.19 Hypocalcemia also stimulates 1 a-hydroxylation, but this effect is achieved largely through a feedback loop (see Chap. 58) involving the parathyroid glands17; hypocalcemia stimulates PTH secretion, which raises serum 1,25(OH)2D concentrations. 1,25(OH)2D enhances calcium absorption, and the increased serum calcium concentration then inhibits PTH secretion (Fig. 54-4). Some experimental studies suggest that the extracellular calcium concentration can directly influence the renal 1 a-hydroxylase.

FIGURE 54-4. Effect of a decline in plasma calcium on circulating parathyroid hormone and 1,25(OH)2D concentrations in humans. Eight patients with Paget disease received plicamycin (25 g/kg) by infusion. This resulted in a sudden reduction of plasma calcium levels followed by an increase in parathyroid hormone and 1,25(OH)2D3 concentrations. (AMP, adenosine monophosphate.) (Modified from Bilezikian JP, Canfield RE, Jacobs JP. Response of 1 a-dihydroxyvitamin D3 to hypocalcemia in human subjects. N Engl J Med 1978; 299:437.)

PHOSPHORUS The blood phosphorus concentration is the third major control factor. In rats and in humans, hypophosphatemia stimulates 1 a-hydroxylase activity and increases the 1,25(OH)2D concentration in serum, whereas hyperphosphatemia inhibits formation of the metabolite (see Fig. 54-4). This adaptive effect does not depend on PTH, but it may require insulin-like growth factor-I.20 OTHER REGULATORS Other factors may directly or indirectly affect the activity of the renal 1 a-hydroxylase enzyme. 1,25(OH)2D3 inhibits its own production in cultured kidney cells.21 In rats, calcitonin stimulates the production of 1,25(OH)2D, apparently through a mechanism different from PTH stimulation.22 Prolactin, insulin, and growth hormone all stimulate 1 a-hydroxylase in mammals. In birds, enhanced estrogen and progesterone secretion during ovulation stimulates 1 a-hydroxylation and increases serum 1,25(OH)2D concentrations. Therefore, these hormones appear to stimulate 1,25(OH)2D3 production to accommodate the increased mineral ion requirement during growth and reproduction. CONTROL OF 24-HYDROXYLATION When blood calcium, phosphorus, and PTH levels are normal, 25(OH)D is converted mainly to 24,25(OH)2D, whereas in the vitamin Ddeficient state, the production of 1,25(OH)2D3 is induced and the synthesis of 24,25(OH)2D3 is suppressed. This reciprocal relationship between the production of these metabolites has been demonstrated elegantly in cultured kidney cells.21 The 24-hydroxylase enzyme activity and expression of its mRNA are acutely induced by 1,25(OH)2D3 in kidney and in other vitamin Dresponsive cells.23 As mentioned earlier, this appears to represent a mechanism by which activity of 1,25(OH)2D3 is limited. MATERNAL-FETAL METABOLISM During human fetal development, ~30 g of calcium is transferred from mother to fetus. Adaptations in maternal vitamin D metabolism during pregnancy are important in supplying this extra calcium to the fetus.24 The increase in intestinal calcium absorption that persists throughout pregnancy is sufficient to sustain a positive calcium balance.25 The circulating concentrations of 1,25(OH)2D and its transporting protein (vitamin D binding protein) increase in parallel during the first two trimesters. During the third trimester, however, total serum 1,25(OH)2D levels continue to increase, whereas vitamin D binding protein concentrations fall slightly.26 This finding suggests that elevation in the biologically important, free 1,25(OH)2D concentration stimulates maternal intestinal calcium absorption. The placental 1 a-hydroxylase enzyme supplies additional 1,25(OH)2D to the mother and child during pregnancy. The assayed concentrations of 1,25(OH)2D in maternal and cord blood are positively correlated, suggesting placental transfer of the metabolite, or stimulation of both fetal and maternal 1 a-hydroxylase by common hormonal factors. Conversely, maternal and fetal 24,25(OH)2D concentrations decrease during this period. Enhanced hepatic synthesis of vitamin Dbinding protein, which would cause a transient decrease in free 1,25(OH)2D levels, may be the initial stimulus for the subsequent increase in 1,25(OH)2D concentrations. PTH and prolactin also probably play a role in the adaptation of vitamin D metabolism during pregnancy. ABSORPTION, TRANSPORT, AND EXCRETION Dietary vitamin D is absorbed with fats in the small intestine and is incorporated into chylomicrons. The efficiency of absorption normally is 60% to 90%, and interruptions of this process, such as in steatorrhea, can lead to vitamin D malabsorption (see Chap. 63). In the vitamin Dreplete state, as much as 50% of the absorbed vitamin may be stored in body fats. Thus, high tissue concentrations of 25(OH)D may persist in patients long after treatment with pharmacologic doses of vitamin D has been discontinued. VITAMIN DBINDING PROTEIN Vitamin D entering the circulation from either dietary or epidermal sources is bound by vitamin Dbinding protein (DBP), an a-globulin of 55,000 daltons that is identical with a serum protein originally described as the group-specific component.27 This protein has a single, high-affinity binding site that will bind vitamin D and all its metabolites, although it has higher affinity for 25(OH)D3, 24,25(OH)2D3, and 25,26(OH)2D3 than for vitamin D3 or 1,25(OH)2D3.28 This order of selectivity may facilitate the partitioning of vitamin D in lipid stores and favors entry of biologically active 1,25(OH)2D3 into target cells. Mice null for the DBP gene develop normally, but demonstrate an increased rate of clearance of 25(OH)D compared with controls,29 which is consistent with the notion that DBP functions primarily as a vitamin D

metabolite carrier protein. The serum concentration of DBP greatly exceeds the concentration of all the known metabolites of the vitamin, so that only 3% to 5% of available binding sites are occupied. Hepatic synthesis of the protein is increased during pregnancy and there is appreciable urinary loss in patients with hypoproteinemia. Nevertheless, its serum concentration is largely unaltered in other disorders of mineral or skeletal homeostasis.30 EXCRETION The pathways for excretion of vitamin D and its metabolites are still poorly understood. Bile appears to be a principal route; the biliary excretion of vitamin D metabolites increases when anti-convulsant drugs are administered.31 There is little urinary excretion of vitamin D. Sulfate conjugates of vitamin D3, with no related biologic activity, have been identified in urine and in breast milk.32

BIOLOGIC ACTIONS OF VITAMIN D

INTESTINE CALCIUM TRANSPORT The most important bioeffect of 1,25(OH)2D3 is to increase the intestinal absorption of calcium. Although the details of the mechanisms underlying this process remain sketchy, several key steps have been characterized.33 In the absence of vitamin D, Ca2+ is able to enter the cell, but little is absorbed as a result of its sequestration in the brush border terminal web. When 1,25(OH)2D3 is given to a vitamin Ddeficient animal, there is a rapid increase in Ca2+ entry over ~30 minutes. A second phase of calcium transport peaks several hours after administration of 1,25(OH)2D3, and termination of the entire transcellular Ca2+ absorption process is not complete for several more hours. During this lag period, calbindin (see later) and other vitamin D induced proteins are synthesized and are believed to participate in the delivery of Ca2+ to the basolateral membrane, where the ion is extruded by the action of the Na+/Ca2+ exchanger. The number of functional Na+/Ca2+ exchangers is increased by vitamin-D treatment. The best characterized vitamin Dinduced protein is a cytoplasmic, 28,000-dalton calcium-binding protein (calbindin 28K) that was first identified in chick intestine.34 This protein has four high-affinity calcium-binding sites, and its dependence on vitamin D distinguishes it from calmodulin and other intracellular calcium-binding proteins. A separate 9000-dalton vitamin Ddependent calcium-binding protein (calbindin 9K), with 2 Ca2+ binding sites, is expressed in mammalian intestine. In vitamin Ddeficient chickens given 1,25(OH)2D orally, both the induction of calcium transport and the appearance of calbindin in villus enterocytes are demonstrable within a few hours after dosing. As indicated earlier, calbindin is thought to play a direct role in calcium transport but it may also mediate specific calcium-dependent processes within the enterocyte. The vitamin Ddependent calbindin 28K also is expressed in kidney,35 brain,36 bone, and other tissues.34 Although the calbindins were initially viewed as vitamin Ddependent proteins, it is now clear that they are expressed in a wide array of tissues in a vitamin Dindependent fashion. It has been suggested that the cal-bindins might serve as a buffer to guard against large fluctuations in intracellular calcium. PHOSPHATE TRANSPORT The impact of vitamin D on phosphate transport was described in the 1960s by Harrison and Harrison using an everted gut sac method. Phosphate, like calcium, is transferred across the intestinal epithelium via a multi-step process.37 An energy-dependent step is required for entry of the negatively charged phosphate ion into the cell, whereas the extrusion step occurs primarily by diffusional processes. Studies in vitamin Ddeficient chicks administered32 P-phosphate suggest that vitamin D exerts effects on both the phosphate entry and exit mechanisms. The phosphate-entry step is Na+-dependent, which is reminiscent of the absorption of a number of nonelectrolytes, such as glucose and amino acids. Several isoforms of the renal Na+/phosphate co-transporters have been identified in intestinal mucosa. In addition, vitamin D induces the synthesis of a Na+/phosphate co-transporter on the basolateral membrane. A Na+ /phosphate co-transporter is expressed on the basolateral membrane; however, it is unclear whether vitamin D affects the synthesis of this protein. EFFECT ON MEMBRANE LIPID 1,25-Dihydroxyvitamin D affects the lipid composition of intestinal brush border epithelial cells. 1,25(OH)2D-induced synthesis of phosphatidylcholine occurs over a time course similar to the 1,25(OH)2D-stimulated uptake of calcium into brush border membrane vesicles.38 These observations gave rise to the concept that 1,25(OH)2D3 may exert some of its actions through nongenomic mechanisms. Consistent with this notion are studies that show rapid influxes in intracellular calcium in several cell types in vitro and acute increases in intestinal calcium uptake in ex vivo preparations of chicken intestine.39 Presumably, these activities would require 1,25(OH)2D3 binding to a membrane receptor. Such a receptor has been proposed but has not yet been identified. SKELETON The most familiar property of the D vitamins is their ability to cure rickets. However, normalization of blood calcium and phosphorus levels by intravenous infusion of calcium into vitamin Ddeficient rats can cure rickets in the absence of circulating 1,25(OH)2D.40 Hence, the main antirachitogenic effects of 1,25(OH)2D3 rely more on its mobilization of calcium and phosphorus at the level of the intestine than on any direct effect it has on bone formation. Nonetheless, treatment of osteoblast cells with 1,25(OH)2D in vitro is associated with marked changes in expression of osteoblast-specific genes.41 1,25(OH)2D3 decreases the rate of synthesis of collagen and citrate decar-boxylase and increases alkaline phosphatase activity and the synthesis of the noncollagenous proteins, osteocalcin, and osteopontin. The precise role of 1,25(OH)2D3 in osteoblast function remains controversial, in part because of the differences in the experimental osteoblast cell models. By contrast, the ability of 1,25(OH)2D to cause bone mineral resorption is uncontested. Studies in vitro, using neonatal calvaria or long bones pre-labeled with45 Ca, have readily demonstrated increases in cell-mediated resorption.42 1,25(OH)2D3 increases the activity of existing osteoclasts and the rate of recruitment from mononu-clear cell precursors.43 Each of these activities appears to occur indirectly, after stimulation of an osteoblast factor called osteoprotegerin, which is a soluble member of the tumor necrosis factor (TNF) receptor family (see Chap. 61). ACTIONS ON OTHER TISSUES In addition to the classic actions of vitamin D on mineral ion homeostasis, many other biologic processes are affected by 1,25(OH)2D3. As mentioned earlier, 1,25(OH)2D3 induces the 24- hydroxylase enzyme activity in kidney and other tissues. An inhibitory effect of 1,25(OH)2D3 on PTH secretion has been demonstrated, thus documenting a feedback loop in the control of 1,25(OH)2D3 synthesis.44 This observation has led to the use of intravenous 1,25(OH)2D3 in the treatment of secondary hyperparathyroidism of chronic renal failure.45 Vitamin Ddeficient animals have reduced insulin secretion and impaired glucose tolerance that can be corrected by the administration of 1,25(OH)2D3.46 An increasingly appreciated action of 1,25(OH)2D3 is its ability to alter the developmental program of a diverse array of cell types. Exposure to near-physiologic concentrations of 1,25(OH)2D3 induces maturation of basal epidermal skin cells into keratinocytes and stimulates mouse myeloid leukemic cells to differentiate into macrophage-like cells.47,48 and 49 As previously mentioned, 1,25(OH)2D3 stimulates the formation of multinucleated osteoclast-like cells in bone marrow cultures50 (Fig. 54-5). These activities are accompanied by alterations in biochemical markers and morphologic changes consistent with the differentiated phenotype. Such pleiotropic actions of 1,25(OH)2D3, together with the multiple-tissue distribution of the vitamin D receptor, suggest that the hormone functions during development and organogenesis.

FIGURE 54-5. Induction of osteo-clast- like cells by 1,25-(OH)2D3 in vitro. Autoradiographs of [125I] calcitonin binding in mouse bone marrow cultures exposed to 1,25-(OH)2D3 for 8 days (A through C) or 3 days (D). Arrows are pointing to binding of [125I] calcitonin to calcitonin receptors. Panel B shows a high-power micrograph of multinucleate cells staining positive for tartrate-resistant acid phosphatase (TRAP) and demonstrating calcitonin binding (indicated by the dense silver grains), two markers for osteoclasts. (From Takahashi N, Akatsu T, Sakasi T, et al. Induction of calcitonin receptors by 1,25-dihydroxy-vitamin D3 in osteoclast-like multinucleated cells formed in mouse bone marrow cells. Endocrinology 1988; 123:1504.)

NONHYPERCALCEMIC VITAMIN D ANALOGS Synthetic vitamin D analogs have been prepared that retain the ability to bind and activate the vitamin D receptor but are less calcemic than 1,25(OH)2D3 when administered to animals. These so-called nonhypercalcemic analogs (which is not a strictly accurate term because they cause hypercalcemia when administered at large doses) have a chemically modified side chain that reduces affinity for the serum vitamin Dbinding protein and accelerates their metabolic degradation.51 This property apparently accounts in part for their selective action on certain tissues (e.g., the parathyroid gland). It is likely that these analogs eventually will be used therapeutically in the treatment of such conditions as secondary hyperparathyroidism of chronic renal failure and hyperproliferative disorders such as psoriasis. BIOCHEMICAL MECHANISM OF ACTION INTRACELLULAR RECEPTORS Tissues that are responsive to vitamin D express a specific intra-cellular receptor for 1,25(OH)2D.52 The fraction of circulating 1,25(OH)2D that is not bound to serum vitamin Dbinding protein diffuses across the target cell membrane and binds to an intracellular receptor protein. The protein from chick intestine has a molecular mass of 60,000 daltons and appears to be located predominantly within the nucleus, although a separate cytoplasmic component may serve to translocate the 1,25(OH)2D ligand to the nuclear compartment. The rank order of receptor binding for a series of vitamin D metabolites correlates well with their order of biologic activities. Biochemical characterization of the vitamin D receptor (VDR) and its molecular cloning has revealed that it is closely related to the thyroid hormone and retinoic acid receptors, which, like the vitamin D receptor, function as ligand-activated transcription factors. Each receptor exhibits functional domains for ligand binding and a highly conserved zinc finger region that mediates DNA binding. All three receptors bind to response elements consisting of the consensus hexamer AGGTCA, organized into combinations of direct and inverted repeat motifs. The specificity of each receptor for a particular target gene, therefore, would depend on the numbers and spacing of these elements, as well as on the presence and abundance of other coregulatory proteins; the retinoic acid X receptor functions as one such coregulator, but there are likely to be others. This complex interplay between the receptors, their cis-acting response elements, and the presence and abundance of specific transacting nuclear proteins would apparently constitute the means for tissue-specific and developmental actions for each hormone. The vitamin D response element has been identified in several 1,25(OH)2D3-responsive genes, including osteocalcin, PTH, osteopontin, and the 25-hydroxy, 24-hydroxylase genes. A molecular model for 1,25(OH)2D3 transcriptional activation of responsive genes is shown in Figure 54-6.

FIGURE 54-6. The molecular mechanism of action of 1,25(OH)2D3. The hormone dissociates from serum vitamin D binding protein (DBP), enters the cell, and interacts with the vitamin D receptor (VDR). Activation by the ligand leads to interaction of the VDR with responsive gene and modulation of gene expression. VDR-Mu, or VDR modulatory unit, comprises one VDR molecule and an associated protein, which is exemplified by but not restricted to a retinoid X receptor iso-form. (a) An early model wherein the VDR, shown located in the cytoplasm, undergoes cytoplasmic to nuclear translocation on ligand activation and eventually binds to the regulatory region of a modulated gene. (b) Modern view of the location of the VDR, wherein the receptor is located in the nucleus and following ligand activation binds to the regulatory region of a vitamin D modulated gene. (From Pike JW. The vitamin D receptor and its gene. In: Feldman D, Glorieux FH, Pike JW, eds. Vitamin D. San Diego: Academic Press, 1997:105.)

RECEPTOR DEFECTS Knowledge of the various components of the 1,25(OH)2D receptor has been helpful in establishing the pathogenesis of vitamin Ddependent rickets type II (see Chap. 63 and Chap. 70). In this rare, autosomal-recessive disorder, severe hypocalcemia and rickets develop early in childhood and are unresponsive to all forms of vitamin D (Fig. 54-7). Affected persons demonstrate reduced receptor-binding capacity or defective nuclear internalization of hormone in skin fibroblasts.53 Some also exhibit alopecia totalis. The genetic basis for dysfunctional vitamin D receptors has been identified in several affected kindreds by analysis of the chromosomal gene. Point mutations in nucleotides encoding amino acids in both ligand binding and zinc finger DNA binding domains have been demonstrated.54 Mice lacking the vitamin D receptor (VDR) develop severe hypocalcemia, defective bone mineralization, and secondary hyperparathyroidism and alopecia, a phenotype that closely resembles that seen in patients with vitamin Ddependent rickets type II.55

FIGURE 54-7. Two siblings with vitamin Ddependent rickets type II and alopecia. (From Rosen JF, Fleischman AR, Finberg L, et al. Rickets with alopecia: an inborn

error of Vitamin D metabolism. J Pediatrics 1979; 94:729.)

Another consequence of altered vitamin D receptor expression may manifest itself at the level of bone mass. Studies have shown that a common allelic variation in the vitamin D receptor locus is a predictor for bone mass. 56,57 For example, bone mineral density was similar in dizygotic twin pairs with the same VDR genotype and much more variable in twin pairs with different VDR genotypes. Thus, the VDR (or a closely linked gene) appears to contribute to the development of peak bone mass, but the mechanism or mechanisms underlying this association remain obscure. However, conflicting results have been obtained by others58; therefore, the exact significance of the VDR locus to the development of peak bone mass remains to be established.

MEASUREMENT OF VITAMIN D METABOLITES

METHODS The measurement of vitamin D in biologic fluids was pioneered with bioassays that measured either the linear calcification rate (line test) or a rise in blood calcium concentration (calcemic assay) in rats after the administration of vitamin D or its 25(OH)D analog. Modern assays, which are less tedious and more specific, use radioligand-binding techniques, physico-chemical methods, or a combination of these.59 25-HYDROXYVITAMIN D In the early 1970s, the first competitive protein-binding assays for 25(OH)D were developed.60,61 This method uses the vitamin D transport protein as a specific binder. Extraction and chromatographic purification of 25(OH)D generally are required before the binding assay can be conducted. It also is possible to measure 25(OH)D by high-pressure liquid chromatography and UV absorbance spectroscopy. The circulating concentration of 25(OH)D generally reflects the amount of vitamin D transmitted to the liver from dietary or epidermal sources, and therefore serves as the best index of overall vitamin D status. There is a seasonal fluctuation in serum 25(OH)D concentrations, with the highest values occurring in late summer and the lowest seen in late winter.62 The normal range for 25(OH)D varies among laboratories, but generally is accepted to be 5 to 65 ng/mL. Frankly low values are found in patients with nutritional vitamin D deficiency and those with intestinal malabsorption syndromes.63 Conversely, serum concentrations as high as 600 ng/mL are found in persons with vitamin D toxicity. 24,25-DIHYDROXYVITAMIN D AND 25,26-DIHYDROXYVITAMIN D Because 24,25-dihydroxyvitamin D and 25,26-dihydroxyvitamin D also have strong affinity for vitamin Dbinding protein, they can be measured using the competitive protein-binding technique if they first are separated individually by chromatography. The normal range for 24,25-dihydroxyvitamin D is 1 to 4 ng/mL; for 25,26-dihydroxyvitamin D, it is slightly lower. In the absence of renal disease, the serum concentrations of these two dihydroxy metabolites reflect the amount of substrate 25(OH)D. 1,25-DIHYDROXYVITAMIN D The development of an assay for 1,25(OH) 2D has relied mainly on the use of its native receptor protein in radioreceptor assays, or on the technique of radioimmunoassay.64,65 The prior separation of 1,25(OH)2D from other metabolites is achieved by chro-matography. The normal range varies somewhat depending on the method used, but is generally accepted to be from 18 to 65 pg/mL. The physiologic status of the individual must be considered carefully when interpreting the serum level. For example, during growth, pregnancy, or lactation, renal 1,25(OH)2D synthesis may be appropriately increased. Hypocalcemia associated with vitamin D deficiency also is accompanied by augmented 1 a-hydroxylase activity. Therefore, it is common to find elevated serum 1,25(OH)2D concentrations in patients undergoing repletion of vitamin D.

ASSESSMENT OF VITAMIN D STATUS

Because of the wide anatomic distribution of the sites of synthesis, metabolism, and action of vitamin D, clinical deficiency can result from disturbances at many levels (Fig. 54-8). Insufficient intake or production of vitamin D can lead to inadequate substrate concentrations for the conversion to 1,25(OH)2D. Disturbances in the metabolism of vitamin D at the level of the liver or kidney can cause deficient production of 1,25(OH)2D. In renal disease, a decline in the functional renal mass combined with increased phosphate retention eventually reduces or entirely eliminates 1,25(OH)2D production. The loss or failure of parathyroid function in hypoparathyroidism or pseudohypoparathyroidism also is associated with low 1,25(OH)2D levels. Serum concentrations are low in vitamin Ddependent rickets type I because of an inherited defect in the renal 1 a-hydroxylase enzyme. End-organ resistance in patients with vitamin Ddependent rickets type II also can result from inherited defects in the receptor mechanism for 1,25(OH)2D. In this disorder, serum 1,25(OH)2D levels are grossly elevated in response to hypocalcemia. For accurate assessment of vitamin D status, it often is essential to consider both 25(OH)D and 1,25(OH)2D concentrations. For example, a low 1,25(OH)2D level need not indicate defective renal production of the metabolite. Instead, it could result from insufficient vitamin D intake (nutritional vitamin D deficiency), indicated by the finding of a low serum 25(OH)D concentration. If, however, 1,25(OH) 2D levels are low in the face of a normal concentration of 25(OH)D, then a disturbance in the renal synthesis of 1,25(OH)2D3 is likely. Further aspects of clinical disturbances in vitamin D metabolism are discussed in Chapter 61, Chapter 63 and Chapter 70.

FIGURE 54-8. Potential clinical disturbances in intake, metabolism, and action of vitamin D. (G.I., gastrointestinal.)

CHAPTER REFERENCES
1. DeLuca HF. Historical overview. In: Feldman D, Glorieux FH, Pike JW, eds. Vitamin D. San Diego: Academic Press, 1997:3. 2. Holick MF. Vitamin D: new horizons for the 21st century. Am J Clin Nutr 1994; 60:619. 2a. Nakamura K, Nashimoto M, Hori Y, Yamamoto M. Serum 25-hydroxyvitamin B concentrations and related dietary factors in peri- and postmeno-pausal Japanese women. Am J Clin Nutr 2000; 71:1161. Chapuy M-C, Meunier PJ. Vitamin D insufficiency in adults and the elderly. In: Feldman D, Glorieux FH, Pike JW, eds. Vitamin D. San Diego: Academic Press, 1997:679. Usui E, Noshiro M, Okuda K. Molecular cloning of cDNA for vitamin D 3 25- hydroxylase from rat liver mitochondria. FEBS Lett 1990; 262(Mar 12;1):135. Okuda K, Usui E, Ohyama Y. Recent progress in enzymology and molecular biology of enzymes involved in vitamin D metabolism. J Lipid Res 1995; 36:1641. Bouillon R, Okamura WH, Norman AW. Structure-function relationships in the vitamin D endocrine system. Endocr Rev 1995; 16:200.

3. 4. 5. 6.

6a.

Bland R, Zehnder P, Hewison M. Expression of 25-hydroxyvitamin D3 1 alpha-hydroxylase along the nephron: new insights into renal vitamin D metabolism. Curr Opin Nephrol Hypertens 2000; 9:17. Kawashima H, Jorika S, Kurokawa K. Localization of 25-hydroxyvitamin D 3 -1 a-hydroxylase and 24-hydroxylase along the rat nephron. Proc Natl Acad Sci U S A 1981; 78:1199. Halloran BP. Cellular growth and differentiation during embryogenesis and fetal development. The role of vitamin D. Adv Exp Med Biol 1994; 352:227. Barbour GL, Coburn JW, Slatopolsky E, et al. Hypercalcemia in a anephric patient with sarcoidosis: evidence for extrarenal generation of 1,25-dihydroxyvitamin D. N Engl J Med 1981; 305:440. Adams TS, Sharma DP, Gacad MA, Singer FR. Metabolism of 25-hydroxy-vitamin D 3 by cultured pulmonary alveolar macrophages in sarcoidosis. J Clin Invest 1983; 72:1856. Takeyama K, Kitanaka S, Sato T, et al. 25-Hydroxyvitamin D3 1 a-hydroxylase and vitamin D synthesis. Science 1997; 277:1827.

7. 8. 9. 10. 11.

11a. Hewison M, Zehnder D, Bland R, Stewart PM. 1 alpha-hydroxylase and the action of vitamin D. J Mol Endocrinol 2000; 25:141. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. Yoon PS, DeLuca HF. Purification and properties of chick renal mitochondrial ferredoxin. Biochemistry 1980; 19:2165. Halloran BP, Portale AA, Lonergan ET, Morris RC Jr. Production and metabolic clearance of 1,25-dihydroxyvitamin D in men: effect of advancing age. J Clin Endocrinol Metab 1990; 70:318. Ohyama Y, Noshiro M, Okuda K. Cloning and expression of cDNA encoding 25-hydroxyvitamin D3 24-hydroxylase. FEBS Lett 1991; 28;278:195. Omdahl J, May B. The 25-hydroxyvitamin D 24-hydroxylase. In: Feldman D, Glorieux FH, Pike JW, eds. Vitamin D. San Diego: Academic Press, 1997:69. Horst RL, Reinhardt TA. Vitamin D metabolism In: Feldman D, Glorieux FH, Pike JW, eds. Vitamin D. San Diego: Academic Press, 1997:13. Fraser DR. Regulation of the metabolism of vitamin D. Physiol Rev 1980; 60:551. Selby PL, Davies M, Marks JS, Mawer EB. Vitamin D intoxication causes hypercalcaemia by increased bone resorption which responds to pamidronate. Clin Endocrinol (Oxf) 1995; 43:531. Slovik DM, Daly MA, Potts JT Jr, Neer RM. Renal 1,25-dihydroxyvitamin D, phosphaturic, and cyclic-AMP responses to intravenous synthetic human parathyroid hormone-(1-34) administration in normal subjects.Clin Endocrinol (Oxf) 1984; 20:369. Gray RW. Evidence that somatomedins mediate the effect of hypophosphatemia to increase serum 1,25-dihydroxyvitamin D3 levels in rats. Endocrinology 1987; 121(2):504. Henry HL, Norman AW. Vitamin D: metabolism and biological actions. Annu Rev Nutr 1984; 4:493. Horiuchi N, Takahashi H, Matsumoto T, et al. Salmon calcitonin-induced stimulation of 1,25-dihydroxycholecalciferol synthesis in rats involving mechanism independent of adenosine 3,5-cyclic monophosphate. Biochem J 1979; 184:269. Chen ML, Boltz MA, Armbrecht HJ. Effects of 1,25-dihydroxyvitamin D3 and phorbol ester on 25-hydroxyvitamin D3 24-hydroxylase cytochrome P450 messenger ribonucleic acid levels in primary cultures of rat renal cells. Endocrinology 1993; 132:1782. Care AD. Vitamin D in pregnancy, the fetoplacental unit and lactation. In: Feldman D, Glorieux FH, Pike JW, eds. Vitamin D. San Diego: Academic Press, 1997:437. Heaney RP, Skillman JG. Calcium metabolism in normal pregnancy. J Clin Endocrinol Metab 1971; 24:661. Bouillon R, Van Assche FA, Van Baelen H, et al. Influence of the vitamin D binding protein on the serum concentration of 1,25-dihydroxyvitamin D. J Clin Invest 1981; 67:689. Daiger SP, Shanfield MS, Calli-Sforza LL. Group-specific components (GC) proteins bind vitamin D and 25-hydroxyvitamin D. Proc Natl Acad Sci U S A 1975; 72:2076. Belsey R, Clark MB, Bernat M, et al. The physiologic significance of plasma transport of vitamin D and metabolites. Am J Med 1974; 57:50. Safadi FF, Thornton P, Magiera H, et al. Osteopathy and resistance to vitamin D toxicity in mice null for vitamin D binding protein. J Clin Invest 1999; 103:239. Cooke NE, Haddad JG. Vitamin D binding proteins. In: Feldman D, Glorieux FH, Pike JW, eds. Vitamin D. San Diego: Academic Press, 1997:87. Hahn JT, Birge SJ, Scharp CK, Avioli LV. Phenobarbital-induced alterations in vitamin D metabolism. J Clin Invest 1972; 51:741. Hollis BW, Lambert PW, Horst RL. Factors affecting the antirachitic sterol content of native milk. In: Holick MF, Anast CS, Gray TK, eds. Calcium, phosphate and vitamin D metabolism in pregnancy and neonate. Amsterdam: Elsevier, 1983:157. Wasserman R. Vitamin D and intestinal absorption of calcium and phosphorus. In: Feldman D, Glorieux FH, Pike JW, eds. Vitamin D. San Diego: Academic Press, 1997:259. Christakos S, Gabrielides C, Rhoten WB. Vitamin Ddependent calcium binding proteins: chemistry, distribution, functional considerations, and molecular biology. Endocr Rev 1989; 10:3. raviso GL, Garrett KP, Clemens TL. 1,25-Dihydroxyvitamin D3 induces synthesis of vitamin Ddependent calcium-binding protein in cultured chick kidney cells. Endocrinology 1987; 120:894. DelValle ME, Vazquez E, Represa J, et al. Immunohistochemical localization of calcium-binding proteins in the human cutaneous sensory corpuscles. Neurosci Lett 1994; 168:247. Lee DB. Mechanisms and regulation of intestinal phosphate transport. Adv Exp Med Biol 1986; 208:207. Matsumota T, Fontain OP, Rasmussen H. Effect of 1,25-dihydroxyvitamin D3 on phospholipid metabolism in chick duodenal mucosa cell. J Biol Chem 1981; 256:3354. Norman AW. Rapid biological actions mediated by 1a,25-dihydroxyvita min D3: a case study of transcaltachia (rapid hormonal stimulation of intestinal calcium transport). In: Feldman D, Glorieux FH, Pike JW, eds. Vitamin D. San Diego: Academic Press, 1997:233. Underwood J, DeLuca HF. Vitamin D is not directly necessary for bone growth and mineralization. Am J Physiol 1984; 246:E493. Raisz LG, Kream BE. Regulation of bone formation (parts I and II). N Engl J Med 1983; 309:29. Aubin JE, Heershe JNM. Vitamin D and osteoblasts. In: Feldman D, Glorieux FH, Pike JW, eds. Vitamin D. San Diego: Academic Press, 1997:313. Takahashi N, Yamana H, Yoshiki S, et al. Osteoclast-like cell formation by osteotropic hormones in mouse bone marrow cultures. Endocrinology 1988; 122:1373. Russell J, Letteriere D, Sherwood LM. Suppression by 1,25-(OH)2D3 of transcription of the preproparathyroid hormone gene. Endocrinology 1986; 119:2864. Slatopolsky E, Weerts C, Thielan J, et al. Marked suppression of secondary hyperparathyroidism by intravenous administration of 1,25-dihydroxy-cholecalciferol in uremic patients. J Clin Invest 1984; 74:2136. Kadowski S, Norman AW. Dietary vitamin D is essential for normal insulin secretion from the perfused rat pancreas. J Clin Invest 1984; 73:759. Abe E, Mayaura C, Sakagami H, et al. Differentiation of mouse myeloid leukemia cells induced by 1,25-dihydroxyvitamin D3. Proc Natl Acad Sci U S A 1981; 78:4990. Hosomi J, Hosoi J, Abe E, et al. Regulation of terminal differentiation of cultured mouse epidermal cells by 1,25-dihydroxyvitamin D3. Endocrinology 1983; 113:1950. Amento EP. Vitamin D and the immune system. Steroids 1987; 49:55. Takahashi N, Akatsu T, Sakasi T, et al. Induction of calcitonin receptors by 1,25-dihydroxyvitamin D3 in osteoclast-like multinucleated cells formed in mouse bone marrow cells. Endocrinology 1988; 123:1504. Boiullon R, Allewaert K, Xiang DZ, et al. Vitamin D analogs with low affinity for the vitamin D binding protein: enhanced in vitro and decreased in vivo activity. J Bone Miner Res 1993; 6:1051. Pike JW. The vitamin D receptor and its gene. In: Feldman D, Glorieux FH, Pike JW, eds. Vitamin D. San Diego: Academic Press, 1997:105. Lieberman UA, Eil C, Marx SJ. Resistance to 1,25-(OH)2-D: association with heterogeneous defects in cultured skin fibroblasts. J Clin Invest 1983; 71:192. Hewison M, Rut AR, Kristjansson K, et al. Tissue resistance to vitamin D without a mutation of the vitamin D receptor gene. Clin Endocrinol (Oxf) 1993; 39:663. Li YC, Pirro AE, Amling M, et al. Targeted ablation of the vitamin D receptor: an animal model of vitamin Ddependent rickets type II with alopecia. Proc Natl Acad Sci U S A 1997; 94:9831. Morrison NA, Yeoman R, Kelly PJ, Eisman JA. Contribution of trans-acting factor alleles to the normal physiological variability: vitamin D receptor gene polymorphism and circulating osteocalcin. Proc Natl Acad Sci U S A 1992; 89:6665. Morrison NA, Qi JC, Tolita A, et al. Prediction of bone density from vitamin D receptor alleles. Nature 1994; 367:284. Cooper GS, Umbach DM. Are vitamin D receptor polymorphisms associated with bone mineral density? A meta-analysis. J Bone Miner Res 1996; 11:1841. Hollis BW. Detection of vitamin D and its major metabolites. In: Feldman D, Glorieux FH, Pike JW, eds. Vitamin D. San Diego: Academic Press, 1997:587. Belsey R, DeLuca HF, Potts JT Jr. Competitive protein-binding assay for vitamin D and 25-OH-vitamin D. J Clin Endocrinol Metab 1971; 33:554. Haddad JG, Chyu KJ. Competitive protein-binding radioassay for 25-hydroxycholecalciferol. J Clin Endocrinol Metab 1971; 32:992. Stamp TCB, Round JM. Seasonal changes in human plasma levels of 25-hydroxyvitamin D. Nature 1974; 247:1563. Preece MA, Tomlinson S, Ribot CA, et al. Studies of vitamin D deficiency in man. Q J Med 1975; 44:575. Brumbaugh PF, Haussler DH, Bursac KM, Haussler MR. Filter assay for 1,25-dihydroxyvitamin D3: utilization of the hormone's target tissue chromatin receptor. Biochemistry 1974; 13:4091. Clemens TL, Hendy GN, Papapoulos SE, et al. Measurement of 1,25-dihydroxycholecalciferol in man by radioimmunoassay. Clin Endocrinol (Oxf) 1979; 11:325.

CHAPTER 55 BONE QUANTIFICATION AND DYNAMICS OF TURNOVER Principles and Practice of Endocrinology and Metabolism

CHAPTER 55 BONE QUANTIFICATION AND DYNAMICS OF TURNOVER

DAVID W. DEMPSTER AND ELIZABETH SHANE Technique of Iliac Crest Biopsy Pretreatment of the Patient with Tetracycline The Site and the Surgical Procedure Preparation and Analysis of the Biopsy Specimen Variables Routinely Measured in the Analysis of the Biopsy Specimen Static Variables Normal Bone Remodeling Hyperparathyroidism Osteomalacia Renal Osteodystrophy Osteoporosis Indications for Biopsy Chapter References

The period since the 1970s has produced tremendous growth in the understanding of the pathophysiology of bone. This has resulted largely from an increased comprehension of normal and abnormal bone structure and from a much clearer grasp of the dynamic cellular processes involved in bone remodeling. Three major advances have greatly facilitated the accumulation of this information. First, the advent of plastic embedding media has made it possible to obtain, routinely, good quality histologic sections of fully mineralized bone. Previously, it was necessary to decalcify all bone specimens before histologic evaluation, and fundamental structural information was literally washed down the sink with the decalcifying fluid. The second important advance was the discovery that the tetracycline antibiotics are permanently incorporated at sites of bone formation, allowing these regions to be visualized and quantitatively analyzed in histologic sections. Finally, a simple, safe, and relatively atraumatic surgical technique for obtaining well-preserved and adequately sized biopsy samples with the patient under local anesthesia was perfected. Consequently, the iliac crest biopsy has become one of the most powerful clinical and research tools available for the study of metabolic bone disease. This chapter describes the surgical procedure for obtaining the specimen, the methods by which the bone sample is prepared and analyzed, the variables that are measured and their clinical relevance, and the indications for biopsy.

TECHNIQUE OF ILIAC CREST BIOPSY

PRETREATMENT OF THE PATIENT WITH TETRACYCLINE The information obtainable from an iliac crest biopsy is significantly increased by the use of in vivo markers of bone formation, such as the tetracycline antibiotics.1 Approximately 3 weeks before the scheduled biopsy, tetracycline is administered for 2 days. This is followed by an interval of 12 drug-free days and then a 4-day course of tetracycline (2:12:4 sequence). The tetracycline is incorporated into bone at sites of new bone formation, binding irreversibly to hydroxyapatite at the mineralization front. When thin biopsy sections are subsequently viewed under violet or blue light, the tetracycline fluoresces so that bright lines (labels) are visible at the formation sites (Fig. 55-1). At sites where bone formation was ongoing during the entire labeling sequence, two tetracycline labels are present, corresponding to the two discrete periods of tetracycline exposure. At sites where formation either commenced or ended in the interval with no tetracycline administration, only single fluorescent labels are visible. At least 5 days should elapse between the last day of tetracycline administration and the date of the biopsy to prevent the last label from diffusing out.

FIGURE 55-1. Double tetracycline labels at a site of active bone formation in an iliac crest biopsy specimen. The labeling sequence was 2:12:4. The first and second labels are indicated. Note that the second label, which was deposited over 4 days, is broader than the first. This may be helpful in distinguishing the labels that have been administered for bone biopsy purposes from those deposited during previous treatment with tetracyclines as antibiotic agents. (B, mineralized bone; O, osteoid; M, marrow.) Epifluorescence microscopy with a mercury light source and a violet excitation filter. Field width = 0.45 mm.

Demeclocycline, tetracycline, and oxytetracycline all are effective labeling agents, although demeclocycline appears to produce more intense fluorescent bands. Demeclocycline, 600 mg per day (4 150-mg tablets), is taken on an empty stomach; dairy products and antacids containing aluminum, calcium, or magnesium impair absorption. Tetracyclines should not be given to children younger than 8 years of age or to pregnant women because it is incorporated into growing teeth and causes discoloration. Although tetracycline is generally well tolerated, some patients react adversely to these drugs with nausea, vomiting, or diarrhea. All patients should avoid direct exposure to sunlight or ultraviolet light during the time of administration because of potential skin phototoxicity. THE SITE AND THE SURGICAL PROCEDURE The anterior iliac crest is the preferred site for several reasons: (a) it is easily accessible, making the surgical procedure simple and safe; (b) the bone sample obtained consists of both trabecular and cortical bone (Fig. 55-2A); (c) the amount of trabecular bone correlates with that of the axial skeleton (vertebral bodies); and (d) a large body of histomorphometric data has been accumulated at this site in both normal persons and patients with a wide variety of bone diseases.2, 3, 4 and 5

FIGURE 55-2. A, Iliac crest biopsy specimen from a 33-year-old normal man. Note the inner and outer cortices (C) and intervening trabecular bone (T). Goldner trichrome stain. (M, marrow.) Field width = 9.6 mm. B, Higher-power view of normal trabecular architecture. Arrows indicate normal osteoid seam. Field width = 2.3

mm.

The biopsy generally is performed with a standard trephine of the type designed by Bordier. The internal diameter of the trephine should be 8 mm to obtain sufficient tissue, minimize compression of the sample, and reduce sampling error.3,4 The patient should be sedated, usually with intravenous meperidine hydrochloride (Demerol) and diazepam (Valium), immediately before the procedure. It is important to provide thorough local anesthesia of the skin, subcutaneous tissue, muscle, and, particularly, the periosteum covering both the lateral and the medial aspects of the ilium. The specimen is obtained through an incision 23 cm long made at a point 2 cm posterior and 2 cm inferior to the anterior superior iliac spine (Fig.55-3). Proper positioning of the trephine is essential for accurate interpretation of the biopsy. The needle should be rotated back and forth with gentle but firm pressure so that it cuts rather than pushes through the ilium. Excessive force may yield a damaged and uninterpretable biopsy specimen. After the procedure, the patient should refrain from excessive activity for 24 hours; a mild analgesic may be necessary.

FIGURE 55-3. Demonstration of anterior iliac crest for Meunier trephine biopsy (2 cm posterior and 2 cm inferior to the anterior superior iliac spine).

When a transiliac crest bone biopsy is performed with adequate sedation and with careful local anesthesia of the skin and periosteum, there is minimal discomfort, and patient acceptance is generally excellent. Significant complications from transiliac bone biopsy are unusual. In an international multi-center study involving 9131 transiliac biopsies, complications were recorded in 64 patients (0.7%).3 The most common complications were pain at the biopsy site that persisted for more than 7 days after the procedure and hematoma; rarer complications included wound infection, fracture through the iliac crest, and osteomyelitis. PREPARATION AND ANALYSIS OF THE BIOPSY SPECIMEN The biopsy should be fixed in 70% ethanol because more aqueous fixatives may leach the tetracycline from the bone. After a fixation period of 47 days, the biopsy is dehydrated in ethanol, cleared in toluene, and embedded in methyl methacrylate. No decalcification steps are performed. Once polymerized, the methyl methacrylate provides a sufficiently hard but flexible supporting matrix to allow good quality undecalcified sections to be cut with a heavy-duty microtome, usually equipped with a tungsten carbide-edged knife. The 5- to 10-m sections are then stained with a variety of dyes to allow good discrimination between calcified bone matrix and noncalcified osteoid (Fig. 55-4 and Fig. 55-5A) and clear visualization of the cellular components of bone and marrow (Fig. 55-5B). Unstained sections also are prepared to allow observation of the tetracycline labels by fluorescence microscopy (see Fig. 55-1).6,7

FIGURE 55-4. Iliac crest biopsy specimen from a patient with primary hyperparathyroidism caused by parathyroid carcinoma. Note extended resorption surface (white arrows) with associated marrow fibrosis and increased osteoid surface (black arrows). Goldner trichrome stain. Field width = 2.3 mm.

FIGURE 55-5. A, Iliac crest biopsy specimen from a patient with renal osteodystrophy. This is an example of mixed bone disease with evidence of secondary hyperparathyroidism (extended resorption and osteoid surface) and osteomalacia (increased width of osteoid seams). (O, osteoid; B, mineralized bone.) Goldner trichrome stain. Field width = 2.3 mm. B, Higher-power view of the area indicated in A to show osteoclasts (black arrows), osteoblasts (white arrows), and extensive marrow fibrosis (MF). Goldner trichrome stain. Field width = 0.47 mm. C, Histochemical staining procedure reveals the presence of aluminum (arrows) at the junction of an osteoid seam and mineralized bone in a patient with renal osteodystrophy who is receiving long-term hemodialysis. (B, mineralized bone; M, marrow.) Field width = 0.52 mm.

The sections are analyzed morphometrically, according to standard stereologic principles, using either simple point-counting techniques or the less time-consuming method of computer-aided planimetry.8,9 and 10

VARIABLES ROUTINELY MEASURED IN THE ANALYSIS OF THE BIOPSY SPECIMEN

An abundance of quantitative information may be obtained from the biopsy. Because the morphometric analysis is a labor-intensive process, the number of variables evaluated depends on whether the biopsy specimen is being analyzed for diagnostic or research purposes. Listed below are eight indices of trabecular bone that are of

particular clinical relevance. A detailed account of the complete analysis of bone biopsy specimens is found in more specialized texts.2,11 The histomorphometric parameters are generally divided into two categories. Static variables provide information on the amount of bone present and the proportion of bone surface engaged in a particular phase of remodeling activity. Dynamic variables yield information on the rate of cell-mediated processes involved in remodeling. This category can be evaluated only in biopsy specimens that have been appropriately labeled with tetracycline. Importantly, inferences about the rate of cellular activities can be made only from an analysis of the dynamic variables. An increase in resorption surface, a static variable, does not necessarily indicate an increase in resorption rate. The bone formation rate may be calculated directly from selected dynamic variables measured in an appropriately labeled biopsy specimen. Resorption rate, on the other hand, can be computed only indirectly, using certain indices of bone formation, in a single biopsy specimen or from two sequential biopsy specimens.11,12 STATIC VARIABLES The following variables are categorized as static. The terms and abbreviations used are consistent with the recommendation of the Bone Histomorphometry Nomenclature Committee of the American Society for Bone and Mineral Research.13 Cancellous bone volume (Cn-BV/TV, %). This is the fraction of a given volume of whole cancellous bone tissue (i.e., bone + marrow) that consists of mineralized and nonmineralized bone. Osteoid volume (OV/BV, %). This is the fraction of a given volume of bone tissue (mineralized bone + osteoid) that is osteoid (i.e., unmineralized matrix). Osteoid surface (OS/BS, %). This is the fraction of the entire trabecular surface that is covered by osteoid seams. Osteoid thickness (O.Th, m). This is the average width of the osteoid seams. Eroded surface (ES/BS, %). This is the fraction of the entire trabecular surface that is occupied by resorption bays (Howship lacunae), including both those with and without osteoclasts. The following variables are categorized as dynamic. Mineral apposition rate (MAR, m per day). This is calculated by dividing the average distance between the first and second tetracycline labels by the time interval (e.g., 12 days) separating them. It is a measure of the linear rate of production of calcified bone matrix by the osteoblasts. Mineralizing surface (MS/BS, %). This is the fraction of trabecular surface bearing double tetracycline labels plus one-half of the singly labeled surface. It is a measure of the proportion of bone surface on which new mineralized bone was being deposited at the time of tetracycline labeling. Bone Formation Rate (BFR/BS, m3/m2 per day). This is the volume of mineralized bone made per unit surface of trabecular bone per year. It is calculated by multiplying the mineralizing surface by the mineral apposition rate. Table 55-1 indicates how each of these variables is altered in different disease states.

TABLE 55-1. Bone Biopsy Variables in a Variety of Disease States

NORMAL BONE REMODELING Bone is remodeled continuously throughout lifeold, worn out bone is replaced by new bone. Approximately 25% of trabecular bone and 3% of cortical bone is replaced annually. The remodeling process is a quantum phenomenon because it occurs in discrete units or packets. Osteoclasts resorb a preprogrammed volume of old bone. Shortly thereafter, the osteoclasts are replaced by osteoblasts, which refill the resorption cavity with a new packet of bone. The group of cells (osteoclasts and osteoblasts) that creates one new packet of bone is called a bone remodeling unit. In normal trabecular bone, ~900 bone remodeling units are activated each day. In cortical bone, the activation frequency is ~180 per day.2,14,15 and 16 HYPERPARATHYROIDISM Elevated circulating levels of parathyroid hormone (PTH) stimulate the activation frequency of bone remodeling units, resulting in increased numbers of osteoclasts in bone and, because of the coupling process, increased numbers of osteoblasts. Consequently, quantitative analysis of a biopsy specimen from a patient with either primary or secondary hyperparathyroidism reveals increases in eroded surface, osteoid surface, and mineralizing surface (see Fig. 55-4).10,17,18 and 19 The extension of surface double labeled by tetracycline has allowed precise determination of the mineral apposition rate, which has been shown to be reduced. Because the osteoid thickness is normal in this disorder, the reduction in mineral apposition rate indicates a decreased rate of organic matrix production by the osteoblasts, rather than a defect in calcification. However, the extension of mineralizing surface overcompensates for the decrease in mineral apposition rate, so that the overall bone formation rate, the product of these two variables, is still increased. Thus, in hyperparathyroidism, bone resorption and formation rates are increased, which is a situation commonly referred to as a high turnover state. Bone turnover is higher in hyperthyroid patients with vitamin D insufficiency.18a Despite the rapid remodeling that occurs in primary hyperparathyroidism, the balance between resorption and formation is generally conserved because cancellous bone volume is normal or may even be increased, and trabecular connectivity is preserved.19 Often accompanying the increased bone turnover rate are increased deposition of immature (woven) bone and marrow fibrosis. These features are particularly prominent in severe secondary hyperparathyroidism. Because the hallmarks of excessive remodeling activity (e.g., increased eroded surface) accumulate in bone over time, the biopsy can be a sensitive indicator of parathyroid gland hyper activity, especially when this is mild or intermittent. However, the biopsy alone does not distinguish between primary and secondary hyperparathyroidism. OSTEOMALACIA Numerous causative mechanisms underlie osteomalacia (see Chap. 63 and Chap. 70). Most of the mechanisms involve processes that decrease the circulating calcium phosphate product below the levels required for bone mineralization. Although calcification of the organic matrix is inhibited, the osteoblasts continue to synthesize and secrete the matrix. An accumulation of unmineralized matrix (osteoid) results (Fig. 55-6). The cancellous bone volume is normal in osteomalacia, but as the osteoid volume is increased, the amount of mineralized bone is actually reduced.

FIGURE 55-6. A, Iliac crest biopsy specimen from a patient with osteomalacia (tumor-induced).22 There is a marked increase in the osteoid volume as a result of increased extent and thickness of osteoid seams. (O, osteoid; B, mineralized bone.) Goldner trichrome stain. Field width = 2.3 mm. B, Abnormal tetracycline deposition in the same patient. Compared with normal double labels seen in Figure 55-1, the labels (L) are broad, dull, and diffused. This pattern of tetracycline uptake is typical of osteomalacia and is thought to result from the presence of increased amounts of immature mineral that is capable of binding tetracycline.20 (B, mineralized bone; O, osteoid; M, marrow.) Epifluorescence microscopy with a mercury light source and a violet excitation filter. Field width = 0.45 mm.

Analysis of the dynamic variables is particularly important in osteomalacia. At some formation sites, mineral is still deposited, but less rapidly than normal, resulting in reduced separation between the double tetracycline labels and, consequently, low values for mineral apposition rate. At other formation sites, calcification may be completely inhibited, decreasing the extent of tetracycline uptake and therefore the mineralizing surface. The fall in both these variables markedly reduces bone formation rate. As regards the static variables, the accumulation of unmineralized matrix is reflected by increased osteoid thickness, osteoid surface, and osteoid volume. If PTH secretion is elevated, the activation frequency of bone remodeling units is enhanced and the biopsy reveals an increase in eroded surface. The elevated remodeling rate, along with the mineralization failure, accentuates the increased deposition of osteoid.10,20,21 and 22 RENAL OSTEODYSTROPHY Chronic renal failure is generally accompanied by phosphate retention and hyperphosphatemia, which lead to a reciprocal decrease in serum ionized calcium concentration and a consequent increase in secretion of PTH. Moreover, as functional renal mass decreases, the plasma 1,25-dihydroxyvitamin D level falls (see Chap. 54 and Chap. 209), leading to impaired intestinal calcium absorption, which contributes to the tendency for hypocalcemia and ultimately may result in reduced bone mineralization. Consequent to these marked changes in calcium and phosphate metabolism, the bone biopsy specimen in renal osteodystrophy may display a heterogeneous picture; bone biopsy has contributed significantly to the current classification of renal bone disease.23,24,25,26,27,28,29,30,31,32,33,34 and 35 Thus, renal osteodystrophy has been subdivided into two broad categories, primarily on the basis of histomorphometric features: one characterized by normal or high bone turnover and a second characterized by low bone turnover. In patients with end-stage renal disease, the most frequently observed biopsy changes arise from the effects of chronic excess PTH secretion on the skeleton, are classified as normal/high turnover, and include (a) osteitis fibrosa, (b) mild hyperparathyroidism, and (c) mixed bone disease. As a group, these forms are characterized histologically by increases in eroded surface, osteoid surface, and mineralizing surface. With osteitis fibrosa, however, there is deposition of woven osteoid and variable amounts of peritrabecular marrow fibrosis, in contrast to the minimal or absent fibrosis observed in mild hyperparathyroidism. States of low bone turnover also are frequently observed in patients undergoing dialysis, albeit less often than high turnover disease. The low turnover states are classified as (d) osteomalacia and (e) aplastic or adynamic bone disease. Patients with osteomalacia have evidence of reduced values for dynamic variables associated with the accumulation of excess unmineralized osteoid, whereas those with aplastic or adynamic disease have reduced bone formation as measured by tetracycline labeling, but normal or reduced osteoid volume. Most symptomatic patients with osteomalacia or aplastic bone disease have a significant extent (>25%) of surfaces covered with stainable aluminum and are considered to have aluminum-related bone disease.24,25,26 and 27 The major sources of this aluminum are the aluminum-containing phosphate binders used to control hyperphosphatemia and dialysate solutions contaminated with aluminum (see Chap. 61 and Chap. 131). Aluminum accumulates at sites of bone formation (see Fig. 55-5C), where it may directly inhibit mineralization, leading to an increase in osteoid thickness and osteoid surface. However, aluminum also is toxic to osteoblasts and may impair their ability both to synthesize and to mineralize bone matrix. These two effects of aluminum reduce both the mineral apposition rate and the mineralizing surface. However, if matrix production also is inhibited, osteoid thickness is not increased. With greater recognition of the potential sources of aluminum and the substitution of phosphate binders containing calcium for those containing aluminum, aluminum-related bone disease is becoming less of a clinical problem. Another form of low-turnover bone disease has been described that is termed idiopathic aplastic or adynamic bone disease and is not characterized by significant aluminum accumulation or staining. Its pathogenesis is unclear but may be related to various therapeutic maneuvers designed to prevent or treat hyperparathyroidism in patients undergoing dialysis, including the use of dialysates with higher calcium concentrations (3.03.5 mEq/L), large doses of calcium-containing phosphate binders, and calcitriol therapy. In general, these patients tend to have few or no symptoms of bone disease, and this disease ultimately may prove to be a histologic rather than a clinically relevant form of bone disorder. It is unknown whether patients with aplastic bone disease are at increased risk for the development of clinical problems in the future. The form of renal bone disease that is classified as mixed disease is characterized histologically by features of both osteitis fibrosa and osteomalacia. Such lesions are observed most commonly in patients who are in transition between osteitis fibrosa and aluminum-related bone disease. OSTEOPOROSIS The most striking feature of bone biopsy specimens from patients with osteoporosis is the reduction in cancellous bone volume (see Chap. 64). Approximately 80% of patients with vertebral crush fractures have values that are lower than normal. The reduction in cancellous bone volume results primarily from progressive loss of entire trabeculae and, to a lesser degree, from thinning of those that remain (Fig. 55-7).36,37,38 39,39a

FIGURE 55-7. Iliac crest biopsy specimen from a patient with post-menopausal osteoporosis. Note the marked reduction in cancellous bone volume and in the thickness of the cortices compared with the normal biopsy specimen shown in Figure 55-2A. Goldner trichrome stain. Field width = 9.6 mm.

Concerning the changes in the other static and dynamic variables in osteoporosis, there is still considerable debate over whether patients can be stratified into high, normal, or low turnover groups. Even if they can, the pathogenetic and clinical significance of this so-called histologic heterogeneity in patients with osteoporosis is unclear. In a study of 50 postmenopausal women with untreated osteoporosis, the investigators identified two subsets of patients, one with normal turnover and one with high turnover, the latter representing 30% of the cases.40 However, this conclusion was based on the finding of a bimodal distribution in osteoid surface. The tetracycline-based bone formation rate, a more reliable measure of turnover rate, showed a normal distribution. Based on the interval between the 10th percentile and the 90th percentile for calculated bone resorption rate in a large group (n = 32) of normal postmenopausal women, another study classified 30% of women with untreated postmenopausal osteoporosis as having high turnover, whereas 64% and 6% had normal and low turnover, respectively.41 When bone formation rate was used as the discriminant variable, 19% were classified as having high turnover, 72% as having normal turnover, and 9% as having low turnover. Alternatively, in two

studies of postmenopausal women with osteoporosis and their normal counterparts, the same wide variation in turnover indices was found in both groups, leading to the conclusion that there were no important subsets of patients with osteoporosis.42,43 These studies, however, confirmed the earlier observations of others that, as a group, women with osteoporosis display a decrease in bone formation rate, and that some patients show profoundly depressed formation with little or no tetracycline uptake.38 From a clinical perspective, the impetus for classifying patients with osteoporosis according to their turnover status stems from the notion that the turnover rate may influence the response to particular therapeutic agents. For example, patients with high turnover rates may respond better to antiresorptive treatments. There is evidence that this is true for calcitonin.44 However, in clinical practice, the biopsy is an impractical way to determine turnover status, and it is likely that biochemical markers of bone resorption and formation will be used increasingly for this purpose. It is important to note that, in most cases, bone biopsy is performed when the disease is in an advanced stage, with multiple fractures already having occurred. It is probable that, in many cases, the disturbances in bone metabolism that led to the reduction in bone mass took place several years before the time of the biopsy and are no longer evident.45 Another confounding factor is that most patients who undergo biopsy for osteoporosis already have been treated with one or more pharmaceutical agents. INDICATIONS FOR BIOPSY Generally, biopsy of the iliac crest is useful only in diffuse diseases of the skeleton. A major use of the bone biopsy is as a diagnostic tool in patients with skeletal disease manifested by bone pain, fractures, or osteopenia of unclear cause or pathogenesis. In such cases, a biopsy specimen can provide important information about a pathologic process. There are only exceptional indications for performing this procedure in patients with localized skeletal disease such as Paget disease of bone, primary bone tumors, or bone metastases involving the iliac crest. The indications for bone biopsy in women with postmenopausal osteoporosis are controversial. It would be difficult to perform biopsies on the many women who have this disease. However, bone biopsy can provide useful information in groups of patients less commonly affected by osteoporosis, such as young men and premenopausal women. Patients with osteopenia or women with postmenopausal osteoporosis should not undergo biopsy merely to measure cancellous bone volume to confirm the diagnosis of osteoporosis. The intraindividual and interindividual variability in cancellous bone volume is too great, and there is too much overlap between cancellous bone volumes in patients with clinical osteoporosis and normal persons to make it useful. However, bone biopsy is useful to exclude subclinical osteomalacia. In one study, 5% of patients with typical crush fracture syndrome had clear-cut histologic evidence of osteomalacia despite normal biochemical and radiologic variables.52 Moreover, the biopsy can be useful in defining, more precisely, the probable mechanisms of bone loss in individual patients with osteoporosis, and may aid in choosing the most appropriate course of action. For example, if the biopsy reveals or confirms a high bone turnover rate, it is important to carefully rule out endocrine disorders, such as hyperthyroidism and hyperparathyroidism. Finally, the biopsy is the best available way to evaluate the effect of various therapeutic maneuvers on bone cell function.53,54 Persons with renal osteodystrophy represent another group of patients in whom a transiliac crest bone biopsy may be extremely helpful. Once again, however, the large number of patients with renal disease precludes the use of this procedure in every case. Generally, if a symptomatic patient has biochemical evidence of secondary hyperparathyroidism (hyperphosphatemia, hypocalcemia, and markedly elevated intact PTH levels), biopsy is unnecessary because it almost certainly will reveal osteitis fibrosa. However, patients with renal disease who have bone pain and fractures but who do not have the biochemical hallmarks typical of secondary hyperparathyroidism should undergo biopsy to determine whether they have osteomalacia or idiopathic aplastic bone disease, and whether aluminum accumulation appears to be the primary etiologic factor. Although biopsy can be useful in the clinical management of certain patients with bone disease, its primary use today is as a powerful research tool. CHAPTER REFERENCES
1. Frost HM. Bone histomorphometry: choice of marking agent and labeling schedule. In: Recker RR, ed. Bone histomorphometry: techniques and interpretation. Boca Raton, FL: CRC Press, 1983:37. 2. Parfitt AM. The physiological and clinical significance of bone histomorphometric data. In: Recker RR, ed. Bone histomorphometry: techniques and interpretation. Boca Raton, FL: CRC Press, 1983:143. 3. Rao DS. Practical approach to bone biopsy. In: Recker RR, ed. Bone histomorphometry: techniques and interpretation. Boca Raton, FL: CRC Press, 1983:3. 4. Bordier P, Matrajt H, Miravet B, Hioco D. Mesure histologique de la masse et de la rsorption des travs osseuse. Pathol Biol (Paris) 1964; 12:1238. 5. Dempster DW. The relationship between the iliac crest bone biopsy and other skeletal sites. In: Kleerekoper M, Krane S, eds. Clinical disorders of bone and mineral metabolism. New York: Mary Ann Liebert, Inc, 1988:247. 6. Baron R, Vignery A, Neff L, et al. Processing of undecalcified bone specimens for bone histomorphometry. In: Recker RR, ed. Bone histomorphometry: techniques and interpretation. Boca Raton, FL: CRC Press, 1983:13. 7. Weinstein RS. Clinical use of bone biopsy. In: Coe FL, Favus MJ, eds. Disorders of bone and mineral metabolism. New York: Raven Press, 1992:455. 8. Parfitt AM. Stereological basis of bone histomorphometry; theory of quantitative microscopy and reconstruction of the third dimension. In: Recker RR, ed. Bone histomorphometry: techniques and interpretation. Boca Raton, FL: CRC Press, 1983:53. 9. Compston J. Bone histomorphometry. In: Arnett TR, Henderson B, eds. Methods in bone biology. London: Chapman and Hall, 1997:177. 10. Malluche HH, Faugere M-C. Atlas of mineralized bone histology. Basel: Karger, 1987. 11. Frost HM. Bone histomorphometry: analysis of trabecular bone dynamics. In: Recker RR, ed Bone histomorphometry: techniques and interpretation. Boca Raton, FL: CRC Press, 1983:109. 12. Eriksen EF. Normal and pathological remodeling of human trabecular bone: three dimensional reconstruction of the remodeling sequence in normals and in metabolic bone disease. Endocr Rev 1986; 7:379. 13. Parfitt AM, Drezner MK, Glorieux FH, et al. Bone histomorphometry: standardization of nomenclature, symbols, and units. J Bone Miner Res 1987; 2:595. 14. Frost HM. Bone remodeling and its relationship to metabolic bone diseases. Springfield, IL: Charles C Thomas Publisher, 1973. 15. Parfitt AM. Bone remodeling: relationship to the amount and structure of bone, and the pathogenesis and prevention of fractures. In: Riggs BL, Melton LJ III, eds. Osteoporosis: etiology, diagnosis, and management. New York: Raven Press, 1988:45. 16. Dempster DW. Bone remodeling. In: Coe FL, Favus MJ, eds. Disorders of bone and mineral metabolism. New York: Raven Press, 1992:355. 17. Melsen F, Mosekilde L, Kragstrup J. Metabolic bone diseases as evaluated by bone histomorphometry. In: Recker RR, ed. Bone histomorphometry: techniques and interpretation. Boca Raton, FL: CRC Press, 1983:265. 18. Parisien M, Silverberg SJ, Shane E, et al. The histomorphometry of bone in primary hyperparathyroidism: preservation of cancellous bone structure. J Clin Endocrinol Metab 1990; 70:930. 18a. Silverberg SJ, Shane E, Dempster DW, Bilezikian JP. The effects of vitamin D insufficiency in patients with primary hyperparathyroidism. Am J Med 1999; 107:561. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. Parisien MV, Mellish RWE, Silverberg SJ, et al. Maintenance of cancellous bone connectivity in primary hyperparathyroidism: trabecular strut analysis. J Bone Miner Res 1992; 7:913. Teitelbaum SL. Pathological manifestations of osteomalacia and rickets. J Clin Endocrinol Metab 1980; 9:43. Jaworksi ZFG. Histomorphometric characteristics of metabolic bone disease. In: Recker RR, ed. Bone histomorphometry: techniques and interpretation. Boca Raton, FL: CRC Press, 1983:241. Siris ES, Clemens TL, Dempster DW, et al. Tumor-induced osteomalacia. Kinetics of calcium, phosphorus, and vitamin D metabolism and characteristics of bone histomorphometry Am J Med 1987; 82:307. Malluche HH, Ritz E, Lange HP, et al. Bone histology in incipient and advanced renal failure. Kidney Int 1976; 9:355. Hodsman AB, Sherrard DJ, Alfrey AC, et al. Bone aluminum and histomorphometric features of renal osteodystrophy. J Clin Endocrinol Metab 1982; 54:539. Boyce BF, Fell GS, Elder HY, et al. Hypercalcemic osteomalacia due to aluminum toxicity. Lancet 1982; 2:1009. Dunstan CR, Hills E, Norman AW, et al. The pathogenesis of renal osteodystrophy: role of vitamin D, aluminum, parathyroid hormone, calcium and phosphorus. Q J Med 1985; 55:127. Parisien M, Charhon SA, Mainetti E, et al. Evidence for a toxic effect of aluminum on osteoblasts: a histomorphometric study in hemodialysis patients with aplastic bone disease. J Bone Miner Res 1988; 3:259. Charhon SA, Berland YF, Olmer MJ, et al. Effects of parathyroidectomy on bone formation and mineralization in hemodialized patients. Kidney Int 1985; 27:426. Sherrard DJ, Hercz G, Pei Y, et al. The spectrum of bone disease in end-stage renal failurean evolving disorder. Kidney Int 1993; 43:435. Coburn JW, Salusky IB. Hyperparathyroidism in renal failure: clinical features, diagnosis, and management. In: Bilezikian JP, Levine MA, Marcuo R, eds. The parathyroids. New York: Raven Press, 1994:721. Slatopolsky E, Delmaz J. Bone disease in chronic renal failure and after renal transplantation. In: Coe FL, Favus MF, eds. Disorders of bone and mineral metabolism. New York: Raven Press, 1992:905. Felsenfeld AJ, Rodriguez M, Dunlay R, Llach F. A comparison of parathyroid gland function in haemodialysis patients with different forms of renal osteodystrophy. Nephrol Dial Transplant 1991; 6:244. Salusky IB, Coburn JW, Brill J, et al. Bone disease in pediatric patients undergoing dialysis with CAPD or CCPD. Kidney Int 1988; 33:975. Moriniere P, Cohen-Solal M, Belbrik S, et al. Disappearance of aluminic bone disease in a long-term asymptomatic dialysis population restricting Al(OH) 3 intake: emergence of an idiopathic adynamic bone disease not related to aluminum. Nephron 1989; 53:93. Hercz F, Pei Y, Greenwood C, et al. Low turnover osteodystrophy without aluminum: the role of suppressed parathyroid function. Kidney Int 1993; 44:860. Meunier PJ, Sellami S, Briancon D, Edouard C. Histological heterogeneity of apparently idiopathic osteoporosis. In: Deluca HF, Frost HM, Jee WSS, et al, eds. Osteoporosis, recent advances in pathogenesis and treatment. Baltimore: University Park Press, 1981:293. Parfitt AM, Matthews CHE, Villanueva AR, et al. Relationships between surface, volume and thickness of iliac trabecular bone in aging and in osteoporosis. Implications for the microanatomic and cellular mechanisms of bone loss. J Clin Invest 1983; 72:1396. Whyte MP, Bergfeld MA, Murphy WA, et al. Postmenopausal osteoporosis; a heterogeneous disorder as assessed by histomorphometric analysis of iliac crest bone from untreated patients. Am J Med 1982; 72:193. Meunier PJ. Assessment of bone turnover by histomorphometry in osteoporosis. In: Riggs BL, Melton LJ III, eds. Osteoporosis: etiology, diagnosis, and management. New York: Raven Press, 1988:317.

39a. Dempster DW. The contribution of trabecular architecture to cancellous bone quality. J Bone Miner Res 2000; 15:20.

40. Arlot ME, Delmas PD, Chappard D, Meunier PJ. Trabecular and endocortical bone remodeling in postmenopausal osteoporosis: comparison with normal postmenopausal women. Osteoporos Int 1990; 1:41. 41. Eriksen EF, Hodgson SF, Eastell R, et al. Cancellous bone remodeling in type I (postmenopausal) osteoporosis: quantitative assessment of rates of formation, resorption, and bone loss at tissue and cellular levels. J Bone Miner Res 1990; 5:311. 42. Kimmel DB, Recker RR, Gallagher JC, et al. A comparison of iliac bone histomorphometric data in post-menopausal osteoporotic and normal subjects. Bone Miner 1990; 11:217. 43. Garcia Carasco M, de Vernejoul MC, Sterkers Y, et al. Decreased bone formation in osteoporotic patients compared with age-matched controls. Calcif Tissue Int 1989; 44:173. 44. Civitelli R, Gonnelli S, Zacchei F, et al. Bone turnover in postmenopausal osteoporosis. Effect of calcitonin treatment. J Clin Invest 1988; 82:1268. 45. Steiniche T, Christiansen P, Vesterby A, et al. Marked changes in iliac crest bone structure in postmenopausal women without any signs of disturbed bone remodeling or balance. Bone 1994; 15:73. 46. Bressot C, Meunier PJ, Chapuy MC, et al. Histomorphometric profile, pathophysiology and reversibility of corticosteroid-induced osteoporosis. Metab Bone Dis Relat Res 1979; 1:303. 47. Dempster DW. Bone histomorphometry in glucocorticoid-induced osteoporosis. J Bone Miner Res 1989; 4:137. 48. Meunier PJ, Coindre JM, Edouard CM, Arlot ME. Bone histomorphometry in Paget's disease; quantitative and dynamic analysis of Pagetic and non-pagetic bone tissue. Arthritis Rheum 1980; 23:1095. 49. Valentin-Opran A, Charhon SA, Meunier PJ, et al. Quantitative histology of myeloma-induced bone changes. Br J Haematol 1982; 52:601. 50. Baron R, Gertner JM, Lang R, Vignery A. Increased bone turnover with decreased bone formation by osteoblasts in children with osteogenesis imperfecta tarda. Pediatr Res 1983; 17:204. 51. Ste-Marie LG, Charhon SA, Edouard C, et al. Iliac bone histomorphometry in adults and children with osteogenesis imperfecta. J Clin Pathol 1984; 37:1081. 52. Meunier PJ. Bone biopsy in diagnosis of metabolic bone disease. In: Cohn DV, Talmage R, Matthews JL, eds. Hormonal control of calcium metabolism. Proceedings of the Seventh International Conference on Calcium Regulating Hormones. Amsterdam: Excerpta Medica, 1981:109. 53. Holland EFN, Chow JWM, Studd JWW, et al. Histomorphometric changes in the skeleton of postmenopausal women with low bone mineral density treated with percutaneous estradiol implants. Obstet Gynecol 1994; 83:387. 54. Marcus R, Leary D, Schneider DL, et al. The contribution of testosterone to skeletal development and maintenance: lessons from the androgen insensitivity syndrome. J Clin Endocrinol Metab 2000; 85:1032.

CHAPTER 56 MARKERS OF BONE METABOLISM Principles and Practice of Endocrinology and Metabolism

CHAPTER 56 MARKERS OF BONE METABOLISM

MARKUS J. SEIBEL, SIMON P. ROBINS, AND JOHN P. BILEZIKIAN Biochemistry of Bone Markers of Bone Metabolism Markers of Bone Formation Markers of Bone Resorption Use of Biochemical Markers of Bone Turnover in Osteoporosis Chapter References

Bone is a metabolically active tissue that throughout life undergoes constant remodeling. Skeletal turnover is achieved by two counteracting processes: bone formation and bone resorption. Whereas bone formation is a function of osteoblast activity, bone resorption is attributed to osteoclast activity. The metabolic and cellular events in the skeleton are regulated by a large number of systemic and local modulators, such as parathyroid hormone, vitamin D, sex hormones, glucocorticoids, calcitonin, prostaglandins, growth factors, and cytokines (see Chap. 49, Chap. 50, Chap. 51, Chap. 54, Chap. 55 and Chap. 173). Under normal conditions, bone formation and resorption are closely coupled to each other, and this balance maintains a stable bone mass. Metabolic bone diseases, in contrast, are characterized by more or less pronounced imbalances in bone turnover (often referred to as uncoupling). The long-term result of such imbalances in bone turnover often is changes in bone mass and structure, which either become clinically symptomatic (e.g., fracture), or may be detected by means of radiographic or densitometric techniques. In addition to these static measures, markers of bone turnover can be used to detect the dynamics of the metabolic imbalance itself. Bone mass and bone turnover are, therefore, complementary parameters in the assessment of skeletal homeostasis. This chapter reviews the basic biochemistry, methodology, and clinical application of the currently known markers of bone turnover (Table 56-1 and Table 56-2). The endocrine regulation of calcium homeostasis and bone metabolism is discussed in Chapter 49 and Chapter 50.

TABLE 56-1. Biomarkers of Bone Formation

TABLE 56-2. Biomarkers of Bone Resorption

BIOCHEMISTRY OF BONE
Bone is composed of ~70% mineral and 30% organic matter. The mineral, primarily in the form of hydroxyapatite [Ca10(PO4)6(OH)2] crystals, is embedded in and aligned with the collagen fibrils, which play an important role in crystal formation. This calcium-collagen composite ensures the two main functions of bone: providing a structural framework and acting as a reservoir for mineral ions. The organic phase of bone is made of cells and a protein matrix, of which ~90% is collagen type I. Bone also contains a large number of different proteins, glycoproteins, and proteoglycans, many of which are negatively charged (Table 56-3). Although their precise functions are not established, these noncollagenous proteins are probably associated with the organization and mineralization of the skeletal matrix.

TABLE 56-3. Biochemical Components of Bone

Collagen is synthesized by osteoblasts as a larger precursor molecule. This procollagen molecule contains the triple helix portion, which bears several hydroxyproline residues, and globular extensions at the N- and the C-terminal ends (i.e., N- and C-terminal propeptides). After secretion of the collagen molecule, these propeptides are removed en bloc by specific proteases at the cell surface. The C-terminal propeptide remains intact and appears as a 100-kDa protein in the blood. The helical collagen molecules spontaneously assemble into fibrils, which are then covalently cross-linked to impart the necessary tensile strength. This process is initiated extracellularly through the action of a single enzyme, lysyl oxidase. Thereafter, reactions of the lysine-derived aldehydes occur spontaneously and culminate in the

formation of trifunctional pyridinium cross-links (Fig. 56-1). Unlike hydroxyproline, which is already present in the newly synthesized protein, the pyridinium cross-links are found exclusively in mature collagens of the established extracellular matrix. Of the two cross-link analogs produced, deoxypyridinoline is located primarily in bone collagen. Pyridinoline occurs in cartilage and other soft tissues as well as in bone (see Chap. 189).

FIGURE 56-1. Lysyl-oxidasederived cross-links. Scheme illustrating the tissue-specific differences in the trifunctional, mature cross-links formed, depending on whether lysine or hydroxylysine residues in the telopeptides are oxidized.

Of the noncollagenous proteins in bone, osteocalcin, or bone Gla protein, is one of the most abundant (~15%). The protein contains 49 amino-acid residues, three of which are g-carboxyglutamic acid (GLA). Whereas the latter are formed in a posttranslational vitamin Kdependent process, transcription of the OC gene itself is controlled by 1,25-dihydroxyvitamin D3. A small proportion of the newly synthesized protein is transported directly into the blood, from which it is eliminated mainly through the kidneys. Other noncollagenous proteins in the extracellular matrix of bone are summarized in Table 56-3. Because many of these components are produced by osteoblasts and by a variety of other cell types, they may be found in nonskeletal tissues as well. This distribution is explained by the fact that most of these proteins play an important role in the general organization of the extracellular matrix, mediating basic processes such as cell attachment and migration, growth, development, and fibril formation. Thus, both osteopontin and bone sialoprotein are found predominantly in mineralizing tissues, where they serve as cell-attachment proteins by means of their intrinsic RGD amino-acid sequences. In bone, the proteins are synthesized by osteoblasts and mediate the attachment of osteoclasts to the mineralized matrix. Approximately 25% of the noncollagenous proteins found in the extracellular matrix of bone are actually derived from the serum and are not osteogenic products (see Table 56-3). Besides the collagenous and noncollagenous proteins of bone, several specific enzymes play an important role in skeletal metabolism. Alkaline phosphatase is a characteristic product of active osteoblasts and young osteocytes. Although the function of this enzyme in bone metabolism is not clear, a close association with bone formation processes (i.e., maturation and mineralization of osteoid) has been established. The main bone-resorbing cell is the osteoclast. Interactions of specific cell-surface components form a secondary, extracellular lysosome with a ruffled border through which hydrogen ions are pumped to produce a low-pH local environment. This process is necessary to demineralize the bone, after which the matrix is removed by a series of acidic proteases that are released by the osteoclasts. One of these enzymes is tartrate-resistant acid phosphatase, which also is found in blood and is considered to be an osteoclast-specific product.

MARKERS OF BONE METABOLISM

Assays of biochemical markers of bone turnover are noninvasive and, when the results are applied and interpreted correctly, are very helpful tools in the assessment of metabolic bone disease. The various serum and urinary components used as markers of bone turnover include enzymes released by bone cells and peptides derived from the skeletal matrix during bone formation or bone resorption. For clinical purposes, bone biomarkers are usually classified according to the metabolic process they are considered to reflect; that is, bone formation or bone resorption (see Table 56-1 and Table 56-2). Some components, such as the serum amino-terminal procollagen type I propeptide and urinary hydroxyproline, are derived from anabolic and catabolic processes and are, therefore, influenced by the rate of bone formation and bone resorption. Other markers, such as the bone isoenzyme of alkaline phosphatase or the pyridinium cross-links of collagen, are more specific to individual metabolic processes. Most of the compounds that are used as markers of skeletal metabolism are not unique to bone, as they also occur in other tissues (see Table 56-1 and Table 56-2). Few or perhaps none of the available markers is absolutely specific for bone. Moreover, most serum and urinary indices are influenced by nonskeletal diseases, such as inflammatory conditions, malignancies, and chronic renal or hepatic failure. Changes in biochemical markers of bone metabolism are, therefore, not disease specific; abnormal results should always be interpreted within the context of the clinical picture. With these notions in mind, one may use bone markers for the following purposes: 1. 2. 3. 4. 5. To evaluate bone turnover in individual patients and disorders To predict future bone loss and hip fractures in larger cohorts To select therapy for individual patients To predict therapeutic response in individual patients To monitor therapeutic response and efficacy in individual patients

MARKERS OF BONE FORMATION All bone formation markers are products or enzymes released by active osteoblasts and are generally measured in serum or plasma. The most commonly used markers of bone formation are alkaline phosphatase, OC, and the propeptides of type I collagen. ALKALINE PHOSPHATASE The alkaline phosphatases form a family of isoenzymes that are found in a multitude of tissues, including bone (e.g., in osteo-blasts), liver, intestines, kidney, and placenta.1 The various iso-forms can be differentiated by their carbohydrate content. Several methods have been established to determine specifically the enzyme fraction derived from osteoblasts, including selective denaturation, inhibition or activation, electrophoretic separation, and immunologic quantification.2 In healthy adults, ~50% of the total serum alkaline phosphatase (TAP) activity is derived from osteoblasts, whereas the other half is usually of biliary-hepatic origin. Particularly in elderly patients, elevated serum TAP activities are more often the result of hepatic afflictions than of skeletal diseases (Table 56-4). Therefore, serum TAP levels should be used as an index of bone formation only if an impairment of liver and biliary function can be excluded. In contrast, the bone-specific isoenzyme of alkaline phosphatase (BAP) is located exclusively in the osteoblast membrane, from which it is released into the circulation on osteoblast activation. Compared to the total enzyme pool, serum BAP levels are clearly less affected by nonskeletal disorders and therefore more specific for changes in bone formation. In the clinical setting, however, the diagnostic sensitivity of serum BAP is not superior to that of serum TAP.2

TABLE 56-4. Factors Affecting the Serum Levels of Total Alkaline Phosphatase

Total alkaline phosphatase is the classic laboratory marker of skeletal activity in Paget disease of bone. Serum levels of the total and bone-specific enzymes are markedly increased when the disease is active, but are normal or only slightly elevated when patients have monostotic, mild polyostotic, or inactive disease. In certain cases, determination of serum BAP instead of TAP may help identify patients with very mild disease. Rather high levels of serum TAP or BAP are typically seen in patients with involvement of the skull, and sometimes in cases with sarcomatous degeneration of pagetic bone. Serial measurements of serum TAP are helpful to monitor a therapeutic response and to detect a recrudescence of disease activity. Primary hyperparathyroidism was previously often associated with significantly elevated levels of serum alkaline phos-phatase, indicating gross bone involvement. However, as the clinical profile of the disorder has evolved toward a predominance of cases with asymptomatic hypercalcemia, marked elevations in serum alkaline phosphatase are rarely seen (see Chap. 58).3 The expectation that serum BAP may be a more sensitive parameter of bone involvement in asymptomatic primary hyperparathyroidism has not yet been met. 2 For the use of serum TAP and/or BAP measurements in osteoporosis, see last section of this chapter. In osteomalacia, total and bone-specific enzyme activities are markedly elevated; the determination of serum alkaline phosphatase is often a clue to the diagnosis of this disorder. In renal osteodystrophy, elevated levels of serum alkaline phos-phatase may be indicative of progressive skeletal involvement and, particularly in patients on chronic hemodialysis, should prompt appropriate diagnostic and therapeutic measures. Serum levels of TAP and BAP may be elevated in primary and secondary skeletal malignancies. High serum activities are often found in osteoblastic bone metastases and after multiple fractures. In contrast, the alkaline phosphatase level is usually low in multiple myeloma. Because of its rather low sensitivity and specificity, serum alkaline phosphatase is not recommended as a tumor marker for clinical use. OSTEOCALCIN Osteocalcin (OC, or bone Gla protein), a small, GLArich peptide synthesized by osteoblasts, is one of the major noncollagenous proteins of the bone matrix.4,5 The molecule is found exclusively in mineralized tissues, and although its precise function is unknown, one obviously important property of OC is its high affinity for hydroxyapatite. Interestingly, OC-deficient knock-out mice have increased cortical and trabecular bone thickness, and mechanically more stable bones than wild-type mice.6 Therefore, during mineralization, OC may act via a negative feedback mechanism. After its secretion by activated osteoblasts, the major fraction of OC is incorporated into the extracellular matrix. However, 15% to 30% of the newly synthesized peptide is released into the general circulation, where it may be detected and quantified by immunoassay (Fig. 56-2). Although serum levels of circulating OC correlate well with the rate of bone formation,7 significant drawbacks are encountered in the practical use of this marker. Serum OC levels are strongly affected by the pronounced thermal instability of the molecule, the specificity of the antibodies, and factors such as hormonal status, renal function, age, and sex4,4a (Table 56-5 and Table 56-6).

FIGURE 56-2. Synthesis and metabolism of osteocalcin. Although the synthesis of osteocalcin is stimulated by vitamin D, g-carboxylation of glutamic acid (Glu) residues depends on vitamin K. Because of the affinity of g-carboxyglutamic acid (GLA) residues to hydroxylapatite, 70% to 80% of the newly synthesized peptide is bound to the mineralizing matrix of bone, and 20% to 30% enters the circulation as intact osteocalcin (iOC). Circulating osteocalcin is rapidly degraded into smaller fragments (fOC), some of which may also be derived from bone resorption. (Vit, Vitamin; N, amino terminus; C, carboxyl terminus.)

TABLE 56-5. Factors Affecting the Serum Levels of Osteocalcin

TABLE 56-6. Technical Characteristics of Bone Biomarkers

With the exception of Paget disease of bone, in which serum OC levels are often normal, most conditions with increased bone and mineral formation are characterized by elevated serum concentrations of OC. This is the case in primary hyperparathyroidism and hyperthyroidism, in which serum OC levels correlate with the extent of bone involvement.8 In healthy women, a transient, two-fold increase of serum OC is seen during early menopause, and individual values appear to correlate with the rate of bone loss in this population.9,10 In contrast, a wide range of values is observed in overt postmenopausal osteoporosis, which may be caused in part by the heterogeneity of the disease. A different situation may be present in the elderly and in patients with senile osteoporosis. In this population, the effect of aging and possibly a deficiency of vitamins K and D may lead to an impairment in the g-carboxylation of OC and ultimately to an increase in the proportion of partially undercar-boxylated circulating OC. Several studies now indicate that the proportion of undercarboxylated OC in serum may be an important determinant of femoral bone mineral density in elderly women, because high serum levels of undercarboxy-lated OC are associated with a low bone density at the hip and an increased risk of hip fractures.11 (See also the last section of this chapter.) Serum OC levels are often elevated in untreated osteomalacia, in which they correlate with histomorphometric indices of osteoid formation.12 However, serum OC levels are usually much less elevated than the corresponding serum TAP or BAP levels. In advanced renal failure (i.e., glomerular filtration rate of <2030 mL/minute/1.73 m2), the accumulation of intact OC and its various immunoreactive fragments may lead to very high serum levels, even in the absence of significant skeletal disease (see Table 56-6). OC levels are not affected by hemodialysis. Because of impaired renal clearance and increased bone turnover, serum OC usually is elevated in patients with renal osteodystrophy. Serum OC levels may be elevated in patients with metastatic bone disease, for whom it has been shown to be a useful marker of therapeutic responsiveness. Compared to the pyridinium cross-links, however, the diagnostic validity of serum OC for metastatic bone disease is low.13,14 Unlike in states of high bone turnover, serum OC levels are reduced in patients with glucocorticosteroid excess, such as Cushing disease or corticosteroid-induced osteoporosis. Also, low levels of serum OC have been correlated with poor survival in patients with stage III multiple myeloma. Serum free and urinary GLA are in part derived from the breakdown of OC and matrix Gla protein. However, many other components, such as certain clotting factors and plasma proteins, contain considerable amounts of GLA. Although increased levels have been found in Paget disease and osteoporosis, neither serum nor urinary GLA has achieved clinical relevance. PROCOLLAGEN TYPE I PROPEPTIDES During the process of collagen synthesis and before fibril formation, the N- and C-terminal procollagen propeptides are cleaved from the newly formed molecule and are released into the circulation. Because collagen and procollagen propeptides are generated stoichiometrically, levels of circulating procollagen propeptides are considered to reflect collagen neosynthesis (Fig. 56-3). Despite significant correlations between bone formation rates and the serum levels of procollagen type I propeptide, tissues other than bone contribute to the serum concentrations of these components15,16 (see Table 56-1). As the procollagen propeptides are cleared from the circulation, mainly through uptake by endothelial cells of the liver, their serum and urinary concentrations are strongly influenced by hepatic function (see Table 56-6).

FIGURE 56-3. Synthesis and metabolism of procollagen propeptides. After the synthesis and secretion of the procollagen molecule by osteo-blasts, the amino- (N-) and carboxy-terminal (C-) propeptides of type I collagen (e.g., PINP, PICP) are cleaved extracellularly from the procollagen molecule by specific proteases. The propeptides are considered to enter the circulation in an equimolar ratio to newly synthesized collagen and are regarded as specific markers of collagen biosynthesis. Similar mechanisms exist for other fibrillar collagens, such as type III collagen (e.g., PIIINP, PIIICP), which is typically found in soft tissues and mainly produced by fibroblasts.

Several polyclonal immunoassays have been developed to determine specifically the N- and C-terminal propeptides of procollagen type I (PINP and PICP, respectively) in body fluids.17,18 A number of different studies have shown good correlations between serum procollagen type I propeptide levels and the rate of bone formation or serum TAP activity.13,16 Increased serum levels of both procollagen propeptides are generally seen in conditions involving somatic growth or enhanced bone turnover. Serum levels of both propeptides are much higher in pre-pubertal children and adolescents than in healthy adults. In children with somatic growth disorders, serum PICP levels correlate with linear growth velocity before and during growth hormone treatment.19 Like most other biochemical markers of bone metabolism, serum propeptide levels are elevated in patients with Paget disease of bone; declining levels are seen after effective treatment. Normal menopause induces only slight changes in serum PICP and somewhat more pronounced elevations in PINP levels. Hormone-replacement therapy may be monitored by serum PICP or PINP, although the effect is usually less pronounced than with other markers of bone metabolism. MARKERS OF BONE RESORPTION Markers of bone resorption are either enzymes released by active osteoclasts (i.e., tartrate-resistant acid phosphatase), or degradation products of the extracellular bone matrix. The latter are mainly collagen-related compounds such as the pyridinium cross-links. Also, noncollagenous proteins such as bone sialoprotein are being investigated as markers of bone resorption. Unlike the formation markers, most resorption markers are routinely measured in urine, a medium that for several reasons is associated with a relatively high degree of imprecision. During the past few years, therefore, immunoassays for the measurement of type I collagen telopeptides have been developed (see later). The data so far

are promising, and these new serum assays most likely will replace the urine-based techniques in the near future. HYDROXYPYRIDINIUM CROSS-LINKS OF TYPE I COLLAGEN AND RELATED STRUCTURES The 3-hydroxypyridinium cross-links, pyridinoline (PYD) and deoxypyridinoline (DPD), are formed during extracellular maturation and are located in the telopeptide region of skeletal (type I) collagens. When mature collagen is broken down, both compounds and parts of the telopeptide region are released into the circulation and are ultimately excreted in the urine (Fig. 56-4). The pyridinium cross-links and the related telopeptides of type I collagen are currently considered the most specific and sensitive markers of bone resorption.20

FIGURE 56-4. Synthesis and metabolism of pyridinium cross-links of collagen. After their extracellular aggregation, the collagen molecules are covalently cross-linked by type- and tissue-specific compounds derived from lysine (see Fig. 56-1). These components are released from the extracellular matrix during bone resorption and are ultimately excreted in the urine.

The pattern of collagen cross-linking is tissue-specific in regard to cross-link type and concentration. Although PYD is found in cartilage, bone, tendons, and vascular tissues, DPD is found almost exclusively in bone and dentin.21,22 Moreover, because the formation of collagen cross-links is a time-dependent, posttranslational process, the source of the pyridinium components is restricted to mature collagens that have already been formed in the extracellular tissue matrix. Unlike urinary hydroxyproline, which may also be derived from the degradation of immature collagens (see later), the urinary excretion of PYD and DPD is not influenced by the breakdown of newly synthesized protein but instead reflects exclusively the degradation of mature collagens, as in bone resorption (Fig. 56-5).

FIGURE 56-5. Comparison of hydroxyproline (OHP) and hydroxypyridinium cross-links in urine. Urinary hydroxyproline is derived from the degradation of mature and newly synthesized collagens. Additional sources of hydroxyproline are dietary intake and partly collagenous proteins, such as elastin and the complement component, C1q. The hydroxypyridinium cross-links are formed later in the process of collagen maturation and are exclusively derived from the breakdown of mature matrix collagens.

Approximately 50% to 60% of the total urinary cross-link pool is peptide-bound, with the cross-link incorporated into collagen peptides of various lengths. The remaining fraction is present in (peptide) free form. Several assays have been developed for the determination of pyridinium cross-link compounds in urine, including high-performance liquid chromatography (HPLC) and immunoassay techniques. The HPLC method makes use of the natural fluorescence of free cross-link compounds and, after complete acid hydrolysis of the urine, measures the total amount of urinary cross-link components.23,24 In contrast, the various antibodies used in the commercially available immunoassays detect either the free cross-link components per se,25,26 and 27 or small collagen peptides containing a cross-linking structure.28,29 The direct immunoassays for free PYD and DPD in urine are based on the fact that the proportion of free to peptide-bound cross-links in urine remains stable over a wide spectrum of bone pathologies. Measurement of the free cross-link fraction, therefore, provides similar information about bone resorption as does the determination of the total cross-link pool. This assumption has been confirmed in a number of clinical studies,25,26 and 27 and assays for free pyridinium cross-links are currently a standard technique in many clinical laboratories throughout the world. Parallel developments have resulted in several immunoassays for the measurement of collagen type I telopeptides in urine and in serum (Fig. 56-6). These assays are based on the fact that the cross-linking of collagen always involves a specific region of the molecule, the amino-terminal (N-) or carboxy-terminal (C-) telopeptide. The various types of assays currently available or under investigation are characterized as follows:

FIGURE 56-6. Structure and proposed epitopes of the carboxy-terminal (CTx; upper panel) and amino-terminal (NTx; lower panel) cross-linked telopeptides of type I collagen. The C-terminal telopeptide of type I collagen (b-CTx) excreted in urine requires only one strand of the octapep-tide Glu-Lys-Ala-His-bAsp-Gly-Gly-Arg (EKAH-bD-GGR), which contains the cross-linking site of the C-terminal type I collagen telopep-tide (K). The transformation of aspartyl to isoaspartyl (bAsp) results from protein aging. In serum, the epitope is thought to be a cross-linked b-isomerized dipeptide of the C-terminal telopeptide region (oval insert). The N-terminal telopeptide of type I collagen (NTx) excreted in urine and serum requires the a-2 chain of type I collagen. The sequence is Gln-Tyr-Asp-Gly-Hyl-Gly-Val-Gly (QYDGKGVG), and is identical for both the urine- and serum-based assay.

C-terminal type I collagen telopeptide (ICTP) in serum29 : Antigenic determinant requires a trivalent cross-link; assay is not as sensitive to normal bone turnover as

urinary cross-link assays but is more sensitive to abnormal bone resorption as in multiple myeloma or metastatic bone disease. C-terminal telopeptide of type I collagen (CTx) in urine or serum30,31,32 and 33: Antigenic determinant is a synthetic octapeptide (EKAHDGGR) containing the cross-linking site of the C-terminal type I collagen telopeptide. Because of protein aging, the aspartyl may be present as isoaspartyl (BD),31 which is also the form recognized in the serum assay.33 N-terminal telopeptide of type I collagen (NTx) in urine and serum28 : Antigenic determinant requires the a-2 chain of type I collagen (QYDGKGVG); antibody reacts with several cross-linking components and is identical for urine- and serum-based assays.34 In healthy individuals, the excretion of PYD and DPD varies within a relatively narrow range, although a significant diurnal rhythm is observed, with highest values in the early morning hours and lowest values at night.35 Menopause is associated with a two- to three-fold increase.36 As these changes occur in healthy, nonosteoporotic women, they are likely to reflect the increase in bone turnover associated with estrogen withdrawal. Very similar results are seen with the immunoassays for either free DPD or the N- and C-terminal telopeptides (CTx, NTx). In active Paget disease of bone, all collagen-related resorption markers are greatly elevated; a rapid decrease in urine or serum concentrations is seen after treatment with bisphospho-nates, known inhibitors of osteoclast function.34,37 Approximately 60% of patients with mild primary hyper-parathyroidism have elevated urinary cross-link concentrations, revealing subclinical bone involvement. After successful parathyroidectomy, urinary levels of DPD fall quickly, preceding the changes in urinary hydroxyproline and serum alkaline phosphatase by ~6 months.38 Overt metastatic bone disease and multiple myeloma are generally associated with marked changes in collagen-related resorption markers.38a In fact, these markers may be used to diagnose and/or monitor the therapeutic response in individual patients13,14 (Fig. 56-7). Serum ICTP is a sensitive marker of osteolytic bone destruction in multiple myeloma and appears to predict disease progression independently from immunoglobulin production.39

FIGURE 56-7. Changes in urinary (upper panel) and serum (lower panel) levels of bone resorption markers in patients with hypercalcemia of malignancy after a single dose of intravenous pamidronate. Note that the magnitude of change is similar for both urinary and serum markers. (U-, urinary; S-, serum; tDPD, total deoxypyridinoline; fDPD, free immunoreactive deoxypyridinoline; NTx, amino-terminal telopeptide of type I collagen; CTx, carboxy-terminal telopeptide of type I collagen; BSP, bone sialoprotein.) (From Woitge HW, Oberwittler H, Farahmand I, et al. New serum markers of bone resorption. J Bone Miner Res 1999; 14: 792.)

For the use of cross-link compounds and related markers in osteoporosis, see the last section of this chapter. HYDROXYPROLINE AND HYDROXYLYSINE GLYCOSIDES Fibrillar collagens are rich in the amino acids hydroxyproline (OHP) and hydroxylysine (OHL), both of which are excreted in the urine after collagen degradation and are considered to be markers of bone resorption. Urinary OHP is a rather nonspe-cific and insensitive index of total body collagen turnover. First, OHP is formed intracellularly as one of the earliest posttransla-tional modifications of the collagen molecule. The amino acid is already present in newly synthesized collagens, and ~10% of the total OHP excreted in urine is likely to be derived from the degradation of immature collagens. Second, 90% to 95% of OHP released from various tissues is metabolized in the liver, and only a small fraction of the total circulating pool is actually excreted and measured in the urine. Third, urinary OHP levels are considerably influenced by dietary protein intake (especially meat) and by the metabolism of other collagenous proteins such as elastin and the complement component C1q (see Fig. 56-5). The amino acid is present in all fibrillar collagens and therefore in other soft tissues, especially in the skin40 (see Table 56-2). The urinary excretion of OHP is increased in states of physiologically high turnover, such as somatic growth and during early menopause, and in high-turnover osteopathies, such as Paget disease of bone, osteomalacia, renal osteodystrophy, hyperthyroidism, multiple myeloma, and metastatic bone disease. Because of its low sensitivity and specificity, urinary OHP is often normal in cases of asymptomatic or mild hyperparathy-roidism and in patients with postmenopausal osteoporosis. OHL may be present in two glycosylated forms: b-1-galactosyl-OHL and b-1,2-glucosyl-galactosyl-OHL. The type and concentration of OHL-glycoside vary among tissues, but the b-1-galactosyl-OHL appears to be a specific component of bone collagen. Unlike OHP, the lysine glycosides are not metabolized and may be more sensitive markers of collagen metabolism than OHP. The urinary levels of glycosylated OHL are not influenced by diet. The use of glycosylated OHL was long hampered by the lack of appropriate methods. With the development of an HPLC technique, this marker is now applicable to clinical studies.41 Data suggest that the determination of urinary OHL glycosides may be a helpful index of bone metabolism in women with postmenopausal osteoporosis and in growth-deficient children.42,43 TARTRATE-RESISTANT ACID PHOSPHATASE Human acid phosphatases form a heterogeneous group of at least five electrophoretically different isoenzymes that are found in prostate, various bloods cells, and osteoclasts. All acid phosphatases are inhibited by (+)-tartrate, except for the band 5 isoenzymes, of which the subtype 5b appears to be found in and secreted by osteoclasts. Serum or plasma tartrate-resistant acid phosphatase (TRAP) activity has been suggested as a marker of osteoclast activity in bone resorption. Circulating levels of TRAP may be determined by electrophoretic, spectrophotometric, and immunoassay procedures.44 Although most assays are easily performed, the pronounced thermal instability of the enzyme has so far precluded use of this marker in routine clinical application. Artificially high values are often seen in hemolytic samples, due to the release of erythrocytic TRAP (see Table 56-6). Serum TRAP activity is elevated in healthy children, during somatic growth, and in a variety of metabolic bone diseases associated with increased bone resorption, such as Paget disease of bone, hyperparathyroidism, multiple myeloma, and metastatic bone disease.45,46 Serum TRAP activity may be elevated in postmenopausal healthy and osteoporotic women, in whom TRAP activity is inversely correlated with bone mineral density.47 In patients on long-term maintenance hemodialysis, serum TRAP levels have been shown to reflect osteoclast activity. BONE SIALOPROTEIN Bone sialoprotein (BSP), a major synthetic product of active osteoblasts and odontoblasts, accounts for 10% of the noncollagenous matrix of bone. The expression of BSP is restricted to mineralized tissue (e.g., bone and dentin), and to the interface of calcifying cartilage. The intact molecule contains an Arg-Gly-Asp (RGD) integrin-recognition sequence, and plays an important role in the supramolecular organization of the extracellular matrix of mineralized tissues. Immunoassays have been developed for the determination of immunoreactive BSP in serum.48,49 Elevated levels of serum BSP were found in children and adolescents as well as in patients with high turnover osteopathies such as Paget disease of bone, primary hyperparathyroidism, active rheumatoid arthritis, metastatic bone disease, and multiple myeloma.50,50a In patients with the latter, high baseline serum BSP levels appear to predict poor survival. Furthermore, in patients with newly diagnosed primary breast cancer, baseline levels of serum BSP above 24 ng/mL are associated with a high risk of early bone metastases.51 Based on clinical data and the rapid reduction of serum BSP levels after intravenous bisphosphonate treatment (see Fig. 56-7), serum BSP levels are currently considered to reflect processes related to bone resorption or osteoclast activity.

USE OF BIOCHEMICAL MARKERS OF BONE TURNOVER IN OSTEOPOROSIS

Since the previous edition of this textbook, major advances have been made with regard to the application of biomarkers of bone turnover in osteoporosis. In particular, a number of studies have shown that bone turnover, as assessed by markers of bone formation or of bone resorption, are predictive of future bone loss and fracture risk. Thus, individuals with high rates of bone turnover have been shown to lose bone at a much faster rate than subjects with normal or low bone turnover52,53 (Fig. 56-8). Furthermore, vertebral fracture rates seem to increase as a direct function of either increased bone turnover or decreased vertebral bone mineral density (BMD).54 A prospective, population-based study of elderly women (older than 75 years) demonstrated that an increase in the urinary excretion of total or free DPD was associated with an increased risk of hip fracture.55 Later analyses of data from the same study revealed that low serum intact OC levels were also associated with an increased risk of hip fracture. Similar results have been reported for the urinary telopeptide markers (b-CTx).56 Importantly, not only were the relative risks similar (as defined by either BMD or marker measurements), but also the combined measurement of hip bone density and of bone markers predicted future hip fractures better than the determination of either bone density or markers alone. This means that in older postmenopausal women, the statistical risk of future hip fracture is highest in subjects with a combination of low bone mass and high rates of bone turnover. In contrast, markers of bone turnover are less useful in the primary diagnosis of vertebral osteoporosis. Usually, a broad overlap in marker levels is seen between osteoporotic and healthy populations, and most markers have insufficient diagnostic power to distinguish between vertebral osteopenia and osteoporosis.36 Markers of bone turnover are, therefore, not useful in the diagnosis of osteoporosis.

FIGURE 56-8. Probability of rapid bone loss as a function of biochemical marker levels as determined in a retrospective study over 14 years. The odds of rapid bone loss increased by 1.8 to 2.0 times for each 1.0 standard deviation (SD) increase of the marker. For example, for serum bone-specific alkaline phosphatase (BAP) (filled bars), the probability of rapid bone loss was 80% when the marker level was 2 SD above the group mean, but only 20% when the level was 2 SD below the mean (n = 200). Note that all markers yielded similar results. (2nd bar from left, osteocalcin [OC]; 3rd bar from left, pyridinoline [PYD]; 4th bar from left, deoxypyridinoline [DPD].) (From Ross PD, Knowlton W. Rapid bone loss is associated with increased levels of biochemical markers. J Bone Miner Res 1998; 13:297.)

An emerging application of biochemical bone markers is their use in therapeutic decision making. This is based on the observation that osteoporotic patients with high bone turnover benefit more from antiresorptive treatment than do patients with low bone turnover.52,57,58 Depending on the baseline rate of bone turnover, both an increase in vertebral BMD and a decrease in bone turnover are equally effective in reducing vertebral fractures in osteoporotic women.54 These results may provide a rationale for the use of bone markers in the selection of therapy and in the prediction of therapeutic response. So far, however, this relationship has been shown only for estrogen or calcitonin treatments, not for treatment with other antiresorptives such as the bisphosphonates or vitamin D. The clinically most relevant application of bone biomarkers is their use in the monitoring of ongoing therapy. Although the efficacy of therapy is assessed primarily by a reduction in incidence of fracture, and secondarily by an increase in bone density, these effects usually occur slowly and within several years of treatment. In contrast, some markers of bone metabolism change as early as 72 hours after intravenous bisphosphonate therapy. Most other interventions result in significant changes within 6 weeks of implementation50,59,60 and 61 (see Fig. 56-7). One should bear in mind, however, that most biochemical markers are characterized by a high degree of variability, and that, depending on the type of marker used, posttherapeutic changes of 30% to 60% are required to reach significance.62 Newer studies indicate that the magnitude of the change in bone marker levels at 3, 4, or 6 months of treatment with hormone-replacement therapy or bisphosphonates correlates with the increase in BMD after 12 or 24 months of treatment.52 Serial measurements of bone markers soon after implementation of therapy may therefore be helpful in deciding whether a patient has responded to a specific treatment regimen. CHAPTER REFERENCES
1. 2. 3. 4. 4a. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. Harri, H. The human alkaline phosphatases: what we know and what we don't know. Clin Chim Acta 1989; 186:133. Woitge HW, Seibel MJ, Ziegler R. Comparison of total and bone specific alkaline phosphatase in skeletal and non-skeletal diseases. Clin Chem 1996; 42:1796. Silverberg SJ, Shane E, DeLaCruz L, et al. Skeletal disease in primary hyperparathyroidism. J Bone Miner Res 1989; 4:283. Lian JB, Gundberg CM. Osteocalcin. Biochemical considerations and clinical applications. Clin Orthop Rel Res 1988; 226:267. Kakonen SM, Hellman J, Karp M, et al. Development and evaluation of three immunofluorometric assays that measure different forms of osteocalcin in serum. Clin Chem 2000; 46:332. Power MJ, Fottrell PF. Osteocalcin: diagnostic methods and clinical applications. Crit Rev Clin Lab Sci 1991; 28:287. Ducy P, Desbois C, Boycem B, et al. Increased bone formation in osteocal-cin-deficient mice. Nature 1996; 382:448. Malluche HH, Faugere MC, Fant P, Price PA. Plasma levels of bone Gla-protein reflect bone formation in patients on chronic maintenance dialysis. Kidney Int 1984; 26:869. Charles P, Mosekilde L, Jensen FT. Primary hyperparathyroidism: evaluated by 47 calcium kinetics, calcium balance, and serum boneGla-protein. Eur J Clin Invest 1986; 16:277. Epstein S, Poser J, McClintock R, et al. Differences in serum bone Gla protein with age and sex. Lancet 1984; 1:307. Chen JT, Hosoda K, Hasumi K, et al. Serum N-terminal osteocalcin is a good indicator for estimating responders to hormone replacement therapy in postmenopausal women. J Bone Miner Res 1996; 11:1784. Vergnaud P, Garnero P, Meunier P, et al. Undercarboxylated osteocalcin measured with a specific immunoassay predicts hip fracture in elderly women. The EPIDOS study. J Clin Endocrinol Metab 1998; 82:719. Demiaux B, Arlot ME, Chapuy MC, et al. Serum osteocalcin is increased in patients with osteomalacia: correlations with biochemical and histomor-phometric findings. J Clin Endocrinol Metab 1992; 74:1146. Pecherstorfer M, Zimmer-Roth I, Schilling T, et al. The diagnostic value of urinary pyridinium cross-links of collagen, serum total alkaline phos-phatase, and urinary calcium excretion in neoplastic bone disease. J Clin Endocrinol Metab 1995; 80:97. Pecherstorfer M, Seibel MJ, Woitge H, et al. Urinary pyridinium crosslinks in multiple myeloma, MGUS and osteoporosis. Blood 1997; 90:3743. Parfitt AM, Simon LS, Villanueva AR, Krane SM. Procollagen type I car-boxyterminal extension propeptide in serum as a marker of collagen biosynthesis in bone. J Bone Miner Res 1987; 5:427. Eriksen EF, Charles P, Meisen F, et al. Serum markers of type 1 collagen formation and degradation in metabolic bone disease: correlation with bone histomorphometry. J Bone Miner Res 1993; 8:127. Melkko J, Niemi S, Risteli L, Risteli J. Radioimmunoassay of the carboxy-terminal propeptide of human type I procollagen. Clin Chem 1990; 36:1328. Pedersen BJ, Bonde M. Purification of human procollagen type I carboxyl-terminal propeptide cleaved as in vivo from procollagen and used to calibrate a radioimmunoassay of the propeptide. Clin Chem 1994; 40:811. Trivedi P, Risteli J, Risteli L, et al. Serum concentrations of the type I and III procollagen propeptides as biochemical markers of growth velocity in healthy infants and children with growth disorders. Pediatr Res 1991; 30:276. Seibel MJ, Baylink F, Farley H, et al. Basic science and clinical application of bone biomarkers. Exp Clin Endocrinol Diabetes 1997; 105:125. Robins SP, Duncan A. Pyridinium crosslinks of bone collagen and their location in peptides isolated from rat femur. Biochim Biophys Acta 1987; 914:233. Eyre D, Dickson IR, Van Ness K. Collagen crosslinking in human bone and articular cartilage. Biochem J 1988; 252:495. Black D, Duncan A, Robins SP. Quantitative analysis of the pyridinium crosslinks of collagen in urine using ion-paired reverse-phase high-performance liquid chromatography. Anal Biochem 1988; 169:197. Pratt DA, Daniloff Y, Duncan A, Robins SP. Automated analysis of the pyridinium crosslinks of collagen in tissue and urine using solid-phase extraction and reversed-phase high-performance liquid chromatography. Ann Biochem 1992; 207:168. Seyedin S, Kung VT, Daniloff YN, et al. Immunoassay for urinary pyridin-oline: the new marker of bone resorption. J Bone Miner Res 1993; 8:635. Seibel MJ, Woitge HW, Scheidt-Nave C, et al. Urinary hydroxypyridinium crosslinks of collagen in population-based screening for overt vertebral osteoporosis: results of a pilot study. J Bone Miner Res 1994; 9:1433. Robins SP, Woitge H, Hesley R, et al. A direct enzyme-linked immunoassay for urinary deoxypyridinoline as a specific marker for measuring bone resorption. J Bone Miner Res 1994; 9:1643. Hansen DA, Weis MA, Bollen AM, et al. A specific immunoassay for monitoring human bone resorption: quantitation of type I collagen cross-linked N-telopeptides in urine. J Bone Miner Res 1992; 7:1251. Risteli J, Elomaa I, Niemi S, Novamo A, Risteli L. Radioimmunoassay for the pyridinoline crosslinked carboxyterminal telopeptide of type I collagen: a new serum marker of bone collagen degradation. Clin Chem 1993; 39:655. Bonde MQP, Fledelius C, Riis BJ, Christiansen C. Immunoassay for quantifying type I degradation products in urine evaluated. Clin Chem 1994; 40:2022.

31. Fledelius C, Johnsen AH, Cloos PAC, et al. Characterization of urinary degradation products derived from type I collagen. Identification of a beta-isomerized Asp-Gly sequence within the C-terminal telopeptide (alpha1) region. J Biol Chem 1997; 272:9755. 32. Bonde M, Fledelius C, Qvist P, Christiansen C. Coated-tube radioimmu-noassay for C-telopeptides of type I collagen to assess bone resorption. Clin Chem 1996; 42:1639. 33. Bonde M, Garnero P, Fledelius C, et al. Measurement of bone degradation products in serum using antibodies reactive with an isomerized form of an 8 amino acid sequence of the C-telopeptide of type I collagen. J Bone Miner Res 1997; 12:1028. 34. Woitge HW, Oberwittler H, Farahmand I, et al. New serum markers of bone resorption. J Bone Miner Res 1999; 14:792. 35. Seibel MJ, Duncan A, Robins SP. Urinary pyridinium crosslinks of collagen provide indices of cartilage and bone involvement in arthritic diseases. J Rheumatol 1989; 16:970. 36. Seibel MJ, Cosman V, Shen V, et al. Urinary hydroxypyridinium crosslinks of collagen as markers of bone resorption and estrogen efficacy in post-menopausal osteoporosis. J Bone Miner Res 1993; 8:881. 37. Robins SP, Black D, Paterson RD, et al. Evaluation of urinary hydroxypyridinium crosslink measurement as resorption markers in metabolic bone disease. Eur J Clin Invest 1991; 21:310. 38. Seibel MJ, Gartenberg F, Silverberg S, et al. Urinary hydroxypyridinium crosslinks of collagen in primary hyperparathyroidism. J Clin Endocrinol Metab 1992; 74:481. 38a. Demers LM, Costa L, Lipton A. Biochemical markers and skeletal metastases. Cancer 2000; 88(2 Suppl):2919. 39. Elomaa I, Virkkunen P, Risteli L, Risteli J. Serum concentrations of the crosslinked carboxyterminal telopeptide of type I collagen (ICTP) is a useful prognostic indicator in multiple myeloma. Br J Cancer 1992; 66:337. 40. Kivirikko KI. Urinary excretion of hydroxyproline in health and disease. Int Rev Connect Tissue Res 1970; 5:93. 41. Moro L, Modricky C, Stagni N, et al. High performance liquid chromatographic analysis of urinary hydroxylysine glycosides as indicators of collagen turnover. Analyst 1984; 109:1621. 42. Moro L, Mucelli RS, Gazzarrini C, et al. Urinary b-1-galactosyl-O-hydroxylysine (GH) as a marker of collagen turnover in bone. Calcif Tissue Int 1988; 42:87. 43. Rauch F, Schonau E, Woitge E, et al. Urinary excretion of hydroxypyridinium cross-links of collagen reflects skeletal growth velocity in normal children. Exp Clin Endocrinol 1994; 102:104. 44. Lau KH, Onishi T, Wergedal JE, et al. Characterization and assay of tartrate-resistant acid phosphatase activity in serum: potential use to assess bone resorption. Clin Chem 1987; 33:458. 45. Stepan JJ, Silinkova E, Havranek T, et al. Relationship of plasma tartrate-resistant acid phosphatase to the bone isoenzyme of serum alkaline phos-phatase in hyperparathyroidism. Clin Chim Acta 1983; 133:189. 46. Colson F, Berny C, Tebib J, et al. Assessment of bone resorption by measuring tartrate-resistant acid phosphatase (TAcP) activity in serum. In: Chris-tiansen C, Overgaard K, eds. Osteoporosis. Copenhagen: Osteopress, 1990:621. 47. De la Piedra C, Torres R, Rapado A, et al. Serum tartrate-resistant acid phosphatase and bone mineral content in postmenopausal osteoporosis. Calcif Tissue Int 1989; 45:58. 48. Saxne T, Zunino L, Heinegard D. Increased release of bone sialoprotein into synovial fluid reflects tissue destruction in rheumatoid arthritis. Arthritis Rheum 1995; 38:82. 49. Karmatschek M, Woitge HW, Armbruster FP, et al. Improved purification of human bone sialoprotein and development of a homologous radioim-munoassay. Clin Chem 1997; 43:2076. 50. Seibel MJ, Woitge HW, Pecherstorfer M, et al. Serum immunoreactive bone sialoprotein as a new marker of bone turnover in metabolic and malignant bone disease. J Clin Endocrinol Metab 1996; 81:3289. 50a. Ogata Y, Nakao S, Kim RH, et al. Parathyroid hormone regulation of bone sialoprotein (BSP) gene transcription is mediated through a pituitary-specific transcription factor-1 (Pit-1) motif in the rat BSP gene promoter. Matrix Biol 2000; 19:395. 51. Diel IJ, Solomayer EF, Siebel MJ, et al. Serum bone sialoprotein in patients with primary breast cancer is a prognostic marker for subsequent bone metastases. Clin Cancer Res 1999; 5:3914. 52. Chesnut CH III, Bell NH, Clark GS, et al. Hormone replacement therapy in postmenopausal woman: urinary N-telopeptide of type I collagen monitors therapeutic effect and predicts response of bone mineral density. Am J Med 1997; 102:29. 53. Ross PD, Knowlton W. Rapid bone loss is associated with increased levels of biochemical markers. J Bone Miner Res 1998; 13:297. 54. Riggs BL, Melton LJ III, O'Fallon WM. Drug therapy for vertebral fractures in osteoporosis: evidence that decreases in bone turnover and increases in bone mass both determine antifracture efficacy. Bone 1996; 18:197S. 55. Van Daele PL, Seibel MJ, Burger H, et al. Case control analysis of bone resorption markers, disability and hip fracture risk: the Rotterdam study. BMJ 1996; 312:482. 56. Garnero P, Hausherr E, Chapuy MC, et al. Markers of bone resorption predict hip fractures in elderly women. The EPIDOS study. J Bone Mineral Res 1996; 11(10):1531. 57. Civitelli R, Gonnelli S, Zacchei F, et al. Bone turnover in postmenopausal osteoporosis. J Clin Invest 1988; 82:1268. 58. Nielsen NM, Von der Recke P, Hansen MA, et al. Estimation of the effect of salmon calcitonin in established osteoporosis by biochemical bone markers. Calcif Tissue Int 1994; 55:8. 59. Kraenzlin ME, Seibel MJ, Trechsel U, et al. The effect of intranasal salmon calcitonin on postmenopausal bone turnover: evidence for maximal effect after 8 weeks of continuous treatment. Calcif Tissue Int 1996; 58:216. 60. Raisz LG, Wiita B, Artis A, et al. Comparison of the effects of estrogen alone and estrogen plus androgen on biochemical markers of bone formation and resorption in postmenopausal women. J Clin Endocrinol Metab 1996; 81:37. 61. Heikkinen AM, Parvianen M, Niskanen L, et al. Biochemical bone markers and bone mineral density during postmenopausal hormone replacement therapy with and without vitamin D 3: a prospective, controlled, randomized study. J Clin Endocrinol Metab 1997; 82:2476. 62. Hannon R, Blumsohn A, Naylor K, Eastell R. Response of biochemical markers of bone turnover to hormone replacement therapy: impact of biological variability. J Bone Miner Res 1998; 13:1124.

CHAPTER 57 CLINICAL APPLICATION OF BONE MINERAL DENSITY MEASUREMENTS Principles and Practice of Endocrinology and Metabolism

CHAPTER 57 CLINICAL APPLICATION OF BONE MINERAL DENSITY MEASUREMENTS

PAUL D. MILLER, ABBY ERICKSON, AND CAROL ZAPALOWSKI Diagnosis of Osteoporosis Using Bone Densitometry: The World Health Organization Criteria Why are Bone Mass Measurements Performed? The Diagnosis of Osteopenia and Osteoporosis The Prediction of Fracture Risk Serial Assessments of Bone Mass Conclusions Chapter References

Bone densitometry is accepted as a useful quantitative measurement for assessing skeletal status and predicting the risk of fragility fractures.1 Low bone mineral density (BMD) is the most important risk factor for fracture. BMD testing is an objective measurement, and extensive data demonstrate that bone mass and future fracture risk are inversely correlated. Low bone mass predicts fracture in the same way that high cholesterol and high blood pressure predict myocardial infarction and stroke, respectively.2 Historical risk factors cannot identify individual patients with low bone mass with adequate certainty.3 This does not discount the importance of assessing risk factors other than BMD for fracture. These risk factors add valuable information required for individual patient management decisions. Also, some risk factors can be modified to help reduce fracture risk.4 This is particularly true in the perimenopausal population and in patients with secondary conditions associated with bone loss. Measurement techniques for quantitating BMD that use radioisotopes as their photon energy source have been replaced by techniques that use dual-energy x-ray sources. Another technique, currently under evaluation, is quantitative ultrasound.4a Dual-energy x-ray absorptiometry (DXA) can be performed at several skeletal sites (e.g., hip, spine, finger, wrist, or heel) and does not require a water bath for the equalization of the soft tissue surrounding bone. DXA uses a dual-energy x-ray system to separate bone from soft tissue. The radiologic and technical aspects for quantitating bone mass and the principles of bone mass measurement physics are well documented.5,6 Two technical terms, accuracy and precision, are important for the understanding of bone mass measurements. Accuracy relates to the ability of a bone mass measurement device to measure the actual amount of bone tissue per unit volume present. It is determined using in situ and/or cadaveric bone samples that are measuredusing the device being testedin grams per centimeter (g/cm), with comparison of these values to the ashed-weight in grams per centimeter of the identical area of bone. Accuracy studies have been performed on most technical methods, and the accuracy ranges from 90% to 99%. Therefore, the accuracy error is 1% to 10%. The lower the accuracy error, the greater the accuracy of the measurement. The greater the accuracy of the device, the closer the measurement is to the actual bone mineral content per unit volume of the bone. Accuracy errors of 1% have very little effect on patient diagnosis if the number of standard deviations from a young normal mean is used for diagnostic criteria. Accuracy errors of 10%, on the other hand, may profoundly affect patient diagnosis (Fig. 57-1). Fortunately, the accuracy errors of the central DXA machines are ~6% and those of peripheral devices are ~3%. Accuracy errors are generally lower in peripheral devices because of decreased soft tissue at peripheral sites.

FIGURE 57-1. The effect of accuracy error on the distribution of T score. A 1% accuracy error has little effect on diagnosis, whereas a 10% accuracy error can change the T score in an individual patient from +1 to l SD. (BMD, bone minderal density.)

Precision, or reproducibility, pertains to the measurement error introduced by repetitive or serial tests. The precision error of the BMD testing devices is very low (1% to 2%). The lower the precision error, the better the precision of a device. The more precise the device, the shorter period of time required for a significant change in BMD to be detected. The metabolic bone community suggests that a serial measurement fall within the 95% confidence interval to be significant, and, therefore, represents a BMD change rather than a change caused by simply measurement error. Hence, for a serial BMD change to be significant, it must be at least 2.8 times the precision error of the bone mass measurement technique. Each densitometry facility should perform daily quality control (in vitro) using a phantom as well as an in vivo precision study for each skeletal site in normal subjects and in subjects with low bone mass to determine the precision error of their individual machine and technologist.

DIAGNOSIS OF OSTEOPOROSIS USING BONE DENSITOMETRY: THE WORLD HEALTH ORGANIZATION CRITERIA
To allow bone densitometry to be used for the identification of asymptomatic individuals at risk for fracture, a paradigm shift in the definition of osteoporosis had to occur. The accepted definition of osteoporosis is a systemic skeletal disease characterized by low bone mass and microarchitectural deterioration of bone tissue, with a consequent increase in bone fragility and susceptibility to fracture risk7 (see Chap. 64). Although this accepted definition includes the terms low BMD, systemic disease, microarchitectural deterioration, and increased bone fragility, only BMD can presently be objectively measured in clinical practice. Before this consensus statement defined osteoporosis, a diagnosis was made on the basis of the presence of a fragility fracture, which usually occurs only in the advanced osteoporotic state. Recently, osteoporosis has been defined on the basis of a reduced level of BMD. This new definition was implemented by the World Health Organization (WHO) Consensus Development Conference on Osteoporosis.2 The goal of the WHO Working Group was to establish a relationship between bone mineral content (BMC) per centimeter measured, at that time, predominantly at the radius and hip, and the prevalence of fractures in elderly postmenopausal white women. It was determined that those women in the lowest quintile of BMD had the greatest prevalence of global fractures. The lowest BMD quintile, when corrected for different instrument calibrations, corresponded to 2.5 standard deviations (SD) below the mean of the reference population. Therefore, the new definition of osteoporosis became a bone mass that is at least 2.5 SD below the mean of a young normal reference population. One major clinical justification for changing the diagnostic criteria for osteoporosis from one of prevalent fragility fractures to one of low BMD are data that established a higher risk for a second fracture once a first fracture has occurred.8 In addition, the combination of low bone mass and a vertebral fracture increased the risk of a hip fracture. This risk is far greater than the risk of a first fracture in elderly individuals who have low bone mass alone.8 The logical corollary of these data is the need to identify individuals with low bone mass before the first fracture. The WHO criteria for the diagnosis of osteopenia and osteoporosis (Table 57-1) are based on a patient's comparison to peak adult bone mass (PABM) and use standardized scores (T score). The WHO Working Group chose to categorize individuals using the number of SD a patient's bone mass is below the mean bone mass of a young normal reference population. Because fracture risk is a gradient, increasing with declining levels of bone mass,9 the WHO created a second diagnostic category of osteopenia (low bone mass, T score >2.5 but <1.0 SD) to alert the clinician that individuals with smaller reductions in bone mass merit attention, particularly if they are postmenopausal or have secondary conditions associated with bone loss. Data suggest that postmenopausal women who are not receiving hormone-replacement therapy (HRT) will predictably lose bone.10,11 As the emphasis on skeletal health shifts from treatment to prevention, the diagnostic category of osteopenia may become increasingly important. Postmenopausal women not receiving therapy to prevent bone loss may unknowingly continue to lose bone, thereby progressively increasing their fracture risk as they age. Hence, postmenopausal women identified as osteopenic may be targeted for prevention strategies to preserve their skeletal mass. Individuals with osteopenia may also experience fragility fractures, particularly when risk factors for fracture are present.12 Such factors include

increased rates of bone turnover, advancing age, increased likelihood of falling, etc. It, therefore, is not surprising that some patients with small reductions in BMD and the presence of other risk factors may have fragility fractures. For this reason, the National Osteoporosis Foundation (NOF) clinical guidelines recommend treating all postmenopausal women with T scores <2.0 SD without risk factors and <1.5 SD with one or more risk factors.13

TABLE 57-1. World Health Organization Criteria for the Diagnosis of Osteopenia and Osteoporosis

Like all diagnostic criteria, there are strengths and limitations to the WHO criteria for the diagnosis of osteoporosis. These are listed in Table 57-2. The WHO criteria provide a practitioner with objective values to be used for the diagnosis of low bone mass, akin to blood pressure for the diagnosis of hypertension. Practitioners have objective criteria that initiate the cognitive process of assessment of low bone mass, leading to intervention. In addition, the WHO criteria stress the importance of making a diagnosis of low bone mass (osteopenia) or osteoporosis before the first fracture occurs.

TABLE 57-2. World Health Organization Criteria for the Diagnosis of Osteopenia/Osteoporosis: Pros and Cons

The limitations presented in Table 57-2 appear to outnumber the strengths. However, this belies the beneficial impact the WHO criteria have had on the prevention and management of osteoporosis. The WHO criteria have facilitated the widespread measurement of bone mass for the identification of individuals at risk for fracture. One of the limitations of the WHO criteria is its restricted data for its application to nonwhite women. The WHO criteria are based on BMD and prevalence of low bone mass and fracture prevalence in postmenopausal white women. Similar data in men and women of other racial groups are not well established. However, some studies have shown that fracture rates in white men per standard deviation (SD) reduction in BMD may parallel that seen in white women14,15 and 16; however, one study suggests that this relationship may not be the same for men.17 Nonetheless, it may be reasonable to use the WHO criteria to approximate fracture rates in elderly white men as well. As prospective fracture data are obtained for other races, the applicability of the WHO criteria can be judged more accurately. The WHO criteria were never intended to be applied to healthy premenopausal women of any race. BMD data for the total hip and femoral neck have been analyzed in men and women of different races with T scores derived from a common database for the hip (NHANES III).18 These data have helped determine the prevalence of osteoporosis at the hip in men and women of different races, using the WHO criteria.18,19 The incorporation of the common hip reference database (NHANES III) into the central dual-energy x-ray absorptiometry (DXA) equipment produced by the three major manufacturers has eliminated the potential for machine-specific misdiagnoses of osteoporosis or osteopenia at the hip.20 Regardless of the database issues as they relate to WHO classifications, low bone mass (g/cm) in any elderly individual is still the most important factor in the determination of fragility fracture risk. Manufacturers' devices that measure skeletal sites other than the hip still use nonuniform young normal reference databases. This is a serious issue in the field of bone densitometry, because one of the basic principles of statistics is that each sample derived from a given population yields a different mean and potentially a different SD. The young normal mean and SD from that mean form the basis of the calculation of the T score (T score = [measured BMD in g/cm2 PABM mean in g/cm2]/PABM SD in g/cm2) or for ultrasound devices (T score = [mean young normal broadband ultrasound attenuation (BUA) or speed of sound (SOS) measured BUA or SOS]/young normal SD of BUA or SOS). Therefore, an individual may be classified quite differently if different young normal sample populations are used to create the reference database to which the patient is being compared.21,22 Dissimilar T scores in the same patient, measured on different devices, is also in part a result of different accuracy errors of various technologies and differing age-dependent rates of bone loss from different skeletal sites.23 These present clinical dilemmas, because the same patient may be classified differently by various bone mass measurement devices. The T score dissimilarities related to the young normal database could be resolved by the creation of a universal reference database for each skeletal site and technique used. In support of this proposal, data have shown that bone mass measured by various heel devices (ultrasound and DXA) and hip DXA result in similar T scores in post-menopausal women when calculated using a universal young normal reference population.24 A universal database represents, at least in part, the long-term solution to this clinical dilemma. A short-term solution has been suggested by a committee of the National Osteoporosis Foundation (NOF) and the International Society for Clinical Densitometry (ISCD). A T score equivalent for different manufacturers' devices can be calculated, based on the prevalence of low bone mass as determined by each device or risk if device-specific fracture data are available. A 50% prevalence in 60- to 69-year-old women is set by any device equal to the established 50% prevalence at the femoral neck established by NHANES III, which occurs at a T score of 1.5 SD. This 50% prevalence cutoff of 1.5 SD seems reasonable, because it captures 60% of the women who will ultimately have a hip or spine fracture; it corresponds to the NOF treatment threshold for postmenopausal women with one or more risk factors (1.5 SD); and it equals global fracture risk prediction by all devices (RR 1.5/SD reduction).25 Thus, the equivalent T score might be 0.8 SD for one device and 1.2 SD for another device. This short-term (T-score equivalent) solution would reduce the number of patient misclassifications resulting from the use of different devices, until a universal database is completed. The NOF/ISCD Committee is also examining the potential of calculating the T-score equivalence of different devices based on risk rather than prevalence. Both methods may be incorporated into the final T-score equivalence calculations, because prevalence approaches allow the clinician to make a diagnosis, and risk approaches assist with the prediction of the disease outcome of fragility fracture. Equating a level of risk to a level of prevalence may also be appealing. A final solution to patient misclassification is to abandon the T score as a diagnostic or intervention tool and instead use the absolute BMD value as it relates to absolute fracture prediction for patient classification. This would require comparing a patient's BMD to instrument-specific fracture data for the calculation of fracture risk. This could also minimize discrepancies observed in individual patients measured on different equipment. Absolute fracture risk may be similar regardless of the technique used.26 The WHO-defined criteria for osteoporosis are made on the basis of the T score (based on young normal bone mass), rather than the Z score (based on age-matched bone mass). Although bone mass declines with advancing age,27 it is no longer logical to assume that bone loss is inevitable. If only those individuals who lose more bone than predicted for their age are termed diseased, then many individuals with increased risk of fracture will be classified as normal and remain untreated. If the prevalence of osteoporosis were defined using age-matched BMD data, there would be no increase in prevalence with advancing age. This is illogical, because the

number of fragility fractures increase with age and fragility fracture is the result of osteoporosis. The use of age-matched Z scores would result in the underdiagnosis of osteoporosis.2 This would damage the credibility of bone densitometry by incorrectly describing elderly patients with fragile skeletons or possibly prevalent fragility fractures as normal for age. Even if 70% of white women older than age 80 years are osteoporotic using the WHO criteria, this may not be an overestimate, because 50% of these women will experience a fragility fracture if untreated.28 Use of the T score for diagnosis recognizes the true magnitude of the osteoporotic problem. Age-matched data are appropriate, however, for comparisons in the adolescent population, before peak adult bone mass is achieved.29 Once peak bone mass is achieved, the goal is to maintain that level of bone mass for the remainder of the patient's life. After peak adult bone mass has been achieved, age-matched Z scores that are < 2.0 SD may indicate the presence of a metabolic process causing bone loss or impeding bone acquisition in addition to aging and/or estrogen deficiency in the elderly population. Aging bone has altered quality and microarchitectural deterioration that cannot presently be clinically measured and may increase fracture risk in the elderly. These unmeasurable bone quality changes in the elderly explain higher fracture risks at equal BMD levels as age increases. As mentioned previously, the WHO criteria were not meant to be applied to healthy, estrogen-replete premenopausal women. Peak adult bone mass is normally distributed. Hence, a healthy, estrogen-replete premenopausal woman with a T score of 1.8 SD may not have lost bone. Because her level falls within the normal bell-shaped curve, this T score could simply represent a low peak adult bone mass. Furthermore, healthy premenopausal women with low peak adult bone mass may not have any greater risk of fragility fracture than age-matched women with normal bone mass. There are scant data in younger women that relate fracture risk to BMD, because premenopausal women rarely fracture and early postmenopausal women fracture only infrequently. Cross-sectional29 and prospective fracture studies in premenopausal women have small numbers with wide confidence intervals.30 Additionally, whereas there are some data to suggest that a subset of healthy, estrogen-replete premenopausal women may lose bone mass before the menopause, the majority of prospective longitudinal studies indicate that premenopausal women do not lose bone mass until the perimenopausal period.31,32 Fracture risk has been shown to be much less for a 55-year-old woman than for a 75-year-old woman at equivalent levels of BMD (Fig. 57-2).33 The inappropriate use of the WHO diagnostic criteria may cause undue fear regarding fracture risk and the inappropriate initiation of pharmacologic therapy in premenopausal women in the absence of data on efficacy and safety. The majority of longitudinal data suggest that healthy, estrogen-replete premenopausal women are not losing bone mass.31 Healthy, estrogen-replete premenopausal women who seek advice after the discovery of low bone mass require a basic assessment to exclude secondary causes of potential bone loss, possibly a biomarker of bone turnover, conservative prevention advice, and a repeat bone mass measurement in 2 years to ensure that bone loss is not occurring. Most importantly, they need prompt protection of their skeleton at the menopause.32 There are firm data that after the age of 50 years, postmenopausal women who are losing bone mass have fracture rates that are higher than those for age-matched postmenopausal women who have the same baseline BMD that, however, remains stable.34

FIGURE 57-2. Increasing age increases fracture risk at equivalent levels of bone mass.

WHY ARE BONE MASS MEASUREMENTS PERFORMED?

There are three clinical applications for bone mass measurement. Each has a distinct value in clinical decision making. The three applications are 1. Diagnosis of osteopenia or osteoporosis (see Chap. 64). 2. Fracture risk prediction. 3. Serial monitoring to measure response to interventions, diseases, or medications that affect bone. Clinicians have at their disposal various devices that measure bone mineral density. These devices are characterized as central or peripheral. Central devices measure the spine or the hip and peripheral devices measure the wrist, heel, finger, or tibia. These devices are listed in Table 57-3.

TABLE 57-3. Techniques for Bone Mass Measurement

For each of the three applications of bone density measurements previously listed, the central and peripheral devices have advantages and disadvantages.35,36 THE DIAGNOSIS OF OSTEOPENIA AND OSTEOPOROSIS The diagnosis of osteopenia (T score of <1.0 but >2.5) is clinically important, independent of fracture-risk assessment, for the initiation of early postmenopausal prevention strategies. This diagnostic category represents the range of bone density for which prevention strategies were designed. It has been shown that the recognition of low bone mass in early post-menopausal women facilitates the acceptance of hormone-replacement therapy (HRT).37,38 This observation will most likely be the same for other preventive interventions in early postmenopausal women. The utility of central or peripheral devices for the detection of osteopenia is dependent on the age of the patient and the particular skeletal site measured. In early menopausal patients, it is more likely that a diagnosis of osteopenia or osteoporosis will be missed if only one skeletal site is measured. In the elderly, bone mass may be more concordant at various skeletal sites. This reduces, but does not abolish, the likelihood of missing a diagnosis of osteopenia or osteoporosis if only one skeletal site is measured. The exception to the ability to use a single measurement in the elderly is the anteroposterior (AP) spine study by DXA. The AP spine measurement by DXA is frequently falsely elevated by osteophytes and/or facet sclerosis of the posterior elements; this may lead to underdiagnosis in this population.39 In the elderly, all peripheral and central measurements, with the exception of the AP spine measurement by DXA, have nearly equivalent value for diagnosing osteopenia or osteoporosis.27,40,41 There are fewer prospective fracture data for lateral spine DXA measurements. However, there is a strong correlation between lateral spine DXA BMD and spine quantitative computerized tomography (QCT) BMD, the latter of which has adequate fracture data. This suggests that low lateral spine DXA values should predict increased fracture risk in the elderly.42,43 BMD is more discordant throughout the skeleton in the early menopausal population than in the elderly (older than 65 years) population.27,40,44 There are at least four

potential explanations for this discordance: 1. 2. 3. 4. Differences in development of PABM at various sites.45 Differences in rates of age-dependent bone loss between cancellous bone and cortical bone after the menopause. Differences in the accuracy of measuring BMD using various techniques. T scores being derived from different young normal databases.

Data suggest that in early menopausal women, if one skeletal site is osteoporotic (T score <2.5 SD), there is a <10% chance that another skeletal site will be normal (T score >1.0 SD).27,46 Studies have not yet clarified how many early postmenopausal women have normal BMD at one site and osteopenia at another. These numbers may not be insignificant. Bone mass measurements can be used to rule in a diagnosis of low bone mass, but cannot be used to rule out a diagnosis (i.e., if they are normal). Hence, many early menopausal women will need additional BMD testing if a single skeletal site is normal to avoid missing an abnormal measurement at another skeletal site. This is particularly true in women with additional risk factors for low BMD or when a diagnosis of osteopenia would change the pharmacologic intervention. Patients with a normal single skeletal site BMD who may need additional testing are described in Table 57-4. These guidelines will minimize the potential for misdiagnosis.47

TABLE 57-4. Patients with a Normal Single Skeletal Site Bone Mass Who May Need Additional Bone Mass Testing

There are other patient populations in which a diagnosis of osteopenia may change intervention strategies. The prevention of bone loss associated with glucocorticoid therapy is recommended in patients with T scores <1.0.48 It may also be appropriate to intervene in other medical conditions associated with bone loss (i.e., hyperthyroidism [see Chap. 42], hyperparathy-roidism [see Chap. 58], posttransplantation, etc.) based on a finding of osteopenia. THE PREDICTION OF FRACTURE RISK It has been established that low bone mass is the most important predictor of fragility fracture. Approximately 80% of the variance in bone strength and resistance to fracture can be explained by bone mineral content per unit volume of bone in animal models.49 Other risk factors are predictive of fracture, only some of which can be modified.50 Risk factors such as increased height (>5 feet 7 inches), low body weight (<127 pounds), smoking, or maternal history of hip fracture are strong predictors of hip fracture. BMD is an objective, quantifiable measurement that offers clinical value akin to measuring a patient's blood pressure or cholesterol level. The relationship between fracture risk and bone density is a gradient rather than a single threshold value. The diagnostic category of osteopenia is valuable clinically because it may facilitate the acceptance of preventive interventions in early menopausal women to reduce their lifetime fracture risk. The measurement of low bone mass may also be used to determine current fracture risk, within 5 years, in elderly women. A 1.0 SD reduction in BMD is associated with an average 1.5-fold relative risk (RR) increase in global fracture risk in postmenopausal women and elderly men. The RR is greater with the same reduction in BMD as age increases.51 Because bone loss is an asymptomatic process, identifying small reductions in bone mass in the early menopausal patients is important to allow early intervention to halt this process. Patients may fracture with minimally reduced BMD (osteopenia), because of the effects of other factors that lead to bone fragility. The rate of bone turnover or the nature of a fall may lead to fracture with only a minimal reduction in BMD.12,52,53 Fracture prediction can be expressed as current fracture risk or lifetime fracture risk. Current risk is the risk of fracture within 3 to 5 years of the bone mass measurement. Current fracture risk can be expressed as RR, absolute risk or incidence, or annual risk. Relative risk, the ratio of two absolute risks, is increased 1.5 to 3 times for each 1.0 SD reduction in BMD.8,54 All the data available in the elderly suggests that fracture risk is not dependent on the skeletal site measured or the technique (central or peripheral) used.25,54,55 This observation is probably related to the concordance of bone mass measurements at different skeletal sites in the elderly population. Thus, in the elderly, low BMD measured at any skeletal site is more likely to represent increased fracture risk at other sites because of a global reduction in BMD and therefore increased fracture risk. The one exception is the prediction of hip fracture risk, which appears to be slightly better if the BMD is measured at the proximal femoral sites.54 This observation does not, however, diminish the strong predictive value of peripheral bone mass for hip fracture. Data on current fracture risk have been obtained only in the elderly because younger women fracture only infrequently and bone mass technology has not existed long enough to follow a sufficient number of younger women for an adequate period of time.8,16,19,54,56 Current fracture risk data should not be applied to younger perimenopausal women or premenopausal women with low peak adult bone mass for the age-related reasons previously stated. After 60 years, advancing age is an independent predictor of fracture. In counseling patients, it is correct to state that the current fracture risk of a 50-year-old woman at a given level of BMD is substantially lower than the current fracture risk of a 70-year-old woman with the same BMD.51 However, the 50-year-old woman with low BMD may have a greater lifetime fracture risk than her 50-year-old counterpart with a normal BMD57,58 and 59 if no measures are taken to preserve skeletal mass. No direct data exist to validate these lifetime risk statistical models.57,58,59 and 60 Certainly if the 50-year-old woman with low bone mass is postmenopausal, she should be counseled about preventive therapeutic options. She should be advised to continue with her usual life activities without excessive fear of fracture. SERIAL ASSESSMENTS OF BONE MASS Bone densitometry is also valuable for serial monitoring of the natural progression of disease processes (such as primary hyperparathyroidism) or for monitoring the response of bone to pharmacologic interventions.60a There is currently a clear advantage of using the central skeleton instead of the peripheral skeleton for serial monitoring. The axial skeleton measured by DXA or QCT consistently demonstrates a greater magnitude of change over time to allow monitoring of the bone response to pharmacologic therapy.11,61,62,63,64 and 65 The femur demonstrates smaller changes in response to pharmacologic interventions, and little or no change is observed at the wrist, finger, or heel. The reason for these discrepancies is unclear.66 It is not the result of any limitations of the technology, because the precision of the peripheral techniques is excellent (1.0%). Several hypotheses have been suggested to explain this observation. There may be differences in the bone marrow environment of the peripheral skeleton versus the central skeleton, the surface area of the bone may be different, or differences in blood flow may exist between the cortical and cancellous bone compartments. No matter which hypothesis is correct, the bone mass of the spine and hip is more responsive to pharmacologic intervention. There are some limited data suggesting that forearm bone mass measurements may be valuable for monitoring the bone response to hormone-replacement therapy in some patients.67 This study needs to be repeated and confirmed. The heel site is not recommended for this purpose for the reasons previously mentioned. To summarize, the issue is not one of precision using peripheral measurements, but instead is related to the slow, small response to pharmacologic interventions seen at peripheral sites. Nevertheless, peripheral measurements have been approved by the Health Care Finance Administration (HCFA) for both diagnosis and longitudinal monitoring. Bone mass measurements using both central and peripheral technologies may be appropriate within the same year if the intent of the two measurements is different (e.g., peripheral for diagnosis and central for monitoring). Guidelines regarding which patients may not need central skeletal testing once they are diagnosed with low bone mass by a peripheral device have been published.68 These latter guidelines are intended to preclude overuse of both technologies yet provide some assurance that a treated patient is being effectively monitored. Biochemical markers of bone turnover can also be used to assess pharmacologic effects and help to guide management decisions. They can also be used to limit the number of patients who receive both peripheral and central testing in the same year.69 In contrast to pharmacologic therapies, the forearm appears to be the best skeletal site for monitoring the effects of excess parathyroid hormone.70 It may also be

valuable in decisions regarding the timing of surgical parathyroidectomy. No matter which BMD technique is used, serial changes can be expressed as the percentage of change (% change)(1st BMD 2nd BMD/1st BMD 100) or absolute change (g/cm2) between the two measurements. Either format is acceptable, providing that the precision of the device and the skeletal site is known. Individual manufacturers report precision errors of 1% to 2% at the AP spine. However, their precision error may be higher (3% to 4%) in the elderly population with low BMD values.71 This is the result of an increased coefficient of variation (precision error) as BMD declines at each skeletal site (Fig. 57-3). In addition, different operators may introduce variability based on their training and skill during positioning or analysis. Therefore, each bone densitometry facility should calculate their individual precision error for each bone densitometry device used. Once the precision error is known, the significance of a serial BMD change can be determined. At the 95% confidence level, a significant BMD change needs to be at least 2.8 times the precision error. For example, if the precision error is 1.5%, then a 4.16% change would be necessary for significance at the 95% confidence level. In contrast, if the precision error is 3%, an 8.3% change would be necessary for significance.6 Misleading conclusions may be drawn based on nonsignificant serial changes, with attendant inappropriate alterations in treatments, if serial measurements are interpreted incorrectly.

FIGURE 57-3. Precision error (CV%) of serial bone mass measurements deteriorates at lower level of bone mass. (BMD, bone mineral density.)

In contrast, absolute change has a small margin of error because the SD of absolute BMD is reasonably constant over a wide range. Any change in BMD that is > 0.04 g/cm2 at the AP spine or 0.05 g/cm2 at the femoral neck is significant at the 95% confidence level. Absolute change is calculated by subtracting the second BMD measurement from the baseline measurement. In most of the medical literature, serial BMD change is expressed as a percentage; however, either method is valid. BMD testing for serial monitoring is generally performed every 12 to 24 months, depending on the disease process or therapeutic intervention. In patients who have a documented response to pharmacologic interventionwhich may be defined as either a gain or no loss in BMDannual BMD measurements may not be necessary after the first year. In patients with a documented response to HRT, repeat BMD measurements every 3 to 5 years may improve the long-term compliance to therapy. Even in elderly women who have previously been documented as estrogen responders, bone mass measurements at 3- to 5-year intervals may document continued response and compliance to therapy. In this elderly population, age-related bone loss may overcome estrogen therapy causing further bone loss.72 Thus, the NOF recommends BMD testing in women older than 65 years of age even if they have no risk factors and are receiving hormone-replacement therapy.13 The frequency of serial monitoring may differ for non-estrogen therapy (e.g., bis-phosphonates and calcitonin). It is very difficult to compare serial changes if the measurements are performed on machines from different manufacturers. It even is sometimes difficult to compare values obtained from different machines made by the same manufacturer. It would be ideal if patients had serial measurements performed on the same machine by the same technician. However, this is unrealistic. The International Bone Densitometry Standards Committee has established a standardized BMD (sBMD), which allows for comparisons to be made between BMD values obtained from different manufacturers' equipment.73 Using the calculated sBMD of the spine and hip for serial comparison reduces but does not eliminate the variance in measurements. As a general rule, the precision error should be increased by 1% for the calculation of percentage change if sBMD is used.

CONCLUSIONS
Bone densitometry has revolutionized the clinical approach to osteoporosis. This technology provides a direct measurement of bone mineral density by which fracture risk can be estimated. If the results of testing are used responsibly and competently, patient care will be enhanced. The measurement of bone mineral density enables physicians and their patients to make informed decisions regarding preventive and therapeutic strategies. It also allows the physician to monitor the longitudinal efficacy of these interventions. CHAPTER REFERENCES
1. 2. 3. 4. Miller PD, Bonnick SL, Rosen CJ, et al. Clinical utility of bone mass measurements in adults: consensus of an international panel. Semin Arthritis Rheum 1996; 25:361. The WHO Study Group. Assessment of fracture risk and its application to screening for postmenopausal osteoporosis. Geneva: World Health Organization, 1994. Hui SL, Slemenda CW, Carey MA, Johnston CC Jr. Choosing between predictors of fracture. J Bone Miner Res 1995; 10:186. Cummings SR. Treatable and untreatable risk factors for hip fracture. Bone 1996; 18:165S.

4a. Peretz A, Penaloza A, Mesquita M, et al. Quantitative ultrasound and dual x-ray absorptiometry measurements of the calcaneus in patients on maintenance hemodialysis. Bone 2000; 27:287. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. Genant HK, Engelke K, Fuerst T, et al. Noninvasive assessment of bone mineral and structure: state of the art. J Bone Miner Res 1996; 11:707. Faulkner KG, von Stetten E, Miller P. Discordance in patient classification using T-scores. J Clin Densitom 1999; 2(3):343. Anonymous. Consensus development conference: diagnosis, prophylaxis and treatment of osteoporosis. Am J Med 1993; 94:646. Ross PD, Davis JW, Epstein RS, Wasnich RD. Pre-existing fractures and bone mass predict vertebral fracture incidence in women. Ann Intern Med 1991; 114:919. Huang C, Ross PD, Wasnich RD. Short-term and long-term fracture prediction by bone mass measurements: a prospective study. J Bone Miner Res 1998; 13:107. Ravn P, Overgaard K, Huang C, et al. Comparison of bone density of the phalanges, distal forearm and axial skeleton in early postmenopausal women participating in the EPIC study. Osteoporosis Int 1996; 6:308. Writing Group for PEPI trial. Effects of hormone therapy on bone mineral density. JAMA 1996; 276:1389. Riis BJ, Hansen MA, Jensen AM, et al. Low bone mass and fast rate of bone loss at menopause: equal risk factors for future fracture: a 15-year follow-up study. Bone 1996; 19:9. Lindsay R. Risk assessment using bone mineral density determination. Osteoporosis Int 1998; 8(S1):28. Lunt M, Felsenberg D, Reeve J, et al. Bone density variation and its effect on risk of vertebral deformity in men and women studied in thirteen European centers: the EVOS study. J Bone Miner Res 1997; 12:1883. Mussolino ME, Looker AC, Madans JH, et al. Risk factors for hip fracture in white men: the NHANES I epidemiological follow-up study. J Bone Miner Res 1998; 13:918. De Laet CEDH, Van Hout BA, Burger H, et al. Hip fracture prediction in the elderly men and women: validation in the Rotterdam study. J Bone Miner Res 1998; 13:1587. Melton LJ III, Atkinson EJ, O'Connor MK, et al. Bone density and fracture risk in men. J Bone Miner Res 1998; 13:1915. Looker AC, Wahner HW, Dunn WL, et al. Proximal femur bone mineral levels of US adults. Osteoporosis Int 1995; 5:389. Melton LJ III. How many women have osteoporosis now? J Bone Miner Res 1995; 10:175. Faulkner KG, Roberts LA, McClung MR. Discrepancies in normative data between Lunar and Hologic DXA system. Osteoporosis Int 1996; 6:432. Ahmed AIH, Blake GM, Rymer JM, Fogelman I. Screening for osteopenia and osteoporosis: do the accepted normal ranges lead to overdiagnosis? Osteoporosis Int 1997; 7:432. Simmons A, Simpson DE, O'Doherty MJ, et al. The effects of standardization and reference values on patient classification for spine and femur dual-energy x-ray absorptiometry. Osteoporosis Int 1997; 7:200. Faulkner K, von Stetten E, Miller P. Discordance in patient classification using T-scores. J Clin Densitometry 1999; 2(3):343. Greenspan SL, Bouxein ML, Melton ME, et al. Precision and discriminatory ability of calcaneal bone assessment technologies. J Bone Miner Res 1997; 12:1303. Marshall D, Johnell O, Wedel H. Meta-analysis of how well measures of bone mineral density predict occurrence of osteoporotic fractures. BMJ 1996; 312:1254. Grampp S, Genant HK, Mathur A, et al. Comparisons of noninvasive bone mineral measurements is assessing age-related loss, fracture discrimination, and diagnostic classification. J Bone Miner Res 1997; 12:697. Arlot ME, Sornay-Rendu E, Garnero P, et al. Apparent pre- and postmeno-pausal bone loss evaluated by DXA at different skeletal sites in women: the OLEFY cohort. J Bone Miner Res 1997; 12:683. Cummings SR, Black DM, Rubin SM. Lifetime risks of hip, Colles' or vertebral fracture and coronary heart disease among white postmenopausal women. Arch Intern Med 1989; 149:2556. Goulding A, Cannan R, Williams SM, et al. Bone mineral density in girls with forearm fractures. J Bone Miner Res 1998; 13:143. Duppe H, Gardsell P, Nilsson B, Johnell O. A single bone density measurement can predict fractures over 25 years. Calcif Tissue Int 1997; 60:171. Riis BJ. Premenopausal bone loss: fact or artifact? Osteoporosis Int 1994; S1:S35. Recker RR, Lappe JM, Davies KM, Kimmel DB. Change in bone mass immediately before menopause. J Bone Miner Res 1992; 7:857. Hui SL, Slemenda CW, Johnston CC Jr. Age and bone mass as predictors of fracture in a prospective study. J Clin Invest 1988; 81:1804. Melton LJ III, Khosla S, Atkinson EJ, et al. Relationship of bone turnover to bone density and fractures. J Bone Miner Res 1997; 12:1083. Miller PD, McClung M. Prediction of fracture risk I: bone density. Am J Med Sci 1996; 312:257. Baran DT, Faulkner KG, Genant HK, et al. Diagnosis and management of osteoporosis: guidelines for the utilization of bone densitometry. Calcif Tissue Int 1997; 61:433. Rubin SM, Cummings SR. Results of bone densitometry affect women's decisions about taking measures to prevent fractures. Ann Intern Med 1992; 116:990. Silverman SL, Greenwald M, Klein RA, Drinkwater BL. Effect of bone density information on decisions about hormone replacement therapy: a randomized trial. Obstet Gynecol 1997; 89:321. Greenspan SL, Maitland-Ramsey L, Myers E. Classification of osteoporosis in the elderly is dependent on site-specific analysis. Calcif Tissue Int 1995; 58:409.

40. Melton LJ III, Chrischilles EA, Cooper C, et al. How many women have osteoporosis? J Bone Miner Res 1992; 7:1005. 41. Mazess RB. Advances in bone densitometry. Ital J Miner Electrolyte Metab 1997; 11:73. 42. Finkelstein JS, Cleary RL, Butler JP, et al. A comparison of lateral versus anterior-posterior spine dual energy x-ray absorptiometry for the diagnosis of osteopenia. J Clin Endocrinol Metab 1994; 78:724. 43. Ross PD, Genant HK, Davis JW, et al. Predicting vertebral fracture incidence from prevalent fractures and bone density among non-black osteoporotic women. Osteoporosis Int 1993; 3:120. 44. Pouilles JM, Tremollieres F, Ribot C. Spine and femur densitometry at the menopause: are both sites necessary in the assessment of the risk of osteoporosis? Calcif Tissue Int 1993; 52:344. 45. Bonnick SL, Nichols DL, Sanborn CF, et al. Dissimilar spine and femoral Z-scores in premenopausal women. Calcif Tissue Int 1997; 61:263. 46. Nelson DA, Molloy R, Kleerekoper M. Prevalence of osteoporosis in women referred for bone density testing: utility of multiple skeletal sites. J Clin Densitometry 1998; 1:5. 47. Miller PD, Bonnick SL, Johnston CC Jr, et al. The challenges of peripheral bone density testing. J Clin Densitometry 1998; 1:1. 48. Reid I. Glucocorticoid-induced osteoporosis: assessment and treatment. J Clin Densitometry 1998; 1:55. 49. Yang R-S, Wang S-S, Lin H-J, et al. Differential effects of bone mineral content and bone area on vertebral strength in a swine model. Calcif Tissue Int 1998; 63:86. 50. Cummings SR, Nevitt MC, Browner WS, et al. Risk factors for hip fracture in white women. N Engl J Med 1995; 332:767. 51. Hui SL, Slemenda CW, Johnston CC Jr. Age and bone mass as predictors of fracture in a prospective study. J Clin Invest 1988; 81:1804. 52. Garnero P, Hausherr E, Chapuy M-C, et al. Markers of bone resorption predict hip fracture in elderly women: the EPIDOS prospective study. J Bone Miner Res 1996; 11:1531. 53. Greenspan SL, Myers ER, Maitland LA, et al. Fall severity and bone mineral density as risk factors for hip fracture in ambulatory elderly. JAMA 1994; 271:128. 54. Cummings SR, Black DM, Nevitt MC, et al. Bone density at various sites for prediction of hip fracture. Lancet 1993; 341:72. 55. Yates AJ, Ross PD, Lydick E, Epstein RS. Radiographic absorptiometry in the diagnosis of osteoporosis. Am J Med 1995; 98(S2A):41S. 56. Hans D, Dargent-Molina P, Scott AM, et al. Ultrasonographic heel measurements to predict hip fracture in elderly women. The EPIDOS prospective study. Lancet 1996; 348:511. 57. Kanis JA. Diagnosis of osteoporosis. Osteoporosis Int 1997; 7(S3):S108. 58. Black DM, Cummings SR, Melton LJ III. Appendicular bone mineral and a woman's lifetime risk of hip fracture. J Bone Miner Res 1992; 7:639. 59. Melton LJ III, Atkinson EJ, O'Fallon WM, et al. Long-term fracture prediction by bone mineral assessed at different skeletal sites. J Bone Miner Res 1993; 8:1227. 60. Huang C, Ross PD, Wasnich RD. Short-term and long-term fracture prediction by bone mass measurements: a prospective study. J Bone Miner Res 1998; 13:107. 60a. Orwoll E, Ettinger M, Weiss S, et al. Alendronate for the treatment of osteoporosis in men. N Engl J Med 2000; 343:604. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. Lufkin EG, Wahner HW, O'Fallon WM, et al. Treatment of postmenopausal osteoporosis with transdermal estrogen. Ann Intern Med 1992; 117:1. Watts NB, Harris ST, Genant HK, et al. Intermittent cyclic etidronate treatment of postmenopausal osteoporosis. N Engl J Med 1990; 323:73. Liberman UA, Weiss SR, Broll J, et al. Effect of oral alendronate on bone mineral density and the incidence of fractures in postmenopausal osteoporosis. N Engl J Med 1995; 333:1437. Black DM, Cummings SR, Karpf DB, et al. Randomized trial of effect of alendronate on risk of fracture in women with existing vertebral fractures. Lancet 1996; 348:1535. McClung M, Clemmesen B, Daifotis A, et al. Alendronate prevents post-menopausal bone loss in women without osteoporosis. Ann Intern Med 1998; 128:253. Rosenthall L, Caminis J, Tenehouse A. Calcaneal ultrasonometry: response to treatment in comparison with dual x-ray absorptiometry measurements of the lumbar spine and femur. Calcif Tissue Int 1999; 64:200. Christiansen C, Lindsay R. Estrogens, bone loss and preservation. Osteoporosis Int 1990; 1:7. Miller PD, Bonnick SL, Johnston CC Jr, et al. The challenges of peripheral bone density testing. Which patients need additional central density skeletal measurement. J Clin Densitometry 1998; 1:211. Greenspan SL, Parker RA, Ferguson L, et al. Early changes in biochemical markers of bone turnover predict the long-term response to alendronate therapy in representative elderly women: a randomized clinical trial. J Bone Miner Res 1998; 13:1431. Silverberg SJ, Shane E, de la Cruz L, et al. Skeletal disease in primary hyperparathyroidism. J Bone Miner Res 1989; 4:283. Faulkner KG, McClung MR. Quality control of DXA instruments in multi-center trials. Osteoporosis Int 1995; 5:218. Cauley JA, Seeley DG, Ensrud K, et al. Estrogen replacement therapy and fracture in older women. Ann Intern Med 1995; 122:9. Steiger P, for the International Committee for Standards in Bone Measurement. Letter to the editor: standardization of spine BMD measurements. J Bone Miner Res 1995; 10:1602.

CHAPTER 58 PRIMARY HYPERPARATHYROIDISM Principles and Practice of Endocrinology and Metabolism

CHAPTER 58 PRIMARY HYPERPARATHYROIDISM

SHONNI J. SILVERBERG AND JOHN P. BILEZIKIAN Incidence Pathophysiology Etiology Clinical Manifestations Bone Disease Renal Complications Other Systemic Effects Hypercalcemic Manifestations Clinical Presentation Laboratory Evaluation Parathyroid Hormone Measurement Other Tests Diagnosis Differential Diagnosis Lithium Use Thiazide Use Coexistence of Two Causes of Hypercalcemia Therapy Surgery Nonsurgical Medical Approaches Primary Hyperparathyroidism During Pregnancy Parathyroid Carcinoma Familial Hypocalciuric Hypercalcemia Pathophysiology Manifestations and Therapy Chapter References

Primary hyperparathyroidism is caused by excessive, abnormally regulated secretion of parathyroid hormone (PTH) from the parathyroid glands. Chronic exposure of its two principal target organs, bone and kidney, to PTH causes hypercalcemia, a major hallmark of the disease. The incidence, pathophysiology, cause, clinical manifestations, diagnostic evaluation, and therapy of primary hyperparathyroidism are discussed in this chapter. Pathology and surgical therapy are covered separately in Chapter 48 and Chapter 62, respectively.

INCIDENCE
The widespread clinical use of the multichannel screening test coincided with a dramatic increase in the incidence of primary hyperparathyroidism.1,2 Before routine determinations of serum calcium were initiated in the early 1970s, primary hyperparathyroidism was an infrequent diagnosis; in the 1990s, it was diagnosed in as many as 1 person of every 1000 members of the general population. The dramatic four- to five-fold increase in apparent incidence that occurred within 10 years of introduction of the multichannel screening test has returned to a relatively stable rate. The incidence of the disease (i.e., recognized cases) now closely approximates the prevalence (i.e., disease detected and undetected) in the population. One study, as yet unconfirmed, suggests a declining incidence of primary hyperparathyroidism.3 Primary hyperparathyroidism occurs at all ages but remains distinctly unusual in children. The peak incidence of primary hyperparathyroidism is in the fifth to sixth decade of life, with a female to male ratio of 3:1. Primary hyperparathyroidism results from a single parathyroid adenoma in 80% of individuals with surgically proven disease. Involvement of more than one parathyroid gland in a different pathologic process, hyperplasia, occurs in most of the remaining individuals. Primary hyperparathyroidism associated with four-gland parathyroid hyperplasia occurs commonly in conjunction with the syndromes of multiple endocrine neoplasia types 1 and 2A (MEN1 and MEN2A) (see Chap. 188). Rarely, patients with primary hyperparathyroidism harbor multiple adenomas. Even less commonly (<1%), pathologic analysis reveals parathyroid carcinoma. The histologic features that differentiate parathyroid adenoma, hyperplasia, and carcinoma are discussed in Chapter 48.

PATHOPHYSIOLOGY
The functional abnormality of primary hyperparathyroidism is incompletely understood. Normally, PTH secretion is tightly regulated by the serum ionized calcium concentration; it is stimulated when the ionized calcium concentration decreases and suppressed when this concentration increases. In primary hyperparathyroidism, this sensitive regulatory control is lost, and the parathyroid glands continue to secrete PTH despite an elevated ionized calcium level. The cellular basis for altered control could be an altered setpoint, so that the parathyroid cell can be suppressed only if the ionized calcium concentration is raised to a level that is higher than normal. Alternatively, the excessive secretion of PTH could result from an increased number of parathyroid cells without a change in the setpoint for calcium. In primary hyperparathyroidism caused by adenoma, both mechanisms are likely to be operative. Conversely, in primary hyperparathyroidism with hyperplasia, the setpoint appears to be closer to normal.4,5 The abnormal secretion of PTH probably arises from the increased number of parathyroid cells (see Chap. 49). Evidence suggests that circulating growth factors may be related to the hyperparathyroidism associated with MEN1.6 Although other regulators of PTH secretion, such as b-adrenergic catecholamines, histamine, and ions other than calcium, may possibly be important factors in the pathogenesis of primary hyperparathyroidism, the evidence suggests that these secretagogues are not important pathophysiologic factors in the establishment or maintenance of the hyperparathyroid state.7 Abnormal regulation of PTH by calcium is found in primary hyperparathyroidism, but the glands are not completely autonomous. The abnormal parathyroid gland is suppressible by some level of calcium; otherwise, all patients with the disorder would have uncontrolled hypercalcemia instead of relatively stable, albeit abnormal levels. Patients with primary hyperparathyroidism who have a relatively high calcium intake (1 g daily) have PTH levels that are lower than patients whose calcium intake is less (400 mg daily).8

ETIOLOGY
The cause of primary hyperparathyroidism is known only for a small subset of patients. Patients who had prior irradiation to the neck region may be at risk years later for the development of primary hyperparathyroidism.9 and 10 The forms of primary hyperparathyroidism associated with familial syndromes have a genetic origin.6,11,12,13,14 and 15 However, most patients with primary hyperparathyroidism do not have a family history of primary hyperparathyroidism or of its associated familial syndromes. Genetic abnormalities that could be linked to sporadic parathyroid tumors have generated great interest; the first to be described is a rearrangement of the cyclin D1/(PRAD 1) oncogene. The gene defect for abnormal parathyroid tissue from some patients with primary hyperparathyroidism has been shown to be repositioned close to the strong tissue-specific enhancer elements of the PTH gene.16 This realignment of DNA associates a growth promoter (cyclin D1) with a regulatory element that normally controls only a hormonal function, namely synthesis of PTH. Under these circumstances, whenever PTH is stimulated, cellular growth is also stimulated. Despite the attractiveness of this concept of a molecular mechanism of monoclonal parathyroid gland growth, only a few parathyroid tumors have been demonstrated to harbor this defect. Nevertheless, cyclin D1 protein levels are increased in as many as 20% of parathyroid adenomas.17 Another attractive molecular mechanism postulated to account for abnormal parathyroid secretion in primary hyperparathyroidism is loss of tumor-suppressor gene function. Abnormalities in the MEN1 gene have been shown to cause the MEN1 syndrome18,19 (see Chap. 188). The view was long held that both copies of this tumor-suppressor gene might well be inactivated and, thus, contribute to the development of sporadic parathyroid adenomas. A somatic mutation in the MEN1 gene has been found to be associated with 15% to 25% of nonfamilial parathyroid adenomas.20,21 and 22 In addition to the MEN1 gene, other tumor-suppressor genes have been implicated in the pathogenesis of sporadic parathyroid tumors. Chromosomal regions of

interest include 1p and lq, 6q, 9p, and 15q.23,24 Mutations in the gene for the calcium-sensing receptor cause two well-known parathyroid disorders.25,26 In neonatal severe hyperparathyroidism, homozygous mutations completely inactivate this gene; thus, uncontrolled hypercalcemia results. In the heterozygous counterpart, familial hypocalciuric hypercalcemia (FHH), a point mutation in one allele leads to a condition that simulates primary hyperparathyroidism but is distinctly different. In sporadic primary hyperparathyroidism, no somatic mutations in the calcium-sensing receptor gene have been detected,27 although in many adenomas, calcium-receptor protein and sometimes even its messenger RNA is reduced.28,29

CLINICAL MANIFESTATIONS
The signs and symptoms of primary hyperparathyroidism result from the hypercalcemia (see Chap. 59) and the hyperparathyroid state. In primary hyperparathyroidism, the classic target organs are the bones and the kidneys. BONE DISEASE In classic primary hyperparathyroidism, the skeleton is involved in a process known as osteitis fibrosa cystica.30 Radiographically, subperiosteal bone resorption is seen in several typical sites (Fig. 58-1). The salt-and-pepper appearance of the skull, resorption of the distal phalanges, and tapering of the distal clavicles are classic features of the resorptive process in bone. Bone cysts and brown tumors also may be seen. Brown tumors are collections of osteoclasts intermixed with poorly mineralized woven bone. Another important radiologic feature of hyperparathyroid bone disease is demineralization caused by prolonged bone resorption. Osteopenia alone may be a radiographic characteristic of primary hyperparathyroidism. The differential diagnosis of osteoporosis should always include an evaluation for primary hyperparathyroidism.

FIGURE 58-1. Radiographic representations of osteitis fibrosa cystica in primary hyperparathyroidism. A, Salt-and-pepper erosions of the skull. B, Cystic bone disease in the clavicle (arrows). C, Subperiosteal bone resorption of the digits. D, Cortical erosions (arrows).

Before the widespread use of the multichannel autoanalyzer in the early 1970s, the incidence of radiologically apparent bone disease in hyperparathyroidism was 10% to 15%. This figure is now much lower (<5%) because of the markedly higher detection of asymptomatic primary hyperparathyroidism in patients. Although grossly evident bone disease is no longer a frequent finding in primary hyperparathyroidism, more sophisticated diagnostic testing shows continued involvement of the skeleton in the hyperparathyroidism. Bone mineral densitometry and quantitative bone histomorphometry, for example, have detected evidence for hyperparathyroidism among patients without otherwise apparent hyperparathyroid bone disease.30 The changes are seen at cortical bone, which is found predominantly in the appendicular skeleton. Thinning of cortical bone can be detected by bone mineral densitometry. Figure 58-2 demonstrates typical preferential loss of cortical bone in primary hyperparathyroidism. In contrast, excess PTH secretion appears to have a relative protective effect against bone loss at sites of cancellous bone, such as vertebral bone. Trabecular plates are typically maintained in number and in connectivity in primary hyperparathyroidism (Fig. 58-3; see Chap. 55).31,32 and 33

FIGURE 58-2. Bone densitometry can reveal mineral loss and bone thinning in patients with primary hyperparathyroidism. Patients' lumbar spine, femoral neck, and radius values are compared with the expected values from age- and sex-matched controls. The divergence from expected value is different at each site (asterisk indicates that the change is significant at the level of p <.0001).

FIGURE 58-3. Primary hyperparathyroidism as seen by electron microscopy. A, Bone from a 45-year-old man is compared with (B) normal bone from an age- and sex-matched normal person. The trabecular plates are conserved in primary hyperparathyroidism, but the cortex is considerably thinner. (Courtesy of Dr. David Dempster.)

Awareness of the potential for bone involvement in patients with asymptomatic primary hyperparathyroidism has led to the routine use of bone mineral densitometry as part of the clinical evaluation of the disease. The results are used to help determine whether the patient should undergo parathyroid surgery. RENAL COMPLICATIONS Nephrolithiasis was and still is the most frequent complication of primary hyperparathyroidism. In older studies, as many as 40% of patients had stone disease.34 Just as the incidence of overt bone disease in primary hyperparathyroidism has decreased, the incidence of nephrolithiasis has also diminished with the changing clinical

profile of the disease.35 The incidence of nephrolithiasis is ~20%. Other renal complications include nephrocalcinosis, a radiographic presentation of diffuse renal calcification, and hypercalciuria (see Chap. 208). Although PTH stimulates distal tubular reabsorption of calcium, the increased filtered load of calcium caused by hypercalcemia may lead to increased renal calcium excretion. As many as 35% to 40% of patients with primary hyperparathyroidism may have a urinary calcium excretion of >250 mg daily. Patients with primary hyperparathyroidism may show only a decrease in the creatinine clearance. However, successful removal of the parathyroid adenoma does not predictably lead to an improvement in the creatinine clearance. OTHER SYSTEMIC EFFECTS Organs other than skeleton and kidneys may be primary targets for complications of primary hyperparathyroidism. Muscle weakness and easy fatigability can occur. The underlying abnormality appears to be muscle atrophy directed toward, but not limited to, type II muscle fibers.36 Atrophy of these fibers can explain the clinical finding of proximal muscle weakness. However, patients with primary hyperparathyroidism no longer demonstrate the classic neuromuscular signs of the disease, probably because so many patients are diagnosed with mild hypercalcemia.37 Nevertheless, the presentation of weakness and easy fatigability may be considered a nonspecific manifestation of primary hyperparathyroidism. The gastrointestinal tract is a focus for symptoms of hypercalcemia and a potential site of complications of primary hyperparathyroidism. Peptic ulcer disease and acute pancreatitis are classically associated with primary hyperparathyroidism. Whether these gastrointestinal disorders reflect a tendency for common diseases to appear coincidentally in some patients or whether an abnormal functional association exists between primary hyperparathyroidism and gastrointestinal disease is uncertain38 (see Chap. 204). Similarly, the increased prevalence of hypertension in patients with primary hyperparathyroidism may not have a pathophysiologic basis. Successful removal of a parathyroid adenoma does not usually change the degree of hypertension, nor does it improve antihypertensive management.39 Gout and pseudogout have occurred in patients with primary hyperparathyroidism, particularly during the postoperative period40 (see Chap. 211). Normocytic, normochromic anemia is observed only in patients with advanced parathyroid disease and corrects itself after cure of the hyperparathyroid state.41 Among the psychiatric manifestations of primary hyperparathyroidism, depression may be prominent. Less serious, but nevertheless important, is a set of cognitive complaints that patients may bring to the physician's attention. A general decrease in attentiveness and a sense of intellectual weariness have been observed among middle-aged individuals with primary hyperparathyroidism (see Chap. 200). This is another area in which associating a causal role for the hyperparathyroid state is difficult.42 Reports of improvement in this area after successful parathyroidectomy are difficult to evaluate because of the paucity of accurate quantitative tools for assessing these features and the lack of adequately controlled clinical trials.43,44 In very rare cases, primary hyperparathyroidism has been described in association with several unusual nonendocrine manifestations. In some cases of familial isolated primary hyperparathyroidism (FIPH), coexisting ossifying jaw tumors have been described.13,15 Parathyroid adenomata in association with Cowden disease has also been reported.45 This is a rare, but well described, syndrome characterized by diffuse gastrointestinal polyposis and oral and cutaneous hamartomas. HYPERCALCEMIC MANIFESTATIONS Other signs and symptoms of primary hyperparathyroidism are not specific features of the disorder but reflect the hypercalcemia itself (see Chap. 59). They are not unique to primary hyperparathyroidism and may be associated with any hypercalcemic state. The extent to which patients demonstrate signs of hypercalcemia (e.g., shortened QT interval on the electrocardiogram) or symptoms of hypercalcemia (e.g., anorexia, nausea, vomiting, constipation, polyuria) depends on the actual level of the serum calcium and its rate of rise.

CLINICAL PRESENTATION
Patients with primary hyperparathyroidism may present with one of several clinical pictures: asymptomatic hypercalcemia; bone or stone disease; other recognized complications, such as neuromuscular, gastrointestinal, articular, hematologic, or central nervous system features; normocalcemic hyperparathyroidism; acute primary hyperparathyroidism; parathyroid carcinoma; MEN1 or MEN2A; FIPH; familial cystic parathyroid adenomatosis; or neonatal hyperparathyroidism.46 The most common form is mild hypercalcemia without any specific signs or symptoms. The increased incidence of primary hyperparathyroidism reflects recognition of a population brought to the physician's attention by the incidental determination of calcium levels. Usually, these asymptomatic patients have serum calcium levels that are not more than 1 mg/dL above the upper limits of normal.47 Other test results, such as those for serum phosphorus and the alkaline phosphatase, are usually normal. In marked contrast to this presentation is an unusual and life-threatening form, acute primary hyperparathyroidism. Patients with this disorder are symptomatic and present with extremely high serum calcium values (>15 mg/dL).48 Approximately 25% of patients reported with this form of primary hyperparathyroidism have a previous history of mild hypercalcemia. Patients with acute primary hyperparathyroidism have extremely high serum PTH levels. In one review, the serum calcium value for 43 patients was 17.5 2.1 mg/dL (mean standard deviation). In marked contrast to patients with asymptomatic primary hyperparathyroidism, virtually all patients were symptomatic. Unlike primary hyperparathyroidism, acute primary hyperparathyroidism involved both bone and kidney in one-half of the patients. More than two-thirds of patients had nephrolithiasis or nephrocalcinosis. The clinician must recognize that life-threatening hypercalcemia may be the result of acute primary hyperparathyroidism, because this is a curable disease. Another clinical presentation of primary hyperparathyroidism is that associated with a limited set of complications. As expected, the skeleton, kidneys, or neuromuscular system is most commonly involved. Primary hyperparathyroidism may also occur in a hereditary syndrome in association with MEN1 or MEN2A or simply as FIPH. Patients with FIPH do not have any manifestations of MEN. Patients from these kindreds often present at a younger age than the typical individual with sporadic primary hyperparathyroidism. Their disease generally affects more than one gland and is often recurrent. Because of the high recurrence rate after removal of a single adenoma, subtotal parathyroidectomy is the preferred approach in such individuals.49,50 Neonatal primary hyperparathyroidism also occurs (see Chap. 70). Another familial form of primary hyperparathyroidism is characterized by distinctive cystic pathologic features.51 Patients from these kindreds typically develop recurrent primary hyperparathyroidism resulting from another cystic adenoma years after successful removal of the first adenoma. No other endocrine tumors have been described in this syndrome, but several fibrous maxillary or mandibular tumors have been seen. Results of the physical examination of patients with primary hyperparathyroidism are not noteworthy. The clinical sign of calcium deposition in the cornea, band keratopathy, is rarely seen (see Chap. 215). Similarly, palpation of an enlarged parathyroid gland is unusual unless parathyroid carcinoma is present.

LABORATORY EVALUATION
The diagnosis of primary hyperparathyroidism depends on laboratory tests. The serum calcium concentration is virtually always elevated. The serum phosphorus value may be decreased, but it is usually normal, although typically in the lower range of normal. Newer markers of bone metabolism are useful to evaluate the presence of underlying bone involvement in cases of primary hyperparathyroidism. Markers of bone formation and resorption tend to be elevated, even in asymptomatic primary hyperparathyroidism. Whether quantification of these markers is useful to predict the extent of PTH-dependent skeletal tumors or whether they are useful predictors of the course of the skeletal involvement is not yet certain, however. The bone formation marker, total alkaline phosphatase activity, which is readily available on routine multichannel panels, continues to be a useful marker in primary hyperparathyroidism,52 and its bone-specific isoenzyme is often clearly elevated in patients with mild primary hyperparathyroidism.53,54 Osteocalcin is also generally increased in patients with primary hyperparathyroidism.54,55 and 56 Bone resorption is reflected in a product of osteoclast activity, tartrate-resistant acid phosphatase (TRAP), and collagen breakdown products such as hydroxyproline, hydroxypyridinium cross-links of collagen, and the small N- and C-telopep-tides of collagen metabolism. The oldest known marker of bone resorption,57,58 urinary hydroxyproline excretion, is frankly elevated in patients with osteitis fibrosa cystica. However, in patients with mild, asymptomatic primary hyperparathyroidism, values are often entirely normal. The hydroxypyridinium cross-links of collagen have replaced hydroxyproline excretion as markers of bone resorption used in evaluating primary hyperparathyroidism (see Chap. 56). These cross-links of collagen (pyridinoline and deoxypyridinoline) are mildly elevated.59 Studies of TRAP levels in primary hyperparathyroidism are limited, although levels have been shown to be elevated.60 Bone sialoprotein is a phosphorylated glycoprotein that makes up b5% to 10% of the noncollagenous protein of bone. It is elevated in high-turnover states and reflects, in part, bone resorption. In primary hyperparathyroidism, bone sialoprotein levels are elevated and correlate with urinary pyridinoline and deoxypyridinoline.61 Although their measurement is not mandatory for the evaluation of the patient with primary

hyperparathyroidism, bone marker levels can provide useful information about the extent to which the skeleton is involved in the hyperparathyroid process. PARATHYROID HORMONE MEASUREMENT The diagnosis of primary hyperparathyroidism is substantiated by measurement of the concentration of PTH. For most patients with primary hyperparathyroidism, assays show frankly elevated levels of PTH associated with the sine qua non of the disease, hypercalcemia. Fragment assays for circulating midregion or COOH-terminal forms of PTH reliably show elevations by radioimmunoassay. Circulating intact PTH can be measured by immunoradiometric or immunochemiluminometric assays (Fig. 58-4; see Chap. 51).62,63 Elevated levels of circulating, intact PTH occur in b90% of patients with primary hyperparathyroidism. The remaining 10% of patients have levels that are in the upper range of normal, concentrations that are inappropriately high in the face of hypercalcemia. If these patients with normal levels of PTH are monitored over time, they invariably show elevated levels.

FIGURE 58-4. Comparison of midregion radioimmunoassay and immunoradiometric assay for parathyroid hormone (PTH) in primary hyperparathyroidism and hypercalcemia associated with malignancy. (From Nussbaum SR, Zahradnik RJ, Lavigne JR, et al. Highly sensitive two-site immunoradiometric assay of parathyrin and its clinical utility in evaluating patients with hypercalcemia. Clin Chem 1987; 33:1364.)

In addition to helping to establish the diagnosis of primary hyperparathyroidism, the assays for PTH help to differentiate it from the other most common cause of hypercalcemia, malignancy-associated hypercalcemia. In malignancy-associated hypercalcemia, even that associated with the production of parathyroid hormonerelated protein (PTHrP), PTH levels are suppressed. This is because PTHrP is not recognized in assays for PTH (see Chap. 52). In virtually all other hypercalcemic states, the parathyroid glands are normally suppressed, and levels of PTH are not detected in the circulation. OTHER TESTS As a reflection of the increased concentration of biologically active PTH in the circulation of patients with primary hyperparathyroidism, total and nephrogenous urinary cyclic adenosine monophosphate (cAMP) levels are elevated.64 However, with clinically proven assays for PTH, urinary cAMP measurement does not offer any additional information. Urinary calcium excretion is not useful for diagnostic purposes but does help in the overall assessment of the patient and in the process of decision making for surgery. Serum 1,25-dihydroxyvitamin D [1,25(OH)2D] may be elevated in primary hyperparathyroidism, reflecting an action of PTH to stimulate the conversion of 25-hydroxyvitamin D [25(OH)D] to 1,25(OH)2D. Further insight into the functional abnormality of primary hyperparathyroidism derives from the observation that patients with this disorder convert more 25(OH)D to 1,25(OH)2D than do normal people.65 The level of 1,25(OH)2D in primary hyperparathyroidism may reflect the actions of PTH on the renal 1a-hydroxylase enzyme. However, the elevated 1,25(OH)2D level in primary hyperparathyroidism is not specific. Many patients with primary hyperparathyroidism do not show elevations in this vitamin D metabolite, and other disorders associated with hypercalcemia (e.g., certain granulomatous diseases, lymphomas, sarcoidosis, vitamin D toxicity) may be associated with elevations in the 1,25(OH)2D concentration. Patients with primary hyperparathyroidism demonstrate characteristic histopathologic changes of the hyperparathyroid state on bone biopsy (see Chap. 55). Presently, bone biopsy has little clinical utility in management of this disease.

DIAGNOSIS
The diagnostic hallmarks of primary hyperparathyroidism are hypercalcemia and an elevated level of PTH. Just as a normal circulating PTH concentration is a relatively unusual finding in primary hyperparathyroidism, a normal level of serum calcium is also most unusual in primary hyperparathyroidism. When this occurs, the circulating serum proteins may be low, and the normal serum calcium value may be associated with an elevated ionized calcium concentration (see Chap. 49). Only in this unusual situation of suspected primary hyperparathyroidismelevated PTH and normal serum calciummay ionized serum calcium values be a valuable adjunct to other diagnostic tests. Nonetheless, primary hyperparathyroidism is seen occasionally with a truly normal serum calcium concentration. The name normocalcemic primary hyperparathyroidism was given to this clinical presentation; over time, however, the serum calcium level does become elevated.66 Presumably, this situation reflects increased parathyroid glandular activity that has not yet caused frank hypercalcemia. Rarely can primary hyperparathyroidism be shown to be present in an individual for whom serum calcium determinations are always normal. In a small number of patients with primary hyperparathyroidism, the circulating PTH concentration is not frankly elevated. Values are instead in the upper range of normal. The PTH level is, however, inappropriately high given the patient's hypercalcemia. Any patient with hypercalcemia should have suppressed levels of PTH unless primary hyperparathyroidism is present. Exceptions to this rule include patients with FHH and those who are taking thiazide diuretics or lithium. Some data also suggest that younger individuals (i.e., younger than 45 years of age) may normally have lower levels of PTH than their older counterparts.67 Thus, a value at the upper end of the normal range in a 40-year-old patient with hypercalcemia is probably indicative of hyperparathyroidism. If the serum calcium level is consistently normal when the PTH concentration is elevated, the physician must consider a secondary cause for the hyperparathyroid state. The distinction between primary and secondary hyperparathyroidism is made readily. Parathyroid glandular overactivity in secondary hyperparathyroidism has a nonparathyroid cause, a condition associated with a depressed serum calcium level. Parathyroid tissues respond to this hypocalcemic stimulus with an increased secretory rate. The increased PTH concentration returns the serum calcium level toward normal. Secondary hyperparathyroidism has many causes (see Chap. 61 and Chap. 63), but usually it is related to renal insufficiency or to gastrointestinal tract disorders in which malabsorption of vitamin D, calcium, or both occurs. All four parathyroid glands respond to hypocalcemia with the development of hyperplasia. In renal disease, if the hypocalcemic signal is uncontrolled and prolonged, the parathyroid tissue may become relatively autonomous, a condition called tertiary hyperparathyroidism, in which extreme elevations of PTH, sometimes >10 times normal, occur with hypercalcemia. Before determination of PTH became the most useful test in the differential diagnosis of hypercalcemia, several other tests and measurements were used that now have only historic interest. These tests include the decreased basal tubular reabsorption of phosphate, the lack of effect of exogenous PTH on the tubular reabsorption of phosphate, an increased chloride/bicarbonate ratio, and the failure to suppress serum calcium with prednisone. None of these maneuvers is of diagnostic value and they are not used to establish the diagnosis of primary hyperparathyroidism.

DIFFERENTIAL DIAGNOSIS
The differential diagnosis of hypercalcemia is covered in Chapter 59, but several points are particularly relevant to the discussion of primary hyperparathyroidism. If previous medical records are available, the patient may be found to have had serum calcium levels at the upper limits of normal before frank hypercalcemia became evident. LITHIUM USE Lithium administration has been associated with hypercalcemia and an apparent hyperparathyroid state. In vitro studies suggest that lithium may alter the calcium

setpoint for calcium-mediated inhibition of PTH secretion.68 Anecdotal clinical reports describe reversible hyperparathyroidism in patients treated with lithium. Confounding variables, such as diuretic therapy or renal failure, preclude definite conclusions in many of these case reports. Lithium treatment has been associated with hypercalcemia, hypermagnesemia, and reduced urinary calcium.69 A controlled study involving normal volunteers assessed lithium effects on PTH secretion. No significant difference was found in serum calcium, plasma PTH, or nephrogenous cAMP measurements after calcium infusion in normal volunteers, off or on lithium therapy.70 These data suggest that clinically relevant lithium levels do not alter the calcium setpoint of PTH release, but evidence suggests that the setpoint for calcium may be altered short term and that PTH levels may rise over time.71 Practically, the diagnosis of primary hyperparathyroidism is on firmer grounds if lithium can be withdrawn and the patient shown to have persistent hypercalcemia over the ensuing several months. THIAZIDE USE Another medication in common use, thiazide diuretics, may be associated with hypercalcemia.72 The mechanism for hypercalcemia is related to several factors: reduced plasma volume, increased proximal tubular reabsorption of calcium, and perhaps activity at the level of the parathyroid glands themselves. Some patients who develop hypercalcemia in association with thiazide diuretic therapy are ultimately shown to have primary hyperparathyroidism. In these patients, the hypercalcemia persists after thiazides are withdrawn. Other patients who develop hypercalcemia during thiazide therapy show a return to normal of the serum calcium and serum PTH levels. The diagnosis of primary hyperparathyroidism cannot be made with confidence in hypercalcemic patients receiving thiazides unless the calcium and parathyroid levels are still elevated 2 to 3 months after the diuretic is discontinued. COEXISTENCE OF TWO CAUSES OF HYPERCALCEMIA Because primary hyperparathyroidism is a relatively common disorder, it may coexist with another disorder associated with hypercalcemia, such as malignancy.

THERAPY
SURGERY In view of the modern clinical profile of primary hyperparathyroidism, not all patients have features that would prompt a recommendation for surgery. The existence of asymptomatic patients with primary hyperparathyroidism has engendered controversy about the need for parathyroidectomy and surgical guidelines. At the Consensus Development Conference on the Diagnosis and Management of Asymptomatic Primary Hyperparathyroidism, a set of surgical guidelines was endorsed73 (Table 58-1). The criteria for surgery include serum calcium concentration >12 mg/dL; any complication of primary hyperparathyroidism (e.g., overt bone disease, nephrolithiasis, nephrocalcinosis, classic neuromuscular disease); marked hypercalciuria (>400 mg per day); reduction in bone density more than two standard deviations below normal at the site of cortical bone, as in the forearm; an episode of acute hyperparathyroidism; and age younger than 50 years (see Table 58-1). Approximately 50% of patients with primary hyperparathyroidism meet one or more of these criteria. This is a significantly greater percentage of patients that have symptomatic primary hyperparathyroidism (2030%). Some patients with asymptomatic primary hyperparathyroidism do meet surgical criteria and are candidates for parathyroid surgery, and in individual cases, surgical guidelines may be altered according to clinical judgment.

TABLE 58-1. Criteria for Surgery in Primary Hyperparathyroidism

Parathyroidectomy leads to the rapid resolution of the biochemical abnormalities of primary hyperparathyroidism.74 Surgery has also been documented to be of clear benefit in reducing the incidence of recurrent nephrolithiasis75,76 and in leading to an improvement in bone mineral density.74 Parathyroidectomy leads to a 12% rise in bone density mainly at the cancellous lumbar spine and femoral neck (Fig. 58-5). This increase is sustained over at least four years after surgery.

FIGURE 58-5. Mean (SE) bone mineral density at three sites in two groups of patients with primary hyperthyroidism. Cumulative percentage change (mean SEM) from baseline by site at year 1, year 4, year 7, and year 10 of follow-up, reported in patients followed with no intervention (hatched bars) and after parathyroidectomy (solid bars). Differences between parathyroidectomy and no intervention groups are shown. (From Silverberg SJ, Gartenberg F, Jacobs TP, et al. Increased bone mineral density following parathyroidectomy in primary hyperparathyroidism. J Clin Endocrinol Metab 1995; 80:729.

A few patients show a densitometric profile that is unusual for patients with asymptomatic primary hyperparathyroidism. These individuals do not show the usual sparing of cancellous bone; rather, they have low bone density at the spine. In these patients, the postoperative increase in vertebral bone density is even more dramatic than that seen in the average patient; this has led to the recommendation that patients with low vertebral bone density also should be considered for parathyroidectomy.77 Special mention should be made of surgery in postmenopausal women with primary hyperparathyroidism.78 Many postmenopausal women with primary hyperparathyroidism have bone mass of the lumbar spine that is not below normal. Bone loss in the cancellous spine, typical of the postmenopausal state, may not be apparent. In these patients, the hyperparathyroid state could conceivably afford a relative protection, and the physician could use this information to proceed with a more conservative approach to management, especially if other guidelines for surgery are not met. On the other hand, the increase in bone density after surgery (at the spine and hip) is also seen in postmenopausal women.74 Issues related to parathyroid surgery and to preoperative localization of parathyroid tissue are also covered in Chapter 62.79 NONSURGICAL MEDICAL APPROACHES In patients who are not candidates for surgery at the time primary hyperparathyroidism is recognized, the course over the next 10 years often is stable (see Fig.

58-5).74,80 Serum calcium and PTH levels, urinary calcium excretion, and other biochemical indices do not appear to show any changes or trends over time. Bone mineral densitometry at all three sites (i.e., forearm, hip, lumbar spine) does not appear to show any unusual changes or trends over time.81,82 Fractures might be expected to be increased in patients followed conservatively, because of the reduction in bone mineral density typically seen in this disease. In fact, an increased incidence of distal radial fractures would be consistent with the known selective effects of PTH to reduce cortical bone mass. Despite publication of a few reports, no conclusion can be drawn as to whether fracture incidence is increased in primary hyperparathyroidism or whether these potential adverse events show site or time dependence.83,84 Relatively little is known about survival in patients who develop primary hyperparathyroidism. The indolent nature of the disease, as well as the difficulty in long-term follow-up, account, in part, for lack of pertinent information about mortality in this disease. However, the Mayo Clinic's review of the records of >400 patients with primary hyperparathyroidism for 30 years indicates that survival, on average, is indistinguishable from the expected longevity from life tables.85 Patients with hypercalcemia in the highest quartile (Ca2+ of 11.216.0 mg/dL) may have had higher mortality; however, this becomes apparent only after 15 years. These findings are different from those of earlier studies in which an increased risk of death from cancer and cardiovascular events was reported.86,87 These differences may be attributed to the apparently much milder disease observed in the Mayo Clinic population, very much like that usually seen in the United States today. Although knowledge of the natural history of primary hyperparathyroidism managed without parathyroid surgery is still incomplete, medical approaches to the management of primary hyperparathyroidism should be considered in nonsurgical patients. Patients with acute primary hyperparathyroidism should be treated the same as any patient with severe hypercalcemia88,89 (see Chap. 59). Long-term management of chronically elevated, mild hypercalcemia centers on adequate hydration and ambulation. If possible, diuretics should be avoided; in particular, thiazide diuretics may worsen the hypercalcemia in some patients. Other diuretics, such as furosemide, may place the patient at risk for dehydration and electrolyte imbalances. General recommendations for diet are not yet certain, and rationales exist for both low- and highcalcium diets in patients with hyperparathyroidism. Diets high in calcium may suppress levels of endogenous PTH. However, high-calcium diets may lead to greater absorption of calcium because of the elevated levels of 1,25(OH)2D in some patients.90,91 The recommendation for a low-calcium diet is based on the notion that less calcium is available for absorption. However, low-calcium diets may predispose patients to further stimulation of endogenous PTH levels. One study showed no effect of dietary calcium on biochemical indices or bone densitometry in patients with primary hyperparathyroidism.92 A normal calcium intake can be followed without adverse effects, except in those patients with elevated 1,25-dihydroxyvitamin D levels. Such patients are advised to be more moderate in their calcium intake to prevent hypercalciuria. Other approaches to the medical management of primary hyperparathyroidism have been considered. Attempts to block PTH secretion with b-adrenergic inhibitors or H2-receptor antagonists have not been successful.7 Oral phosphate, which has been used for many years in primary hyperparathyroidism, lowers serum calcium by 0.5 to 1.0 mg/dL in most patients. The average dosage is 1 to 2 g daily in divided doses. Phosphate appears to have several mechanisms of action. It may inhibit calcium absorption from the gastrointestinal tract; it prevents calcium mobilization from bone; and it may also impair the production of 1,25(OH)2D. The risks of the long-term use of phosphate in the management of patients with primary hyperparathyroidism are unknown. One concern is the possibility of ectopic calcification in soft tissues when the normal solubility product of Ca2+ PO43 is exceeded (normally ~40). Phosphate therapy is contraindicated when renal insufficiency or hyperphosphatemia is present. If phosphate is to be used, the serum levels of calcium and phosphate should be monitored at regular intervals. Moreover, the long-term use of phosphate in patients with primary hyperparathyroidism promotes a further increase in PTH.93 The possibility that some of the symptoms and signs of primary hyperparathyroidism are caused by PTH itself and not by hypercalcemia raises questions about further increasing PTH levels in conjunction with phosphate therapy. If sufficient concern exists about lowering the serum calcium level in patients with asymptomatic primary hyperparathyroidism, parathyroid surgery remains the treatment of choice. Estrogen therapy has been proposed as a means of lowering serum calcium, especially because prevalence of primary hyperparathyroidism among women is increased in the post-menopausal years. Estrogens have well-known but not clearly understood antagonist actions on PTH-induced bone resorption. The serum calcium does tend to fall by ~0.5 to 1.0 mg/dL in women receiving estrogens.94,95 and 96 However, PTH and phosphorus levels do not change. More studies are required to better delineate the role of estrogen therapy. To date, no available data exist on the role of selective estrogen-receptor modulators (SERMs) in the treatment of primary hyperparathyroidism. Because the antiresorptive actions of these drugs are similar to those of estrogens, SERMs might be predicted to lower serum calcium levels in postmenopausal women with primary hyperparathyroidism in a fashion similar to that seen with estrogen. This hypothesis remains to be tested. Calcitonin may have a potential use in treating primary hyperparathyroidism, but no controlled trial has been conducted to test the efficacy of calcitonin for management of this condition. Data from the use of calcitonin in other states of increased bone turnover indicate that it may never become a useful long-term therapy for primary hyperparathyroidism. The bisphosphonates represent a class of important antiresorbing agents that may emerge as a useful approach to the medical management of primary hyperparathyroidism. Oral etidronate is not useful, and the effect of oral clodronate and alendronate is limited in duration of effect.97 Of the third-generation bisphosphonates, risedronate has been shown to have some efficacy in preliminary studies of patients with primary hyperparathyroidism.98 Finally, specifically targeted medical therapy for primary hyperparathyroidism is under active investigation. Calcimimetic agents that target the calcium-sensing receptor on the parathyroid cell99 have shown early promise in animal and in vitro studies.100,101 Data from human investigations are also encouraging. In early studies, one such calcimimetic has been shown to lower serum calcium and PTH levels both in a patient with parathyroid carcinoma and in a group of postmenopausal women with mild primary hyperparathyroidism.102,103

PRIMARY HYPERPARATHYROIDISM DURING PREGNANCY

Rarely, primary hyperparathyroidism becomes evident during pregnancy.104 Hyperparathyroidism during pregnancy used to be associated with an increased incidence of fetal death. Perinatal and neonatal complications were also thought to be increased in the hypocalcemic infant whose endogenous PTH production is suppressed by maternal hypercalcemia. Neonatal hypocalcemia and tetany can be the first sign of primary hyperparathyroidism in the mother. Although systematically collected and controlled data are unavailable, experience has indicated that primary hyperparathyroidism during pregnancy can be managed successfully without resorting to surgery.105,106 The management of primary hyperparathyroidism during pregnancy is controversial.107 Some clinicians advocate surgery during the second trimester to reduce fetal risk of chronic hypercalcemia during the gestational period. Others advocate a much more conservative approach.

PARATHYROID CARCINOMA
Parathyroid carcinoma is a rare form of primary hyperparathyroidism.108,109 These patients often have hypercalcemia that is more severe than that usually seen in primary hyperparathyroidism. Serum calcium values >14 mg/dL may suggest a parathyroid malignancy. However, parathyroid carcinoma is uncommon, occurring in <0.5% of patients with primary hyperparathyroidism. Making the clinical diagnosis of parathyroid carcinoma before surgery is often difficult. Along with the high serum calcium level, the PTH level tends to be markedly elevated. Patients with parathyroid carcinoma often have coexistent complications related to the skeleton and the kidney. This combination is unusual in benign primary hyperparathyroidism. Another striking feature may be a palpable neck mass. The clinical features of palpable neck mass, marked hypercalcemia, high concentrations of PTH, and coexistent skeletal resorption and kidney stones should alert the clinician to the possibility of parathyroid carcinoma.110 The diagnosis of parathyroid carcinoma may not be made until the time of surgery. The gross appearance of the parathyroid tissue may indicate the diagnosis. At times, the histologic features of parathyroid carcinoma may be typical (see Chap. 48), or it may be difficult to diagnose. Surgery is the mainstay of therapy for parathyroid carcinoma. With the clinical presentation of benign disease changing, the case may well be that parathyroid cancer also will be diagnosed much earlier in its natural history. Patients with parathyroid carcinoma might be expected to show biochemical evidence for somewhat more active disease than seen usually in the benign disorder, but not like the older classical descriptions. The diagnosis would rest primarily with the histologic examination of the parathyroid tissue removed at the time of surgery.111 The management of parathyroid carcinoma that has metastasized to local structures or to more distant sites, such as liver or lung, is difficult. Because this malignancy causes symptoms and signs resulting from the hypercalcemia and not necessarily from enlargement of the tumor, surgery has a role in the management of metastatic disease. Removal of local metastases or nodules in the lung has been associated with prolonged survival. Because parathyroid carcinoma tissue is not responsive to feedback of ambient hypercalcemia, debulking the tumor with removal of functional tissue is helpful in controlling hypercalcemia. When the disease is no longer amenable to surgical intervention, management becomes limited to the control of hypercalcemia. In the long run, hypercalcemia is responsible for the morbidity and mortality of the disease. However, parathyroid carcinoma is an indolent malignancy, and patients often survive for years after the diagnosis is made. The probability of 5-year survival in two series approached 60%.110

As mentioned earlier (see Nonsurgical Medical Approaches), calcimimetic agents may hold some promise for the future in the treatment of this disease. One patient was treated with such an agent to control life-threatening hypercalcemia with good results.103 Whether calcimimetic agents would have any efficacy in treating the tumor per se, or only in controlling its manifestations (hypercalcemia), is unknown. Finally, one report describes immunotherapy in the treatment of parathyroid carcinoma.112 One patient was treated with autoantibodies raised against PTH, with a resulting decrease in hypercalcemia.

FAMILIAL HYPOCALCIURIC HYPERCALCEMIA

FHH is a benign disorder associated with hypercalcemia.113 Differentiating individuals with FHH from those with primary hyperparathyroidism is extremely difficult. This disorder is a familial disease that has an autosomal dominant pattern of transmission. Penetrance is high, and the hypercalcemia often manifests in affected individuals in childhood. The disease is characterized by hypercalcemia and relative hypocalciuria. The hypocalciuria may be marked, but it is not absolutely discriminatory when compared with urinary calcium levels in patients with primary hyperparathyroidism. Subtotal parathyroidectomy does not cure the hypercalcemia. Laboratory findings for patients with FHH include mild hypercalcemia but normal serum levels of phosphate and 1,25(OH)2 D. The PTH concentration is usually normal, although in several cases, it has been elevated. The serum magnesium concentration is normal to moderately elevated. Typically, the renal calcium clearance/creatinine clearance ratio is <0.01. Family screening is often necessary to differentiate FHH from primary hyperparathyroidism. PATHOPHYSIOLOGY The pathophysiologic basis of FHH is not clear. The normal serum PTH level and urinary cAMP level suggest that this is not a disorder of parathyroid hyperactivity. Provocative tests to raise the calcium further lead to responses that suggest a set-point error. 5 In this respect, patients with FHH are different from patients with primary hyperparathyroidism. Mutations in the calcium-sensing receptor (CaR) have clearly been shown to be the cause of FHH in almost all cases. Many families with FHH have now been identified who have inactivating mutations in the CaR.114,115 and 116 The kindreds generally harbor unique heterozygous mutations, most of which are missense. These mutations can be found in any region of the receptor. When there are two mutations, the disorder presents in infancy as neonatal severe hypercalcemia. Although most patients with FHH express a defect in the CaR, which is located on 3q13-q21,117 two other genes for FHH have been described. In one family, the disorder was traced to the short arm of chromosome 19.118 In another kindred, the disorder is characterized by progressive elevations in PTH, hypophosphatemia, and osteomalacia without assignment to either chromosome 3 or 19.119 MANIFESTATIONS AND THERAPY Familial hypocalciuric hypercalcemia is not associated with signs and symptoms of primary hyperparathyroidism. Nephrolithiasis and evidence for hyperparathyroid bone disease are not found, nor do other potential complications of primary hyperparathyroidism occur. The disorder usually appears to be relatively benign. It may be associated with pancreatitis, hypertension, gallstones, and chondrocalcinosis, but the pathophysiologic bases of these relationships are unclear. Because of the benign natural history of FHH and because subtotal parathyroidectomy does not cure the disorder, these patients should not undergo neck exploration. After the diagnosis is made, the wise course is to observe these patients but not to engage in any aggressive intervention. Family screening is also recommended. CHAPTER REFERENCES
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. Heath H III, Hodgson SF, Kennedy MA. Primary hyperparathyroidism: incidence, morbidity, and potential economic impact in a community. N Engl J Med 1980; 302:189. Mundy GR, Cove DH, Fisken R. Primary hyperparathyroidism: changes in the pattern of clinical presentation. Lancet 1980; 1:1317. Wermers RA, Khosla S, Atkinson EJ, et al. The rise and fall of primary hyperparathyroidism in Rochester, Minnesota; a population based study. Ann Intern Med 1997; 126:433. Brown EM, Broadus AE, Brennan MF, et al. Direct comparison in vivo and in vitro of suppressibility of parathyroid function by calcium in primary hyperparathyroidism. J Clin Endocrinol Metab 1979; 48:604. Khosla S, Ebeling PR, Firek AF, et al. Calcium infusion suggests a set-point abnormality of parathyroid gland function in familial benign hypercalcemia and more complex disturbances in primary hyperparathyroidism. J Clin Endocrinol Metab 1993; 76:715. Friedman E, Larsson C, Amorosi A, et al. Multiple endocrine neoplasia type I: pathology, pathophysiology, molecular genetics, and differential diagnosis. In: Bilezikian JP, Levine MA, Marcus R, eds. The parathyroids. New York: Raven Press, 1994:647. Heath H III. Biogenic amines and the secretion of parathyroid hormone and calcitonin. Endocr Rev 1980; 1:319. Insogna KL, Mitnick ME, Stewart AS, et al. Sensitivity of the parathyroid hormone1,25-dihydroxyvitamin D axis to variations in calcium intake in patients with primary hyperparathyroidism. N Engl J Med 1985; 313:1126. Melton LJ. Therapeutic radiation and hyperparathyroidism. A case-control study in Rochester, Minn. Arch Intern Med 1989; 149:1887. Parfitt AM, Braunstein GD, Katz A. Radiation-associated hyperparathyroidism: comparison of adenoma growth rates, inferred from weight and duration of latency, with prevalence of mitosis. J Clin Endocrinol Metab 1993; 77:1318. Wassif WS, Moniz CF, Friedman E, et al. Familial isolated hyperparathyroidism: a distinct genetic entity with an increased risk of parathyroid cancer. J Clin Endocrinol Metab 1993; 77:1485. Metz DC, Jensen RT, Bale AE, et al. Multiple endocrine neoplasia type I: clinical features and management. In: Bilezikian JP, Levine MA, Marcus R, eds. The parathyroids. New York: Raven Press, 1994:591. Teh BT, Farnebo F, Kristoffersson U, et al. Autosomal dominant primary hyperparathyroidism and jaw tumor syndrome associated with renal hamartomas: linkage to 1q21-q32 and loss of the wild-type allele in renal hamartomas. J Clin Endocrinol Metab 1996; 81:4204. Teh BT, Kytola S, Farnebo F, et al. Mutation analysis of the MEN1 gene in MEN1, familial acromegaly and familial isolated hyperparathyroidism. J Clin Endocrinol Metab 1998; 83:2621. Teh BT, Farnebo F, Twigg S, et al. Familial isolated hyperparathyroidism maps to 1q21-q32 in a subset of families. J Clin Endocrinol Metab 1998; 83:2114. Rosenberg CL, Motokura T, Kronenberg HM, Arnold A. Coding sequence of the overexpressed transcript of the putative oncogene PRADI/cyclin D1 in two primary human tumors. Oncogene 1993; 8:519. Hsi ED, Zukerberg LR, Yang W-I, Arnold A. Cyclin D1/PRAD 1 expression in parathyroid adenomas: an immunohistochemical study. J Clin Endocrinol Metab 1996; 81:1736. Chandrasekharappa SC, Guru SC, Manickam P, et al. Position cloning of the gene for multiple endocrine neoplasia type 1. Science 1997; 276:404. European Consortium on MEN I Gene. Hum Mol Genet 1997; 6:1177. Heppner C, Kester MB, Agarwal SK, et al. Somatic mutations of the MEN 1 gene in parathyroid tumors. Nat Genet 1997; 16:375. Farnebo F, Teh BT, Kytola S, et al. Alterations of the MEN1 gene in sporadic parathyroid tumors. J Clin Endocrinol Metab 1998; 83:2627. Carling T, Correa P, Hessman O, et al. Parathyroid MEN 1 gene mutations in relation to clinical characteristics of nonfamilial primary hyperparathyroidism. J Clin Endocrinol Metab 1998; 83:2960. Tahara H, Smith AP, Gaz RD, et al. Genomic localization of novel candidate tumor suppressor gene loci in human parathyroid adenomas. Cancer Res 1996; 56:599. Tahara H, Smith AP, Gaz RD, Arnold A. Loss of chromosome arm 9p DNA and analysis of the p16 and p15 cyclindependent kinase inhibitor genes in human parathyroid adenomas. J Clin Endocrinol Metab 1996; 81:3663. Brown EM, Steinham B. In vivo and in vitro characterization of neonatal hyperparathyroidism resulting from a de novo, heterozygous mutation in the CaSR gene: normal maternal calcium as a cause of hyperparathyroidism in familial hypocalciuric hyperparathyroidism. J Clin Invest 1997; 99:88. Pollack MR, Seidman CE, Brown EM. Three inherited disorders of calcium sensing. Medicine 1996; 75:115. Hosokawa Y, Pollak MR, Brown EM, Arnold A. Mutational analysis of the extracellular Ca2+-sensing receptor gene in human parathyroid tumors. J Clin Endocrinol Metab 1997; 80:3107. Kifor O, Moore FD Jr, Wang P, et al. Reduced immunostaining for the extracellular Ca2+-sensing receptor in primary and uremic secondary hyperparathyroidism. J Clin Endocrinol Metab 1996; 81:1598. Gogusev J, Duchambon P, Hory B, et al. Depressed expression of calcium receptor in parathyroid gland tissue of patients with primary hyperparathyroidism. Kidney Int 1997; 51:328. Bilezikian JP, Silverberg SJ, Gartenberg F, et al. Clinical presentation of primary hyperparathyroidism. In: Bilezikian JP, Marcus R, Levine MA, eds. The parathyroids. New York: Raven Press, 1994:457. Parisien M, Silverberg SJ, Shane E, et al. The histomorphometry of bone in primary hyperparathyroidism: preservation of cancellous bone. J Clin Endocrinol Metab 1990; 70:930. Parisien M, Mellish RW, Silverberg SJ, et al. Maintenance of cancellous bone connectivity in primary hyperparathyroidism: trabecular and strut analysis. J Bone Miner Res 1992; 7:913. Parisien M, Cosman F, Mellish RWE, et al. Bone structure in postmenopausal hyperparathyroid, osteoporotic and normal women. J Bone Min Res 1995; 10:1393. Mallette LE, Bilezikian JP, Heath DA, Aurbach GD. Hyperparathyroidism: a review of 52 cases. Medicine (Baltimore) 1974; 53:127. Silverberg SJ, Shane E, Jacobs TP, et al. Nephrolithiasis and bone involvement in primary hyperparathyroidism. Am J Med 1990; 89:327. Patten BM, Bilezikian JP, Mallette LE, et al. The neuromuscular disease of hyperparathyroidism. Ann Intern Med 1974; 80:182. Turken SA, Cafferty M, Silverberg SJ, et al. Neuromuscular involvement in mild, asymptomatic primary hyperparathyroidism. Am J Med 1989; 87:553. Linos DA, Van Heerden JA, Abboad CF, Edis AJ. Primary hyperparathyroidism and peptic ulcer disease. Arch Surg 1978; 113:384. Scholz DA. Hypertension and hyperparathyroidism. Arch Intern Med 1977; 137:1123. Bilezikian JP, Aurbach GD, Connor TB, et al. Pseudogout following parathyroidectomy. Lancet 1973; 1:445. Mallette LE. Anemia in hypercalcemic hyperparathyroidism: renewed interest in an old observation. Arch Intern Med 1977; 137:572. Ljunghall S, Jakobsson S, Joborn C, et al. Longitudinal studies of mild primary hyperparathyroidism. J Bone Miner Res 1991; 6(Suppl 2):S111. Solomon BL, Schaaf M, Smallridge RC. Psychologic symptoms before and after parathyroid surgery. Am J Med 1994; 96:101. Kleerekoper M, Bilezikian JP. Parathyroidectomy for nontraditional features of primary hyperparathyroidism. Am J Med 1994; 96:99. Hamby LS, Lee EY, Schwartz RW. Parathyroid adenoma and gastric carcinoma as manifestations of Cowden's disease. Surgery 1995; 118:115. Mallette LE. The functional and pathologic spectrum of parathyroid abnormalities in hyperparathyroidism. In: Bilezikian JP, Levine MA, Marcus R, eds. The parathyroids. New York: Raven Press, 1994:423. Potts JT. Proceedings of the NIH Consensus Development Conference on Diagnosis and Management of Asymptomatic Primary Hyperparathyroidism. J Bone Miner Res 1991; 6(Suppl 2):S1. Fitzpatrick LA, Bilezikian JP. Acute primary hyperparathyroidism: a review of 48 patients. Am J Med 1987; 82:275. Huang SM, Duh QY, Shaver J, et al. Familial hyperparathyroidism without MEN. World J Surg 1997; 21:22. Barry MK, van Heerdon JA, Grant CS, et al. Is familial hyperparathyroidism a unique disease? Surgery 1997; 122:1028. Mallette LE, Malini S, Rappaport MP, Kirkland JL. Familial cystic parathyroid adenomatosis. Ann Intern Med 1987; 107:54. Moss DW. Perspectives in alkaline phosphatase research. Clin Chem 1992; 38:2486. Silverberg SJ, Deftos LJ, Kim T, Hill CS. Bone alkaline phosphatase in primary hyperparathyroidism. J Bone Mineral Res 1991; 6:A624. Duda RJ, O'Brien JF, Katzman JA, et al. Concurrent assays of circulating bone Gla-protein and bone alkaline phosphatase: effects of sex, age, and metabolic bone disease. J Clin Endocrinol

Metab 1988; 5:1. 55. Price PA, Parthemore JG, Deftos LJ. New biochemical marker for bone metabolism. Measurement by radioimmunoassay of bone Gla-protein in the plasma of normal subjects and patients with bone disease. J Clin Invest 1980; 66:878. 56. Deftos LJ, Parthemore JG, Price PA. Changes in plasma bone Gla-protein during treatment of bone disease. Calcif Tissue Int 1982; 34:121. 57. Deftos LJ. Bone protein and peptide assays in the diagnosis and management of skeletal disease. Clin Chem 1991; 37:1143. 58. Delmas PH. Biochemical markers of bone turnover: methodology and clinical use in osteoporosis. Am J Med 1991; 91:169. 59. Seibel MJ, Gartenberg F, Silverberg SJ, et al. Urinary hydroxypyridinium cross-links of collagen in primary hyperparathyroidism. J Clin Endocrinol Metab 1992; 74:481. 60. Kraenzlin ME, Lau KHW, Liang L, et al. Development of an immunoassay for human serum osteoclastic tartrate-resistant acid phosphatase. J Clin Endocrinol Metab 1990; 71:442. 61. Seibel MJ, Woigte HW, Pecherstorfer M, et al. Serum immunoreactive bone sialoprotein as a new marker of bone turnover in metabolic and malignant diseases. J Clin Endocrinol Metab 1996; 81:3289. 62. Kao PC, van Heerden JA, Grant CS, et al. Clinical performance of parathyroid hormone immunometric assays. Mayo Clin Proc 1992; 67:637. 63. Nussbaum SR, Potts JT Jr. Advances in immunoassays for parathyroid hormone: clinical applications to skeletal disorders of bone and mineral metabolism. In: Bilezikian JP, Levine MA, Marcus R, eds. The parathyroids. New York: Raven Press, 1994:157. 64. Broadus AE. Nephrogenous cyclic AMP as a parathyroid function test. Nephron 1979; 23:136. 65. Locascio V, Adami S, Galvanini G, et al. Substrate-product relation of 1-hydroxylase activity in primary hyperparathyroidism. N Engl J Med 1985; 313:1123. 66. Wills MR, Pak CYC, Hammond WG, Bartter FC. Normocalcemic primary hyperparathyroidism. Am J Med 1969; 47:384. 67. Segre GV. Personal communication. 68. Brown E. Lithium induces abnormal calcium-regulated PTH release in dispersed bovine parathyroid cells. J Clin Endocrinol Metab 1981; 52:1046. 69. Christiansen C, Baastrup P, Transbol I. Development of primary hyperparathyroidism during lithium therapy: longitudinal study. Neuropsychobiology 1980; 6:280. 70. Spiegel A, Rudorfer M, Marx S, Linnoila M. The effect of short term lithium administration on suppressibility of parathyroid hormone secretion by calcium in vivo. J Clin Endocrinol Metab 1984; 59:354. 71. Mallette LE, Khouri K, Zengotita H, et al. Lithium treatment increases intact and midregion parathyroid hormone and parathyroid volume. J Clin Endocrinol Metab 1989; 68:654. 72. Middler S, Pak CY, Murad F, Bartter F. Thiazide diuretics and calcium metabolism. Metabolism 1973; 22:1438. 73. National Institutes of Health Consensus Development Conference statement on primary hyperparathyroidism. J Bone Miner Res 1991; 6(Suppl 2):S9. 74. Silverberg SJ, Gartenberg F, Jacobs TP, et al. Increased bone mineral density following parathyroidectomy in primary hyperparathyroidism. J Clin Endocrinol Metab 1995; 80:729. 75. Klugman VA, Favus M, Pak CYC. Nephrolithiasis in primary hyperparathyroidism. In: Bilezikian JP, Levine MA, Marcus R, eds. The parathyroids. New York: Raven Press, 1994:505. 76. Deaconson TF, Wilson SD, Lemann J. The effect of parathyroidectomy on the recurrence of nephrolithiasis. Surgery 1987; 215:241. 77. Silverberg SJ, Locker FG, Bilezikian JP. Vertebral osteopenia: a new indication for surgery in primary hyperparathyroidism. J Clin Endocrinol Metab 1996; 81:4007. 78. Bilezikian JP. Primary hyperparathyroidism: another important metabolic bone disease of women. J Women Health 1994; 3:21. 79. Thule P, Thakore K, Vansant J, et al. Preoperative localization of parathyroid tissue with technetium-99m Sestamibi 123 I subtraction scanning. J Clin Endocrinol Metab 1994; 78:77. 80. Parfitt AM, Rao DS, Kleerekoper M. Asymptomatic primary hyperparathyroidism discovered by multichannel biochemical screening: clinical course and considerations bearing on the need for surgical intervention. J Bone Miner Res 1991; 6(Suppl 2):S97. 81. Sudhaker RD, Wilson RJ, Kleerekoper M, Parfitt AM. Lack of biochemical progression or continuation of accelerated bone loss in mild asymptomatic primary hyperparathyroidism. J Clin Endocrinol Metab 1988; 67:1294. 82. Silverberg SJ, Gartenberg F, Jacobs TP, et al. Longitudinal bone density measurements in untreated primary hyperparathyroidism. J Clin Endocrinol Metab 1995; 90:723. 83. Melton U III, Atkinson EJ, O'Fallon WM, Heath H. Risk of age-related fractures in patients with primary hyperparathyroidism. Arch Intern Med 1992; 152:2269. 84. Wilson RJ, Rao DS, Ellis B, Kleerekoper M, Parfitt AM. Mild asymptomatic primary hyperparathyroidism is not a risk factor for vertebral fractures. Ann Intern Med 1988; 109:959. 85. Wermers RA, Khosla S, Atkinson EJ, et al. Survival after the diagnosis of primary hyperparathyroidism. Am J Med 1998; 104:115. 86. Hedhack G, Tisell LE, Bengtsson BA, et al. Premature death in patients operated on for primary hyperparathyroidism. World J Surgery 1990; 14:829. 87. Palmer M, Adami HO, Krusemo UB, Ljunghall S. Increased risk of malignant diseases after surgery for primary hyperparathyroidism: a nationwide cohort study. Am J Epidemiol 1988; 127:1031. 88. Bilezikian JP. Management of acute hypercalcemia. N Engl J Med 1992; 326:1196. 89. Bilezikian JP. Management of hypercalcemia. J Clin Endocrinol Metab 1993; 77:1445. 90. Kaplan RA, Haussler MR, Deftos LJ, et al. The role of 1,25-dihydroxyvitamin D in the mediation of intestinal hyperabsorption of calcium in primary hyperparathyroidism and absorptive hypercalciuria. J Clin Invest 1977; 59:756. 91. Broadus AE, Horst RL, Lang R, et al. The importance of circulating 1,25-dihydroxyvitamin D in the pathogenesis of hypercalciuria and renal-stone formation in primary hyperparathyroidism. N Engl J Med 1980; 302:421. 92. Locker FG, Silverberg SJ, Bilezikian JP. Optimal dietary calcium intake in primary hyperparathyroidism. Am J Med 1997; 102:543. 93. Broadus AE, Magee JS, Mallette LE, et al. A detailed evaluation of oral phosphate therapy in selected patients with primary hyperparathyroidism. J Clin Endocrinol Metab 1983; 56:953. 94. Marcus R, Madvig P, Crim M, et al. Conjugated estrogens in the treatment of postmenopausal women with hyperparathyroidism. Ann Intern Med 1984; 100:633. 95. Selby PL, Peacock M. Ethinyl estradiol and norethindrone in the treatment of primary hyperparathyroidism in post menopausal women. N Engl J Med 1986; 314:1481. 96. Stock JL. Medical management of primary hyperparathyroidism. In: Bilezikian JP, Levine MA, Marcus R, eds. The parathyroids. New York: Raven Press, 1994:519. 97. Hamdy NAT, Gray E, McCloskey E, et al. Clodronate in the medical management of hyperparathyroidism. Bone 1987; 8(Suppl 1):569. 98. Reisner CA, Stone MD, Hoshiy DJ, et al. Acute changes in calcium homeostasis during treatment of primary hyperparathyroidism with risedronate. J Clin Endocrinol Metab 1993; 77:1067. 99. Brown EM, Gamba G, Riccardi D, et al. Cloning and characterization of an extracellular calcium-sensing receptor from bovine parathyroid. Nature 1993; 366:575. 100. Rogers KV, Fox J, Nemeth EF. The calcimimetic compound NPS467 reduces plasma calcium in a dose-dependent and stereospecific manner. J Bone Miner Res 1993; 8(Suppl 1):S180. 101. Steffey ME, Fox J, Van Wagenen BC, et al. Calcimimetics: structurally and mechanistically novel compounds that inhibit hormone secretion from parathyroid cells. J Bone Miner Res 1993; 8(Suppl 1):S175. 102. Silverberg SJ, Bone HG 3rd, Marriott TB, et al. Short term inhibition of parathyroid hormone secretion by a calcium receptor agonist in primary hyperparathyroidism. N Engl J Med 1997; 337(21):1506. 103. Collins MT, Skarulis MC, Bilezikian JP, et al. Treatment of hypercalcemia secondary to parathyroid carcinoma with a novel calcimimetic agent. J Clin Endocrinol Metab 1998; 83:1083. 104. Kristofferson A, Dahlgren S, Lithner F, Jarhult J. Primary hyperparathyroidism in pregnancy. Surgery 1985; 97:326. 105. Kelly TR. Primary hyperparathyroidism during pregnancy. Surgery 1991; 110:1028. 106. Haenel LC IV, Mayfield RK. Primary hyperparathyroidism in a twin pregnancy and review of fetal/maternal calcium homeostasis. Am J Med Sci 2000; 319:191. 107. Ficinski ML, Mestman JH. Primary hyperparathyroidism during pregnancy. Endocr Pract 1996; 2:362. 108. Wynne AG, van Heerden J, Carney JA, Fitzpatrick LA. Parathyroid carcinoma: clinical and pathologic features in 43 patients. Medicine (Baltimore) 1992; 71:197. 109. Kytola S, Farnebo F, Obara T, et al. Patterns of chromosomal imbalances in parathyroid carcinomas. Am J Pathol 2000; 157:579. 110. Shane E, Bilezikian JP. Parathyroid carcinoma. In: Williams CJ, Krikorian JC, Green MR, Raghavan D, eds. Textbook of uncommon cancer. New York: John Wiley and Sons, 1988:763. 111. Silverberg SJ, Lowenfeld A, Perzin K, et al. Parathyroid cancer presents with a new phenotype in the 1990's. J Bone Mineral Res 1997; 12(Suppl 1):S522. 112. Bradwell AR, Harvey TC. Control of hypercalcaemia of parathyroid carcinoma by immunisation. Lancet 1999; 353:370. 113. Heath H III. Familial benign (hypocalciuric) hypercalcemia. A troublesome mimic of mild primary hyperparathyroidism. Endocrinol Metab Clin North Am 1989; 18:723. 114. Pollak MR, Brown EM, Chou YHW, et al. Mutations in the human Ca 2+-sensing receptor gene cause familial hypocalciuric hypercalcemia and neonatal severe hyperparathyroidism. Cell 1993; 75:1297. 115. Thakker RV. Molecular genetics of parathyroid disease. Curr Opin Endocrinol Diabetes 1996; 3:521. 116. Hendy GN, D'Souza-Li L, Yang B, et al. Mutations of the calcium-sensing receptor (CASR) in familial hypocalciuric hypercalcemia, neonatal severe hyperparathyroidism, and autosomal dominant hypocalcemia. Hum Mutat 2000; 16:281. 117. Chou YHW, Brown EM, Levi T, et al. The gene responsible for familial hypocalciuric hypercalcemia maps to chromosome 3q in four unrelated families. Nat Genet 1992; 1:295. 118. Heath H III, Jackson CE, Otterud B, Leppert MF. Genetic linkage analysis in familial benign (hypocalciuric) hypercalcemia: evidence for locus heterogeneity. Am J Hum Genet 1993; 53:193. 119. Trump D, Whyte M, Wooding C, et al. Linkage studies in a kindred from Oklahoma, with familial benign (hypocalciuric) hypercalcemia (FBH) and developmental elevations in serum parathyroid hormone levels, indicate a third locus for FBH. Hum Genet 1995; 96:183.

CHAPTER 59 NONPARATHYROID HYPERCALCEMIA Principles and Practice of Endocrinology and Metabolism

CHAPTER 59 NONPARATHYROID HYPERCALCEMIA

ANDREW F. STEWART Signs and Symptoms of Hypercalcemia Differential Diagnosis of Hypercalcemia Malignancy-Associated Hypercalcemia Endocrinopathies Associated with Hypercalcemia Medications Granulomatous Disorders Immobilization Milk-Alkali Syndrome Parenteral Nutrition Familial Hypocalciuric Hypercalcemia Hypophosphatemia Renal Failure Idiopathic Hypercalcemia of Infancy Hyperproteinemia Manganese Intoxication Advanced Chronic Liver Disease Treatment of Hypercalcemia Saline Infusion Loop Diuretics Restriction of Calcium and Vitamin D Intake Calcitonin Plicamycin Glucocorticoids Phosphorus Bisphosphonates Gallium Nitrate Other Measures Chapter References

Primary hyperparathyroidism is the most common cause of hypercalcemia among nonhospitalized patients (see Chap. 58). This chapter focuses on nonparathyroid causes of hypercalcemia, the most common of which is malignancy. Ionized extracellular fluid calcium has quantitatively important interfaces with four physiologic compartments: serum proteins (predominantly albumin), the skeleton, the gastrointestinal tract, and the kidneys. A disorder at the level of any one of these compartments may cause hypercalcemia. For example, a malignancy that is metastatic to the skeleton may cause hypercalcemia primarily through skeletal resorption (resorptive hypercalcemia). Alternatively, sarcoidosis appears to lead to hypercalcemia largely because of intestinal calcium hyperabsorption (absorptive hypercalcemia). Use of thiazide diuretics results in hypercalcemia mainly because of excessive renal calcium reabsorption (renal hypercalcemia). Finally, serum protein abnormalities may lead to elevations in total, but not ionized, serum calcium (factitious hypercalcemia). Combinations of renal, absorptive, resorptive, and protein-binding components all may contribute to a given patient's hypercalcemia. For example, hypercalcemia in a dehydrated patient with sarcoidosis may include all four components. Nevertheless, the grouping and conceptualization of the types of hypercalcemia into this pathophysiologic framework aids in both diagnosing and treating individual cases of hypercalcemia.

SIGNS AND SYMPTOMS OF HYPERCALCEMIA

Hypercalcemia, through its effect on cellular sodium-potassium adenosine triphosphatase, leads to hyperpolarization of cell membranes. This hyperpolarization or refractoriness of cell membranes is particularly apparent in nervous and muscle tissue. Neurologic manifestations of hypercalcemia include the spectrum of metabolic central nervous system dysfunction, which involves apathy, drowsiness, and depression, progressing to obtundation and coma (see Chap. 200). Skeletal muscle and neurologic dysfunction may present as weakness (see Chap. 210). Smooth-muscle hyperpolarization may be manifest as bowel hypomotility and constipation. The most reproducible cardiac conduction abnormality is a shortening of the QTc interval. Hypercalcemia interferes with vasopressin-induced urinary concentrating ability (see Chap. 206 and Chap. 208). Thus, a clinical syndrome of nephrogenic diabetes insipidus is common in patients with hypercalcemia; it is manifested by polyuria, polydipsia, and dehydration. A reduced glomerular filtration rate is particularly common in patients with hypercalcemia. The azotemia is multifactorial, reflecting reductions in renal blood flow caused by nephrogenic diabetes insipidusinduced dehydration and hypercalcemia-induced afferent arteriolar vasocon-striction, as well as by nephrocalcinosis and, in some cases, obstructive uropathy resulting from nephrolithiasis. Nephrocalcinosis is particularly likely to occur in patients with nonparathyroid hypercalcemia, perhaps because their serum phosphate concentrations are normal to elevated in comparison to those of patients with primary hyperparathyroidism. Calcium phosphate salts are insoluble at alkaline pH values. Nephrocalcinosis is believed to result from calcium phosphate precipitation in the alkaline renal medulla. Calcium phosphate precipitation may occur in other alkaline environments, including the cornea (rarely observed on physical examination as band keratopathy), gastric mucosal cells, and the pancreas (leading to pancreatitis). Gastrointestinal complaints such as nausea, anorexia, vomiting, abdominal pain, and constipation are common in patients with hypercalcemia (see Chap. 204). In the older literature, hypercalcemia was reported to be associated with peptic ulcer disease, particularly when primary hyperparathyroidism was present. This apparent correlation may relate to the coexistence of gastrin-secreting islet cell tumors with parathyroid adenomas or hyperplasia (multiple endocrine neoplasia type 1 [MEN1]), to related development of the milk-alkali syndrome, or to the simple coincidence of two common syndromes. The suggestion has been made that peptic ulcer disease may result from calcium-induced gastrin secretion or gastric acid hypersecretion.

DIFFERENTIAL DIAGNOSIS OF HYPERCALCEMIA

Nonparathyroid disorders that may lead to hypercalcemia are listed in Table 59-1. The most common of these is malignancy, which is second in frequency only to primary hyperparathyroidism among ambulatory patients with hypercalcemia and is considerably more common than hyperparathyroidism among hospitalized patients with hypercalcemia. TABLE 59-1. Nonparathyroid Causes of Hypercalcemia Cause MALIGNANCY Local osteolytic hypercalcemia Humoral hypercalcemia of malignancy 1,25-dihydroxyvitamin Dmediated hypercalcemia Ectopic parathyroid hormone secretion Unusual variants ENDOCRINOPATHIES Thyrotoxicosis Pheochromocytoma Addisonian crisis VIPoma syndrome MEDICATIONS Vitamins D and A (intoxication) Lithium Reference 121 121 3234 35,36 9,37 3840 4143 44, 45 46 4754 5557

Thiazide diuretics Estrogens/antiestrogens Growth hormone Theophylline Foscarnet GRANULOMATOUS DISORDERS Sarcoidosis Berylliosis Wegener granulomatosis Tuberculosis Histoplasmosis Coccidioidomycosis Nocardiasis Candidiasis Cat scratch fever Eosinophilic granuloma Crohn disease Silicone implants, paraffin injection IMMOBILIZATION PLUS Juvenile skeleton Malignancy Paget disease Primary hyperparathyroidism Renal failure MILK-ALKALI SYNDROME PARENTERAL NUTRITION FAMILIAL HYPOCALCIURIC HYPERCALCEMIA HYPOPHOSPHATEMIA RENAL FAILURE IDIOPATHIC HYPERCALCEMIA OF INFANCY HYPERPROTEINEMIA MANGANESE INTOXICATION ADVANCED CHRONIC LIVER DISEASE MALIGNANCY-ASSOCIATED HYPERCALCEMIA

58,59 60,61 62,63 64 65 69,70,113 114 68 71 115 116 117 118 119 120 72,121 76,77

80 81,83 Chap. 58 84 Chap. 70 85,86 87,88 89

Malignancy-associated hypercalcemia may arise through four general mechanisms: local osteolysis, humoral hypercalcemia of malignancy (HHM), excessive production of 1,25-dihydroxyvitamin D [1,25(OH)2D], and paraneoplastic secretion of parathyroid hormone (PTH). LOCAL OSTEOLYTIC HYPERCALCEMIA The first type of hypercalcemia to be well described clinically was that which accompanies multiple myeloma and breast cancer.1 Most cases of lymphoma-induced hypercalcemia probably also belong in this group. This category comprises ~20% of patients with malignancy-associated hypercalcemia. Typically, affected patients display widespread skeletal tumor involvement as evidenced by radiographs, bone radionuclide scans, bone marrow biopsies, and postmortem examinations (Fig. 59-1). Histologically, the bone marrow space is infiltrated by tumor cells, and bone surfaces are lined by numerous active osteoclasts. Because of this histologic picture, tumor cells within the marrow space are widely believed to secrete local factors that stimulate osteoclastic bone resorption. With this apparent local mechanism in mind, these patients may be referred to as having local osteolytic hypercalcemia (LOH). The nature of the bone-resorbing or osteoclast-activating factors secreted by the tumor cells remains incompletely defined. Several factors have been proposed, including tumor necrosis factor a,2 lymphotoxin,3 interleukin-1,4 PTH-related protein (PTHrP),5,6 and 7 interleukin-6,8 and prostaglandin E29 all of which stimulate osteoclastic bone resorption in in vitro systems. Direct bone resorption of devitalized bone in vitro also has been demonstrated by breast cancer cells.10 Possibly, in some instances, tumors may directly phagocytose bone. Biochemical studies of patients with LOH characteristically reveal increases in fractional calcium excretion, normal or near-normal serum phosphate concentrations and renal phosphate threshold values, decreases in circulating immunoreactive PTH and 1,25(OH)2D concentrations, and reductions in nephrogenous cyclic adenosine monophosphate (cAMP) excretion11,12 (Fig. 59-2 and Fig. 59-3). Bone resorption markers, such as hydroxyproline cross-links and N-telopeptide, are increased.13

FIGURE 59-1. Multiple myeloma. A, Bone marrow aspirate showing extensive infiltration with myeloma cells. The cells and nuclei are of varying sizes. The nuclei are eccentrically placed, and the nucleoli are prominent. B, Skull radiograph of a hypercalcemic patient with multiple myeloma, showing multiple lytic lesions (arrowheads). (A, Courtesy of Dr. Geraldine P. Schecter.)

FIGURE 59-2. Biochemical profiles in patients with cancer and normal calcium values (cancer controls), patients with primary hyperparathyroidism (HPT), patients with humoral hypercalcemia of malignancy (HHM), and patients with local osteolytic hypercalcemia (LOH). (NcAMP, nephrogenous cyclic adenosine monophosphate excretion; GF, glomerular filtration; TMP/GFR, renal phosphorus threshold; 1,25(OH)2D, 1,25-dihydroxyvitamin D.) (Adapted from Stewart AF, Horst R, Deftos LJ, et al. Biochemical evaluation of patients with cancer-associated hypercalcemia: evidence for humoral and nonhumoral groups. N Engl J Med 1980; 303:1377.)

FIGURE 59-3. Serum immunoreactive parathyroid hormone (PTH) levels obtained using a midregion PTH immunoassay in the patient groups described in Figure 59-2. Note that PTH concentrations correlate with the serum calcium concentration in patients with primary hyperparathyroidism (HPT) but not in patients with humoral hypercalcemia of malignancy (HHM) or local osteolytic hypercalcemia (LOH); also, although PTH values are normal in the two cancer hypercalcemia groups, they are indistinguishable, suggesting that the normal PTH values are related to immunoassay methodology and not physiology. (From Godsall JW, Burtis WJ, Insogna KL, et al. Nephrogenous cyclic AMP, adenylate cyclase-stimulating activity, and the humoral hypercalcemia of malignancy. Recent Prog Horm Res 1986; 42:705.)

HUMORAL HYPERCALCEMIA OF MALIGNANCY A second group of patients with malignancy-associated hypercalcemia includes those with HHM. This group comprises ~80% of patients with malignancy-associated hypercalcemia in unselected series.11,12,13,14 and 15 The syndrome was first described in 1941. In contrast to the tumors encountered in patients with LOH, those with HHM typically have squamous carcinomas of any site (including the head and neck, lung, esophagus, cervix, vulva, and skin); renal carcinomas; transitional carcinomas of the bladder; ovarian carcinomas; and human T-cell lymphoma/leukemia.11,12,13,14,15 and 16 In addition, although skeletal metastases are the cause of hypercalcemia in most women with breast cancer, a significant minority (perhaps as many as 30%) have a humoral form of hypercalcemia.17,18 The evidence that hypercalcemia is humoral in origin includes the observations that (a) these patients typically display hypercalcemia in the setting of few or no skeletal metastases,11,12 (b) hypercalcemia reverses with the resection of primary tumors (e.g., of the lung or kidney) that do not involve the skeleton,19 and (c) skeletal biopsies from these patients reveal striking osteoclastic activation in the absence of tumor invasion of marrow.20 From the initial description of this syndrome in 1941 through the early 1980s, the humoral factor responsible for this syndrome was variously believed to be PTH (causing ectopic hyperparathyroidism or pseudohyperparathyroidism), vitamin Drelated phytosterols, or prostaglandins of the E series. It is now clear that the majority of cases of HHM result from overproduction of PTHrP by the tumor (see Chap. 52).20a The evidence that PTHrP causes HHM can be summarized as follows. First, PTHrP is produced by and has been purified from tumors associated with HHM.21,22,23 and 24 Second, PTHrP is structurally homologous with PTH, binds to PTH receptors, and mimics the effects of PTH in bone and kidney.21,25 Third, infusion of PTHrP into laboratory animals and humans reproduces the major features of the syndrome.21,26,27 Fourth, PTHrP concentrations in the circulation are elevated in patients with HHM but not in patients with other types of hypercalcemia or in patients with cancer who do not have hypercalcemia.14,15 Finally, the infusion of anti-PTHrP antisera into animals with HHM reverses the syndrome.28,29 Similar to patients with primary hyperparathyroidism, patients with HHM display hypercalcemia, a reduction in the renal phosphorus threshold (maximal rate of tubular resorption of phosphorus/glomerular filtration rate), accelerated osteoclastic bone resorption, and increases in nephrogenous (i.e., renal tubulederived) cAMP excretion (see Fig. 59-2 and Fig. 59-3).11,13,20 Unlike patients with primary hyperparathyroidism, however, patients with HHM display reductions in circulating concentrations of 1,25(OH)2D, reductions in immunoreactive PTH, and reductions of osteoblastic bone formation.11,13,20 Markers of bone resorption (hydroxyproline, pyridinoline cross-links, N-telopeptide) are increased, whereas markers of bone formation (alkaline phosphatase, osteocalcin) are reduced.13 Immunoreactive PTHrP concentrations in the circulation are elevated in HHM.11,15,18 Although this point is controversial, patients with HHM appear to have a higher fractional excretion of calcium than do patients with primary hyperparathyroidism.11,30 From a pathophysiologic standpoint, hypercalcemia results primarily from accelerated net bone resorption. Intestinal calcium absorption is reduced,31 as would be predicted by the reduced plasma 1,25(OH)2D concentrations.11,13 Hypercalcemia occurs when the rate of calcium mobilization from the skeleton exceeds the ability of the kidney to clear this calcium load. Although enormous progress has been made during the past decade in understanding the pathophysiology underlying HHM, three unanswered questions remain. First, why is osteoblastic activity uncoupled from the increase in osteoclastic activity in patients with HHM,20 and why does this differ from the situation observed in patients with primary hyperparathyroidism, in whom a coupled increase occurs in both osteoblastic and osteoclastic activity?20 Second, why are plasma 1,25(OH)2D concentrations elevated in primary hyperparathyroidism but reduced in HHM?11,13 Third, do differences truly exist in fractional calcium excretion and distal tubular calcium handling in patients with HHM and primary hyperparathyroidism,11,30 and if so, why? EXCESSIVE PRODUCTION OF 1,25-DIHYDROXYVITAMIN D A third subtype of malignancy-associated hypercalcemia has been defined in patients with lymphoma. Approximately 30 such patients have been described.32,33 and 34 No clear unifying pattern of tumor histology is found. Circulating 1,25(OH)2D concentrations in these patients are dramatically elevated (Fig. 59-4). The belief is that the serum 1,25(OH)2D elevations result from excessive and unregulated conversion of circulating 25-hydroxyvitamin D [25(OH)D] to 1,25(OH)2D by the lymphomatous tissue, and that the 1,25(OH)2D elevations lead to hypercalcemia through intestinal calcium hyperabsorption and bone resorption. Nephrogenous cAMP excretion and serum PTH values are reduced.32,33 and 34

FIGURE 59-4. Serum calcium and plasma 1,25-dihydroxy vitamin D [1,25(OH)2D] concentrations in two patients with lymphoma. Note that in contrast to patients with humoral hypercalcemia of malignancy and local osteolytic hypercalcemia (see Fig. 59-2), plasma 1,25(OH)2D concentrations are strikingly elevated. These values reversed after glucocorticoid therapy in one patient (left panel) and resection of a solitary splenic lymphoma in the other patient (right panel). (From Rosenthal ND, Insogna KL, Godsall JW, et al. 1,25 dihydroxyvitamin Dmediated humoral hypercalcemia in malignant lymphoma. J Clin Endocrinol Metab 1985; 60:29.)

PARANEOPLASTIC (ECTOPIC) SECRETION OF PARATHYROID HORMONE As noted earlier, HHM was once thought to result almost exclusively from the ectopic secretion of PTH. With the advent of second-generation and third-generation PTH immunoassays and of molecular techniques for the measurement of PTH mRNA, it became apparent in the 1980s that HHM was not caused by ectopic secretion of PTH. The discovery of PTHrP and the development of sensitive and specific PTHrP immunoassays have now demonstrated that HHM results from systemic secretion

of PTHrP by tumors. The concept of hyperparathyroidism due to ectopic secretion of PTH was abandoned. Interestingly, however, seven cases of true ectopic secretion of PTH have been documented.35,36 The tumors in these cases contained actual PTH mRNA and secreted authentic PTH. The tumors responsible were a small cell carcinoma of the lung, a small cell carcinoma of the ovary, a thymoma, a primitive neuroendocrine tumor, squamous carcinomas of the lung, and a clear cell carcinoma of the ovary. PTH levels declined with resection of the responsible tumors. In one case, the ectopic production of PTH was found to result from a gene rearrangement in tumor DNA that placed the coding region of the PTH gene under the control of a promoter for an ovarian gene product.35 From a clinical standpoint, these observations have three implications. First, ectopic secretion of PTH, although extremely rare, does occur. Second, affected patients display elevations in circulating authentic PTH, even when measured by modern, highly specific PTH immunoassays. Third, this diagnosis should be considered before surgery in all patients with otherwise typical primary hyperparathyroidism and should be actively excluded in patients with primary hyperparathyroidism who have failed to respond to parathyroidectomy. MISCELLANEOUS CAUSES OF MALIGNANCY-ASSOCIATED HYPERCALCEMIA In some cases of malignancy-associated hypercalcemia, the mechanism cannot be easily categorized. For example, the author has described a young woman with an ovarian dysgerminoma who had low nephrogenous cAMP values but who clearly had a humoral form of hypercalcemia of malignancy, because no bone metastases were apparent and her hypercalcemia reversed with resection of her tumor.37 Certain common tumors are rarely associated with hypercalcemia. Examples include adenocarcinomas of the prostate, colon, and stomach, and small cell carcinomas of the lung. The discovery of hypercalcemia in these patients should alert the physician to causes other than malignancy, to the presence of a second primary neoplasm, or to an incorrect histologic diagnosis. Finally, hyperparathyroidism caused by a parathyroid adenoma or hyperplasia may coexist in a patient with cancer. Some authors have suggested that primary hyperparathyroidism actually is more common in patients with cancer than in the general population. In a study of 133 patients with malignancy and hypercalcemia, 8 proved to have primary hyperparathyroidism as the cause.12 ENDOCRINOPATHIES ASSOCIATED WITH HYPERCALCEMIA THYROTOXICOSIS Mild hypercalcemia, as assessed by ionized serum calcium measurement, occurs in as many as 47% of patients with thyrotoxicosis,38,39 but total serum calcium values exceeding 11.0 mg/dL are rare. Commonly, the serum alkaline phosphatase level is increased. In a few cases, the patients have coexisting Graves disease and primary hyperparathyroidism, but usually thyrotoxicosis is believed to be the cause of the hypercalcemia. The circulating 1,25(OH)2D level falls in thyrotoxicosis.40 Because intestinal calcium absorption is reduced and osteoclastic activity is increased in hyperthyroidism, high circulating levels of thyroxine and triiodothyronine are believed to produce hypercalcemia through excessive osteoclastic activity. In these patients, b-adrenergic blockade may reduce the hypercalcemia. The diagnosis is established when the hypercalcemia reverses with successful treatment of the hyperthyroidism. PHEOCHROMOCYTOMA Patients with pheochromocytoma may have mild to severe hypercalcemia.41,42 and 43 The mechanisms responsible for this hypercalcemia may theoretically include (a) hyperparathyroidism occurring as part of the MEN2A syndrome, (b) hyperparathyroidism occurring as a result of catecholamine-mediated PTH secretion from the parathyroid glands, (c) ectopic PTH secretion from the pheochromocytoma, (d) catecholamine-induced bone resorption, and (e) secretion by the pheochromocytoma of PTHrP. Among these possibilities, strong evidence exists to support only the first and the last. In pheochromocytoma, hypercalcemia caused by primary hyperparathyroidism in the context of the MEN2A syndrome has been reported most commonly. In these patients, as expected, hypercalcemia persists after the resection of the adrenal tumor; however, a few patients with pheochromocytoma have been described whose hypercalcemia reversed after surgery. From a clinical standpoint, patients discovered to have hypercalcemia and a pheochromocytoma should have the pheochromocytoma resected before consideration of parathyroidectomy, because the hypercalcemia rarely reverses after adrenalectomy, and performance of a parathyroidectomy in a patient with a pheochromocytoma may result in hypertensive crisis. As a corollary, patients with hypertension who have well-documented primary hyperparathyroidism should be screened for a pheochromocytoma before parathyroidectomy for these same reasons (see Chap. 86 and Chap. 188). ADDISONIAN CRISIS Mild hypercalcemia has been described rarely in patients with addisonian crisis.44,45 Their hypercalcemia usually is mild (11.012.0 mg/dL) and could be the result of hemoconcentration and hyperalbuminemia. Rehydration and glucocorticoid therapy readily normalize the serum calcium level. In some patients, excessive renal calcium resorption, caused by dehydration and enhanced skeletal mobilization, may be present.45 Most of these patients were described several decades ago when tuberculosis was the predominant cause of adrenal insufficiency; this raises the interesting, but undocumented, possibility that the glucocorticoid-reversible hypercalcemia was related not to adrenal insufficiency but to tuberculosis. VASOACTIVE INTESTINAL PEPTIDESECRETING TUMORS Benign and malignant islet cell tumors that secrete vasoactive intestinal peptide (VIP) present with a characteristic clinical picture including watery diarrhea, hypokalemia, and achlorhydria (VIPoma syndrome, pancreatic cholera, or WDHA syndrome; see Chap. 220). Approximately 50% of these patients have hypercalcemia, which occasionally is severe.46 Although the hypercalcemia may result from primary hyperparathyroidism occurring as part of the MEN1 syndrome, the fact that it sometimes reverses with surgical resection of the VIPoma suggests that VIP or another islet cell product is responsible. MEDICATIONS VITAMIN D Vitamin D intoxication is an unusual but frequently overlooked cause of hypercalcemia.47,48 The dosage of vitamin D required to produce hypercalcemia is large, usually at least 50,000 U three times weekly. This dose is available only by prescription (most multivitamins contain 400 U per tablet); thus, vitamin D intoxication usually occurs inadvertently in the setting of vitamin D therapy for osteoporosis, hypoparathyroidism, or renal osteodystrophy. It also may occur in patients taking self-administered megavitamin therapy. Moreover, vitamin D intoxication may occur in patients receiving any of the available analogs of vitamin D, such as 1,25(OH)2D (see Chap. 54). Vitamin D intoxication resulting from excessive addition of vitamin D to milk also may occur.49,50 Frequently, hypercalcemia in patients with vitamin D intoxication goes unrecognized, with the symptoms of nausea, vomiting, lethargy, and weakness being attributed to a viral syndrome. Hence, patients being treated with pharmacologic dosages of vitamin D should be warned of the possibility of hypercalcemia should flu-like symptoms develop. In vitamin D toxicity, hypercalcemia develops largely because of excessive intestinal calcium absorption; however, accelerated bone resorption also plays a role, because patients with vitamin D intoxication are in negative calcium balance, and vitamin D is known to resorb bone in vitro. Decreased renal clearance of calcium also is important in most patients with vitamin D intoxication because of hypercalcemia-induced reduction in the glomerular filtration rate. The diagnosis of vitamin D intoxication is made on clinical grounds. Because dosages of vitamin D sufficient to cause hypercalcemia can be obtained only by prescription, most patients who are vitamin D intoxicated are taking their vitamin D for osteoporosis, hypoparathyroidism, or chronic renal failure in excessive dosages or at too frequent dosing intervals. In all patients taking pharmacologic dosages of vitamin D, plasma concentrations of 25(OH)D are markedly elevated, regardless of the serum calcium value, and thus may not be diagnostic. Plasma 1,25(OH)2D concentrations are not markedly elevated in patients with vitamin D intoxication and often are normal or low, as would be expected in the setting of reduced circulating PTH concentrations. Azotemia may be present, and the serum phosphate value may be normal or increased. The serum alkaline phosphatase level is normal. A diagnosis of vitamin D intoxication is secured when hypercalcemia reverses with hydration, a low-calcium diet, and glucocorticoid therapy, and does not recur after dosage reduction or discontinuation of therapy with the offending vitamin D compound. Hypercalcemia usually reverses rapidly (within days) in patients intoxicated with 1,25(OH)2D after use of the medication is discontinued, reflecting its short biologic half-life. Conversely, hypercalcemia may persist for weeks after the discontinuation of vitamin D use, reflecting accumulation in, and prolonged release from, adipose

tissue. VITAMIN A Hypercalcemia has been described in rare patients with vitamin A intoxication.51,52 Usually, these persons have ingested vitamin A at a dosage of 50,000 IU per day for weeks to months. The recommended daily allowance is 5000 IU per day. The serum alkaline phosphatase concentration may be elevated. Although this syndrome has been reported only rarely in the past, the current widespread use of vitamin A from health food stores and of vitamin A congeners such as cis-retinoic acid in dermatologic disorders53 and chemotherapy54 suggests that it may be encountered more frequently in the future. Vitamin A is presumed to cause hypercalcemia through excessive bone resorption; cis-retinoic acid causes bone resorption in vitro. The diagnosis is made on clinical grounds, with heavy reliance on a history of excessive vitamin A intake, associated findings (alopecia, dermatitis, hepatic dysfunction), and reversal of hypercalcemia after cessation of vitamin intake. Measurements of circulating vitamin A and retinyl ester are useful to establish the diagnosis. LITHIUM Lithium therapy has been associated with hypercalcemia.55,56 and 57 In one study from Denmark, 12 of 96 patients taking lithium were noted to have hypercalcemia. In vitro and in vivo evidence suggests that lithium may alter the parathyroid cell set-point for PTH secretion. Cessation of lithium therapy completely reverses the hypercalcemia in many patients. The hypercalcemia persists in other patients, however, and primary hyperparathyroidism eventually is diagnosed. Although the hypercalcemia encountered in these patients is mild (10.511.5 mg/dL), the question usually arises as to whether this mild hypercalcemia is contributing to, or perpetuating, the psychiatric symptoms and findings that originally required the lithium therapy. This can be resolved only by discontinuing or reducing the dosage of lithium. THIAZIDE DIURETICS Use of thiazide diuretics (but not loop diuretics) has been associated with mild hypercalcemia.58 Cessation of thiazide therapy reverses the hypercalcemia in many cases, although it persists in some patients. Thiazide diuretics stimulate renal tubular calcium reabsorption, a finding that has led to their use in the treatment of renal leak hypercalciuria and hypoparathyroidism. However, the kidney is not likely to be the only organ involved in thiazide-mediated hypercalcemia, because hypercalcemia may develop in anephric patients treated with chlorthalidone.59 As with lithium-induced hypercalcemia, the diagnosis is confirmed by the reversal of hypercalcemia after the cessation of thiazide therapy. ESTROGENS AND ANTIESTROGENS Estrogen and antiestrogen therapy causes hypercalcemia in as many as one-third of patients with breast cancer.60,61 This estrogen flare hypercalcemia apparently occurs only in the setting of extensive skeletal metastases. Interestingly, if the hypercalcemia can be controlled with hydration and other measures, and administration of the estrogen/antiestrogen can be continued, a hypercalcemic response to estrogen/antiestrogen therapy may be a favorable predictor of response to hormonal therapy. The mechanism of the estrogen/antiestrogen flare remains obscure, but it usually responds to hydration, diuresis, and administration of glucocorticoids and agents that inhibit bone resorption. GROWTH HORMONE Growth hormone anabolic therapy used in the setting of the intensive care unit (in treatment of burns, respiratory failure, poor wound healing) or in the treatment of acquired immunodeficiency syndrome (AIDS) has been associated with hypercalcemia in the 11 to 12 mg/dL range.62,63 THEOPHYLLINE Large doses of theophylline, used in the treatment of asthma, have caused hypercalcemia. The specific mechanism is unknown.64 FOSCARNET Foscarnet is an antiviral agent used in the AIDS setting. It has been reported to cause hypercalcemia.65 GRANULOMATOUS DISORDERS Each of the granulomatous diseases listed in Table 59-1 has been associated with hypercalcemia. Other granulomatous diseases (e.g., chronic granulomatous disease of the young, Crohn disease), however, have not. Clearly 1,25(OH)2D not only is involved in mineral metabolism but also is a lymphokine produced by, and acting on, subsets of T cells, B cells, and macrophages.66,67 Although the normal immunologic role of 1,25(OH)2D has been only partially characterized, excessive systemic production of 1,25(OH)2D is clearly a feature of tuberculosis,68 sarcoidosis69,70 (Fig. 59-5), histoplasmosis,71 the granulomatous reaction to silicone implants,72 and other granulomatous disorders (see Table 59-1). In contrast to the normal situation, in which the renal conversion of circulating 25(OH)D to 1,25(OH)2D is tightly regulated, in patients with these granulomatous disorders, 1,25(OH)2D production appears to be largely, and perhaps exclusively, substrate dependent. Thus, exposure to even small amounts of dietary vitamin D or to solar or other ultraviolet radiation leads to the expected physiologic increase in circulating 25(OH)D, which is then converted in an unregulated fashion to 1,25(OH)2D. Although in most of the examples cited, evidence for this extrarenal conversion is inferential, direct conversion of 25(OH)D to 1,25(OH)2D has been demonstrated by sarcoidosis-derived pulmonary alveolar macrophages73 and by lymph node homogenates.74 Hypercalcemia appears to result from 1,25(OH)2D-mediated intestinal calcium hyperabsorption, 1,25(OH)2D-mediated bone resorption (patients may remain hypercalcemic and hypercalciuric on very-low-calcium diets), and hypercalcemia-induced reductions in the glomerular filtration rate.

FIGURE 59-5. Radiograph of a patient with sarcoidosis and hypercalcemia. A typical-appearing, diffuse, bilateral, reticulonodular infiltrate is seen, predominantly in the midlung regions.

Because of the increases in 1a-hydroxylase activity [the enzyme that converts 25(OH)D to 1,25(OH)D], patients with these disorders may be exquisitely sensitive to small increases in dietary vitamin D intake, to sunlight or other forms of ultraviolet radiation, and to dietary calcium intake. Patients with granuloma-induced 1,25(OH)2D elevations appear to be, from a physiologic standpoint, the benign equivalent of those with 1,25(OH)2D-secreting malignant lymphomas. Generally, patients with these disorders have widespread pulmonary and extra-pulmonary disease, and the diagnosis is obvious at presentation. Exceptions do exist, however. The author's practice is to measure circulating 1,25(OH)2D values in patients with unexplained hypercalcemia. If the values are elevated, and if other causes for 1,25(OH)2D elevation (lymphoma, primary hyperparathyroidism, pregnancy, absorptive hypercalciuria) are not apparent, then a systematic search for pulmonary, renal, hepatic, ocular, and bone marrow granulomas is made. Occasionally, a glucocorticoid suppression test is warranted to exclude primary hyperparathyroidism (i.e., failure of the calcium to be suppressed to normal levels in patients with hyperparathyroidism).75 If this is done, care should be taken to establish a stable baseline; that is, serum calcium values should be monitored and should be stable for 3 to 4 days before glucocorticoid administration is begun. In this test, hydrocortisone is administered orally, 40 mg every 8

hours for 10 days, and a serum calcium evaluation is performed each morning. IMMOBILIZATION Weightlessness (as occurs in space flight) and complete immobilization (as occurs in spinal cord injury and other neuromuscular disorders, extensive casting for skeletal fractures, or prolonged bed rest associated with a chronic debilitating illness) regularly lead to calcium mobilization from the skeleton. Hypercalcemia rarely occurs in patients whose underlying bone turnover rate is normal or low. However, in persons with elevated bone turnover rates, such as young adults, adolescents, and children76,77 and 78; rare patients with Paget disease (who have focal increases in bone turnover); patients with malignancy with or without bone metastases; and patients with primary or secondary hyperparathyroidism, immobilization can produce or aggravate hypercalcemia. An example of the latter would be an elderly woman with mild primary hyperparathyroidism (serum calcium value, 10.6 mg/dL) who becomes bedridden and develops further hypercalcemia (serum calcium value, 12 or 13 mg/dL). As with hypercalcemia occurring in other settings, hypercalcemia associated with immobilization may present as lethargy and depression. In a young adult with quadriplegia, the depression may be misinterpreted as an appropriate psychologic response to a devastating injury, and the hypercalcemia may go undetected and untreated. In particular, abdominal pain syndromes caused by hypercalcemia may be difficult to evaluate in patients with spinal cord injury and have led to unnecessary laparotomy. Immobilization leads to rapid demineralization of the skeleton. In young adults who are completely immobilized, as much as 30% of the skeleton may be lost in the first few weeks to months. Although the mechanism by which immobilization leads to accelerated bone resorption remains unclear, bone histologic studies have shown that it is associated with an increase in osteoclastic activity and a decrease in osteoblastic activity.79 This uncoupling of bone cell activity is reminiscent of that encountered in malignancy-associated hypercalcemia, and explains the hypercalcemia, hypercalciuria, increased urinary hydroxyproline excretion, and normal circulating alkaline phosphatase values typically encountered in these patients. Interestingly, previous parathyroidectomy appears to prevent skeletal mineral loss in animal models of immobilization, suggesting that the skeleton must be primed by PTH for immobilization to be followed by increases in osteoclastic activity. Bone resorption continues for as long as 12 to 18 months and then ceases spontaneously, despite continued immobilization. Nephrolithiasis (infection or triple phosphate stones) is a frequent finding in patients with spinal cord injury. This occurs in part because of urinary tract infections resulting from stasis and long-term or intermittent bladder catheterization, regardless of whether hypercalcemia is present. The incidence of nephrolithiasis among patients with spinal cord injuries, both for infection stones and for calcium oxalate stones, is higher in those who have hypercalciuria than in those who do not.78 In immobilization-induced hypercalcemia, parathyroid function is suppressed, as indicated by measurement of low circulating immunoreactive PTH and 1,25(OH)2D concentrations and by low nephrogenous cAMP excretion.76 The diagnosis is made in the presence of the biochemical findings just described, together with a clinical picture of complete immobilization. MILK-ALKALI SYNDROME Hypercalcemia was described in the 1920s as a result of peptic ulcer treatment regimens that included excessive amounts of milk or cream (gallons per day) and absorbable antacids such as sodium bicarbonate.80,81 The original milk-alkali syndrome presented as moderate to severe hypercalcemia associated with advanced renal failure, hyperphosphatemia, and a systemic alkalosis. This variety of milk-alkali syndrome has largely disappeared, as this form of antiulcer regimen has disappeared. In more modern times, hypercalcemia has occurred principally in people taking large quantities of absorbable calcium-containing antacids for the treatment of esophagitis or peptic symptoms. Typically, patients report having taken > 4 g of elemental calcium per day in the form of calcium carbonate tablets (e.g., 200-mg tablets) or liquid calcium carbonate antacids. In contrast to the intestinal absorption of calcium that 1,25(OH)2Dregulated occurs when dietary calcium intake is normal (~1000 mg per day), ingestion of large amounts of dietary calcium can lead to passive, unregulated, and inappropriate calcium absorption. This latter variety of milk-alkali syndrome is frequently overlooked by physicians for two reasons. First, patients often fail to volunteer the fact that they are using large amounts of calcium carbonate, because they do not consider it a medication. Second, physicians fail to specifically ask patients with hypercalcemia if they are using antacids. Hypercalcemia leads to polyuria, polydipsia, mental status changes, dehydration, prerenal reductions in the glomerular filtration rate; if hypercalcemia, hyperphosphatemia, and dehydration persist, nephrocalcinosis and permanent renal impairment may occur. The diagnosis is made by excluding other causes of hypercalcemia and documenting calcium intakes in excess of 4 g per day of elemental calcium. Treatment includes the discontinuation of the calcium antacid, hydration, and education regarding the large amounts of calcium that can be ingested unintentionally. PARENTERAL NUTRITION Hypercalcemia may develop in two groups of patients receiving parenteral hyperalimentation. The first group includes patients in whom hypercalcemia develops while they are receiving large amounts (>300 mg per day) of intravenous calcium, particularly when renal compromise is present. In this group, a reduction in parenteral calcium administration reverses the hypercalcemia and hypercalciuria. In the second group of patients, hypercalcemia may develop after long-term parenteral hyperalimentation, despite only modest parenteral calcium administration, and hypercalcemia and hypercalciuria may persist after the cessation of parenteral calcium administration.81,82 and 83 Affected patients may have symptomatic bone disease, and most have a histologic picture on bone biopsy that is suggestive of osteomalacia. Osteomalacia in these patients may result in part from aluminum intoxication. Aluminum has been delivered to these patients in the past through antacids and as a component of the amino-acid hydrolysates used in parenteral nutrition preparations. With the recognition of aluminum intoxication in these patients and the elimination of aluminum from hyperalimentation solutions, this syndrome appears to be disappearing. FAMILIAL HYPOCALCIURIC HYPERCALCEMIA The hypercalcemia that occurs in patients with familial hypocalciuric hypercalcemia, also known as familial benign hypercalcemia (see Chap. 58), has been attributed to diminished renal calcium clearance (excessive tubular reabsorption of calcium) as well as failure of parathyroid gland calcium sensing, both occurring most commonly as a result of inactivating mutations in the calcium-sensing receptor.83a HYPOPHOSPHATEMIA In animals, hypophosphatemia leads to osteoclast stimulation, inhibition of mineralization, hypercalciuria, and hypercalcemia. This is true even when animals have been rendered vitamin D deficient and surgically hypoparathyroid. In humans, severe hypophosphatemia leads to hypercalciuria, presumably reflecting both increases in osteoclastic activity and hypophosphatemia-induced increases in renal 1,25(OH)2D production, with consequent enhanced intestinal calcium absorption. Although hypercalcemia resulting exclusively from hypophosphatemia has not been described in humans, it seems likely that when hypophosphatemia complicates hypercalcemia of any cause, it exacerbates the hypercalcemia. Hypophosphatemia is particularly common among patients with hypercalcemia, regardless of the underlying cause, for the following reasons: Food intake may be reduced due to vomiting or anorexia; dietary phosphorus absorption may be limited by the use of phosphate-binding antacids; hypercalcemia per se is phosphaturic; PTH and PTHrP reduce the renal phosphate threshold; and agents that are used to treat hypercalcemia (calcitonin, loop diuretics, saline infusion) induce phosphaturia. RENAL FAILURE Hypercalcemia has been described occasionally in patients with acute tubular necrosis, especially during the polyuric phase. In one careful study of patients with acute tubular necrosis resulting from rhabdomyolysis, patients initially were hypocalcemic because of severe hyperphosphatemia but later became hypercalcemic with the reversal of oliguria and hyperphosphatemia.84 The hypercalcemia was transient and may have resulted from (a) mobilization of soft-tissue calcium phosphate salts deposited during the early hyperphosphatemic period, and (b) rebound hypercalcemia from the transiently elevated PTH and 1,25(OH)2D values that occurred during the early hypocalcemic period. Hypercalcemia often develops in patients with chronic renal failure who are receiving hemodialysis. Hypercalcemia is difficult to evaluate in such patients because the usual biochemical values [renal phosphate and calcium excretion, serum PTH, serum 1,25(OH)2D, and nephrogenous cAMP] are difficult to interpret in this setting. Nevertheless, reductions in dietary calcium intake and vitamin D supplements may correct the hypercalcemia, a result which suggests that increased intestinal calcium absorption in the face of diminished renal clearance may account for hypercalcemia in some patients. Immobilization of patients receiving hemodialysis may lead to hypercalcemia, presumably through acceleration of the underlying high rates of bone turnover. Tertiary hyperparathyroidism (the evolution of autonomous parathyroid-dependent hypercalcemia in patients with renal failure and chronic secondary hyperparathyroidism) may occur. Dialysis against a high-calcium bath also

may lead to hypercalcemia. Finally, hypercalcemia may occur as part of the aluminum bone disease that forms an important aspect of renal osteodystrophy. IDIOPATHIC HYPERCALCEMIA OF INFANCY The complex of disorders known as idiopathic hypercalcemia of infancy is discussed in Chapter 70. HYPERPROTEINEMIA Hyperalbuminemia may lead to hypercalcemia because albumin is the major circulating calcium-binding protein. The hypercalcemia is usually mild, and patients in whom the diagnosis is suspected generally are severely dehydrated. Rapid rehydration often corrects the hypercalcemia before careful biochemical evaluation can be performed. In practice, this form of hypercalcemia can be diagnosed by the determination of a normal ionized calcium value in a patient with mildly elevated total serum calcium values, markedly elevated serum albumin levels, and none of the other causes for hypercalcemia listed in Table 59-1. Rarely, patients with multiple myeloma have been described who display elevations in the total serum calcium concentration (sometimes striking elevations) but normal ionized calcium values.85,86 This factitious hypercalcemia is caused by unusual immunoglobulins with an extremely high binding capacity for calcium. Hypercalciuria, electrocardiographic changes, and symptoms of hypercalcemia are absent in these unusual patients. Although hypercalcemia caused by a myeloma calcium-binding protein is unusual, it should be considered before attributing hypercalcemia in patients with multiple myeloma to the more common cause, osteolytic bone disease. MANGANESE INTOXICATION Hypercalcemia has been reported to occur in patients with manganese intoxication.87,88 Exposure to manganese in doses sufficient to cause intoxication has been reported in miners, in welders, and in people drinking water from manganesecontaminated wells. The mechanisms responsible for the development of hypercalcemia are uncertain. The hypercalcemia may be mild (e.g., 11.5 mg/dL) or severe (e.g., 20 mg/dL). ADVANCED CHRONIC LIVER DISEASE Mild to severe hypercalcemia has been described in one series of 11 patients with advanced chronic liver disease awaiting liver transplantation.89 All the patients had severe jaundice (the mean serum bilirubin value was 28.7 mg/dL), and most had hypoalbuminemia and prolonged prothrombin times and were in mild renal failure (the mean serum creatinine level was 2.8 mg/dL). Hypercalcemia was associated with normal to reduced plasma 1,25(OH)2D, 25(OH)D, and PTH concentrations. Ionized serum calcium values were elevated. The cause of the syndrome appeared to be multifactorial, reflecting contributions from renal failure, immobilization, total parenteral nutrition, and vitamin D supplementation.

TREATMENT OF HYPERCALCEMIA
Principles of therapy for hypercalcemia should be based on an understanding of its cause, knowledge of its pathophysiology, and careful assessment of the symptom complex in individual patients. The treatment of mild hypercalcemia in a patient with asymptomatic primary hyperparathyroidism (i.e., without nephrolithiasis or apparent skeletal disease) is controversial (see Chap. 58). Similarly, mild hypercalcemia associated with thyrotoxicosis requires no treatment because it resolves spontaneously with treatment of the hyperthyroidism. The hypercalcemia in patients with familial hypocalciuric hypercalcemia almost always is asymptomatic and requires no therapy. Some may argue that severe hypercalcemia in a comatose patient with terminal cancer is best left untreated. The treatment of hypercalcemia in a patient with significant symptomatology may be an urgent problem. The severity of symptoms of hypercalcemia can be related to the absolute calcium concentration; the rate of development of hypercalcemia (the faster the increase, the greater the symptoms); and the duration of the hypercalcemia. These points, as well as the patient's overall status and prognosis, should be considered before commencing therapy. Therapy is best determined with an understanding of the pathophysiology of a given patient's hypercalcemia. For example, the absorptive hypercalcemia of sarcoidosis is best treated with dietary calcium restriction, whereas the resorptive hypercalcemia of immobilization and malignancy is most appropriately treated with measures aimed at diminishing osteoclastic activity. Specific treatment of most of the disorders listed in Table 59-1 is self-evident. Therapy for the hypercalcemia associated with primary hyperparathyroidism, renal osteodystrophy, and neonatal disorders is discussed in Chapter 58, Chapter 61 and Chapter 70, respectively. Treatment of the underlying disorder responsible for hypercalcemia is an obvious, but frequently overlooked, goal, particularly in patients with malignancy-associated hypercalcemia. In these patients, antitumor therapy should be planned and initiated early in the hospital course, because other measures are effective only transiently and may be associated with toxicity. Usually, long-term control of malignancy-associated hypercalcemia can be accomplished only by successful antitumor therapy. SALINE INFUSION Renal calcium excretion is enhanced by saline infusion, because of both the increase in glomerular filtration rate and the role of the filtered load of sodium in blocking proximal tubular calcium reabsorption.90 The rate of infusion should be adjusted to the clinical situation. The uncertain cardiovascular status of these patients, many of whom are elderly, dictates caution. Infusion rates of 200 to 400 mL per hour of 0.9% saline are commonly used. The status of hydration should be checked frequently by physical signs (rales, skin turgor, mucous membrane hydration, blood pressure) and by hemodynamic monitoring when appropriate. Aggressive saline infusion may lead to hypokalemia, hypophosphatemia, and hypomagnesemia. These ions should be measured and replaced as necessary. LOOP DIURETICS Loop diuretics appear to inhibit calcium reabsorption at the level of the thick ascending limb of the Henle loop.90 In theory, use of agents such as furosemide should enhance the calciuresis of a saline infusion. Furosemide in dosages of 40 mg or more, given intravenously or orally, is widely used for this purpose. This drug should not be administered to dehydrated patients until rehydration is complete, because it may further reduce the glomerular filtration rate and, thereby, reduce the filtered load of calcium, decreasing renal calcium clearance even more. The drug is particularly useful if the patient has coexistent congestive heart failure. Again, careful surveillance of a patient's fluid, electrolyte, and renal status during therapy is critical. RESTRICTION OF CALCIUM AND VITAMIN D INTAKE Dietary calcium and vitamin D restriction and avoidance of exposure to sunlight and other ultraviolet light sources is appropriate in patients with absorptive causes of hypercalcemia, such as sarcoidosis and other granulomatous diseases during their active stages.90 Dietary calcium intake should be kept at <400 mg per day, the use of vitamin D supplements should be discontinued, and the patient should be advised to limit sun exposure as much as possible. These same guidelines probably apply to the occasional patient with 1,25(OH)2D-mediated hypercalcemia associated with lymphoma.32,33 and 34 Similarly, vitamin D intoxication and the milk-alkali syndrome should be treated with dietary calcium restriction. Conversely, dietary calcium intake should not be reduced in patients with primarily resorptive, that is, malignancy-associated, hypercalcemia. Plasma 1,25(OH)2D values are reduced in these patients, and dietary calcium absorption is inefficient. Unpalatable calcium-restricted diets are unnecessary and may further the nutritional deterioration and general misery of these patients. CALCITONIN Most causes of acute hypercalcemia requiring aggressive treatment are related to accelerated bone resorption. Therefore, agents that inhibit osteoclastic function are important aspects of therapy. Bone resorption in vivo and in vitro can be inhibited by calcitonin. Calcitonin use has been disappointing in treating moderate to severe hypercalcemia because it does not work in all patients. The decrease in serum calcium is generally small, and most patients escape from the osteoclast-inhibiting effects of calcitonin in several days. Furthermore, the treatment is extremely expensive.90,91 Nevertheless, given its low incidence of side effects and the rapidity with which the serum calcium level falls after administration of the hormone (2 to 4 hours), it is useful in some clinical settings. The dosage of salmon calcitonin suggested by the manufacturer is 4 IU/kg every 12 hours given subcutaneously or intramuscularly. Patients should undergo skin testing for hypersensitivity before use, as described in the package insert. Although interest has been shown in, and an argument made for, the synergistic effects of calcitonin and glucocorticoids, evidence supporting the efficacy of a combined regimen is not compelling (see Chap. 53).

PLICAMYCIN Plicamycin (Mithramycin) was once a staple of hypercalcemia management, but it has now largely been replaced by the bisphosphonates. It inhibits osteoclastic bone resorption in a predictable and impressive fashion.90,92 Its mechanism of action on bone cells is unknown. The nadir in serum calcium after a dose of plicamycin occurs at ~48 hours. The usual dosage is 25 g/kg given intravenously over 4 to 6 hours, repeated as necessary every 2 to 7 days. Despite the reliability and efficacy of plicamycin, its use may be limited by the associated toxicity in the form of transaminase elevations, proteinuria, azotemia, bone marrow suppression, or a failure of platelet aggregation. These side effects were more commonly reported in the older literature describing the use of larger dosages of plicamycin.93 They occur rarely when the lower dosage is used for a limited duration. Patients without liver, bone marrow, or renal problems generally tolerate the agent well. In patients with preexisting disease of the marrow, kidney, or liver, if the agent is used at all, the dosage should be decreased to 12.5 g/kg and the number of repeated infusions should be minimized. Plicamycin is best reserved for patients with moderate to severe hypercalcemia that threatens the central nervous system or renal function and for whom bisphosphonates are not an option. Most often, it is used in patients with malignancy-associated hypercalcemia. It also has been used in patients with acute primary hyperparathyroidism, patients with parathyroid carcinoma, and patients with severe Paget disease who have normal or elevated calcium levels, to limit bone resorption. GLUCOCORTICOIDS Glucocorticoids have classically been used in the treatment of hypercalcemia associated with multiple myeloma, lymphoma, breast cancer, Addison disease, granulomatous disorders, and vitamin D intoxication.75,90,94 The oral dosages used have been ~25 mg of cortisone acetate to as high as 20 mg of prednisone, or their parenteral equivalents, given 3 to 4 times per day. The hypocalcemic response may take 7 to 10 days to become apparent. The mechanism of response to glucocorticoids includes tumorlytic (or granulomalytic) effects, inhibition of 1a-hydroxylase activity (in granulomatous disorders and lymphomas), direct inhibition of intestinal calcium transport (independent of corticosteroid effects on vitamin D metabolism), and inhibition of osteoclastic activity. Side effects of glucocorticoids (demineralization, generalized catabolic effects, immunosuppression), the slow response time, and interference with chemotherapeutic regimens have made these agents only moderately useful in treating patients with malignancy-associated hypercalcemia. Glucocorticoids do not reverse PTH-mediated hypercalcemia. Whether they are effective in treating HHM, mediated by PTHrP, remains an open question.75 PHOSPHORUS Phosphorus depletion exacerbates the hypercalcemia associated with most of the disorders described in Table 59-1. For this reason, if hypophosphatemia occurs, it should be corrected in patients with hypercalcemia using oral phosphorus. Detailed guidelines for phosphorus repletion have been reported.95 Generally, in hospitalized patients with malignancy-associated hypercalcemia, 250 mg (8 mmol) of phosphorus is given orally four times a day if the serum phosphate value is <3.0 mg/dL. The use of phosphate-binding antacids should be discontinued if possible. Meticulous care should be taken in following up serum calcium, phosphate, and creatinine values in patients with hypercalcemia who are treated with phosphorus preparations, because renal failure resulting from calcium phosphate precipitation (nephrocalcinosis) may occur abruptly. The use of phosphorus should be discontinued at the first sign of a rise in creatinine values or when serum phosphate values exceed 4.0 mg/dL. Oral phosphorus should be used with similar caution in the treatment of outpatients with hypercalcemia caused by primary hyperparathyroidism (see Chap. 58). Intravenous phosphorus has been used in patients with malignancy-associated hypercalcemia, usually with dramatic decreases in serum calcium levels but frequently with the abrupt onset of renal failure.96 Intravenous phosphorus administration is potentially dangerous and should be reserved for the rare patients with life-threatening hypercalcemia who cannot be treated using any other approach. Again, great care should be taken in following up serum calcium, phosphate, and creatinine values, and the minimum dose required to normalize the serum phosphate level should be administered. BISPHOSPHONATES The treatment of HHM, LOH, and parathyroid carcinoma has been facilitated by the advent of bisphosphonate therapy.90,97,98 and 99 Originally used as industrial detergents, bisphosphonates were found to inhibit osteoclastic bone resorption, the predominant or sole pathophysiologic mechanism underlying hypercalcemia of malignancy. The cellular mechanism responsible for this inhibition is not clear. The bisphosphonates are pyrophosphate analogs with a P-C-P bond that is resistant to hydrolysis by alkaline phosphatase. They bind to hydroxyapatite in bone. Bisphosphonates are poorly absorbed from the gastrointestinal tract (only 12% of an oral dose is absorbed) and, thus, are more effective when given intravenously. In the United States, two bisphosphonates, etidronate (1-hydroxyethylidene bisphosphonic acid) and pamidronate (3-amino-1-hydroxypropylidene bisphosphonic acid), are available for intravenous administration for malignancy-associated hypercalcemia. Pamidronate enjoys wider usage because of its single-day dosing and its greater potency. Unlike etidronate, pamidronate,100,101,102 and 103 a second-generation bisphosphonate, does not inhibit bone mineralization or increase phosphate levels at dosages in the therapeutic range. It is 100-fold more potent than etidronate and can be given intravenously in 250 mL of saline or 5% dextrose over the course of 4 to 24 hours in doses of 60 or 90 mg. The serum calcium concentration begins to decline 1 to 2 days after the infusion, with the nadir occurring 3 to 6 days later. Normocalcemia may last for 1 to 8 weeks.91 The most common side effect is a transient fever (1C to 2C above normal) beginning 24 to 48 hours after the infusion and lasting up to 3 days. This effect is usually observed only after the first dose. Rarely, first-dose effects of leukopenia and lymphopenia may occur. Hypophosphatemia, hypomagnesemia, and mild hypocalcemia may occur. Several newer, more potent bisphosphonates are under development and will become available for clinical use in the next several years. Their relative lack of side effects, together with their reliability and effectiveness, make the bisphosphonates the therapeutic agents of choice in most patients with malignancy-associated hypercalcemia. GALLIUM NITRATE Gallium nitrate, originally used as an antineoplastic agent, was found incidentally to cause hypocalcemia in previously normocalcemic patients.104,105 and 106 Although the drug appears to be effective, the need for continuous intravenous infusion and the potential for nephrotoxicity limit its use. OTHER MEASURES To the extent that an immobilized patient can perform active weight-bearing exercise, this is desirable. The response to mobilization in such patients may be rapid (within days) and dramatic. If parenteral calcium is being administered to patients with hypercalcemia, it should be withdrawn; in this regard, the presence of calcium supplements in parenteral hyperalimentation solutions is frequently overlooked. Hemodialysis and peritoneal dialysis may cause dramatic reductions in serum calcium levels in patients with hypercalcemia who are in renal failure, particularly when low-calcium or zero-calcium dialysate is used.107,108 and 109 Although most patients with malignancy-associated hypercalcemia are not candidates for dialysis, a few are. Dialysis also has been used to treat patients in hypercalcemic crisis, including patients with severe primary hyperparathyroidism.109 Prostaglandin synthetase inhibitors (aspirin and indomethacin) have been used to treat malignancy-associated hypercalcemia. Although prostaglandins clearly are important in bone metabolism, the use of the inhibiting agents in hypercalcemia has been disappointing.110 Several experimental therapies for hypercalcemia are being examined. Ethiophos (WR-2721) is a chemoprotective agent that has inhibitory effects on bone resorption and on PTH secretion.111 Thionaphthene-2-carboxylic acid and related compounds, which also appear to inhibit bone resorption, have been shown to reverse hypercalcemia in a rat model of HHM.112 CHAPTER REFERENCES
1. Gutman AB, Tyson TL, Gutman EB. Serum calcium, inorganic phosphatase activity in hyperparathyroidism, Paget's disease, multiple myeloma and neoplastic disease of the bones. Arch Intern Med 1936; 57:379. 2. Thompson BM, Mundy GR, Chambers TJ. Tumor necrosis factors alpha and beta induce osteoblastic cells to stimulate osteoclastic bone resorption. J Immunol 1987; 138:775. 3. Garrett RI, Durie BGM, Nedwin GE, et al. Production of the bone resorbing cytokine lymphotoxin by cultured human myeloma cells. N Engl J Med 1987; 317:526. 4. Cozzolino F, Torcia M, Aldinucci D, et al. Production of interleukin-1 by bone marrow myeloma cells. Blood 1989; 74:387. 5. Firkin F, Seymour JF, Watson AM, et al. Parathyroid hormone-related protein in hypercalcemia associated with hematological malignancy. Br J Hematol 1996; 94:486. 6. Horiuchi T, Miyachi T, Arai T, et al. Raised plasma concentrations of parathyroid hormonerelated protein in hypercalcemic multiple myeloma. Horm Metab Res 1997; 29:469.

7. Guise TA, Yin JJ, Taylor SD, et al. Evidence for a causal role of parathyroid hormone-related protein in the pathogenesis of human breast cancer-mediated osteolysis. J Clin Invest 1996; 98:1544. 8. Bataille R, Jourdan M, Zhang X-G, Klein B. Serum levels of interleukin-6, a potent myeloma cell growth factor, as a reflection of disease in plasma cell dyscrasias. J Clin Invest 1989; 84:2008. 9. Metz SA, McRae JR, Robertson RP. Prostaglandins as mediators of paraneoplastic syndromes. Metabolism 1981; 30:299. 10. Eilon G, Mundy GR. Direct resorption of bone by human breast cancer cells in vitro. Nature 1978; 276:726. 11. Stewart AF, Horst R, Deftos LJ, et al. Biochemical evaluation of patients with cancer-associated hypercalcemia: evidence for humoral and nonhumoral groups. N Engl J Med 1980; 303:1377. 12. Godsall JW, Burtis WJ, Insogna KL, et al. Nephrogenous cyclic AMP, adenylate cyclase-stimulating activity, and the humoral hypercalcemia of malignancy. Recent Prog Horm Res 1986; 40:705. 13. Nakayama K, Fukumoto S, Takeda S, et al. Differences in bone and vitamin D metabolism between primary hyperparathyroidism and malignancy-associated hypercalcemia. J Clin Endocrinol Metab 1996; 81:607. 14. Burtis WJ, Brady TG, Orloff JJ, et al. Immunochemical characterization of circulating parathyroid hormone-related protein in patients with humoral hypercalcemia of malignancy. N Engl J Med 1990; 322:1106. 15. Budayr AA, Nissenson RA, Klein RF, et al. Increased serum levels of a parathyroid hormonelike protein in malignancy-associated hypercalcemia. Ann Intern Med 1989; 111:807. 16. Motokura T, Fukumoto S, Matsumoto T, et al. Parathyroid hormone related protein in adult T-cell leukemia-lymphoma. Ann Intern Med 1989; 111:484. 17. Isales C, Carcangiu ML, Stewart AF. Hypercalcemia in breast cancer: a reassessment of the mechanism. Am J Med 1987; 82:1143. 18. Grill V, Ho P, Body JJ, et al. Parathyroid hormone-related protein: elevated levels in both humoral hypercalcemia of malignancy and hypercalcemia complicating metastatic breast cancer. J Clin Endocrinol Metab 1991; 73:1309. 19. Plimpton CH, Gellhorn A. Hypercalcemia in malignant disease without evidence of bone destruction. Am J Med 1956; 21:750. 20. Stewart AF, Vignery A, Silvergate A, et al. Quantitative bone histomorphometry in humoral hypercalcemia of malignancy: uncoupling of bone cell activity. J Clin Endocrinol Metab 1982; 55:219. 20a. Rabbani SA. Molecular mechanism of action of parathyroid hormone related peptide in hypercalcemia of malignancy: therapeutic strategies. Int J Oncol 2000; 16:197. 21. Halloran BD, Nissenson BP. Parathyroid hormone-related protein: normal physiology and its role in cancer. Boca Raton, FL: CRC Press, 1992. 22. Burtis WJ, Wu T, Bunch C, et al. Identification of a novel 17,000 dalton PTH-like adenylate cyclase stimulating protein from a tumor associated with humoral hypercalcemia of malignancy. J Biol Chem 1987; 262:7151. 23. Moseley JM, Kubota M, Diefenbach-Jagger H, et al. Parathyroid hormone-related protein purified from a human lung cancer cell line. Proc Natl Acad Sci U S A 1987; 84:5048. 24. Strewler GJ, Stern PH, Jacobs JW, et al. Parathyroid hormone-related protein from human renal carcinoma cells. J Clin Invest 1987; 80:1803. 25. Orloff JJ, Reddy DR, dePapp AE, et al. Parathyroid hormone-related protein as a prohormone: posttranslational processing and receptor interactions. Endocr Rev 1994; 15:40. 26. Stewart A, Mangin M, Wu T, et al. A synthetic human parathyroid hormone-like protein stimulates bone resorption and causes hypercalcemia in rats. J Clin Invest 1988; 81:596. 27. Everhart-Caye M, Inzucchi SE, Guinness-Henry J, Mitnick MA, Stewart AF. Parathyroid hormone-related protein(1-36) is equipotent with parathyroid hormone(1-34) in humans. J Clin Endocrinol Metab 1996; 81:199 28. Kukreja SC, Shevrin DH, Wimbiscus SA, et al. Antibodies to PTH-related protein lower serum calcium in athymic mouse models of malignancy-associated hypercalcemia due to human tumors. J Clin Invest 1988; 82:1798. 29. Henderson J, Bernier S, D'Amour, et al. Effect of passive immunization against PTH-like peptide and PTH in hypercalcemia tumor-bearing rats and normocalcemic controls. Endocrinology 1990; 127:1310. 30. Bonjour J-P, Phillipe J, Guelpa G. Bone and renal components in hypercalcemia of malignancy and response to a single infusion of clodronate. Bone 1988; 9:123. 31. Coombes RC, Ward MK, Greenberg PB, et al. Calcium metabolism in cancer. Cancer 1976; 38:211. 32. Breslau NA, McGuire JL, Zerwekh JR, et al. Hypercalcemia associated with increased serum calcitriol levels in three patients with lymphoma. Ann Intern Med 1984; 100:1. 33. Rosenthal NR, Insogna KL, Godsall JW, et al. 1,25 dihydroxyvitamin D-mediated humoral hypercalcemia in malignant lymphoma. J Clin Endocrinol Metab 1985; 60:29. 34. Adams JS, Fernandez M, Gacad MA, et al. Vitamin D metabolitemediated hypercalcemia and hypercalcemia in patients with AIDS- and non-AIDS-associated lymphoma. Blood 1989; 73:235. 35. Nussbaum SR, Gaz RD, Arnold A. Hypercalcemia and ectopic secretion of parathyroid hormone by an ovarian carcinoma with rearrangement of the gene for PTH. N Engl J Med 1990; 323:1324. 36. Nielsen PK, Rasmussen AK, Feldt-Rasmussen U, et al. Ectopic production of intact parathyroid hormone by a squamous cell lung carcinoma in vivo and in vitro. J Clin Endocrinol Metab 1996; 81:3793. 37. Stewart AF, Broadus AE, Schwartz PE, et al. Hypercalcemia in gynecologic neoplasms. Cancer 1982; 49:2389. 38. Burman KD, Monchik JM, Earll JM, Wartofsky L. Ionized and total serum calcium and parathyroid hormone in hyperthyroidism. Ann Intern Med 1976; 84:668. 39. Ross DS, Nussbaum SR. Reciprocal changes in parathyroid hormone and thyroid function after radioiodine treatment of hyperthyroidism. J Clin Endocrinol Metab 1989; 68:1216. 40. Peerenboom H, Keck E, Kruskeniper HL, Strohmeyer G. The defect in intestinal calcium transport in hyperthyroidism and its response to therapy. J Clin Endocrinol Metab 1984; 59:936. 41. Stewart AF, Hoecker J, Segre GV, et al. Hypercalcemia in pheochromocytoma: evidence for a novel mechanism. Ann Intern Med 1985; 102:276. 42. Miller SS, Sizemore GW, Sheps SG, Tyce GM. Parathyroid function in patients with pheochromocytoma. Ann Intern Med 1975; 82:372. 43. Mune T, Katakami H, Kato Y. Production and secretion of PTH-related protein in pheochromocytoma. J Clin Endocrinol Metab 1993; 76:757. 44. Pederson KO. Hypercalcemia in Addison's disease. Acta Med Scand 1967; 181:691. 45. Muls E, Bouillon R, Boelart J, et al. Etiology of hypercalcemia in a patient with Addison's disease. Calcif Tissue Int 1982; 34:523. 46. Verner JV, Morrison AB. Endocrine pancreatic islet disease with diarrhea. Arch Intern Med 1974; 133:492. 47. Haussler MR, McCain TA. Basic and clinical concepts related to vitamin D metabolism and action. N Engl J Med 1977; 297:974. 48. Markowitz ME, Rosen JF, Smith C, DeLuca HF. 1,25-Dihydroxyvitamin D3- treated hypoparathyroidism: 35 patient years in 10 children. J Clin Endocrinol Metab 1982; 55:727. 49. Jacobus CH, Holick MF, Shao Q. Hypervitaminosis D associated with drinking milk. N Engl J Med 1992; 326:1173. 50. Holick MF, Shau Q, Liu WW, Chen TC. The vitamin D content of fortified milk and infant formula. N Engl J Med 1992; 326:1178. 51. Hofman KJ, Milne FJ, Schmidt C. Acne, hypervitaminosis A, and hypercalcemia. S Afr Med J 1978; 54:579. 52. Ragavan W, Smith JE, Bilezikian JP. Vitamin A toxicity and hypercalcemia. Am J Med Sci 1982; 283:161. 53. Valente JP, Elias AN, Weinstein GD. Hypercalcemia associated with oral isotretinoin in the treatment of severe acne. JAMA 1983; 250:1899. 54. Villablanca JG, Khan AA, Avramis VI, et al. Phase I trial of cis-retinoic acid in children with neuroblastoma following bone marrow transplantation. J Clin Oncol 1995; 13:894. 55. Christiansen C, Baastrup PC, Lindgreen P, Transbol I. Endocrine effects of lithium. Acta Endocrinol (Copenh) 1978; 88:528. 56. Spiegel AM, Rudorfer MV, Marx SJ, Linnoila M. The effect of short-term lithium administration on suppressibility of parathyroid hormone secretion by calcium in vivo. J Clin Endocrinol Metab 1984; 59:354. 57. Haden ST, Stoll AL, McCormick S, et al. Alterations in parathyroid dynamics in lithium-treated subjects. J Clin Endocrinol Metab 1997; 82:2844. 58. Porter RH, Cox BG, Heaney D, et al. Treatment of hypoparathyroid patients with chlorthalidone. N Engl J Med 1978; 298:577. 59. Koppel MH, Massry SG, Shinaberger JH, et al. Thiazide-induced rise in serum calcium and magnesium in patients on maintenance hemodialysis. Ann Intern Med 1970; 72:895. 60. Legha SS, Powell K, Buzdar AU, Blumenachein GR. Tamoxifen-induced hypercalcemia in breast cancer. Cancer 1981; 47:2803. 61. Valentin-Opran A, Eilon G, Saez S, Mundy GR. Estrogens and antiestrogens stimulate release of bone resorbing activity in cultured human breast cancer cells. J Clin Invest 1985; 75:726. 62. Knox JB, Demling RH, Wilmore DW, Sarraf P, Santos AA, Hypercalcemia associated with the use of growth hormone in an adult surgical intensive care unit. Arch Surg 1995; 130:442. 63. Sakoulas G, Tritos NA, Lally M, et al. Hypercalcemia in an AIDS patient treated with growth hormone. AIDS 1997; 11:1353. 64. McPherson ML, Prince SR, Atamer ER, et al. Theophylline-induced hypercalcemia. Ann Intern Med 1986; 105:52. 65. Gayet S, Ville E, Durand JM, et al. Foscarnet-induced hypercalcemia in AIDS. AIDS 1997; 11:1068. 66. Tsoukas CD, Provvedini DM, Manolagas SC. 1,25-Dihydroxyvitamin D 3: a novel immunoregulatory hormone. Science 1984; 224:1438. 67. Amento EP. Vitamin D and the immune system. Steroids 1987; 49:55. 68. Gkonos PJ, London R, Hendler ED. Hypercalcemia and elevated 1,25(OH)2D levels in a patient with end stage renal disease and active tuberculosis. N Engl J Med 1984; 311:1683. 69. Adams JS, Singer FR, Gacad MA, et al. Isolation and structural identification of 1,25-dihydroxyvitamin D produced by cultured alveolar macrophages in sarcoidosis. J Clin Endocrinol Metab 1985; 60:960. 70. Peris P, Font J, Grau JM, et al. Calcitriol-mediated hypercalcemia and increased interleukins in a patient with sarcoid myopathy. Clin Rheumatol 1999; 18:488. 71. Murray JJ, Helm CR. Hypercalcemia in disseminated histoplasmosis. Am J Med 1985; 78:881. 72. Kozemy GA, Barbato AL, Bansal VK, et al. Hypercalcemia associated with silicone-induced granulomas. N Engl J Med 1984; 311:1103. 73. Adams JS, Sharma OP, Gacad MA, Singer FR. Metabolism of 25-hydroxyvitamin D3 by cultured pulmonary alveolar macrophages in sarcoidosis. J Clin Invest 1983; 72:1856. 74. Mason RS, Frankel TI, Chan YL, et al. Vitamin D conversion by sarcoid lymph node homogenate. Ann Intern Med 1984; 100:59. 75. Watson L, Moxham J, Fraser P. Hydrocortisone suppression test and discriminant analysis in differential diagnosis of hypercalcemia. Lancet 1980; 1:1320. 76. Stewart AF, Alder M, Byers CM, et al. Calcium homeostasis in immobilization: an example of resorptive hypercalciuria. N Engl J Med 1982; 306:1136. 77. Bergstrom WH. Hypercalciuria and hypercalcemia complicating immobilization. Am J Dis Child 1978; 132:553. 78. Tori JA, Kewalramani LS. Urolithiasis in children with spinal cord injury. Paraplegia 1979; 16:357. 79. Minaine P, Meunier P, Edouard C, et al. Quantitative histological data on disuse osteoporosis. Calcif Tissue Res 1974; 17:57. 80. Beall DP, Scofield H. Milk-alkali syndrome associated with calcium carbonate consumption. Medicine 1995; 74-89. 81. Klein GL, Horst RL, Norman AW, et al. Reduced serum 1,25-dihydroxyvitamin D during long-term total parenteral nutrition. Ann Intern Med 1981; 94:638. 82. Shike M, Sturtridge WC, Tam CS, et al. A possible role of vitamin D in the genesis of parenteral nutrition-induced metabolic bone disease. Ann Intern Med 1981; 95:560. 83. Ott SM, Maloney NA, Klein GL, et al. Aluminum is associated with low bone formation in patients receiving chronic parenteral nutrition. Ann Intern Med 1983; 98:910. 83a. Schwarz P, Larsen NE, Lonborg Friis IM, et al. Familial hypocalciuric hypercalcemia and neonatal severe hyperparathyroidism associated with mutations in the human Ca 2+-sensing receptor gene in three Danish families. Scand J Clin Lab Invest 2000; 60:221. 84. 85. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. Llach F, Felsenfeld AJ, Haussler MR. The pathophysiology of altered calcium metabolism in rhabdomyolysis-induced acute renal failure. N Engl J Med 1981; 305:117. Merlini G, Fitzpatrick IA, Siris ES, et al. A human myeloma immunoglobulin G binding four moles of calcium associated with asymptomatic hypercalcemia. J Clin Immunol 1984; 4:185. John R, Oleesky D, Issa B, et al. Pseudohypercalcemia in two patients with IgM paraproteinemia. Ann Clin Biochem 1997; 34:694. Chandra SV, Seth PK, Mankeshu JK. Manganese poisoning: clinical and biochemical observations. Environ Res 1974; 7:374. Chandra SV, Shukla GS, Srivastava RS. An exploratory study of manganese exposure to welders. Clin Toxicol 1981; 18:407. Gerhardt A, Greenberg A, Reilly JJ, Van Theil DH. Hypercalcemia: a complication of advanced chronic liver disease. Arch Intern Med 1987; 147:274. Yang KH, Stewart AF. Treatment of hypercalcemia. In: Mazzaferri EL, Bar RS, Kreisberg RA, eds. Advances in endocrinology and metabolism, vol 4. St. Louis: Mosby, 1993:305. Binstock ML, Mundy GR. Effect of calcitonin and glucocorticoids in combination on the hypercalcemia of malignancy. Ann Intern Med 1980; 93:269. Perlia CP, Gubisch NJ, Cootter J, et al. Mithramycin treatment of hypercalcemia. Cancer 1970; 25:389. Kennedy BJ. Metabolic and toxic effects of mithramycin during tumor therapy. Am J Med 1970; 49:494. Zerwekh JE, Pak CYC, Kaplan RA, et al. Pathogenic role of 1,25 dihydroxyvitamin D in sarcoidosis and absorptive hypercalciuria: different response to prednisolone therapy. J Clin Endocrinol Metab 1980; 51:381. Lentz RD, Brown DM. Treatment of severe hypophosphatemia. Ann Intern Med 1978; 89:941. Goldsmith RS, Ingbar SH. Inorganic phosphorus in the treatment of hypercalcemia of diverse etiologies. N Engl J Med 1966; 274:1. Bilezikian JP. Management of hypercalcemia. J Clin Endocrinol Metab 1993; 77:1445. Bilezikian JP. Management of acute hypercalcemia. N Engl J Med 1992; 326:1196. Body JJ. Current and future directions in medical therapy: hypercalcemia. Cancer 2000; 88(12 Suppl):3054. Fitton A, McTavish D. Pamidronate: a review of its pharmacological properties and therapeutic efficacy in resorptive bone disease. Drugs 1991; 41:289. Thiebaud D, Jaeger PH, Jacquet AF, et al. Dose-response in the treatment of hypercalcemia of malignancy by a single infusion of the bisphosphonate AHPrBP. J Clin Oncol 1988; 6:762. Nussbaum SR, Younger J, VandePol CJ, et al. Single-dose intravenous therapy with pamidronate for the treatment of hypercalcemia of malignancy: comparison of 30-, 60-, and 90-mg dosages. Am J Med 1993; 95:297.

103. 104. 105. 106. 107. 108. 109. 110. 111. 112. 113. 114. 115. 116. 117. 118. 119. 120. 121.

Gucalp R, Ritch P, Wiernik PH, et al. Comparative study of pamidronate disodium and etidronate disodium in the treatment of cancer-related hypercalcemia. J Clin Oncol 1992; 10:134. Todd PA, Fitton A. Gallium nitrate: a review of its pharmacological properties and therapeutic potential in cancer-related hypercalcemia. Drugs 1991; 42:261. Warrell RP, Isreal R, Frisone M, et al. Gallium nitrate for acute treatment of cancer-related hypercalcemia. A randomized, double blind comparison to calcitonin. Ann Intern Med 1988; 108:669. Warrell RP, Murphy WK, Schulman P, et al. A randomized double-blind study of gallium nitrate compared with etidronate for acute control of cancer-related hypercalcemia. J Clin Oncol 1991; 9:1467. Hegbrun PJ, Selby PL, Peacock M, et al. Peritoneal dialysis in the management of severe hypercalcemia. BMJ 1980; 280:525. Bayat-Moktari F, Palmieri GMA, Momuddin M, Pourmand R. Parathyroid storm. Arch Intern Med 1980; 140:1092. Kaiser W, Biesenbach G, Kramar R, Zazgornik J. Kalziumfreie hamodialyse-stellenwert in der therapie der hyperkalzamischen krise. Klin Wochenschr 1989; 67:86. Brenner DE, Harvey HA, Lipton A, Demers L. A study of prostaglandin E2, parathormone and response to indomethacin in patients with hypercalcemia of malignancy. Cancer 1982; 49:556. Glover DJ, Shaw L, Glick JH, et al. Treatment of hypercalcemia in parathyroid cancer with WR-2721, S-2(3-aminopropylamino)ethylphosphorothi-otic acid. Ann Intern Med 1985; 103:55. Johannssen AJ, Onkelinx C, Rodan GA, Raisz LR. Thionaphthene-2-carboxylic acid: a new antihypercalcemic agent. Endocrinology 1985; 117:1508. Zeimer HJ, Greenaway TM, Slavin J, et al. Parathyroid hormonerelated protein in sarcoidosis. Am J Pathol 1998; 152:17. Bosch X, Lopez-Soto A, Morello A, et al. Vitamin D metabolitemediated hypercalcemia in Wegener's granulomatosis. Mayo Clin Proc 1997; 72:440. Lee JC, Catanzaro A, Parthemore JG, et al. Hypercalcemia in disseminated coccidiomycosis. N Engl J Med 1977; 297:431. Dockrell D, Poland G. Hypercalcemia in a patient with hypoparathyroidism and Nocardia asteroides infection. Mayo Clin Proc 1997; 72:757. Kantarjian HM, Saad MR, Estey EH, et al. Hypercalcemia in disseminated candidiasis. Am J Med 1983; 74:721. Bosch X. Hypercalcemia due to endogenous overproduction of active vitamin D in identical twins with catscratch disease. JAMA 1998; 279:532 Jurney TH. Hypercalcemia in a patient with eosinophilic granuloma. Am J Med 1984; 76:527. Bosch X. Hypercalcemia due to endogenous overproduction of 1,25-dihydroxyvitamin D in Crohn's disease. Gastroenterology 1998; 114:1061. Albitar S, Genin R, Fen-Chong M, et al. Multisystem granulomatous injuries 28 years after paraffin injections. Nephrol Dial Transplant 1997; 12:1974.

CHAPTER 60 HYPOPARATHYROIDISM AND OTHER CAUSES OF HYPOCALCEMIA Principles and Practice of Endocrinology and Metabolism

CHAPTER 60 HYPOPARATHYROIDISM AND OTHER CAUSES OF HYPOCALCEMIA

SUZANNE M. JAN DE BEUR, ELIZABETH A. STREETEN, AND MICHAEL A. LEVINE Mechanisms of Hypocalcemia Hypoalbuminemia Ionized and Bound Fractions Effects of Parathyroid Hormone and Vitamin D Signs and Symptoms of Hypocalcemia Signs of Chronic Hypocalcemia Specific Causes of Hypocalcemia Hypoparathyroidism Parathyroid Hormone Resistance Pseudohypoparathyroidism Type 1 Pseudohypoparathyroidism Type 2 Pseudohypoparathyroidism with Skeletal Responsiveness Bioineffective Parathyroid Hormone Other Causes of Hypocalcemia Hyperphosphatemia Rhabdomyolysis Cancer Disorders of Vitamin D Metabolism Differential Diagnosis of Hypocalcemia Serum Parathyroid Hormone Levels Parathyroid Hormone Infusion Measurement of Serum Vitamin D Metabolites Therapy for Hypocalcemia Emergency Therapy Long-Term Therapy Treatment with Calcium Salts Therapy with Vitamin D or its Metabolites Vitamin D Toxicity Influence of Other Drugs Therapy During Pregnancy Chapter References

MECHANISMS OF HYPOCALCEMIA
Calcium is present in serum in three forms: bound to serum protein (4045%), complexed to inorganic anions (510%), and ionized (~4550%).1 Although total serum calcium is most commonly measured, the ionized fraction is most important physiologically. HYPOALBUMINEMIA Albumin is the major calcium-binding serum protein, and hypoalbuminemia, rather than a decrease in the concentration of ionized calcium, accounts for most cases of low total serum calcium in hospitalized patients. Because reliable direct measurement of ionized serum calcium is not always readily available, a number of algorithms based on albumin or total protein concentrations2,3 have been proposed for the correction of total serum calcium. None of these correction factors should be regarded as absolutely accurate, but they are useful as general indicators of the concentration of ionized calcium in serum. One widely used algorithm estimates that total serum calcium declines by ~0.8 mg/dL for each 1-g/dL decrease in albumin concentration, without a change in ionized calcium. IONIZED AND BOUND FRACTIONS Sudden changes in the distribution of calcium between ionized and bound fractions may cause symptoms of hypocalcemia, even in patients who have normal hormonal mechanisms for the regulation of the ionized calcium concentration. Increases in the extracellular fluid concentration of anions such as phosphate, citrate, bicarbonate, or edetic acid increase the proportion of bound calcium and decrease ionized calcium until intact regulatory mechanisms normalize ionized calcium. Extracellular fluid pH also affects the distribution of calcium between ionized and bound fractions. Acidosis increases the ionized calcium, whereas alkalosis decreases it.1 EFFECTS OF PARATHYROID HORMONE AND VITAMIN D The concentration of extracellular ionized calcium is tightly regulated by parathyroid hormone (PTH) and 1,25-dihydroxyvitamin D [1,25(OH)2D; calcitriol]. PTH has direct effects on bone to regulate calcium exchange at osteocytic sites and to enhance osteoclast-mediated bone resorption. In the kidney, PTH directly enhances distal tubular reabsorption of calcium, decreases the proximal tubular reabsorption of phosphate, and stimulates the metabolic conversion of 25-hydroxyvitamin D [25(OH)D)] to 1,25(OH)2D, the active vitamin D metabolite (Fig. 60-1). The 1,25(OH)2D acts on bone to enhance bone resorption and on the gastrointestinal mucosa to increase absorption of dietary calcium (see Chap. 49, Chap. 50, Chap. 51, Chap. 52, Chap. 53 and Chap. 54). Clinical disorders causing hypocalcemia occur if the production of biologically active PTH or 1,25(OH)2D is impaired or if target organ responses to these hormones are abnormal, either because of a specific biochemical defect or because of generalized target organ damage (Table 60-1).

FIGURE 60-1. Schema for effects of parathyroid hormone (PTH). Secondary dominant hypoparathyroidism-deafness-renal dysplasia (HDR) hyperparathyroidism develops in response to the Hypocalcemia of vitamin D deficiency. (1,25(OH)2D, 1,25-dihydroxyvitamin D; 25(OH)D, 25-hydroxyvitamin D.)

TABLE 60-1. Biochemical Characteristics of Hypocalcemic Disorders

In the hypoparathyroid states in which PTH secretion or action is deficient, the normal effects of PTH on bone and kidney are absent; the efflux of calcium from bone is diminished, and distal renal tubular calcium reabsorption is impaired. In the absence of PTH action, the proximal tubular reabsorption of phosphate is enhanced and hyperphosphatemia is common. The deficiency of PTH action and the hyperphosphatemia result in decreased renal production of 1,25(OH)2D and impaired intestinal calcium absorption. Hypoparathyroidism, therefore, is characterized by a decreased entry of calcium into the extracellular fluid compartment from bone, kidney, and intestine, and is associated with hypocalcemia and hyperphosphatemia. In states of vitamin D deficiency or vitamin D insensitivity, hypocalcemia is caused by decreases in intestinal calcium absorption. The 1,25(OH)2D is a potent stimulator of bone resorption, and its absence may also decrease the availability of calcium from bone. Because the parathyroid glands are intact in the vitamin Ddeficient states, hypocalcemia induces secondary hyperparathyroidism, and renal phosphate clearance is enhanced. Thus, hypocalcemia in vitamin D deficiency results from decreased calcium absorption and a limited availability of calcium from bone despite secondary hyperparathyroidism; characteristically, it is accompanied by hypophosphatemia.

SIGNS AND SYMPTOMS OF HYPOCALCEMIA

Ionized calcium, rather than total calcium, is the primary determinant of symptoms in patients with hypocalcemia. A low extracellular fluid ionized calcium concentration enhances neuromuscular excitability, an effect that is potentiated by hyperkalemia and hypomagnesemia. Substantial variation is seen among patients in the severity of symptoms. Those with chronic hypocalcemia sometimes have few, if any, symptoms of neuromuscular irritability despite quite low total serum calcium concentrations. Patients with acute hypocalcemia often do have symptoms, although no absolute level of serum calcium exists at which symptoms predictably occur. Most patients have at least mild symptoms of circumoral numbness, paresthesias of the distal extremities, ormuscle cramping. Symptoms of fatigue, hyperirritability, anxiety, and depression are common. Severe manifestations of hypocalcemia include carpopedal spasm, laryngospasm, and focal or sometimes life-threatening generalized seizures (which must be distinguished from the generalized tonic muscle contractions that occur in severe tetany). Clinical signs of the neuromuscular irritability associated with latent tetany include Chvostek sign and Trousseau sign. Chvostek sign is elicited by tapping the facial nerve just anterior to the ear to produce ipsilateral contraction of the facial muscles. Slightly positive reactions occur in 10% to 30% of normal adults4; thus, this sign cannot be considered diagnostic of hypocalcemia unless one knows that it previously was absent. Trousseau sign is present if carpal spasm is induced by pressure ischemia of nerves in the upper arm during the inflation of a sphygmomanometer above systolic blood pressure for 3 to 5 minutes.5 Both of these signs can be absent even in patients with definite hypocalcemia. Hypocalcemia is also associated with nonspecific electroencephalographic changes, increases in intracranial pressure, and papilledema. Prolongation of the corrected QT interval on the electrocardiogram is a useful sign of significant hypocalcemia (Fig. 60-2), but other causes of QT prolongation exist. Cardiac dysfunction that reversed with treatment of the hypocalcemia has been reported. This may range from subclinical impairment of cardiac performance that is noted only with exercise6 to life-threatening cardiac failure.7,8 The somatosensory-evoked potential recovery period may be a tool for assessing the effects of and recovery from hypocalcemia.9

FIGURE 60-2. Electrocardiogram of patient with hypocalcemia, demonstrating a long QT interval. The QT interval, which comprises the duration of ventricular depolarization and repolarization, is measured from the beginning of the QRS complex to the end of the T wave (arrows). The interval increases with a decreasing heart rate and may be corrected (QTc) by measuring the RR interval and using the following formula: QTc = QT/(R R). Alternatively, one may consult a nomogram. Serum calcium affects the second or plateau phase (ST segment) of the ventricular action potential; hypocalcemia results in a prolonged QTc interval. In this patient, the QT interval is ~600 msec and the QTc is ~525 msec. The upper limit of normal is 440 msec. (Because the exact end of the T wave may be difficult to determine, some clinicians use the Qa Tc interval, which comprises the onset of the QRS complex to the apex of the T wave.) (Courtesy of Dr. Steven Singh.)

SIGNS OF CHRONIC HYPOCALCEMIA Chronic hypocalcemia is associated with other signs. Ectodermal findings such as dry skin, coarse hair, and brittle nails are common, but they are frequently overlooked. Dental and enamel hypoplasia and absence of adult teeth indicate that hypocalcemia has been present since childhood10 (see Chap. 217). The pattern of dental abnormality may help date the onset of hypocalcemia. Calcification of the frontal lobes and basal ganglia can occur in all forms of hypoparathyroidism and are now detected with computed tomographic scanning even when routine skull radiographs do not demonstrate intracerebral calcification.11 Occasionally, the calcification of the basal ganglia is associated with parkinsonism or chorea, and the prevalence of dystonic reactions to phenothiazines is reported to be high in hypoparathyroid patients.12 Subcapsular cataracts are common in untreated hypoparathyroidism and are best seen with slit-lamp examination (see Chap. 215). Treatment may reverse or decrease the progression of the cataracts. Rickets and osteomalacia, although not characteristic, do occur occasionally in hypoparathyroidism after prolonged hypocalcemia.13,14 Patients with longstanding hypoparathyroidism have been reported to have significantly increased bone mineral density whether they are treated15,16 or untreated.17 Other features exist that are characteristic for the particular disorders which cause hypocalcemia. Recognition of these specific features can be helpful in the differential diagnosis of hypocalcemic states.

SPECIFIC CAUSES OF HYPOCALCEMIA

HYPOPARATHYROIDISM A biochemical state of functional hypoparathyroidism occurs either because of failure of secretion of PTH or, less commonly, failure of PTH action at its target tissues.

The clinical forms and characteristics of hypoparathyroidism are described below and in Table 60-2.

TABLE 60-2. Causes of Hypocalcemia

SURGICAL HYPOPARATHYROIDISM Hypoparathyroidism occurs most commonly as a result of parathyroid or thyroid surgery or after radical surgery for laryngeal or esophageal carcinoma. The resulting hypoparathyroidism can be transient or permanent and sometimes may not develop for many years. In some patients, a chronic state of decreased parathyroid reserve may exist18 in which hypocalcemia becomes apparent only when mineral homeostasis is stressed further by other factors such as pregnancy, lactation, or illness. Transient Hypocalcemia after Parathyroid Surgery. Hypocalcemia frequently occurs after removal of a hyperfunctioning parathyroid adenoma because of deficient secretion of PTH by the remaining previously suppressed parathyroid tissue. Hypoparathyroidism is usually transient, because the normal parathyroid glands recover function quickly (generally within 1 week), even after long-term suppression. Transient postoperative hypocalcemia may be exaggerated or prolonged in those patients who have significant preexisting hyperparathyroid bone disease. In these patients the surgically induced reduction of previously elevated plasma PTH results in an increased movement of plasma calcium (and phosphorus) into remineralizing hungry bones.19 Treatment with calcium and a short-acting vitamin D metabolite may be required until the bones heal. Permanent Hypoparathyroidism after Parathyroid Surgery. Permanent hypoparathyroidism after an initial neck exploration for primary hyperparathyroidism is rare and develops in ~1% of patients. The incidence is greatly increased with repeated neck surgery for recurrent or persistent hyperparathyroidism, after subtotal parathyroidectomy for parathyroid hyperplasia, or when surgery is performed by an inexperienced surgeon. Hypocalcemia after Thyroid Surgery. After thyroid surgery, the incidence of permanent hypoparathyroidism varies widely, depending on the underlying thyroid lesion and the extent of the procedure, as well as on the experience of the surgeon. Hypoparathyroidism may result from direct injury, inadvertent removal, or devascularization of the parathyroid glands. Permanent hypoparathyroidism is distinctly unusual after a hemithyroidectomy and should be relatively uncommon even after total thyroidectomy. 20,21 However, up to 33% of patients who undergo a total thyroidectomy for cancer may develop transient postoperative parathyroid insufficiency.22 Transient hypocalcemia occurs in approximately one-third of patients who undergo a subtotal thyroidectomy for thyrotoxicosis. The fall in plasma calcium level generally occurs within 24 to 48 hours after surgery and can be sufficient to produce symptoms of tetany. The mechanism of this hypocalcemia is not well understood. Frequently, hyperthyroidism is associated with increased bone turnover and resorption, elevated plasma ionized calcium levels, and suppressed parathyroid function. Although the proposal has been made that hypocalcemia occurs as calcium moves into remineralizing hungry bones after reduction of thyroid hormone levels,23 the early development of hypocalcemia appears to be inconsistent with this abnormality. Some patients may have unappreciated damage to the parathyroid glands18; whether thyroidectomy causes hypocalcemia by producing hypercalcitonemia is disputed.24,25 Clearly, whatever the initiating cause, the secretory response of the suppressed parathyroid glands is inadequate to maintain a normal plasma calcium.24,25 IDIOPATHIC HYPOPARATHYROIDISM The term idiopathic hypoparathyroidism describes a heterogeneous group of rare disorders that share in common the deficient secretion of PTH. Although most cases are sporadic, the familial occurrence of idiopathic hypoparathyroidism has been reported. Within these families hypoparathyroidism may occur as part of a complex autoimmune disorder (see Chap. 197) associated with multiple endocrine deficiencies (i.e., type 1 polyglandular syndrome)26 or in association with diverse developmental abnormalities (e.g., nephropathy, lymphedema, nerve deafness, or tetralogy of Fallot).27,28,29,30,31,32,33,34,35,36,37,38,39,40,41 and 42 The pleiotropic nature of many of these various syndromes suggests that the genetic basis of PTH deficiency is not related to a specific defect intrinsic to the parathyroid gland. Type 1 Polyglandular Syndrome. Type 1 polyglandular syndrome may be sporadic or familial with an autosomal recessive inheritance pattern. The classic triad of this syndrome is hypoparathyroidism, adrenal insufficiency, and mucocutaneous candidiasis (HAM). The recognition that affected patients may have additional components has led to the suggestion that a more inclusive term be used to describe the syndrome: autoimmune polyendocrinopathycandidiasisectodermal dystrophy (APECED).43 The syndrome is first recognized in early childhood, although a few individuals have developed the condition after the first decade of life. The clinical onset of the three principal components of the syndrome typically follows a predictable pattern, in which mucocutaneous candidiasis first appears at a mean age of 5 years, followed by hypoparathyroidism at a mean age of 9 years and adrenal insufficiency at a mean age of 14 years.43 Patients may not manifest all three elements of the triad. Alopecia, keratoconjunctivitis, malabsorption and steatorrhea, gonadal failure, pernicious anemia, chronic active hepatitis, thyroid disease, and insulin-requiring diabetes mellitus occur in some patients.26 Antibodies directed against the parathyroid, thyroid, and adrenal glands are demonstrable in many patients,44 and a T-cell abnormality has been described.45,46 The presence of antibodies may not correlate well with the clinical findings. In those cases that have been examined pathologically, complete parathyroid atrophy or destruction has been demonstrated. In some patients, treatment of hypoparathyroidism has been complicated by apparent vitamin D resistance, possibly related to coexistent hepatic disease or steatorrhea, or both. Mutations in the gene AIRE (autoimmune regulator gene) have been identified in patients with APECED. AIRE, located on chromosome 21 (21q22.3), encodes a nuclear protein containing zinc-finger motifs; this suggests that it may play a role as a transcriptional regulator.47,48 AIRE is expressed in tissues important in the development and regulation of the immune system (e.g., the thymus and lymph nodes). Although many different mutations have been reported throughout the world,48a the majority of patients harbor either the R257X mutation or a 13-base-pair deletion in exon 8, each resulting in a truncated AIRE protein.49,50 The function of the AIRE protein remains unknown. The hypothesis has been made that the AIRE protein may regulate the immune response via B- and T-cell stimuli.47 Functional analysis of the AIRE protein and further molecular analysis of the AIRE gene will elucidate the molecular pathogenesis of APECED. Isolated Hypoparathyroidism. Isolated idiopathic hypoparathyroidism, in which PTH deficiency is unassociated with other endocrine disorders or developmental defects, is usually sporadic, but it may occur on a familial basis. Most commonly, the onset of isolated hypoparathyroidism is between the ages of 2 and 10 years, although it may first be recognized in adult life, and the onset may be at any age up to the eighth decade. Females are affected twice as often as males. A high incidence of parathyroid antibodies is seen in patients with isolated idiopathic hypoparathyroidism, and some cases may be examples of incomplete expression of the type 1 polyglandular syndrome (APECED, see earlier). Some patients may possess antibodies that inhibit the secretion of PTH,51 rather than cause parathyroid gland destruction.52 In other cases that have been examined pathologically, fatty replacement53 or atrophy with fatty infiltration and fibrosis 54 has been described. Familial Isolated Hypoparathyroidism. In rare instances, isolated hypoparathyroidism may be familial, with the PTH deficiency being inherited by autosomal dominant, autosomal recessive, or X-linked modes of transmission.55 The age at onset covers a broad range (1 month to 30 years), and the condition is often recognized first in the child, rather than in the parent. parathyroid antibodies are absent. As the preproPTH gene is located on the short arm of chromosome 11, molecular genetic studies of familial isolated hypoparathyroidism have focused first on kindreds in which inheritance of hypoparathyroidism is consistent with an autosomal mode of transmission. In one pedigree in which hypoparathyroidism was inherited in an autosomal dominant manner, DNA sequencing revealed a point mutation (TC) in exon 2 that resulted in the substitution of arginine (CGT) for cysteine (TGT) in the leader sequence of preproPTH.56 The substitution of a charged amino acid in the hydrophobic core of the leader sequence inhibits processing of the mutant preproPTH molecule by signal peptidase56 and is presumed to impair translocation of the mutant and normal proteins across the plasma membrane of the endoplasmic reticulum.

An abnormality of the preproPTH gene has also been found in a consanguineous family with autosomal recessive hypoparathyroidism.57 Affected members of this family are homozygous for a single base transversion (GC) at the exon 2intron 2 boundary. This mutation leads to abnormal processing of the nascent preproPTH mRNA and results in mature transcripts in which exon 1 is spliced to exon 3 (i.e., exon skipping).57 Although the molecular pathophysiology of hypoparathyroidism has been defined in these two families, detailed analyses of the preproPTH gene have failed to disclose defects in affected members of other autosomal recessive and dominant kindreds.57 Familial hypoparathyroidism is also inherited as an X-linked disorder that is, of course, unrelated to the preproPTH gene. Using a battery of X chromosome gene markers, linkage studies of two large multigenerational families with X-linked hypoparathyroidism have localized the defective gene to the region Xq26-27.58,59 The defective gene or genes in this syndrome appear to be important for parathyroid cell development or function.58 The early onset of hypocalcemia and the absence of parathyroid tissue at autopsy in individuals with this disorder are consistent with an important role for this genetic locus in the embryologic development of the parathyroid glands.60 AUTOSOMAL DOMINANT HYPOCALCEMIA: A MODEL OF HUMAN DISEASE DUE TO DEFECTIVE G PROTEINCOUPLED RECEPTOR SIGNALING New insights into the molecular pathology of hypoparathyroidism have come from the cloning and characterization of the calcium-sensing receptor (CaR), a cell surface protein that binds extracellular calcium.61 The CaR is a member of the superfamily of heptahelical receptor proteins that are coupled via signal-transducing G proteins to a variety of intracellular signal effectors (e.g., enzymes such as adenylate cyclase and phospholipase C). G proteincoupled receptors (GPCRs) detect extracellular signals as diverse as light, odorants, hormones, growth factors, neurotransmitters, and ions, and interact with heterotrimeric guanine nucleotidebinding proteins (G proteins), which couple the extracellular receptors to intracellular effector enzymes and ion channels. GPCR mutations may be activating or inactivating. The CaR is present in a variety of tissues, but its expression on parathyroid and renal tubular cells is the basis for its role in regulating serum and urinary calcium levels. Expression of the CaR on the surface of parathyroid cells is required for calcium-sensitive regulation of PTH secretion, and changes in the number or activity of CaRs can alter the calcium setpoint for PTH secretion.62 Although the precise molecular basis for calcium-dependent regulation of PTH secretion remains unknown, the cloning of the CaR has led to the identification of the initial steps in the signaling pathway. Binding of extracellular calcium to CaRs on the surface of the parathyroid cell leads to activation of Gq, a G protein that stimulates phospholipase C activity. Phospholipase C hydrolyzes phosphoinositides, leading to the generation of inositol 1,4,5-triphosphate, which releases intracellular calcium from storage in the endoplasmic reticulum, and diacylglycerol, which activates protein kinase C.63 The calcium-dependent activation of these second messenger pathways ultimately inhibits secretion of PTH (see Chap. 50). Heterozygous mutations in the CaR gene that lead to constitutive (ligand-independent) activation of CaRs (i.e., gain of function mutations) have been identified in several kindreds with autosomal dominant hypocalcemia, a syndrome associated with low serum PTH and relative hypercalciuria.64 In other cases, linkage of hypocalcemia to the chromosomal locus for the CaR gene (3q13.3-21) has indirectly implicated this gene in familial hypoparathyroidism.65 Preliminary studies have identified similar activating mutations of the CaR in cases of sporadic hypoparathyroidism. In both familial and sporadic cases, each proband has a unique mutation. These observations suggest that mutation of the CaR gene may be the most common cause of genetic hypoparathyroidism. Further confirmation of the important role that the CaR plays in regulating PTH secretion derives from studies of patients with the contrasting syndrome of familial (benign) hypocalciuric hypercalcemia (FHH), an autosomal dominant disorder associated with excessive secretion of PTH due to reduced sensitivity of the parathyroid glands to extracellular calcium.66,67 and 68 Heterozygous mutations that result in loss of function of the CaR are present in most patients with FHH. Although FHH is typically a benign condition, in unusual cases, an affected subject may develop severe neonatal hyperparathyroidism, a life-threatening hypercalcemic disorder that is generally associated with inheritance of defective CaR genes from both parents.69,70 Defects in other GPCRs have been identified as the bases for a variety of human diseases. Several syndromes of inherited hormonal resistance result from germline loss of function mutations in GPCR, including the PTH/parathyroid hormonerelated protein (PTHrP),71,72 thyroid-stimulating hormone (TSH),73,74 luteinizing hormone (LH),75,76 growth hormonereleasing hormone (GHRH),77,78,79 and 80 adrenocorticotropic hormone (ACTH),81,82 V2 vasopressin,83 follicle-stimulating hormone (FSH),84 and gonadotropin-releasing hormone (GnRH)85 receptors (Table 60-3). With the exception of the CaR, inheritance of these resistance syndromes is recessive. Gain of function mutations may be either germline or sporadic and are typically heterozygous. Somatic activating mutations in the TSH receptor result in ligand-independent stimulation and formation of hyperfunctioning thyroid nodules.86 When similar TSH receptoractivating mutations occur in the germline, familial nonauto-immune hyperthyroidism results.87,88 Activating mutations of PTH/PTHrP89 and LH receptors,90,91 which also have been identified (see Table 60-3), result in ligand-independent activation of hormone-sensitive signaling pathways. As previously noted for the CaR, many GPCRs have been associated with both gain of function and loss of function mutations, with generation of contrasting clinical syndromes. Another relevant example is the PTH/PTHrP receptor. Heterozygous activating mutations result in Jansen metaphyseal chondrodysplasia, a rare form of dwarfism with PTH-independent hypercalcemia and abnormal endochondral bone formation secondary to constitutive activity of the PTH/PTHrP receptor.89 Homozygous inactivating mutations in the PTH/PTHrP receptor cause a rare lethal disorder of accelerated endochondral bone maturation known as Blomstrand chondrodysplasia.71,72

TABLE 60-3. G ProteinCoupled Receptor Mutations and Related Endocrine Disorders

DEVELOPMENTAL DISORDERS OF THE PARATHYROID GLAND Hypoparathyroidism may result from agenesis or dysgenesis of the parathyroid glands. The most well-described example of parathyroid gland dysembryogenesis is the DiGeorge syndrome (DGS), in which maldevelopment of the third and fourth branchial pouches is frequently associated with congenital absence of not only the parathyroids but also the thymus. Because of thymic aplasia, T-cellmediated immunity is impaired, and affected infants have an increased susceptibility to recurrent viral and fungal infections. Maldevelopment of the first and fifth branchial pouches occurs frequently as well, producing characteristic facial anomalies (Fig. 60-3), including hypertelorism, antimongoloid slant of the eyes, low-set and notched ears, short philtrum of the lip, and micrognathia (first branchial pouch) or aortic arch abnormalities, such as right-sided arch, truncus arteriosus, or tetralogy of Fallot. Most cases of branchial pouch dysembryogenesis are sporadic, but familial occurrence with apparent autosomal dominant inheritance has been described.30,31 Although most children with DGS die of infections or cardiac failure by the age of 6 years, survival into adolescence or adulthood is possible when the syndrome is only partially expressed. Molecular mapping studies have demonstrated an association between the syndrome and deletions involving 22q1132,33 and 34 or 10p.35,36 Large deletions of genetic material at 22q11 result in hemizygosity for genes located in this region, and they are associated with contiguous gene deletion syndromes, which include not only the DGS but also the overlapping conotruncal anomaly and velocardiofacial syndromes. More than 20 candidate genes from the deleted region in DGS have been identified, and none has been shown to cause DGS.92 Evidence both from the hand2 knock-out mouse, which has characteristics that resemble DGS,93 and from a DGS patient with a small deletion in the region of the ubiquitin fusion degradation 1 (UFD1L) gene has implicated heterozygous loss of function of this gene as the basis for the DGS phenotype.93,94 Underexpression of UFD1L may lead to cell death in the pharyngeal arches in embryonic development. UFD1L regulates the accumulation of protein substrates via finely tuned protein degradation. Accumulation of excess protein may be toxic and lead to aberrant cell signaling, abnormal cellular proliferation, or apoptosis of the pharyngeal arches where UFD1L is expressed.92 Further investigation demonstrating mutations in the UFD1L gene alone in patients with DGS are needed to validate UFD1L as the DGS gene.

FIGURE 60-3. An 18-month-old boy with DiGeorge syndrome. Initially, he presented on the third day of life with a seizure that did not respond to phenobarbital. On examination, mild hypertelorism (increased inter-pupillary distance); slight antimongoloid slant of the eyes (outer can-thus lower than inner canthus); asymmetric, malformed ears that were low set; and a short philtrum (infranasal groove) were noted. The serum calcium level was 6.8 mg/dL, the serum phosphate level was 6.8 mg/dL, and the chest radiograph revealed a right-sided aortic arch. No thymus shadow was present. The seizures responded to calcium gluconate therapy. The subsequent course was characterized by upper respiratory tract infections and episodes of otitis media. When the patient was seen again at the time of the photograph, retardation of growth and mental development were observed. Note the broad nose, cupid bow mouth, and mandibular hypoplasia. The serum calcium concentration was 8.2 mg/dL, and the serum phosphate level was 7.8 mg/dL. Calcium gluconate therapy was discontinued without a further drop in serum calcium levels or a recurrence of seizures. (From Kretschmer R, Say B, Brown D, Rosen F. Congenital aplasia of the thymus gland [DiGeorge's syndrome]. N Engl J Med 1968; 279:1295.)

Hypoparathyroidism is also associated with several other less well understood developmental syndromes, including the autosomal dominant hypoparathyroidism-deafness-renal dysplasia (HDR) syndrome associated with sensorineural deafness and renal dysplasia, which has been linked to deletions on 10p 13-1427,95,95a that result in loss of the GATA3 gene.95b Mutations in the GATA3 gene, which encodes a zinc-finger transcription factor, are also a cause of HDR.95 The syndrome of lymphedema, prolapsing mitral valve, brachydactyly, and nephropathy29; and severe growth failure and dysmorphic features that include microcephaly, beaked nose, and micrognathia40,42 as well as the Kenney-Caffey syndrome (short stature, osteosclerosis, basal ganglion calcifications, ophthalmic defects) are also associated with hypoparathyroidism.60 Studies in mice show that loss of the GCMB gene, which encodes a transcription factor, results in hypoparathyroidism.95c Thus, this gene is the first specific regulator of parathyroid embryogenesis. Similar inactivating mutations in the GCMB gene have been described in patients with neonatal isolated hypoparathyroidism.95d TRANSIENT HYPOPARATHYROIDISM OF THE NEONATE Shortly after birth, a physiologic fall occurs in the serum calcium concentration, and many normal infants may have serum calcium levels <8 mg/dL during the first 3 weeks of life. Hypocalcemia in the neonate can be divided into early hypocalcemia, starting within the first 24 to 72 hours of life before feedings have been given, and late hypocalcemia, usually appearing after several days to weeks of feeding. Early Neonatal Hypocalcemia. Early neonatal Hypocalcemia represents an exaggeration of the normal fall in serum calcium concentration and theoretically is due to deficient release of PTH by immature parathyroid glands. Prematurity, low birth weight, hypoglycemia, maternal diabetes, difficult delivery, and respiratory distress syndrome are frequently associated findings. Symptoms may be absent, or irritability, muscular twitching, or convulsive seizures may occur. Although the course is self-limited, symptomatic infants should be treated with oral or intravenous calcium. Transient congenital hypoparathyroidism can also occur in infants with DGS.95e A more severe form of transient neonatal hypoparathyroidism and tetany may also occur in children born to mothers with hyperparathyroidism or hypercalcemia. In these infants, parathyroid activity has been more profoundly suppressed in utero by maternal hypercalcemia. Late Neonatal Hypocalcemia. Late neonatal hypocalcemia may develop 4 to 6 days (or later) after birth and may be considered a transient form of relative immaturity of renal phosphorus handling or of the renal adenylate cyclase system. Hypocalcemia may be precipitated by a high-phosphate diet and appears to occur particularly in those infants who are fed with artificial foods such as cow's milkbased formulas.96 In these infants, the renal response to PTH is inadequate and hypocalcemia results. The reduction of serum calcium ion concentration is probably secondary to elevated serum phosphate levels and should stimulate parathyroid gland activity. This form of hypocalcemia is the most common cause of seizures in the newborn period. A spontaneous recovery of normal mineral homeostasis typically occurs after a few weeks, but the serum calcium levels of symptomatic infants can be increased within 1 to 2 days by feeding a supplemented milk mixture with a high (3:1 to 4:1) calcium/phosphorus ratio. OTHER FORMS OF PARATHYROID GLAND DYSFUNCTION Irradiation and Drugs. The parathyroid glands appear remarkably resistant to damage by a great many toxic agents and processes. Transient hypoparathyroidism has been associated with ingestion of large quantities of alcohol.97 The administration of iodine-131 for the treatment of benign or malignant thyroid disease or for the deliberate induction of hypoparathyroidism has only rarely caused permanent, symptomatic hypoparathyroidism. Similarly, parathyroid gland function is altered only occasionally by most chemotherapeutic or cytotoxic agents. Notable exceptions include asparaginase, which causes parathyroid necrosis in rabbits, and ethiofos, a radio- and chemoprotector that causes a dose-dependent and reversible inhibition of PTH secretion.98,99 Along with its effects on the parathyroid gland, ethiofos blocks the ability of osteoclasts to respond to hormonal stimuli. Thus, a significant component of its calcium-reducing effect derives from the ability of this agent to inhibit osteoclast-directed bone resorption and calcium release from bone. Infiltrative Disease of the Parathyroids. Parathyroid gland function may also be impaired by infiltrative processes. Iron overload caused by hemochromatosis (see Chap. 131) or transfusion therapy is frequently associated with significant parathyroid gland iron deposition and, occasionally, clinical hypoparathyroidism. Moreover, a similar pathologic picture has been described in one patient with Wilson disease and increased copper storage who developed symptomatic hypoparathyroidism.100 Pathologic involvement of the parathyroid glands can also occur in metastatic neoplasia, miliary tuberculosis, amyloidosis, and syphilis, but clinical hypoparathyroidism rarely occurs in these conditions. Magnesium Deficiency. Reversible alterations of parathyroid gland function and PTH secretion are associated with magnesium depletion. Modest declines in magnesium slightly increase PTH secretion. However, as magnesium deficiency becomes more severe, PTH secretion becomes inappropriately low,101,102 and hypocalcemia with tetany may ensue. With even more severe magnesium depletion, resistance to the action of PTH occurs, which is reversible with magnesium repletion (see Chap. 68).103 By contrast, magnesium replacement does not appear to reverse the hypoparathyroidism associated with burns and hypomagnesemia.103a Hypermagnesemia. Hypermagnesemia also may cause hypocalcemia.104 This situation is commonly encountered in obstetric practice when high-dose magnesium infusions are used for the treatment of toxemia or premature labor. Because the hypocalcemia is accompanied by significant hypermagnesemia, neuromuscular irritability should be less than that expected when similar calcium concentrations occur with normal magnesium; clinical tetany usually does not occur.

PARATHYROID HORMONE RESISTANCE

PSEUDOHYPOPARATHYROIDISM TYPE 1 The term pseudohypoparathyroidism (PHP) describes a heterogeneous syndrome characterized by biochemical hypoparathyroidism (i.e., hypocalcemia and hyperphosphatemia), increased plasma levels of PTH, and peripheral unresponsiveness to the biologic actions of PTH. Thus PHP differs substantially from true hypoparathyroidism; in contrast with the latter condition, PTH secretion is excessive and the parathyroid glands are hyperplastic.105 PHP was the first recognized human disease to be ascribed to diminished responsiveness to a hormone by otherwise normal target organs (subsequently, many others have been described) (see below).106 Albright's initial description of the blunted calcemic and phosphaturic response to PTH administration in patients with PHP provided the basis for his original hypothesis that the disorder is due to target organ resistance.107 MOLECULAR BASIS FOR PSEUDOHYPOPARATHYROIDISM: A MODEL FOR HUMAN G PROTEIN DISEASE

Characterization of the molecular basis for PHP commenced when research showed that cyclic adenosine monophosphate (cAMP) mediates the actions of PTH on kidney and bone and that PTH infusion in humans leads to a significant increase in urinary excretion of cAMP.108,109 When these observations were applied to the study of PHP, affected persons were found to have a blunted urinary response of nephrogenous cAMP to PTH infusion in comparison with normal persons and patients with other forms of hypoparathyroidism (Fig. 60-4).109 These findings formed the basis for the most reliable test presently available for the diagnosis of PHP type 1. Moreover, these results suggested that PTH resistance is caused by a defect in the plasma membranebound PTH receptoradenylate cyclase complex that produces cAMP. Evidence is accumulating that some actions of PTH may be cAMP independent, with signal transduction through inositol phospholipids, intracellular calcium mobilization, and protein kinase C activation (Fig. 60-5).110,111,112 and 113 This may explain why some patients with PHP who have blunted urinary cAMP responses after PTH infusion still have evidence for PTH-mediated effects on renal tubular calcium handling114,115 and skeletal remodeling (see later).

FIGURE 60-4. Cyclic adenosine monophosphate (AMP) excretion in urine in response to the injection of parathyroid hormone (300 USP units of bovine parathyroid extract infused from 9:00 to 9:15 a.m.). (Reproduced from Chase LR, Melson GL, Aurbach GD. Pseudohypo-parathyroidism: defective excretion of 3'5'-AMP in response to parathyroid hormone. J Clin Invest 1969; 48:1836, by permission of the authors and the American Society for Clinical Investigation.)

FIGURE 60-5. Schematic outline of the adenylate cyclasecyclic adenosine monophosphate (cAMP) system. Hs and Hi denote stimulatory and inhibitory agents, respectively; Rs and Ri stimulatory and inhibitory receptors; and Gs and Gi the stimulatory and inhibitory guanine nucleotidebinding regulatory proteins. The rate of conversion of substrate adenosine triphosphate (ATP) to product cAMP by the catalytic (C) unit of adenylate cyclase is regulated by the interactions of Gs and Gi. Details of the interactions are described in the text.

The adenylate cyclase system is far more complex than originally suspected, consisting of at least three types of proteins embedded in the plasma membrane (i.e., receptors, G proteins, and adenylate cyclase [Fig. 60-6; see Chap. 4]). 116 The signal-transducing G proteins are comprised of three subunits (abg) that are encoded by 16a, 6b, and 12g genes. The a subunits are loosely associated with tightly coupled bg dimers; this facilitates great combinatorial variability and allows G proteins to interact with as many as 1000 different GPCRs and effector proteins.117 G proteins have been divided into four subfamilies: Gs, Gi, Gq, and G12, based on structural and functional similarities. Members of the G s family activate adenylate cyclase and members of the Gq family activate phospholipase C; members of the Gi subfamily inhibit adenylate cyclase (Gi) or participate in visual (Gt1 and Gt2) or gustatory (G gust) neurosensory processes. G13 stimulates the exchange of sodium and hydrogen ions and cytoskeletal rearrangements.117,118 and 119 The activity of the heterotrimeric protein is regulated by the binding and hydrolysis of guanosine triphosphate (GTP) by the Ga subunit. When Gb is bound to GDP, it is inactive and loosely associates with the bg dimer. Receptor activation results in release of GDP by the heterotrimeric abg complex, binding of GTP to Ga, and dissociation of the Ga-GTP complex from the bg dimer and receptor. Both the Ga-GTP chain and free bg dimer can regulate downstream effectors. Termination of receptor signaling results from the hydrolysis of GTP to guanosine diphosphate (GDP), allowing the bg dimer to reassociate with Ga-GDP.117

FIGURE 60-6. Cell surface receptors for parathyroid hormone (PTH) appear to be coupled to two classes of G proteins. Gs mediates stimulation of adenylate cyclase (AC) and the production of cyclic adenosine monophosphate (cAMP), which in turn activates protein kinase A (PKA). Gq stimulates phospholipase C (PLC) to form the second messengers inositol 1,4,5-trisphosphate (IP3) and diacylglycerol (DAG) from membrane-bound phosphatidylinositol 4,5-bisphosphate (PIP2). IP3 increases intracellular calcium (Ca2+) and DAG stimulates protein kinase C (PKC) activity. Each G protein consists of a unique a chain and a bg dimer. (ATP, adenosine triphosphate; PTH-R, parathyroid hormone receptor.) (From Bilezikian JP, Levine MA, Marcus R, eds. The parathyroids: basic and clinical concepts. New York: Raven Press, 1994:782.)

Mutations in G proteins tend to affect a variety of tissues, as demonstrated by the occurrence of multihormonal resistance in PHP type 1a. By contrast, receptor mutations are generally limited to one or a few tissues in which the defective receptor is expressed (e.g., defects in the CaR produce a disturbance in calcium sensing that appears limited to parathyroid and kidney cells). The determinants of phenotypic expression of the G protein and GPCR disease are the range of expression of the gene (tissue distribution), the developmental timing of the mutation (germline vs. somatic), and the nature of the mutation (loss or gain of function). PSEUDOHYPOPARATHYROIDISM TYPE 1A Albright's original description of PHP focused on PTH resistance in this disorder. Resistance to PTH alone would be consistent with a defect in the cell surface receptor specific for PTH. However, some patients with PHP type 1 are resistant to multiple hormones whose effects are mediated by cAMP and have additional abnormalities, such as hypothyroidism and hypogonadism,120 mental retardation,121 and defective olfaction.122 These patients, whose disorder is referred to as PHP type 1a, have ~

50% reduction in the activity of Gs (Fig. 60-7). This reduction occurs in all tissues that have been examined (including erythrocytes, platelets, fibroblasts, transformed lymphocytes, and, in one case, renal cortex).123,124 and 125 A generalized deficiency of Gs activity apparently results in a reduced ability of hormones and neurotransmitters to activate adenylate cyclase in diverse tissues and thereby leads to widespread hormone resistance (Fig. 60-8).

FIGURE 60-7. Gs activity in pseudohypoparathyroidism (PHP) type 1a and in pseudopseudohypoparathyroidism (pseudoPHP). Gs activity was measured in erythrocyte membrane extracts by complementation with Gs-deficient membranes from S49 cyc cells. The resultant adenylate cyclase activity is expressed as a percentage of a pooled normal human erythrocyte membrane standard. (From Levine MA, Jap T, Mauseth RS, et al. Activity of the stimulatory guanine nucleotidebinding protein is reduced in erythrocytes from patients with pseudohypoparathyroidism and pseudopseudohypoparathyroidism: biochemical, endocrine, and genetic analysis of Albrights hereditary osteodystrophy in six kindreds. J Clin Endocrinol Metab 1986; 62:497.)

FIGURE 60-8. Model showing some of the multiple receptors that are coupled by means of Gs to activation of adenylate cyclase. Gene mutations that reduce Gs a activity commonly impair transmembrane signal transduction processes and result in hormone resistance. By contrast, activating mutations of Gsa produce constitutive signal transduction in the absence of hormone or neurotransmitters and can result in autonomous cell function and proliferation. (ADH, antidiuretic hormone; PTH, parathyroid hormone; TSH, thyroid-stimulating hormone; LH, luteinizing hormone; ACTH, adrenocorticotropic hormone; R, receptor; C, catalytic unit of adenylate cyclase; ATP, adenosine triphosphate; cAMP, cyclic adenosine monophosphate.)

Patients with PHP type 1a show a unique phenotype characterized by round facies, short stature, obesity, brachydactyly (short metacarpal and metatarsal bones), heterotopic subcutaneous ossification, and bony exostoses (Fig. 60-9). Mental retardation may be present. In 1942, Albright and coworkers107 first observed these unusual developmental defects, subsequently termed Albright hereditary osteodystrophy (AHO), in three patients during the initial description of PHP.

FIGURE 60-9. A, Young woman with Albright hereditary osteodystrophy. Note the short stature and rounded facies. B, Brachydactyly, particularly manifest in many patients as shortened fourth fingers. A dimpling of the fourth and fifth knuckle region may be seen when a fist is made. C, The radiograph reveals characteristic shortening of the fourth metacarpal, as well as shortening of the fifth metacarpal.

Ten years after the description of PHP, Albright and colleagues126 described a patient with a habitus typical of AHO but who lacked biochemical evidence of target organ resistance to PTH. This normocalcemic variant of AHO was termed pseudopseudohypoparathyroidism (pseudoPHP) to call attention to the physical similarity with AHO yet indicate the metabolic dissimilarity (PTH responsiveness) of this disorder. PseudoPHP is genetically related to PHP. Early clinical observations of AHO kindreds in which several affected members had only AHO (i.e., pseudoPHP), whereas others had PTH resistance as well (i.e., PHP) first suggested that the two disorders might reflect variability in expression of a single genetic lesion. Further support for this hypothesis derives from biochemical studies which indicate that patients with pseudoPHP who are related to patients with PHP type 1a have equivalent functional Gsa deficiency (see Fig. 60-7)127 and identical gene defects (see later). Therefore, it seems reasonable to use the term Albright hereditary osteodystrophy to simplify description of this syndrome and to acknowledge the common clinical and biochemical characteristics that patients with PHP type 1a and pseudoPHP share. The inheritance of Gsa deficiency in patients with AHO127,128 first led to the speculation that the primary defect in this disorder involves the Gsa gene. The primary structure of the human Gsa protein has been deduced from characterization of complementary129 and genomic130 DNA clones. Gs a is encoded by exons 1 to 13 of the GNAS1 gene (Fig. 60-10),131,132,133,134 and 135 a complex gene that has been mapped to chromosome 20q13.2 q13.3 in humans.136 The observation that patients with AHO have reduced137,138 or normal137 levels of Gsa mRNA has suggested that Gsa deficiency might arise from a variety of genetic mutations. Distinct heterozygous mutations in the GNAS1 gene, including missense mutations,131,133,134,139,140 and 141 point mutations in sequences required for efficient splicing,132 insertions,142 and small deletions132,133,134 and 135,143,144,145,146 and 147 have been found in most kindreds studied (see Fig. 60-10); these findings imply that new mutations sustain this disorder in the population. A four-base-pair deletion in exon 7, which has been identified in several unrelated kindreds, appears to be a mutational hot spot.135,146,147,148 and 149 Most patients with AHO have genetic defects that impair the synthesis of Gsa protein and, therefore, have Gsa deficiency. In other patients, mutations in the GNAS1 gene lead to synthesis of dysfunctional proteins. These studies provide a molecular basis for Gsa deficiency and confirm that transmission of GNAS1 gene defects accounts for the autosomal dominant inheritance of AHO (see Fig. 60-10). The delineation of these defects does not explain the often striking variability in expression of the biochemical or clinical phenotype, however. For example, despite a generalized Gsa deficiency, only some Gsa-coupled pathways show reduced hormone responsiveness (e.g., to PTH, TSH, gonadotropins), whereas other pathways are apparently unaffected (those for ACTH in the adrenal and vasopressin in the renal

medulla) (see Fig. 60-8).

FIGURE 60-10. Mutations in the Gsa gene in Albright hereditary osteodystrophy. The upper diagram depicts the human Gsa gene, which spans over 20-kilobase pairs and contains 13 exons and 12 introns. Deletions (denoted by D), missense (denoted by asterisk), nonsense, and splice-site mutations are noted in the appropriate location in the GNAS1 gene. The lower panel depicts the amino acid changes in the protein as a result of the missense mutations in the gene. (bp, base pairs.)

One possible interpretation of variable hormonal responsiveness is that haploinsufficiency of Gsa is tissue specific; that is, in some tissues a 50% reduction in Gsa is still sufficient to facilitate normal signal transduction. However, this explanation leaves unanswered the even more intriguing question of why some subjects with Gsa deficiency have hormone resistance (PHP type 1a), whereas others lack hormone resistance (pseudoPHP). Analysis of published pedigrees has indicated that in most cases maternal transmission of Gsa deficiency leads to PHP type 1a, whereas paternal transmission of the defect leads to pseudoPHP,150,151 and 152 findings that have implicated genomic imprinting of the GNAS1 gene as a possible regulatory mechanism.151 Studies have indeed confirmed that the GNAS1 gene is imprinted, but in a far more complex manner than had been anticipated. Two upstream promoters, each associated with a large coding exon, lie 35 kilobases (kb) upstream of GNAS1 exon 1. These promoters are only 11 kb apart, yet show opposite patterns of allele-specific methylation and monoallelic transcription. The more 5' of these exons encodes NESP55, which is expressed exclusively from the maternal allele. By contrast, the XLas exon is paternally expressed.153,154 Despite the simultaneous imprinting in both the paternal and maternal directions of the GNAS1 gene, expression of Gsa appears to be biallelic in all human tissues examined.153,154 and 155 The lack of access to relevant tissues in patients with PHP type 1a has hindered studies of Gsa expression and stimulated attempts to develop suitable animal models. Two groups have succeeded in developing mice in which one gnas gene is disrupted, thereby generating murine models of PHP type 1a.156,157 Although these mice have reduced levels of Gsa protein, they lack many of the features of the human disorder. Biochemical analyses of these heterozygous gnas knock-out mice indicate that Gsa expression is similarly reduced in most tissues, regardless of whether maternal or paternal transmission of the defective allele has occurred. In some tissues (e.g., renal cortex), however, Gsa expression is less in mice with maternal inheritance of the defective allele than in mice with paternal inheritance of the defective allele. Accordingly, mice that inherit the defective gnas gene maternally express too little Gsa protein in the renal proximal tubular cells to facilitate normal PTH stimulation of adenylate cyclase. By contrast, the 50% reduction in Gsa expression that occurs in other tissues may account for more variable and moderate hormone resistance in these sites (e.g., the thyroid). Further studies are necessary to confirm that tissue-specific imprinting is the basis for these differences in Gsa expression. OTHER G PROTEIN DISEASES G protein defects are responsible for a number of other human diseases, most of which are rare endocrine disorders158,159 (Table 60-4). The McCune-Albright syndrome is characterized by the clinical triad of autonomous hyperfunction of one or more endocrine glands, irregularly bordered caf-au-lait lesions, and fibrous dysplasia. The variable expression of these features in different tissues results from the mosaic distribution of cells containing a postzygotic somatic mutation in the GNAS1 gene that activates Gsa protein. These mutations replace arginine201, the site of modification by cholera toxin, by either cysteine or histidine and thereby impair the ability of Gsa to hydrolyze GTP. As occurs in cholera, the modified Gsa protein has enhanced activity160,161 and is able to activate adenylate cyclase in the absence of hormonal stimulation. In McCune-Albright syndrome, the presence of the activated Gsa protein in tissues in which cAMP is trophic leads to cellular proliferation and autonomous hyperfunction (see Fig. 60-8) and can result in hyperthyroidism, precocious puberty, and growth hormone excess.159 The characterization of a unique missense mutation in the GNAS1 gene in two unrelated males with AHO and precocious puberty provides an unusual mechanism by which one mutation may cause both hormone resistance and constitutive signaling.134 An Arg366 to Ser missense mutation in the G5 region of the protein, which is important for guanyl nucleotide exchange, results in a temperature-sensitive form of Gsa that is rapidly degraded at 37C, thereby leading to hormone resistance in many target tissues. However, the abnormal Gsa chain is stable at 32C, and therefore accumulates in the testes, where the Arg366 to Ser substitution leads to constitutive activation of adenylate cyclase and precocious puberty.

TABLE 60-4. G Protein Mutations in Human Disease

Activating mutations in Gsa have been identified in up to 40% of sporadic somatotropic tumors in acromegaly. As in the McCune-Albright syndrome, mutations in Gsa arginine201 and glutamine227 impair the ability of Gsa to hydrolyze GTP and result in persistent stimulation of adenylate cyclase.162 This constitutively active mutant Gsa has been termed the gsp oncogene; it mimics the intracellular signaling triggered by GHRH and stimulates release of growth hormone and proliferation of somatotropes. A few autonomously hyperfunctioning thyroid adenomas harbor similar activating Gsa mutations,163 and gain of function mutations in Gi2 have been identified in adrenal cortical adenomas and the endocrine tumors of the ovary.164 PSEUDOHYPOPARATHYROIDISM TYPE 1B Some subjects with PHP type 1 lack features of AHO. These patients typically show hormone resistance that is limited to PTH target organs (see Fig. 60-8) and have normal Gsa activity.123,124 This variant is termed PHP type 1b.165 Although patients with PHP type 1b fail to show a nephrogenous cAMP response to PTH, they often manifest skeletal lesions similar to those that occur in patients with hyperparathyroidism. These observations have suggested that at least one intracellular signaling pathway coupled to the PTH receptor may be intact in patients with PHP type 1b (see Fig. 60-6). The molecular basis for PTH resistance in PHP type 1b has not been clearly defined. Specific resistance to PTH and normal activity of Gsa in available cells have implicated decreased expression or function of the PTH/PTHrP receptor as the cause of the hormone resistance. Despite extensive analysis of the complementary DNA,166,167 coding exons and exon/intron boundaries,168 and the promoter regions of the PTH/PTHrP receptor gene,169,170 no mutations have been identified in PHP type 1b patients, and linkage analysis studies have excluded unsequenced regions of the gene.170a Additional evidence against the notion that the defects in the PTH/PTHrP receptor cause PHP type 1b come from studies of mice171 and humans71,72 who are heterozygous for inactivation of the gene encoding the PTH/PTHrP receptor. Remarkably, loss of one PTH/PTHrP receptor gene does not cause PTH resistance or hypocalcemia. Finally, inheritance of two defective type PTH/PTHrP

receptor genes results in Blomstrand chondrodysplasia, a lethal genetic disorder characterized by advanced endochondral bone maturation.71,72 Thus, the molecular defect in PHP type 1b is likely to reside in another gene or genes that regulate expression or activity of the PTH/PTHrP receptor. A linkage analysis of four unrelated PHP type 1b kindreds established linkage to 20q13.3, the region that includes the Gsa gene (GNAS1).172 The chromosomal region most tightly linked to the phenotype in these kindreds is not GNAS1 itself but an area centromeric to the Gsa gene. Although these data are consistent with the existence of a second gene involved in mineral ion homeostasis in close proximity to GNAS1, defects in GNAS1 or the promoter region that might limit G sa deficiency to the kidney cannot be excluded. PSEUDOHYPOPARATHYROIDISM TYPE 1C In a few patients with PHP type 1, resistance to multiple hormones occurs in the absence of a demonstrable defect in Gs or Gi120 (see Fig. 60-8). The nature of the lesion in such patients is unclear, but it could be related to some other general component of the receptor-adenylate cyclase system, such as the catalytic unit.173 Alternatively, these patients could have functional defects of Gs (or Gi) that do not become apparent in the assays presently available. GENETICS Genetic studies of AHO and other forms of PTH resistance have been hampered by incomplete clinical descriptions of affected patients and inadequate characterization of their biochemical defects. The inheritance of AHO has been controversial. X-linked, autosomal dominant, and autosomal recessive inheritance of AHO has been proposed128; however, the description of father-to-son transmission of AHO with Gsa deficiency provided strong evidence against an X-linked mode of inheritance.174 The identification of defects in the GNAS1 gene in subjects with AHO has provided compelling evidence for autosomal dominant inheritance of AHO, including PHP type 1a and pseudoPHP. A striking feature of AHO is the occurrence of individuals with PHP type 1a and pseudoPHP in the same family who have identical mutations in the GNAS1 gene and show similar reductions in function or expression of Gsa. Perhaps even more remarkable is the observation that pseudoPHP and PHP type 1a do not occur in the same generation. Furthermore, the phenotypic expression of Gsa deficiency appears to become more severe with each successive generation.175 This pattern had originally been attributed to genomic imprinting,150,151 as reviews of published reports of AHO kindreds had indicated that maternal transmission of Gsa deficiency led to PHP type 1a, whereas paternal transmission led to pseudoPHP.150,151 Although the GNAS1 locus is within an imprinted region of the genome, several lines of evidence argue against a model of simple imprinting as the mechanism for variable phenotype (see earlier). First, analysis of cells from patients with PHP type 1a and pseudoPHP have shown that these cells contain equivalent amounts of Gsa protein.127 Second, the description of both maternal and paternal transmission of PHP type 1a176 in a single multiplex family with Gsa deficiency177 has suggested that the parental origin of the GNAS1 mutation may be unrelated to phenotype. Third, although transcripts encoding NESP55 and XLas are derived from only the maternal or the paternal allele, respectively (see earlier), Gsa transcripts are actively transcribed from both GNAS1 genes in all fetal177 and adult human tissues analyzed,153,154 indicating that both Gsa alleles are expressed in a wide variety of tissues. This suggests that either tissue-specific imprinting or other mechanisms may be responsible for the variable expression of GNAS1 gene defects observed within members of the same kindred. The inheritance of other forms of PHP is less well characterized. Inheritance of PHP type 1b is most consistent with an autosomal dominant pattern.172 PSEUDOHYPOPARATHYROIDISM TYPE 2 A few patients have been described in whom renal resistance to PTH is manifested by a reduced phosphaturic response to exogenous PTH, despite a normal increase in urinary cAMP excretion.178 These observations suggest that the PTH receptor-adenylate cyclase complex is functionally normal but that PTH resistance arises from an inability of intracellular cAMP to initiate the chain of metabolic events that results in the ultimate expression of PTH action. Although no supportive data are yet available, a defective cAMP-dependent protein kinase A has been postulated.178 Pseudohypoparathyroidism type 2 is a clinically heterogeneous disorder without a clear genetic or familial basis. Fewer than 25 patients with the disorder have been reported, and biochemical characterization in many cases has been incomplete. The frequent observation that normalization of the serum calcium concentration can correct the defective phosphaturic response to PTH has suggested that the metabolic defect in this disorder may be a failure of PTH to activate a calcium-dependent second messenger system. In addition, apparent PTH resistance in some of these cases could be the result of severe hypocalcemia caused by an unrecognized acquired deficiency of vitamin D or its metabolites.179 PSEUDOHYPOPARATHYROIDISM WITH SKELETAL RESPONSIVENESS Although, generally, the assumption has been that PTH resistance extends to the skeleton, this remains unproved. Evidence that bone cells are unresponsive to PTH in patients with PHP has been inferred from the attendant hypocalcemia in the disorder that is not corrected with the administration of PTH. However, bone demineralization occurs frequently in patients with PHP types 1a and 1b,180 although clinically significant hyperparathyroid bone disease with radiologic or histologic features of severe osteitis fibrosa cystica is rare. Whether these latter patients represent the extreme end of a spectrum of bone demineralization or a qualitatively different disease mechanism is unclear. In PHP, a deficiency of 1,25(OH)2D may underlie the presence of the hypocalcemia and the apparent PTH unresponsiveness of the bone homeostatic mechanism in PHP; this conflicts with the premise that bone cells are intrinsically resistant to the actions of PTH. BIOINEFFECTIVE PARATHYROID HORMONE The secretion of an abnormal, biologically inactive PTH molecule is a theoretical cause of functional hypoparathyroidism. Individuals with such a disorder might be expected to have elevated levels of circulating immunoreactive PTH and should respond normally to exogenous PTH, thus distinguishing them from patients with PHP. To date, a few cases have been reported in which the biochemical data were supportive of such a mechanism. In one previously reported case of bioineffective PTH, repeat study revealed clearly reduced or absent plasma levels of PTH in radioimmunoassays that were midmolecule specific or carboxy-terminal specific, despite symptomatic hypocalcemia.55 A second form of bioineffective PTH has been proposed as a mechanism to explain resistance to PTH in some patients with PHP.181,182 and 183 In these patients an inhibitor is presumed to be secreted by the parathyroid gland with the characteristics of a PTH antagonist. Thus, subjects show elevated circulating levels of PTH by radioimmunoassay, absent or reduced levels of PTH by a sensitive cytochemical bioassay, and a defective urinary cAMP response to exogenous PTH. The nature of this putative inhibitor remains elusive. Whether such an inhibitor is an abnormal form of PTH or even a product of the parathyroid gland is not clear from available data. Conceivably this inhibitor may provide the pathophysiologic basis for PTH resistance for some patients with PHP; however, the possibility that a circulating PTH inhibitor arises because of the primary biochemical defect is equally likely.

OTHER CAUSES OF HYPOCALCEMIA

HYPERPHOSPHATEMIA Hypocalcemia may also occur because of acute or chronic hyperphosphatemia. Plasma phosphate, when elevated, complexes with calcium to cause subsequent extraskeletal calcifications; if chronically maintained, this elevation depresses renal 1a-hydroxylase activity and leads to reduced production of 1,25(OH)2D. Hyperphosphatemia may result from iatrogenic administration of phosphate (either parenterally, orally, or rectally) or release from cells (e.g., after rapid tumor lysis during chemotherapy for Burkitt lymphoma or acute lymphoblastic leukemia). RHABDOMYOLYSIS Hypocalcemia also occurs in rhabdomyolysis and in acute severe illnesses (pancreatitis, sepsis, burns).103a,184 In these disorders, the cause of hypocalcemia is unclear, but it may be related to cell lysis with hyperphosphatemia or to movement of calcium into soft tissues with an inadequate acute response of the PTHvitamin D homeostatic system. CANCER Hypocalcemia frequently occurs in patients with cancer, but it usually represents a reduction in total, rather than ionized, calcium because of the effect of

hypoproteinemia. Conversely, true reductions in ionized calcium occur in patients with cancer, particularly of breast and prostate, who have osteoblastic metastases.185,186 Hypocalcemia in these patients has been attributed to increased calcium flux into the osteoblastic lesions. Hypocalcemia of unknown cause is commonly associated with acute illness and carries a poor prognosis. Frequently, these patients are experiencing stress related to sepsis.187 DISORDERS OF VITAMIN D METABOLISM Abnormalities in vitamin D metabolism cause reduced intestinal calcium absorption; moreover, PTH-induced mobilization of calcium from bone to blood is impaired. Both of these factors may produce hypocalcemia if compensatory actions of PTH or 1,25(OH)2D are inadequate. ACQUIRED VITAMIN D DEFICIENCIES Deficiencies of vitamin D or its active metabolite, 1,25(OH)2D, are accompanied by hypocalcemia and secondary (adaptive) hyperparathyroidism. Deficiency of vitamin D may arise because of inadequate sunlight exposure, dietary deficiency, or malabsorption. A specific deficiency of 25(OH)D may develop in some patients with severe hepatobiliary disease. The principal example of an acquired deficiency of 1,25(OH)2D is chronic renal insufficiency (see Chap. 61). Several drugs, most notably corticosteroids, barbiturates, and phenytoin, can interfere with vitamin D metabolism or action, and occasionally they can cause clinical hypocalcemia and secondary hyperparathyroidism. HEREDITARY DISORDERS OF VITAMIN D METABOLISM Hypocalcemia and secondary hyperparathyroidism can be biochemical manifestations of hereditary disorders of vitamin D function. Deficient 1,25(OH)2D formation and action are causes of hypocalcemia and osteomalacia in patients who have either 1a-hydroxylase deficiency (vitamin Ddependent rickets type I)188 or hereditary resistance to 1,25(OH)2D (vitamin Ddependent rickets type II), respectively189 (see Chap. 63).

DIFFERENTIAL DIAGNOSIS OF HYPOCALCEMIA

The history and physical examination are quite important in the evaluation of hypocalcemia. The goal of the bedside examination is to characterize the severity of symptoms and the duration of hypocalcemia, to detect underlying medical conditions, and to search for specific features of the various states of hypoparathyroidism and other disorders of mineral metabolism. Often, the clinical evaluation provides a tentative diagnosis that can be confirmed by appropriate laboratory tests. An understanding of the causes of hypocalcemia (see Table 60-2) provides the basis for the clinical evaluation. If the patient is asymptomatic and the hypocalcemia has been detected by blood tests performed for other reasons, one should first estimate the likelihood of true ionized hypocalcemia by examining the concentrations of serum proteins. Occasionally, the measurement of ionized calcium by a reliable laboratory is helpful. Acute changes in the distribution of calcium between ionized and bound fractions may be indicated by a history of hyperventilation (respiratory alkalosis), recent massive blood transfusion (citrate), trauma with rhabdomyolysis, tumor lysis after chemotherapy, or phosphate administration by enema or intravenous infusion. Transient hypoparathyroidism can often be anticipated (in neonates and in patients with parathyroid suppression after treatment of hypercalcemia); therefore, it is usually not a diagnostic problem. Reversible hypoparathyroidism associated with hypomagnesemia may be suggested by a history of alcoholism, gastrointestinal disease, or treatment with drugs that cause renal magnesium wasting. Chronic genetic or acquired hypoparathyroidism often is associated with characteristic features, such as family history, abnormal habitus, other endocrine disturbances, or a history of damage to the parathyroid glands. Steatorrhea, liver disease, or renal insufficiency may suggest a deficiency of vitamin D metabolites (see Chap. 63). One occasionally may be confronted with a patient in tetany for whom no previous medical history is available. Even if urgent therapy is deemed necessary, a blood sample should be obtained quickly, before the administration of calcium, for the determination of levels of calcium, phosphorus, blood urea nitrogen or creatinine, albumin, electrolytes, and magnesium, and for PTH radioimmunoassay. The results of these determinations can direct the subsequent evaluation. In the absence of renal insufficiency, and if no history of an acute phosphate load is present, hyperphosphatemia is often an indication of deficient PTH secretion or abnormal PTH action. If the patient's nutritional status is normal, then hypophosphatemia may be suspected of being associated with secondary hyperparathyroidism in states of abnormal vitamin D action. SERUM PARATHYROID HORMONE LEVELS Specific endocrine tests often are helpful in the evaluation. Because normally hypocalcemia is a potent stimulus for the secretion of PTH, patients with intact parathyroid gland function should be expected to have high circulating levels of PTH during hypocalcemic episodes. Low levels of serum PTH, and, with some assays, even inappropriately normal levels despite hypocalcemia, should prompt consideration of the various forms of hypoparathyroidism (see Table 60-2). Given the clinical history, duration of hypocalcemia, and physical findings, the cause of deficient parathyroid gland function can usually be determined. Because a number of different PTH assays are available (see Chap. 51, Chap. 58 and Chap. 61), correct interpretation depends on familiarity with the results to be expected in the particular assay. Caution should be exercised in the interpretation of the PTH assay if the patient has renal insufficiency. If PTH levels are appropriately elevated in a hypocalcemic patient, then pseudohypoparathyroidism, disorders of vitamin D metabolism or action, and a variety of other conditions (see Table 60-2) need to be considered. PARATHYROID HORMONE INFUSION Pseudohypoparathyroidism can sometimes be suggested by the characteristic habitus or the finding of hyperphosphatemia. The confirmation of the diagnosis depends on the demonstration of a deficient target organ response to PTH. This is usually accomplished by demonstrating a lack of the normal brisk increase in urinary excretion of cAMP and phosphate after an infusion of PTH.190 Deficient activity of the adenylate cyclase coupling protein (Gs) in cell membranes from those patients with pseudohypoparathyroidism type 1a can also be demonstrated with specialized techniques available in some centers. Under some circumstances, analyzing the response of plasma cAMP to an intravenous infusion of synthetic human PTH (134) is more convenient, particularly in patients in whom collection of urine is not possible or practical. Although this testing protocol can differentiate patients with PHP type 1a from normal subjects and patients with pseudoPHP or hypoparathyroidism,191,192 normal changes in plasma cAMP are less dramatic than the changes in urinary cAMP. The observation of an apparently normal plasma cAMP response to PTH in a 3-month-old child who later developed clinical and biochemical evidence of PHP type 1a, including an abnormal plasma cAMP response to PTH, suggests that testing be delayed until after 6 months of age.193 MEASUREMENT OF SERUM VITAMIN D METABOLITES The measurement of vitamin D metabolites can be helpful in documenting abnormalities of vitamin D metabolism or action as causes of hypocalcemia. Low serum 25(OH)2D levels suggest nutritional deficiency, malabsorption, or liver disease. Low serum 1,25(OH)2D levels may be seen in patients who have a specific defect of renal 1a-hydroxylase and in patients with renal failure. Low serum levels of 1,25(OH)2D are also found in patients with hypoparathyroidism and pseudohypoparathyroidism, so results must be interpreted in the light of other tests of PTH secretion and action. Very high levels of 1,25(OH)2D in patients with hypocalcemia may indicate resistance to the action of calcitriol. Other causes of hypocalcemia with appropriate elevations of PTH are numerous (see Table 60-2) but are often acute in onset, with a definite precipitating event that can be determined by a careful history.

THERAPY FOR HYPOCALCEMIA

Before treating hypocalcemia, one should recall that hypoalbuminemia is the most common cause of low total serum calcium concentration; it often occurs in patients with malnutrition or other debilitating diseases. In such patients, nutritional support, rather than treatment of hypocalcemia, is appropriate. If hypoalbuminemia cannot account for the observed decrease in total calcium, or if the ionized calcium level is found to be low, an evaluation and treatment of the hypocalcemia are indicated.

Underlying disorders, such as hypomagnesemia or malabsorption, should be sought and corrected whenever possible. EMERGENCY THERAPY Urgent therapy with intravenous calcium is required for patients who have acute hypocalcemic crisis manifested by severe tetany, laryngospasm, or convulsions. In adults, 10 to 20 mL of 10% calcium gluconate (93 mg elemental calcium per 10 mL, 4.65 mEq/10 mL) may be infused over 10 minutes. For children, 2 mg of elemental calcium per kilogram of body weight (~0.2 mL/kg of 10% calcium gluconate) may be used. Patients who require immediate treatment for tetany and patients who have symptoms of hypocalcemia without tetany are often candidates for intravenous calcium infusion over a longer period. Elemental calcium (1015 mg/kg) may be infused over 6 to 8 hours while serum calcium levels are monitored to assess the adequacy of the administered dose. Calcium salts are extremely irritating; therefore, care should be taken to avoid extravasation of the dose, and dilution of the calcium in at least 50 to 100 mL of electrolyte solution is recommended. Certain anions, such as bicarbonate and phosphate, cause the precipitation of calcium. LONG-TERM THERAPY Maintenance therapy for hypoparathyroidism consists of supplemental administration of oral calcium or vitamin D, or both. Because patients with hypoparathyroidism lack the renal effects of PTH, they usually have hypercalciuria at serum calcium concentrations within the normal range.194 Therefore, the serum calcium concentration should be maintained in the range of 8 to 9 mg/dL. This is sufficiently high to prevent most symptoms of hypocalcemia and to retard the progression of cataracts, but low enough that hypercalciuria is uncommon. Nevertheless, episodes of hypercalciuria and hypercalcemia can occur even in patients who have taken stable doses of calcium and vitamin D for years. By contrast, hypercalciuria is unusual in patients with pseudohypoparathyroidism. TREATMENT WITH CALCIUM SALTS A relatively high dose of a calcium salt (15002500 mg elemental calcium as a daily supplement) is recommended to provide a level of ingested calcium that is adequate to minimize fluctuations caused by variable dietary calcium intake. Patients can also be educated to consume a consistent quantity of calcium from dietary sources each day. Calcium content differs considerably among the available oral preparations195 (see Chap. 236). Because therapy for hypoparathyroidism must be individualized to obtain a desired serum calcium level, few absolute guidelines exist for choosing one calcium salt over another. However, the patient may have a preference, and compliance can be maximized by suggesting the most tolerable dosage form. Calcium carbonate and calcium phosphate are poorly absorbed by patients with achlorhydria; optimal absorption requires that these salts be taken with meals.196 THERAPY WITH VITAMIN D OR ITS METABOLITES In some patients with decreased parathyroid reserve, oral calcium supplementation may be sufficient to alleviate symptoms of hypocalcemia. In most patients, however, treatment with vitamin D or one of its metabolites is required. In patients with some disorders of vitamin D metabolism, treating with a precursor metabolite is reasonable, so that the ability of PTH to regulate conversion of precursors to 1,25(OH)2D can be preserved as a defense against drug toxicity. In hypoparathyroidism, PTH action is absent and regulated conversion of precursors to 1,25(OH)2D does not occur. Therefore, the choice of vitamin D or another metabolite may be based on other considerations. Vitamin D2 (ergocalciferol) is effective in treating patients with all forms of hypoparathyroidism and is the least expensive of the available preparations. The dosage required for the treatment of hypoparathyroidism is usually 50,000 to 100,000 U per day (1.252.5 mg per day), but dosage must be individualized. During treatment with vitamin D2, serum levels of 25(OH)D can be measured, if necessary, to assess compliance and drug absorption.197 Vitamin D2 is lipophilic, and a period of several weeks is required for its full effect to be established on beginning therapy. Likewise, if it must be discontinued, a prolonged period will be required for its effect to decrease. More active metabolites, such as calcitriol and 1a-OH-cholecalciferol (available outside the United States), have a more rapid onset of action, and a shorter period of time is required for their effects to decrease after discontinuation. However, these metabolites are more expensive, and frequent monitoring is necessary to avoid toxicity. Dihydrotachysterol is also active in patients who lack PTH action, and its moderate (1- to 4-week) duration of action and low cost offer a satisfactory compromise to vitamin D2 and calcitriol. A disadvantage is that plasma levels of dihydrotachysterol cannot be monitored. The assessment of compliance may be more difficult when the patient is receiving a shortacting drug. VITAMIN D TOXICITY Drug toxicity is a major concern whenever vitamin D or one of its metabolites is prescribed. For all of the agents, the therapeutic dose is close to the dose at which toxicity occurs; hence, therapy must be individualized and closely monitored. Frequent testing should include a determination of the serum calcium concentration, the urinary calcium excretion, and the serum creatinine level. Early toxicity can be recognized as asymptomatic hypercalciuria; more severe vitamin D intoxication is associated with asymptomatic or symptomatic hypercalcemia. Renal function can deteriorate during episodes of toxicity, and repeated episodes of vitamin D toxicity can lead to progressive renal insufficiency. If hypercalcemia occurs, both vitamin D and supplemental calcium should be discontinued and the patient should be treated, if necessary, for vitamin D intoxication (see Chap. 59). Those patients taking short-acting vitamin D metabolites can be rapidly switched to another dose form; patients taking vitamin D2 may need to limit calcium intake for a prolonged period, until body stores of vitamin D decrease. INFLUENCE OF OTHER DRUGS A number of commonly used drugs have important effects in patients who are treated with vitamin D and calcium. Thiazide diuretics enhance renal calcium reabsorption and may precipitate hypercalcemia in a patient taking vitamin D and calcium.198 Treatment of patients who have mild hypoparathyroidism with sodium restriction and a long-acting thiazide diuretic has been reported to increase total and ionized serum calcium levels, even in patients not treated with vitamin D supplements199; however, the effect of thiazides to decrease renal calcium excretion is modest compared with the effect of vitamin D to increase intestinal calcium absorption.200 Patients with hypoparathyroidism are markedly sensitive to the calciuric effects of loop diuretics and may become hypocalcemic after treatment with such drugs.201 Glucocorticosteroids antagonize the actions of vitamin D metabolites. Anticonvulsants and other inducers of hepatic microsomal oxidases may increase the clearance of vitamin D metabolites. Treatment with phosphate-binding antacids to decrease the level of serum phosphate is rarely necessary in patients with hypoparathyroidism, because renal phosphate clearance increases as serum calcium rises.202 Parathyroid allotransplantation has been attempted in some patients who have complicated hypoparathyroidism, with occasional success.203 Long-term immunosuppression is required to maintain a functional graft, however, so that most patients with hypoparathyroidism are not candidates for such therapy. THERAPY DURING PREGNANCY Neonatal hyperparathyroidism has been reported in the infants of hypocalcemic mothers; therefore, women with hypocalcemia should be appropriately treated during pregnancy. Because the mineral demands of the fetus change significantly during the course of pregnancy, patients should be monitored closely and treatment with a short-acting metabolite of vitamin D should be begun, so that dosage adjustments are immediately effective. CHAPTER REFERENCES
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. Moore EW. Ionized calcium in normal serum, ultrafiltrates, and whole blood determined by ion-exchange electrodes. J Clin Invest 1970; 49:318. Ladenson JH, Lewis JW, Boyd JC. Failure of total calcium corrected for protein, albumin, and pH to correctly assess free calcium status. J Clin Endocrinol Metab 1978; 46:986. Berry EM, Gupta MM, Turner SJ, Burns RR. Variations in plasma calcium with induced changes in plasma specific gravity, total protein, and albumin. BMJ 1973; 4:640. Hoffman E. The Chvostek sign: a clinical study. Am J Surg 1958; 96:33. Simpson JA. The neurologic manifestations of idiopathic hypoparathyroidism. Brain 1952; 75:76. Wong CK, Lau CP, Cheng CH, et al. Hypocalcemic myocardial dysfunction: short and long-term improvement with calcium replacement. Am Heart J 1990; 120:381. Levine SN, Rheams CN. Hypocalcemic heart failure. Am J Med 1975; 78:1022. Rimailho A, Bouchard P, Schaison G, et al. Improvement of hypocalcemic cardiomyopathy by correction of serum calcium level. Am Heart J 1985; 109:611. Kanda F, Jinnai J, Fujita T. Somatosensory evoked potentials in patients with hypocalcaemia after parathyroidectomy. J Neurol 1988; 235:136. Nikiforuk G, Fraser D. Chemical determinants of enamel hypoplasia in children with disorders of calcium and phosphate homeostasis. J Dent Res 1979; 58(B):1014. Illum F, Dupont E. Prevalence of CT-detected calcification in the basal ganglia in idiopathic hypoparathyroidism and pseudohypoparathyroidism. Neuroradiology 1985; 27:32. Schaaf M, Payne C. Dystonic reactions to prochlorperazine in hypoparathyroidism. N Engl J Med 1966; 275:991. Drezner ML, Neelan FA, Haussler M, et al. 1,25-Dihydroxycholecalciferol deficiency: the probable cause of hypocalcemia and metabolic bone disease in pseudohypoparathyroidism. J Clin Endocrinol Metab 1976; 42:621. Schutt-Aine JC, Young MA, Pescovitz DH, et al. Hypoparathyroidism: a possible cause of rickets. J Pediatr 1985; 106:255. Abugassa S, Nordenstrom J, Eriksson S, Sjoden G. Bone mineral density in patients with chronic hypoparathyroidism. J Clin Endocrinol Metab 1993; 76:1617. Shukla S, Gillespy T III, Thomas WC Jr. The effect of hypoparathyroidism on the aging skeleton. J Am Geriatr Soc 1990; 38:884. Orr-Walker B, Harris R, Holdaway IM, et al. High peripheral and axial bone densities in a postmenopausal woman with untreated hypoparathyroidism. Postgrad Med J 1990; 66:1061.

18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48.

Wade JSH, Fourman P, Deane L. Recovery of parathyroid function in patients with transient hypoparathyroidism after thyroidectomy. Br J Surg 1965; 52:493. Dent CE. Some problems of hyperparathyroidism. BMJ 1962; 2:1419. Edis AJ. Prevention and management of complications associated with thyroid and parathyroid surgery. Surg Clin North Am 1979; 59:83. Lor JM Jr. Complications in management of thyroid cancer. Semin Surg Oncol 1991; 7:120. Gann DS, Paone JF. Delayed hypocalcemia after thyroidectomy for Graves' disease is prevented by parathyroid autotransplantation. Ann Surg 1979; 190:508. Michie W, Stowers JM, Duncan T, et al. Mechanism of hypocalcemia after thyroidectomy for thyrotoxicosis. Lancet 1971; 1:508. Watson CG, Steed DL, Robinson AG, Deftos LJ. The role of calcitonin and parathyroid hormone in the pathogenesis of postthyroidectomy hypocalcemia. Metabolism 1981; 30:588. Percival RC, Hargreaves AW, Kanis JA. The mechanism of hypocalcemia following thyroidectomy. Acta Endocrinol (Copenh) 1985; 109:220. Neufield RB, Maclaren N, Blizzard R. Two types of autoimmune Addison's disease associated with different polyglandular autoimmune syndromes. Medicine (Baltimore) 1981; 60:355. Barakat AY, D'Albora JB, Martin MM, Jose PA. Familial nephrosis, nerve deafness, and hypoparathyroidism. J Pediatr 1977; 91:61. Benson PF, Parsons V. Hereditary hypoparathyroidism presenting with oedema in the neonatal period. QJM 1964; 130:197. Dahlberg PJ, Borer WZ, Newcomer KL, Yutac WR. Autosomal or X-linked recessive syndrome of congenital lymphedema, hypoparathyroidism, nephropathy, prolapsing mitral valve, and brachytelephalangy. Am J Med Genet 1983; 16:99. Raatikka M, Ropola J, Tuteri L, et al. Familial third and fourth pharyngeal pouch syndrome with truncus arteriosus: DiGeorge syndrome. Pediatrics 1981; 67:173. Rohn RD, Leffell MS, Leadem P, et al. Familial third-fourth pharyngeal pouch syndrome with apparent autosomal dominant transmission. J Pediatr 1984; 105:47. de la Chapelle A, Herra R, Kiovisto M, Aula P. A deletion in chromosome 22 can cause DiGeorge syndrome. Hum Genet 1981; 57:253. Kelley RI, Zackai FH, Emanuel BS, et al. The association of the DiGeorge anomalad with partial monosomy of chromosome 22. J Pediatr 1982; 101:197. Driscoll DA, Budarf ML, Emanuel BS. A genetic etiology for DiGeorge syndrome: consistent deletions and microdeletions of 22q11. Am J Hum Genet 1991; 50:924. Monaco G, Pignata C, Rossi E, et al. DiGeorge anomaly associated with 10p deletion. Am J Med Genet 1991; 19:215. Lai MMR, Scriven PN, Ball C, Berry AC. Simultaneous partial monosomy 10p and trisomy 5q in a case of hypoparathyroidism. J Med Genet 1992; 29:586. Spinner MW, Blizzard RM, Childs B. Clinical and genetic heterogeneity in idiopathic Addison's disease and hypoparathyroidism. J Clin Endocrinol Metab 1968; 28:795. Winter WE, Silverstein JH, Barrett DJ, Kiel E. Familial DiGeorge syndrome with tetralogy of Fallot and prolonged survival. Eur J Pediatr 1984; 141:171 Shaw NJ, Haigh D, Lealmann GT, et al. Autosomal recessive hypoparathyroidism with renal insufficiency and developmental delay. Arch Dis Child 1991; 66:1191. Sanjad SA, Sakati NA, Abu-Osba YK, et al. A new syndrome of congenital hypoparathyroidism, severe growth failure, and dysmorphic features. Arch Dis Child 1991; 66:193. Billous RW, Murty G, Parkinson DB, et al. Brief report: autosomal dominant familial hypoparathyroidism, sensorineural deafness, and renal dysplasia. N Engl J Med 1992; 327:1069. Richardson RJ, Kirk JM. Short stature, mental retardation, and hypoparathyroidism: a new syndrome. Arch Dis Child 1990; 65:1113. Ahonen P, Myllarniemi S, Sipila I, Perheentupa J. Clinical variation of autoimmune polyendocrinopathycandidiasisectodermal dystrophy (APECED) in a series of 68 patients. N Engl J Med 1990; 322:1829. Blizzard RM, Chee D, Davis W. The incidence of parathyroid and other antibodies in the sera of patients with idiopathic hypoparathyroidism. Clin Exp Immunol 1966; 1:119. Verghese MW, Ward FE, Eisenbarth GS. Lymphocyte suppressor activity in patients with polyglandular failure. Hum Immunol 1981; 3:173. Jackson RA, Haynes BF, Burch WM, et al. 1a+ T cells in new onset Graves' disease. J Clin Endocrinol Metab 1984; 59:187. Nagamine K, Peterson P, Scott HS, et al. Positional cloning of the APECED gene. Nat Genet 1997; 17:393. The Finnish-German APECED Consortium. An autoimmune disease, caused by mutations in a novel gene featuring two PHD-type zinc-finger domains. Nat Genet 1997; 17:399.

48a. Ishii T, Suzuki Y, Ando N, et al. Novel mutations of the autoimmune regulator gene in two siblings with autoimmune polyendocrinopathy-candidiasis-ectodermal dystrophy. J Clin Endocrinol Metab 2000; 85(8):2922. 49. Wang CY, Davoodi-Semiromi A, Huang W, et al. Characterization of mutations in patients with autoimmune polyglandular syndrome type 1 (APS1). Hum Genet 1998; 103:681. 50. Ward L, Paquette J, Seidman E, et al. Severe autoimmune polyendocrinopathy candidiasis-ectodermal dystrophy in an adolescent girl with a novel AIRE mutation: response to immunosuppressive therapy. J Clin Endocrinol Metab 1999; 84:844. 51. Posillico JT, Wortsman J, Srikanta S, et al. Parathyroid cell surface auto-antibodies that inhibit parathyroid hormone secretion from dispersed human parathyroid cells. J Bone Miner Res 1986; 1:475. 52. Brandi ML, Aurbach GD, Fattorossi A, et al. Antibodies cytotoxic to bovine parathyroid cells in autoimmune hypoparathyroidism. Proc Natl Acad Sci U S A 1986; 83:8366. 53. Drake TG, Albright F, Bauer W. Chronic idiopathic hypoparathyroidism: report of 6 cases with autopsy findings in one. Ann Intern Med 1934; 12:1751. 54. Treusch JV. Idiopathic hypoparathyroidism: follow-up study including autopsy findings of a case previously reported. Ann Intern Med 1962; 56:484. 55. Ahn TG, Antonarakis SE, Kronenberg HM, et al. Familial isolated hypoparathyroidism: a molecular genetic analysis of 8 families with 23 affected persons. Medicine (Baltimore) 1986; 65:73. 56. Arnold A, Horst SA, Gardella TJ, et al. Mutation of the signal peptide encoding region of the preproparathyroid hormone gene in familial isolated hypoparathyroidism. J Clin Invest 1990; 86:1084. 57. Parkinson DB, Shaw NJ, Himsworth RL, Thakker RV. Parathyroid hormone gene analysis in autosomal hypoparathyroidism using an intragenic tetranucleotide (AAAT)n polymorphism. Hum Genet 1993; 91:281. 58. Thakker RV, Davies KE, Whyte MP, et al. Mapping the gene causing X-linked recessive idiopathic hypoparathyroidism to Xq26-Xq27 by linkage studies. J Clin Invest 1990; 86:40. 59. Trump D, Dixon PH, Mumm S, et al. Localization of X-linked recessive idiopathic hypoparathyroidism to a 1.5 Mb region on Xq26-q27. J Med Genet 1998; 35:905. 60. Thakker RV. Molecular basis of PTH underexpression. In: Bilezikian JP, Raisz LG, Rodan GA, eds. Principles of bone biology. San Diego: Academic Press, 1996:837. 61. Brown EM, Gamba G, Riccardi D, et al. Cloning and characterization of an extra-cellular Ca 2+-sensing receptor from bovine parathyroid. Nature 1993; 366:575. 62. Brown EM. Physiology and pathophysiology of the extracellular calcium-sensing receptor. Am J Med 1999; 106:238. 63. Arthur JM, Collinsworth GP, Gettys TW, et al. Specific coupling of a cation-sensing receptor to G protein alpha-subunits in MDCK cells. Am J Physiol 1997; 273: F129. 64. Pearce SH, Williamson C, Kifor O, et al. A familial syndrome of Hypocalcemia with hypercalciuria due to mutations in the calciumsensing receptor [see comments]. N Engl J Med 1996; 335:1115. 65. Finegold DN, Armitage MM, Galiani M, et al. Preliminary localization of a gene for autosomal dominant hypoparathyroidism to chromosome 3q13. Pediatr Res 1994; 36:414. 66. Pollak MR, Brown EM, Chou YW, et al. Mutations in the human Ca2+-sensing receptor gene cause familial hypocalciuric hypercalcemia and neonatal severe hyperparathyroidism. Cell 1993; 75:1297. 67. Pearce SH, Bai M, Quinn SJ, et al. Functional characterization of calcium-sensing receptor mutations expressed in human embryonic kidney cells. J Clin Invest 1996; 98:1860. 68. Pollak MR, Seidman CE, Brown EM. Three inherited disorders of calcium sensing. Medicine (Baltimore) 1996; 75:115. 69. Pollak MR, Brown EM, Estep HL, et al. Autosomal dominant hypocalcemia caused by a Ca 2+-sensing receptor gene mutation. Nat Genet 1994; 8:303. 70. Bai M, Quinn S, Trivedi S, et al. Expression and characterization of inactivating and activating mutations in the human Ca 2+o-sensing receptor. J Biol Chem 1996; 271:19537. 71. Zhang P, Jobert AS, Couvinea A, Silve C. A homozygous inactivating mutation in the parathyroid hormone/parathyroid hormone related peptide receptor causing Blomstrand chondrodysplasia. J Clin Endocrinol Metab 1998; 83:3365. 72. Jobert AS, Zhang P, Couvineau A, et al. Absence of functional receptors for parathyroid hormone and parathyroid hormone-related peptide in Blomstrand chondrodysplasia. J Clin Invest 1998; 102:34. 73. Abramowicz MJ, Duprez L, Parma J, et al. Familial congenital hypothyroidism due to inactivating mutation of the thyrotropin receptor causing profound hypoplasia of the thyroid gland. J Clin Invest 1997; 99:3018. 74. Sunthornthepvarakul T, Gottschalk ME, Hayashi Y, Refetoff, S. Brief report: resistance to thyrotropin caused by mutations in the thyrotropin-receptor gene [see comments]. N Engl J Med 1995; 332:155. 75. Kremer H, Kraaji R, Toledo SPA, et al. Male pseudohermaphroditism due to homozygous missense mutation of the luteinizing hormone receptor gene. Nat Genet 1995; 9:160. 76. Latronico AC, Anasti J, Arnhold IJP, et al. Testicular and ovarian resistance to luteinizing hormone caused by inactivating mutations of the luteinizing hormone receptor gene. N Engl J Med 1996; 507. 77. Wajnrajch MP, Gertner JM, Harbison MD, et al. Nonsense mutation in the human growth hormone-releasing hormone receptor causes growth failure analogous to the little (lit) mouse. Nat Genet 1996; 12:88. 78. Netchine I, Talon P, Dastot F, Vitaux F, Goossens M, Amselem S. Extensive phenotypic analysis of a family with growth hormone (GH) deficiency caused by a mutation in the GH-releasing hormone receptor gene. J Clin Endocrinol Metab 1998; 83:432. 79. Salvatori R, Hayashida CY, Aguiar-Oliveira MH, et al. Familial dwarfism due to a novel mutation of the growth hormone-releasing hormone receptor gene. J Clin Endocrinol Metab 1999; 84:917. 80. Maheshwari HG, Silverman BL, Dupuis J, Baumann G. Phenotype and genetic analysis of a syndrome caused by an inactivating mutation in the growth hormone-releasing hormone receptor: dwarfism of Sindh. J Clin Endocrinol Metab 1998; 83:4065. 81. Weber A, Toppari J, Harvey RD, et al. Adrenocorticotropin receptor gene mutations in familial glucocorticoid deficiency: relationships with clinical features in four families. J Clin Endocrinol Metab 1995; 80:65. 82. Tsigos C, Arai K, Hung W, Chrousos GP. Hereditary isolated glucocorticoid deficiency is associated with abnormalities of the adrenocorticotropin receptor gene. J Clin Invest 1993; 92:2458. 83. Rosenthal W, Antaramian A, Gilbert S, Birnbaumer M. Nephrogenic diabetes insipidus. A V2 vasopressin receptor unable to stimulate adenylyl cyclase. J Biol Chem 1993; 268:13030. 84. Aittomaki K, Lucena JLD, Pakarinen P, et al. Mutation in the follicle-stimulating hormone receptor gene causes hereditary hypergonadotrophic ovarian failure. Cell 1995; 82:959. 85. De Roux N, Young J, Misrahi M, et al. A family with hypogonadotrophic hypogonadism and mutations in the gonadotrophin releasing hormone receptor. N Engl J Med 1997; 337:1597. 86. Parma J, Duprez L, Van Sande J, et al. Somatic mutations in the thyrotropin receptor gene cause hyperfunctioning thyroid adenomas. Nature 1993; 365:649. 87. Refetoff S, Sunthornthepvarakul T, Gottschalk ME, Hayashi, Y. Resistance to thyrotropin and other abnormalities of the thyrotropin receptor. Recent Prog Horm Res 1996; 51:97. 88. Duprez L, Parma J, Van Sande J, et al. Germline mutations in the thyrotropin receptor gene cause non-autoimmune autosomal dominant hyperthyroidism. Nat Genet 1994; 7:396. 89. Schipani E, Parfitt AM, Jensen GS, et al. Constitutively activated receptors for parathyroid hormone and parathyroid hormone related peptide in Jansen's metaphyseal chondrodysplasia. N Engl J Med 1996; 335:708. 90. Shenker A, Laue L, Kosugi S, et al. A constitutively activating mutation of the luteinizing hormone receptor in familial male precocious puberty. Nature 1993; 365:652. 91. Yano K, Hidaka A, Saji M, et al. A sporadic case of male-limited precocious puberty has the same constitutively activating point mutation in the luteinizing hormone/choriogonadotropin receptor gene as familial cases. J Clin Endocrinol Metab 1994; 79:1818. 92. Baldini A. Is the genetic basis of DiGeorge syndrome in HAND? Nat Genet 1999; 21:246. 93. Yamagishi H, Garg V, Matsuoka R, et al. A molecular pathway revealing a genetic basis for human cardiac and craniofacial defects. Science 1999; 283:1158. 94. Pizzuti A, Novelli G, Rattie A, et al. UFD1L, a developmentally expressed ubiquitination gene, is deleted in CATCH 22 syndrome. Hum Mol Genet 1997; 6:259. 95. Watanabe, Mochizuki H, Kohda N, et al. Autosomal dominant familial hypoparathyroidism and sensorineural deafness without renal dysplasia. Eur J Endocrinol 1998; 139:631. 95a. Lichtner P, Konig R, Hasegawa T, et al. An HDR (hypoparathyroidism, deafness, renal dysplasia) syndrome locus maps distal to the DiGeorge syndrome region on 10p13/14. J Med Genet 2000; 37(1):33. 95b. Van Esch H, Groenen P, Nesbit MA, et al. GATA3 haplo-insufficiency causes human HDR syndrome. Nature 2000; 406(6794):419. 95c. Gunther T, Chen ZF, Kim J, et al. Genetic ablation of parathyroid glands reveals another source of parathyroid hormone. Nature 2000; 406(6792):199.

95d. Ding C, Buckingham B, Levine MA. Neonatal hypoparathyroidism attributable to homozygous partial deletion of the human glial cell missing gene-B. Proceedings of the annual meeting of the Endocrine Society Toronto, 2000:CA409 (Abstr). 95e. Garcia-Garcia E, Camacho-Alonso J, Gomez-Rodriguez MJ, et al. Transient congenital hypoparathyroidism and 22q11 deletion. J Pediatr Endocrinol Metab 2000; 13(6):659. 96. Baltrop D. Neonatal hypocalcemia. Postgrad Med 1975; 51(Suppl 3):18. 97. Laitinen K, Tahtela R, Valimanki M. The dose-dependency of alcohol-induced hypoparathyroidism, hypercalciuria, and hypermagnesuria. Bone Miner 1992; 19:75. 98. Attie MF, Fallon MD, Spar B, et al. Bone and parathyroid inhibitory effects of S-2(3-aminopropylamino) ethylphosphorothioic acid. J Clin Invest 1985; 75:1191.

99. Hirschel-Scholz S, Caverzasio J, Bonjour J-P. Inhibition of parathyroid hormone secretion and parathyroid hormoneindependent diminution of tubular calcium reabsorption by WR-2721, a unique hypocalcemic agent. J Clin Invest 1985; 76:1851. 100. Carpenter TO, Carnes DL Jr, Anast CS. Hypoparathyroidism in Wilson's disease. N Engl J Med 1983; 309:873. 101. Duran MJ, Borst GC, Osburne RC, Eil C. Concurrent renal hypomagnesemia and hypoparathyroidism with normal parathormone respon-siveness. Am J Med 1984; 76:151. 102. Anast CS, Mohs JM, Kaplan SL, Burns TW. Evidence for parathyroid failure in magnesium deficiency. Science 1972; 177:606. 103. Estep H, Shaw WA, Watlington C, et al. Hypocalcemia due to hypomagnesemia and reversible parathyroid hormone unresponsiveness. J Clin Endocrinol Metab 1969; 29:842. 103a. Klein GL, Langman CB, Herndon DN. Persistent hypoparathyroidism following magnesium repletion in burn-injured children. Pediatr Nephrol 2000; 14(4):301. 104. 105. 106. 107. 108. 109. 110. 111. 112. 113. 114. 115. 116. 117. 118. 119. 120. 121. 122. 123. 124. 125. 126. 127. 128. 129. 130. 131. 132. 133. 134. 135. 136. 137. 138. 139. 140. 141. 142. 143. 144. 145. 146. 147. 148. 149. 150. 151. 152. 153. 154. 155. 156. 157. 158. 159. 160. 161. 162. 163. 164. 165. 166. 167. 168. 169. 170. Cholst IN, Steinberg SF, Tropper PJ, et al. The influence of hypermagnesemia on serum calcium and parathyroid hormone levels in human subjects. N Engl J Med 1984; 310:1221. Kerr D, Hosking DJ. Pseudohypoparathyroidism: clinical expression of PTH resistance. QJM 1987; 247:889. Verhoeven GFM, Wilson JD. The syndromes of primary hormone resistance. Metabolism 1979; 28:253. Albright F, Burnett CH, Smith PH, Parson W. Pseudohypoparathyroidism an example of Seabright-Bantam syndrome. Endocrinology 1942; 3:922. Chase LR, Aurbach GD. Parathyroid function and the renal excretion of 3'5'-adenylic acid. Proc Natl Acad Sci U S A 1967; 55:518. Chase LR, Melson GL, Aurbach GD. Pseudohypoparathyroidism: defective excretion of 3'5'-AMP in response to parathyroid hormone. J Clin Invest 1969; 48:1832. Hruska KA, Moskowitz D, Esbrit P, et al. Stimulation of inositol triphosphate and diacylglycerol production in renal tubular cells by parathyroid hormone. J Clin Invest 1987; 79:230. Reid IR, Civitelli R, Halstead LR, et al. Parathyroid hormone acutely elevates intracellular calcium in osteoblast-like cells. Am J Physiol 1987; 252:E45. Cole JA, Eber SL, Poelling RE, et al. A dual mechanism for regulation of kidney phosphate transport by parathyroid hormone. Am J Physiol 1987; 253:E221. Yamaguchi DT, Hahn TJ, Iida-Klein A, et al. Parathyroid hormone-activated calcium channels in an osteoblast-like clonal osteosarcoma cell line: cAMP-dependent and cAMP-independent calcium channels. J Biol Chem 1987; 262:7711. Yamamoto M, Takuwa Y, Masuko S, Ogata E. Effects of endogenous and exogenous parathyroid hormone on tubular reabsorption of calcium in pseudohypoparathyroidism. J Clin Endocrinol Metab 1988; 66:618. Stone MD, Hosking DJ, Garcia-Himmelstine C, et al. The renal response to exogenous parathyroid hormone in treated pseudohypoparathyroidism. Bone 1993; 14:727. Gilman AG. G proteins and dual control of adenylate cyclase. Cell 1984; 36:577. Farfel Z, Bourne H, Iiri T. The expanding spectrum of G protein diseases. N Engl J Med 1999; 340:1012. Hamm H. The many faces of G protein signalling. J Biol Chem 1998; 273:669. Sprang SR. G protein mechanisms: insights from structural analysis. Annu Rev Biochem 1997; 66:639. Levine MA, Downs RW Jr, Moses AM, et al. Resistance to multiple hormones in patients with pseudohypoparathyroidism: association with deficient activity of the guanine nucleotide regulatory protein. Am J Med 1983; 74:545. Farfel Z, Friedman E. Mental deficiency in pseudohypoparathyroidism type 1 is associated with Ns-protein deficiency. Ann Intern Med 1986; 105:197. Weinstock RS, Wright HN, Spiegel AM, et al. Olfactory dysfunction in humans with deficient guanine nucleotide-binding protein. Nature 1986; 322:635. Levine MA, Downs RW Jr, Singer MJ, et al. Deficient activity of guanine nucleotide regulatory protein in erythrocytes from patients with pseudohypoparathyroidism. Biochem Biophys Res Commun 1980; 94:1319. Farfel Z, Brickman AS, Kaslow HR, et al. Deficiency of receptor-cyclase coupling protein in pseudohypoparathyroidism. N Engl J Med 1980; 303:237. Downs RW, Levine MA, Drezner MK, et al. Deficient adenylate cyclase regulatory protein in renal membranes from a patient with pseudohypoparathyroidism. J Clin Invest 1983; 71:231. Albright F, Forbes AP, Henneman PH. Pseudopseudohypoparathyroidism. Trans Assoc Am Physicians 1952; 65:337. Levine MA, Jap T, Mauseth RS, et al. Activity of the stimulatory guanine nucleotide-binding protein is reduced in erythrocytes from patients with pseudohypoparathyroidism and pseudopseudohypoparathyroidism: biochemical, endocrine, and genetic analysis of Albright's hereditary osteodystrophy in six kindreds. J Clin Endocrinol Metab 1986; 62:497. Fitch N. The identification and inheritance of Albright's hereditary osteodystrophy. Am J Med Genet 1982; 11:11. Bray P, Carter A, Simons C, et al. Human cDNA clones for four species of Gsa signal transduction protein. Proc Natl Acad Sci U S A 1986; 83:8893. Kozasa T, Itoh H, Tskukamoto T, Kaziro Y. Isolation and characterization of the human Gsa gene. Proc Natl Acad Sci U S A 1988; 85:2081. Patten JL, Johns DR, Valle D, et al. Mutation in the gene encoding the stimulatory G protein of adenylate cyclase in Albright's hereditary osteodystrophy. N Engl J Med 1990; 322:1412. Weinstein LS, Gejman PV, Friedman E, et al. Mutations of the Gs alpha-subunit gene in Albright hereditary osteodystrophy detected by denaturing gradient gel electrophoresis. Proc Natl Acad Sci U S A 1990; 87:8287. Miric A, Vechio JD, Levine MA. Heterogeneous mutations in the gene encoding the alpha subunit of the stimulatory G protein of adenylyl cyclase in Albright hereditary osteodystrophy. J Clin Endocrinol Metab 1993; 76:1560. Iiri T, Herzmark P, Nakamoto JM, et al. Rapid GDP release from Gsa in patients with gain and loss of function. Nature 1994; 371:164. Weinstein LS, Gejman PV, de Mazancourt P, et al. A heterozygous 4-bp deletion mutation in the Gsa gene (GNAS1) in a patient with Albright hereditary osteodystrophy. Genomics 1992; 13:1319. Levine MA, Modi WS, O'Brien SJ. Mapping of the gene encoding the alpha subunit of the stimulatory G protein of adenylyl cyclase (GNAS1) to 20q13.2-q13.3 in human by in situ hybridization. Genomics 1991; 11:478. Levine MA, Ahn TG, Kaufman K, et al. Genetic deficiency of the alpha-subunit of Gs as the molecular basis for Albright's hereditary osteodystrophy. Proc Natl Acad Sci U S A 1988; 85:617. Carter A, Bardin C, Collins R, et al. Reduced expression of multiple forms of the alpha subunit of the stimulatory GTP-binding protein in pseudohypoparathyroidism type 1a. Proc Natl Acad Sci U S A 1987; 84:7266. Jan de Beur SM, Deng ZC, Ding CL, Levine MA. Amplification of GC rich exon 1 of GNAS1 and identification of three novel missense mutations. (Abstract). Endocrine Soc Abstracts 1998; Abs. OR7-6:62. Warner DR, Gejman PV, Collins RM, Weinstein LS. A novel mutation adjacent of switch III domain of Gs a in a patient with pseudohypoparathyroidism. Mol Endocrinol 1997; 11:1718. Warner DR, Weng G, Yu S, et al. A novel mutation in the Switch 3 region of Gsa in a patient with Albright hereditary osteodystrophy impairs GDP binding and receptor activation. J Biol Chem 1998; 273:23976. Farfel Z, Iiri T, Shapira H, et al. Pseudohypoparathyroidism, a novel mutation in the bg contact region of Gsa impairs receptor stimulation. J Biol Chem 1996; 271:19653. Shapira H, Mouallem M, Shapiro MS, Weisman Y, Farfel Z. Pseudohypoparathyroidism type 1a: two new heterozygous frameshift mutations in exon 5 and 10 of the Gsa gene. Hum Genet 1996; 97:73. Fischer JA, Egert F, Werder E, Born W. An inherited mutation associated with functional deficiency of the alpha subunit of the guanine nucleotide-binding protein Gs in pseudo- and pseudopseudohypoparathyroidism. J Clin Endocrinol Metab 1999; 83:935. Luttikhuis ME, Wilson LC, Leonard JV, Trembath RC. Characterization of a de novo 43-bp deletion of the Gs alpha gene (GNAS1) in Albright hereditary osteodystrophy. Genomics 1994; 21:455. Yu S, Yu D, Hainline BE, et al. A deletion hotspot in exon 7 of the Gs alpha gene (GNAS1) in patients with Albright hereditary osteodystrophy. Hum Mol Genet 1995; 4:2001. Walden U, Weissortel R, Corria Z, et al. Stimulatory guanine nucleotide binding protein subunit 1 mutation in two sibling with pseudohypoparathyroidism type 1a and mother with pseudopseudohypoparathyroidism. Eur J Pediatr 1999; 158:200. Ahmed SF, Dixon PH, Bonthron DT, et al. GNAS1 mutational analysis in pseudohypoparathyroidism. Clin Endocrinol 1998; 49:525. Yokoyama M, Takeda K, Iyota K, et al. A 4 base pair deletion mutation of the Gsa gene in a Japanese patient with pseudohypoparathyroidism. J Endocrinol Invest 1996; 19:236. Wilson LC, Oude Luttikhuis ME, Clayton PT, et al. Parental origin of Gs alpha gene mutations in Albright's hereditary osteodystrophy. J Med Genet 1994; 31:835. Davies SJ, Hughes HE. Imprinting in Albright's hereditary osteodystrophy. J Med Genet 1993; 30:101. Nakamoto JM, Sandstrom AT, Brickman AS, et al. Pseudohypoparathyroidism type 1a from maternal but not paternal transmission of a Gsa gene mutation. Am J Med Genet 1998; 77:261. Hayward BE, Kamiya M, Strain L, et al. The human GNAS1 gene is imprinted and encodes distinct paternally and biallelically expressed G proteins. Proc Natl Acad Sci U S A 1998; 95:10038. Hayward BE, Moran V, Strain L, Bonthron DT. Bidirectional imprinting of a single gene: GNAS1 encodes maternally, paternally, and biallelically derived proteins. Proc Natl Acad Sci U S A 1998; 95:15475. Campbell R, Gosden CM, Bonthron DT. Parental origin of transcription from the human GNAS1 gene. J Med Genet 1994; 31:607. Yu S, Yu D, Lee E, et al. Variable and tissue-specific hormone resistance in heterotrimeric Gs protein alpha-subunit (Gsalpha) knockout mice is due to tissue-specific imprinting of the Gsa gene. Proc Natl Acad Sci U S A 1998; 95:8715. Schwindinger WF, Earnest KA, McCarthy EF, Levine MA. Mice with deficiency of Gas demonstrate a renal but not a skeletal phenotype. American Society for Bone and Mineral Research 1998:S115(Abstr). Speigel A. Mutations in G proteins and G protein-coupled receptors in endocrine disease. J Clin Endocrinol Metab 1996; 81:2434. Ringel MD, Schwindinger WF, Levine MA. Clinical implications of genetic defects in G proteins. Medicine 1996; 171. Schwindinger WF, Francomano CA, Levine MA. Identification of a mutation in the gene encoding the alpha subunit of the stimulatory G protein of adenylyl cyclase in McCune-Albright syndrome. Proc Natl Acad Sci U S A 1992; 89:5152. Weinstein LS, Shenker A, Gejman PV, et al. Activating mutations of the stimulatory G protein in the McCune-Albright syndrome. N Engl J Med 1991; 325:1688. Landis CA, Masters SB, Spada A, et al. GTPase inhibiting mutations activate the alpha chain of Gs and stimulate adenylyl cyclase in human pituitary tumours. Nature 1989; 340:692. Russo D, Arturi F, Wicker R, et al. Genetic alterations in thyroid hyperfunctioning adenomas. J Clin Endocrinol Metab 1995; 80:1347. Lyons J, Landis CA, Griffith H, et al. Two G protein oncogenes in human endocrine tumors. Science 1990; 249:655. Silve C, Santora A, Breslau N, et al. Selective resistance to parathyroid hormone in cultured skin fibroblasts from patients with pseudohypoparathyroidism type 1b. J Clin Endocrinol Metab 1986; 62:640. Fukumoto S, Suzawa M, Takeuchi Y, et al. Absence of mutations in parathyroid hormone (PTH)/PTH-related protein receptor complementary deoxyribonucleic acid in patients with pseudohypoparathyroidism type 1b. J Clin Endocrinol Metab 1996; 81:2554. Ding CL, Usdin TB, LaBuda MC, Levine MA. Molecular genetic analysis of pseudohypoparathyroidism 1B: exclusion of the genes encoding the type 1 and type 2 PTH receptors. (Abstract). J Bone Miner Res 1996; 11: S302. Schipani E, Weinstein LS, Bergwitz C, et al. Pseudohypoparathyroidism type 1b is not caused by mutations in the coding exons of the human parathyroid hormone (PTH)/PTH-related peptide receptor gene. J Clin Endocrinol Metab 1995; 80:1611. Bettoun JD, Minagawa M, Kwan MY, et al. Cloning and characterization of the promoter regions of the human parathyroid hormone (PTH)/PTH-related peptide receptor gene: analysis of deoxyribonucleic acid from normal subjects and patients with pseudohypoparathyroidism type 1b. J Clin Endocrinol Metab 1997; 82:1031. Minagawa M, Watanabe T, Minamitani K, et al. Analysis of P3 promoter of human PTH/PTHrp receptor gene in pseudohypoparathyroidism type 1b. Bone 1998; 23, S462.

170a. Jan de Beur SM, Ding CL, LaBuda MC, et al. Pseudohypoparathyroidism 1b: exclusion of parathyroid hormone and its receptors as candidate disease genes. J Clin Endocrinol Metab 2000; 85:2239. 171. Lanske B, Karapalis AC, Lee K, et al. PTH/PTHrP receptor in early development and Indian hedgehog-related bone growth. Science 1996; 273:663. 172. Juppner H, Schipani E, Bastepe M, et al. The gene responsible for pseudohypoparathyroidism type 1b is paternally imprinted and maps in four unrelated kindreds to chromosome 20q13.3. Proc Natl Acad Sci U S A 1998; 95:11798. 173. Barrett D, Breslau NA, Wax MB, et al. A new form of pseudohypoparathyroidism with abnormal catalytic adenylate cyclase. Am J Physiol 1989; 257:E277. 174. Van Dop C, Bourne HR, Neer RM. Father to son transmission of decreased Ns activity in pseudohypoparathyroidism type 1a. J Clin Endocrinol Metab 1984; 59:825. 175. Levine MA. Pseudohypoparathyroidism. In: Avioli LV, Krane SM, eds. Metabolic bone disease and clinically related disorders, 3rd ed. San Diego: Academic Press, 1998:507. 176. Schuster V, Kress W, Kruse K. Paternal and maternal transmission of pseudohypoparathyroidism type 1a in a family with Albright hereditary osteodystrophy: no evidence of genomic imprinting. (Letter). J Med Genet 1994; 31:84. 177. Schuster V, Eschenhagen T, Kruse K, et al. Endocrine and molecular biological studies in a German family with Albright hereditary osteodystrophy. Eur J Pediatr 1993; 152:185. 178. Drezner MK, Neelon FA, Lebovitz HE. Pseudohypoparathyroidism type II: a possible defect in the reception of the cyclic AMP signal. N Engl J Med 1973; 280:1056. 179. Gascon-Barre M, Haddad P, Provencher SJ, et al. Chronic hypocalcemia of vitamin D deficiency leads to lower intracellular calcium concentrations in rat hepatocytes. J Clin Invest 1994; 93:2159.

180. Breslau NA, Moses AM, Pak CYC. Evidence for bone remodeling but lack of calcium mobilization in response to parathyroid hormone in pseudohypoparathyroidism. J Clin Endocrinol Metab 1983; 57:638. 181. Nagant de Deuxchaines C, Fischer JA, Dambacher MA, et al. Dissociation of parathyroid hormone bioactivity and immunoreactivity in pseudohypoparathyroidism type 1. J Clin Endocrinol Metab 1981; 53:1105. 182. Loveridge N, Fischer JA, Nagant de Deuxchaines C, et al. Inhibition of cytochemical bioactivity of parathyroid hormone by plasma in pseudohypoparathyroidism type 1. J Clin Endocrinol Metab 1982; 54:1274. 183. Mitchell J, Goltzman D. Examination of circulating parathyroid hormone in pseudohypoparathyroidism. J Clin Endocrinol Metab 1985; 61:328. 184. Chernow B, Zaloga G, McFadden E, et al. Hypocalcemia in critically ill patients. Crit Care Med 1982; 10:848. 185. Raskin P, McClain CJ, Medsger TA. Hypocalcemia associated with metastatic bone disease: a retrospective study. Arch Intern Med 1973; 132:539. 186. Smallridge RC, Wray HL, Schaaf M. Hypocalcemia with osteoblastic metastases in a patient with prostatic carcinoma: a cause of secondary hyperparathyroidism. Am J Med 1981; 71:184. 187. Desai TK, Carlson RW, Geheb MA. Prevalence and clinical implications of hypocalcemia in acutely ill patients in a medical intensive care setting. Am J Med 1988; 84:209. 188. Kitanaka S, Takeyama KI, Murayama A, et al. Inactivating mutations in the 25 hydroxyvitamin D 3 1 -hydroxylase gene in patients with pseudovitamin D-deficiency rickets. N Engl J Med 1998; 338:653. 189. Malloy PJ, Wesley Pike J, Feldman D. The vitamin D receptor and the syndrome of hereditary 1,25-dihydroxyvitamin D resistant rickets. Endocr Rev 1999; 20:156. 190. Mallette LE, Kirkland JL, Gagel RF, et al. Synthetic human parathyroid hormone-(1-34) for the study of pseudohypoparathyroidism. J Clin Endocrinol Metab 1988; 67:964. 191. Sohn HE, Furukawa Y, Yumita S, et al. Effect of synthetic 1-34 fragment of human parathyroid hormone on plasma adenosine 3',5'-monophosphate (cAMP) concentrations and the diagnostic criteria based on the plasma cAMP response in Ellsworth-Howard test. Endocrinol Jpn 1984; 31:33. 192. Stirling HF, Darling JA, Barr DG. Plasma cyclic AMP response to intravenous parathyroid hormone in pseudohypoparathyroidism. Acta Paediatr Scand 1991; 80:333. 193. Barr DG, Stirling HF, Darling JA. Evolution of pseudohypoparathyroidism: an informative family study. Arch Dis Child 1994; 70:337. 194. Peacock M, Robertson WG, Nordin BEC. Relation between serum and urinary calcium with particular reference to parathyroid activity. Lancet 1969; 1:384. 195. Nicar NJ, Pak CYC. Calcium bioavailability from calcium carbonate and calcium citrate. J Clin Endocrinol Metab 1985; 61:391. 196. Recker R. Calcium absorption and achlorhydria. N Engl J Med 1985; 313:70. 197. Mason RS, Posen S. The relevance of 25-hydroxycalciferol measurements in the treatment of hypoparathyroidism. Clin Endocrinol (Oxf) 1979; 10:265. 198. Parfitt AM. Thiazide-induced hypercalcemia in vitamin Dtreated hypoparathyroidism. Ann Intern Med 1972; 77:557. 199. Porter RH, Cox BG, Heaney D, et al. Treatment of hypoparathyroid patients with chlorthalidone. N Engl J Med 1978; 298:577. 200. Newman GH, Wade M, Hosking DJ. Effect of bendrofluazide on calcium reabsorption in hypoparathyroidism. Eur J Clin Pharm 1984; 27:41. 201. Gabow PA, Hanson TJ, Popovitzer MM, Schrier RW. Furosemide-induced reduction in ionized calcium in hypoparathyroid patients. Ann Intern Med 1977; 86:579. 202. Eisenberg E. Effects of serum calcium level and parathyroid extracts on phosphate and calcium excretion in hypoparathyroid patients. J Clin Invest 1965; 44:942. 203. Duarte B, Mozes MF, John E, et al. Parathyroid allotransplantation in the treatment of complicated primary hypoparathyroidism. Surgery 1985; 98:1072.

CHAPTER 61 RENAL OSTEODYSTROPHY Principles and Practice of Endocrinology and Metabolism

CHAPTER 61 RENAL OSTEODYSTROPHY

KEVIN J. MARTIN, ESTHER A. GONZALEZ, AND EDUARDO SLATOPOLSKY Pathogenesis Osteitis Fibrosa Retention of Phosphorus Altered Production of Calcitriol Altered Parathyroid Function Skeletal Resistance to the Calcemic Effect of Parathyroid Hormone Impaired Renal Metabolism of Circulating Parathyroid Hormone Osteomalacia and Adynamic Bone b2-Microglobulin and Bone Clinical Features Abnormalities of Laboratory Tests Assays for Parathyroid Hormone Bone Histology Radiologic Abnormalities Prevention, Diagnosis, and Management Dietary Phosphorus Restriction Calcium Supplementation Use of Vitamin D Sterols Use of Calcium and Vitamin D Sterols in Aluminum-Associated Osteomalacia Parathyroidectomy Aluminum Accumulation Renal Transplantation Chapter References

The association of skeletal disease with renal failure has been recognized for many years. The term renal osteodystrophy is used to encompass the many complex disorders of the skeleton that occur in patients with chronic renal disease and includes disorders with high bone turnover such as osteitis fibrosa as well as disorders with abnormally low bone turnover such as adynamic bone and osteomalacia; in addition, osteoporosis and, in children, growth retardation may be seen. Areas of osteosclerosis as well as cystic bone lesions may also be present. These disorders may occur together to variable degrees in any given patient. Thus, to prevent and treat bone disease in patients with renal failure, an understanding of the pathogenesis, diagnosis, and treatment of the various types of bone disease is essential. The prevalence of the various types of bone disease encountered in patients with end-stage renal disease differs somewhat according to the modality of dialysis used, as illustrated in Figure 61-1.

FIGURE 61-1. The spectrum of histologic abnormalities of bone in patients with end-stage renal disease on hemodialysis or peritoneal dialysis. (Adapted from Sherrard DJ, Hercz G, Pei Y, et al. The spectrum of bone disease in end-stage renal failure: an evolving disorder. Kidney Int 1993; 43:436.)

PATHOGENESIS
OSTEITIS FIBROSA Osteitis fibrosa results from high circulating levels of biologically active parathyroid hormone (PTH). Elevated levels of immunoreactive PTH (iPTH) and hyperplasia of the parathyroid glands are among the most consistent abnormalities of mineral metabolism in patients with chronic renal disease.1,2 The elevation of serum PTH occurs in patients with only mild (~20%) reductions in renal function. The principal factors involved in the pathogenesis of the secondary hyperparathyroidism of chronic renal disease include (a) retention of serum phosphorus because of a reduced glomerular filtration rate (GFR), (b) altered production of [1,25(OH)2D], (c) abnormal parathyroid growth and function, and (d) resistance to the calcemic effect of PTH. RETENTION OF PHOSPHORUS Considerable evidence supports the importance of phosphorus retention in the genesis of secondary hyperparathyroidism.3,4 and 5 In normal subjects, the oral administration of elemental phosphorus (1 g) causes an increase in serum phosphorus, a fall in serum ionized calcium levels, and an increase in serum iPTH. Moreover, the long-term administration of a high-phosphorus diet can result in parathyroid hyperplasia and increased levels of iPTH. In experimentally induced renal disease, if dietary phosphorus is restricted in proportion to the decrease in renal function, secondary hyperparathyroidism can be prevented or markedly attenuated for at least 2 years6 (Fig. 61-2). These findings have been confirmed in patients with mild to moderate renal insufficiency. Even in more advanced renal insufficiency, phosphorus restriction has resulted in substantial reductions in serum iPTH, although the levels remain above normal.

FIGURE 61-2. Serial determinations of parathyroid hormone (PTH) in dogs with chronic renal disease (CRD). The animals received a normal diet (OO), a diet with phosphorus content reduced in proportion to the decrement in glomerular filtration rate (), or this latter diet plus 25-hydroxyvitamin D3 [25(OH)D3], 20 g three times per week for 2 years (DD). (Modified from Rutherford WE, Bordier P, Marie P. Phosphate control and 25-hydroxycholecalciferol administration in preventing

experimental renal osteodystrophy in the dog. J Clin Invest 1977; 60:332.)

Studies have begun to refine our understanding of the mechanisms by which phosphorus retention alters parathyroid activity, as illustrated in Figure 61-3. Phosphorus directly increases the secretion of PTH7,8 and also appears to influence the growth of the parathyroid cells9 as well as to indirectly affect parathyroid function. Furthermore, the effects of phosphorus on parathyroid cell growth have been shown to occur early in the course of experimentally induced renal insufficiency.10

FIGURE 61-3. The consequences of hyperphosphatemia (for full description see text). (Pi, inorganic phosphate; PTH, parathyroid hormone.)

The phosphorus retention that decreases the levels of ionized calcium (either by causing the formation of complexes with calcium that precipitate in soft tissues or, alternatively, by decreasing the production of calcitriol) also contributes to the genesis of hyperparathyroidism.11 Thus, the decreased production of calcitriol may lead to hypocalcemia by decreasing intestinal absorption of calcium or by altering PTH secretion (see later). Hyperphosphatemia also leads to resistance to the actions of PTH, as manifested by decreased release of calcium from bone, as well as to resistance to the actions of calcitriol. However, neither hypocalcemia nor hyperphosphatemia is a universal finding in patients with early renal insufficiency,12 and experimental studies have shown that hyperparathyroidism may occur without any hypocalcemia.13 Support for the role of decreased levels of calcitriol is provided by the observation that hyperparathyroidism can be prevented by the administration of calcitriol immediately after the induction of renal insufficiency.14 ALTERED PRODUCTION OF CALCITRIOL Because the kidney is the major site of production of calcitriol14 (see Chap. 54), low levels of calcitriol may accentuate the abnormalities of mineral metabolism seen in patients with chronic renal disease. Thus, the fact that intestinal absorption of calcium is decreased in patients with far-advanced renal insufficiency is explained by the concomitant low levels of serum calcitriol. Considerable evidence exists that abnormal vitamin D metabolism contributes to the altered mineral metabolism in patients with advanced renal insufficiency. Plasma levels of calcitriol in adult patients with mild to moderate decreases in GFR (e.g., 4060 mL per minute) are relatively normal, whereas patients with more advanced renal insufficiency have low levels of this hormone.15,16 These observations agree with studies that show that intestinal absorption of calcium is not decreased in patients with renal insufficiency until the GFR is <30% of normal.17,18 Because PTH increases the production of calcitriol, however, even normal levels of calcitriol may be inappropriately low for the prevailing metabolic state. An additional important role of low levels of calcitriol in the genesis of secondary hyperparathyroidism may be its influence on the regulation of PTH secretion at the level of the parathyroid gland. ALTERED PARATHYROID FUNCTION Several lines of evidence suggest the presence of intrinsic abnormalities in the parathyroid glands resected from patients with renal insufficiency and severe hyperparathyroidism. Membranes prepared from hyperplastic parathyroid glands obtained from patients with chronic renal insufficiency are less susceptible to the inhibition of adenylate cyclase activity by calcium.19 Thus, the setpoint for calcium (i.e., the calcium concentration required to inhibit enzyme activity by 50%) may be elevated in the hyperfunctioning parathyroid glands. Similar findings were obtained for calcium-regulated PTH secretion from isolated parathyroid cells from hyperfunctioning human parathyroid glands.20 These in vitro data show that a higher concentration of calcium is required to inhibit PTH secretion in cells from the hyperplastic glands. This abnormal regulation of PTH secretion by calcium can potentially represent the consequences of low levels of calcitriol, which can have several important effects on parathyroid function, as illustrated in Figure 61-4. The finding that the parathyroid glands express vitamin D receptors was followed by the demonstration that calcitriol directly decreases PTH secretion in vitro.21,22,23,24,25,26 and 27 Clinical studies confirmed the effects of calcitriol on the parathyroid by demonstrating that intravenous administration of calcitriol markedly suppresses PTH secretion in uremic patients on hemodialysis.28 Furthermore, the effect of calcitriol may extend to restoring the abnormal calcium setpoint to normal.29 This latter effect, however, is not seen in all studies, perhaps due to differences in the relative degree of hyperparathyroidism.30 The consequences of low levels of calcitriol on the parathyroid may become amplified because calcitriol receptors are decreased in the parathyroid glands of hyperparathyroid patients31 and animals.32,33 The low levels of calcitriol may lead to decreased expression of calcitriol receptors in the parathyroid, because research has shown that calcitriol receptors can be up-regulated by the administration of calcitriol.34 Calcitriol also regulates the growth of parathyroid cells associated with the altered expression of replication-associated oncogenes.35 Thus, decreased levels of calcitriol may potentially lead to enhanced parathyroid cell growth. These observations are consistent with the findings of altered distribution of calcitriol receptors in the parathyroid glands from uremic subjects, in whom disordered regulation of PTH secretion and nodular hyperplasia occur.36 In some patients, the nodular hyperplasia has been shown to be due to monoclonal expansion of parathyroid cells.37 Parathyroid glands from patients with severe hyperparathyroidism also have decreased expression of the calcium-sensing receptor; this likely contributes to the abnormal regulation of PTH secretion and, possibly, parathyroid cell growth.38,39

FIGURE 61-4. The direct and indirect effects of low levels of calcitriol on parathyroid gland activity and function. (PTH, parathyroid hormone.)

SKELETAL RESISTANCE TO THE CALCEMIC EFFECT OF PARATHYROID HORMONE That the calcemic response to PTH is less than normal in patients with renal insufficiency indicates that this phenomenon may be another cause of hypocalcemia and a further stimulus to PTH secretion in azotemic patients. This has been demonstrated not only by the administration of exogenous PTH but also by the studies showing a delayed recovery from induced hypocalcemia in patients with renal insufficiency, despite a greater augmentation in serum PTH levels. The pathogenesis of the apparent skeletal resistance to the calcemic effect of PTH remains unclear. The suggestion has been made that this abnormality is due to lower levels of calcitriol.40,41 However, other evidence suggests that this phenomenon also may be due to desensitization of the skeletal response to PTH because of the high levels of circulating PTH.42,43 Moreover, hyperphos-phatemia may limit the release of calcium from bone.

IMPAIRED RENAL METABOLISM OF CIRCULATING PARATHYROID HORMONE In patients with chronic renal disease, an increased rate of PTH secretion is the major factor responsible for high plasma PTH levels. Nonetheless, the kidney is important for the peripheral metabolism of PTH.44,45 The metabolic clearance rate for PTH is prolonged in renal insufficiency; thus, decreased removal of this hormone from the circulation may accentuate the hyperparathyroidism of renal failure. OSTEOMALACIA AND ADYNAMIC BONE Osteomalacia, manifested by increased osteoid seams and an abnormal or decreased mineralization front in bone, has not been frequently encountered since the realization that the major factor in its pathogenesis was the accumulation of aluminum in patients with chronic renal failure.46,47 Initially, the source of this metal was thought to be the dialysate, because in geographic areas where the aluminum content of water was high, the incidence of osteomalacia decreased when the aluminum was removed46; however, the most important source of aluminum was found to be the ingestion of aluminum-containing antacids, which were used to complex dietary phosphorus. These are now used infrequently because other phosphate binders have become available to complement dietary phosphorus restriction. Accumulation of iron at the mineralization front has also been described and may also give rise to osteomalacia on bone histology.48 A disorder of bone characterized by extremely low bone turnover in the absence of aluminum, known as adynamic bone, is being increasingly recognized in patients on long-term dialysis and is now a major type of abnormal bone histology in patients with renal disease (see Fig. 61-1).49 Symptoms referable to the skeletal system, such as bone pain or fractures, are relatively uncommon with this type of bone histology.49 The pathogenesis of this abnormality is not well understood. Possibly it represents physiologic suppression of bone turnover as a consequence of the increased use of calcitriol and calcium salts as phosphorus binders, together with dialysate calcium concentrations that result in a positive calcium transfer to the patient. This would account for the increased incidence of this type of bone histology in patients on continuous ambulatory peritoneal dialysis (CAPD) and the associated suppression of the blood levels of PTH. Other patients with this syndromelow bone turnover without demonstrable aluminuminclude patients with diabetes mellitus and elderly patients. b2-MICROGLOBULIN AND BONE Patients on long-term dialysis may develop juxtaarticular radiolucent cysts in bone, destructive arthropathies, and carpal tunnel syndrome. These abnormalities appear to be due to deposition of b2-microglobulin in tissues in the form of amyloid fibrils.50 (See also refs. 50a and 50b.) The factors responsible for its deposition are not known. The increase of plasma b2-microglobulin in dialysis patients reflects the increased production rates and decreased removal by dialysis.

CLINICAL FEATURES
Symptoms of renal osteodystrophy usually appear only in patients with advanced renal failure, although the abnormalities in mineral metabolism occur early in renal insufficiency. Numerous signs and symptoms may appear. The classic symptom of renal osteodystrophy is bone pain, which may be so severe that the patient becomes bedridden. This may occur whether the major abnormality of bone is osteomalacia or osteitis fibrosa. More commonly, the bone pain is vague and localized to the lower back, hips, or legs. Low back pain may arise from collapse of a vertebral body. Chest pain may arise from spontaneous rib fractures. Muscle weakness is often a major associated symptom. The muscle enzymes are normal, and elec-tromyographic findings are nonspecific. In severe cases, electron microscopy of muscle shows local disorganization of myofibrils and dispersion of Z bands.51 Itching is common in patients with renal osteodystrophy, especially in those with severe hyperparathyroidism. This has been attributed to the deposition of calcium in the skin. Peripheral ischemic necrosis and vascular calcifications may also be found. Acute joint pain and periarthritis, owing to the deposition of hydroxyapatite crystals, may also occur concomitantly with severe hyperparathyroidism. Tendon rupture may ensue from the abnormal collagen metabolism of uremia and has been associated with b2-microglobulin accumulation. In children, skeletal deformities, bowing of the tibia and femur, and slipped epiphysis are not uncommon. Growth retardation is usual and is due to the combination of malnutrition, acidosis, and osteomalacia. The correction of these abnormalities may improve growth.

ABNORMALITIES OF LABORATORY TESTS

One of the earliest changes in patients with chronic renal failure is increased plasma levels of PTH. As the disease progresses, hypocalcemia may occur. Although hyperphos-phatemia is not seen until the GFR is <25 mL per minute, one should remember that serum phosphorus remains normal as a consequence of the development of hyperparathyroidism, which occurs early in the course of renal insufficiency. The magnitude of the hyperphosphatemia depends on the amount of phosphorus ingested, the intestinal absorption, and the excretion of phosphorus by the diseased kidney. Hypermag-nesemia occurs with a GFR of <15 mL per minute; the administration of magnesium-containing antacids must be monitored closely. Serum alkaline phosphatase levels are often increased with osteitis fibrosa, osteomalacia, or mixed lesions. Metabolic acidosis commonly is present when the GFR is <25 mL per minute.52 ASSAYS FOR PARATHYROID HORMONE Because hyperparathyroidism plays a major role in renal osteodystrophy, serum PTH must be measured to monitor the degree of hyperparathyroidism and the effects of treatment. Specific two-site immunoassays for intact PTH have become the most widely used and are readily available.53 The desired range for levels of intact PTH that are associated with normal bone turnover in patients with end-stage renal disease appears to be 150 to 300 pg/mL, or 2.5- to 5-fold greater than the upper limit of normal for the intact PTH assays; this higher range presumably represents the need for higher levels of the hormone to overcome the skeletal resistance to PTH discussed previously.54,55 BONE HISTOLOGY An examination of undecalcified sections of bone is required for the precise diagnosis of renal osteodystrophy. Osteitis fibrosa is a manifestation of secondary hyperparathyroidism. This condition manifests with increased osteoclast number and increased resorption surface. Double tetracycline labeling shows an increase in bone turnover and an increased quantity of woven osteoid (see Chap. 55). The deposition of calcium phosphate may explain the osteosclerosis in some uremic patients. Double tetracycline labeling is mandatory for the diagnosis of osteomalacia, characterized by an increase in osteoid and an abnormal or decreased mineralization front. Bone turnover is decreased in osteomalacia as well as in the adynamic bone lesions of renal osteodystrophy. The lesions of osteitis fibrosa and osteomalacia can occur alone or in combination. Special stains may directly indicate the presence of aluminum. RADIOLOGIC ABNORMALITIES The main radiographic feature of hyperparathyroidism is increased bone resorption, most commonly seen in the subperiosteal surfaces of the hands, neck of femur, and clavicle (Fig. 61-5). These radiologic features correlate well with serum PTH levels and evidence of osteitis fibrosa by bone histology.56,57 Skull radiographs show a granular mottled appearance (Fig. 61-6). Osteosclerosis, owing to increased thickness and increased number of trabeculae, accounts for the classic radiographic appearance of the spine, called rugger-jersey spine (Fig. 61-7). Looser zones or pseudofractures may be seen in osteomalacia (see Chap. 63). Although radiographs are useful for the diagnosis of severe hyperparathyroidism, the less distinct changes of osteomalacia or adynamic bone may require bone biopsy for definitive diagnosis. Extraskeletal calcifications may also be found by these radiologic studies (Fig. 61-8). Juxtaarticular radiolucent cysts are suggestive of b2-microglobulin amyloid deposition.

FIGURE 61-5. Chest radiograph of a patient with severe renal osteodystrophy secondary to chronic renal disease as a child. Note the deformity of the rib cage. The inset (top left) demonstrates severe osteitis fibrosa cystica.

FIGURE 61-6. Abnormalities of skull radiograph in a patient with secondary hyperparathyroidism noted on bone biopsy. The skull has a diffuse granular appearance with focal areas of sclerosis.

FIGURE 61-7. Radiograph of lateral lumbar spine of a patient with chronic renal failure and secondary hyperparathyroidism showing a rugger-jersey spine.

FIGURE 61-8. Radiograph of a foot of a patient with chronic renal failure and persistent hyperphosphatemia. The auto-arteriogram is the result of extensive vascular calcification.

PREVENTION, DIAGNOSIS, AND MANAGEMENT

An understanding of the pathogenesis of renal osteodystrophy leads to a number of maneuvers that are effective in its prevention and treatment; these are summarized in Table 61-1.

TABLE 61-1. Treatment of Renal Osteodystrophy at Various Stages of Renal Insufficiency

DIETARY PHOSPHORUS RESTRICTION The cornerstone for the prevention and treatment of renal osteodystrophy is the restriction of phosphorus and the prevention of hyperphosphatemia. The usual phosphorus intake by normal adults is 1 to 1.6 g per day, from meat and dairy products. The dietary restriction of these foods is useful, but as renal failure becomes advanced, additional strategies to limit phosphate absorption are required. This is achieved by the use of calcium salts, calcium carbonate or calcium acetate, or sevelamer, a noncalcium-containing phosphate-binding polymer, taken with meals, which bind the ingested phosphorus in the gastrointestinal tract and prevent its absorption.58,59,60,61 and 62 The goal of this treatment is to reduce the serum phosphate level to ~5 mg/dL. Serum phosphate levels should be closely monitored to avoid hypophosphatemia. Although aluminum-containing antacids are extremely effective in controlling serum phosphorus levels, these agents are not often used except for short-term administration because of the toxicity of aluminum in causing dialysis dementia and osteomalacia.46,63 The administration of antacids such as calcium carbonate, calcium acetate, or sevelamer must be adjusted according to the approximate phosphorus content of the meal. Thus, a small breakfast (5075 mg of phosphorus) may require small amounts of a phosphate binder, whereas a larger meal (600 mg of phosphorus) may require increased amounts. Calcium citrate should not be used, because aluminum absorption may be increased.

CALCIUM SUPPLEMENTATION Because intestinal calcium absorption is decreased in patients with advanced renal failure, and because their diets are usually low in calcium, the supplementation of calcium intake is required.64,65 The minimum requirement is 1.5 g per day of elemental calcium; larger doses may be given, especially if hypocalcemia persists. As discussed previously, large doses of calcium carbonate or calcium acetate are also useful as phosphate binders for the control of hyperphosphatemia and prevention of hyperparathyroidism; if such doses of calcium are given, the levels of serum calcium and phosphate must be monitored frequently to avoid hypercalcemia and an elevation of the calcium-phosphorus product. USE OF VITAMIN D STEROLS Because low levels of calcitriol play a major role in the pathogenesis of hyperparathyroidism in chronic renal disease, the administration of calcitriol has an important role in its treatment. In some countries, 1a-hydroxycholecalciferol, which is metabolized to calcitriol, is used. The administration of calcitriol is effective in achieving suppression of PTH levels with significant improvements in skeletal histology.66 The use of calcitriol may vary according to the degree of renal insufficiency and the degree of hyperparathyroidism. In early renal failure, the oral administration of calcitriol, 0.25 to 0.5 g per day, may be effective in minimizing hyperparathyroidism. However, that therapy should be monitored closely so that hypercalcemia and hypercalciuria do not occur, because a risk exists that renal function may be compromised. In end-stage renal disease, larger doses of calcitriol may be required. In patients on hemodialysis, the intravenous administration of calcitriol is associated with marked suppression of elevated PTH values and improvement in renal osteodystrophy. The doses required are 1 to 3 g given intravenously with each dialysis. As before, levels of calcium and phosphorus and their products must be monitored closely to avoid severe hypercalcemia, an elevated calcium phosphate product, and metastatic calcification. The intermittent administration of large doses of calcitriol orally (25 g twice per week) has also been shown to be effective, and this pulse dosing can achieve a significant reduction in PTH levels, comparable to that achieved with intravenous administration.67,68 and 69 This mode of administration is also applicable to patients on CAPD.70 The increased efficacy of intermittent administration, whether by intravenous or oral routes, may be due to the high levels achieved in serum, which allow increased delivery to peripheral tissues, including the parathyroid glands.71 In the parathyroid, the induction of the vitamin D receptor by calcitriol may allow the glands to become more responsive to the prevailing levels of calcitriol and calcium. If the degree of hyperparathyroidism is severe (e.g., intact PTH of >2000 pg/mL), suppression of PTH by calcitriol may prove difficult. This may be explained by the relative lack of vitamin D receptor and calcium-sensing receptor in nodular hyperplastic parathyroid glands; thus, the administered calcitriol or elevations in serum calcium would be less effective in suppressing PTH biosynthesis. Such large glands may require excision. An analog of calcitriol with less calcemic effects, 19-nor-1a-25-dihydroxy-vitamin D2 (paricalcitol) is available for the treatment of hyperparathyroidism in renal failure.72 Other analogs, 22-oxacalcitriol, falecalcitriol, and 1a-hydroxy vitamin D2, are in clinical trials for the same purpose.73,74 and 75 USE OF CALCIUM AND VITAMIN D STEROLS IN ALUMINUM-ASSOCIATED OSTEOMALACIA Although serum calcium should be monitored in all patients treated with large doses of calcium or calcitriol, patients with aluminum-associated bone disease or low bone turnover, such as adynamic bone, are particularly prone to develop hypercalcemia. Presumably, this tendency to hypercalcemia arises because the increased calcium absorbed in the intestine cannot be deposited in bone in which normal mineralization is defective. Thus, patients who show mild hypercalcemia and markedly lower levels of iPTH than are found in those with severe osteitis fibrosa should be treated with caution; if severe hypercalcemia occurs at relatively low dosages of calcium or calcitriol, adynamic bone or aluminum-associated osteomalacia should be considered. PARATHYROIDECTOMY Although the treatment regimens outlined earlier are often effective in improving mineral homeostasis, reversing the symptoms of bone disease, and reducing the serum levels of PTH, occasionally these measures are not totally effective, and parathyroidectomy may be indicated. When parathyroidectomy is considered, substantial evidence of severe hyperparathyroidism should be present, such as very high serum levels of PTH (as are found with severe osteitis fibrosa) or severe osteitis fibrosa on bone biopsy. Aluminum-associated osteomalacia should be excluded. Other features that are an indication for parathyroidectomy include a high PTH and (a) severe hypercalcemia, especially if symptomatic; (b) intractable pruritus; (c) progressive extraskeletal calcification with elevated calcium-phosphate product; (d) progressive skeletal pain and fractures; and (e) calciphylaxis (ischemic lesions of the skin and soft tissue and vascular calcifications).76,77 and 78 Importantly, hypercalcemia may occur in patients who lack evidence of secondary hyperparathyroidism, such as in aluminum intoxication, or those with adynamic bone. Parathyroidectomy should not be undertaken unless firm evidence of hyperparathyroidism is found. Prior to surgery, parathyroid imaging is helpful.79 The most common surgical procedure is to identify all four parathyroid glands and to remove three totally plus a portion of the remaining gland, leaving 80 to 100 mg of parathyroid tissue. Alternatively, total parathyroidectomy is performed, coupled with autotransplantation of parathyroid tissue to the forearm.80 When the latter procedure is performed, the availability of cryopreservation facilities to store the remaining tissue is desirable so that further implants may be performed if the initial graft fails or is insufficient. Because hypocalcemia may occur postoperatively, beginning treatment with calcitriol before surgery usually is desirable. Postoperative hypocalcemia is controlled with intravenous and oral administration of calcium supplements and calcitriol. Serum phosphorus may also decrease after surgery, so that temporary discontinuance of phosphorus binders is required. Although alkaline phosphatase may rise postoperatively, a fall toward normal may indicate that skeletal remineralization is slowing and that calcium and vitamin D therapy may be reduced. ALUMINUM ACCUMULATION The main sources of aluminum for patients with renal failure are the dialysis water and aluminum-containing antacids. Appropriate water purification should be routine and a dialysis fluid aluminum concentration of <10 g/L should be maintained. Deferoxamine has been found to be useful both for diagnosing and treating aluminum accumulation.81,82 and 83 A rise in serum aluminum of >150 g/L in response to administration of deferoxamine together with a PTH level of <200 pg/mL suggests that the patient is at risk for bone disease secondary to aluminum accumulation.84 The definitive diagnosis is made by bone biopsy with special aluminum staining. If the diagnosis is confirmed, all sources of aluminum must be sought and removed. Deferoxamine therapy may be used to remove large amounts of aluminum (also see Chap. 131). RENAL TRANSPLANTATION Many of the factors responsible for the altered divalent cation metabolism of chronic renal failure are corrected by successful renal transplantation. However, new problems may arise. Hyperparathyroidism gradually decreases after transplantation; PTH levels generally approach normal within 4 to 6 months if renal function is adequate. Sometimes, hyperparathyroidism may persist and hypercalcemia may occur, especially if the hyperparathyroidism was severe and of long duration. Hypophosphatemia may occur after transplantation because of this persistent hyperparathyroidism, because of antacid therapy, or because of an acquired renal tubular defect independent of PTH. The management of mild hypercalcemia and hypophosphatemia usually is conservative, with oral phosphate therapy and careful follow-up. If hypercalcemia persists or is associated with decreasing renal function, and PTH levels remain elevated, parathyroidectomy may be required. Long-term therapy with corticosteroids in renal transplant recipients has been associated with aseptic necrosis of bone and cortico-steroid-induced osteopenia (see Chap. 64 and Chap. 211). CHAPTER REFERENCES
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. Reiss E, Canterbury JM, Egdahl RH. Measurement of serum parathyroid hormone in renal insufficiency. Trans Assoc Am Physicians 1968; 81:104. Arnaud CD. Hyperparathyroidism and renal failure. Kidney Int 1973; 4:89. Slatopolsky E, Caglar S, Pennell IP, et al. On the pathogenesis of hyperparathyroidism in chronic experimental insufficiency in the dog. J Clin Invest 1971; 50:492. Slatopolsky E, Delmez JA. Pathogenesis of secondary hyperparathyroidism. Am J Kidney Dis 1994; 23:229. Slatopolsky E, Bricker NS. The role of phosphorus restriction in the prevention of secondary hyperparathyroidism in chronic renal disease. Kidney Int 1973; 4:141. Rutherford WE, Bordier P, Marie P. Phosphate control and 25-hydroxy-cholecalciferol administration in preventing experimental renal osteodys-trophy in the dog. J Clin Invest 1977; 60:332. Slatopolsky E, Finch J, Denda M, et al. Phosphorus restriction prevents parathyroid gland growth. High phosphorus directly stimulates PTH secretion in vitro. J Clin Invest 1996; 97:2534. Almaden Y, Canalejo A, Hernandez A, et al. Direct effect of phosphorus on PTH secretion from whole rat parathyroid glands in vitro. J Bone Miner Res 1996; 11:970. Naveh-Many T, Rahamimov R, Livni N, Silver J. Parathyroid cell proliferation in normal and chronic renal failure rats. The effects of calcium, phosphate, and vitamin D. J Clin Invest 1995; 96:1786. Denda M, Finch J, Slatopolsky E. Phosphorus accelerates the development of parathyroid hyperplasia and secondary hyperparathyroidism in rats with renal failure. Am J Kidney Dis 1996; 28:596. Tanaka Y, DeLuca HF. The control of 25-hydroxyvitamin D metabolism by inorganic phosphorus. Arch Biochem Biophys 1973; 159:566. Portale AA, Booth BE, Halloran BP, Morris RC Jr. Effect of dietary phosphorus on circulating concentrations of 1,25-dihydroxyvitamin D 3 and immunoreactive parathyroid hormone in children with moderate renal insufficiency. J Clin Invest 1984; 73:1580. Lopez-Hilker S, Gelceram T, Chen Y, et al. Hypocalcemia may not be essential for the development of secondary hyperparathyroidism in chronic renal failure. J Clin Invest 1986; 78:1079. Lopez-Hilker S, Dusso AS, Rapp NS, et al. Phosphorus restriction reverses hyperparathyroidism in uremia independent of changes in calcium and calcitriol. Am J Physiol 1990; 259:F432.

15. Slatopolsky E, Gray R, Adams ND, et al. The pathogenesis of secondary hyperparathyroidism in early renal failure. In: Norman A, ed. Fourth International Workshop on Vitamin D. Berlin: deGruyter, 1979:1209. 16. Cheung AK, Manolagas SC, Catherwood BD, et al. Determination of serum 1,25(OH) 2D3 levels in renal disease. Kidney Int 1983; 24:104. 17. Coburn JW, Koppel MH, Brickman AS, Massry SG. Study of intestinal absorption of calcium in patients with renal failure. Kidney Int 1973; 3:264. 18. Malluche HH, Werner E, Ritz E. Intestinal absorption of calcium and whole-body calcium retention in incipient and advanced renal failure. Miner Electrolyte Metab 1978; 1:263. 19. Bellorin-Font E, Martin KJ, Freitag JJ, et al. Altered adenylate cyclase kinetics in hyperfunctioning human parathyroid glands. J Clin Endocrinol Metab 1981; 52:499. 20. Brown EM, Wilson RE, Eastman RC, et al. Abnormal regulation of parathyroid hormone release by calcium in secondary hyperparathyroidism due to chronic renal failure. J Clin Endocrinol Metab 1982; 54:172. 21. Brumbaugh PF, Hughes MR, Haussler MR. Cytoplasmic and nuclear binding components for 1,25-dihydroxyvitamin D 3 in chick parathyroid glands. Proc Natl Acad Sci U S A 1975; 72:4871. 22. Chertow BS, Baylink DJ, Wergedal MH, et al. Decrease in serum immunoreactive parathyroid hormone in rats and in parathyroid hormone secretion in vivo by 1,25-dihydroxycholecalciferol. J Clin Invest 1975; 56:668. 23. Au WYW, Bukowsky A. Inhibition of PTH secretion by vitamin D metabo-lites in organ cultures of rat parathyroids. Fed Proc 1976; 35:530. 24. Dietel M, Dorn G, Montz R, Altenahr E. Influence of vitamin D3, 1,25-dihydroxyvitamin D3 and 24,25-dihydroxyvitamin D3 on parathyroid hormone secretion, adenosine 3',5'-monophosphate release, and ultrastructure of parathyroid glands in organ culture. Endocrinology 1979; 105:237. 25. Silver J, Russell J, Sherwood LM. Regulation by vitamin-D metabolites of messenger ribonucleic-acid for preproparathyroid hormone in isolated bovine parathyroid cells. Proc Natl Acad Sci U S A 1985; 82:4270. 26. Silver J, Navey-Many T, Mayer H, et al. Regulation by vitamin D metabo-lites of parathyroid hormone gene transcription in vivo in the rat. J Clin Invest 1986; 78:1296. 27. Russell J, Lettieri D, Sherwood LM. Suppression by 1,25-(OH) 2D3 of transcription of the parathyroid hormone gene. Endocrinology 1986; 119:2864. 28. Slatopolsky E, Weerts C, Thielan J, et al. Marked suppression of secondary hyperparathyroidism in intravenous administration of 1,25-dihydroxy-cholecalciferol in uremic patients. J Clin Invest 1984; 74:2136. 29. Delmez AJ, Tindira C, Grooms P, et al. Parathyroid hormone suppression by intravenous 1,25-dihydroxyvitamin D: a role for increased sensitivity to calcium. J Clin Invest 1989; 83:1349. 30. Ramirez JA, Goodman WG, Gornbein J, et al. Direct in vivo comparison of calcium-regulated parathyroid hormone secretion in normal volunteers and patients with secondary hyperparathyroidism. J Clin Endocrinol Metab 1993; 76:1489. 31. Korkor AB. Reduced binding of [3H] 1,25-dihydroxyvitamin D 3 in the parathyroid glands of patients with renal failure. N Engl J Med 1987; 316:1573. 32. Brown AJ, Dusso A, Lopez-Hilker S, et al. 1,25-(OH)2D receptors are decreased in parathyroid glands from chronically uremic dogs. Kidney Int 1989; 35:19. 33. Menke J, Hugel U, Zlotkowski A, et al. Diminished parathyroid 1,25(OH) 2D3 receptors in experimental uremia. Kidney Int 1987; 32:350. 34. Naveh-Many T, Marx R, Keshet E, et al. Regulation of 1,25-dihydroxyvitamin D 3 receptor gene expression by 1,25-dihydroxyvitamin D3 in the parathyroid in vivo. J Clin Invest 1990; 86:1968. 35. Kremer R, Bolivar I, Goltzman D, Hendy GN. Influence of calcium and 1,25-dihydroxycholecalciferol on proliferation and proto-oncogene expression in primary cultures of bovine parathyroid cells. Endocrinology 1989; 125:935. 36. Fukada N, Tanaka H, Tominaga Y, et al. Decreased 1,25-dihydroxyvitamin D 3 receptor density is associated with a more severe form of parathyroid hyperplasia in chronic uremic patients. J Clin Invest 1993; 92:1436. 37. Arnold A, Brown MF, Urena P, et al. Monoclonality of parathyroid tumors in chronic renal failure and in primary parathyroid hyperplasia. J Clin Invest 1995; 95:2047. 38. Kifor O, Moore FD Jr, Wang P, et al. Reduced immunostaining for the extra-cellular Ca 2+-sensing receptor in primary and uremic secondary hyperparathyroidism [see comments]. J Clin Endocrinol Metab 1996; 81:1598. 39. Gogusev J, Duchambon P, Hory B, et al. Depressed expression of calcium receptor in parathyroid gland tissue of patients with hyperparathyroidism. Kidney Int 1997; 51:328. 40. Somerville PJ, Kaye M. Resistance to parathyroid hormone in renal failure: role of vitamin D metabolites. Kidney Int 1978; 14:245. 41. Massry SG, Stein R, Garty J, et al. Skeletal resistance to the calcemic action of parathyroid hormone in uremia: role of 1,25(OH) 2D3. Kidney Int 1976; 9:467. 42. Galceran T, Martin KJ, Morrissey JJ, Slatopolsky E. Role of 1,25-dihydroxy-vitamin D on the skeletal resistance to parathyroid hormone. Kidney Int 1987; 32:801. 43. Rodruigez M, Felsenfeld AJ, Llach F. Calcemic response to parathyroid hormone in renal failure: role of calcitriol and the effect of parathyroidectomy. Kidney Int 1991; 40:1063. 44. Hruska KA, Kopelman R, Rutherford WE, et al. Metabolism of immunore-active parathyroid hormone in the dog: the role of the kidney and the effects of chronic renal disease. J Clin Invest 1975; 56:39. 45. Martin KJ, Hruska KA, Lewis J, et al. The renal handling of parathyroid hormone: role of peritubular uptake and glomerular filtration. J Clin Invest 1977; 60:808. 46. Ward MK, Feest TG, Ellis HA, et al. Osteomalacic dialysis osteodystrophy: evidence for a water borne aetiological agent, probably aluminum. Lancet 1978; 1:841. 47. Gonzlez EA, Martin KJ. Aluminum and renal osteodystrophy: a diminishing clinical problem. Trends Endocrinol Metab 1992; 3:371. 48. Phelps KR, Vigorita VJ, Bansal M, Einhorn TA. Histochemical demonstration of iron but not aluminum in a case of dialysis-associated osteomalacia. Am J Med 1988; 84:775. 49. Pei Y, Hercz G, Greenwood C, et al. Non-invasive prediction of aluminum bone disease in hemo- and peritoneal dialysis patients. Kidney Int 1992; 41:1374. 50. Gejyo F, Yamada T, Odani S, et al. A new form of amyloid protein associated with chronic hemodialysis was identified as b-2 microglobulin. Biochem Biophys Res Commun 1985; 129:701. 50a. Moe SM, Singh GK, Bailey AM. Beta-2 microglobulin induces MMP-1 but not TIMP1 expression in human synovial fibroblasts. Kidney Int 2000; 57:2023. 50b. Lornoy W, Becaus I, Billiouw JM, et al. On-line haemodiafiltration. Remarkable removal of beta2-microglobulin. Long-term clinical observations. Nephrol Dial Transplant 2000; 15(Suppl 1):49. 51. Schoenfeld PJ, Martin JA, Barnes B, Teitelbaum SL. Amelioration of myopathy with 25-dihydroxyvitamin D3 therapy (25 OH D3) in patients on chronic hemodialysis. Abstract Book. Third Workshop on Vitamin D. Asilomar, California, 1977:160. 52. Hakim RM, Lazarus JM. Biochemical parameters in chronic renal failure. Am J Kidney Dis 1988; 11:238. 53. Nussbaum SR, Zahradnik RJ, Lavigne JR, et al. Highly sensitive two-site immunoradiometric assay of parathyrin, and its clinical utility in evaluating patients with hypercalcemia. Clin Chem 1987; 33:1364. 54. Wang M, Hercz G, Sherrard DJ, et al. Relationship between intact 1-84 parathyroid hormone and bone histomorphometric parameters in dialysis patients without aluminum toxicity. Am J Kidney Dis 1995; 26:836. 55. Quarles LD, Lobaugh B, Murphy G. Intact parathyroid hormone overestimates the presence and severity of parathyroid-mediated osseous abnormalities in uremia. J Clin Endocrinol Metab 1992; 75:145. 56. Hruska KA, Teitelbaum SL, Kopelman R, et al. The predictability of the his-tologic features of uremic bone disease by non-invasive techniques. Metab Bone Dis Relat Res 1978; 1:39. 57. Ritz E, Malluche HH, Bommar J, et al. Metabolic bone disease in patients on maintenance hemodialysis. Nephron 1974; 12:393. 58. Morniere PH, Roussel A, Tahiri Y, et al. Substitution of aluminum hydroxide by high doses of calcium carbonate in patients on chronic hemodialysis: disappearance of hyperaluminemia and equal control of hyperparathyroidism. Proc Eur Dial Transplant Assoc 1982; 19:784. 59. Slatopolsky E, Weerts C, Lopez-Hilker S, et al. Calcium carbonate is an effective phosphate binder in dialysis patients. N Engl J Med 1986; 315:157. 60. Salusky IB, Coburn JW, Foley J, et al. Calcium carbonate as a phosphate binder in children on dialysis. Kidney Int 1985; 27:185A. 61. Sheikh MS, Maquire JA, Emmett M, et al. Reduction of dietary phosphorus absorption by phosphorus binders. J Clin Invest 1989; 83:66. 62. Slatopolsky EA, Burke SK, Dillon MA, and the RenaGel Study Group. Rena-Gel, a non-absorbed calcium- and aluminum free phosphate binder, lowers serum phosphorus and parathyroid hormone. Kidney Int 1999; 55:299. 63. Fournier AE, Johnson WJ, Taves DR, et al. Etiology of hyperparathyroidism and bone disease during chronic hemodialysis. I. Association of bone disease with potentially etiologic factors. J Clin Invest 1971; 50:592. 64. Kopple JD, Coburn JW. Metabolic studies of low protein diets in uremia. II. Calcium, phosphorus and magnesium. Medicine 1973; 52:597. 65. Clarkson EM, Eastwood JB, Koutsaimanis KG, de Wardener HE. Net intestinal absorption of calcium in patients with chronic renal failure. Kidney Int 1973; 3:258. 66. Cannella G, Bonucci E, Rolla D, et al. Evidence of healing of secondary hyperparathyroidism in chronically hemodialyzed uremic patients treated with long-term intravenous calcitriol. Kidney Int 1994; 46:1124. 67. Indridason OS, Quarles LD. Oral versus intravenous calcitriol: is the route of administration really important? Curr Opin Nephrol Hypertens 1995; 4:307. 68. Tsukamoto Y, Nomura M, Maurmo F. Pharmacological parathyroidectomy by oral 1,25(OH)2D3 pulse therapy. Nephron 1989; 51:130. 69. Tsukamoto Y, Nomura M, Takahashi Y, et al. The oral 1,25(OH) 2D3 pulse therapy in hemodialysis patients with severe secondary hyperparathyroidism. Nephron 1991; 57:23. 70. Martin KJ, Ballal HS, Domoto DT, et al. Pulse oral calcitriol for the treatment of hyperparathyroidism in patients on continuous ambulatory peritoneal dialysis: preliminary observations. Am J Kidney Dis 1992; 19:540. 71. Reichel H, Szabo A, Uhl J, et al. Intermittent versus continuous administration of 1,25-dihydroxyvitamin D3 in experimental renal hyperparathyroidism. Kidney Int 1993; 44:1259. 72. Martin KJ, Gonzlez EA, Gellens M, et al. 19-Nor-1-a-25-dihydroxyvitamin D 2 (paricalcitol) safely and effectively reduces the levels of intact PTH in patients on hemodialysis. J Am Soc of Nephrol 1998; 10:1427. 73. Kurokawa K, Akizawa T, Suzuki M, et al. Effect of 22-oxacalcitriol on hyperparathyroidism of dialysis patients: results of a preliminary study. Nephrol Dial Transplant 1996; 11:121. 74. Tan AU Jr, Levine BS, Mazess RB, et al. Effective suppression of parathyroid hormone by 1 alpha-hydroxy-vitamin D2 in hemodialysis patients with moderate to severe secondary hyperparathyroidism. Kidney Int 1997; 51:317. 75. Akiba T, Marumo F, Owada A, et al. Controlled trial of falecalcitriol versus alfacalcidol in suppression of parathyroid hormone in hemodialysis patients with secondary hyperparathyroidism. Am J Kidney Dis 1998; 32:238. 76. Bleyer AJ, Choi M, Igwemezie B, et al. A case control study of proximal cal-ciphylaxis. Am J Kidney Dis 1998; 32:376. 77. Coates T, Kirkland GS, Dymock RB, et al. Cutaneous necrosis from calcific uremic arteriolopathy. Am J Kidney Dis 1998; 32:384. 78. Llach F. Calcific uremic arteriolopathy (calciphylaxis): an evolving entity? Am J Kidney Dis 1998; 32:514. 79. Ambrosoni P, Olaizola I, Heuguerot C, et al. The role of imaging techniques in the study of renal osteodystrophy. Am J Med Sci 2000; 320:90. 80. Wells SA Jr, Gumells JC, Shelburne JD, et al. Transplantation of the parathyroid glands in man: clinical indications and results. Surgery 1975; 78:34. 81. Milliner DS, Nebeker HG, Ott SM, et al. Use of the desferrioxamine infusion test in the diagnosis of aluminum-related osteodystrophy. Ann Intern Med 1984; 101:775. 82. Ackrill P, Day JP, Gargstang FM, et al. Treatment of fracturing renal osteo-dystrophy by desferrioxamine. Proc Eur Dial Transplant Assoc 1982; 19:203. 83. Malluche HH, Smith AL, Abreo K, Faugere M-C. The use of desferrioxamine in the management of aluminum accumulation in bone in patients with renal failure. N Engl J Med 1984; 311:140. 84. Pei Y, Hercz G, Greenwood C, et al. Non-invasive prediction of aluminum bone disease in hemo- and peritoneal dialysis patients. Kidney Int 1992; 41:1374.

CHAPTER 62 SURGERY OF THE PARATHYROID GLANDS Principles and Practice of Endocrinology and Metabolism

CHAPTER 62 SURGERY OF THE PARATHYROID GLANDS

GERARD M. DOHERTY AND SAMUEL A. WELLS, JR. Standard Initial Operative Management Promising Techniques for Initial Parathyroid Exploration Concise Parathyroidectomy Radioguided Parathyroidectomy Videoscopic Parathyroidectomy Techniques for Localization of Abnormal Parathyroid Tissue in Patients with Persistent or Recurrent Hyperparathyroidism Strategy at Reoperation for Recurrent or Persistent Hyperparathyroidism Mediastinal Exploration Parathyroid Autotransplantation Immediate Versus Delayed Autografting Parathyroid Carcinoma Chapter References

Hyperparathyroidism occurs in ~1 in 1000 persons in the United States. Successful parathyroid surgery requires an understanding of the anatomy of the head and neck area as well as the embryology and pathology of the parathyroid glands (see Chap. 48).

STANDARD INITIAL OPERATIVE MANAGEMENT

Under general anesthesia, the patient is positioned supine with arms at the sides and the neck extended to maximally expose the anterior cervical region. A curvilinear incision is made low in the neck and extended through the platysma muscle. The strap muscles are separated in the midline to expose the underlying thyroid gland. After mobilization and medial retraction of the respective thyroid lobe, the recurrent laryngeal nerve and inferior thyroid artery are identified, and the search for the parathyroid glands begins. The goal of the initial neck exploration for hyperparathyroidism is first to identify all four parathyroid glands and then to remove those that are enlarged. This requires exploration of both sides of the neck to identify all parathyroid glands. Stopping the exploration after finding one normal gland and one abnormal gland on one side of the neck leads to an increased incidence of missed tumors.1 The size of the parathyroid gland should be the factor that determines resection, because the pathologists' interpretations of adenoma or hyperplasia are misleading. In most patients, only one parathyroid gland is enlarged. The large parathyroid glands should be excised, and the normal-appearing parathyroid glands should be left in situ. The surgical decision should not be based only on the histologic interpretation. In addition, tests of gland density (water immersion test) and histochemical staining for cytoplasmic lipid content are unreliable indicators of parathyroid gland behavior. In most patients with single-gland disease, removal of the enlarged parathyroid gland is curative.2 Infrequently, patients with hyperparathyroidism harbor two or three enlarged parathyroid glands, all of which should be removed. In one study, 76 patients with hyperparathyroidism caused by two or three enlarged parathyroid glands were evaluated from 12 to 140 months after resection of only the large parathyroids.3 Persistent or recurrent postoperative hypercalcemia developed in eight patients (10.5%) from 1 to 133 months postoperatively. Thus, these patients appear to have an increased incidence of persistent or recurrent hypercalcemia compared with patients with hyperparathyroidism caused by single-gland disease. A difficult management problem occurs in patients with hyperparathyroidism caused by four-gland enlargement (parathyroid hyperplasia).4 This entity occurs in 10% to 15% of cases. Primary hyperplasia includes the rare water clear cell hyperplasia and the more common chief cell hyperplasia, which may occur alone or in association with certain familial endocrinopathies, including multiple endocrine neoplasia types 1 (MEN1) and 2A (MEN2A) (see Chap. 188). These patients are especially difficult to manage, and the postoperative results are less satisfactory than for patients undergoing surgery for one-, two-, or three-gland enlargement. Patients with parathyroid hyperplasia may be managed by subtotal parathyroidectomy (removal of three and one-half glands) or, alternatively, total parathyroidectomy and heterotopic autotransplantation of parathyroid tissue to the forearm. The former approach is more commonly used and leaves the remaining parathyroid tissue with its native blood supply. The latter approach has the advantage of removing all known parathyroid tissue from the neck and grafting it to a heterotopic site to simplify reoperation in case of recurrent hyperparathyroidism. Both operations are dependent on a meticulous initial neck exploration to identify all parathyroid glands. After subtotal parathyroidectomy for nonfamilial parathyroid hyperplasia, the incidence of recurrent hypercalcemia is 0% to 16%, and the incidence of permanent hypoparathyroidism is 4% to 5%.5,6,7 and 8 In patients with MEN1, however, subtotal parathyroidectomy has led to recurrent hypercalcemia in 26% to 66% of patients at long-term follow-up (mean time to recurrence, >5 years). 9,10 and 11 Total parathyroidectomy is associated with a significant incidence of permanent hypoparathyroidism (536%) and possibly a higher incidence of graft-dependent recurrent hyperparathyroidism (familial, 2064%; nonfamilial, 20%).11,12 However, reoperation for recurrent hypercalcemia is simpler if the parathyroid tissue is in the forearm rather than in the neck. For patients with nonfa-milial parathyroid hyperplasia, either approach is acceptable; however, patients with MEN1 or MEN2A should have total parathyroidectomy with autotransplantation. Occasionally, only three parathyroid glands are found or, if four glands are seen, none is enlarged. The surgeon then must explore the thyrothymic ligament and superior mediastinum through the cervical incision, even removing a portion of the thymus if a lower parathyroid gland is missing. The lateral neck, from the thymus to the pharynx, should be explored for an undescended parathymus. Rarely, an enlarged parathyroid gland is located within a crevice in the thyroid, between nodules, or it may be embedded within the thyroid parenchyma. Intraoperative ultrasonography may be useful to identify ectopic parathyroid tissue either within the thyroid or elsewhere in the neck. As a last resort, the surgeon should consider resecting the thyroid lobe on the side of the missing parathyroid gland. The superior parathyroid glands are more constant in location but are far posterior and may be overlooked. The upper glands, when enlarged, may descend into the tracheoesophageal groove or into the posterior mediastinum, where they may be extracted by the examining finger. Confirming all abnormal parathyroid tissue by biopsy, and confirming each parathyroid gland in patients with more than one abnormal gland, is wise. However, one should not completely resect a normal parathyroid gland. If no enlarged parathyroid gland is identified after a thorough examination of the neck, the operation should be terminated. Postoperatively, the biochemical diagnosis of primary hyperparathyroidism should be confirmed. Localization studies should then be performed before repeat neck exploration. Localization procedures are not generally indicated in patients undergoing initial full neck exploration for hyperparathyroidism because the experienced endocrine surgeon identifies four parathyroid glands and successfully performs corrective surgery in nearly all cases.13

PROMISING TECHNIQUES FOR INITIAL PARATHYROID EXPLORATION

Three technological advancements have made selective exploration strategies viable in the initial exploration of the neck. These advances are (a) the ability to rapidly (in 1015 minutes) and accurately measure the serum level of PTH; (b) the improved ability to localize abnormal parathyroid tissue with technetium Tc 99m sestamibi (99mTc sestamibi) scanning, using either preoperative imaging or intraoperative handheld gamma probe localization; and (c) the improvement of videoscopic surgical instruments. A variety of strategies have been described that use combinations of these technologies to treat primary hyperparathyroidism. As the standard full neck exploration cures ~95% of patients, these newer strategies are designed to match those outcome results, with less surgical trauma. The long-term evaluations of these approaches remain to be accomplished; however, the early results are very promising. CONCISE PARATHYROIDECTOMY The term concise parathyroidectomy was first used to describe a unilateral exploration approach.14 This approach uses preoperative 99mTc sestamibi imaging to select one side of the neck for exploration. During the operation, parathyroid hormone (PTH) levels are measured before and after the removal of the localized abnormal parathyroid tissue. If the PTH level falls sufficiently after the resection, then the exploration is terminated. If the PTH level does not fall, then the neck is explored further.15 This approach avoids bilateral neck exploration in a significant number of patients, and may shorten both operative time and the risk of recurrent laryngeal nerve injury on the side contralateral to the abnormal parathyroid tissue. The more limited exploration may also facilitate the use of local or regional anesthesia with sedation, avoiding the morbidity of general anesthesia. RADIOGUIDED PARATHYROIDECTOMY An alternative strategy depends on the ability to localize the abnormal parathyroid tissue using 99m Tc sestamibi scanning and a handheld gamma probe for intraoperative localization.16 This technique can be performed with or without concurrent intraoperative PTH monitoring. Because the handheld probe can be used to localize the abnormal parathyroid gland in many cases, very small incisions and local anesthesia can be used to identify and remove the localized parathyroid gland. Prospective trials are currently under way at several institutions to evaluate this technique as well as the utility of combining the radioguided approach with

intraoperative PTH monitoring to avoid persistent hyperparathyroidism in patients with multiple gland disease. VIDEOSCOPIC PARATHYROIDECTOMY A minimally invasive approach to the removal of the parathyroid glands has been reported using videoscopic surgical techniques. This approach is currently applied in selected centers, and always in combination with preoperative 99m Tc sestamibi scanning and intraoperative PTH monitoring.17 The technique has been performed from a variety of anatomic approaches, but to date includes visualization of just one parathyroid gland and resection using small (25 mm) videoscopic instruments, through an operative space created by a balloon expander. The value of this technique awaits the results of further experience.

TECHNIQUES FOR LOCALIZATION OF ABNORMAL PARATHYROID TISSUE IN PATIENTS WITH PERSISTENT OR RECURRENT HYPERPARATHYROIDISM
Several noninvasive techniques have been developed for localizing hyperfunctioning parathyroid tissue. Barium swallow, cine-esophagography, and thyroid scanning are rarely helpful, whereas other procedures such as computed tomographic (CT) scans, magnetic resonance imaging, 99mTc sestamibi scanning, and high-resolution real-time ultrasonography identify enlarged parathyroid glands in 40% to 60% of patients studied.18,19 and 20 CT scans, 99mTc sestamibi scans, and sestamibi/iodine subtraction single photon emission computed tomography (SPECT)20a are most useful for identifying lesions in the neck and mediastinum (Fig. 62-1). Sometimes, these techniques are coupled with needle aspiration of an identified mass followed by assay of the aspirate for PTH. Markedly increased PTH values, compared with an aspirate from muscle, document the presence of parathyroid tissue.

FIGURE 62-1. Sestamibi scan. The technetium-99m sestamibi scintigram shows concentration of the nuclide in ectopic parathyroid tissue in the anterior mediastinum (arrow). B, Computed tomographic scan of the mediastinum in the same patient as shown in A reveals a soft-tissue mass anterior to the brachiocephalic artery consistent with parathyroid adenoma (arrow).

Invasive localization techniques include selective thyroid arte-riography and selective venous sampling with assay of plasma levels of PTH.19,21 Selective arteriography (Fig. 62-2) identifies hyperfunctioning parathyroid tissue in 50% to 70% of cases. Arte-riography should be performed only by an experienced angiogra-pher, because the technique may have significant morbidity. In rare patients for whom there are significant reasons to avoid reop-eration, angiographic ablation of hyperfunctioning parathyroid tissue may be appropriate.22 Selective thyroid venous sampling with PTH assay is useful in lateralizing the side of the neck in which hyperfunctioning parathyroid tissue resides (Fig. 62-3). The combined techniques of arteriography and selective venous catheterization can accurately localize hyperfunctioning tissue in 70% to 85% of reoperative cases. For patients who have previously undergone exploration by an experienced endocrine surgeon, most clinicians would use noninvasive and then (if no two studies concur) invasive localization studies.

FIGURE 62-2. Selective arteriography. Late phase of an inferior thyroid arteriogram. A left lower parathyroid adenoma stain is indicated by the four small arrows. The normal thyroid gland blush is superior and is identified as such by superimposition of the thyroid scan. The curved arrow denotes the draining inferior thyroid vein. (From Wells SA Jr. The parathyroid glands. In: Sabiston DC Jr, ed. Textbook of surgery. Phila-delphia: WB Saunders, 1972:656.)

FIGURE 62-3. Selective thyroid venous sampling in a patient with a retained hyperfunctioning right inferior parathyroid gland. Plasma parathyroid hormone concentrations in samples obtained at different points in the venous circulation. Veins are jugular (J), innominate (I), superior vena cava (SVC), superior thyroid (ST), left inferior thyroid (LIT), and right inferior thyroid (RIT). MBL is mean background level; sites of sampling indicated by ; adjacent numbers indicate plasma PTH concentration in ng/mL. (Adapted from Potts J, et al. Parathyroid re-exploration sequence, synthesis, immunoassay studies. Am J Med 1971; 50:639.)

STRATEGY AT REOPERATION FOR RECURRENT OR PERSISTENT HYPERPARATHYROIDISM

Parathyroid reoperation may be necessary in cases of persistent or recurrent primary hyperparathyroidism. Persistent primary hyperparathyroidism is the more common cause of postoperative hypercalcemia and represents a continuation of an elevated serum calcium level through the immediate postoperative period or its development within 6 months of operation. This usually is the result of a missed hyperfunctioning parathyroid gland or of inadequate resection of hyperfunctioning parathyroid tissue in a patient with multiglandular hyperparathyroid-ism. Presumably, recurrent primary hyperparathyroidism results from hyperfunction of a previously normal parathyroid gland, incomplete resection of a parathyroid, incomplete resection of hyperfunctional parathyroid tissue, or hypertrophy of autotrans-planted abnormal parathyroid tissue. The severity of the parathyroid disease should be carefully assessed to justify the increased risk of parathyroid reoperation. Reexploration is usually limited to patients

with a serum calcium concentration of >12 mg/dL or patients who have renal stones and in whom deteriorating renal function, skeletal fractures, peptic ulcer disease, hypertension, neuromuscular dysfunction, or other complications of primary hyperparathyroidism are present, regardless of the degree of hypercalcemia. In a totally asymptomatic patient with a serum calcium level of <11 mg/dL and no evidence of skeletal or renal disease, a nonoperative course should be considered. Before repeat neck exploration, the surgeon should review previous surgical notes, pathology reports, and histologic sections of all resected tissue. Reoperation is generally deferred until localization tests can be performed. If initially generalized parathyroid gland enlargement was found and either a three-gland or a three-and-one-half-gland parathyroidectomy was performed and confirmed histologically, one may undertake repeat neck exploration without performing localization procedures, on the assumption that the remaining gland or gland remnant is the cause of hypercalcemia. The same course may be undertaken if the previous neck exploration has been performed by an inexperienced parathyroid surgeon. In 75% to 90% of patients undergoing repeat surgery for primary hyperparathyroidism, the hyperfunctioning tissue can be reached through a cervical incision (Fig. 62-4A). Therefore, the neck should be explored first. Even if a parathyroid tumor is identified in the mediastinum by localization techniques, it may be possible to extract it through a cervical incision. The use of a thymectomy retractor can facilitate this mediastinal exploration.23 Repeat neck exploration is technically more difficult than the initial surgery and has a greater risk of recurrent laryngeal nerve damage and permanent hypoparathyroidism. Hypoparathyroidism can be prevented by cryopreserving a portion of abnormal parathyroid tissue. The reported success rate in repeat parathyroid surgery varies from 60% to 91%.

FIGURE 62-4. A, Schematic representation of location, at reexploration, of 104 missing parathyroid glands. The sites in which the missing parathyroid glands were most frequently located were the superior posterior mediastinum, the mediastinal thymus, the superior thyroid pole, the thymic tongue, and the inferior portion of the thyroid gland. Thus, 81% of the missing glands were reached by way of the neck and 19% by way of the mediastinum. B, Surgical technique of median sternotomy to reach an enlarged parathyroid gland in the thymus. (A from Wang CA. Parathyroid re-exploration: a clinical and pathological study of 112 cases. Ann Surg 1977; 186:140.)

MEDIASTINAL EXPLORATION
The mediastinum is explored through either a cervical incision23 or a median sternotomy (see Fig. 62-4B). If the enlarged parathyroid gland is visible in the thymus, it should be selectively excised. If no parathyroid is evident, the entire thymus should be removed and sectioned. The parathyroid tumor may not be identified until the pathologist reviews the permanent sections of the submitted tissue. The surgeon should also explore the entire mediastinum around the great vessels and the pericardium. After an enlarged parathyroid gland is identified, if doubt exists that normal parathyroid tissue remains in the neck, the resected parathyroid gland should be sliced into 1-mm 3-mm pieces and cryopreserved in liquid nitrogen.24 If the patient has been rendered hypoparathyroid, then the viably frozen tissue can be autotransplanted into a forearm muscle.

PARATHYROID AUTOTRANSPLANTATION
Clinical circumstances in which parathyroid autotransplantation is useful include radical thyroid surgery with either unavoidable or inadvertent devascularization of parathyroid tissue, renal osteodystrophy, primary parathyroid hyperplasia, and repeat surgery for recurrent or persistent hyperparathyroidism. The decision to autotransplant parathyroid tissue depends on the individual patient. In the case of inadvertent devascularization of normal parathyroid tissue, such as during radical thyroid surgery, immediate autotransplantation is indicated. The devascularized gland is removed and placed in cold saline. A frozen section is obtained to confirm that the tissue is parathyroid. After thyroidectomy, the parathyroid tissue is sliced into 1-mm 3-mm slivers and immediately implanted into the sternocleidomas-toid muscle. Each implantation site contains one or two parathyroid fragments and is marked with a silk suture. Some investigators question the need to perform this technique if other in situ parathyroid glands appear viable. However, assessing viability by macroscopic appearance is difficult. If any question exists regarding adequacy of the blood supply to a parathyroid, the gland should be removed and transplanted. Reduction of the hyperplastic parathyroid tissue mass is indicated in patients with renal osteodystrophy who are refractory to medical management and who have fractures or severe bone pain, extraosseous soft tissue or vascular calcification, or intractable pruritus (see Chap. 61). The role of parathyroid surgery in the management of secondary hyperparathyroidism has been complicated by the recognition of aluminum-related osteodystrophy (see Chap. 61 and Chap. 131). All patients with severe secondary hyperparathyroidism (tertiary hyperparathyroidism) should be evaluated for the presence of this entity. Evidence suggests that, for these patients, parathyroid surgery is not the treatment of choice and actually may further predispose to aluminum bone disease. However, in patients with severe renal osteodystrophy that is not related to aluminum and in whom the control of secondary hyperparathyroidism is impossible by an aggressive medical approach, surgery may be necessary. In this setting, subtotal parathyroidectomy is generally indicated.25 Total parathyroidectomy with heterotopic autografting4,26,27 and 28 is an equivalent or superior option because cervical reexploration for recurrent disease, with the risk of hypoparathyroidism and recurrent laryngeal nerve injury, is precluded, and the bone symptom improvements may be more durable.29 A thorough neck exploration is undertaken, with removal of all parathyroid glands. Each gland should have a small portion removed for frozen-section examination. These glands are then placed in cold sterile tissue culture medium and sliced into 1-mm 3-mm pieces. After closure of the cervical incision, ~20 pieces of tissue are implanted into the brachioradialis muscle of the nondominant forearm. In cases in which fewer than four parathyroid glands are identified despite a thorough neck exploration, the harvested parathyroid tissue is cryopreserved, and parathyroid grafting is delayed. The patient should then be monitored postoperatively for evidence of parathyroid hyperfunction or hypofunction. If prolonged hypocalcemia develops, forearm engraftment under local anesthesia can be carried out at a convenient time. Persistent hypercalcemia indicates that an inadequate amount of parathyroid tissue was resected. A 10% to 40% incidence of persistent or recurrent hyperparathyroidism is reported in patients managed by subtotal parathyroidectomy for renal osteodystrophy.30 Cervical reexploration results in postoperative hypoparathyroidism in approximately one-third of these patients. With immediate or delayed autografting, graft failure is reported to occur in 5% of patients; hyperfunction of the heterotopic autograft is readily managed by partial graft resection under local anesthesia.

IMMEDIATE VERSUS DELAYED AUTOGRAFTING

The decision to use immediate versus delayed autografting depends on the potential risk of graft loss during the freezing and thawing process, versus the risk of graft-dependent hypercalcemia. The incidence of graft-dependent hypercalcemia is 30% within 3 months to 2 years after the autografting of fresh hyperfunctioning parathyroid tissue.12 Also, cases of recalcitrant or recurrent graft-dependent hypercalcemia have been reported.31 Whether persistent parathyroid hyperfunction is caused by pathologic tissue left in situ or by the hyperfunctioning graft may be unclear. Assay for forearm PTH levels may aid in this differentiation. For these reasons, some have advocated delayed autografting using in vitrotested cryopreserved parathyroid tissue. If delayed transplantation is the decided course, fresh tissue removed at the time of operation is divided into 1-mm 3-mm fragments and cryopreserved.24 Long-term tissue viability has been demonstrated by successful engraftment after 9 to 48 months of storage. During subsequent transplantation, the tissue is thawed, and fragments are implanted into the forearm muscle pockets as a heterotopic autograft. A single sliver of parathyroid tissue is implanted into an individual muscle pocket. The musculature and fascia above the graft are then enclosed with a single 5-0 nonabsorbable suture, which provides a marker of the graft site. Implantation must be carried out without hemorrhage into the muscle pocket, and care must be taken not to impale the graft tissueeither will result in graft failure.

PARATHYROID CARCINOMA
Hyperparathyroidism caused by parathyroid carcinoma is rare; the reported incidence is from 0.5% to 4% (see Chap. 48 and Chap. 58). This diagnosis should be suspected when (a) a palpable neck mass is present, (b) the serum calcium level exceeds 14 mg/dL, (c) the serum PTH level is markedly elevated, and (d) a previously unoperated patient is hoarse from a recurrent laryngeal nerve invasion. At surgery, parathyroid carcinomas appear white and very firm, unlike adenomas, which are reddish brown and soft. The initial operation must be aggressive yet meticulous, with en bloc resection of the parathyroid tumor and all adjacent invaded tissues, including the ipsilateral thyroid lobe. Care must be taken not to rupture the capsule and spill the tumor. Radical neck dissection is reserved for patients with clinically overt cervical node metastases. Distinguishing early parathyroid carcinoma from atypical parathyroid adenoma by histologic and clinical criteria may be difficult. In these cases, DNA cytometry can be helpful, because an aneuploid pattern and higher nuclear DNA content are typical of carcinomas but not of adenomas. For recurrent disease, repeated resection of local cervical implants and distant metastases is important in palliative management. Tumor recurrence is generally apparent within 6 months to 3 years of surgery and may denote an incurable process. However, select patients appear to benefit greatly from re-resection of locally recurrent tumors. CHAPTER REFERENCES
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. Duh Q-Y, Uden P, Clark OH. Unilateral neck exploration for primary hyperparathyroidism: analysis of a controversy using a mathematical model. World J Surg 1992; 16:654. Brasier A, Wang C-A, Nussbaum S. Recovery of parathyroid hormone secretion after parathyroid adenectomy. J Clin Endocrinol Metab 1988; 66:495. Wells SA, Leight GS, Hensley M, Dilley WG. Hyperparathyroidism associated with the enlargement of two or three parathyroid glands. Ann Surg 1985; 202:533. Cope O, Keynes WM, Roth SI, Castleman B. Primary chief cell hyperplasia of the parathyroid glands: a new entity in the surgery of hyperparathyroidism. Ann Surg 1958; 148:375. Rudberg C, Akerstrm G, Palmer M, et al. Late results of operation for primary hyperparathyroidism in 441 patients. Surgery 1986; 99:643. Edis AJ, van Heerden JA, Scholz DA. Results of subtotal parathyroidectomy for primary chief cell hyperplasia. Surgery 1979; 86:462. Bonjer HJ, Bruining HA, Birkenhager JC, et al. Single and multigland disease in primary hyperparathyroidism: clinical follow-up, histopathology, and flow cytometric DNA analysis. World J Surg 1992; 16:737. Proye C, Carnaille B, Quievreux J-L, et al. Late outcome of 304 consecutive patients with multiple gland enlargement in primary hyperparathyroidism treated by conservative surgery. World J Surg 1998; 22:526. Rizzoli R, Green J III, Marx SJ. Primary hyperparathyroidism in familial multiple endocrine neoplasia type I: long-term follow-up of serum calcium levels after parathyroidectomy. Am J Med 1985; 78:467. Kraimps J, Duh Q-Y, Demeure M, Clark O. Hyperparathyroidism in multiple endocrine neoplasia syndrome. Surgery 1992; 112:1080. Hellman P, Skogseid B, Oberg K, et al. Primary and reoperative parathyroid operations in hyperparathyroidism of multiple endocrine neoplasia type 1. Surgery 1998; 124:993. Wells SA Jr, Farndon JR, Dale JK, et al. Long-term evaluation of patients with primary parathyroid hyperplasia managed by total parathyroidectomy and heterotopic autotransplantation. Ann Surg 1980; 192:451. Consensus Development Conference Panel. Diagnosis and management of asymptomatic primary hyperparathyroidism: Consensus Development Conference statement. Ann Intern Med 1991; 114:593. Carty S, Worsey M, Virji M, et al. Concise parathyroidectomy: the impact of preoperative SPECT 99m Tc sestamibi scanning and intraoperative quick parathormone assay. Surgery 1997; 122:1107. Boggs JE, Irvin GL 3rd, Molinari AS, Deriso GT. Intraoperative parathyroid hormone monitoring as an adjunct to parathyroidectomy. Surgery 1996; 120:954. Norman J, Chheda H, Farrell C. Minimally invasive parathyroidectomy for primary hyperparathyroidism: decreasing operative time and potential complications while improving cosmetic results. Ann Surg 1998; 64:391. Miccoli P, Bendinelli C, Vignali E, et al. Endoscopic parathyroidectomy: report of an initial experience. Surgery 1998; 124:1077. Miller DL, Doppman JL, Shawker TH, et al. Localization of parathyroid adenomas in patients who have undergone surgery. Part I. Noninvasive imaging methods. Radiology 1987; 162:133. Levin KE, Gooding GAW, Okerlund M, et al. Localizing studies in patients with persistent or recurrent hyperparathyroidism. Surgery 1987; 102:917. Wei J, Burke G, Munsberger A. Prospective evaluation of the efficacy of technetium 99m sestamibi and iodine 123 radionuclide imaging of abnormal parathyroid glands. Surgery 1992; 112:1111.

20a. Neumann DR, Esselstyn CB Jr, Madera AM. Sestamibi/iodine subtraction single photon emission computed tomography in reoperative secondary hyperparathyroidism. Surgery 2000; 128:22. 21. Miller DL, Doppman JL, Krudy AG, et al. Localization of parathyroid adenomas in patients who have undergone surgery. Part II. Invasive procedures. Radiology 1987; 162:138. 22. Doherty GM, Doppman JL, Miller DL, et al. Results of a multidisciplinary strategy for management of mediastinal parathyroid adenoma as a cause of persistent primary hyperparathyroidism. Ann Surg 1992; 215:101. 23. Wells SA Jr, Cooper JD. Closed mediastinal exploration in patients with persistent hyperparathyroidism. Ann Surg 1991; 214:555. 24. Wells SA, Gunnells JC, Gutman RA, et al. The successful transplantation of frozen parathyroid tissue in man. Surgery 1977; 81:86. 25. Stanbury S, Lumb G, Nicholson W. Selective subtotal parathyroidectomy for renal hyperparathyroidism. Lancet 1960; 1:793. 26. Cordell LJ, Maxwell JG, Warden GD. Parathyroidectomy in chronic renal failure. Am J Surg 1979; 138:951. 27. Dubost C, Drueke T, Jeaneau P, et al. Hyperparathyroidie secondaire: para-thyroidectomie subtotale ou totale avec autotransplantation parathyroidienne. Nouvelle Presse Mdicale 1980; 9:2709. 28. Niederle B, Roka R, Brennan MF. The transplantation of parathyroid tissue in man: development, indications, techniques, and results. Endocr Rev 1982; 3:345. 29. Rothmund M, Wagner PK, Schark C. Subtotal parathyroidectomy versus total parathyroidectomy and autotransplantation in secondary hyperparathyroidism: a randomized trial. World J Surg 1991; 15:745. 30. Rothmund M, Wagner P. Reoperations for persistent and recurrent secondary hyperparathyroidism. Ann Surg 1988; 207:310. 31. D'Avanzo A, Parangi S, Morita E, et al. Hyperparathyroidism after thyroid surgery and autotransplantation of histologically normal parathyroid glands. J Am Coll Surg 2000; 190:546.

CHAPTER 63 OSTEOMALACIA AND RICKETS Principles and Practice of Endocrinology and Metabolism

CHAPTER 63 OSTEOMALACIA AND RICKETS

NORMAN H. BELL Definition Pathogenesis Clinical Presentation Radiographic Findings Types of Osteomalacia Nutritional Deficiency Vitamin D-25-Hydroxylase Deficiency Gastrointestinal and Hepatic Diseases Hypoparathyroidism Pseudohypoparathyroidism Chronic Renal Insufficiency Vitamin D1a-Hydroxylase Deficiency Hypophosphatemic Rickets Tumor-Induced Osteomalacia Hereditary Hypophosphatemic Rickets with Hypercalciuria Vitamin DDependent Rickets Type II Anticonvulsant-Induced Osteomalacia Renal Tubular Acidosis Mineralization Defects Hypophosphatasia Treatment Chapter References

DEFINITION
Osteomalacia is a disorder of mineralization of newly formed organic matrix that occurs in adults. Rickets is a disease of children in which abnormal calcification of matrix also occurs. Further, defective mineralization of cartilage takes place in the epiphyseal cartilage growth plate so that disorganization of cellular development ensues, which leads to widening of the ends of long bones and, possibly, retardation of growth and skeletal deformities.

PATHOGENESIS
For mineralization to take place normally, newly formed osteoid must be normal both qualitatively and quantitatively, the concentrations of calcium and phosphate in extracellular fluid must be sufficient, the activity of alkaline phosphatase must be adequate, the pH at the site of calcification must be optimal, and excess concentration of inhibitors of calcification must be prevented. Bone formation takes place in two steps. The organic matrix or osteoid derived from osteoblasts, or bone-forming cells, is laid down and then undergoes a process of maturation that requires 10 to 15 days before mineralization takes place.1 Whereas osteomalacia could result from abnormal structure of organic matrix so that calcification cannot take place normally, it much more frequently is caused by alterations in mineral metabolism in which reduced serum concentrations of calcium or phosphate, or both, occur.2 Normal mineralization is incompletely understood (see Chap. 50). Calcium and phosphate in extracellular fluid are supersaturated, and metastatic calcification is prevented by pyrophosphate and possibly other substances that include peptides. Bone mineral is first deposited as amorphous calcium phosphate that eventually undergoes conversion to hydroxyapatite [Ca10(PO4)6(OH)2]. Calcium in extracellular fluid space is taken up by mitochondria in osteoblasts and transported to matrix vesicles for mineralization. The phosphate that is used in mineralization is made available by alkaline phosphatase that is present in the matrix vesicles. Deposition and eventual conversion of amorphous calcium phosphate to hydroxyapatite occur at gaps, hole zones, between the distal ends of two molecules of collagen. The rate of bone formation and calcification can be measured accurately by histomorphometric techniques, which require double tetracycline labeling (see Chap. 55). Tetracyclines are deposited in the skeleton at the mineralization front in a band-like pattern. They fluoresce and are easily visualized on histologic sections under a fluorescence microscope. After two brief courses of administration, with an intervening interval, the appositional growth rate of the skeleton can be estimated in biopsy samples of the iliac crest by determining the distance between the two fluorescent bands. In normal adults, this distance averages ~1 m per day.3 In osteomalacia, the distance is reduced. In some patients with osteomalacia, mineralization of the skeleton is so abnormal that only a single ill-defined band of fluorescence can be discerned. Newly synthesized matrix that is abnormally mineralized appears as a wide osteoid seam. The two cardinal histomorphometric features of osteomalacia are the wide osteoid seam and the reduced calcification rate; these two findings are essential for the diagnosis, because wide osteoid seams occur without abnormal mineralization in bone disease associated with hyperthyroidism, Paget disease, and primary hyperparathyroidism.4

CLINICAL PRESENTATION
Sometimes patients with osteomalacia have no symptoms, so that the diagnosis is not readily apparent early in the course of the disease. When present, symptoms include diffuse skeletal pain and muscle weakness. The pain often is described as dull and aching and is worsened by activity. It is present in the lower back and hips or at sites where fractures have taken place. Bone tenderness may be present on palpation. Fractures most commonly occur in the ribs, vertebrae, and long bones, and may lead to skeletal deformities. Muscle weakness, often present in patients with osteomalacia, may be severe and associated with wasting.5 Because weakness usually involves the proximal muscle groups, particularly of the lower extremities, it may contribute to the characteristic waddling gait. Numerous factors appear to play a role in the development of the muscle weakness associated with osteomalacia. Whereas myopathic muscle changes sometimes are evident by electromyography, denervation rarely is seen on muscle biopsy.6 However, neurogenic atrophy or type II muscle fiber atrophy occasionally is observed.7 Hypophosphatemia itself can cause muscle weakness; therefore, the weakness can be improved simply by correction of the hypophosphatemia. Secondary hyperparathyroidism may be important in the pathogenesis of the neuromuscular disease in osteomalacia, because clinical and laboratory findings similar to those that occur in patients with osteomalacia occasionally are found in patients with primary hyperparathyroidism.

RADIOGRAPHIC FINDINGS
The most common radiographic change in osteomalacia is reduced skeletal density, a nonspecific finding with little diagnostic value. More helpful are the less frequently occurring coarsening of trabecular pattern and the Looser zones or pseudofractures.8,9 The loss of distinctness of trabeculae in vertebral bodies is attributed to the loss of secondary trabeculae and to inadequate mineralization of osteoid. Because of softening, the vertebral bodies in more advanced disease may become concave, termed codfish vertebrae. Conversely, the vertebral disks are large and biconvex. Compression fractures of the spine occur but are less common than in osteoporosis. Looser zones are narrow lines of radiolucency that usually transect and lie either at right angles or obliquely to the cortical margins of bones. They typically are bilateral and symmetric. Common sites include the axillary margins of the scapulae, the lower ribs, the superior and inferior pubic rami, the inner margins and neck of the proximal femora, and the posterior margins of the proximal ulnae. Multiple, bilateral, and symmetric pseudofractures in a patient with osteomalacia are called Milkman syndrome.10,11 The pathogenesis of Looser zones is thought by some to be stress fractures that are repaired by the laying down of inadequately mineralized osteoid. Because the pseudofractures often lie adjacent to arteries, others attribute the lesions to mechanical erosion caused by arterial pulsations, a concept that is supported by the arteriographic demonstration that arteries frequently overlie sites of the pseudofractures.8,12 True fractures can occur at these weakened areas. Deformities of the skeleton often occur with rickets during childhood, some of which may persist into adulthood.13 Skeletal abnormalities occur at sites of rapid growth. Because the rate of growth of different bones varies with age, the skeletal changes of rickets vary with age and may indicate the age of onset of the disease. Because of rapid growth, the skull is particularly affected in neonates. Softening of the cranium, or cra-niotabes, may be associated with parietal flattening, frontal bossing, and widened sutures. Rapid growth of the arms and rib cage in early childhood is reflected in the widening of the forearm at the wrist, and the thickening of the costochondral junctions that produces the rachitic rosary (see Chap. 70). Indentations of the lower ribs at the site of attachment of the diaphragm are called Harrison grooves. With increased growth of long bones, bowing of the lower extremities may result as a consequence of weight-bearing. Deformities of the pelvis occur and can

produce major problems at the time of weight-bearing. Because of secondary hyperparathyroidism, skeletal changes resulting from increases in circulating parathyroid hormone (PTH) may be present. These include subperiosteal resorption of the phalanges and resorption of the distal ends of long bones such as the clavicle and humerus. In patients with osteomalacia, bone mass, determined by single- and dual-photon absorptiometry, is reduced,14 and bone scans show increased uptake of technetium 99m pyrophosphate by long bones and wrists, and prominence of the calvaria and mandible.15 Less prominent is beading of the costochondral junctions and marked uptake of the tracer by the sternum and its margins, the tie sternum. Moreover, pseudofractures appear as hot spots. 15 Sometimes pseudofractures may be evident only on radiographs and other times only on bone scan. Hot spots may be erroneously diagnosed as metastatic lesions.

TYPES OF OSTEOMALACIA
Osteomalacia can result from abnormalities in vitamin D metabolism, phosphate deficiency, various mineralization defects, and states of rapid bone formation (Table 63-1; see Chap. 70).

TABLE 63-1. Etiology of Osteomalacias

NUTRITIONAL DEFICIENCY Normally, vitamin D3 derived by dermal production from 7-dehydrocholesterol is the major source of the vitamin (see Chap. 54). Because exposure to sunlight is adequate and dairy products are fortified with vitamin D, osteomalacia resulting from nutritional deficiency is rarely seen in the United States. When it does occur, lack of exposure to sunlight usually is responsible. Conversely, nutritional deficiency of vitamin D is more common in other parts of the world. In the United Kingdom, osteomalacia occurs in immigrant Indians and Paki-stanis.16,17 and 18 Ethnic traditions and dietary patterns contribute to the development of osteomalacia in the Asian population in England. Endogenous production of vitamin D3 is limited in women who remain indoors and wear traditional clothing, because of diminished exposure to the sun. Chupatti flour, a dietary staple of these population groups, has a high phytate content in the wheat and binds calcium, causing increased fecal excretion. Lignin, a component of wheat fiber, binds to bile acids, preventing their absorption.19 Vitamin D, which normally forms micelles with bile acids, a requirement for its absorption by the intestine, may be bound by the lignin-bile acid complex instead and not be absorbed. Removal of the chupatti flour from the diet corrects abnormal vitamin D and mineral metabolism. In addition, fortification of the diet with vitamin D can correct the deficiency. In the United States and other countries, rickets occurs in newborns who are breast-fed because the content of vitamin D and 25-hydroxyvitamin D [25(OH)D] in human milk is not adequate.20 In the northern hemisphere, blacks, Pakistanis, and Asian Indian infants are particularly at risk for rickets because maternal serum vitamin D and 25(OH)D values are low as a consequence of increased skin pigmentation and diminished dermal synthesis of vitamin D. Asian Indians and Pakistanis, but not blacks, are at risk for vitamin D deficiency and osteomalacia in later life.21 However, osteomalacia caused by nutritional vitamin D deficiency does not occur in blacks. In developed countries, osteomalacia occurs in the elderly, who appear to be the population at greatest risk for this disorder. This is especially true for individuals who are housebound or institutionalized. An age-related decline in the dermal synthesis of 7-dehydrocholesterol,22 inadequate intake of vitamin D, and impaired production of 25(OH)D in the liver and 1,25-dihydroxyvitamin D [1,25(OH)2D] in the kidney23 also may be contributing factors. VITAMIN D-25-HYDROXYLASE DEFICIENCY Rarely, rickets occurs as a consequence of deficiency of vitamin D-25-hydroxylase.24,25 The disease apparently is transmitted genetically as an autosomal recessive trait. Affected children show evidence of rickets during early childhood. Clinical findings include irritability, seizures, growth retardation, and poor motor development. Hypocalcemia of variable degree, hypophosphatemia, elevated serum alkaline phosphatase and serum immunoreactive PTH (iPTH) values, low serum 25(OH)D values, and normal or elevated serum levels of 1,25 dihydroxyvitamin D [1,25(OH)2D] are found. The diagnosis is made as follows: demonstration of a low serum 25(OH)D value; poor response to physiologic doses of vitamin D; correction of biochemical, clinical, and radiographic changes of rickets after treatment with physiologic doses of 25(OH)D; the lack of hepatic, gastrointestinal, or renal disease; and no history of consumption of anticonvulsants. GASTROINTESTINAL AND HEPATIC DISEASES Intestinal malabsorption associated with diseases of the small intestine, hepatobiliary tree, and pancreas is the most common cause of vitamin D deficiency in the United States.26 Disorders of the small intestine that may cause malabsorption and osteomalacia include celiac disease or sprue, regional enteritis, scle-roderma, multiple jejunal diverticula, and the blind loop syndrome. In some cases, the bone disease is more evident than the gastrointestinal disease. Impaired absorption of calcium as well as vitamin D may contribute to the development of osteomalacia. Although vitamin D undergoes an enterohepatic circulation, intestinal loss of endogenous vitamin D has not been demonstrated to be important in the pathogenesis of osteomalacia in enteric diseases. In adults, osteomalacia is a common complication of intestinal bypass surgery for the treatment of morbid obesity. In these patients, histologic changes of osteomalacia are more common than are radiographic changes.27 Osteomalacia is a complication of partial gastrectomy, especially in association with the Billroth II operation, which excludes the duodenum.28 Deficient intake of vitamin D, postgastrectomy steatorrhea, and diminished exposure to sunlight and calcium absorption are contributing factors. The incidence of postgastrectomy osteomalacia will probably decline with the advent of more effective medical treatment of peptic ulcer disease. Osteomalacia occurs as a complication of hepatocellular biliary disorders29 but is infrequent in pancreatic disorders. Biliary obstruction or parenchymal disease of the liver can diminish the synthesis of 25(OH)D and interfere with the intestinal absorption of vitamin D and calcium. In some patients, the defect in vitamin D-25-hydroxylation may be so great that vitamin D is an ineffective means of treatment and 25(OH)D must be administered. HYPOPARATHYROIDISM Osteomalacia only rarely occurs in patients with hypoparathyroidism (see Chap. 60). Bone pain suggests the diagnosis, and in some instances, because radiographs of the skeleton may be unremarkable, the diagnosis can be made only by histomorphometric analysis of bone biopsy material.30 Hypocalcemia and low or low-normal serum 1,25(OH)2D usually are present and appear to be important in the pathogenesis of the bone disease. Because PTH is the major regulator of the renal synthesis of 1,25(OH)2D, the low serum 1,25(OH)2D value and hypocalcemia result from its deficiency. Some patients can be treated effectively with vitamin D, whereas others require treatment with 1,25(OH)2D.30 PSEUDOHYPOPARATHYROIDISM In pseudohypoparathyroidism (see Chap. 60), resistance to PTH results in hypocalcemia, retention of phosphate, and low serum 1,25(OH)2D values.30,31 and 32 Patients

with pseudohypoparathyroidism rarely have osteomalacia.30 Hypocalcemia and a low serum 1,25(OH)2D value are important factors in the pathogenesis of the bone disease.30 As in hypoparathyroidism, some patients can be treated with vitamin D and others with 1,25(OH)D.30 CHRONIC RENAL INSUFFICIENCY Sometimes osteomalacia occurs in patients with chronic renal failure (see Chap. 61). Several factors may contribute. These include low serum 1,25(OH)2D value resulting from loss of renal tissue and the enzyme 25(OH)D1a-hydroxylase,33 retention of inhibitors of calcification, metabolic acidosis, and formation of abnormal collagen matrix. More commonly, skeletal changes of osteitis fibrosa cystica occur as a consequence of secondary hyperparathyroidism, sometimes in association with osteomalacia.34 Subperiosteal resorption may be present on the medial aspect of the middle phalanges. Resorption also may be evident at other sites, including the medial margins of the femoral necks and the inner aspects of the proximal tibiae. Osteomalacia may develop as a consequence of excessive intake of aluminum-containing phosphate binders,35 as a result of dialysis-induced phosphate depletion36 or renal phosphate wasting after renal transplantation.37 Osteomalacia may be caused by aluminum derived from water that is used for hemodialysis or from aluminum hydroxide gels that are used to bind phosphate (see Chap. 61 and Chap. 131). Aluminum is deposited in bone at the mineralization front and may impair calcification and cause osteomalacia.38 In aluminum-induced osteomalacia, serum iPTH is normal or minimally elevated, and the bone disease does not respond to treatment with vitamin D or its analogs. Long-term treatment with the chelating agent deferoxamine is effective in removing aluminum from bone.39 The osteomalacia can then be treated by calcium together with either 25(OH)D or 1,25(OH)2D. VITAMIN D1-a-HYDROXYLASE DEFICIENCY Vitamin D1a-hydroxylase deficiency, or vitamin Ddependent rickets type I, is an inborn error of vitamin D metabolism that is transmitted genetically as an autosomal recessive trait.40,41 Affected children appear to be normal at birth and show evidence of the clinical and biochemical changes of rickets during the first year of life. Hypocalcemia is a consistent finding, together with hypophosphatemia and elevated serum iPTH and alkaline phosphatase values. The diagnosis is made by demonstration of a normal or elevated serum 25(OH)D level and normal or low serum 1,25(OH)2D level in association with a low serum calcium level. Some patients are treated successfully with large doses of vitamin D or 25(OH)D but readily respond to physiologic doses of 1,25(OH)2D.40,41 When 25(OH)D1a-hydroxylase, which had been cloned from human keratinocytes, was sequenced, it was found to consist of 508 amino acids with an N-terminal mitochondrial signal sequence and a heme-binding region.42 One patient had deletion/frameshift mutations at codons 211 and 231 that produced a stop codon after amino acid 233; this resulted in a severely truncated protein, which would not bind heme or express 25(OH)D1a-hydroxylase activity. More recent studies have demonstrated mutations at amino acids 65, 189, 212, 241, 389, 409, 429, 438, 453, 497, 958, 1921, and 1984 (the latter three outside the coding region), which are associated with diminished bioactivity (Fig. 63-1).43 The gene contains duplicated 7 base-pair (bp) sequences encoding residues 438 and 442. In a number of patients, a third 7 bp copy was found that altered the downstream reading frame and created a premature TGA stop signal.43 Deletion of glycine 958 is the most common mutation. The disease in these patients is attributed to two autosomal recessive mutations, each inherited from one of the parents.42,43

FIGURE 63-1. Mutations of the vitamin D1 -hydroxylase gene that account for the disease vitamin D 1a-hydroxylase deficiency. 7 bp refers to a third copy of a normally duplicated 7-bp sequence that causes a premature TGA stop signal. is deleted. Note that three mutations are outside the coding region. X is a stop codon.

HYPOPHOSPHATEMIC RICKETS Hypophosphatemic rickets or osteomalacia is characterized by hypophosphatemia and renal wasting of phosphate.44,45 The primary defect is one of impaired reabsorption of filtered phosphate by the proximal renal tubule. Two separate mechanisms exist for phosphate transport in the renal tubule: a PTH-dependent component that is responsible for two-thirds of the net tubular reabsorption, and a PTH-independent component that is regulated by the serum calcium and is responsible for the remaining one-third of the net reabsorption.46 The PTH-dependent component is lacking in men and is partially deficient in women with hypophosphatemic osteomalacia. Defective phosphate transport also is present in the intestinal mucosa, suggesting that the deficit is generalized and not restricted to the kidney.47 Hypophosphatemic rickets usually is familial, but occasionally is sporadic. It is transmitted as an X-linked dominant trait. Onset occurs between 1 and 1.5 years of age and is associated with delayed or abnormal dentition. Screening studies in families indicate that fasting hypophosphatemia is the most common manifestation of the disorder. Serum calcium and serum iPTH levels usually are normal. The clinical spectrum is broad, varying from patients who have hypophosphatemia and no apparent bone disease (usually girls) to patients who have symptoms and severe bone disease.39,40 Patients with the disorder are short and stocky and have bowed legs, particularly when treatment has been inadequate.47a The degree of hypophosphatemia and the severity of the bone disease are not correlated. However, diminished growth rate is attributed to hypophosphatemia. Defective phosphate transport by the kidney and intestine and the production of a phosphaturic substance have been implicated in the pathophysiology of the disorder. However, crosstransplantation experiments in the X-linked hypophosphatemic (Hyp) mouse, a model for the human disease, demonstrated that the Hyp-mouse phenotype is not corrected or transferred by renal transplantation.48 These findings support the likelihood of a humoral factor rather than an intrinsic renal or intestinal abnormality in the etiology of hypophosphatemic rickets. A type II sodium/phosphate cotransporter in the brush border membranes is the rate-limiting, physiologic mediator of proximal renal tubular phosphate reabsorption. Expression of sodium/phosphate cotransporter mRNA and protein is reduced in proximal renal tubular brush border membranes of Hyp mice with impaired renal tubular reabsorption of phosphate.49 The renal production of 1,25(OH)2D is impaired in patients with hypophosphatemic osteomalacia. Serum values are inappropriately low for the degree of hypophosphatemia and respond poorly to administration of PTH.50 The defect appears to be coupled to the abnormality in renal phosphate transport. Plasma 1,25(OH)2D values vary directly with the renal tubular maximum of phosphate glomerular filtration rate (TmP/GFR) in patients with X-linked hypophosphatemia, normal individuals, and patients with other defects in phosphate transport, including tumor-induced osteomalacia (Fig. 63-2).51 Although the renal production of 1,25(OH)2D and the renal tubular reabsorption of phosphate probably are linked, treatment with 1,25(OH)2D does not reverse the renal phosphate wasting, and normalization of the serum phosphate value does not enhance the renal production of 1,25(OH)2D. Interestingly, serum 1,25(OH)2D and TmP/GFR are elevated in patients with tumoral calcinosis, a disease that appears to be the mirror image of X-linked hypophosphatemia and tumor-induced osteomalacia in which the renal tubular reabsorption of phosphate is abnormally increased.

FIGURE 63-2. Relation between plasma 1,25 dihydroxyvitamin D [1,25(OH)2D] and renal tubular maximum of phosphate glomerular filtration rate (TmP/GFR) in normal subjects and in patients with hypophosphatemic rickets (XLH), tumor-induced osteomalacia (TIO), and tumoral calcinosis. Note that values tend to be low in XLH and TIO, and high in tumoral calcinosis. Also note the highly significant positive correlation between plasma 1,25(OH)2D and renal TmP/GFR. These results point to the possibility that renal tubular reabsorption of phosphate may be a major determinant of the renal production of 1,25(OH)2D. (From Drezner MK. Understanding the pathogenesis of X-linked hypophosphatemic rickets: a requisite for successful therapy. In: Zackson DA, ed. A CPC series: cases in metabolic bone disease. New York: Triclinica Publishing, 1987:1.)

Mutations in the PEX gene, a phosphate-regulating endopeptidase, have been shown to be the cause of X-linked phosphatemic rickets in patients and in the Hyp mouse.52 It has been proposed that the PEX gene product acts to increase a phosphaturic factor, phosphatonin, that inhibits the renal tubular reabsorption of phosphate. Preliminary studies with cultured osteoblasts from normal and Hyp mice have shown that PEX and phosphatonin production are maturationally dependent and that aberrant degradation of phosphatonin by PEX may be involved in the pathogenesis of the disease.53 The treatment of hypophosphatemic osteomalacia is lifelong administration of phosphate and 1,25(OH)2D. The bone disease can be completely reversed with adequate therapy.52 However, sometimes hypercalcemia and nephrocalcinosis are complications. TUMOR-INDUCED OSTEOMALACIA In tumor-induced osteomalacia (oncogenous osteomalacia), renal phosphate wasting and osteomalacia occur in association with a variety of benign or malignant neoplasms and disappear with removal or irradiation of the tumors.54 Lesions responsible for the disease often are of mesenchymal origin, are vascular, show foci of new bone formation, and include sarcomas, hemangiomas, and giant cell tumors of bone as well as carcinoma of the breast and prostate. Osteomalacia associated with fibrous dysplasia of bone, neurofibromatosis,55 and linear nevus sebaceous syndrome may be associated with tumors as well.54 The clinical and biochemical features are typical of osteomalacia; patients often have generalized muscle weakness and bone pain. Serum calcium values usually are normal to slightly reduced, serum phosphate values are low, and serum alkaline phosphatase and urinary calcium values are elevated. Serum 1,25(OH)2D values and calcium absorption may be low, and contribute to the pathogenesis of the bone disease. The hypophosphatemia probably results from one or more factors produced by the tumors that alter the renal tubular reabsorption of phosphate in the proximal tubule and interfere with the renal production of 1,25(OH) 2D.55a Evidence for this is the demonstration of phosphaturic activity in tumor extracts, the occurrence of renal phosphate wasting and diminished 25(OH)D1a-hydroxylase activity in tumor-bearing athymic nude mice, and reduced activity of the enzyme in kidney cell cultures exposed to extracts of tumors.52,54 The pathophysiology of the disease may be similar to that of X-linked hypophosphatemic osteomalacia.54 Increased expression of normal PEX mRNA and protein were found in tumors of two patients with tumor-induced osteomalacia.56 Thus, mutations of the PEX gene did not account for the disease in these two patients. These findings illustrate the complexity of the relationship between PEX and hypophosphatemia. Treatment of tumor-induced osteomalacia is resection or irradiation of the lesion. If this is not possible, treatment with phosphate and 1,25(OH)2D may be used (see Chap. 219). HEREDITARY HYPOPHOSPHATEMIC RICKETS WITH HYPERCALCIURIA Hereditary hypophosphatemic rickets with hypercalciuria is a rare autosomal recessive disorder characterized by hypophosphatemic rickets, increased serum 1,25(OH)2D, normocalcemia, suppression of serum PTH and urinary cyclic adenosine 3',5'-monophosphate, and hypercalciuria.57 Onset occurs during infancy and childhood. Clinical findings include muscle weakness, short stature, disproportionately short lower limbs, genu varum or valgum, anterior external bowing of the femur, coxa vara, radiologic signs of rickets, and diminished bone density. Relatives with mild hypophosphatemia and no bone disease are considered to have a milder form of the disease.58 The disorder is attributed to a primary defect in the renal tubular reabsorption of phosphate that results in hypophosphatemia and stimulation of renal 25(OH)D1a-hydroxylase, increases in serum 1,25(OH)2D, intestinal absorption of cal- cium, inhibition of secretion of PTH, and hypercalciuria. VITAMIN DDEPENDENT RICKETS TYPE II Vitamin Ddependent rickets type II results from resistance of target organs to 1,25(OH)2D59,60 and is usually caused by mutations of the vitamin D receptor. The development of the bone disease can occur at any time from infancy to adolescence. It is rare, and the occurrence is sporadic or familial. Sometimes it results from a consanguineous marriage and is transmitted as an autosomal recessive trait. Infants with the familial disease may have permanent alopecia, which usually is a sign of increased severity. In these patients, serum 1,25(OH)2D values may be as high as 700 pg/mL. Studies of cultured skin fibroblasts from patients with the disorder demonstrate a wide variety of abnormalities in the uptake and nuclear binding of 3H-labeled 1,25(OH)2D. These findings indicate considerable genetic heterogeneity in the disorder.60 Five different types of intracellular defects have been identified: (a) hormone-binding negative, (b) defect in hormone-binding capacity, (c) defect in hormone-binding affinity, (d) deficient nuclear localization, and (e) normal or near-normal binding of 1,25(OH)2D to the vitamin D receptor but diminished binding of the hormone receptor to DNA.60 As noted later, subsequent studies indicate that the heterogeneity is caused by genetic mutations at different sites of the vitamin D receptor.61,62 The 1,25(OH)2D binds to specific intracellular receptors in target cells, and the receptorhormone complex acts in trans by binding to specific cis-acting DNA sequences in the promoter region of hormone-responsive genes to modulate transcription. The gene for the human vitamin D receptor has been cloned and sequenced.63 The receptor consists of a hydrophilic DNA-binding domain that includes two zinc fingers, a hinge, and a hydrophobic hormone-binding domain. As shown in Figure 63-3, 19 sites of mutations in the receptor have been described to date: nine in the DNA-binding domain, nine in the hormone-binding domain, and one in the hinge region. Single-point mutations at sites 30, 33, 35, 70, and 73 that include one of the two zinc fingers, or at sites 44, 45, 46, 50, and 80 lying between or near the two zinc fingers in the DNA-binding domain, result in a vitamin D receptor of normal size with normal binding to 1,25(OH)2D and diminished binding of the 1,25(OH)2D receptor complex to DNA (type V). Stop codons at sites 30, 73, 152, 259, and 295 result in a truncated vitamin D receptor that is unable to bind 1,25(OH)2D or promote gene expression (type I). Finally, vitamin Ddependent rickets type II can occur in patients with a normal vitamin D receptor, which has been found to be caused by inadequate calcium intake.64,64a

FIGURE 63-3. Sites of mutations reported for the vitamin D receptor gene in patients with vitamin D dependent rickets type II.

In general, patients with type III (deficient affinity) and type IV (deficient nuclear localization) defects respond to high doses of vitamin D or its metabolites with complete remission, whereas patients with the type II (low-capacity) defect and most patients with type I (receptor-negative) and type V (receptor-positive) defects respond either poorly or not at all to such treatment.60 However, these patients can be treated with prolonged administration of intravenous calcium.65 ANTICONVULSANT-INDUCED OSTEOMALACIA Osteomalacia occasionally occurs in patients with epilepsy who are receiving anticonvulsant therapy, especially phenobarbital and phenytoin.66,67 Other anticonvulsants also are implicated.67a Whereas the incidence of bone disease is low, hypocalcemia, diminished intestinal absorption of calcium, and increased serum alkaline phosphatase are more common. The clinical spectrum ranges from patients with reduced bone mass who have no symptoms to patients with hypocalcemia, clinically apparent bone disease, fractures, and pseudofractures. The administration of more than one anticonvulsant drug increases the incidence of bone disease. Anticonvulsant drugs alter vitamin D metabolism. This alteration includes increased hepatic conversion of vitamin D to 25(OH)D and increased hepatic conversion of vitamin D and 25(OH)D to more polar biologically inactive metabolites.68 As a consequence, serum 25(OH)D values are reduced. Serum 1,25(OH)2D values are normal; however, the bone disease may not be caused by abnormal vitamin D metabolism. In animals, phenytoin inhibits the intestinal absorption of calcium, and both phenytoin and phenobarbital inhibit the mobilization of calcium from bone in vitro. Thus, the drugs may cause bone disease by blocking the actions of 1,25(OH)2D in target organs. The administration of vitamin D corrects and prevents the biochemical and radiographic abnormalities of osteomalacia and decreases the incidence of fractures.69 RENAL TUBULAR ACIDOSIS Osteomalacia occurs in renal tubular acidosis70 and in the acidosis that occurs after ureterosigmoidoscopy.71 Chronic acidosis may cause bone disease by decreasing the conversion of amorphous calcium phosphate to hydroxyapatite at the mineralization front and by causing renal phosphate wasting. Systemic acidosis enhances the dissolution of bone and results in hypercalciuria.11 Acidosis enhances the response of bone to PTH in vitro and may contribute to skeletal loss of calcium. In humans, systemic acidosis does not compromise the metabolism of vitamin D, that is, the renal production of 1,25(OH)2D. In patients with renal tubular acidosis and osteomalacia, serum calcium values are normal, serum phosphate values are normal or reduced, and serum alkaline phosphatase values are elevated. Because of hypercalciuria, nephrocalcinosis and renal lithiasis occur. Some patients have salt wasting, so that secondary aldosteronism and hypokalemia develop. Under these circumstances, muscle weakness is profound, and an erroneous diagnosis of poliomyelitis may be made. Osteomalacia from chronic acidosis can be treated successfully with bicarbonate.70 Patients with hypokalemia can be treated with potassium. Vitamin D and calcium may be given at the outset to hasten healing of the bone disease but usually are not required for long-term treatment. MINERALIZATION DEFECTS Bisphosphonates are compounds that have the structure P-C-P and are analogs of pyrophosphate P-O-P. The drugs are taken up selectively by skeletal tissue. In the case of etidronate, when it is administered in high doses, it may inhibit mineralization, causing osteomalacia.72,73 Aluminum has been implicated in the development of the osteomalacia and fractures that occur in patients who are receiving total parenteral nutrition and amino acids in the form of casein hydrolysate.74 Aluminum occurs in high concentrations in casein hydrolysate and also in the osseous tissue of these patients. In patients who have osteomalacia while receiving treatment with total parenteral nutrition, serum iPTH values are either abnormally low or in the lower range of normal, and the rate of bone formation is reduced. The pathogenesis apparently is similar to that of the aluminum-related osteomalacia that occurs in patients with chronic renal failure. The bone disease improves when the use of aluminum-containing preparations is avoided in patients who are receiving total parenteral nutrition.75 HYPOPHOSPHATASIA Hypophosphatasia is a rare disease characterized by abnormally low alkaline phosphatase values in the serum and skeleton, abnormal mineralization of the skeleton with rickets and osteomalacia, increased phosphoethanolamine and pyrophosphate values in the blood and urine, and premature loss of deciduous teeth. It is transmitted as an autosomal recessive trait and is prevalent in inbred populations. The disorder is usually caused by missense mutations of the tissue nonspecific alkaline phosphatase gene, particularly in severe forms of the disease.76 Mutations that have been reported are outlined in Figure 63-4. Many patients are compound heterozygotes and inherit one or more mutations from each parent. Mutations occur along almost the entire length of the alkaline phosphatase gene.

FIGURE 63-4. Mutations of tissue nonspecific alkaline phosphate that account for abnormal bone and mineral metabolim in hypophosphatasia.

Three types of hypophosphatasia occur: infantile, childhood, and adult.77 Whereas in most patients the disorder is diagnosed in infancy or childhood, in others it is not recognized until adulthood. Neonatal or infantile hypophosphatasia is most common, and the diagnosis is evident before the age of 6 months and sometimes even in utero. Impaired mineralization of the skeleton, increased intracranial pressure, hypercalcemia, hypercalciuria, and nephrocalcinosis generally are present. The skeletal disease is so severe that <50% of infants survive. Childhood hypophosphatasia is associated with premature loss of deciduous teeth, increased susceptibility to infection, and retarded growth. A generalized deossification with a coarse trabecular pattern is present on radiographs, together with bowing deformities and fractures. Craniosynostosis is common. Irregular

epiphyses and islands of radiolucency are present in the shafts of bones. Spontaneous healing usually occurs. Adult hypophosphatasia is rare, and patients may have fractures that are delayed in healing. They also may have histories of rickets and loss of deciduous teeth in childhood, and loss of density of the skeleton. Radiographs show a coarse trabecular pattern, Looser zones and subperiosteal bone formation. The diagnosis is established by the findings of low serum alkaline phosphatase values and elevated serum and urine phosphoethanolamine values. No effective treatment exists. Vitamin D administration produces improvement in some patients, but the treatment may cause vitamin D intoxication. Treatment with phosphate improves mineralization of the skeleton in some instances when given in daily doses of 1.25 to 3.0 g as neutral phosphate.

TREATMENT
In patients with osteomalacia, the goals of treatment are (a) to correct hypocalcemia and the related symptoms and to prevent the consequences of hypocalcemia, including seizures and cataracts; (b) to correct and prevent skeletal deformities and changes resulting from secondary hyperparathyroidism; and (c) to prevent hypercalcemia, hypercalciuria, and renal complications, including stone formation, nephrocalcinosis, and renal damage. Vitamin D2 and several of its derivatives are available in the United States. Vitamin D is marketed in 50,000-IU (1.25-mg) capsules for oral administration, in sesame oil (500,000 IU or 12.5 mg/mL) for injection, and in propylene glycol (250 IU or 6.25 g/drop; 8000 IU/mL) for oral administration. The advantages of vitamin D are that the cost is modest and it often is effective even in patients with abnormal vitamin D metabolism. The disadvantages of vitamin D are that several weeks may be required before optimal therapeutic effectiveness is achieved, the therapeutic dose is near the toxic dose, and the bioactivity persists after its administration is discontinued. The metabolite 25(OH)D3 is available in capsules of 20 and 50g. The drug may be particularly useful in patients with hepatic disease and impaired synthesis of 25(OH)D3. Its onset of action is more rapid than that of vitamin D, and it has similar disadvantages. However, the half-life of 25(OH)D3, 2 to 3 weeks, may be shorter than that of vitamin D2, which is stored in fat, so that the biologic effects after cessation of administration may not be as long-lasting as those of vitamin D2. The metabolite 1,25(OH)2D3 is marketed as capsules of 0.25 and 0.50 g. The advantages of the drug are its rapid onset of action and rapid disappearance of its biologic effects after discontinuation. Its half-life is <6 hours. One disadvantage is that hypercalcemia may occur after long-term treatment during which the abnormal calcium metabolism has been stabilized.78 The hypercalcemia can be treated by discontinuing administration of the drug and is prevented by decreasing the dose. Hypercalcemia occurs fairly frequently, so patients must be followed up closely. Treatment with 1,25(OH)2D3 is most useful in diseases in which its synthesis by the kidney is impaired. Sometimes it is of value in disorders that have resistance to its effects at the cellular level. In these cases, higher doses are required. Dihydrotachysterol is available as tablets of 0.125, 0.2, and 0.4 mg. The drug has a rapid onset of action and a relatively short duration of biologic action after cessation of administration. The 180-degree rotation of the A ring permits the hydroxyl group in the 3 position to act as a pseudohydroxyl group. The drug becomes bioactive after it undergoes 25-hydroxylation in the liver. Because it does not need to be hydroxylated in the 1a position, dihydrotachysterol is of potential value in treating the same diseases for which 1,25(OH)2D3 is used. Calcium supplements should be administered in divided doses, because calcium absorption is abnormally low in most patients with osteomalacia. The administration of calcium decreases the amount of vitamin D or its analogs that is required for treatment. The content of elemental calcium varies among preparations, so the amount that provides 1 g of elemental calcium also varies (calcium carbonate = 40% elemental calcium; calcium chloride = 36%; calcium lactate = 13%; and calcium glu-conate = 9%). From 1 to 2 g per day of elemental calcium should be administered to adults in divided doses. The dosage should be modified depending on the severity of disease. Deficiency of vitamin D is treated by the administration of 5000 IU (125 g) to 10,000 IU (250 g) per day of vitamin D2 until healing of the bone disease occurs. The dosage then is decreased to 400 IU (10 g) per day, the recommended daily requirement. Larger doses of 50,000 to 100,000 IU (1250 to 2500 g) per day may be required in patients with gastrointestinal or hepatic disease. Treatment with 25(OH)D3 is indicated when the hepatic vitamin D-25-hydroxylase is markedly impaired. Vitamin D2, 50,000 to 100,000 IU (1250 to 2500 g) per day or more, is often effective in treating osteomalacia resulting from diseases that are associated with renal insufficiency, hypoparathyroidism, and pseudohypoparathyroidism, but it may not be effective in the treatment of bone disease associated with vitamin Ddependent rickets type I, hypophosphatemic rickets, and tumor-induced osteomalacia. In these disorders, g per day, usually is effective. The 1,25(OH)2D3, 0.5 to 3.0 requirements of vitamin D, 25(OH)D3, and 1,25(OH)2D3 vary from patient to patient, so that the dosage for a given individual must be determined by trial and error. Phosphate supplements, 2 to 4 g per day in divided oral doses, are essential for the treatment of hypophosphatemic rickets to heal the bone disease, along with as much as 4 g per day of 1,25(OH)2D3. Lifelong treatment is required. Removal or irradiation of the tumor is necessary to treat tumor-induced osteomalacia. Sometimes osteomalacia associated with vitamin Ddependent rickets type II responds to treatment with vitamin D.60 However, in instances of a more profound target-organ defect, administration of 1,25(OH)2D3 is necessary. Dosages as high as 15 to 20 g per day may be required, and even these may not be effective. Under these circumstances, long-term parenteral administration of calcium can be used to heal the bone disease.65 Osteomalacia associated with renal tubular acidosis is treated initially with 5000 to 10,000 IU (125 to 250 g) per day of vitamin D3. The acidosis is corrected by treatment with sodium bicarbonate. Once the bone disease heals, however, vitamin D is not required. Sometimes hypercalcemia and hypercalciuria occur during treatment with vitamin D and its metabolites. Patients may have no symptoms, or they may have anorexia, nausea, vomiting, weight loss, headache, constipation, polyuria, polydipsia, and altered mental status. The abnormal calcium metabolism is characterized by increased intestinal absorption and enhanced release of calcium from skeletal tissue. A decline in renal function, nephrocalcinosis, nephrolithiasis, urinary tract infections, and even death may ensue. Patients who are receiving long-term treatment with vitamin D and its analogs require careful follow-up at intervals of 4 to 6 weeks with measurement of serum and urinary calcium and serum creatinine. Careful evaluation of patients is important, because no means exist to predict when or in whom vitamin D intoxication will develop (see Chap. 59). The most effective means of treating these side effects and complications is prevention. When intoxication does occur, administration of the drug and of calcium supplements should be discontinued, and fluids, 3 to 4 L per day, should be given. When intoxication is severe, the use of prednisone or salmon calcitonin may be necessary. The dosage of either vitamin D or its derivatives should be reduced, or another drug should be substituted. CHAPTER REFERENCES
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. Frost HM. Tetracycline-based histological analysis of bone remodeling. Calcif Tissue Res 1969; 3:211. Frame B, Parfitt AM. Osteomalacia: current concepts. Ann Intern Med 1978; 89:966. Lee WR. Bone formation in Paget's disease. A quantitative microscopic study using tetracycline markers. J Bone Joint Surg 1967; 49B:146. Parfitt AM. The physiologic and clinical significance of bone histomorpho-metric data. In: Recker RR, ed. Bone histomorphometry: techniques and interpretation. Boca Raton, FL: CRC Press, 1983:143. Schott GD, Wills MR. Muscle weakness in osteomalacia. Lancet 1976; 1:626. Prineas JW, Stuart AM, Henson RA. Myopathy in metabolic bone disease. BMJ 1965; 1:1034. Mallette LE, Patten BM, King WE. Neuromuscular disease in secondary hyperparathyroidism. Ann Intern Med 1975; 82:474. Steinbach HL, Kolb FO, Crane JT. Unusual roentgen manifestations of osteomalacia. AJR Am J Roentgenol 1959; 82:875. Steinbach HL, Noetzli M. Roentgen appearance of the skeleton in osteomalacia and rickets. AJR Am J Roentgenol 1964; 91:955. Milkman LA. Multiple spontaneous idiopathic symmetrical fractures. AJR Am J Roentgenol 1934; 32:622. Albright F, Reifenstein EC Jr. The parathyroid glands and metabolic bone disease. Baltimore: Williams & Wilkins, 1948:1. Goldring SR, Krane SM. Disorders of calcification: osteomalacia and rickets. In: DeGroot LJ, ed. Endocrinology. New York: Grune & Stratton, 1979; 2:853. Park EA. The Blackader lecture on some aspects of rickets. Can Med Assoc J 1932; 26:3. Wahner HA. Single- and dual-photon absorptiometry in osteoporosis and osteomalacia. Semin Nucl Med 1987; 27:305. Fogelman I, McKillop JA, Bessent RG, et al. The role of bone scanning in osteomalacia. J Nucl Med 1978; 19:245. Stamp TCB, Exton-Smith AN, Richens A. Classical rickets and osteomalacia in Britain. Lancet 1976; 2:308. Holmes AM, Enoch BA, Taylor JL, Jones ME. Occult rickets and osteomalacia amongst the Asian immigrant population. QJM 1973; 42:125. Pietrek J, Preece MA, Windo J, et al. Prevention of vitamin-D deficiency in Asians. Lancet 1976; 1:1145. Eastwood MA, Girdwood RH. Lignin: a bile-salt sequestrating agent. Lancet 1968; 2:1170. Mitra D, Bell NH. Racial, geographic, genetic, and body habitus effects on vitamin D metabolism. In: Feldman D, Glorieux FH, Pike JW, eds. Vitamin D. San Diego: Academic Press, 1997:521. Awumey EMK, Mitra DA, Hollis BW, et al. Vitamin D metabolism is altered in Asian Indians in the Southern United States: a Clinical Research Center study. J Clin Endocrinol Metab 1998; 83:169. MacLaughlin J, Holick MF. Aging decreases the capacity of human skin to produce vitamin D 3. J Clin Invest 1985; 76:1536. Gallagher JC, Riggs BL, Esiman JA, et al. Intestinal calcium absorption and serum vitamin D metabolites in normal subjects and osteoporotic patients. J Clin Invest 1979; 64:729. Casella SJ, Reiner BJ, Chen TC, et al. A possible genetic defect in 25-hydroxylation as a cause of rickets. J Pediatr 1994; 124:929.

25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47.

Walter Ntzenadel W, Mehis O, Klaus G. A new defect in vitamin D metabolism. J Pediatr 1995; 126:76. Coburn JW, Brickman AS, Hartenblower DL. Clinical disorders. In: Nor-man AW, Schaeffer P, eds. Vitamin D and problems related to uremic bone disease. New York: de Gruyter, 1975:219. Parfitt AM, Podenphant J, Villanueva AR, Frame B. Metabolic bone disease with and without osteomalacia after intestinal bypass surgery. Bone 1985; 6:211. Eddy RL. Metabolic bone disease after gastrectomy. Am J Med 1971; 50:442. Compston JE. Hepatic osteodystrophy: vitamin D metabolism in patients with liver disease. Gut 1986; 27:1073. Epstein S, Meunier PJ, Lambert PW, et al. 1a,25-Dihydroxyvitamin D3 corrects osteomalacia in hypoparathyroidism and pseudohypoparathyroid-ism. Acta Endocrinol (Copenh) 1983; 103:241. Chase LR, Melson GL, Aurbach GD. Pseudohypoparathyroidism: defective excretion of 3',5'cAMP in response to parathyroid hormone. J Clin Invest 1969; 48:1832. Levine MA, Downs RW Jr, Moses AM, et al. Resistance to multiple hormones in patients with pseudohypoparathyroidism: association with deficient activity of guanine nucleotide regulating protein. Am J Med 1983; 74:545. Haussler MR, Baylink DJ, Hughes MR, et al. The assay of 1a,25-dihydroxys-vitamin D 3: physiologic and pathologic modulation of circulating hormone levels. Clin Endocrinol (Oxf) 1976; 5:151S. Coburn JW, Sherrard DJ, Ott SM, et al. Bone disease in uremia: a reappraisal. In: Norman AW, Schaefer K, Herrath DV, Grigoleit H-G, eds. Vitamin D: chemical, biochemical and clinical endocrinology of calcium metabolism. New York: de Gruyter, 1982:835. Ravid M, Robson M. Proximal myopathy caused by iatrogenic depletion. JAMA 1976; 236:1380. Pierdes AM, Ellis HA, Kerr DNS. Phosphate-deficiency during regular hae-modialysis. Lancet 1976; 2:746. Moorhead JF, Wills MR, Ahmet KY, et al. Hypophosphataemic osteomalacia after cadaveric renal transplantation. Lancet 1974; 1:694. Hodsman AB, Sherrard DJ, Alfrey AC, et al. Bone aluminum and histomorpho-metric features of renal osteodystrophy. J Clin Endocrinol Metab 1982; 54:539. Brown DJ, Dawborn JK, Ham KN, Xipell JM. Treatment of dialysis osteomalacia with desferrioxamine. Lancet 1982; 2:343. Fraser D, Kooh SW, Kind HP, et al. Pathogenesis of hereditary vitamin D-dependent rickets (an inborn error of vitamin D metabolism involving defective conversion of 25-hydroxyvitamin D to 1,25-dihydroxyvitamin D). N Engl J Med 1973; 289:817. Scriver CR, Reade TM, DeLuca HF, Hamstra AJ. Serum 1,25-dihydroxy-vitamin D levels in normal subjects and in patients with hereditary rickets or bone disease. N Engl J Med 1978; 299:976. Fu GK, Lin D, Ahang MYH, et al. Cloning of human 25-hydrovitamin D 1a-hydroxylase and mutations causing vitamin Ddependent rickets type I. Mol Endocrinol 1997; 11:1961. Wang JT, Lin C-J, Burridge SM, et al. Genetics of vitamin D 1a-hydroxylase deficiency in 17 families. Am J Hum Genet 1998; 63:1694. Graham JB, McFalls VW, Winters RW. Familial hypophosphatemia with vitamin D resistant rickets. III. Three additional kindreds of sex-linked dominant type with a genetic analysis of four such families. Am J Hum Genet 1959; 11:311. Lobaugh B, Burch WM Jr, Drezner MK. Abnormalities of vitamin D metabolism and action in the vitamin D resistant rachitic and osteomalacic diseases. In: Kumar R, ed. Vitamin D basic and clinical aspects. Boston: Martinus Nijhoff, 1984:665. Glorieux FH, Scriver CR. Loss of a parathyroid hormone-sensitive component of phosphate transport in X-linked hypophosphatemia. Science 1972; 175:997. Short EM, Binder HJ, Rosenberg LE. Familial hypophosphatemic rickets: defective transport of inorganic phosphate by intestinal mucosa. Science 1973; 179:700.

47a. Miyamoto J, Koto S, Hasegawa Y. Final height of Japanese patients with x-linked hypophosphatemic rickets: effect of vitamin D and phosphate therapy. Endocr J 2000; 47:163. 48. Nesbitt T, Coffman TM, Griffiths R, Drezner MK. Crosstransplantation of kidneys in normal and Hyp mice: evidence that the Hyp mouse phenotype is unrelated to an intrinsic renal defect. J Clin Invest 1992; 89:1453. 49. Collins JF, Scheving LA, Ghishan FK. Decreased transcription of the sodium-phosphate transporter gene in the hypophosphatemic mouse. Am J Physiol 1995; 269:F439. 50. Lyles KW, Drezner MK. Parathyroid hormone effects on serum 1,25-dihy-droxy-vitamin D levels in patients with X-linked hypophosphatemic rickets: evidence for abnormal 25-hydroxyvitamin D-1-hydroxylase activity. J Clin Endocrinol Metab 1982; 54:638. 51. Drezner MK. Understanding the pathogenesis of X-linked hypophosphatemic rickets: a requisite for successful therapy. In: Zackson DA, ed. A CPC series: cases in metabolic bone disease. New York: Triclinica Publications, 1987:1. 52. Drezner MK. Clinical disorders of phosphate homeostasis. In: Feldman D, Glorieux FH, Pike JW, eds. Vitamin D. San Diego: Academic Press, 1997:733. 53. Nesbitt T, Hamrick JL, Quarles LD, Drezner MK. Coordinated developmental regulation of PEX and phosphatonin: evidence for the pathophysio-logical role of PEX in XLH. J Bone Miner Res 1997; 12:S113. 54. Drezner MK. Tumor associated rickets and osteomalacia. In: Favus M, Christakos S, Gagel RF, et al., eds. Primer on the metabolic bone diseases and disorders of mineral metabolism, 3rd ed. Philadelphia: Lippincott Raven Publishers, 1996:319. 55. Konishi K, Nakamura M, Yamakawa H, et al. Case report: hypophos-phatemic osteomalacia in von Recklinghausen neurofibromatosis. Am J Med Sci 1991; 301:322. 55a. Kumar R. Tumor-induced osteomalacia and the regulation of phosphate homeostasis. Bone 2000; 27:333. 56. Panda D, Lipman ML, Henderson JE, et al. Cloning of human PEX cDNA from tumors causing hypophosphatemic osteomalacia. J Bone Miner Res 1997; 12:S113. 57. Tieder M, Modai D, Samuel R, et al. Hereditary hypophosphatemic rickets with hypercalciuria. N Engl J Med 1985; 312:611. 58. Tieder M, Modai D, Shaked U, et al. Idiopathic hypercalciuria and hereditary hypophosphatemic rickets; two phenotypical expressions of a common genetic defect. N Engl J Med 1987; 316:125. 59. Brooks MH, Bell NH, Love L, Stern PH, et al. Vitamin Ddependent rickets type II: resistance of target organs to 1,25-dihydroxyvitamin D. N Engl J Med 1978; 298:996. 60. Liberman UA, Marx SJ. Vitamin D dependent rickets. In: Favus M, Christa-kos S, Gagel RF, et al., eds. Primer on the metabolic bone diseases and disorders of mineral metabolism, 2nd ed. New York: Raven Press, 1993; 274. 61. Whitfield GK, Selznick SH, Haussler CA, et al. Vitamin D receptors from patients with resistance to 1,25-dihydroxyvitamin D 3: point mutations confer reduced transactivation in response to ligand and impaired interaction with the retinoid X receptor heterodimeric partner. Mol Endocrinol 1996; 10:1617. 62. Malloy PJ, Pike JW, Feldman D. Hereditary 1,25-dihydroxyvitamin D resistant rickets. In Feldman D, Glorieux D, Pike JW, eds. Vitamin D. San Diego: Academic Press, 1997:765. 63. Baker AR, McDonnell DP, Hughes M, et al. Cloning and expression of full-length cDNA encoding human vitamin D receptor. Proc Natl Acad Sci U S A 1988; 85:3294. 64. Giraldo A, Pino W, Garcia-Ramirez LF, et al. Vitamin D dependent rickets type II and normal vitamin D receptor cDNA sequence. A cluster in a rural area of Cauca, Colombia, with more than 200 affected children. Clin Gen 1995; 48:57. 64a. Sierra OL, Alarcon Y, Penaloza LN, et al. Fractional intestinal calcium absorption in children with rickets and controls from Cauca, Columbia. J Bone Miner Res 2000; Su105(abstract). 65. Balsan S, Barabdian M, Larchet M, et al. Long-term nocturnal calcium infusions can cure rickets and promote normal mineralization in hereditary resistance to 1,25-dihydroxyvitamin D. J Clin Invest 1986; 77:1661. 66. Sotaniemi EA, Hakkarainen HK, Puranen JA, Lahti RO. Radiologic bone changes and hypocalcemia with anticonvulsant therapy in epilepsy. Ann Intern Med 1972; 77:389. 67. Winnacker JL, Yeager H, Saunders JA, et al. Rickets in children receiving anticonvulsant drugs. Am J Dis Child 1977; 131:286. 67a. Karaaslan Y, Haznedaroglu S, Ozturk M. Osteomalacia associated with car-bamazepine/valproate. Ann Pharmacother 2000; 34:264. 68. 69. 70. 71. 72. 73. 74. 75. Hahn JT. Drug-induced disorders of vitamin D and mineral metabolism. Clin Endocrinol Metab 1980; 9:107. Sherk HH, Cruz M, Stambaugh J. Vitamin D prophylaxis and the lowered incidence of fractures in anticonvulsant rickets and osteomalacia. Clin Orthop Rel Res 1977; 129:251. Richards P, Chamberlain MJ, Wrong OM. Treatment of osteomalacia of renal tubular acidosis by sodium bicarbonate alone. Lancet 1972; 2:994. Cunningham J, Fraher LJ, Clemens TL, et al. Chronic acidosis with metabolic bone disease. Am J Med 1982; 73:199. Russell RGG, Smith R, Preston C, et al. Diphosphonates in Paget's disease. Lancet 1974; 1:894. Khairi MRA, Altman RD, DeRosa GP, et al. Sodium etidronate in the treatment of Paget's disease of bone. Ann Intern Med 1977; 87:656. Ott SM, Maloney NA, Klein GL, et al. Aluminum is associated with low bone formation in patients receiving chronic parenteral nutrition. Ann Intern Med 1983; 98:910. Vargas JH, Klein GL, Ament ME, et al. Metabolic bone disease of total parenteral nutrition: course after changing from casein to amino acids in parenteral solutions with reduced aluminum content. Am J Clin Nutr 1988; 48:1070. 76. Henthorn PS, Raducha M, Fedde KN, et al. Different missense mutations at the tissue-nonspecific alkaline phosphatase gene locus in autosomal recessively inherited forms of mild and severe hypophosphatasia. Proc Natl Acad Sci U S A 1992; 89:9924. 77. Whyte MP. Hypophosphatasia. In: Scriver CR, Beaudet AL, Sly WS, Valle D, eds. The metabolic and molecular bases of inherited disease, 7th ed. New York: McGraw-Hill, 1995:4095. 78. Bell NH, Stern PH. Hypercalcemia and increases in serum hormone value during prolonged administration of 1a,25-dihydroxyvitamin D. N Engl J Med 1978; 298:1241.

CHAPTER 64 OSTEOPOROSIS Principles and Practice of Endocrinology and Metabolism

CHAPTER 64 OSTEOPOROSIS
ROBERT LINDSAY AND FELICIA COSMAN Epidemiology Postmenopausal and Age-Related Primary Osteoporosis Pathophysiology and Etiologic Considerations Determinants of Maximum Bone Mass Determinants and Pathogenesis of Bone Loss Other Risk Factors Clinical Features Intervention in Postmenopausal Osteoporosis Other Forms of Primary Osteoporosis Osteoporosis in Men Juvenile Osteoporosis Idiopathic Osteoporosis in Young Adults Regional Osteoporosis Reflex Sympathetic Dystrophy Transient Osteoporosis of the Hip Regional Migratory Osteoporosis Other Osteolytic Syndromes Corticosteroid-Induced Osteoporosis Primary Hyperparathyroidism Hyperthyroidism Amenorrhea and Anorexia Nervosa Hyperprolactinemia Type 1 (Insulin-Dependent) Diabetes Acromegaly Osteogenesis Imperfecta Other Genetic Disorders Alcoholism Multiple Myeloma and Other Malignancies Pregnancy and Lactation Gastrointestinal Disease Obstructive Liver Disease Heparin-Induced Osteoporosis Anticonvulsant Therapy Chapter References

Osteoporosis, the most common bone disease in clinical practice, is a skeletal disorder characterized by a reduction in bone mass with accompanying microarchitectural damage that increases bone fragility and the risk of fracture.1,2 Histologically, the cortices are thinned and porous, and the trabeculae are fewer, thinner, and less connected. Clinically the disease is important because of the fractures that occur as a consequence.

EPIDEMIOLOGY
The incidence of osteoporosis increases with age. It becomes widely prevalent in the elderly, in whom it is a major public health problem, causing >1.3 million fractures annually and costing more than $13 billion per year in the United States.1,3 Peak bone mass occurs at ~18 to 30 years of age, after which bone is progressively lost. This phenomenon is seen in all populations, regardless of race, sex, economic development, or geographic location.4 Fracture rates vary in different groups; they are higher in whites and Asians than in blacks, and usually are higher among women than among men. The incidence of limb fractures doubles approximately every 5 to 8 years after the fifth decade among women but does not increase substantially until after the seventh decade among men. The most common fracture sites are in the thoracic and lumbar vertebral bodies (i.e., crush or wedge fractures), the neck and intertrochanteric regions of the femur (i.e., hip fractures), and the distal radius (i.e., Colles fractures); however, in osteoporosis, fracture of any bone is possible (Fig. 64-1).

FIGURE 64-1. Incidence rates for vertebral, Colles, and hip fractures in women. (From Wasnich RD. In: Favus MJ, et al. Primer on the metabolic bone diseases and disorders of mineral metabolism. Philadelphia: Lippincott-Raven Publishers, 1996:249.)

The prevalence of hip fractures in the United States reaches 15% by age 80 and exceeds 25% by age 90, with a female/male predominance of ~2:1 among whites.4 In other populations, however, such as those of Singapore and Hong Kong, and the South African Bantu, the incidence of hip fractures in men may equal or exceed that in women.5 Approximately 300,000 hip fractures occur in the United States every year. The incidence has been rising, with a 40% increase reported between 1970 and 1980. However, data suggest that this rise has plateaued.6 The mortality rate associated with hip fracture and its complications is 12% to 20% within the first year, and a large proportion of patients are mildly or severely disabled permanently.1,2 The true prevalence of vertebral fracture is unknown because of the variable definitions of vertebral fracture and variable clinical expressions. More than 25% of all women older than 65 years of age may have sustained a crush fracture, many of which are asymptomatic. The female/male predominance of vertebral fracture is estimated to be between 2:1 and 8:1. 7,8 Like hypertension, osteoporosis has many causes and pathogenetic mechanisms. The primary form of the disease is the most common and can be further classified into several types. Secondary osteoporosis is seen in many endocrine, genetic, and other illnesses (Table 64-1), but the most common form is that associated with the therapeutic use of corticosteroids.

TABLE 64-1. Classification of Osteoporosis

POSTMENOPAUSAL AND AGE-RELATED PRIMARY OSTEOPOROSIS

PATHOPHYSIOLOGY AND ETIOLOGIC CONSIDERATIONS More than 80% of osteoporosis cases occur among postmenopausal and aging populations. Although some differences in pathogenetic mechanisms have been described,9,10 the overlap between populations seen clinically is sufficient for them to be considered together. Fracture risk at any age is determined primarily by skeletal mass,11,12 which depends on the maximal mass achieved at maturity and on the subsequent rate and duration of bone loss. Various interrelated factors are thought to control the two processes.13 The frequency of falling and factors predisposing to falls are also determinants of fracture risk, particularly for limb fractures.14 Thus, the frail elderly are more likely to sustain fractures than those who maintain their health, coordination, and strength. Reduced bone elasticity and decreased dissipation of force over bone by weaker and less coordinated muscle contractions also add to fracture risk. The presence of vertebral fractures has been found to be an independent predictor of further vertebral fracture risk,15 and also increases the risk of fractures at other sites, notably the hip.16 The occurrence of any fracture after the age of 40 doubles the risk of further osteoporotic fractures in later life.17 DETERMINANTS OF MAXIMUM BONE MASS Maximum bone mass is achieved some time after puberty (age 1830) and is primarily determined genetically.18,19 Race, gender, and body size are the principal genetic determinants; consequently, maximum bone mass probably is under polygenic control. The inequality between the sexes is probably not evident prepubertally,20 but is clear after transition through puberty, with total bone mass becoming significantly greater among males.21,22 and 23 The positive correlation between body size and bone mass is well established.23 Bone density is also a function of skeletal size, because the usual two-dimensional techniques employed do not adequately correct for the size of the bone. Gender differences in true density are less than those in bone mass in young adults, and in some studies are found to be insignificant.23 Racial differences in both mass and density are apparent in young adults, with black populations generally having a 5% to 10% greater bone mass.24 In cross-sectional studies these racial differences are maintained with increased age, whereas gender differences become yet greater. Investigations into the genetic control of bone mass and density have focused on several candidate genes that, by virtue of the known function of their products, have some biologic credibility.19 Thus, investigators have evaluated polymorphisms in the vitamin D receptor; the estrogen receptor (ER-a); the gene for type I collagen, osteocalcin, and the interleukins; and receptors and receptor antagonists. Although all have provided some promising data, each contributes only a small amount to the variability seen in bone density or bone turnover. Although multiple genetic factors clearly play an important role in determining peak bone mass, other influences are also important. Adequate nutrition, health, and exercise in childhood can modify the likelihood that any individual will reach his or her genetic potential.25,26 Nutritional factors include not only calcium but calories and protein. Furthermore, a normal progression through puberty and the associated growth spurt as well as normal menstrual function are critical to adult skeletal health. Osteoporosis has been dubbed a disease that presents in the geriatric population but often has its roots in the pediatric population. DETERMINANTS AND PATHOGENESIS OF BONE LOSS In women, bone loss begins between 20 to 30 years of age, at least in the hip. Bone mass may be more stable in the spine.27 This premenopausal loss of bone in the hip may be related to marginal estrogen deficiency or decreased physical activity.28 Bone loss in the spine probably does not occur until after the fifth decade of life and is more dependent on the decreasing production of estrogen as women approach the menopause.28,29 Bone loss in the spine in men also may begin at this age but occurs at a slower rate and without the accelerated loss that is typically seen in menopausal women. Net loss of bone may be a universal phenomenon with increasing age, at least in humans.30 Whether such bone loss with increasing age occurs in other animal species is unclear. This is an important issue, because unique facets of human life may have consequences for skeletal health. The rate of bone loss is determined by endocrine, environmental, and various local factors, and probably also by genetic influences. Eskimos, for example, have rapid age-related bone loss and a high frequency of fracture. 31 Those traits probably have both environmental and genetic causes. The adult skeleton undergoes a process of remodeling by which old bone is removed and replaced by new, young bone.32 Before menopause, the bone that is removed is replaced by an equal amount of new bone tissue, and the frequency with which bone remodeling units are initiated is constant. With estrogen deficiency in women, and with increasing age in both sexes, changes occur in the remodeling process.33 Across menopause, an increase is seen in the number of new bone remodeling units initiated per unit time. This creates a transient loss of bone mass. For continued loss to occur, however, an imbalance must exist between formation and resorption within each remodeling unit.34 This can occur for one of several reasons: (a) more bone may be removed by more aggressive osteoclasts, (b) insufficient new bone may be laid down by new osteoblasts, (c) a combination of these two may occur, or (d) aggressive osteoclasts may penetrate trabeculae, removing the template on which new bone may be laid down. Variable combinations of these effects probably occur. In postmenopausal women, the rate at which new remodeling sites are initiated increases. The suggestion has been made that the osteoclast population is hyperactive, leading to excessive bone resorption.35 A combination of these effects results in a net bone loss within each remodeling site and a stochastic probability of increased trabecular penetration. In glucocorticoid-induced osteoporosis, the same effects occur but are exacerbated by impaired osteoblast activity.36 This remodeling process takes place on the surface of cancellous bone at the endocortical junction and within the cortex of bone where the completed remodeling process is seen as Haversian systems. The fact that the majority of the activity occurs on the surface means that, when situations of imbalance occur, the loss of tissue is greatest at high surface areas and, within sites of equal surface, is highest at sites within red marrow in juxtaposition.37 Consequently, in any situation in which excessive bone loss occurs, bone loss occurs first and most markedly in cancellous bone, which contributes 80% of the surface area of the skeleton but only 20% of the skeletal mass. In healthy individuals, the process of bone loss is insidious, occurring over many years in an asymptomatic fashion. The average negative calcium balance of 30 to 50 mg per day after three decades can cause as much as 30% to 50% loss of skeletal mass. Superimposed on these slow changes are the endocrine changes at menopause, in which a loss of estrogen produces increased bone remodeling.33,34,35,36,37 and 38 After spontaneous or surgical menopause, vertebral bone loss rates of 3% to 5% per year have been recorded for the first 5 to 10 years. Thereafter, the process of bone loss becomes slower, but it continues in many people until old age. Some indications are found that bone loss may accelerate in the very old individual.39 Menopause-related changes in bone remodeling are clearly triggered by the declining estrogen level.40 The mechanisms by which estrogen deficiency induces bone loss are only now beginning to be understood. Estrogen receptors, both ER-a and ER-b have been detected in a variety of cells within marrow and on cells of osteoblast and osteoclast lineage.41,42 and 43 However, whether the estrogen effects are mediated through ER-a or ER-b in the skeleton is unknown. Multiple target cells for estrogens probably exist within the skeleton, including macrophages, monocytes, lymphocytes, and other cells within marrow, as well as osteoblasts, preosteoclasts, and osteoclasts. In addition, osteocytes may be target cells for estrogen. Consequently, the fact that estrogen actions depend on a variety of cell-cell interactions, and a release of a variety of local cytokines and growth factors, is not surprising.44 Thus, implicated in the estrogen actions on the skeleton are interleukins 1, 6, and 11, prostaglandin E2, tumor necrosis factor-a (TNF-a, transforming growth factor-b (TGF-b, insulin-like growth factor-I (IGF-I), and perhaps also IGF-II.45,46,47,48,49,50,51,52,53,54,55,56 and 57 Moreover, direct effects of estrogen may include control of cell life span for osteoblasts, osteoclasts, and/or osteocytes, and control of the synthesis of new osteoclasts via osteoclast differentiation factor (or RANK ligand), which could be the final effectors of estrogen action in the recruitment process of new osteoclasts.58,59 and 60 Effects on other estrogen-responsive genes (e.g.,

c-phos and c-jun) have not been integrated into these processes.61 The consequence of the changes in the bone turnover with estrogen deficiency can be detected by the increased mean and median levels of bone biochemical markers in serum and urine.62,63,64,65,66,67,68,69,70,71,72,73 and 74 The ranges of these marker levels are also increased; many individuals still have levels of biochemical markers within the expected premenopausal range, whereas others have levels that are distinctly elevated. Thus, the suggestion has been made that measurement of biochemical marker levels may provide a useful test for determining the rates of bone loss. However, although several studies have examined the relationship between baseline biochemical marker values and changes in bone mineral density (BMD) in untreated postmenopausal women, in general, the relationships are modest; biochemical marker levels are inadequate for clinical use in determining whether individual patients will lose bone rapidly.70,71 and 72 Although increased levels of biochemical markers do indicate the increased remodeling activity within the skeleton, the inherent intraindividual variability of marker levels probably limits their usefulness.73 The consequence of this increased skeletal remodeling is loss of bone mass. As noted, loss of bone occurs earliest and most rapidly in areas of cancellous bone, which can be seen in the vertebral body. The risk of fracture increases with declining bone mass.74,75,76,77,78,79,80 and 81 That women who lose bone at the greatest rate will have the greatest deterioration in skeletal architecture,82 and subsequently be at even higher risk of fracture regardless of bone density, seems logical. This suggests that biochemical markers should correlate with the risk of fracture. Several studies have now indicated that biochemical marker levels do predict the risk of fracture in several populations.83,84,85 and 86 In addition, the suggestion has been made that biochemical markers of bone resorption predict the risk of fracture independently of bone mass and that the two effects are additive.83,87 However, this combination of tests has yet to be applied clinically. Bone loss continues among postmenopausal women and becomes more apparent in men older than age 60 years.39,88 Some have argued that this process of bone loss in older individuals is related to features other than estrogen deficiency.89,90 Data have shown, however, that rates of bone loss in this population are dependent on endogenous estrogen production and that estrogen intervention prevents bone loss in the elderly, just as it does in women in the early years after the menopause.91 Thus, these findings suggest that at least part of the process of bone loss in elderly individuals is estrogen sensitive. The report of a man with estrogen resistance due to absence of the estrogen receptor who presented with osteoporosis and continued growth because of failure of the epiphyses to fuse suggests that, in men, estrogen is responsible for control of skeletal homeostasis and may also be important in determining rates of bone loss.92 Declining ability to aromatize androgens to estrogen may also contribute to bone loss in aging men.93 One alternative hypothesis regarding bone loss after menopause is that a primary renal leak of calcium occurs, with increased obligatory loss of calcium that is offset by an increased skeletal remodeling required to maintain serum calcium.94 This hypothesis has some biologic plausibility, because hypercalciuria is a known risk factor for osteoporosis,95 particularly among men.96 Some studies have suggested that the increased renal loss of calcium may be estrogen dependent in women and, thus, a feature of postmenopausal bone loss.94 Thus estrogen deficiency may affect mineral metabolism. The net effect is to increase the likelihood that bone will be resorbed more easily by any stimulus to remodeling.97 Bone is less resistant to the resorbing effects of parathyroid hormone (PTH) infusion in the estrogen-deficient state, confirming this hypothesis.98 OTHER RISK FACTORS A large number of risk factors have been implicated in post-menopausal osteoporosis (Table 64-2).99 Some of these affect bone mass and its loss, thereby altering fracture risk, whereas others modify the risk of fracture independently of bone mass. Some risk factors may have effects on both bone mass and fracture risk. One example is cigarette use, which may affect bone turnover and the age of menopause, and, through its harmful effects on general health, may independently raise fracture risk.100 More conveniently, from the clinician's viewpoint, risk factors are usually categorized conceptually into those that are modifiable and those that are not. Although all are important in determining an individual's likelihood of fracture, it is the former that must be addressed when making treatment decisions.

TABLE 64-2. Risk Factors, Diseases, and Drugs Associated with an Increased Risk of Osteoporosis in Adults

Nonmodifiable Risk Factors. Of the factors that influence risk and cannot be changed, age is the single most important.101,102 and 103 In addition, gender, a family history of osteoporosis, and a past history of fracture affect risk.101,102,103 and 105 In most societies, fractures occur more commonly in women, and in white and Asian groups in most Western countries, female gender doubles fracture risk. The same appears true for the black population, although the absolute risk is significantly lower for blacks than for those of other races.106 In the United States, the risk of fracture among black women is roughly equal to that among white males. Not surprisingly, persons with a family history of hip fracture in a parent are at approximately twice the risk of others for hip fracture.101 Data suggest that those with a family history of osteoporosis have lower peak bone mass, and parental bone mass is correlated with bone mass in offspring.107,108,109 and 110 The clinical experience that patients who present with fractures are more likely to fracture in the future has been confirmed by several formal studies.101,111 The view is now generally accepted that those who present with a history of fracture are at approximately twice the risk of future fracture, an increase in risk that is independent of bone mass. The occurrence of vertebral fractures dramatically increases the risk of subsequent vertebral fractures by five-fold for one fracture and by 12-fold for two or more fractures, while doubling the risk for other osteoporotic fractures.111 Modifiable Risk Factors. Although the results of studies evaluating those factors that might be modifiable in any individual patient have varied, some broad agreement exists regarding evaluations for fracture risk. Furthermore, a large number of chronic diseases and medications have been associated with an increased risk of osteoporosis (see Table 64-2). NUTRITION. Good nutrition is as important for skeletal health as it is for general health.112 During growth, an adequate supply of calories, protein, and minerals is a prerequisite for the attainment of peak skeletal mass.18,26,27,112,113 Although the effect has not been proven by long-term studies, undernutrition in childhood leads to a skeletal deficit that cannot be repaired during adult life and presumably, therefore, would lead to an increase in the risk of osteoporotic fracture in later life. During adult life, an adequate intake of calcium is required to ensure that serum calcium can be maintained within the normal range (Table 64-3). Inadequate intake necessitates the use of calcium from the skeleton to maintain levels of serum calcium and leads to a consequent increase in bone turnover that results in a loss of skeletal mass.38 The effects of inadequate calcium intake are likely to be less among premenopausal women with adequate ovarian function than among women in an estrogen-deficient state.114,115 and 116 In the early postmenopausal years, the effects of estrogen deficiency are sufficiently potent that calcium supplementation has only modest effects.117 As individuals age, however, the role of calcium appears to become more important.118 Thus, studies of calcium supplementation conducted among older individuals show significant beneficial effects on bone mass and the incidence of fractures.119,120,121,122,123,124,125,126 and 127 The message to patients must be that calcium intake is important at all ages.

TABLE 64-3. Adequate Intake Levels for Calcium and Vitamin D

The role of vitamin D in the pathogenesis of osteoporosis has been debated for some time.128 Although clearly vitamin D deficiency leads to osteomalacia in adults, whether lesser degrees of vitamin D deprivation can lead to osteoporosis is unclear. Certain populations frequently have circulating levels of 25-dihydroxyvitamin D [25(OH)D] that are considered to be in the borderline range (5 to 20 ng/mL).129,130 and 131 Because costs and risks are minimal, recommending supplementation to these groups seems prudent. These populations include the elderly, the institutionalized, and those who are chronically ill, especially those with malabsorption and diseases causing diminished ambulation, such as multiple sclerosis.132,133,134,135 and 136 For these groups, ingestion of 400 to 800 U per day of cholecalciferol is safe and cost effective. The suggestion has also been made that the reduced intestinal calcium absorption seen in osteoporosis is the primary defect and part of a more general resistance to the actions of 1,25-hydroxyvitamin D [1,25(OH)2D], which, in turn, could be related to a reduced receptor density.9 Evidence also exists of a reduced supply of the active metabolite, 1,25(OH)2D, perhaps related to an age-dependent decline in renal function.13 Superimposed on these metabolic features is the decline in physical activity with age, which may play an important role in determining rates and duration of bone loss, especially in the hip. The hypothesis argues that the skeleton has an inbuilt mechanostat, which sets bone mass according to the degree of stress placed on the skeleton. This is most obvious in the increase of bone mass during growth and in the loss of bone that occurs after disuse. A smaller loss could be precipitated by the gradually reduced stress of lessening physical activity, as is often seen in the elderly. In any individual some combination of these events may conspire to produce a negative skeletal balance. Superimposed on these age-related changes are secondary causes of bone loss and, in women, the effects of estrogen deficiency at menopause. An often underrecognized nutritional factor in the pathogenesis of osteoporosis is sodium intake.137 A high intake of sodium results in an obligatory loss of calcium in the urine in excess of the normal urinary losses. This phenomenon, continued over years (especially if not compensated for by an adequate calcium intake), would be expected to exacerbate other risk factors for osteoporosis. Whether limitation of sodium intake can reduce bone loss has not been determined. Other nutritional factors that have been implicated in the pathogenesis of osteoporosis include excessive intake of protein, especially animal protein.138 The suspicion is that this is the effect of the acid load of such diets. Excess caffeine intake has also been implicated, either because it increases urinary calcium excretion or because it leads to a minor diminution of calcium absorption. However, demonstrating a relationship between caffeine intake and fracture is difficult, partly because of the difficulty of determining the level of caffeine intake, partly because of the confounding effects of other risk factors,138,139 and partly because caffeine appears to have such a minor effect on skeletal metabolism. Phosphorus intake, such as in cola drinks, also has been implicated, but has a very minor effect on bone metabolism. The major importance of this latter nutrient is that phosphorus-containing drinks often substitute for calcium-containing ones, such as milk. Perhaps the most controversial nutritional risk factor is alcohol. No doubt exists that excess alcohol intake is detrimental to the skeleton140,141; a high intake is a risk factor for osteoporosis among men. The role of moderate intakes is less clear, however. Some data suggest that consumption of one to two alcoholic drinks per day may have a positive effect on bone mass, but because those data are from epidemiologic studies, other biases cannot be excluded.142 The current calcium recommendations should be adequate to offset effects for patients with even moderately high intakes of protein, caffeine, and salt, as are often seen in the U.S. population. LIFESTYLE. Data from several studies suggest that cigarette use is a significant risk factor for osteoporotic fracture 101,142 (Fig. 64-2). Several mechanisms exist by which this effect might be mediated. Cigarette use results in a menopause that is earlier by as much as 2 years and lowers estradiol levels in both pre- and postmenopausal women.143 In some animal models of osteoporosis, exposure to cigarette tobacco has direct toxic effects on the skeleton. Cigarette use may also be associated with decreased physical activity. Finally, the effects of cigarettes on general health may result in frail individuals who are more at risk for falls and subsequent injury.144 Consequently, some of the risk conferred by cigarette use is independent of its effects on bone.101 Among the elderly, current cigarette use doubles the risk of hip fracture.101

FIGURE 64-2. A, Mean ( standard error of the mean) adjusted annual rates of bone mineral density (BMD) change of the femoral neck, total body, and lumbar spine in smokers (hatched bars) and nonsmokers (white bars). Rates are adjusted for baseline BMD, weight, age, gender, supplementation status (calcium plus vitamin D or placebos), and dietary calcium intake. Values at the femoral neck are -0.714 0.285% per year for 31 smokers versus +0.038 0.084% per year for 355 nonsmokers, p <.02; for the total body, -0.360 0.101% per year for 31 smokers versus -0.152 0.030% per year for 354 nonsmokers, p <.05; and at the spine, +0.260 0.252% per year for 29 smokers versus +0.593 0.074% per year for 339 nonsmokers, p = .21. B, Mean ( standard error of the mean) adjusted fractional calcium absorption in smokers (hatched bar) and nonsmokers (white bar). Means are adjusted for gender, age, supplementation status (placebo or calcium plus vitamin D), and dietary calcium and vitamin D intakes. Values are 12.9 0.8% for 23 smokers and 14.6 0.2% for 310 nonsmokers, p <.05. (From Krall EA, Dawson-Hughes B. Smoking increases bone loss and decreases intestinal calcium absorption. J Bone Miner Res 1999; 14:215.)

PHYSICAL ACTIVITY. The popular belief exists that physical activity is important for good skeletal health. Cross-sectional data clearly show dramatic effects on the skeleton at the extremes of activity.145,146,147 and 148 Complete loss of activity causes loss of bone mass, just as muscle mass is lost. Highly active individuals have greater skeletal mass than average; this may be specific to the underlying bone used in the activity (e.g., forearm bone in tennis players).148 Within the average range of physical activity in the general population, however, demonstrating marked effects of changing activity on bone mass has been difficult. Most studies do suggest modest effects, with the changes in activity producing bone mass differences of 1% to 2%. Metaanalyses confirm decreased bone loss in exercisers compared with sedentary individuals. Most would agree that activity has an important role to play in the prevention of osteoporotic fracture. Among the elderly, this may be related to the effects of maintaining activity on the risk of falls and subsequent injury.149,150,151 and 152 Thus, patients with or at risk of osteoporosis should incorporate exercise into their therapy. CLINICAL FEATURES In general, osteoporosis has no symptoms, and the clinical features with which it presents are those of its complications, that is, fractures. Vertebral fracture is the most common clinical manifestation of osteoporosis. It is often asymptomatic and identified incidentally on a chest radiograph; only 10% to 30% of vertebral fractures present with acute pain. Multiple vertebral fractures (crush fractures) lead to increased dorsal kyphosis, loss of height, and dowager's hump (Fig. 64-3). When fractures are

symptomatic, acute pain may develop suddenly after minimal force or strain from routine activities such as bending, reaching, or lifting. This pain subsides within a few weeks, but may be followed by a period of chronic, dull, diffuse pain, which may last for months to years. In part, this chronic pain is related to the changing biomechanical competence of the spine. Progressive vertebral body collapse, with loss of thoracolumbar height, may cause the lower ribs to rest on the iliac crest, which may produce some relief of back pain but may result in lower rib and flank pain. This altered body configuration compresses pulmonary and abdominal organs and may be associated with exertional dyspnea and gastrointestinal distress, including early satiety, bloating, and constipation.153,154,155,156,157,158 and 159 Psychosocial consequences (fear, depression, irritability, and loneliness) often follow multiple vertebral fractures. Neurologic signs and symptoms of nerve root or spinal cord compression are seldom caused by osteoporotic vertebral fractures and should prompt a search for other causes of fracture.

FIGURE 64-3. Radiographic appearance of the dorsal spine (A) and the lumbar spine (B) showing anterior wedging and vertebral lapse, radiolucency, and biconcavities.

Hip fractures (e.g., subcapital femoral neck, intertrochanteric, and subtrochanteric) occur in 15% of women and 5% of men by age 80 years4 (Fig. 64-4). Surgical intervention is almost always necessary if patients are expected to return to an ambulatory status. Hip fracture is a common cause of nursing home admission. Because of the surgery and postsurgical complications, hip fractures are associated with a 5% to 20% excess mortality in the first year after fracture.4,14,160,161,162,163,164,165 and 166 Approximately 30% to 50% of patients with hip fracture never return to their level of activity before the fracture. Distal radius fractures (i.e., Colles fractures) cause less morbidity than hip or vertebral fracture (see Fig. 64-4). However, wrist fractures increase in frequency at an earlier age than do fractures of the hip or spine.4,167,168,169 and 170 Any individual who fractures a wrist should be considered possibly at risk for osteoporosis and should be considered for evaluation. Because osteoporosis is a systemic disease, fractures of any bone can occur, and fracture of any bone after the age of 40 should raise a suspicion of osteoporosis in any individual.

FIGURE 64-4. A, A 54-year-old osteoporotic woman with typical Colles fracture caused by a fall on the outstretched hand. There is a fracture of the distal end of the radius (arrowhead), with angulation and impaction of the distal fragment. B, A 75-year-old osteoporotic man with a sub-capital femoral fracture (arrowheads). Notice that the broken-off head of the femur is completely displaced from the neck. The usual treatment of a fracture with this degree of displacement is surgical excision of the femoral head and replacement with a metallic prosthesis. C, A 77-year-old man with intertrochanteric femoral fracture. The fracture line (arrowheads) extends from the center of the greater trochanter (GT) to the lesser trochanter (LT). In this instance, the LT has broken off as a separate fragment. The treatment of this fracture is by internal fixation with a metallic device. (Courtesy of Dr. Bahman Sadr, Washington, DC.)

HISTORY AND PHYSICAL EXAMINATION Postmenopausal osteoporosis is a diagnosis of exclusion. Clinical evaluation is specifically directed toward determining if secondary causes of the disease are present. Evaluation of all prior fractures, current or past renal stone or other renal disease, thyroid disease, glucocorticoid use, Cushing syndrome, diabetes, hypogonadism, immobilization, liver disease, intestinal disease, and gastric or intestinal surgery should be made. Back pain should be evaluated, as it might be the presenting symptom of vertebral osteoporosis. The consideration of possible systemic symptoms of malignancy, especially of multiple myeloma, is essential in any patient. In women, the age of menopause, menarche, and reproductive history are important. A history of excessive alcohol or caffeine consumption and of cigarette smoking should be elicited. Lifelong calcium intake and exercise history should be determined. Nutritional calcium status is easily estimated. The family's medical history should be ascertained, particularly with respect to hip and wrist fractures and significant height loss or change in thoracic shape. Physical evaluation should include a general physical examination with measurement of height, weight, and arm span (which gives an estimate of original height) and with a search for signs of secondary diseases. Suspicion of secondary disease should be particularly high in patients who are neither postmenopausal women nor elderly. LABORATORY TESTS All laboratory values, including indices of mineral metabolism, are usually normal in patients with primary or postmenopausal osteoporosis, except in those with recent fractures, who may show elevations in serum alkaline phosphatase related to fracture healing. Laboratory evaluation could include measurement of serum calcium, phosphorus, alkaline phosphatase, and thyroid-stimulating hormone levels, complete blood count, erythrocyte sedimentation rate, tests of renal and hepatic function, and serum protein electrophoresis, depending on clinical suspicion. If adrenocortical hyperfunction is suspected, a urinary free cortisol test or an overnight dexamethasone suppression test should be performed. When serum calcium is high, the serum PTH level should be obtained. If questions exist regarding the nutritional status in elderly patients, patients with other chronic diseases, or patients on antiepileptic drugs, the measurement of the 25(OH)D level may be helpful. Some recommend supplementation with cholecalciferol in these circumstances without necessarily obtaining a biochemical measurement. BONE MASS MEASUREMENT Osteoporosis is defined as a disease of bone mass in which bone mass or density has fallen below the expected range for young adults. The analogy is the upper limits of normal for blood pressure or cholesterol, which are taken as indicators of hypertension or hypercholesterolemia. These risk factors for the clinical catastrophes of stroke or myocardial infarction are in common use in clinical practice. Bone mass measurement, which is at least as good a predictor of the clinical outcome, merits similar use.74,75,76,77,78,79,80 and 81 Two sensitive and noninvasive tests are available for measuring bone density in clinical practice. The first, dual-energy x-ray absorptiometry (DXA), which is used to measure appendicular and axial sites and can be used to measure total body bone mass, is the most precise method available, produces minimal radiation exposure, and has become widely available. The second, quantitative computed tomography (QCT), has the advantage of measuring only trabecular bone and may be the most sensitive indicator of cancellous bone loss after menopause; however, this advantage is outweighed by poor precision in comparison to other techniques and much greater radiation exposure (see Chap. 57). Several techniques that measure peripheral bone have the advantage of small size, which allows them to be easily housed in an office. These techniques include peripheral CT, peripheral dual-energy x-ray absorptiometry, single-energy x-ray absorptiometry, and ultrasonography. (The advantage of ultrasonography is the lack of

x-ray exposure and the possibility that it might measure structural characteristics of the bone that contribute to skeletal fragility, although this remains to be proven.) Bone densitometry is used to determine the risk of fracture. An inverse relationship is seen between density and the likelihood that a patient will experience fracture.74,75,76,77,78,79,80 and 81 The risk of hip fracture increases 2.5-fold for every one standard deviation decline in bone density when measured at the hip. Measurements of bone density at the hip allow estimation of risk of fractures at other sites, for which the relative risk is ~1.5 for every standard deviation change in bone density. Measurements of bone density at other sites are less effective in identifying those patients most likely to experience hip fracture, and because hip fracture is the most serious consequence of osteoporosis, measurements of the hip are generally preferred in predicting risk. However, for the early postmenopausal woman whose major concern may be vertebral fractures, measurements of the spine may be more appropriate. Measurements of the spine are often uninterpretable in the elderly because of facet joint disease and extraosseous calcification, so that measurements of the hip are preferable. When measurements of the hip and spine are not available, peripheral bone measurements may be substituted. However, although guidelines exist for intervention based on bone density measurements of the central skeleton, such guidelines do not exist for measurements of the peripheral skeleton. Bone density measurement has become an almost routine clinical tool in evaluation of patients for risk of osteoporosis and evaluation of those who have already had fractures. For asymptomatic individuals, bone densitometry should be offered to every woman by the age of 65. For women who are postmenopausal but younger than age 65 years, if risk factors other than menopause exist, bone densitometry should be recommended. All persons who present with fracture at >40 years should be considered for bone densitometry. Medicare guidelines allow repeat bone densitometry at intervals of ~2 years after the progression of the disease or 1 year after a new treatment is initiated. Bone density measurements must change by at least 3% to 4% in the spine and at least 5% to 6% in the hip to be considered significant in individual patients. TESTS OF BONE TURNOVER Several tests are available to measure either breakdown products from the skeleton or specific enzymes or proteins synthesized by osteoblasts. Levels of these markers of bone turnover have several possible uses in the management of patients with osteoporosis. First, they may provide some information on the baseline status of bone remodeling, which, when increased as occurs in postmenopausal women, adds to the risk of osteoporotic fracture. Second, these tests can be used to monitor the response to treatment. The tests available in the United States are listed in Table 64-4. These tests should be used only in selected patients, because the results may be difficult to interpret due to significant inter- and intraindividual variability.

TABLE 64-4. Biochemical Markers of Bone Turnover

In unusual situations, when osteoporosis presents in young individuals, a bone biopsy may be used to evaluate cellular activity on the surface of the skeleton. Such situations are, however, increasingly unusual. RADIOLOGIC FEATURES Greater than 30% loss of bone mass has to occur before detectable changes can be seen on plain radiograph. In the spine, individual trabeculae are thinned and, because the horizontal trabeculae are preferentially lost, vertical trabeculae appear more prominent. The vertebral bodies eventually appear as empty shells, and changes in shape begin to occur. Schmorl nodes, caused by the indentation of the disk into the vertebral body, are not pathognomonic of osteoporosis but certainly raise clinical suspicion.171 The classic wedge-shaped deformity is characterized by reduced anterior vertebral height (see Fig. 64-3). This decrement is thought to be significant when the anterior height becomes >15% lower than that of the immediate craniad vertebra, or alternatively, more than three standard deviations below the expected mean for the anterior height of the vertebrae itself. Biconcavity or codfish vertebra indicates the collapse of the central vertebral body and is particularly common in the lumbar spine. Posterior wedging is less common and may suggest a destructive lesion, especially if posterior extrusion of a fragment toward the cord is seen. A true crush fracture is a decrease in all aspects of vertebral height (anterior, central, and posterior). Vertebral fractures seldom occur above the level of T4 and their frequency is maximal between T8 and L3. Single vertebral fractures are often asymptomatic; however, single severe wedge fractures at the level of T7 can be particularly symptomatic because they produce an especially prominent kyphotic angulation. Radiographs can also be useful to rule out other bone diseases, such as Paget disease, in patients with bone pain and/or fractures. Radiographs taken after wrist or hip injury show the typical changes of fractures at those sites (see Fig. 64-4).171 Radionuclide scans may be used to make the diagnosis of fracture at any site for which the routine radiograph is not definitive. Radionuclide scans and magnetic resonance images can also be used in certain clinical settings to rule out diseases that produce hip pain, such as avascular necrosis of the femoral head. INTERVENTION IN POSTMENOPAUSAL OSTEOPOROSIS The first approach to the prevention or treatment of osteoporosis is risk reduction.172 Improving calcium intake and increasing physical activity are the most frequently adopted strategies. Indeed, for their general effects on health and their modest cost and low risk, these are strategies that should be recommended to all. In addition, elimination of cigarette use and moderation of alcohol intake are valuable. Limiting doses of drugs that adversely affect the skeleton and reviewing environmental safety concepts to reduce falling risk, particularly for elderly individuals, are also important. CALCIUM INTAKE All individuals should be advised of the recommended intake of calcium. The recommendations of the National Academy of Sciences (see Table 64-3) are somewhat lower than those previously suggested by the National Institutes of Health but are probably more easily achieved for those with average intakes (~50% of recommended intake).173 The improvement in intake should be obtained by improving diet. In the United States, this primarily means increasing consumption of dairy products, a move that is resisted by many on the grounds that it will result in increased fat or calorie intake. Opting for nonfat or low-fat products helps avoid this problem. Lactose intolerance is a problem that increases in frequency with age, although not all who claim lactose intolerance have true lactase deficiency. Those who cannot increase dietary calcium sufficiently in this way can consume calcium-fortified foods such as fruit juices, which have as much calcium per ounce as milk. Furthermore, many calcium supplement preparations are available. When advising about supplements, the clinician must remember that the aim is to increase intake to the recommended levels. A simple assessment of basal intake can be made easily and supplements can be added as required. For most individuals, an extra intake of 500 to 750 mg per day of elemental calcium is sufficient. When supplements are used, the least expensive is usually a calcium carbonate preparation with a USP label (which ensures solubility). Calcium supplements should be taken in divided doses. Calcium carbonate should be taken at the end of a meal when the acid load of the meal will assist the availability of the calcium (this is not required when calcium citrate is used). Although some suggest that calcium should be taken at bedtime to reduce the nocturnal increase in bone remodeling, no available data indicate a greater beneficial effect. Such increases in calcium intake are important at all ages and may improve peak bone mass and slow premenopausal loss. Although this strategy by itself does not prevent the loss of bone that is driven by estrogen deficiency, improving calcium intake among the elderly clearly reduces the risk of fracture significantly, perhaps by as much as 50%.112,113,114,115,116,117,118,119,120,121,122,123,124,125 and 126 PHYSICAL ACTIVITY The general belief that increased physical activity improves bone mass is based mainly on cross-sectional data.148,149,150,151 and 152 Longitudinal data suggest that in adults the effects of physical activity on bone density are modest, although most studies show some increase.173,174,175,176,177,178,179,180,181 and 182 Whether the effects of

physical activity on bone modify fracture risk is not known and cannot easily be tested. Exercise has other effects on the organism that are likely to yield positive effects on fracture risk (e.g., effects on the neuromuscular system). Improved reaction time and strength might be expected to reduce the risk of falls and, thereby, the risk of injury. Physical activity is a habit. To obtain the maximum benefit from any activity program, the earlier in life that it is begun and the longer in life it is continued, the better. Exercise has many health benefits beyond its effects on the skeleton; thus, the argument for increased activity at all ages is only enhanced by the skeletal effects that might accrue directly or indirectly. No specific exercise prescription for osteoporosis exists, and components of strength training as well as activities that improve cardiovascular performance are required. The authors recommend a pragmatic approach, recognizing that adherence to a program is more important than the program itself. The activities that the individual enjoys are recorded as part of the history-taking, and the patient is encouraged to increase their use. In addition, use of the buddy system is stressed, because participation with a friend makes adherence more likely. PHARMACOLOGIC INTERVENTION Agents available in the United States include bone-specific drugs (bisphosphonates, calcitonin) and agents that have more general effects (e.g., estrogens and tissue-selective estrogens).183,184 Each has different advantages and disadvantages, but their availability allows treatment to be tailored to individual needs. Estrogen. Hormone-replacement therapy (HRT; the terms hormone-replacement therapy and estrogen-replacement therapy are generally used interchangeably) has been available for >50 years and is considered to be the gold standard for the prevention and treatment of osteoporosis.183,184 Most available data support the conclusion that HRT is an effective intervention for osteoporosis prevention and treatment.185,186,187,188,189,190,191 and 192 In terms of prevention of vertebral fracture, one long-term primary prevention study has reported an ~75% protection rate over 10 years.185 Small secondary prevention studies, in which patients with osteoporosis were treated with hormones indicate an ~50% reduction in fracture risk.193 Epidemiologic data indicate an ~50% reduction in overall risk of hip fractures.194,195,196,197,198,199,200,201,202 and 203 One study suggested that the effect may be even greater among older women (older than 65 years) who have taken estrogen for >10 years.203 A 75% reduction in the risk of hip fracture was evident only among those women with a long history of estrogen therapy who were currently taking treatment. In a clinical trial evaluating the effects of estrogen on myocardial infarction in patients with established heart disease, no evidence was found of an effect of estrogen use on nonspinal fractures.204 Because the study was not specifically designed as an osteoporosis study, its importance is uncertain. However, it does suggest that some of the findings regarding estrogen effects on fractures may be related to the healthy user phenomenon205 that is, women who receive HRT are healthier and are more likely to have healthy lifestyles (i.e., no tobacco use, higher exercise activity, better diets). One study of 460 early menopausal women suggested that HRT did reduce the risk of nonspine fractures by ~60% compared to placebo.199 Further controlled studies are necessary to determine the effect of HRT on nonspinal osteoporosis-related fractures. Considerable clinical trial data indicate that estrogen intervention in women after menopause (surgical or natural) reduces the rate of bone remodeling to premenopausal levels and prevents the loss of bone mass.184,185,186,187,188,189,190,191 and 192,206,207,208,209,210,211,212,214 and 215 Most studies have followed patients for 2 to 3 years.185,208 HRT appears to be effective for as long as it is given, although some bone loss may occur in older individuals, in whom bone remodeling is being influenced by factors other than estrogen deficiency. Some data indicate that even in such aged individuals bone loss is partially sensitive to estrogen status and may be influenced by the low level of endogenous estrogen production.216,217 Thus, even in those older than 70 years of age, HRT can produce positive effects on the skeleton. Intervention with HRT is effective only while estrogen is being taken.203 Many individuals may not fill their prescriptions, and, of those that do, many (probably ~50%) fail to complete one year of treatment; thus, many fail to take estrogen for sufficiently long to derive the health benefits.218 For women who have passed a natural menopause, HRT is generally prescribed as a combination of estrogen and a progestin.219 The progestin may be given cyclically (younger women) or continuously (older women). The cyclic regimens are associated with more regular menstrual-type vaginal bleeding, whereas the continuous regimens induce amenorrhea. While continuous regimens are often associated with irregular bleeding during the early months of treatment, most women cease bleeding within 6 to 12 months. This bleeding pattern needs to be explained to the patient, because it can result in serious patient concerns about the safety of the medicine. Several different preparations of estrogen are available, and combined preparations are available that obviate the need for two separate medications. Progestins need not be taken by women who do not have a uterus. Controversy continues about the dosage of estrogen required to achieve a bone-sparing effect.191,220,221 Available data suggest that 0.625 mg per day conjugated equine estrogen (CEE) or its equivalent is >90% effective. In a metaanalysis of the estrogen studies performed with and without calcium supplementation, the effect of estrogen on bone density was found to be enhanced by adequate calcium intake.222 A controlled clinical trial evaluating the effect of 0.3 mg per day CEE (with adequate calcium and vitamin D intake) confirmed that, in the presence of good nutrition, this lower dosage of estrogen is sufficient. Similar results have been obtained with other estrogen products at equivalent dosages.220,223 Estrogens can be administered by oral, transdermal, buccal, vaginal, percutaneous, or subcutaneous routes. Provided that enough estrogen is supplied, an effect on bone remodeling and bone density occurs, although no studies on fractures have been published other than for oral or transdermal administration. No cross-comparison studies have been undertaken; thus, equivalency is based on studies of each compound in different populations, with differing trial designs. Estrogens have multiple tissue effects in the body and are still prescribed principally for the relief of menopausal symptoms.219,224 Consequently, much of the use is relatively short-term and insufficient to prevent osteoporosis. Although the hypothesis has been put forward that estrogen use is associated with reductions in the risk of myocardial infarction among postmenopausal women,224 the data are epidemiologic.205 In fact, one secondary prevention study (a controlled clinical trial) undertaken in women with established heart disease showed a transient increase in cardiac events during the early treatment and no favorable effect until 4 to 5 years had passed.204 This study has caused some reexamination of the data supporting estrogen use for heart disease. Estrogen also may have a protective effect against Alzheimer disease.225,226 and 227 Controlled clinical trial data do suggest that use of estrogens may improve cognitive function among postmenopausal women; however, the data regarding dementia are very preliminary.228 Estrogen use has been associated with an increased risk of breast cancer in some epidemiologic studies, findings that potentially limit long-term use of estrogens for prevention of osteoporosis.229 As might be expected, the prescription of estrogens for postmenopausal women is complicated, and often patients refuse this pharmacologic intervention. Bisphosphonates. Bisphosphonates are analogs of pyrophosphate in which the oxygen is replaced by a carbon atom,230 yielding a stable compound that resists metabolism and allows two side chains to be added to the carbon atom. These agents have varied affinities for bone and differing activities for inhibition of bone resorption. Controlled clinical trials with etidronate demonstrated increased bone density but failed to provide definitive fracture data.230a Thus, etidronate is marketed in the United States for Paget disease, hypercalcemia, and myositis ossificans but not for use in treatment of osteoporosis. Alendronate was the first of the bisphosphonates approved in the United States for osteoporosis treatment and prevention. Clinical trials demonstrated marked reductions in bone remodeling and increases in BMD of spine and hip231,232 and 233 (Fig. 64-5). These remodeling modifications were accompanied by reductions in the risk of fractures. In a large multicenter study, patients with prevalent vertebral fractures experienced reductions in the risk of vertebral fractures (~50%) and a reduction in hip and forearm fractures of ~50%.232 Patients with low bone density but without prevalent fractures experienced no overall reduction of clinical fractures (although radiographically demonstrated vertebral fractures were reduced by 44%)233; however, when the analysis was limited to those who had osteoporosis by bone density criteria, a significant reduction in clinical fractures was seen. This suggests that alendronate is best used for either those who have fractures at presentation or those who have osteoporosis.

FIGURE 64-5. Mean ( standard deviation) percentage changes in bone mineral density from baseline to 48 months in patients receiving alendronate therapy and in

those receiving a placebo. (From Cummings SR, Black DM, Thompson DE, et al. Effect of alendronate on risk of fracture in women with low bone density but without vertebral fractures: results from the Fracture Intervention Trial. JAMA 1998; 280:2077.)

For osteoporosis, the dose of alendronate approved for prevention is 5 mg and for treatment is 10 mg. A modest but significantly greater effect on BMD is seen with 10 mg, but either dose is probably equally effective232 (based on the fracture intervention trials in which the 5-mg dose was used for the first two years, with a change to 10 mg only for the latter portion of the study232,233). In the United States, both doses cost the same; therefore, the 10-mg dose is often preferred as the initial dose. Bisphosphonates are poorly absorbed (<1% of the ingested dose absorbed on an empty stomach).230 Food interferes with absorption. Consequently, alendronate must be taken in the morning on an empty stomach with water (68 ounces) 30 minutes before any food is consumed. The patient must also remain in the upright position (standing or sitting) to avoid reflux of the tablet into the esophagus. For some patients this rigid schedule is difficult, although most seem to manage once the rationale is explained. Esophagitis has been reported, particularly with the 10-mg dose.234,235 Amino-containing bisphosphonates are irritating (although little evidence of this was observed in the alendronate clinical trials).231,232 and 233 The incidence of such problems may be as much as 15% to 40%, but most cases are minor and can be resolved by temporarily discontinuing the drug for a week before resuming treatment. If symptoms recur, treatment should be stopped, and the patient should be changed to an alternative drug or the dose should be lowered to 5 mg. Rarely, a more serious esophageal erosion with ulceration can occur. This may lead to perforation (incidence 1%). Usually, esophageal side effects occur in the early months of treatment. Use of alendronate should be avoided in those with active upper gastrointestinal disease. Occasionally, patients report some general aches and pains during the early months of alendronate treatment. The cause is unclear and the symptoms usually are minor and self-limiting. Risedronate is approved for treatment and prevention of osteoporosis and corticosteroid-induced osteoporosis. Risedronate has a potent effect on bone mass, and turnover. Data show a rapid and sustained reduction in vertebral fracture occurrence and a reduction in all non-spine fractures, including those of the hip in patients with osteoporosis.236,237 Other bisphosphonates include tiludronate and, intravenously, pamidronate. The major clinical problem with bisphosphonates, as with other therapeutic agents used in the treatment of largely asymptomatic conditions (e.g., hypertension, hypercholesterolemia), is obtaining long-term compliance. The median duration of bisphosphonate use in the United States is ~6 months,238 which clearly seems inadequate to obtain long-term protection against fractures. This is similar to the median duration of use of HRT and calcitonin, again not providing long-term protection. Thus, a major advance in patient management would be to lengthen adherence to treatment. Calcitonin. Salmon calcitonin (SCT), originally given in the form of a subcutaneous injection, currently is most frequently used as an intranasal spray. Intranasal administration of SCT (200 units per day in a single spray, given to alternate nostrils on successive days) is approved for treatment of osteoporosis in the United States. Controlled clinical trials have confirmed that this dosage reduces bone turnover, although less than do the bisphosphonates. The effect on bone mass is also less. However, administration of this dosage for 5 years reduces vertebral fractures by 36%, a result that more closely matches the reduction in fracture risk with bisphosphonates than would be predicted from calcitonin's effects on bone mass or turnover.239 The major advantage to calcitonin is safety. No long-term safety concerns have arisen, either in clinical trials or from long experience with subcutaneous therapy. Only nasal irritation may occur, and this is usually mild. The major disadvantage is related to the modest effects on surrogate markers of efficacy, because the changes in bone mass or turnover are insufficient in most individuals to detect whether the treatment is having an effect. No data indicate that this agent significantly reduces the risk of fractures other than those of the vertebrae. Selective Estrogen-Receptor Modulators. Most data concerning efficacy of estrogens at reducing fracture risk are observational. Furthermore, even if all of the alleged beneficial effects of estrogen are confirmed (prevention of spine and other osteoporotic fractures, reduction of heart disease, and blunting of the cognitive decline in Alzheimer disease), concern still exists about breast cancer in many women. Agents are needed that can achieve some of the possible benefits of estrogen without the adverse side effects. Selective estrogen-receptor modulators (SERMs) bind to the estrogen receptor but have tissue-selective effects (estrogenic action in some tissues and estrogen antagonist action in others). Different SERMs have their own specific phenotypic profiles with differential potency and effect on different organ systems.240 The two most important SERMs currently available for postmenopausal women are tamoxifen and raloxifene. Only raloxifene has been approved for prevention of osteoporosis and is currently being evaluated for treatment of osteoporosis. Tamoxifen has beneficial effects on bone density and bone turnover in postmenopausal women. A breast cancer prevention trial involving 13,388 pre- and post-menopausal women showed a reduction in the risk of classic osteoporotic fractures in the tamoxifen group, although the reductions in hip and Colles fracture was not quite significant.241 Overall, the trend toward reduced clinical fracture was ~21% in women older than 50 years. Tamoxifen reduces bone density in premenopausal women and increases it in post-menopausal women.242 The use of tamoxifen is indicated in breast cancer and in women at increased risk for breast cancer. Postmenopausal women taking this agent can probably be considered to be on a bone-active drug. In such circumstances, monitoring with tests of BMD and/or bone turnover markers might be useful. Tamoxifen clearly reduces the risk of breast cancer in high-risk women, but the effects on heart and cere-brovascular disease are unclear, and tamoxifen use is associated with a number of benign uterine conditions as well as uterine malignancy. Unfortunately, as is the case for raloxifene and estrogens, the risk of venous thromboembolic disease is also increased (approximately three-fold relative risk).240 Raloxifene has been studied in several osteoporosis prevention trials as well as an osteoporosis treatment trial called MORE (Multiple Outcomes of Raloxifene Evaluation). In both the prevention and treatment cohorts, raloxifene was associated with small increments of bone density at all skeletal sites and reductions in bone turnover (as measured by both bone formation and bone resorption variables). Data from the 3-year point showed that the drug reduced the risk of vertebral fracture by 40% to 50%.243 However, no reduction in the risk of nonspinal fractures was observed. Raloxifene apparently has beneficial effects on serum lipoproteins (similar to those of tamoxifen240); however, the ultimate outcomes for heart and cerebrovascular disease are unknown. Raloxifene's reduction of the risk of invasive breast cancer by 76% over 3.5 years is quite compelling but needs to be confirmed in ongoing clinical trials. The drug does not increase the risk of uterine cancer.244 Both tamoxifen and raloxifene can increase the prevalence of hot flashes in postmenopausal women. However, absolute risk is still quite low in the normal ambulatory population.240 Experimental Therapies. Fluoride (sodium fluoride), originally used at dosages of 40 to 60 mg per day, has been in clinical development for a number of years; marked increases occurred in bone mass, most obvious in the spine. However, little evidence was seen for a reduction in fracture risk, even in the spine.245,246 In addition, uncontrolled clinical data indicated that fluoride therapy is associated with an increase in the risk of hip fracture. Use of lower-dose fluoride (slow-release fluoride), investigated in clinical trials, has led to increased bone mass that is more controlled, with a concomitant reduction in the risk of vertebral fractures.247 Currently, fluoride is not approved as therapy for osteoporosis in the United States. PTH increases vertebral bone mass when delivered by daily subcutaneous injection, either as the intact hormone or as the 1 34 amino-acid peptide.248 In studies in which PTH alone was used during 1 year of therapy, little if any effect was seen on bone mass in the hip. Indeed, total body bone mass declined by ~2%. These changes are probably related to activation of bone remodeling. PTH has also been administered in combination with antiresorptive therapy: Long-term studies in combination with HRT showed marked (1530%) increases in spinal BMD with accompanying, although smaller, increases in BMD of the hip and in total body bone mass (Fig. 64-6). In two studies, the risk of vertebral fractures appeared to be reduced.249,249a Furthermore, a substantial effect of PTH on all non-spine fractures was seen in the larger of these two studies.249a

FIGURE 64-6. Changes in bone mass during 3 years of therapy for group receiving parathyroid hormone (hPTH) plus estrogen and group receiving estrogen only. Level of statistical significance of differences between groups is indicated as follows: , p <.02; *, p <.001; , p <.02. (Lindsay R, Nieves J, Formica C, et al. Randomized controlled study of effect of parathyroid hormone on vertebral-bone mass and fracture incidence among postmenopausal women on oestrogen with osteoporosis. Lancet 1997; 350:550.)

OTHER FORMS OF PRIMARY OSTEOPOROSIS

OSTEOPOROSIS IN MEN Although osteoporosis is more common in women, it is not unusual in men. One-fifth of all hip fractures and approximately one-seventh of all vertebral fractures are seen in men. One in six men has a hip fracture by the age of 90.7,8 The lower incidence of osteoporosis in men may be attributed to the higher peak bone mass attained (~10% higher) and to the absence of a menopause equivalent in men. Adult men with a history of constitutional delay in puberty have lower radial and spinal bone mineral densities.250 Overall, when osteoporosis is encountered in a non-elderly man, the two most frequent causes are alcohol abuse and hypogonadism.43,44,134,135,251 Heavy smoking also is common in young men with osteoporosis. Preliminary data suggest that alendronate therapy might improve bone mass in osteoporotic men. JUVENILE OSTEOPOROSIS Juvenile osteoporosis is a rare, self-limiting disease of prepubertal children.252 These patients commonly present between the ages of 8 and 14 with the acute onset of back pain secondary to vertebral compression fractures. Spontaneous remission usually occurs after 2 to 6 years. Juvenile osteoporosis must be differentiated from osteogenesis imperfecta, Cushing syndrome, leukemia, and other disorders of bone marrow. Laboratory values are generally normal, as in the adult with osteoporosis. Although the temporal relation of juvenile osteoporosis to puberty implies an endocrine cause, the primary defect is unknown. Biopsy data suggest a relatively uncontrolled activity of metaphyseal osteoclasts. Analgesic therapy, early mobilization, supportive physical therapy, and preventive measures against further bone loss are used. Currently no treatment approved by the Food and Drug Administration exists for osteoporosis in children. Sex hormone therapy is not effective (and not advisable), and the use of bisphosphonates253 should not be recommended because of their long skeletal half-lives. IDIOPATHIC OSTEOPOROSIS IN YOUNG ADULTS Idiopathic osteoporosis, an uncommon condition, occurs between the ages of 30 and 50 and is more frequent in men than in women. The axial skeleton is most severely affected, and vertebral fractures are common. Serum laboratory values are normal, but hypercalciuria and consequent nephrolithiasis are often seen. The high urinary calcium losses may be associated with mild secondary hyperparathyroidism. The bone histology often shows active remodeling with excessive osteoclast activity or decreased bone formation.254 Exclusion of a late-onset form of osteogenesis imperfecta, even by biopsy, may be difficult. The treatment of this condition is the same as for other forms of osteoporosis, although it may be more rapidly progressive and disabling.254,255 REGIONAL OSTEOPOROSIS The classic example of regional osteoporosis is that following disuse or immobilization of a limb. Total immobilization in patients with fractures, motor paralysis, or chronic disabling ailments is associated with rapid and generalized loss of bone. Immobilization is linked to a rapid and significant increase in bone resorption and a slight increase in bone formation followed after a few weeks by a marked reduction in formation. The result is a high-turnover form of osteoporosis characterized by hypercalciuria and hyperphosphaturia.145,146 and 147 Occasionally, especially in the first few weeks, mild hypercalcemia and hyper-phosphatemia may develop. The process may be self-limited, and a new steady state may be reached. However, severe bone loss of ~50% of skeletal mass sometimes occurs. Weight-bearing and positive-stress maneuvers may help to decrease the rate of bone loss, but additional vascular, neurogenic, and humoral factors certainly play a role in the process. Radiologic evidence of osteoporosis may appear within 2 months of the immobilization, and a regional distribution over the areas immobilized is characteristic. If the immobilization time is not so prolonged as to result in permanent loss of bone, remineralization can occur on mobilization. Rheumatoid arthritis and infectious arthritis also may produce a localized osteoporosis adjacent to the affected joints (Fig. 64-7). Sometimes, regional osteoporosis occurs without a history of immobilization (Fig. 64-8).

FIGURE 64-7. Regional osteoporosis in a 64-year-old man with relative immobilization caused by severe rheumatoid arthritis. Bilateral, symmetric loss of the joint spaces of the knees is seen, with large subcortical bone cysts. Marked juxta-articular osteoporosis is present. The patient had a generalized diminution of bone density on radiography. In such cases, the hypothesis has been that the inflammatory process produces bone-resorbing growth factors. (The patient had saphenous vein surgery related to a coronary artery bypass graft procedure.)

FIGURE 64-8. Regional osteoporosis of the right foot of a 57-year-old white woman. The patient had no history of trauma or immobilization of the involved limb, and the

cause was obscure. (R, right; L, left.)

REFLEX SYMPATHETIC DYSTROPHY Reflex sympathetic dystrophy (RSD) is a distinct entity that is also called Sudeck atrophy, osteodystrophy, causalgia, acute bony atrophy, posttraumatic osteoporosis, and shoulder-hand syndrome.256 It follows conditions such as myocardial infarction, trauma, limb surgery, neoplasms, calcific tendinitis, vasculitis, and degenerative disease of the cervical spine. Overactivity of the sympathetic nervous system, with secondarily increased blood flow and activation of sensory fibers, has been suggested as the pathogenetic mechanism. The clinical picture is variable, although it commonly involves the shoulder and the hand with swelling, vasomotor changes (i.e., spasm or dilatation), and hyperesthesia. The symptoms are usually unilateral, but bilateral involvement can occur. The resolution of swelling and vasomotor changes occurs within several months, with the appearance of skin pigmentation, hyperhidrosis, hypertrichosis, skin and nail atrophic changes, and contracture. These phenomena may persist for years or gradually disappear. Radiographs and bone density measurements reveal localized osteoporosis, and soft-tissue swelling is evident in RSD patients. Bone resorption is characteristically seen. Radionuclide imaging reveals increased uptake, which is probably related to increased vascularity. Magnetic resonance imaging tests are normal or nonspecific.257 Histologic examination reveals accelerated bone resorption with excessive osteoclast activity, and intense synovitis is often evident. Treatment is supportive, although corticosteroids have been useful in alleviating pain and decreasing radionuclide uptake in bones and joints. Calcitonin has been used with good symptomatic response, and attempts have been made to attenuate the sympathetic tone with variable success.258 TRANSIENT OSTEOPOROSIS OF THE HIP Transient osteoporosis of the hip, first described in patients in the third trimester of pregnancy, was subsequently reported in young and middle-aged men and nonpregnant women.259 The hip pain begins spontaneously, without an antecedent history of trauma or infection. Radiographic films show osteoporosis of the periarticular bones of the hip joint, especially in the femoral head. No diminution of the joint space occurs. Bone density measurement of the hip may show some localized bone loss. Magnetic resonance imaging and radionuclide bone scan may be valuable in establishing the diagnosis and in follow-up care.260 The clinical course is self-limited, with complete regression of the radiographic and clinical signs and symptoms within months to 1 year. The cause is unknown, and the treatment is supportive and aimed at symptomatic relief. REGIONAL MIGRATORY OSTEOPOROSIS In the syndrome of regional migratory osteoporosis, local pain and swelling rapidly develop in the ankle, foot, or knee and are accompanied by the radiographic appearance of localized osteoporosis in those areas.261 Spontaneous recovery occurs after weeks or months, but with recurrence in other regions of the same or opposite extremity. These events may be repeated over a few years. Radionuclide bone scanning reveals increased activity in involved areas. The cause is unknown. The differential diagnosis includes rheumatoid arthritis, but involvement of the joint space is lacking, as is symmetric involvement of multiple joints. Other monoarticular syndromes, such as septic arthritis, pigmented villonodular synovitis, and idiopathic synovial osteochondromatosis, must be considered in the differential diagnosis; however, these conditions follow a course characterized by rapid joint destruction with only minimal secondary osteoporosis. The treatment of regional migratory osteoporosis is supportive, and includes use of antiinflammatory agents and occasional short courses of glucocorticoids. Calcitonin therapy has been associated with some clinical improvement.

OTHER OSTEOLYTIC SYNDROMES

A number of rare syndromes can lead to some degree of bone lysis. These osteolytic syndromes can result from infectious, neoplastic, traumatic, metabolic, vascular, congenital, and genetic causes. Hyperparathyroidism and Paget disease can cause significant osteolysis as well. The most common form of secondary osteoporosis is cortico-steroid-induced osteoporosis262; however, many other drugs, other endocrinopathies, and genetic diseases produce osteoporotic-like syndromes in which bone mass is reduced sufficiently to increase the risk of fracture. Frequently, secondary forms of osteoporosis coexist with the primary form of the disease. Recognition of the secondary form is important because treatment of the primary condition or elimination of the drug providing the insult to the skeleton can improve the patient's outcome. CORTICOSTEROID-INDUCED OSTEOPOROSIS Corticosteroid-induced osteoporosis is the most common form of secondary osteoporosis. It may also be the presenting sign of Cushing syndrome. In such situations, the clinical onset may be heralded by a cluster of vertebral and rib fractures. Usually, the patient has overt signs of Cushing syndrome. The more common form of corticosteroid-induced osteoporosis occurs in individuals who require long-term pharmacologic doses of steroid.263 The higher incidence in women of rheumatic diseases that may require corticosteroid therapy (e.g., rheumatoid arthritis, systemic lupus erythematosus [SLE], temporal arteritis, multiple sclerosis), coupled with the relatively compromised skeletal status of the aging female, explains why corticosteroid-induced osteoporosis is more common among women. Nonetheless, corticosteroid-induced osteoporosis can occur in both sexes and in connection with any disorder (e.g., chronic lung disease requiring long-term corticosteroid therapy). Prospective studies have revealed that bone loss can be rapid, with significant loss occurring within the first 3 to 6 months of therapy. Surprisingly, however, in some individuals bone loss does not occur, even on very high dosages of cortico-steroid. In other individuals, bone mass may be lost on prednisone regimens as low as 5 mg per day. Bone loss also occurs in patients with Addison disease on standard corticosteroid-replacement therapy. Corticosteroid-induced bone loss occurs throughout the skeleton but is generally greater in trabecular than in cortical bone. One-third or more of corticosteroid-treated patients have vertebral fractures, and the risk of hip fracture is 50% greater than in controls. Glucocorticoid excess causes bone loss by several mechanisms. These drugs inhibit the synthesis of bone matrix by the osteoblast and slow down the recruitment of newer osteoblasts. Histomorphometric studies have shown decreased mineral apposition and decreased width of trabecular packets. Osteocalcin levels in the blood are reduced within 1 day of initiation of glucocorticoid therapy and generally remain suppressed for as long as the therapy continues.264 In addition, bone resorption may be increased. Secondary hyperparathyroidism may be present, for which several potential causes exist: (a) glucocorticoids reduce the efficiency of calcium absorption by a direct effect on the intestinal cells, (b) glucocorticoids may increase obligatory urinary calcium loss, and (c) density of the 1,25(OH)2D receptor in the cell may be reduced. The compensatory secondary hyperparathyroidism that follows is thought to be responsible for accelerating bone loss. In addition, glucocorticoid therapy may reduce circulating levels of estradiol or testosterone through inhibitory effects on the adrenal gland and perhaps also direct effects on the ovary and intestine, exacerbating sex hormone deficiency. The principles of treatment for glucocorticoid-induced osteoporosis are similar to those for postmenopausal osteoporosis. Limitation of dosage and duration of corticosteroid therapy are important. Alternate-day regimens may be beneficial. Patients on corticosteroids should receive calcium supplementation (1.2 g per day) and vitamin D (800 U per day). Physical activity should be encouraged, if possible, because the adverse effects of immobilization may exacerbate the situation. Whether administration of calcitriol is beneficial remains inconclusive.265 Controlled clinical trials have demonstrated preservation of bone mass using bisphosphonate therapy. Furthermore, two separate clinical trials (one using daily doses of alendronate and the other, intermittent cyclical therapy with etidronate) showed a decreased risk of vertebral compression deformity in postmenopausal women. Risedronate also increases bone mass and reduces vertebral fracture occurence in women with steroid-induced osteoporosis. HRT therapy may be appropriate for postmenopausal women, provided it is not contraindicated by the primary diagnosis (e.g., SLE). Patients with corticosteroid-induced osteoporosis perhaps might benefit most from use of an anabolic agent capable of stimulating osteoblast synthesis of new bone. Controlled clinical trials of PTH therapy demonstrated marked increments in bone mass in patients with a variety of rheumatologic disorders who were on baseline treatment with HRT. However, PTH is not currently marketed for use in the United States. PRIMARY HYPERPARATHYROIDISM

Although the classic bone disease of primary hyperparathyroidism is osteitis fibrosa cystica, generalized skeletal demineralization without osteitis may be seen. Bone densitometry has uncovered a higher prevalence of subclinical skeletal involvement, usually a preferential loss of cortical bone, in patients with primary hyperparathyroidism151 (see Chap. 58). Indeed, spinal BMD might be specifically preserved in mild forms of the disorder. HYPERTHYROIDISM Hyperthyroidism consistently increases bone turnover, and mild hypercalcemia occurs in ~10% of cases. Both bone formation and resorption are usually increased. A mild and prolonged hyperthyroid state may aggravate an underlying process of bone loss, and the correction of hyperthyroidism may lead to some restoration of bone mass. Hypothyroidism is prevalent in the elderly, and overreplacement with thyroid hormone may accelerate bone loss and uncover clinical osteoporosis.266 Because the daily requirements for thyroid hormone may decrease with age, a periodic reassessment of the replacement dose is appropriate in this age group. AMENORRHEA AND ANOREXIA NERVOSA Hypogonadal states are associated with osteoporosis.267,268 These include primary gonadal disorders, such as in Turner and Klinefelter syndromes, and secondary gonadal failure, as in patients with pituitary tumors and hypothalamic pituitary insufficiency. The latter may be associated with exercise-induced or stress-induced amenorrhea. Many studies have shown that rapid and profound bone loss occurs with estrogen deficiency in younger women. For example, a 5% to 10% loss in spinal and femoral bone density occurs within 6 months of gonadotropin-releasing hormone agonist therapy.269 Similarly, rapid bone loss is seen with exercise-induced amenorrhea.270,271 In patients with anorexia nervosa, bone loss is related to the accompanying amenorrhea but is not fully corrected with estrogen treatment alone, a finding that indicates the importance of nutritional and perhaps genetic factors.272 HYPERPROLACTINEMIA When hyperprolactinemia is associated with infertility, aggressive management is the rule, and in women, an attempt should be made to restore menses. When fertility is not important, any decision to withhold bromocriptine therapy for a microadenoma should include consideration of the concomitant hypogonadism and the subsequent increased risk of osteoporosis. Moreover, some studies have suggested a direct negative effect of prolactin on bone, independent of the presence of hypogonadism.273 An oral contraceptive containing estrogen should be given to patients with untreated microadenomas associated with amenorrhea or oligomenorrhea. TYPE 1 (INSULIN-DEPENDENT) DIABETES A reduction in bone mass is seen in patients with type 1 (insulin-dependent) diabetes mellitus.274 This may become detectable as early as 2 to 3 years after the onset of clinical diabetes. A similar reduction in bone mass is not usually seen in diabetes mellitus type 2. Hypercalciuria is common in type 1 diabetics, but plasma levels of PTH and vitamin D metabolites are generally normal. Subtle changes have been reported in some patients, however, suggesting a relative deficiency of vitamin D metabolites and possibly calcium malabsorption. Insulin deficiency may reduce bone formation by means of diminished synthesis of collagen by the osteoblasts. In patients with type 1 diabetes, hypercalciuria and reduced tubular reabsorption of phosphate are improved by strict glucose control. ACROMEGALY Osteoporosis in patients with acromegaly is related in part to a concomitant hypogonadal state. Nevertheless, a unique form of high-turnover bone disease, which predominantly affects cortical bone, appears to be connected to the increased serum levels of growth hormone (presumably mediated through insulin-like growth factors).275 Radiographically, new periosteal bone formation and a characteristic increased anteroposterior diameter of the vertebral bodies, with significant osteophyte formation, are noted. In the absence of hypogonadism, trabecular bone involvement is uncommon. Hyperphosphatemia, a well-known feature of acromegaly, probably results from growth hormoneinduced stimulation of renal tubular reabsorption of phosphate. Growth hormone also stimulates the production of 1,25(OH)2D and directly stimulates calcium absorption, both of which may lead to hypercalciuria.276 OSTEOGENESIS IMPERFECTA Osteogenesis imperfecta is a common heritable disorder of bone, found in 1 of 20,000 live births. This disease can be diagnosed by a skin biopsy followed by fibroblast culture and collagen analysis. A blood test for certain genotypes is also available. There may be forms of the disease (tarda) that do not manifest abnormalities until later in life. Other phenotypic features can include blue sclerae, arcus senilis, and dental and hearing disorders. The disappearance of sex hormone protection with the menopause may cause a second rise in fracture incidence among women with the disorder; these women may benefit from HRT.277 OTHER GENETIC DISORDERS Other genetic disorders involving defective collagen synthesis or collagen cross-linking may be associated with osteoporosis. These include homocystinuria, Ehlers-Danlos syndrome, Marfan syndrome, and Menkes syndrome (Fig. 64-9). Congenital anemias that cause bone marrow hyperplasia, such as thalassemia and sickle cell anemia, frequently are accompanied by osteoporosis. Glycogen storage diseases, hypophosphatasia, hemochromatosis, and familial dysautonomia (i.e., Riley-Day syndrome) are associated with osteoporosis (see Chap. 66).

FIGURE 64-9. A 46-year-old man with Ehlers-Danlos syndrome, a heritable connective tissue disorder. The patient had marked osteoporosis. The patient had soft and atrophic skin that was hyperextensible and marked hypermobility of the joints.

ALCOHOLISM An association is seen between marked alcohol consumption and the development of osteoporosis. Whether mild or moderate alcohol consumption is associated with a negative effect on the skeleton is unknown; one large study suggested that such an intake is not detrimental.277a Alcohol abuse can cause calcium and vitamin D malabsorption secondary to pancreatic or liver disease and may inhibit calcium absorption directly. It may also result in hypercalciuria through a direct renal effect. The poor nutritional habits of alcoholics frequently result in low calcium intake. Alterations in liver function may produce abnormalities in vitamin D metabolism, with reduced 25(OH)D and 1,25(OH)2D levels. PTH secretion or function may be impaired directly by ethanol or indirectly by concomitant hypo-magnesemia. Moreover, gonadal function is often impaired in alcoholics, with deficient testosterone production in men. Alcohol appears to exert a direct, toxic depressant effect on osteoblast function. Several studies have demonstrated a reduction in skeletal mass (mostly in trabecular bone) in alcoholics. Some histologic studies have found evidence of osteomalacia, although this usually occurs in the presence of gross malnutrition. In healthier alcohol abusers, bone histology shows a profound reduction in bone formation and resorption, which strongly suggests a direct effect of alcohol on bone cells.

MULTIPLE MYELOMA AND OTHER MALIGNANCIES Multiple myeloma is often associated with generalized osteoporosis. Spinal crush fractures, back pain, and diffuse skeletal demineralization seen on radiographic films are often the initial clinical features of the illness. The production of osteoclast-activating factors (or humoral hypercalcemic factors) by plasma cells may mediate the demineralization. Diffuse osteoporosis may also occur as a result of bone marrow infiltration by carcinoma cells from other malignancies. Moreover, the use of chemotherapeutic agents such as methotrexate and other antimetabolites has been associated with the development of osteoporosis.278 PREGNANCY AND LACTATION A specific and rare syndrome of osteoporosis has been associated with pregnancy and lactation, during which fractures may occur.279 Pregnancy and lactation are generally periods during which calcium is passed from the mother to the fetus and infant. The skeleton of the newborn infant contains almost 30 g of calcium; another 40 g are passed on through the milk of the mother over 3 months of lactation. However, pregnancy is accompanied by regulatory defense mechanisms, including increased calcium absorption and possibly increased calcitonin secretion. Whether similar mechanisms occur during lactation is unknown, but maternal skeletal mass may increase during pregnancy in preparation for the calcium demands of lactation.280 Nulliparity and prolonged lactation are often cited as risk factors for osteoporosis. In the osteoporosis syndrome that develops during pregnancy and lactation, a deficiency of 1,25(OH)2D has been suggested. Bone histology studies have not revealed evidence for osteomalacia or an increase of bone turnover. The syndrome is self-limiting, and fractures may not occur over prolonged periods of follow-up despite subsequent pregnancies. GASTROINTESTINAL DISEASE A variety of gastrointestinal diseases are associated with osteoporosis and osteomalacia due to malabsorption of calcium or vitamin D. Whether osteoporosis or osteomalacia dominates depends on the severity of the malabsorption process. Subtotal gastrectomy, particularly of the Billroth II type, is associated with the development of osteomalacia and osteoporosis in 5% to 10% of cases many years after the surgery. The mechanism may be that of calcium malabsorption, similar to that which occurs with achlorhydria and old age. The role of lactase deficiency with milk intolerance and resultant calcium deficiency may also be important in the development of osteoporosis. OBSTRUCTIVE LIVER DISEASE Chronic obstructive liver disease may interfere with the enterohepatic circulation of vitamin D metabolites and predispose the patient to osteomalacia.281 Osteoporosis, more often than osteomalacia, is associated with primary biliary cirrhosis, and histomorphometric studies reveal inactive remodeling with depressed bone formation. HEPARIN-INDUCED OSTEOPOROSIS Chronic heparin administration (15,000 U or more daily) for 6 months or longer has been associated with the development of osteoporosis and increased fractures.282 The changes may be reversible with cessation of therapy. Heparin activates lysosomal bone resorption as shown in vitro. It may also inhibit bone formation. Moreover, systemic mastocytosis has been associated with generalized demineralization in this disorder. Abnormally proliferating mast cells are thought to produce heparin, although a role for histamine and other mediators has not been excluded.283 ANTICONVULSANT THERAPY Diphenylhydantoin has been associated with a syndrome of osteomalacia that is thought to be related to abnormal vitamin D metabolism in the liver.284 A smaller effect has also been reported with phenobarbital. Phenytoin and carbamazepine may also have more subtle direct effects on bone cells, resulting in osteopenia, particularly in the hip region. Increases in the markers of bone formation and resorption have been seen.285 CHAPTER REFERENCES
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. Consensus development conference: prophylaxis and treatment of osteoporosis. Osteoporos Int 1991; 1:114. Peck WA. Consensus development conference: diagnosis, prophylaxis, and treatment of osteoporosis. Am J Med 1993; 94:646. Melton LJ III. How many women have osteoporosis now? J Bone Miner Res 1995; 10:175. Melton LJ III, Riggs BL. Epidemiology of age-related fractures. In: Avioli LV, ed. The osteoporotic syndrome. New York: Grune & Stratton, 1987:1. Chalmers J, Ho KC. Geographical variations in senile osteoporosis. The association with physical activity. J Bone Miner Res 1970; 52B:667. Wallace WA. The increasing incidence of fractures of the proximal femur: an orthopedic epidemic. Lancet 1983; 1:1413. Cooper C, Melton LJ. Epidemiology of osteoporosis. Trends Endocrinol Metab 1992; 3:224. Kanis JA, McCloskey EV. Epidemiology of vertebral osteoporosis. Bone 1992; 13(Suppl 2):S1. Riggs BL, Melton LJ III. Medical progress: involutional osteoporosis. N Engl J Med 1986; 314:1676. Richelson LS, Wahner HW, Melton LJ III, Riggs BL. Relative contribution of aging and estrogen deficiency to postmenopausal bone loss. N Engl J Med 1985; 311:1273. Cummings SR, Black DM, Nevitt MC, et al. Bone density at various sites for prediction of hip fractures. Lancet 1993; 341:72. Ross PD, Davis JW, Vogel JM, Wasnich RD. A critical review of bone mass and the risk of fractures in osteoporosis. Calcif Tissue Int 1990; 46:149. Dempster DW, Lindsay R. Pathogenesis of osteoporosis. Lancet 1993; 341:797. Kelsey JL, Browner WS, Seeley DG, et al. Risk factors for fractures of the distal forearm and proximal humerus. Am J Epidemiol 1992; 135:477. Ross PD, Davis JW, Epstein RS, Wasnich RD. Pre-existing fractures and bone mass predict vertebral fracture incidence in women. Ann Intern Med 1991; 114:919. Cummings SR. Treatable and untreatable risk factors for hip fracture. Bone 1996; 18(Suppl 3):165S. Ross PD, Genant HK, Davis J, et al. Predicting vertebral fracture incidence from prevalent fractures and bone density in older women. Osteoporos Int 1993; 3:120. Sowers MR, Galuska DA. Epidemiology of bone mass in premenopausal women. Epidemiol Rev 1993; 15:374. Sambrook N, Kelly PJ, White CP, et al. Genetic determinants of bone mass. In: Marcus R, Feldman D, Kelsey J, eds. Osteoporosis. San Diego: Academic Press, 1996:477. Southard R, Morris J, Mahan J, et al. Bone mass in healthy children measurement with quantitative DXA. Radiology 1991; 179:735. Katzman DK, Bachrach LK, Carter DR, Marcus R. Clinical and anthropometric correlates of bone mineral acquisition in healthy adolescent girls. J Clin Endocrinol Metab 1991; 73:1332. Bonjour JP, Theintz G, Buchs B, et al. Critical years and stages of puberty for spinal and femoral bone mass accumulation during adolescence. J Clin Endocrinol Metab 1991; 73:555. Gilsanz V, Boechat I, Roe TF, et al. Gender differences in vertebral body sizes in children and adolescents. Radiology 190; 1994:673. Gilsanz V, Roe TF, Mora S, et al. Changes in vertebral bone density in black girls and white girls during childhood and puberty. N Engl J Med 1991; 325:1597. Matkovic V, Fontana D, Tominac C, et al. Factors that influence peak bone mass formation: a study of calcium balance and the inheritance of bone mass in adolescent females. Am J Clin Nutr 1990; 52:878. Bonjour JP, Rizzoli R. Bone acquisition in adolescence In: Marcus R, Feldman D, Kelsey J, eds. Osteoporosis. San Diego: Academic Press, 1996:477. Steinberg KK, Frein-Titulaer LW, DePuey EG, et al. Sex steroids and bone density in premenopausal and perimenopausal women. J Clin Endocrinol Metab 1989; 69:533. Sowers MF. Premenopausal reproductive and hormonal characteristics and the risk for osteoporosis. In: Marcus R, Feldman D, Kelsey J, eds. Osteoporosis. San Diego: Academic Press, 1996:477. Sowers M, Crutchfield M, Bandekar R, et al. Bone mineral density and its change in pre- and perimenopausal white women: the Michigan bone health study. J Bone Miner Res 1998; 13(7):1134. Meema HE. Cortical bone atrophy and osteoporosis as a manifestation of aging. AJR Am J Roentgenol 1963; 89:1287. Mazess RB, Mather W. Bone mineral content of North Alaskan Eskimos. Am J Clin Nutr 1974; 27:916. Parfitt AM. Skeletal heterogeneity and the purposes of bone remodeling: implications for the understanding of osteoporosis. In: Marcus R, Feldman D, Kelsey J, eds. Osteoporosis. San Diego: Academic Press, 1996:477. Heaney RP, Recker RR, Saville PD. Menopausal changes in bone remodeling. J Lab Clin Med 1978; 92:964. Eriksen EF. Normal and pathological remodeling of human trabecular bone: three dimensional reconstruction of the remodeling sequence in normals and in metabolic disease. Endocr Rev 1986; 4:379. Parfitt AM. Age-related structural changes in trabecular and cortical bone: cellular mechanism and biomechanical response. Calcif Tissue 1984; 36:8123. Chiodini I, Carnevale V, Torlontano M, et al. Alterations of bone turnover and bone mass at different skeletal sites due to pure glucocorticoid excess: study in eumenorrheic patients with Cushing's syndrome. J Clin Endocrinol Metab 1998; 83(6):1863. Krempien B, Lemminger FM, Ritz E, Weber E. The reaction of different skeletal sites to metabolic bone diseasea micromorphometric study. Klin Wochenscher 1978; 56:755. Heaney RP, Recker RR, Saville PD. Menopausal changes in calcium balance. J Lab Clin Med 1978; 92:953. Hui SL, Zhou L, Evans R, et al. Rates of growth and loss of bone mineral in the spine and femoral neck in white females. Osteoporos Int 1999; 9:200. Johnston CC Jr, Hui SL, Witt RM, et al. Early menopausal changes in bone mass and sex steroids. J Clin Endocrinol Metab 1985; 61:905. Kusec V, Virdi AS, Prince R, Triffitt JT. Localization of estrogen receptor-a in human and rabbit skeletal tissues. J Clin Endocrinol Metab 1998; 83(7):2421. Onoe Y, Miyaura C, Ohta H, Nozawa S, Suda T. Expression of estrogen receptor b in rat bone. Endocrinology 1997; 138:4509. Arts J, Kuiper GGJM, Janssen JMMF, et al. Differential expression of estrogen receptors a and b mRNA during differentiation of human osteoblast SV-HFO cells. Endocrinology 1997; 138:5067. Mundy GR, Boyce BF, Yoneda T, et al. Cytokines and bone remodeling. In: Marcus R, Feldman D, Kelsey J, eds. Osteoporosis. San Diego: Academic Press, 1996:477. Horowitz MC. Cytokines and estrogens in bone: anti-osteoporotic effects. Science 1993; 260:626. Girasole G, Jilka RL, Passeri G, et al. 17b-estradiol inhibits interleukin-6 production by bone marrow derived stromal cells and osteoblasts in vitro: a potential mechanism for the antiosteoporotic effect of estrogens. J Clin Invest 1992; 89:883. Cheleuitte D, Mizuno S, Glowacki J. In vitro secretion of cytokines by human bone marrow: effects of age and estrogen status. J Clin Endocrinol Metab 1998; 83:2043. Cohen-Solal ME, Boitte F, Bernard-Poenaru O, et al. Increased bone resorbing activity of peripheral monocyte culture supernatants in elderly women. J Clin Endocrinol Metab 1998; 83:1687. Chen MM, Yeh JK, Aloia JF. Anabolic effect of prostaglandin E 2 in bone is not dependent on pituitary hormones in rats. Calcif Tissue Int 1998; 63:236. Kodama Y, Takeuchi Y, Suzawa M, et al. Reduced expression of interleukin-11 in bone marrow stroma cells of senescence-accelerated mice (SAMP6): relationship to osteopenia with enhanced adipogenesis. J Bone Miner Res 1998; 13:1370. Shimizu T, Mehdi R, Yoshimura Y, et al. Sequential expression of bone morphogenetic protein, tumor necrosis factor, and their receptors in bone-forming reaction after mouse femoral marrow

ablation. Bone 1998; 23:127. 52. Raisz LG. Local and systemic factors in the pathogenesis of osteoporosis. N Engl J Med 1988; 318:818. 53. Pacifici R, Rifas L, Teitelbaum S, et al. Spontaneous release of interleukin-1 from human blood monocytes reflects bone formation in idiopathic osteoporosis. Proc Natl Acad Sci USA 1987; 84:4616. 54. Pacifici R, Brown C, Puscheck E, et al. Effect of surgical menopause and estrogen replacement on cytokine release from human blood mononuclear cells. Proc Natl Acad Sci USA 1991; 88:514. 55. Gowen M, Wood DD, Ihrie EJ, et al. An interleukin-1 like factor stimulates bone resorption in vitro. Nature 1989; 306:378. 56. Stock JL, Coderre JA, MacDonald B, Rosenwasser LJ. Effects of estrogen in vivo and in vitro on spontaneous interleukin-1 release by monocytes from postmenopausal women. J Clin Endocrinol Metab 1989; 68:364. 57. Jilka RL, Hangoe G, Girasole G. Increased osteoclast development after estrogen loss: mediation by interleukin-6. Science 1992; 257:88. 58. Matsuzaki K, Udagawa N, Takahashi N, et al. Osteoclast differentiation factor (ODF) induces osteoclast-like cell formation in human peripheral blood mononuclear cell cultures. Biochem Biophys Res Commun 1998; 246:199. 59. Mizuno A, Amizuka N, Irie K, et al. Severe osteoporosis in mice lacking osteoclastogenesis inhibitory factor/osteoprotegerin. Biochem Biophys Res Commun 1998; 247:610. 60. Yasuda H, Shima N, Nakagawa N, et al. Osteoclast differentiation factor is a ligand for osteoprotegerin/osteoclastogenesis-inhibitory factor and is identical to TRANCE/RANKL. Proc Natl Acad Sci U S A 1998; 95:3597. 61. Turner RT, Riggs BL, Spelsberg TC. Skeletal effects of estrogen. Endocr Rev 1994; 15:275. 62. Uebelhart D, Schlemmer A, Johansen JS, et al. Effect of menopause and hormone replacement therapy on the urinary excretion of pyridinium crosslinks. J Clin Endocrinol Metab 1991; 72:367. 63. Hassager C, Fabbri-Mabelli G, Christiansen C. The effect of the menopause and hormone replacement therapy on serum carboxyterminal propeptide of type I collagen. Osteoporos Int 1993; 3:50. 64. Calvo MS, Eyre DR, Gundberg CM. Molecular basis and clinical application of biological markers of bone turnover. Endocr Rev 1996; 17(4):333. 65. Hansen M. Assessment of age and risk factors on bone density and bone turnover in healthy premenopausal women. Osteoporos Int 1994; 4:123. 66. Mazzuoli G, Minisola S, Valtorta C, et al. Changes in mineral content and biochemical bone markers at the menopause. Isr J Med Sci 1985; 21:875. 67. Stepan JJ, Pospichal J, Presl J, Pacovsky V. Bone loss and biochemical indices of bone remodeling in surgically induced postmenopausal women. Bone 1987; 8:279. 68. Seibel MJ, Cosman F, Shen V, et al. Urinary hydroxypyridinium crosslinks of collagen as markers of bone resorption and estrogen efficacy in post-menopausal osteoporosis. J Bone Miner Res 1993; 8:881. 69. Garnero P, Sornay-Rendu E, Chapuy MC, Delmas PD. Increased bone turnover in late postmenopausal women is a major determinant of osteoporosis. J Bone Miner Res 1996; 11(3):337. 70. Cosman F, Nieves J, Wilkinson C, et al. Bone density change and biochemical indices of skeletal turnover. Calcif Tissue Int 1996; 58:236. 71. Blumsohn A, Eastell R. The performance and utility of biochemical markers of bone turnover: do we know enough to use them in clinical practice? Ann Clin Biochem 1997; 34:449. 72. Rosen CJ, Chesnut III CH, Mallinak NJS. The predictive value of biochemical markers of bone turnover for bone mineral density in early postmenopausal women treated with hormone replacement or calcium supplementation. J Clin Endocrinol Metab 1997; 82(6):1904. 73. Hannon R, Blumsohn A, Naylor K, Eastell R. Response of biochemical markers of bone turnover to hormone replacement therapy: impact of biological variability. J Bone Miner Res 1998; 13(7):1124. 74. Schott AM, Cormier C, Hans D, et al. How hip and whole-body bone mineral density predict hip fracture in elderly women: the EPIDOS prospective study. Osteoporos Int 1998; 8:247. 75. Cummings SR, Black DM, Nevitt MC, et al. Bone density at various sites for prediction of hip fractures. Lancet 1993; 341:72. 76. Wasnich RD, Ross PD, Heilburn LK, et al. Prediction of postmenopausal fracture risk with use of bone mineral measurements. Am J Obstet Gynecol 1985; 153:745. 77. Melton LJ III, Atkinson EJ, O'Fallon WM, et al. Long-term fracture prediction of bone mineral assessed at different sites. J Bone Miner Res 1993; 10:1227. 78. Washnich RD, Ross PD, Heilbrun LK, Vogel JM. Prediction of postmenopausal fracture risk with use of bone mineral measurements. Am J Obstet Gynecol 1985; 153:745. 79. Nevitt MC, Johnell O, Black DM, et al. For the Study of Osteoporotic Fractures Research Group. Bone mineral density predicts non-spine fractures in very elderly women. Osteoporos Int 1994; 4:325. 80. Gardsell P, Johnell O, Nilsson BE. Predicting fractures in women by using forearm bone densitometry. Calcif Tissue Int 1989; 44:235. 81. Hui SL, Slemenda CW, Johnston CC. Baseline measurements of bone mass predict fracture in white women. Ann Intern Med 1989; 111:355. 82. Dempster DW, Shane E, Horbett W, Linday R. A simple method for correlative light and scanning electron microscopy of human iliac crest bone biopsies: qualitative observations in normal and osteoporotic subjects. J Bone Miner Res 1986; 1:15. 83. Garnero P, Hausherr E, Chapuy MC, et al. Markers of bone resorption predict hip fracture in elderly women: the EPIDOS prospective study. J Bone Miner Res 1996; 11:1531. 84. Van Daele PLA, Seibel MJ, Burger H, et al. Case-control analysis of bone resorption markers, disability, and hip fracture risk: The Rotterdam Study. BMJ 1996; 312:482. 85. Akesson K, Ljunghall S, Jonsson B, et al. Assessment of biochemical markers of bone metabolism in relation to the occurrence of fracture: a retrospective and prospective population-based study of women. J Bone Miner Res 1995; 10:1823. 86. Ross PD, Wasnich RD, Knowlton WK. Skeletal alkaline phosphatase (Tandem-R Ostase) measurements predict vertebral fractures: a prospective study. (Abstract). J Bone Miner Res 1997; 12(Suppl 1):S150. 87. Delmas PD. The role of markers of bone turnover in the assessment of fracture risk in postmenopausal women. Osteoporos Int 1998; 8(Suppl 1):S32. 88. Looker AC, Wahner HW, Dunn WL, et al. Updated data on proximal femur bone mineral levels of US adults. Osteoporos Int 1998; 8:468. 89. Riggs BL, Melton LJ III. Evidence for two distinct syndromes of involutional osteoporosis. Am J Med 1983; 75:899. 90. Cummings SR, Browner WS, Bauer D, et al. For the Study of Osteoporotic Fractures Research Group. Endogenous hormones and the risk of hip and vertebral fractures among older women. N Engl J Med 1998; 339:733. 91. Ettinger B, Pressman A, Sklarin P, et al. Associations between low levels of serum estradiol, bone density, and fractures among elderly women: the Study of Osteoporotic Fractures. J Clin Endocrinol Metab 1998; 83:2239. 92. Smith EP, Boyd J, Frank GR, et al. Estrogen resistance caused by a mutation in the estrogen-receptor gene in a man. N Engl J Med 1994; 331:1056. 93. Center JR, Nguven TV, White CP, Eisman JA. Male osteoporosis predictors: sex hormones and calcitropic hormones. J Bone Miner Res 1997; 12(S1):S368. 94. Nordin BEC, Need AG, Morris HA, et al. Evidence for a renal calcium leak in postmenopausal women. J Clin Endocrinol Metab 1991; 72(2):401. 95. Sakhaee K, Nicar MJ, Glass K, Pak CYC. Postmenopausal osteoporosis as a manifestation of renal hypercalciuria with secondary hyperparathyroidism. J Clin Endocrinol Metab 1985; 61:368. 96. Coe FL, Canterbury JM, Firpo JJ, Reiss E. Evidence for secondary hyper-parathyroidism in idiopathic hypercalciuria. J Clin Invest 1973; 52:134. 97. Heaney RP. A unified concept of osteoporosis. Am J Med 1965; 39:377. 98. Cosman F, Shen V, Xie F, Seibel M, et al. Estrogen protection against bone resorbing effects of parathyroid hormone infusion: assessment by use of biochemical markers. Ann Intern Med 1993; 118:337. 99. Kanis JK, ed. Osteoporosis. London: Blackwell Healthcare Communications, 1997. 100. Eddy DM, Johnston CC Jr, Cummings SR, et al. Osteoporosis: review of the evidence for prevention, diagnosis, and treatment and cost-effectiveness analysis. Osteoporos Int 1998; 8(Suppl 4):S1. 101. Cummings SR, Nevitt MC, Browner WS, et al. For the Study of Osteoporotic Fractures Research Group. Risk factors for hip fracture in white women. N Engl J Med 1995; 332:767. 102. Johnell O, Gullberg B, Kanis JA, et al. Risk factors for hip fracture in European womenthe MEDOS study. J Bone Miner Res 1995; 10(11):1802. 103. Dargent-Molina P, Favier F, Grandjean H, et al. For the EPIDOS Study Group. Fall-related factors and risk of hip fracture: the EPIDOS prospective study. Lancet 1996; 348:145. 104. Cummings SR, Kelsey JL, Nevitt MC, O'Dowd KJ. Epidemiology of osteoporosis and osteoporotic fractures. Epidemiol Rev 1985; 7:178. 105. Lindsay R. Estrogen deficiency. In: Riggs BL, Melton III LJ, eds. Osteoporosis. Philadelphia: LippincottRaven Publishers, 1995:133. 106. Farmer ME, White LR, Brody JA. Race and sex differences in hip fracture incidence. Am J Public Health 1984; 74:1374. 107. Garn SM. Population differences and family line differences in cortical thickness. In: The earlier gain and the later loss of cortical bone, in nutritional perspective. Springfield, IL: Charles C Thomas, 1970. 108. Seeman E, Hopper JL, Bach LA, et al. Reduced bone mass in daughters of women with osteoporosis. N Engl J Med 1989; 320:554. 109. Lutz J. Bone mineral, serum calcium and dietary intakes of mother/daughter pairs. Am J Clin Nutr 1986; 44:99. 110. Tylavsky FA, Bortz AD, Hancock R, Anderson JJB. Familial resemblance of radial bone mass between premenopausal mothers and their college-age daughters. Calcif Tissue Int 1989; 45:265. 111. Ross PD, Davis JW, Epstein RS, Wasnich RD. Pre-existing fractures and bone mass predict vertebral fracture incidence in women. Ann Intern Med 1991; 114:919. 112. Lloyd T, Andon MD, Rollings N, et al. Calcium supplementation and bone mineral density in adolescent girls. JAMA 1993; 270:841. 113. Johnston CC Jr, Miller JZ, Slemenda CW, et al. Calcium supplementation and increases in bone mineral density in children. N Engl J Med 1992; 327:82. 114. Matkovic V, Fontana D, Tominac C, et al. Factors which influence peak bone mass formation: a study of calcium balance and the inheritance of bone mass in adolescent females. Am J Clin Nutr 1990; 52:878. 115. Kanders B, Dempster DW, Lindsay R. Interaction of calcium, nutrition, and physical activity on bone mass in young women. J Bone Miner Res 1988; 3:145. 116. Baran D, Sorensen A, Grimes J, et al. Dietary modification with dairy products for preventing vertebral bone loss in premenopausal women: a three year prospective study. J Clin Endocrinol Metab 1989; 70:264. 117. Elders PJM, Netelendo JC, Lips P, et al. Calcium supplementation reduces vertebral bone loss in postmenopausal women: a controlled trial in 248 women between 46 and 55 years of age. J Clin Endocrinol Metab 1991; 73:533. 118. Heaney RP. Calcium intake requirement and bone mass in the elderly. J Lab Clin Med 1982; 100:309. 119. Holbrook TL, Barrett-Connor E, Wingard DL. Dietary calcium and risk of hip fracture: 14 year prospective population study. Lancet 1988; 2:1046. 120. Dawson-Hughes B, Dallal GE, Krall EA, et al. A controlled trial of the effect of calcium supplementation on bone density in postmenopausal women. N Engl J Med 1990; 323:878. 121. Tillyard MW, Spears GFS, Thomson J, Dovey S. Treatment of postmenopausal osteoporosis with calcitriol or calcium. N Engl J Med 1992; 326. 122. Recker RR, Hinders S, Davies KM, et al. Correcting calcium nutritional deficiency prevents spine fractures in elderly women. J Bone Miner Res 1996; 11:1961. 123. Aloia JF, Vaswani A, Yeh JK, et al. Calcium supplementation with and without hormone replacement therapy to prevent postmenopausal bone loss. Ann Intern Med 1994; 120:97. 124. Reid IR, Ames RW, Evans MC, et al. Long-term effects of calcium supplementation on bone loss and fractures in postmenopausal women. A randomized controlled trial. Am J Med 1995; 98:331. 125. Chapuy MC, Arlot ME, Duboeuf F, et al. Vitamin D3 and calcium to prevent hip fractures in elderly women. N Engl J Med 1992; 327:1637. 126. Kanis JA, Johnell O, Gullberg B, et al. Evidence for efficacy of drugs affecting bone metabolism in preventing hip fracture. BMJ 1992; 305:1124. 127. Dawson-Hughes B, Dallal GE, Krall EA, et al. A controlled trial of the effect of calcium supplementation on bone density in postmenopausal women. N Engl J Med 1990; 323:878. 128. Parfitt AM, Gallagher JC, Heaney RP, et al. Vitamin D and bone health in the elderly. Am J Clin Nutr 1982; 36:1014. 129. O'Dowd KJ, Clemens TL, Kelsey JL, Lindsay R. Exogenous calciferol (vitamin D) and vitamin D endocrine status among elderly nursing home residents in the New York City area. J Am Geriatr Soc 1993; 41:414. 130. Nieves J, Cosman F, Herbert J, et al. High prevalence of vitamin D deficiency and reduced bone mass in multiple sclerosis. Neurology 1994; 44:1687. 131. Thomas MK, Lloyd-Jones DM, Thadhani RI, et al. Hypovitaminosis D in medical inpatients. N Engl J Med 1998; 338:777. 132. Komar L, Nieves J, Cosman F, et al. Calcium homeostasis of an elderly population upon admission to a nursing home. J Am Geritatr Soc 1993; 41:1057. 133. Tsai KS, Heath HH III, Kumar R, Riggs BL. Impaired vitamin D metabolism with aging in women: possible role in pathogenesis of senile osteoporosis. J Clin Invest 1984; 73:1668. 134. Silverberg SJ, Shane E, de la Cruz L, et al. Abnormalities in parathyroid hormone secretion and 1,25-dihydroxyvitamin D formation in women with osteoporosis. N Engl J Med 1989; 320(5):277. 135. Villareal DT, Civitelli R, Chines A, Avioli LV. Subclinical vitamin D deficiency in postmenopausal women with low vertebral bone mass. J Clin Endocrinol Metab 1991; 72:628. 136. Khaw K, Sneyd M, Compston J. Bone density, parathyroid hormone and 25-hydroxyvitamin D concentrations in middle aged women. BMJ 1992; 305:273. 137. Nordin BEC, Need AG, Morris HA, Horowitz M. The nature and significance of the relationship between urinary sodium and urinary calcium in women. J Nutr 1993; 123:1615. 138. Heaney RP, Recker RR. Effects of nitrogen, phosphorus and caffeine on calcium balance in women. J Lab Clin Med 1982; 99:46. 139. Barrett-Connor E, Chang JC, Edelstein SL, et al. Coffee-associated osteoporosis offset by daily milk consumption: the Rancho Bernardo study. JAMA 1994; 271:280. 140. Bikle DD, Genant HK, Cann C, et al. Bone disease in alcohol abuse. Ann Intern Med 1985; 103:42. 141. Peris P, Pares A, Guanabens N, et al. Reduced spinal and femoral bone mass and deranged bone mineral metabolism in chronic alcoholics. Alcohol Alcohol 1992; 27:619. 142. Seeman E. The effects of tobacco and alcohol use on bone. In: Marcus R, Feldman D, Kelsey J, eds. Osteoporosis. San Diego: Academic Press, 1996:477. 143. Hopper JL, Seeman E. Bone density in twins discordant for tobacco use. N Engl J Med 1994; 330:387. 144. Hirota Y, Hirohata T, Fukuda K. Association of intake, cigarette smoking and occupational status with the risk of idiopathic osteonecrosis of the femoral head. Am J Epidemiol 1993; 137:530. 145. Krolner B, Toft B. Vertebral bone loss: an unheeded side-effect of therapeutic bed rest. Clin Sci 1983; 64:537. 146. Mazess RB, Whedon JD. Immobilization and bone. Calcif Tissue Int 1983; 35:265. 147. Vogel JM, Whittle MW. Bone mineral changes; the second manned Skylab mission. Aviat Space Environ Med 1976; 47:396. 148. Huddleson AL, Rockwell D, Kurland DN, Harrison B. Bone mass in lifetime tennis athletes. JAMA 1980; 244:1107.

149. 150. 151. 152. 153. 154. 155. 156. 157. 158. 159. 160. 161. 162. 163. 164. 165. 166. 167. 168. 169. 170. 171. 172. 173. 174. 175. 176. 177. 178. 179. 180. 181. 182. 183. 184. 185. 186. 187. 188. 189. 190. 191. 192. 193. 194. 195. 196. 197. 198. 199. 200. 201. 202. 203. 204. 205. 206. 207. 208. 209. 210. 211. 212. 213. 214. 215. 216. 217. 218. 219. 220. 221. 222. 223. 224. 225. 226. 227. 228. 229. 230.

Krolner B, Toft B, Nielsen SP, Tondevold E. Physical exercise as prophylaxis against involutional vertebral bone loss: a controlled trial. Clin Sci 1983; 64:541. Smith EL, Reddan W, Smith PE. Physical activity and calcium modalities for bone mineral increase in aged women. Med Sci Sports Exerc 1981; 13:60. Gutin B, Kasper MJ. Can vigorous exercise play a role in osteoporosis prevention? A review. Osteoporos Int 1992; 2:55. Krall EA, Dawson-Hughes B. Walking is related to bone density and rates of bone loss. Am J Med 1994; 96:20. Greendale GA, Barrett-Connor E. Outcomes of osteoporotic fractures. In: Marcus R, Freedman D, Kelsey J, eds. Osteoporosis. San Diego: Academic Press, 1996. Ensrud KE, Black DM, Harris F, et al. Correlates of kyphosis in older women. The fracture intervention trial research group. J Am Geriatr Soc 1997; 45(6):682. Finsen V. Osteoporosis and back pain among the elderly. Acta Med Scand 1998; 223:443. Bergenudd H, Nilson B, Uden A, Wilner S. Bone mineral content, gender, body posture, and build in relation to back pain in middle age. Spine 1989; 14:577. Nicholson PT, Haddaway MW, Davie MWJ, Evans SF. Vertebral deformity, bone mineral density, back pain, and height loss in unscreened women over 50 years. Osteoporos Int 1993; 3:300. Leidig G, Minne HW, Sauer P, et al. A study of complaints and their relation to vertebral destruction in patients with osteoporosis. Bone Miner 1990; 8:217. Lyles KW, Gold DT, Shipp KM, et al. Association of osteoporotic vertebral compression fractures with impaired functional status. Am J Med 1993; 94:595. Mossey JM, Mutran E, Knott K, Craik R. Determinants of recovery 12 months after hip fracture: the importance of psychosocial factors. Am J Public Health 1989; 79(3):279. US Congress, Office of Technology Assessment. Hip fracture outcomes in people age 50 and overbackground paper OTA-BP-H-120. Washington: US Government Printing Office, 1994. Miller CW. Survival and ambulation following hip fracture. J Bone Joint Surg 1978; 60-A:930. Marottoli RA, Berkman LF, Cooney LM. Decline in physical function following hip fracture. J Am Geriatr Soc 1992; 40:861. Greendale GA, Barrett-Connor E. Outcomes of osteoporotic fractures. In: Marcus R, Freedman D, Kelsey J, eds. Osteoporosis. San Diego: Academic Press, 1996. White BL, Fisher WD, Laurin CA. Rate of mortality for elderly patients after fracture of the hip in the 1980s. J Bone Joint Surg 1987; 69-A:1335. Poor G, Atkinson EJ, O'Fallon WM, Melton LJ III. Determinants of reduced survival following hip fractures in men. Clin Orthop 1995; 319:260. Kaukonen JP, Karaharju EO, Porras M, et al. Functional recovery after fracture of the distal forearm. Ann Chir Gynaecol 1988; 77:27. Gartland JJ, Werley CW. Evaluation of healed Colles fractures. J Bone Joint Surg 1951; 33A:895. Field J, Warwick D, Bannister GC, et al. Long-term prognosis of displaced Colles fracture: a 10 year prospective review. Injury 1992; 23:529. Altissimi M, Antenucci R, Fiacca C, et al. Long-term results of conservative treatment of fractures of the distal radius. Clin Orthop Relat Res 1986; 206:202. Stevenson JC, Marsh MS, Lindsay R. An atlas of osteoporosis, 2nd ed. London: Parthenon Publishing Group, 1999. National Osteoporosis Foundation. Physician's guide to prevention and treatment of osteoporosis. Belle Mead, New Jersey: Excerpta Medica, 1998. Chow R, Harrison JE, Notarius C. Effect of two randomized exercise programs on bone mass of healthy postmenopausal women. BMJ 1987; 295:1441. Friedlander AL, Genant HK, Sadowsky S, et al. A two-year program of aerobics and weight training enhances bone mineral density of young women. J Bone Miner Res 1995; 10:574. Grove KA, Londeree BR. Bone density in postmenopausal women: high impact vs. low impact exercise. Med Sci Sports Exerc 1992; 24:1190. Hatori M, Hasegawa A, Adachi H, et al. The effects of walking at the anaerobic threshold level on vertebral bone loss in postmenopausal women. Calcif Tissue Int 1993; 52:411. Nelson ME, Fiatarone MA, Morganti CM, et al. Effects of high-intensity strength training on multiple risk factors for osteoporotic fractures. A randomized controlled trial. JAMA 1994; 272:1909. Simkin A, Ayalon J, Leichter I. Increased trabecular bone density due to bone-loading exercises in postmenopausal osteoporotic women. Calcif Tissue Int 1987; 40:59. Snow-Harter C, Bouxsein ML, Lewis BT, et al. Effects of resistance and endurance exercise on bone mineral status of young women: a randomized exercise intervention trial. J Bone Miner Res 1992; 7:761. Bassey EJ, Ramsdale SJ. Weight-bearing exercise and ground reaction forces: a 12-month randomized controlled trial of effects on bone mineral density in healthy postmenopausal women. Bone 1995; 16:469. Sinaki M, Wahner HW, Offord DP, Hodgson SF. Efficacy of nonloading exercises in prevention of vertebral bone loss in postmenopausal women: a controlled trial. Mayo Clin Proc 1989; 64:762. Dalsky GP, Stocke KS, Ehsani AA, et al. Weight-bearing exercise training and lumbar bone mineral content in postmenopausal women. Ann Intern Med 1998; 108:824. Eastell R. Drug therapy: treatment of postmenopausal osteoporosis. N Engl J Med 1998; 338:736. Lindsay R. Prevention and treatment of osteoporosis. Lancet 1993; 341:801. Lindsay R, Hart DM, Forrest C, Baird C. Prevention of spinal osteoporosis in oophorectomized women. Lancet 1980; 2:1151. Albright F. The effect of hormones on osteogenesis in man. Recent Prog Horm Res 1947; 1:293. Lindsay R, Aitken JM, Anderson JB, et al. Long-term prevention of post-menopausal osteoporosis by estrogen. Lancet 1976; 1(7968):1038. Lindsay R, Hart DM, Purdie P, et al. Comparative effects of estrogen and a progestogen on bone loss in postmenopausal women. Clin Sci Mol Med 1978; 54:193. Christiansen C, Rodbro P. Does postmenopausal bone loss respond to estrogen replacement therapy independent of bone loss rate? Calcif Tissue 1983; 35:720. Christiansen C, Christiansen MS, McNair P. Prevention of early postmenopausal bone loss: conducted 2 years study in 315 normal females. Eur J Clin Invest 1980; 10:273. Davis ME, Lanzl LH, Cos AB. Detection, prevention and retardation of postmenopausal osteoporosis. Obstet Gynecol 1970; 36:187. Robert R, Recker K, Davies M, et al. The effect of low-dose continuous estrogen and progesterone therapy with calcium and vitamin D on bone in elderly women. Ann Intern Med 1999; 130:897. Lufkin EG, Wahner HW, O'Fallon WM, et al. Treatment of postmenopausal osteoporosis with transdermal estrogen. Ann Intern Med 1992; 117:1. Paganini-Hill A, Ross RK, Gerkins VR, et al. Menopausal estrogen therapy and hip fractures. Ann Intern Med 1981; 95:28. Hutchinson TA, Polansky JM, Fienstein AR. Postmenopausal oestrogens protect against fracture of hip and distal radius. Lancet 1979; 2(8145):705. Kreiger N, Kelsey JL, Holford TR. An epidemiological study of hip fracture in postmenopausal women. Am J Epidemiol 1982; 116:141. Weiss NS, Szekely DR, Dallas R, et al. Endometrial cancer in relation to patterns of menopausal estrogen use. JAMA 1979; 242:261. Ettinger B, Genant HK, Cann CE. Long-term estrogen therapy prevents bone loss and fracture. Ann Intern Med 1985; 102:319. Komulainen MH, Kroger H, Tuppurainen MT, et al. HRT and vitamin D in prevention of non-vertebral fractures in postmenopausal women; a 5 year randomized trial. Maturitas 1998; 13:45. Burch JC, Byrd BF, Vaughn WK. The effects of long-term estrogen on hysterectomized women. Am J Obstet Gynecol 1974; 118:778. Kiel DP, Felson DT, Anderson JJ. Hip fracture and the use of estrogens in postmenopausal women. N Engl J Med 1987; 317:1169. Williams AR, Weiss NS, Ure C, et al. Effect of weight, smoking, and estrogen use on the risk of hip and forearm fractures in postmenopausal women. Obstet Gynecol 1982; 60:695. Cauley JA, Seeley DG, Ensrud K, et al. For the Study of Osteoporotic Fractures Research Group. Estrogen replacement therapy and fractures in older women. Ann Intern Med 1995; 122:9. Hulley S, Grady D, Bush T, et al. Randomized trial of estrogen plus progestin for secondary prevention of coronary heart disease in postmenopausal women. Heart and Estrogen/progestin Replacement Study (HERS) Research Group. JAMA 1998; 280:605. Barrett-Connor E. Postmenopausal estrogen and prevention bias. Ann Intern Med 1991; 115:455. Horsman A, Gallagher JC, Simpson M, Nordin BEC. Prospective trial of estrogen and calcium in postmenopausal women. BMJ 1977; 2:789. Lindsay R, Hart DM, Clark DM. The minimum effective dose of estrogen for prevention of postmenopausal bone loss. Obstet Gynecol 1984; 63:759. Nachtigall LE, Nachtigall RH, Nachtigall RD. Estrogen replacement therapy I: a 10-year prospective study in the relationship to osteoporosis. Obstet Gynecol 1979; 53:277. Quigley MET, Martin PL, Burnier AM, Brooks P. Estrogen therapy arrests bone loss in elderly women. Am J Obstet Gynecol 1987; 156:1516. Recker RR, Saville PD, Heaney RP. The effect of estrogens and calcium carbonate on bone loss in postmenopausal women. Ann Intern Med 1977; 87:649. Riggs BL, Jowsey J, Kelly PJ, et al. Effect of sex hormones on bone in primary osteoporosis. J Clin Invest 1969; 48:1065. Meema S, Bunker ML, Meema HE. Preventive effect of estrogen on post-menopausal bone loss. Arch Intern Med 1975; 135:1436. Riis B, Thomsen K, Strom V, Christiansen C. The effect of percutaneous estradiol and natural progesterone on postmenopausal bone loss. Am J Obstet Gynecol 1987; 156:61. Jensen GF, Christiansen C, Transbol I. Treatment of postmenopausal osteoporosis: a controlled therapeutic trial comparing estrogen/gestagen, 1,25-dihydroxy-vitamin D 3 and calcium. Clin Endocrinol 1982; 16:515. Ettinger B, Genant HK, Cann CE. Postmenopausal bone loss is prevented by treatment with low-dosage estrogen and calcium. Ann Intern Med 1978; 106:40. Khosla S, Melton LJ 3rd, Atkinson EJ, et al. Relationship of serum sex steroid levels and bone turnover markers with bone mineral density in men and women: a key role for bioavailable estrogen. J Clin Endocrinol Metab 1998; 83(7):2266. Cummings SR, Browner WS, Bauer D, et al. Endogenous hormones and the risk of hip and vertebral fractures among older women. N Engl J Med 1998; 339(11):733. Ravnikar VA. Compliance with hormone replacement therapy: are women receiving the full impact of hormone replacement therapy preventive health benefits? Women's Health Issues 1992; 2:75. Ettinger B. Overview of estrogen replacement therapy: a historical perspective. Proc Soc Exp Biol Med 1998; 217:2. Genant HK, Lucas J, Weiss S, et al. Low-dose esterified estrogen therapy: effects on bone, plasma estradiol concentrations, endometrium, and lipid levels. Estratab Osteoporosis Study Group. Arch Intern Med 1997; 157:2609. Arnold JS, Bartley MH, Tont SA, Jenkins DP. Effects of hormone therapy on bone mineral density: results from the Postmenopausal Estrogen/Progestin Interventions (PEPI) trial. The Writing Group for the PEPI Trial. JAMA 1996; 276:1389. Nieves JW, Komar L, Cosman F, Lindsay R. Interaction between antiresorptive therapy and calcium intake: review and analysis. Am J Clin Nutr 1998; 67:18. Speroff L, Rowan J, Symons J, et al. The comparative effect on bone density, endometrium, and lipids of continuous hormones as replacement therapy (CHART Study). A randomized controlled trial. JAMA 1996; 276:1397. Santoro NF, Nananda F, Eckman MH, et al. Therapeutic controversy. Hormone replacement therapywhere are we going? J Clin Endocrinol Metab 1999; 84:1798. Henderson VW. The epidemiology of estrogen replacement therapy and Alzheimer's disease. Neurology 1997; 48:S27. Yaffe K, Sawaya G, Lieberburg I, Grady D. Estrogen therapy in postmenopausal women: effects on cognitive function and dementia. JAMA 1998; 279:688. Paganini-Hill A. Does estrogen replacement therapy protect against Alzheimer's disease? Osteoporos Int 1997; 7(Suppl 1):S12. Haskell SG, Richardson ED, Horwitz RI. The effect of estrogen replacement therapy on cognitive function in women: a critical review of the literature. J Clin Epidemiol 1997; 50:1249. Collaborative Group on Hormonal Factors in Breast Cancer. Breast cancer and hormone replacement therapy: collaborative re-analysis of data from 51 epidemiological studies of 52,705 women with breast cancer and 108,411 women without breast cancer. Lancet 1997; 350:1047. Papapoulos S. Bisphosphonates. Pharmacology and use in the treatment of osteoporosis. In: Marcus R, Freedman D, Kelsey J, eds. Osteoporosis. San Diego: Academic Press, 1996:1209.

230a.Harris SJ, Watts NB, Jackson RD, et al. Four year study of intermittent cyclic etidronate treatment of postmenopausal osteoporosis: three years of blinded treatment followed by one year of open treatment. Am J Med 1993; 95:557. 231. 232. 233. 234. 235. 236. 237. 238. 239. 240. 241. 242. 243. 244. 245. 246. Liberman UA, Weiss SR, Broll J. Effect of oral alendronate on bone mineral density and the incidence of fractures in postmenopausal osteoporosis. N Engl J Med 1995; 333:1437. Black DM, Cummings SR, Karpf DB. Randomised trial of effect of alendronate on risk of fracture in women with existing vertebral fractures. Lancet 1996; 348:1535. Cummings SR, Black DM, Thompson DE, et al. Effect of alendronate on risk fracture in women with low bone density but without vertebral fractures. JAMA 1998; 280:2077. Peter CP, Handt LK, Smity SM. Esophageal irritation due to alendronate sodium tablets: possible mechanisms. Dig Dis Sci 1998; 43:1998. De Grouen PC, Lubbe DF, Hirsch LJ, et al. Esophagitis associated with the use of alendronate. N Engl J Med 1996; 35:1016. Harris ST, Watts NB, Genant HK, et al. Effects of risedronate treatment on vertebral and nonvertebral fractures in women with postmenopausal osteoporosis. A randomized controlled trial. JAMA 1999; 282:14. Miller P, Roux C, McClung M, et al. Risedronate reduces hip fractures in patients with low femoral neck bone mineral density. Arthritis & Rheumatism 1999; 42:2. Ettinger B, Pressman A, Schein J, et al. Alendronate use among 812 women: prevalence of gastrointestinal complaints, noncompliance with patient instructions, and discontinuation. J Managed Care Pharm 1998:488. Chestnut CH III, Silverman S, Andriano K, et al., for the PROOF Study Group. A randomized trial of nasal spray salmon calcitonin in postmenopausal women with established osteoporosis: the prevent recurrence of osteoporotic fractures study. Am J Med 2000; 109:267. Cosman F, Lindsay R. Selective estrogen receptor modulators: clinical spectrum. Endocr Rev 1999; 20:418. Fisher B, Costantino JP, Wickerham DL, et al. Tamoxifen for prevention of breast cancer: report of the National Surgical Adjuvant Breast and Bowel Project P-1 Study. J Natl Cancer Inst 1998; 90:1371. Powles TJ, Hickish T, Kanis JA, et al. Effect of tamoxifen on bone mineral density measured by dual energy x-ray absorptiometry in healthy pre-menopausal and postmenopausal women. J Clin Oncol 1996; 14:78. Delmas PD, Ensrud KE, Harris S, et al. Raloxifene therapy for 3 years reduces the risk of incident vertebral fractures in postmenopausal women. Paper presented at: 26th European Symposium on Calcified Tissues; 1999; The Netherlands. Cummings SR, Eckert S, Krueger KA, et al. The effect of raloxifene on risk of breast cancer in postmenopausal women. Results from the MORE randomized trial. JAMA 1999; 281:2189. Riggs BL, O'Fallon WM, Lane A, et al. Clinical trial of fluoride therapy in postmenopausal osteoporotic women: extended observations and additional analysis. J Bone Miner Res 1994; 9:265. Kleerekoper M, Mendlovic DB. Sodium fluoride therapy of postmenopausal osteoporosis. Endocr Rev 1993; 14:312.

247. Pak CYC, Sakhaee K, Piziak V, et al. Slow-release sodium fluoride in the management of postmenopausal osteoporosis. A randomized clinical trial. Ann Intern Med 1994; 120:625. 248. Cosman F, Lindsay R. Is parathyroid hormone a therapeutic option for osteoporosis? Calcif Tissue 1998; 62:475. 249. Lindsay R, Nieves JW, Formica C, et al. Randomized controlled study of effect of parathyroid hormone on vertebral bone mass and fracture incidence among postmenopausal women on estrogen with osteoporosis. Lancet 1997; 350:550. 249a.Neer R. Recombinant human PTH reduces the risk of spine and non-spine fractures in psotmenopausal osteoporosis. Toronto, Canada: Endocrine Society 2000. 250. 251. 252. 253. 254. 255. 256. 257. 258. 259. 260. 261. 262. 263. 264. 265. 266. 267. 268. 269. 270. 271. 272. 273. 274. 275. 276. 277. Finklestein JS, Neer RM, Biller BMK, et al. Osteopenia in men with history of delayed puberty. N Engl J Med 1992; 326:600. Jackson JA, Kleerekoper M. Osteoporosis in men: diagnosis, pathophysiology and prevention. Medicine (Baltimore) 1990; 69:137. Teotia M, Teotia SPS, Singh RK. Idiopathic juvenile osteoporosis. Am J Dis Child 1979; 133:894. Hoekman J, Papapoulos SE, Peters ACB, Bijvoet OLM. Characteristics and bisphosphonate treatment of a patient with juvenile osteoporosis. J Clin Endocrinol Metab 1985; 61:952. Zerwekh JE, Sakhaee K, Breslau NA, et al. Impaired bone formation in male idiopathic osteoporosis: further reduction in the presence of concomitant hypercalciuria. Osteoporos Int 1992; 2:128. Ljunghalls S, Johansson AG, Burmen P, et al. Low levels of insulin-like growth factor I (IGF-I) in male patients with idiopathic osteoporosis. J Intern Med 1992; 232:59. Kozin F, McCarty DJ, Simms J, Genant H. The reflex sympathetic dystrophy syndrome. I. Clinical and histologic studies: evidence for bilaterality, response to corticosteroid and articular movement. Am J Med 1976; 60:321. Koch E, Hofer HO, Sialer G, et al. Failure of MR imaging to detect reflex sympathetic dystrophy of the extremities. AJR Am J Roentgenol 1991; 56(1):113. Ford SR, Forrest WH Jr, Eltherington L. The treatment of reflex sympathetic dystrophy with intravenous regional bretylium. Anesthesiology 1988; 66:137. Schils J, Piraino D, Richmond BJ, et al. Transient osteoporosis of the hip: clinical and imaging features. Cleve Clin J Med 1992; 59:483. Potter H, Moran M, Schneider R, et al. Magnetic resonance imaging in the diagnosis of transient osteoporosis of the hip. Clin Orthop 1992; 280:223. Kim SM, Desai AG, Krakovitz M, et al. Scintigraphic evaluation of regional migratory osteoporosis. Clin Nucl Med 1989; 14:36. Lukert BP, Raisz LG. Glucocorticoid-induced osteoporosis: pathogenesis and management. Ann Intern Med 1990; 112:352. Meunier PJ. Is steroid induced osteoporosis preventable? N Engl J Med 1993; 323:1781. Cosman F, Nieves J, Herbert J, et al. High dose glucocorticoids in multiple sclerosis patients exert direct effects on the kidney and skeleton. J Bone Miner Res 1994; 9:1097. Tillyard MW, Spears GFS, Thomson J, Dovey S. Treatment of postmenopausal osteoporosis with calcitriol or calcium. N Engl J Med 1992; 326. Greenspan SL, Greenspan FS. The effect of thyroid hormone on skeletal integrity. Ann Intern Med 1999; 130:750. Shore RM, Chesney RW, Mazess RB, et al. Skeletal demineralization in Turner syndrome. Calcif Tissue Int 1982; 34:519. Smith DAS, Walker MS. Changes in plasma steroids in bone density in Klinefelter syndrome. Calcif Tissue Int 1977; 22(Suppl):225. Rico H, Arnanz F, Revilla M. Total and regional bone mineral content in women treated with GnRH agonists. Calcif Tissue Int 1993; 52:354. Warren MP. Amenorrhea in endurance runners. J Clin Endocrinol Metab 1992; 75:393. Slemenda CW, Johnson CC. High intensity activities in young women: site specific bone mass effects among female figure skaters. Bone Miner 1993; 20:125. Rigotti NA, Neer RM, Skates SJ, et al. The clinical course of osteoporosis in anorexia nervosa: a longitudinal study of cortical bone mass. JAMA 1991; 265:1133. Schlechte JA, Sherman B, Martin R. Bone density in amenorrheic women with and without hyperprolactinemia. J Clin Endocrinol Metab 1983; 56:1120. Soejima K, Landing BH. Osteoporosis in juvenile-onset diabetes mellitus: morphometric and comparative studies. Pediatr Pathol 1986; 6:289. Riggs BL, Randall RV, Wahner HW, et al. The nature of the metabolic bone disorder in acromegaly. J Clin Endocrinol Metab 1972; 34:911. Gray RW, Garthwaite TL. Activation of renal 1,25-dihydroxyvitamin D synthesis by phosphate deprivation: evidence for a role of growth hormone. Endocrinology 1983; 116:1989. Paterson CR, McAllion S, Stellman JL. Osteogenesis imperfecta after the menopause. N Engl J Med 1984; 310:1694.

277a.Ganry O, Baudoin C, Fardellone P. Effect of alcohol intake on bone mineral density in elderly women: The EPIDOS study. Am J Epidemiol 2000; 151:773. 278. Schwartz AM, Leonides JC. Methotrexate osteopathy. Skeletal Radiol 1984; 11:13. 279. Smith R, Stevenson JC, Winearls CJ, et al. Osteoporosis of pregnancy. Lancet 1985; 1:1178. 280. Hillman L, Sateesna S, Haussler M, et al. Control of mineral homeostasis during lactation: interrelationships of 25-hydroxyvitamin D, 24,25-dihydroxy D, 1,25-dihydroxy D, parathyroid hormone, calcitonin, prolactin, and estradiol. Am J Obstet Gynecol 1981; 139:471. 281. Long RG, Meinhard E, Skinner RK, et al. Clinical, biochemical, and histological studies of osteomalacia, osteoporosis, and parathyroid function in chronic liver disease. Gut 1978; 19:85. 282. Avioli LV. Heparin-induced osteopenia: an appraisal. Adv Exp Med Biol 1975; 52:375. 283. Chines A, Pacifici R, Avioli LV, et al. Systemic mastocytosis presenting as osteoporosis: a clinical and histomorphologic study. J Clin Endocrinol Metab 1991; 72:140. 284. Hahn TJ, Hendin BA, Scharp CR, et al. Serum 25-hydroxycalciferol levels and bone mass in children on chronic anticonvulsant therapy. N Engl J Med 1975; 292:550. 285. Valimaki MJ, Tiihone M, Laitinen K, et al. Bone mineral density measured by dual energy x-ray absorptiometry and novel markers of bone formation and resorption in patients on antiepileptic drugs. J Bone Miner Res 1994; 9:631.

CHAPTER 65 PAGET DISEASE OF BONE Principles and Practice of Endocrinology and Metabolism

CHAPTER 65 PAGET DISEASE OF BONE

ETHEL S. SIRIS Etiology Pathology Clinical Presentation Diagnosis Treatment Bisphosphonates Calcitonin Plicamycin (Mithramycin) Nonsteroidal Antiinflammatory Drugs Surgical Intervention Chapter References

Paget disease of bone is a localized disorder of skeletal remodeling in which abnormally increased rates of both bone resorption and new bone formation change bone architecture. Typically, pagetic bone shows a characteristic radiographic appearance (Fig. 65-1). The affected bone is often larger, generally less compact, more vascular, and more susceptible to deformity and fracture than normal bone. Paget disease occurs predominantly in persons older than 40 years, affecting men slightly more often than women. In the United States, up to 2% to 3% of the population in their 50s and 60s have this bone disease, and the frequency rises in the ensuing decades.1a Most patients are asymptomatic, although a significant minority do experience signs and symptoms related to the area of involved bone, the degree of increased bone turnover, and the structural changes that result at affected sites.

FIGURE 65-1. Radiographic findings in pagetic bone. A, Typical cotton wool appearance of the skull. Areas of extensive blastic change are present, and the cortex of the bone is markedly thickened. B, A normal left tibia and fibula contrasted with a pagetic, bowed right tibia. The affected tibia is larger, has areas of increased density as well as localized hyperlucent foci, and is bowed. C, Evidence of the localized nature of Paget disease, showing a normal proximal half of the humerus with pagetic involvement in the entire distal half. A sharp line of demarcation exists between normal and involved bone at the midshaft point. This radiograph also demonstrates the poor architectural quality of the pagetic bone. D, Relatively early changes of Paget disease in the right hemipelvis include thickening of the ilioischial and ilio-pectineal lines as well as sclerosis in the ilium and ischium, in contrast to the normal findings in the left hemipelvis. E, Extensive Paget disease involving the entire lumbosacral spine (and visible pelvis), with marked coarsening of trabecular markings. Secondary degenerative changes with loss of disc spaces are seen as well. (R, right; L, left.) (A through C from Siris ES, Canfield RE, Jacobs TP. Paget's disease of bone. Bull N Y Acad Med 1980; 56:285.)

ETIOLOGY
The precise etiology of Paget disease remains unknown; genetic and environmentalspecifically viralfactors play a role. Geographic and ethnic epidemiologic data document the common occurrence of Paget disease in some parts of the world and its rarity in others. Earlier studies had indicated that Paget disease was present in 4% to 5% of older Britons.1 A focus of the disorder occurred in the Lancashire area, where as many as 8% of the older population showed radiographic evidence of this condition.2 Interestingly, more recent epidemiologic studies from the United Kingdom and New Zealand suggest both a decreasing frequency and decreasing severity of the disease.2a,2b Nonetheless, it is relatively common in those countries, as well as in Australia and throughout western Europe (with the exception of Scandinavia, where it is infrequently seen).3 The disorder is uncommon in both southern Asia (i.e., the Indian subcontinent), eastern Asia (i.e., China, Japan), and sub-Saharan Africa. In the United States, it occurs primarily in people of Anglo-Saxon or European descent but is seen in blacks, although apparently to a smaller degree than in whites.4 Paget disease is often present in more than one member of a family; this is noted in up to 40% of cases.5,6 and 7 This value may be an underestimate, because many patients with the condition never know they have it. Early studies analyzing pedigrees of several affected kindreds suggest autosomal dominant inheritance.8 Familial aggregation studies in the United States indicate that first-degree relatives of patients with Paget disease have seven times the risk of developing it than do persons without such a family history. The risk increases if the affected relative had deforming disease or was younger than 50 years old when the disease was diagnosed.9 Linkage of Paget disease to HLA is not conclusive.10,11,12 and 13 Specific susceptibility loci for Paget disease may occur on chromosome 18q, the apparent site of the gene responsible for a similar but rare bone dysplasia called familial expansile osteolysis.14 Studies have reported some families with Paget disease that display genetic linkage to this site,15,16 but such linkage has not been found in others.16 Genetic heterogeneity is very likely. Another area of research to elucidate the etiology of Paget disease involves the possible role of one or more viral agents. At sites of active Paget disease, osteoclasts are abnormally increased in both number and size, containing up to 20 times the usual number of nuclei per cell.17 Initial studies described viral nucleo-capsidlike structures in both the nuclei and cytoplasm of pagetic osteoclasts18,19 (Fig. 65-2). Such virus-like inclusions are not found in osteoclasts in normal bone, although they have been reported in other bone diseases such as osteopetrosis.20 Some subsequent studies using immunofluorescence techniques detected nucleo-capsid protein antigens of both measles and respiratory syncytial virus antigens in these cells,21,22 whereas others implicated canine distemper virus.23 Each of these is one of the paramyxoviruses, a group of viruses for which the members promote the fusion of infected cells and the formation of multinucleated giant cells.24 Conflicting results have emerged, however, when RNA extraction and polymerase chain reaction techniques have been used to search for viral identification in tissue from pagetic patients.25,26,27,28 and 29 Thus, a role for a viral cause as at least part of the etiology of Paget disease is suggested, but clearly not proven.

FIGURE 65-2. Electron micrograph showing the characteristic inclusions seen in an osteoclast nucleus from pagetic bone. A paracrystalline arrangement of ~12.5-nm filaments is noted (arrows). Scattered, nonoriented filaments throughout the karyoplasm are also seen. The filaments are not membrane bound or attached to any nuclear structure. 80,000 (Courtesy of Dr. Barbara Mills.)

Measles virus transcripts have been detected in both bone marrow mononuclear cells28 and in peripheral blood cells,29 including both osteoclast precursors and more primitive hematopoietic stem cells. Because Paget disease is a localized process, the ubiquitous finding of osteoclast precursors bearing evidence of measles virus throughout the circulation is puzzling. One possible approach to explaining this comes from the observation that increased concentrations of interleukin-6 are produced by pagetic osteoclasts and that they express high levels of interleukin-6 receptor on their surfaces.30,31 This suggests that interleukin-6 is an autocrine/paracrine factor that influences and enhances osteoclast activity. Perhaps such cytokines in the marrow microenvironment may influence the differentiation of osteoclast precursors, limiting the extent of pagetic involvement to localized sites once the initial lesion occurs. However, how that initial lesion is established is unknown.

PATHOLOGY
The initial abnormality in Paget disease is a marked increase in the rate of bone resorption at localized sites, mediated by large and numerous pagetic osteoclasts. In response to the increased bone resorption, large numbers of osteoblasts are recruited to these sites and promote a compensatory increase in new bone formation. The osteoblasts probably are inherently normal, although a possible pathologic role for these cells in the initiation of the pagetic process has not been completely excluded. The earliest phase of Paget disease is represented radiographically by an advancing lytic wedge or blade of grass lesion in a long bone, or by osteoporosis circumscripta, as seen in the skull. The next phase is a mixture of the initial abnormality, increased bone resorption, and a compensatory increase of new bone formation. Both osteoclasts and osteoblasts are found at active sites in this stage. Presumably because of the rapidity of the process, newly deposited collagen fibers are laid down in a haphazard, rather than a linear, fashion, creating a mosaic pattern in bonea more primitive woven boneinstead of the normal lamellar pattern. The bone marrow becomes infiltrated by an excess of fibrous connective tissue, and a marked increase in blood vessels is seen, which causes the bone to become hyper-vascular. Usually, the bone matrix is normally mineralized, although areas of reduced mineralization (widened osteoid seams) can be found in isolated areas of some bone biopsy specimens.5 Eventually, the hypercellularity at a given locus may decrease, leaving only the sclerotic mosaic bone (burned out disease) without evidence of active bone turnover. Usually, all different phases are seen concomitantly in different areas of pagetic involvement in a given patient. Figure 65-3 compares the appearance of normal and pagetic bone by scanning electron microscopy, and Figure 65-4A shows the pattern of woven bone in Paget disease.

FIGURE 65-3. Scanning electron micrographs of trabecular bone from iliac crest biopsies taken from a normal person (A) and from a patient with Paget disease (B). The normal biopsy specimen consists of a honeycomb-like network of trabecular plates and cylinders surrounding the marrow cavities. This architecture is lost in the specimen from the patient with Paget disease, in which the trabeculae are coarse and the architecture is disorganized. Note the extensive imprints of osteo-clastic bone resorption in the center portion of the specimen of pagetic bone. (Courtesy of Dr. David Dempster.)

FIGURE 65-4. These photomicrographs, using polarized light, demonstrate the typical pattern of woven bone in a biopsy from a pagetic iliac crest before treatment (A) and after bisphosphonate therapy (B). The first biopsy specimen demonstrates the chaotic and more primitive pattern of collagen deposition that is characteristic of Paget disease. The posttreatment specimen shows areas in which a restoration toward a more lamellar pattern has occurred (arrows). (Courtesy of Dr. Pierre Meunier.)

The increased bone resorption increases the urinary excretion of products of the degradation of bone collagen, including total hydroxyproline and more specific bone markers such as pyridinoline, deoxypyridinoline, and both the N- and C-telopeptides of collagen. The increased osteoblastic activity is associated with a rise in the serum alkaline phosphatase level. Because overall bone resorption and bone formation in most patients are coupled, these values rise in proportion to each other and are useful markers of metabolic activity. Moreover, as specific treatment is instituted, decreases in these indices serve as evidence of a therapeutic response. Serum calcium is typically normal in Paget disease. However, when a patient with extensive and active disease is immobilized and loses the weight-bearing stimulus to new bone formation, hypercalcemia or clinically significant hyper-calciuria may occur. In an ambulatory, otherwise healthy patient with Paget disease, the discovery of an elevated serum calcium level usually leads to a diagnosis of coexisting primary hyperparathyroidism. No data exist to support the idea that the two conditions are causally linked in such people, and the presence of both disorders in the same person is considered a clinical coincidence.32 Some studies have suggested, however, that secondary hyperparathyroidism may emerge in a subset of patients with active Paget disease because of markedly increased calcium requirements during periods of intensive new bone formation.33,34 In patients with either primary or secondary hyperparathyroidism, increased levels of parathyroid hormone stimulate pagetic osteoclasts to even higher levels of activity and may cause a worsening of the underlying bone remodeling abnormality. When primary hyperparathy-roidism coexists, surgical removal of the parathyroid adenoma is often associated with an improvement of the Paget disease. The treatment of coexisting secondary hyperparathyroidism involves maintaining an adequate intake of calcium and vitamin D.

CLINICAL PRESENTATION
The localized nature of Paget disease leads to a variable clinical presentation. Frequently, only one or two bones are involved; alternatively, many bones may be affected. The most commonly involved sites include the pelvis, spine, skull, femur, and tibia. The bones of the upper extremities, the clavicles, scapulae, and ribs are somewhat less commonly involved, and the hands and feet are seldom affected. Clinical symptoms and complications of Paget disease depend greatly on the sites of involvement, the relation of pagetic bone to adjacent structures, and the extent of ongoing metabolic activity. Most patients with Paget disease are asymptomatic, and the diagnosis is an incidental finding, made when either an elevated serum alkaline phosphatase level is discovered or a radiograph showing classic pagetic changes is obtained for some unrelated reason. Other patients have symptoms referable to the skeleton, but these may reflect a range of causes, including problems with the bone itself, joint dysfunction adjacent to pagetic bone, neurologic abnormalities, and changes associated with increased blood flow. Depending on the site of involvement and the nature of the skeletal alterations that are occurring, a simple complaint such as bone pain may be difficult to characterize. For example, patients with skull involvement may describe headaches or tightness across the scalp, or they may note a sensation of warmth, presumably reflecting the increased blood flow through pagetic bone and surrounding soft tissues. A symptom of pain in the back in the setting of one or more enlarged pagetic vertebrae may come from the bone itself, but is probably due to degenerative changes of the joints that are exacerbated by the presence of Paget disease, with or without spinal nerve impingement and muscle spasm. Pagetic involvement of the pelvis or proximal femur may include the hip joint, leading to pain in the groin or in the thigh, often radiating to the knee. This common pattern is a major source of discomfort in patients with Paget disease at this site.

Bone deformity may be the patient's presenting problem. Enlargement or bowing of the femur or tibia is a readily apparent finding, as are skull enlargement with classic frontal bossing, asymmetry of the pelvic or shoulder girdle areas, and kyphosis (Fig. 65-5). When the femur or tibia on one side of the body has a bowing deformity caused by Paget disease, adjacent joints are subjected to mechanical stresses from the curvature; in addition, the bowed limb is typically shortened, increasing the likelihood of back pain or contralateral joint dysfunction and pain resulting from the abnormal gait. These symptoms are among the most common. As with the skull, a sensation of increased heat in a bowed extremity is a frequently stated complaint. Often this increased warmth is felt by the examiner's hand.

FIGURE 65-5. A, Extensive Paget disease in an elderly man, showing considerable deformity of the lower extremities. Outward bowing of the femoral, tibial, and fibular bones and anterior bowing of the shins (saber shins) are seen. The patient also has thoracic kyphosis. Considerable secondary degenerative arthritis of the knees and hips is present. (The nodule on the abdomen is a sebaceous cyst.) B, Severe pagetic involvement of the skull, showing deformity and enlargement of the frontal bone. The patient complained of a throbbing discomfort of the head, and the overlying skin was very warm.

Fracture through pagetic bone is a more serious complication. Pagetic fractures may be either traumatic or pathologic, especially, but not exclusively, involving long bones with active areas of advancing lytic disease. Usually these fractures heal normally, but the rate of nonunion may be as high as 10%.35 The most serious and probably most common fractures involve the femoral shaft or subtrochanteric area.5,35 Some of these fractures are the result of progression from small cortical fissure fractures. Rarely, a fracture occurs in association with an area of malignant degeneration (i.e., through an osteogenic sarcoma in pagetic bone). When Paget disease affects the axial skeleton, it may be associated with neurologic complications. Patients with skull involvement may experience hearing loss of either a sensorineural or a conductive type, or both. Typically, cochlear dysfunction is present and, less commonly, direct eighth nerve compression occurs.36 Vertigo, Bell palsy, and other cranial nerve abnormalities are seen less often. Extensive skull involvement may lead to a softening of the base of the skull, causing platybasia or frank basilar impression. The slow sinking of the skull onto the odontoid process may lead to narrowing of the foraminal space and potentially to an obstructive hydrocephalus. Radiographic criteria for basilar invagination are often met in cases with massive skull involvement, but neurologic sequelae from this are relatively uncommon. When they do occur, however, these sequelae are serious and life-threatening. Brainstem or cerebellar compression or obstruction of cerebrospinal fluid flow from blockage of the fourth ventricle or the foramina of Magendie and Luschka may lead to fairly rapid and progressive neurologic decompensation. The use of computed tomographic scanning and magnetic resonance imaging has made the prompt diagnosis of ventricular enlargement much easier and allows rapid and appropriate neurosurgical intervention. Figure 65-6 shows an example of this complication.

FIGURE 65-6. These two photographs are sections of a computed tomographic scan of a pagetic skull, obtained from a patient with radiographic evidence of basilar invagination in whom coma and a unilateral dilated pupil had developed acutely. A, The thickened cortex of the skull is demonstrated, and marked dilatation of the lateral ventricles reveals evidence of severe hydrocephalus. B, Virtual encasement of the brainstem region by increased amounts of pagetic bone is shown (arrows). Chronic obstruction to cerebrospinal fluid flow had developed in this patient as a result of pagetic bone encroachment. A ventricular shunt was performed immediately, with rapid resolution of her acute, central nervous system decompensation. (From Siris ES, Canfield RE, Jacobs TP. Paget's disease of bone. Bull N Y Acad Med 1980; 56:285.)

Involvement of one or more vertebral bodies also may be associated with neurologic signs and symptoms. Sometimes radiculopathy from nerve root compression, particularly in the lumbar region, is observed. Lumbar spinal stenosis and vertebral compression deformities may be seen. Less common, but more serious, is the occurrence of spinal cord compression syndromes, especially at the narrowed thoracic region but also in the lumbar region. Occasionally, the injury to the cord is thought to result not from direct compression by bone, but rather from ischemia of the neural structures, owing to a diversion of blood to the extremely vascular pagetic bone (i.e., a steal syndrome). 37 Progressive development of sensory changes and motor weakness below the affected level may be slow or relatively rapid, and always requires immediate attention. The cardiovascular complications of Paget disease include increased cardiac output related to increased blood flow in the thickened and vascular bone. These patients also are prone to the development of vascular calcifications. The aortic valve, in particular, has been shown to develop calcification in parallel with advancing stages of the disease.38 Malignant degeneration in pagetic bone occurs with an incidence of <1%.3 The development of increasing pain at a pagetic site may be the first evidence of this rare but extremely grave complication. In 50 years of experience at the Mayo Clinic, the site of malignant change most commonly was the pelvis, with the femur and humerus next in frequency.39 Most of the tumors were osteosarcomas, although fibrosarcomas and chondrosar-comas also occurred. Death resulting from local extension of tumor or from pulmonary metastases occurs in most cases within 1 to 3 years, because no effective treatment regimens are available. Benign giant cell tumors also may arise in pagetic bone, typically presenting as localized protuberances at the involved site that may show lytic changes on radiography. These tumors often initially are thought to represent osteosarcomas, but after biopsy they prove to be clusters of large osteoclasts, which may represent reparative granulomas.40 Many of these giant cell tumors are extremely sensitive to corticosteroids and may literally melt away after a course of dexamethasone.41 An interesting geographic or familial clustering of giant cell tumor cases among patients with Paget disease from the region of Avellino, Italy, has been described.41

DIAGNOSIS
The diagnosis of Paget disease is based on a combination of the medical history and physical examination (particularly when increased heat or bone deformity is present), laboratory tests (measurement of serum alkaline phosphatase, and in some instances, a urinary marker of bone resorption), and radiographic studies (bone scans and conventional radiographs). Bone biopsy is indicated only occasionally in this disorder, primarily as a means of distinguishing between Paget disease and malignant conditions. Perhaps the most specific and straightforward diagnostic test in the evaluation of Paget disease is a radiograph of an involved site, because the changes in the appearance of bone are characteristic (see Fig. 65-1). These include enlargement or expansion of bone, coarsening of trabecular markings, and pathognomonic patterns of lytic and sclerotic change. A bone scan is not a specific diagnostic test for Paget disease but is extremely useful, primarily for its sensitivity in identifying active sites of involvement. Radiographs of these areas then can be obtained to confirm the diagnosis and to provide information on the appearance of the process.

For example, radiographs of an involved femur will reveal the extent of sclerotic versus lytic change, provide data on the status of the hip joint, identify areas of fissure fractures if present, and demonstrate bowing. The serum alkaline phosphatase level is probably the most useful blood test in the diagnostic evaluation. A markedly elevated serum alkaline phosphatase level in a healthy older patient with no evidence of liver disease and a normal serum calcium level probably is due to Paget disease. The degree of elevation of the alkaline phosphatase level is generally a measure of metabolic activity. For reasons not well understood, the highest levels usually are seen in patients whose involved sites include the skull. Minimally elevated levels of alkaline phosphatase in patients with a single site of pagetic change (e.g., a tibia or a single vertebra) may represent progressive disease at that site. Conversely, extensive skeletal involvement identified radiographically and a minimal alkaline phosphatase increase denote relatively inactive or burned out disease in most instances. In patients with coexistent liver disease, serum bone-specific alkaline phosphatase is the most helpful marker. Although a urinary marker of bone resorption reflects more directly the initiating bone-resorptive lesion, its measurement is not usually required for diagnosis. Occasionally, however, a patient may present with radiographic evidence of Paget disease and a normal serum alkaline phosphatase level but an elevated level of urinary resorption marker, presumably indicating ongoing lytic pagetic change. Measurement of serum osteocalcin is not helpful in Paget disease. Problems with differential diagnosis occur infrequently because of the characteristic radiographic appearance in established Paget disease. Occasionally, however, a patient may have bone pain, elevated levels of serum alkaline phosphatase and a urinary resorption marker, a positive bone scan result, and less than typical osteolytic or sclerotic changes on radiography. In this situation, uncertainty may exist as to whether the disorder is Paget disease or metastatic cancer to bone. Old radiographic films and laboratory reports may be helpful in such cases; if the serum alkaline phosphatase level and radiograph of the involved bone were normal a year earlier, Paget disease is much less likely, particularly in a 60- or 70-year-old person. Occasionally, someone with long-standing Paget disease develops nonspecific radiographic changes in a new area. Because the spread of Paget disease to new sites years after its initial appearance is distinctly unusual, the clinical suspicion of another process should be high. When this type of diagnostic problem is encountered, bone biopsy may be indicated.

TREATMENT
Issues regarding therapy for Paget disease include questions about the indications for specific medical treatment, anticipated benefits from these treatments, choices among the antipagetic agents, and the utility of nonspecific therapies (e.g., analgesics, nonsteroidal antiinflammatory drugs). In patients with a range of signs and symptoms, the localized nature of this disorder makes highly individualized treatment decisions necessary. The goals of treatment are to relieve symptoms and to attempt to prevent future complications. The currently available therapies for Paget disease in the United States include the bisphosphonate agents etidronate, pamidronate, alendronate, tiludronate, and risedronate, as well as salmon calcitonin and plicamycin. All appear to act by decreasing osteoclastic bone resorption. The first-choice therapy is a potent bisphosphonate, but parenteral salmon calcitonin may be an alternative for patients who are not candidates for bisphosphonate therapy. Plicamycin has largely been replaced by the potent newer bis-phosphonates, especially intravenous pamidronate. Initial and early reductions in the levels of urinary resorption markers, followed by decreases in serum alkaline phosphatase (reflecting a subsequent slowing of new bone formation) nearly always result from appropriate treatment with these agents. Consequently, a variety of pagetic symptoms may be ameliorated. Moreover, bone biopsies taken before and after courses of treatment provide evidence for normalization of bone morphology (see Fig. 65-4B). Thus, the goal of relieving symptoms is one that often can be met. Some data suggest that the prevention of future complications through an arrest or slowing of active pagetic bone turnover may be possible in some patients, especially with the newer bisphosphonates that have the capacity to achieve a biochemical remission for extended periods of time in most patients.42 Therefore, the indications for treatment include (a) symptoms referable to Paget disease and (b) asymptomatic but active involvement (i.e., associated with elevated bone turnover markers) in areas of the skeleton where future serious complications might arise. Examples of this second indication include marked skull or vertebral involvement, which carries the threat of future neurologic problems; involvement of bones adjacent to major joints, such as the hip, knee, or ankle, with the potential for disabling joint problems; or pagetic change in a long bone, particularly in the lower extremity, so that a future bowing deformity is possible with disease progression over time. These issues are especially important in younger patients who may live with the disorder for many years. On the other hand, a small area of pagetic change in the calvarium or asymptomatic disease in the pelvis not involving the hip joint may be observed without treatment, particularly in elderly patients. The kinds of symptoms that are likely to respond to treatment also must be considered. The early clinical trials with bisphosphonates made it clear that a substantial placebo effect occurs from virtually any therapy. However, when the experience with treatment is analyzed carefully, several points become apparent. The reduction in the activity of Paget disease often decreases bone pain, pagetic headache, excessive warmth, some arthritic complaints, and, in some instances, the signs and symptoms associated with neurologic compression syndromes. In the relatively rare patient with high-output cardiac failure associated with extensive Paget disease (usually a person with underlying heart disease), treatment of the Paget disease may lead to some amelioration of cardiac symptoms. Hearing loss would not be expected to reverse itself (although progression might be slowed), bowed limbs may continue to bow with weight-bearing, and severely narrowed joint spaces will remain arthritic, although perhaps become less painful. BISPHOSPHONATES Five bisphosphonate compounds are used to treat Paget disease in the United States, including the first oral agent used for the disorder, etidronate; the only intravenous preparation, pamidronate; and three newer oral agents, tiludronate, alendronate, and risedronate. These are discussed in the order in which they became available. Etidronate. Etidronate became available at approximately the same time as salmon calcitonin.43,44 Indeed, these two agents were the mainstays of therapy for this disorder for nearly 20 years, after which the newer and more potent bisphosphonates emerged. Etidronate is given orally at a dose of 5 mg/kg per day (400 mg in most patients), taken with a small amount of water or clear juice at a point midway through a 4-hour fast (e.g., between two meals at least 4 hours apart or at bedtime at least 2 hours after any food has been ingested). It is given for a treatment course of 6 months' duration, typically achieving ~50% reduction in elevated indices of pagetic bone turnover in most patients, as well as a relief of those symptoms likely to respond to decreased turnover as previously noted. Thus, in patients with mild disease, indices may return to normal and remain that way with continued symptom relief for several years after a single 6-month course.45 However, individuals with initial elevations of serum alkaline phosphatase more than three to four times the upper limit of normal typically continue to have biochemically elevated turnover indices despite the 50% or so reduction, even if symptoms improve. Such patients are typically retreated at the same dose in repeated cycles of 6 months on, 6 months off therapy, a program that has maintained a reduced level of turnover for up to 10 years in some patients,45 although others have experienced a loss of effectiveness with repeated cycles.46 The failure of indices to return to normal or near normal in most of the more severely affected patients has been suggested as one explanation for the finding of disease progression in many patients treated with etidronate.47 Interestingly, higher dosages of etidronate for longer treatment intervals produce greater degrees of biochemical suppression than the recommended regimen, but clearly such therapy often leads to clinically significant osteomalacia.43 This is the basis for the recommendation that the 400-mg dose be given for no longer than 6 months, followed by at least a 6-month etidronate-free interval, an approach that is generally quite safe. All bisphosphonates have the capacity to impair mineralization of newly forming bone if enough is given for a long enough time period (the process reverses itself once the agent is stopped), but, of the currently available bisphosphonates, only etidronate at excessive dosages is likely to do so. Etidronate has also been shown to be associated with progression of lytic disease in some patients with this manifestation of the disorder48; thus, the drug should not be used in individuals with advancing lytic disease in weight-bearing bone or in patients with healing fractures. As is noted later, the newer bisphosphonates do not appear to impair mineralization at the therapeutic doses that are highly effective in treating this disease. Pamidronate. The first of the more potent bisphosphonates to become available in the United States is intravenously administered pamidronate. This agent has been extensively studied using a variety of different regimens for its administration,49,50,51,52 and 53 all of which are effective. Overall, pamidronate use clearly is associated with reduction in symptoms, improvement in lytic disease, and normalization of serum alkaline phosphatase for a year or more after a given number of infusions (dose and frequency may be individualized to the characteristics of the given patient). The package insert for the use of pamidronate in Paget disease calls for a dose of 30 mg daily for 3 days, each given in 500 mL of 5% dextrose in water over 4 hours, because this was the protocol used in the clinical trial that led to Food and Drug Administration clearance. This regimen actually resulted in an ~50% reduction in elevated indices. However, clinical experience suggests that this approach is cumbersome and not at all optimal for achieving the substantial benefits of this agent. In patients with relatively mild or monostotic disease, a single infusion of 60 or 90 mg infused over 2 to 3 hours (volume must be adequate to avoid superficial phlebitis at the infusion site, so the rapidity of the treatment depends on how quickly the patient can receive 400 to 500 mL of fluid) usually produces a normal serum alkaline phosphatase level with a prolonged biochemical remission before indices again rise above normal and retreatment is required.53 In persons with more severe disease (i.e., serum alkaline phosphatase values 5 to 10 times normal or higher) who rarely achieved normal values with etidronate or calcitonin, 60 to 90 mg can be given weekly or biweekly (depending on doctor and patient convenience) for three or four doses, with a reassessment of the indices 6 to 8 weeks after the final dose. After this, additional infusions can be given if needed (i.e., if the chemistries are still significantly elevated) to attempt to bring the alkaline phosphatase into the normal range over the next few weeks to months. Although some patients do not achieve normal indices even after use of very high doses,54 most do or at least come much closer to normal than was possible with the older agents. Typically, the remission lasts a year or more before indices rise above normal; this means that the patient can be followed with an alkaline phosphatase measurement every 6 months. Because retreatment will be given when values rise slightly above normal, the total dose for retreatment may be a fraction of what was initially required. Asymptomatic localized mineralization defects have been reported with as little as 180 mg of pamidronate,55 but this is a rare finding in the literature. More significant, perhaps, has

been the suggestive evidence that some degree of secondary resistance to pamidronate may occur with repeat treatment courses, requiring the use of higher dosages or a change to another, more potent bisphosphonate.56 The main side effect of pamidronate is a transient acute-phase reaction with fever and myalgias a day after the first-ever infusion. Tiludronate. Three oral bisphosphonates are available that are more potent than etidronate and that do not appear to be associated with mineralization defects at clinically effective doses. Tiludronate is somewhat more potent than etidronate and is given at a dosage of 400 mg per day for 3 months, with a follow-up observation period off therapy of 3 months, during which the serum alkaline phosphatase level continues to decline to its posttreatment nadir. This approach has reduced the serum alkaline phosphatase level by 50% to 60% in ~60% of patients after 3 months and resulted in a normal serum alkaline phosphatase in 24% to 35% of subjects in clinical trials.57,58 The agent is given in the form of two 200-mg tablets (ingested at the same time) taken with 6 to 8 oz of water at least 2 hours before or after consumption of food. Calcium supplements must not be taken for at least 2 hours after tiludronate is taken. Generally, the drug is well tolerated, with only mild upper gastrointestinal tract upset observed in a minority of patients. Patients should be followed with measurement of serum alkaline phosphatase every 3 to 6 months and should be retreated once indices rise above the prior nadir. Alendronate. Alendronate is several hundred times more potent than etidronate and substantially more potent than pamidronate. It is given in a dosage of 40 mg daily for 6 months. The agent must be taken on arising, on an empty stomach, with 8 oz of plain (not mineral) water. The patient should remain upright and should not eat or drink anything but water (including no other medications) for the next 30 minutes. This regimen has produced a normal serum alkaline phosphatase in 48% to 63% of patients with moderately severe disease59,60; among patients whose serum alkaline phosphatase normalized, in approximately half, levels were still within normal limits 24 months after the end of the initial 6-month course of treatment. Analysis of bone biopsy specimens in these studies indicated no evidence of mineralization impairment and revealed normal bone deposition at sites of newly forming bone after successful treatment. Healing of lytic fronts is also seen with alendronate. In the clinical trials, up to 17% of patients had generally mild upper gastrointestinal tract side effects with rare esophageal erosions. Serum alkaline phosphatase can be measured every 4 to 6 months or so, with retreatment once levels rise above normal or above the posttreatment nadir. Risedronate. Risedronate, which is more potent than alendronate and pamidronate, is given at a dosage of 30 mg per day for 2 months as the basic regimen. It is taken in a manner identical to that for alendronate (see earlier). Its use for 2 to 3 months in clinical trials by patients with moderate to severe disease resulted in a normal serum alkaline phosphatase level in 50% to 70% of patients; most of those who attained normal values maintained these values for up to 16 months after the end of the 2-month course. 61,62 Like alendronate, risedronate causes no impairment of mineralization at therapeutic doses and promotes normal bone deposition. Approximately 15% of patients have mild upper gastrointestinal tract side effects. Both pamidronate and risedronate have been associated with iritis in rare instances. Because normalization of biochemical indices and a prolonged remission is a major goal of treatment, the choice of bis-phosphonate should take into account each agent's likelihood of achieving that goal. With any potent bisphosphonate (e.g., pamidronate, alen-dronate, risedronate), confirming that the patient is vitamin D replete and is receiving adequate oral calcium is important; to this end, a wise course is to offer 500 mg of calcium twice daily and 400 to 800 U of vitamin D each day to avoid hypocalcemia and transient secondary hyperparathyroidism. CALCITONIN The biochemistry and physiology of calcitonin secretion are discussed elsewhere (see Chap. 53). The form of calcitonin that is clinically available is synthetic salmon calcitonin. Because calcitonin is a peptide hormone, it cannot be taken orally. The form approved for the treatment of Paget disease in the United States is administered parenterally. Use of the nasal spray form of salmon calcitonin in Paget disease has been studied,63 but its utility for most patients has not been determined. Salmon calcitonin therapy is ordinarily initiated at a dose of 100 U injected subcutaneously by the patient each day. As early as 2 to 3 weeks later, symptomatic benefit may be appreciated, and any benefit that is likely to occur should be apparent by the end of 3 months. Therapeutic responses to calcitonin include the reduction of both bone pain and excessive warmth. Radiographic improvement with healing of lytic areas and improvement of neurologic complications has been reported. Serum alkaline phosphatase and urinary hydroxyproline values generally decline to ~50% of pretreatment levels by the end of the first 6 months of treatment,64 after which the calcitonin dosage often may be reduced to 50 to 100 U three times a week, with maintenance of symptomatic and biochemical benefit. The discontinuation of treatment after an initial 1- to 2-year course may be followed by another year or more of persistent disease suppression, but a return of the biochemical indices to pretreatment values usually occurs after this.65 Thus, depending on the patient's age, extent of disease, and symptoms, an indefinite course of treatment at the lowest effective dosage (e.g., typically recommended to be 50 U three times a week) may be indicated. After a prolonged period of treatment with salmon calcitonin, biochemical indices may begin to rise again in some patients, often back to pretreatment levels. In some cases, this escape from effectiveness reflects the development of high titers of neutralizing antibodies to the salmon hormone.66 However, in other instances titers are not increased, and the suggestion has been made that a down-regulation of calcitonin receptors at target cells may occur, with a resulting decreased therapeutic response. Treatment with salmon calcitonin at full dosages is expensive, and sometimes the need for parenteral administration elicits problems with continued patient compliance. Side effects, which occur in 10% to 20% of patients, consist of nausea, less commonly vomiting or diarrhea, and occasionally marked cutaneous flushing. For this reason, treatment may be initiated using only 50% of the desired dose each day for the first week, with an increase to the full dose if the agent is well tolerated; side effects may be less troublesome if the agent is given at bedtime; in many patients, they diminish substantially as treatment continues. Calcitonin causes a mild and usually asymptomatic decrease (~1 mg/dL) of serum calcium 4 to 6 hours after injection in patients with active Paget disease, but this change in the serum calcium level diminishes as the pagetic process responds to therapy. PLICAMYCIN (MITHRAMYCIN) Plicamycin is a cytotoxic compound that binds to DNA and inhibits RNA synthesis. It is an effective therapy for the hyper-calcemia of malignancy by virtue of its inhibition of osteoclast-mediated bone resorption. For this same reason, it is an effective means of treating Paget disease.67 However, its significant systemic toxicity (nausea, vomiting, elevation of hepatocellular enzymes, and, less commonly, reversible nephrotoxicity, as well as thrombocytopenia and a hemorrhagic diathesis) markedly limits its use. It is rarely needed today because of the availability of potent bisphosphonates such as pamidronate, alendronate, and risedronate. Nevertheless, in the rare patient who needs treatment and who fails to respond to such agents, a course of plicamycin (12.5 to 20 g/kg given intravenously over 4 to 8 hours every second to third day for five to ten total doses) may be of significant value. NONSTEROIDAL ANTIINFLAMMATORY DRUGS Many of the most troublesome symptoms for patients with Paget disease result from a secondary arthritis involving the hip, knee, ankle, or spine. Analgesics such as aspirin or acetam-inophen may provide pain relief, and the use of nonsteroidal antiinflammatory compounds or COX-2 inhibitors can be extremely effective in alleviating symptoms caused by arthritic complications of Paget disease. In patients with metabolically inactive Paget disease (i.e., normal serum alkaline phosphatase levels) but with significant joint symptoms, one of these agents may be the only required therapy. In cases in which Paget disease is active, these compounds may serve as useful adjuncts to either a bisphosphonate or calcitonin. SURGICAL INTERVENTION Several of the mechanical complications of Paget disease can best be managed surgically.68 Joint replacement may be the only means of eliminating severe pain, particularly in the hip, but should be undertaken only by an orthopedic surgeon who has had experience in operating on vascular, pagetic bone.69 Other elective orthopedic procedures include internal stabilization of an impending fracture, particularly when the lytic changes are progressive and pain develops, and, occasionally, osteotomy of a severely bowed limb when a serious gait disturbance or pain unrelieved by medical therapy is interfering with daily function. In general, elective surgery should be delayed until disease activitymanifested by the level of serum alkaline phosphataseis reduced to the lowest possible level through the use of a course of intravenous pamidronate or, if time permits, oral alendronate or risedronate. If the alkaline phosphatase level is reduced to normal or at least to under two to three times the upper limit of normal, hypervascularity of the pagetic bone should be substantially decreased, limiting the potential for excessive bleeding at surgery. Sometimes neurosurgical intervention is necessary as therapy for rapidly progressive spinal cord compression or for neurologic decompensation caused by obstructive hydrocephalus. In acute situations, high doses of dexamethasone (up to 16 mg per day in divided doses) in the hours before surgery also may decrease vascularity and provide some initial degree of decompression.

CHAPTER REFERENCES
1. Barker DJP. The epidemiology of Paget's disease. Br Med Bull 1984; 40:396. 1a.Altman RD, Block DA, Hochberg MC, Murphy WA. Prevalence of pelvic Paget's disease of bone in the United States. J Bone Miner Res 2000; 15:461. 2. Barker DJP, Chamberlain AT, Guyer PB, Gardner MJ. Paget's disease of bone: the Lancashire focus. BMJ 1980; 280:1105. 2a.Cooper C, Shafheutle K, Dennison E, et al. The epidemiology of Paget's disease in Britain: is the prevalence decreasing? J Bone Miner Res 1999; 14:192. 2b.Cundy T, McAnulty K, Wattie D, et al. Evidence for secular changes in Paget's disease. Bone 1997; 20:69. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. Barry HC. Paget's disease of bone. Edinburgh: E & S Livingstone, 1969. Guyer PB, Chamberlain AT. Paget's disease of bone in two American cities. BMJ 1980; 280:985. Singer FR. Paget's disease of bone. New York: Plenum Publishing, 1977. Siris ES, Canfield RE, Jacobs TP. Paget's disease of bone. Bull N Y Acad Med 1980; 56:285. Morales-Piga AA, Rey-Rey JS, Corres-Gonzalez J, et al. Frequency and characteristics of familial aggregation of Paget's disease of bone. J Bone Miner Res 1995; 10:663. McKusick VA. Heritable disorders of connective tissue. St. Louis: CV Mosby, 1972:718. Siris ES, Ottman R, Flaster E, Kelsey JL. Familial aggregation of Paget's disease of bone. J Bone Miner Res 1991; 6:495 Singer FR, Mills BG, Park MS, et al. Increased HLA-DQw1 antigen frequency in Paget's disease of bone. Proceedings of the 7th Annual Meeting of the American Society for Bone and Mineral Research. Washington, DC: 1985:128. Tilyard MW, Gardner RJM, Milligan L, et al. A probable linkage between familial Paget's disease and the HLA foci. Aust N Z J Med 1982; 12:498. Foldes J, Shamir S, Brautbar C, et al. HLA-D antigens and Paget's disease of bone. Clin Orthop 1991; 266:301. Singer FR, Siris ES, Knieriem A, et al. The HLA DRB1*1104 gene frequency is increased in Ashkenazi Jews with Paget's disease of bone. J Bone Miner Res 1996; 11:S369. Hughes AE, Shearmen AM, Weber JL, et al. Genetic linkage of familial expansile osteolysis to chromosome 18q. Hum Mol Genet 1994; 3:359. Cody JD, Singer FR, Roodman GD, et al. Genetic linkage of Paget disease of the bone to chromosome 18q. Am J Hum Genet 1997; 61:1117. Haslam SI, Van Hul W, Morales-Piga A, et al. Paget's disease of bone: evidence for a susceptibility locus on chromosome 18q and for genetic heterogeneity. J Bone Miner Res 1998; 13:911. Rasmussen H, Bordier P. The physiological and cellular basis of metabolic bone disease. Baltimore: Williams & Wilkins, 1974:293. Mills BG, Singer FR. Nuclear inclusions in Paget's disease of bone. Science 1976; 194:201. Rebel A, Malkani K, Basle M, Bregeon C. Nuclear inclusions in osteoclasts in Paget's bone disease. Calcif Tissue Res 1976; 21(Suppl):113. Mills BG, Yabe H, Singer FR. Osteoclasts in human osteopetrosis contain viral-nucleocapsid-like nuclear inclusions. J Bone Miner Res 1988; 3:101. Rebel A, Basle M, Pouplard A, et al. Viral antigens in osteoclasts from Paget's disease of bone. Lancet 1980; 2:344. Mills BG, Singer FR, Weiner LP, et al. Evidence for both respiratory syncytial virus and measles virus antigens in the osteoclasts of patients with Paget's disease of bone. Clin Orthop 1984; 183:303. Gordon MT, Anderson DC, Sharpe PT. Canine distemper virus localized in bone cells of patients with Paget's disease. Bone 1991; 12:195. Fraser KB, Martin SJ. Measles virus and its biology. New York: Academic Press, 1978. Gordon MT, Mee AP, Anderson DC, Sharpe PT. Canine distemper virus transcripts sequenced from pagetic bone. Bone Miner 1992; 19:159. Ralston SH, Digiovine FFS, Gallacher SJ, et al. Failure to detect paramyxo-virus sequences in Paget's disease of bone using the polymerase chain reaction. J Bone Miner Res 1991; 6:1243. Birch MA, Taylor W, Fraser WD, et al. Absence of paramyxovirus RNA in cultures of pagetic bone cells and in pagetic bone. J Bone Miner Res 1994; 9:11. Reddy SV, Singer FR, Roodman GD. Bone marrow mononuclear cells from patients with Paget's disease contain measles virus nucleocapsid messenger ribonucleic acid that has mutations in a specific region of the sequence. J Clin Endocrinol Metab 1995; 80:2108. Reddy SV, Singer FR, Mallette L, Roodman GD. Detection of measles virus nucleocapsid transcripts in circulating blood cells from patients with Paget's disease. J Bone Miner Res 1996; 11:1602. Roodman GD, Kurihara N, Oksaki Y, et al. IL-6. A potential autocrine/paracrine factor in Paget's disease of bone. J Clin Invest 1992; 89:46. Hoyland JA, Freemont AJ, Sharpe PT. Interleukin-6, IL-6 receptor, and IL-6 nuclear factor gene expression in Paget's disease. J Bone Miner Res 1993; 9:75. Posen S, Clifton-Bligh P, Wilkinson W. Paget's disease of bone and hyperpara-thyroidism: coincidence or causal relationship? Calcif Tissue Res 1978; 26:107. Meunier PJ, Coindre JM, Edouard CM, Arlot ME. Bone histomorphometry in Paget's disease: quantitative and dynamic analysis of pagetic and non-pagetic bone tissue. Arthritis Rheum 1980; 23:1095. Siris ES, Clemens TP, McMahon D, et al. Parathyroid function in Paget's disease of bone. J Bone Miner Res 1989; 4:75. Barry HC. Orthopedic aspects of Paget's disease of bone. Arthritis Rheum 1980; 23:1128. Monsell EM, Bone HG, Cody DD, et al. Hearing loss in Paget's disease of bone: evidence of auditory nerve integrity. Am J Otol 1995; 16:27. Herzberg L, Bayliss E. Spinal cord syndrome due to non-compressive Paget's disease of bone: a spinal artery steal phenomenon reversible with calcitonin. Lancet 1980; 2:13. Strickberger SA, Schulman SP, Hutchins GM. Association of Paget's disease of bone with calcific aortic valve disease. Am J Med 1987; 82:953. Wick MR, Siegal GP, Unni KK, et al. Sarcomas of bone complicating osteitis deformans (Paget's disease). Am J Surg Pathol 1981; 5:47. Upchurch KS, Simon LS, Schiller AL, et al. Giant cell reparative granuloma of Paget's disease of bone: a unique clinical entity. Ann Intern Med 1983; 98:35. Jacobs TP, Michelsen J, Polay JS, et al. Giant cell tumor in Paget's disease of bone. Familial and geographic clustering. Cancer 1979; 44:742. Siris ES. Clinical review. Paget's disease of bone. J Bone Miner Res 1998; 13:1061. Canfield R, Rosner W, Skinner J, et al. Diphosphonate therapy of Paget's disease of bone. J Clin Endocrinol Metab 1977; 44:96. Khairi MRA, Altman RD, DeRosa GP, et al. Sodium etidronate in the treatment of Paget's disease of bone. Ann Intern Med 1977; 87:656. Siris ES, Canfield RE, Jacobs TP, et al. Clinical and biochemical effects of EHDP in Paget's disease of bone: patterns of response to initial treatment and to long-term therapy. Metab Bone Dis Relat Res 1981; 3:301. Altman R. Long-term follow-up of therapy with intermittent etidronate disodium in Paget's disease of bone. Am J Med 1985; 79:583. Meunier PJ, Vignot E. Therapeutic strategy in Paget's disease of bone. Bone 1995; 17:489S. Nagant de Deuxchaisnes C, Rombouts-Lindeman C, Huaux JP, et al. Roent-genologic evaluation of the action of the diphosphonate EHDP and of combined therapy (EHDP and calcitonin) in Paget's disease of bone. In: MacIntyre I, Szelke M, eds. Molecular endocrinology. Amsterdam: Elsevier, 1979:405. Cantrill JA, Buckler HM, Anderson DC. Low dose intravenous 3-amino-1-hydroxypropylidene-1,1-bisphosphonate (APD) for the treatment of Paget's disease of the bone. Ann Rheum Dis 1986; 45:1012. Gallacher SJ, Boyce BF, Patel U, et al. Clinical experience with pamidronate in the treatment of Paget's disease of bone. Ann Rheum Dis 1991; 50:930. Harinck HI, Papapoulos SE, Blanksma HJ, et al. Paget's disease of bone: early and late responses to three different modes of treatment with amino-hydroxypropylidene biphosphonate (APD). BMJ 1987; 295:1301. Patel S, Stone MD, Coupland C, Hosking DJ. Determinants of remission in Paget's disease of bone. J Bone Miner Res 1993; 8:1467. Thiebaud D, Jaeger P, Gobelet C, et al. A single infusion of AHPrBP (APD) as treatment of Paget's disease of bone. Am J Med 1988; 85:207. Cundy T, Wattie D, King AR. High dose pamidronate in the management of Paget's disease of bone. Calcif Tissue Res 1996; 58:6. Adamson BB, Gallacher SJ, Byars J, et al. Mineralisation defects with pamidronate therapy for Paget's disease. Lancet 1993; 342:1459. Gutteridge DH, Ward LC, Stewart GO, et al. Paget's disease: acquired resistance to one aminobisphosphonate with retained response to another. J Bone Miner Res 1999; 14:79. Roux C, Gennari C, Farrerons J, et al. Comparative prospective, double-blind, multicenter study of the efficacy of tiludronate and etidronate in the treatment of Paget's disease of bone. Arthritis Rheum 1995; 38:851. McClung M, Tou CPK, Goldstein NH, Picot C. Tiludronate therapy for Paget's disease of bone. Bone 1995; 17:493S. Siris E, Weinstein RS, Altman R, et al. Comparative study of alendronate versus etidronate for the treatment of Paget's disease of bone. J Clin Endo-crinol Metab 1996; 81:961. Reid IR, Nicholson GC, Weinstein RS, et al. Biochemical and radiologic improvement in Paget's disease of bone treated with alendronate: a randomized, placebo-controlled trial. Am J Med 1996; 171:341. Siris ES, Chines AA, Altman RD, et al. Risedronate in the treatment of Paget's disease: an open-label, multicenter study. J Bone Miner Res 1998; 13:103. Miller PD, Brown JP, Siris ES, et al. A prospective, multicenter, randomized, double-blind comparison of risedronate and etidronate in the treatment of Paget's disease of bone. Am J Med 1999; 106:513. Nagant de Deuxchaisnes C, Devogelaer JP, Huaux JP, et al. Effect of a nasal spray of salmon calcitonin in normal subjects and in patients with Paget's disease. Proceedings of the 7th Annual Meeting of the American Society for Bone and Mineral Research. Washington, DC: 1985:369. DeRose J, Singer FR, Avramides A, et al. Response of Paget's disease to porcine and salmon calcitonins. Am J Med 1974; 56:858. Avramides A, Flores A, DeRose J, Wallach S. Paget's disease of bone: observations after cessation of long-term synthetic salmon calcitonin treatment. J Clin Endocrinol Metab 1976; 42:459. Haddad J, Caldwell JG. Calcitonin resistance: clinical and immunological studies in subjects with Paget's disease of bone treated with porcine and salmon calcitonins. J Clin Invest 1972; 51:3133. Ryan WG, Schwartz TB, Perlia CP. Effects of Mithramycin on Paget's disease of bone. Ann Intern Med 1969; 70:549. Sochart DH, Porter ML. Charnley low-friction arthroplasty for Paget's disease of the hip. J Arthroplasty 2000; 15:210. McDonald DJ, Sim FH. Total hip arthroplasty in Paget's disease. A follow-up note. J Bone Joint Surg Am 1987; 69:766.

CHAPTER 66 RARE DISORDERS OF SKELETAL FORMATION AND HOMEOSTASIS Principles and Practice of Endocrinology and Metabolism

CHAPTER 66 RARE DISORDERS OF SKELETAL FORMATION AND HOMEOSTASIS

MICHAEL P. WHYTE Osteopetrosis Pycnodysostosis Progressive Diaphyseal Dysplasia (Camurati-Engelmann Disease) Infantile Cortical Hyperostosis (Caffey Syndrome) Endosteal Hyperostosis Fluorosis Osteopoikilosis Melorheostosis Mixed Sclerosing Bone Dystrophy Fibrodysplasia (Myositis) Ossificans Progressiva Fibrous Dysplasia Osteogenesis Imperfecta Hypophosphatasia Axial Osteomalacia Fibrogenesis Imperfecta Ossium Metaphyseal and Spondylometaphyseal Dysplasias Idiopathic Juvenile Osteoporosis Idiopathic Osteolysis Ectopic Calcification Ischemic Necrosis Chapter References

Although they are individually rare, many unusual disorders of skeletal formation and homeostasis exist.1,2,3,4,5,6 and 7 Some entities are mere radiographic curiosities; others are lethal. Some are associated with well-defined endocrine or metabolic abnormalities, and the endocrinologist may be consulted for his or her expertise, especially regarding metabolic bone disease. Cumulatively, the number of affected individuals is significant. Accordingly, this chapter provides a review of several of these conditions.

OSTEOPETROSIS
Osteopetrosis (marble bone disease) is one of many disorders that cause high bone mass (Table 66-1).6 Two principal forms are well characterized: an autosomal dominant (benign) type with few or no symptoms, and an autosomal recessive (malignant) type that is typically fatal during the first decade of life if untreated.8,9 Both of these forms of osteopetrosis have been genetically mapped,10,11 but the gene defects are partly known. An intermediate form of osteopetrosis also exists; it is inherited as an autosomal recessive trait and presents during childhood with some of the clinical manifestations of the malignant type, but its impact on life expectancy is unclear.12 Also, an autosomal recessive syndrome of osteopetrosis with renal tubular acidosis and cerebral calcification is seen; this form is due to an inborn error of metabolism characterized by deficiency of the carbonic anhydrase II (CA II) isoenzyme.13 At least nine forms of osteopetrosis are found in humans; most have a genetic basis.8,9,14

TABLE 66-1. Disorders That Cause High Bone Mass

All forms of osteopetrosis in humans result from a failure of osteoclasts to resorb skeletal tissue, including the cartilage deposited during endochondral bone formation (primary spongiosa).15 The precise cause of osteoclast malfunction is understood only for CA II deficiency, for which several mutations in the CA II gene have been characterized,13 and for a subset of patients with malignant osteopetrosis who have defects in a gene that encodes a vacuolar proton pump.13a A considerable variety of defects in osteoclast formation and function will likely be found to explain these clinically and genetically heterogeneous conditions,8,9,15,16 for which animal models increasingly provide insight concerning potential etiology and pathogenesis.17,18 In the malignant form of osteopetrosis, symptoms begin during infancy.19 Often, an early observation is nasal stuffiness attributable to malformation of paranasal and mastoid sinuses. Subsequently, patients fail to thrive and often develop palsies of the facial, optic, or oculomotor nerves from compression by narrow cranial foramina; delayed dentition; bone fragility; and infection with spontaneous bruising and bleeding secondary to myelophthisic anemia with extramedullary hematopoiesis.8,9 Some develop hydrocephalus.20 Short stature, a large head, frontal bossing, nystagmus, adenoid appearance, hepatosplenomegaly, and genu valgum may be apparent. Death results from bronchopneumonia, sepsis, hemorrhage, or severe anemia.8,9 Children with intermediate osteopetrosis are of short stature and may have macrocephaly, ankylosed teeth that predispose them to osteomyelitis of the jaw, and recurrent fractures. 12 Cranial nerve deficits can occur. Some develop progressive myelophthisic anemia. In the benign form of osteopetrosis, increased bone density can appear during childhood, but reportedly in two radiographic patterns.21 Although many affected individuals are asymptomatic, the long bones are brittle, and occasionally pathologic fractures occur. Osteomyelitis of the jaw, osteoarthritis, and in rare cases deafness and facial palsy are potential complications.22 CA II deficiency typically manifests in infancy or early childhood with fracture or failure to thrive; short stature, psychomotor delay, and visual or auditory impairment may be present.13 Hematologic abnormalities are usually absent. Most patients have a mixed renal tubular acidosis; hypokalemic paralysis has occurred in several affected individuals. The disorder seems compatible with a long life. Other, even rarer, types of osteopetrosis may have devastating or benign outcomes.8,9 Routine biochemical parameters of mineral homeostasis are usually normal in osteopetrosis. In some severely affected infants or children, hypocalcemia with secondary hyperparathyroidism and elevated serum 1,25-dihydroxyvitamin D levels can occur. Although the cause is poorly understood, acid phosphatase and creatine kinase (brain isoenzyme) are often aberrantly increased or detectable, respectively, in patient sera.8 In the malignant form, dysfunction of circulating monocytes and granulocytes is seen.23 Defective osteoclast action manifests radiographically as abnormalities in skeletal growth, modeling, and remodeling causing a generalized symmetrical increase in bone mass.5 In the axial skeleton, sclerosis of the base of the skull, underpneu-matization of the paranasal and mastoid sinuses, and a mask sign on anteroposterior view of the skull (especially prominent in children) are present. On lateral view of the spine, endo-bones (bone-within-bone) or a rugger-jersey appearance is seen. In the appendicular skeleton, erosion of the distal phalanges may be present. The classic change in the distal femur, an Erlenmeyer flask deformity, is shown in Figure 66-1. Skeletal scintigraphy may disclose fractures or osteomyelitis.

FIGURE 66-1. Osteopetrosis. The classic radiographic changes in the long bones are shown here in a distal femur and proximal tibia. They include increased radiodensity, loss of distinction between cortex and medullary space, modeling defects that result in flask-shaped tubular bones (especially in children), and radiodense transverse striations and longitudinal bands in the metaphyses.

Although the radiographic features can be diagnostic, the principal pathogenetic disturbance provides a histologic finding that is characteristic for all forms of osteopetrosis. Persistence of primary spongiosa is reflected by the presence of unresorbed islands or bars of cartilage encased within bone (Fig. 66-2). In the malignant form, osteoclasts are generally numerous, and the marrow space is crowded with fibrous tissue (see Fig. 66-2). In other forms of osteopetrosis, osteoclasts are present in increased, normal, or decreased numbers.8

FIGURE 66-2. Osteopetrosis. The characteristic histopathologic change in all forms of osteopetrosis reflects the failure of osteoclasts to resorb primary spongiosa; that is, lightly stained cartilage bars (white arrows) remain within darkly stained (well-mineralized) trabecular bone. As shown here in the malignant form, numerous osteoclasts (black arrows) and marked accumulation of fibrous tissue (FT) in the medullary space are also present. (Masson trichrome stain, 160).

Because intensive treatment is necessary for malignant osteopetrosis, whereas the prognosis for the intermediate form and for CA II deficiency is more benign,13,24,25 and 26 correct diagnosis is crucial.19 Infants or young children with CA II deficiency can have radiographic changes consistent with the malignant form of osteopetrosis, yet sequential studies may reveal gradual spontaneous resolution of their high bone mass and its complications.27 Bone marrow transplantation from HLA-identical normal donors to patients with malignant osteopetrosis has been followed by marked clinical improvement (including the reversal of hematologic, radiographic, histologic, and sometimes neurologic abnormalities).24,25 and 26,28 Demonstration after transplantation that osteoclasts, but not osteoblasts, are of donor origin is consistent with defective osteoclast-mediated bone resorption in osteopetrosis and development of these polykaryons from fusion of cells of monocyte/macrophage lineage produced by marrow hematopoietic stem cells.15,18 Although the responses have not been as dramatic or as well documented compared with those of the bone marrow recipients, favorable biochemical changes have been reported in a few patients with malignant osteopetrosis treated with large oral doses of 1,25-dihydroxyvitamin D3 (calcitriol) together with a low-calcium diet.9,18,26 Calcitriol potentiates osteoclast activity. Long-term trials, however, are needed. Interferon-g-1b is effective for malignant osteopetrosis and is especially appropriate for patients who are not good candidates for transplantation.29 Glucocorticoid treatment helps the hematologic problems.30 Transfusion of CA IIreplete erythrocytes to one woman with CA II deficiency did not correct her systemic acidosis.31 Hyperbaric oxygenation has been used for osteomyelitis. Surgical decompression of the facial and optic nerves has been described.20 Prenatal diagnosis of the malignant form of osteopetrosis may be possible by conventional radiographs late in the pregnancy.32 Successful diagnosis using ultrasonography is occasionally reported.33 CA II deficiency can be detected in utero by mutation analysis.34

PYCNODYSOSTOSIS
Pycnodysostosis, an autosomal recessive disorder, perhaps affected the French impressionist painter Henri de Toulouse- Lautrec.2,6 Patients present in childhood with disproportionate short stature, a relatively large cranium, frontooccipital prominence, small facies, obtuse mandibular angle, small chin, dental malocclusion, high palate, exophthalmos, blue-tinged sclerae, a pointed and beaked nose, and short, clubbed fingers with hypoplastic nails. Recurrent fractures, with deformity, scoliosis, kyphosis, and genu valgum, may occur. Generalized uniform osteosclerosis increases with age but is not accompanied by the modeling defects of osteopetrosis.5 Unlike in osteopetrosis, the fontanelles remain open (especially the anterior; Fig. 66-3), radiodense striations and endobones do not occur, wormian bones may be present near the parietal sutures, the mandibular angle is obtuse, and hypoplasia of the facial bones, clavicles, and terminal phalanges is present.

FIGURE 66-3. Pycnodysostosis. The skull of this 7-year-old child shows characteristic markedly widened cranial sutures, a flattened mandibular angle, several wormian bones at the occiput, and increased radiodensity at the base.

Defects in the cathepsin K gene have been found to cause Pycnod- ysostosis.35 Treatment for associated growth hormone deficiency seems helpful.36 Fractures of the long bones generally heal well.

PROGRESSIVE DIAPHYSEAL DYSPLASIA (CAMURATI-ENGELMANN DISEASE)

Progressive diaphyseal dysplasia is a sclerosing bone disorder that is inherited in an autosomal dominant pattern, but with variable penetrance.1,6,37 Patients have defects in the gene that encodes transforming growth factor B.37a New bone formation occurs progressively along both the periosteal and endosteal surfaces of major tubular bones and is associated with pain and gait disturbance. During childhood, the patient experiences leg pains and often has a limp.38,39 Muscle wasting and a paucity of subcutaneous fat in the extremities is common, and the gait may be waddling and broad-based. In severe cases, the axial skeleton (including the cranium) is affected. Patients may have an enlarged head, prominent forehead, proptosis, cranial nerve palsies, and a characteristic body habitus (Fig. 66-4). Sometimes puberty is delayed due to hypothalamic hypogonadotropic hypogonadism. Skeletal symptoms may remit during adolescence. Especially rare, mild, sporadic cases have been referred to as Ribbing disease.37 Such patients seem to represent new, relatively benign expressions of the disorder.

FIGURE 66-4. Progressive diaphyseal dysplasia. This 17-year-old girl has the typical appearance of severe disease. She is of normal height yet is prepubescent, has marked thickening of the long bones, and has a paucity of muscle mass and subcutaneous tissue in the extremities.

Routine studies of bone and mineral metabolism are typically normal. With severe disease, a modest hypocalcemia and hypocalciuria can be presentfindings that appear to reflect markedly positive calcium balance.39 Serum alkaline phosphatase activity and the erythrocyte sedimentation rate may be increased. Radiographic changes feature gradual, generally symmetrical, diaphyseal and metaphyseal sclerosis; the epiphyses are spared.5 Periosteal and endosteal new bone formation increase the outer diameter and cause medullary stenosis, respectively, of affected long bones (Fig. 66-5). The tibiae and femora are most frequently involved. When the skull is affected, progressive diaphyseal dysplasia may be difficult to distinguish from mild forms of craniodiaphyseal dysplasia.5 Although bone scintigraphy and radiographic findings are usually concordant, in some patients bone scanning may be unremarkable, whereas radiographic study shows distinct abnormalities.40 Here, the disease process may be quiescent. The converse situation also occurs and probably reflects active but early disease. Biopsy of affected skeletal tissue reveals periosteal and endosteal new bone that is dis-organized (woven) peripherally but undergoing maturation and cancellous compaction toward the cortex.37

FIGURE 66-5. Progressive diaphyseal dysplasia. There is marked cortical thickening of the tibial diaphysis from both periosteal (white arrows) and endosteal (black arrow) new bone formation.

Glucocorticoid therapygenerally low-dose, alternate-day treatment with prednisoneusually relieves bone pain; histologic improvement has been described.41 Courses of pamidronate may also be effective.42 Surgical removal of especially painful areas of cortical bone can be helpful (personal observation).37

INFANTILE CORTICAL HYPEROSTOSIS (CAFFEY SYNDROME)

Infantile cortical hyperostosis is a self-limited, focal skeletal disorder in which exuberant periosteal new bone causes transient swellings that involve adjacent musculature, fascia, and connective tissues. The swellings slowly resolve but often recur and can cause deformity.5,7 Autosomal dominant transmission with variable expression and incomplete penetrance can occur.43 Symptoms generally begin abruptly before 5 months of age. Fever, irritability, anorexia, pallor, and pleurisy may be present. Soft-tissue swellingsprecursors of the periosteal new bone formationare tender and hard but generally not warm. Some infants appear to have pseudoparalysis, as they splint affected bones. Erb palsy may be diagnosed erroneously before radiographic changes are noted.44 The clinical course is variable; manifestations may subside gradually over months, but occasionally the disease remains intermittently active for years. Patients may develop limb-length inequality or synostosis; some die of secondary infection. Caffey syndrome is diagnosed radiologically.5 The erythrocyte sedimentation rate, serum alkaline phosphatase activity, and leukocyte count may be elevated during episodes of swelling. Some patients have mild anemia. Skeletal involvement can be unifocal or may follow a sequential pattern of multifocal disease.5 The mandible, clavicles, and ribs are involved most frequentlysometimes symmetrically. The skull, scapulae, and long bones may also be affected. Characteristically, diaphyseal cortical hyperostosis begins as new bone within the soft-tissue mass. The swelling then joins with the subjacent osseous tissue, causing bone enlargement, with dense thickening of the cortex secondary to periosteal elevation. Histologically, the periosteum is inflamed and thickened. Immature lamellar bone is present initially but undergoes maturation. Marrow spaces may be filled with vascular fibrous tissue. Dystrophic calcification can occur. Destructive lesions can involve the skull. Infantile cortical hyperostosis generally resolves by 1 year of age. No specific medical or surgical treatment exists, although glucocorticoid therapy can cause prompt improvement. Patients should be observed and managed for symptoms. Sedation for irritability may be necessary.

ENDOSTEAL HYPEROSTOSIS
Endosteal hyperostosis is inherited either as a clinically severe autosomal recessive disorder (sclerosteosis or van Buchem disease) or as a relatively benign autosomal dominant condition (Worth type).6 The clinical and radiographic manifestations, however, are similar.1,6,45 During puberty, progressive asymmetric enlargement of the mandible occurs. Occasionally, facial palsy or deafness is caused by cranial nerve entrapment, but the head circumference is normal. Serum alkaline phosphatase activity may be increased in van Buchem disease. Radiographic abnormalities include sclerosis of the axial skeleton and the skull (where the table may be wide), mandibular enlargement, and thickening of the diaphyses of long bones due to selective endosteal hyperostosis (bones remain properly shaped).5 Some cases reported as van Buchem disease have features suggesting sclerostenosis or craniodiaphyseal dysplasia.46 Previously, sclerostenosis had been

differentiated from van Buchem disease because patients with the former condition were of excessive height and had syndactyly; however, both conditions map to chromosome 17q12-q21; thus, they may be allelic disorders.47 No effective medical treatment exists.

FLUOROSIS
Chronic ingestion or inhalation of large amounts of fluoride causes fluorosis.48 In the United States, prolonged administration of sodium fluoride as a therapy for postmenopausal osteoporosis represents an increasingly unlikely cause. Mottled discoloration of the teeth, diffuse bone pain reflecting an underlying osteomalacia, and a plantar fascia syndrome characterized by painful feet may be presenting features. Biochemical diagnosis by assay of the fluoride ion in 24-hour urine collections can be obtained through commercial laboratories. Early radiographic signs include irregularity at the point of attachment of muscles to the iliac spine and calcification of ligaments in the pelvis, vertebrae, and interosseous membranes.5 Later, generalized osteosclerosis can occur. Histopathologic studies may reveal osteomalacia (fluoride stimulates osteoid synthesis by osteoblasts, yet mineral deposition is impaired). Treatment consists of protection from fluoride exposure and, if osteomalacia is present, calcium supplementation to mineralize the excess osteoid (see Chap. 63 and Chap. 131).

OSTEOPOIKILOSIS
Osteopoikilosis (spotted bones) is a radiographic curiosity.5,6 If unrecognized, however, it may lead to investigation for metastatic disease, mastocytosis, tuberous sclerosis, or other disorders. It is inherited as an autosomal dominant trait and manifests as numerous well-defined, small, homogeneous, symmetric, round or oval, periarticular foci of osteosclerosis (Fig. 66-6). Bone scanning, however, is unremarkable.49 In some kindreds focal, asymptomatic accumulation of dermal elastin (dermatofibrosis lenticularis disseminata) is found, and the disorder is called Buschke-Ollendorff syndrome.50

FIGURE 66-6. Osteopoikilosis. Characteristic features shown here include the spotted appearance of the pelvis and meta-epiphyseal regions of the femora.

OSTEOPATHIA STRIATA Osteopathia striata is characterized radiographically by symmetric, regular, dense bony striations in the metaphyseal regions of tubular bones (Fig. 66-7) and by fan-like linear projections in the iliac wings.5 It is inherited as an autosomal dominant trait and is a radiographic curiosity unless associated with ectodermal manifestations (i.e., focal dermal hypoplasia, Goltz syndrome) or with additional skeletal lesions. Bone scanning is unremarkable.49

FIGURE 66-7. Osteopathia striata. Typical longitudinal striations are present in the metaphyseal regions of the femur, tibia, and fibula. (From Whyte MP, Murphy WA. Osteopathia striata associated with familial dermopathy and white forelock: evidence for postnatal development of osteopathia striata. Am J Med Genet 1980; 5:227.)

MELORHEOSTOSIS
Melorheostosis, a sporadic, asymmetric hyperostosis that usually affects the long bones, can cause pain and intermittent swelling of joints and limit mobility. Deformity and limb-length discrepancy may result from shortening of muscles, ligaments, and tendons.51 Often, cutaneous changes are present over subsequently involved skeletal areas. No sex preference is seen; symptoms usually begin before 20 years of age. The disorder may progress, but it is commonly present in only one limb, generally a lower extremity, of which one or several bones are affected. Radiographic examination reveals a dripping-candle-wax appearance to the cortex (Fig. 66-8). Bone scanning shows increased radionuclide uptake in involved areas.49 No effective medical treatment exists. Some patients require surgery to relieve symptoms from contractions or neurovascular entrapment.

FIGURE 66-8. Melorheostosis. Characteristic patchy cortical sclerosis, which gives a dripping-candle-wax appearance to the cortex (arrows), is present in most of the proximal tibia.

MIXED SCLEROSING BONE DYSTROPHY

Rarely, a patient may have various combinations of osteopathia striata, melorheostosis, osteopoikilosis, cranial sclerosis, diaphyseal dysplasia, or other defects in bone formation.52 Individuals with melorheostotic features or cranial sclerosis can have symptoms from compression of neurovascular tissue.53 This is a sporadic condition.

No known medical treatment exists.

FIBRODYSPLASIA (MYOSITIS) OSSIFICANS PROGRESSIVA

Fibrodysplasia (myositis) ossificans progressiva (FOP) features accumulation of true bone in the fibrous connective tissue of striated muscle, ligaments, tendons, and fascia; however, smooth muscle is spared.2,4,5 and 6 Patients have congenital anomalies of the digits (75% incidence of bilateral microdactyly of the great toe with synostosis and hypoplasia of the phalanges), allowing a presumptive diagnosis at birth.5 The pathogenesis involves endochondral bone formation in soft tissues.54 Most cases appear to represent new mutations for this autosomal dominant disorder (advanced paternal age seems to be a causal factor). The severity of FOP can be quite variable.55 Ectopic bone formation generally begins during the first decade of life as a painful, warm swelling (commonly in the sternocleidomastoid muscle) that regresses over 1 to 2 months, leaving a bony residuum. Trauma can provoke the appearance of lesions. Episodes occur with considerable irregularity, which makes assessment of attempted therapies or prognosis difficult. The shoulder girdle, upper arms, spine, and pelvis may be affectedgenerally in that sequence. Accumulation of ectopic bone eventually limits range of motion, especially in the neck and shoulders; thoracic deformity, due to involvement of the intercostal and paravertebral muscles, causes constriction of the chest. Restrictive pulmonary disease and recurrent pneumonia can follow. Patients may die from adolescence onward from restrictive lung disease, infection, or inanition owing to masseter muscle disease, although a long life is possible.4 Routine biochemical studies are generally normal. Radiographic features include ossification of an entire muscle or the appearance of a plate of ectopic bone that outlines a fascial plane.5 Metaphyseal exostoses may be present (Fig. 66-9). In the spine, fusion of the apophyseal joints is followed by fusion of the vertebrae. Radiographically, the differential diagnosis includes calcinosis universalis, dermatomyositis, tumoral calcinosis, and other disorders of mineral homeostasis that produce ectopic calcification. Histologic examination reveals true bone (i.e., mature bone with trabeculae and haversian systems, sometimes including marrow).

FIGURE 66-9. Fibrodysplasia ossificans progressiva. Typical struts of ectopic bone (white arrows) are seen in the periarticular soft tissues of the shoulder of this 17-year-old boy. Exostoses (black arrow) are common.

No medical therapy is established for FOP.56 Administration of disodium etidronate, a bisphosphonate that in large doses inhibits bone formation, produces equivocal results. Warfarin treatment to disrupt g-carboxylation of skeletal calcium-binding protein reportedly decreases urinary g-carboxyglutamic acid levels but does not limit the ectopic calcification.57 Treatment of FOP is supportive. Physical therapy to maintain joint mobility may exacerbate the condition or provoke new lesions. Trauma to muscle (e.g., surgery, intramuscular injections) should be avoided whenever possible. Sometimes, surgical release of the jaw is successful and may be essential to improve nutrition. Dental procedures should be completed early in life. FOP lesions are a form of ectopic endochondral bone formation.54 The temporal-spatial appearance of the lesions is reminiscent of the distribution of the decapentaplegia mutation in Drosophila species. In FOP, a disturbance in circulating levels of bone morphogenetic protein (BMP) has been reported.58 One patient has been shown to have a defect in the gene that encodes noggin, a BMP inhibitor.58a

FIBROUS DYSPLASIA
Fibrous dysplasia, a sporadic unifocal or multifocal developmental skeletal disorder, affects either sex and often causes fractures and deformities.59 The pathogenetic abnormality is an expanding fibrous lesion of bone-forming mesenchyme.4 Monostotic disease typically presents during the second or third decade of life; polyostotic disease manifests before 10 years of age.60 Long bones or the skull are usually affected. Before adolescence, expansile skeletal lesions may fracture, cause deformity, and sometimes entrap nerves. Affected bones occasionally become sarcomatous (incidence of <1%), especially with involvement of the facial bones or the femur.61 Pregnancy may reactivate previously quiescent lesions. Some patients have characteristic hyperpigmented skin macules called caf-au-lait spots (Fig. 66-10) and endocrine hyperfunction (i.e., McCune-Albright syndrome).59 The endocrinopathy is usually pseudoprecocious puberty in girls and less commonly thyrotoxicosis, Cushing syndrome, acromegaly, hyperprolactinemia, or hyperparathyroidism.62 Rarely, patients with McCune-Albright syndrome also have renal phosphate wasting and hypophosphatemic bone disease that may represent a form of oncogenic rickets or osteomalacia63 (see Chap. 63, Chap. 67, Chap. 70 and chap. 219).

FIGURE 66-10. A, Fibrous dysplasia. Caf-au-lait spots are typically ipsilateral to and near the skeletal lesions. These caf-au-lait macules have irregular (coast of Maine) margins unlike in neurofibromatosis, in which they have smooth (coast of California) borders. B, This 13-year-old boy, however, has bilateral skeletal disease that resulted in characteristic deformity of the face on the contralateral side.

Although serum alkaline phosphatase activity can be increased in fibrous dysplasia, blood calcium and phosphate levels are usually normal. Monostotic disease is more common than polyostotic involvement, but any bone can be affected.5 The femur, tibia, ribs, and facial bones are most frequently involved. Small bones are affected in 50% of the patients with polyostotic disease. Lesions in the long bones occur in the metaphyses and diaphyses. They are generally discrete, sometimes lobulated or trabeculated, areas of radiolucency characteristically with thin cortices and a ground-glass appearance (Fig. 66-11).

FIGURE 66-11. Fibrous dysplasia. Classic features in long bones include an expansile lesion in the diaphysis associated with thin but intact cortices (arrows) and an overall ground-glass appearancechanges present in the tibia of this 16-year-old boy with polyostotic disease who also developed hypophosphatemic osteomalacia.

Monostotic and polyostotic lesions have similar histologic findings. They are well defined but not encapsulated. Characteristically, spindle-shaped fibroblasts form a swirl in marrow spaces with haphazardly arranged trabeculae consisting of immature woven bone (Fig. 66-12). Cartilage tissue is present especially in polyostotic disease. Cystic areas, which are lined by multinucleated giant cells, may also be found. The lesion is reminiscent of the changes seen in hyperparathyroidism, but unlike in osteitis fibrosa cystica, mature osteoblasts are absent.

FIGURE 66-12. Fibrous dysplasia. Characteristic histopathologic findings include whorls of abundant fibrous tissue (FT) in the medullary space, trabecular bone (TB) composed of woven collagen, and, as shown in the inset, osteoclastosis (arrows) yet absence of osteoblasts.

The cause of McCune-Albright syndrome and fibrous dysplasia is somatic mosaicism for activating mutations in the gene encoding the Gsa subunit of the G protein that couples hormone receptors to adenylate cyclase.64 Endocrine abnormalities in McCune-Albright syndrome are due to end-organ hyperactivity rather than premature hypothalamic maturation with secretion of releasing hormones.62 No established medical treatment exists for the skeletal disease; however, promising responses have been reported with intravenous administration of pamidronate.65 Spontaneous improvement does not occur. Skeletal lesions may progress, and new ones may appear; however, in most patients with mild disease, lesions are quiescent. Girls with precocious puberty have responded favorably to the aromatase inhibitor testolactone.66 Although painful stress or fissure fractures may be difficult to detect, generally fractures heal well. The potential for development of endocrinopathies or skeletal sarcomas should be remembered. Calcitriol and inorganic phosphate supplementation may be helpful for associated hypophosphatemic bone disease.

OSTEOGENESIS IMPERFECTA
Osteogenesis imperfecta (OI; brittle-bone disease) is characterized by osteopenia with recurrent fractures and deformity, dental disease, and premature hearing loss67 (see Chap. 70 and Chap. 189). Clinical severity is extremely variable, ranging from still-birth to deafness or fracture late in adult life. The pathogenesis involves quantitative and often qualitative abnormalities in the biosynthesis of type I collagen, the most abundant protein in bone. Connective tissue in ligaments, skin, sclerae, and dentin is also affected, because type I collagen is normally present in these structures as well. Classification into several clinical types has been useful, but the understanding of OI has been greatly improved by the identification of > 250 distinct mutations in the genes that encode the a1 and a2 procollagen chains which inter-twine to form the type I collagen heterotrimer2,3,68 (see Chap. 189). Patients with OI may also be troubled by ligamentous laxity causing joint hypermobility, excessive diaphoresis, and bruising. They are predisposed to pulmonary infections if they have severe thoracic deformity from vertebral collapses and kyphoscoliosis. Mitral valve clicks are common, but significant cardiac disease is unusual. Hearing loss affects approximately half of patients younger than 30 years of age and nearly all who are older. Deafness may be conductive, sensorineural, or of mixed pathogenesis. The differential diagnosis of multiple fractures in the pediatric population includes the battered baby syndrome and congenital indifference to pain. Generalized osteopenia in children may be due to idiopathic juvenile osteoporosis, Cushing disease, or many other disorders.4 The classification scheme of Sillence, which emphasizes the apparent mode of inheritance and severity of the skeletal manifestations of OI, is referenced frequently68 (see Chap. 189); however, essentially all forms of OI are inherited as autosomal dominant disorders.68 Type I OI is associated with blue sclerae (especially obvious in childhood), osteopenia with recurrent fracture but often normal stature, and deafness during early adulthood (30% of cases). Dentinogenesis imperfecta may be absent or present in different kindreds (types IA and IB, respectively). Joint laxity or hernias can also occur. Affected women, who may first present with fractures late in adult life, can be mistakenly thought to have postmenopausal or involutional osteoporosis, and radiographic findings may not distinguish the disorders. Approximately one-third of OI patients represent new mutations.68 Histomorphometric assessment of undecalcified specimens of bone reveal an increased number of cortical osteocytes in some patients.68a Patients with Sillence types II, III, or IV OI are affected more severely. Phenotypic features include a triangle-shaped face, high-pitched voice, deformities of the long bones of the extremities (yet normal-appearing hands and feet), pectus excavatum or carinatum, joint hyperextensibility, excessive sweating, and hernias. Even patients with the most severe skeletal manifestations are generally of normal intelligence.67,68 Routine biochemical studies are usually unremarkable; elevated serum alkaline phosphatase activity and urine hydroxy-proline levels occur in some patients. Hypercalciuria is common in severely affected children.69 This finding may be explained by poor skeletal growth despite effective absorption of dietary calcium. Characteristic radiographic changes occur in severely affected patients.5 Generalized osteopenia and deformity of long bones from recurrent fractures are present, and vertebrae may be collapsed. Periosteal bone formation is defective, which retards the circumferential growth of bone and causes a gracile appearance in ribs and long bones in the limbs (Fig. 66-13). In some severely affected infants with an especially poor prognosis, severe micromelia (major long bones appear short but thick in their external diameter) and marked bowing deformities are present. Wormian bones may be noted (Fig. 66-14), and the pelvis can be triradiate shaped. Radiographic abnormalities may change considerably during growth.5 Popcorn calcifications are rare developmental lesions in the epiphyses and metaphyses of major long bones (predominantly at the knees and ankles) that may be due to traumatic fragmentation of growth plate cartilage. They are observed during childhood, appear to resolve at adolescence, and may severely limit skeletal growth. Understandably, patients are predisposed to osteoarthritis. Fractures are often transverse, yet they heal at normal ratessometimes with exuberant callus formation if motion persists at the fracture site. Platybasia can also occur, occasionally with basilar impression.67,68

FIGURE 66-13. Osteogenesis imperfecta. In the more severely affected patients (here a 4.5-year-old boy), long bones are characteristically gracile (narrow diameter) with thin cortices (arrows); they fracture easily, resulting in bowing deformity.

FIGURE 66-14. Osteogenesis imperfecta. Wormian bones (arrows), although not pathognomonic of this disorder, are often found in the lambdoidal suture of the posterior occiput.

Undecalcified specimens of bone may reveal abnormal skeletal matrix, especially in more severely affected patients. Disorganized, woven bone can be abundant. Polarized-light microscopy may demonstrate thin collagen bundles in areas of lamellar bone. The number of osteocytes in cortical bone can be increased, a finding that perhaps reflects decreased activity of individual osteoblasts, although the overall rate of bone turnover is rapid. Histomorphometric studies after in vivo tetracycline labeling support this hypothesis.4,68a The perinatal form of OI (type II) can result from large rear-rangements or even point mutations within the genes encoding the pro-a1 or pro-a2 chains of type I procollagen.2,3,68 The nascent collagen molecule undergoes excessive posttranslational modification that, in turn, results in diminished secretion of unstable bone matrix (see Chap. 189). In the milder forms (e.g., type I), the amount of type I procollagen synthesized is decreased, but a variety of secondary quantitative defects in noncollagenous bone proteins occur as well, reflecting a more profound defect in the development of hard tissues. Although remarkable progress has been made in our understanding of the heterogeneous genetic defects in OI, no established medical therapy exists.68 Intravenous infusions of pamidronate seem promising in uncontrolled trials.70,70a Treatment is supportive and often requires expert dental care with orthopedic and rehabilitation therapy for the recurrent fractures and limb deformities, kyphosis, and scoliosis. Support groups (e.g., Osteogenesis Imperfecta Foundation, Inc.) are an important source of comfort and current information for patients and their families. Genetic counseling should be offered and periodically updated, because progress in this area has been important.68 Although very rarely some cases represent an autosomal recessive condition, gonadal mosaicism can account for new-onset OI in affected siblings. Prenatal diagnosis by a variety of techniques, including ultrasonography in severe cases and molecular testing, has been successful.71

HYPOPHOSPHATASIA
Hypophosphatasia is an inborn error of metabolism characterized clinically by premature tooth loss and rickets or osteomalacia, and biochemically by subnormal serum alkaline phosphatase activity.72 Delineation of this condition has shown that alkaline phosphatase functions importantly as an ectoenzyme in skeletal mineralization. Four clinical forms, classified according to the patient's age when skeletal disease is first noted (perinatal, infantile, childhood, adult), are generally reported. Actually, the severity of hypophosphatasia is a continuum that ranges from stillbirth to hypophosphatasemia in asymptomatic adults with unremarkable radiologic and histopathologic studies of bone.73,74 When hypophosphatasia manifests at birth, it may be confused with a severe form of OI because of profound hypomineralization of the skeleton, or with cleidocranial dysplasia because of wide cranial sutures.5 If the onset is within the first 6 months of life (infantile form), only hypotonia may be noted perinatally, but subsequently failure to thrive may occur as patients develop clinically apparent rickets and hypercalcemia, recurrent pneumonias, and sometimes seizures or periodic apnea. Onset later in infancy or during childhood may cause short stature, rachitic deformity, early loss of deciduous teeth, and premature cranial synostosis.72,73 In the adult form, a history of dental disease often precedes painful recurrent stress fractures in the feet and pseudofractures (Looser zones) in the proximal femora.74 The severe forms are inherited as autosomal recessive traits.72 The adult form can be transmitted as an autosomal dominant trait, but with variable penetrance.74 Odontohypophosphatasia refers to the disorder in patients (primarily children) who have only dental manifestations. More than 58 different mutations in the gene that encodes the isoenzyme form of alkaline phosphatase found in bone have been detected in individuals with hypophosphatasia.75 The diagnosis requires a finding of subnormal serum alka-line phosphatase activity for the patient's age. (Note: Several clinical situations cause hypophosphatasemia, e.g., glucocorticoid therapy, starvation, vitamin D intoxication, scurvy, hypothyroidism, milk-alkali syndrome, magnesium deficiency, massive transfusion, severe anemia, or improper collection of blood.) Sometimes, clinical laboratories neglect to report a lower limit of serum alkaline phosphatase activity or fail to provide age-appropriate normal ranges. Increased levels of plasma pyridoxal 5'-phosphate or urinary phosphoethanolamine (natural substrates for alkaline phosphatase) give supportive biochemical evidence.76 Inorganic pyrophosphate (an endogenous inhibitor of bone mineralization) is increased in blood and urine, but the assay is a research procedure. Radiographic findings in children often provide more information than merely the suggestion of some form of rickets. Along with rachitic changes, characteristic focal tongues of metaphyseal radiolucency are seen (Fig. 66-15). Adults can be osteopenic and have unhealed stress fractures of the feet or elsewhere. Looser zones may be present in the femora and, although unexplained, are often situated laterally rather than medially as in other forms of osteomalacia. Chondrocalcinosis and additional changes consistent with pyrophosphate arthropathy result from articular accumulation of calcium pyrophosphate.77 As in all forms of rickets and osteomalacia, a generalized defect in skeletal mineralization is present in hypophosphatasia. However, no pathognomonic histopathologic changes in bone are seen. Histochemical studies demonstrate deficient alkaline phosphatase activity.73 Evidence of secondary hyper-parathyroidism is generally absent. Conversely, histologic examination of teeth can be an excellent aid for diagnosis and characteristically shows absence or hypoplasia of cementum.

FIGURE 66-15. Hypophosphatasia. Projections (tongues) of radiolucency from a widened growth plate into the metaphysis (arrows) are a characteristic feature of the rickets of hypophosphatasia, demonstrated here in the forearm of a 2-year-old boy.

No established medical therapy is available for hypophos-phatasia. Hypercalcemia in infants may be improved by dietary calcium restriction (glucocorticoid or calcitonin treatment has also been used). Nevertheless, progressive skeletal demineralization often follows. Calcium supplementation or administration of vitamin D or one of its bioactive metabolites may cause hypercalcemia and hypercalciuria. Inorganic phosphate supplementation to enhance urinary excretion of inorganic pyrophosphate and enzyme-replacement therapy using intravenous infusion of alkaline phosphatases have been disappointing.72 Infusion of pooled normal plasma was followed by remarkable transient skeletal remineralization in a boy with the infantile form.78 Bone marrow transplantation seemed beneficial for another infant.79 A trial of vitamin B6 should be given for unexplained seizures.76 Optimal orthopedic management of femoral fractures and pseudofractures in adults may require insertion of intramedullary rods.80 Craniotomy may be necessary to correct premature synostosis. Prenatal diagnosis of the severe perinatal (lethal) form is possible with ultrasonography and by assay of alkaline phosphatase activity in chorionic villus samples.81 Molecular diagnosis is becoming feasible.

AXIAL OSTEOMALACIA
Axial osteomalacia features defective bone mineralization despite normal serum levels of calcium and phosphate and increased alkaline phosphatase activity. Men seem to be affected more frequently than women; in one study,82 this disorder was found to be familial. The condition may be discovered incidentally; occasionally, patients present with a vague, chronic axial skeletal pain. Radiographic studies reveal osteo-sclerosis of the axial skeleton in which the trabecular pattern is coarse, as in other forms of osteomalacia.5 The appendicular skeleton appears unaffected, and Looser zones have not been reported. Features of ankylosing spondylitis may be present. Histopathologic studies of bone reveal osteoidosis and failure of tetracycline deposition but no evidence of secondary hyper-parathyroidism. Osteoblasts are inactive-appearing (flat) lining cells.82 Because the disorder seems to reflect a defect in the bone tissue, use of vitamin D and mineral supplementation is not recommended. No effective medical therapy exists.

FIBROGENESIS IMPERFECTA OSSIUM

Fibrogenesis imperfecta ossium is a sporadic, idiopathic disorder in which pathologic changes seem to affect collagen within the skeleton only. Symptoms begin in adult life. There is progressive pain, recurrent fractures, and then weakness, muscle atrophy, and contractures. Patients become immobilized and debilitated. Marked bone tenderness is elicited by palpation. Serum calcium and phosphate levels are unremarkable, although alkaline phosphatase activity may be increased. Radiographic abnormalities occur throughout the skeleton, except the skull.5 Although generalized osteopenia (most marked in the axial skeleton and periarticular regions) with a paucity of spongy bone is present, trabeculae appear coarse and dense. An abnormal trabecular pattern replaces cortical bone. Deformity secondary to fracture may be present, but bony contours are usually normal. Vertebral end plates can appear sclerotic and mimic a rugger-jersey spine. The radiographic differential diagnosis includes advanced osteitis fibrosa cystica, Paget bone disease, and axial osteomalacia.5 Histopathologic studies are important and show that the birefringent pattern normally formed by collagen fibrils is absent in lamellar bone. Electron microscopy reveals a random, tangled pattern of thin collagen fibrils.83 Increased numbers of osteoid seams do not take up tetracycline labels. The pathogenesis seems to involve a defect in bone matrix rather than in mineral metabolism. No effective medical therapy is known.

METAPHYSEAL AND SPONDYLOMETAPHYSEAL DYSPLASIAS

Not all disorders in children that result in widened growth plates and metaphyseal irregularity are forms of rickets (i.e., generalized disturbance of mineralization of skeletal matrix).5 Metaphy-seal or spondylometaphyseal dysplasias may mimic rickets radiographically5; however, the biochemical aberrations associated with rachitic disorders are absent.84 Serum levels of calcium, phosphate, and alkaline phosphatase activity are normal, and biopsy of skeletal tissue away from the metaphyses (after in vivo tetracycline labeling) confirms that bone remodeling is unremarkable. Instead, these disorders involve dysplastic processes occurring in growth plates. They are heritable with various patterns of genetic transmission.1,4,7,84 Mutations in the genes that encode types II, IX, and X collagen have been discovered. The most common metaphyseal dysplasia, the Schmid type, is a form of short-limb dwarfism. Patients present with short stature and occasionally with rachitic-like deformities of the legs. They grow steadily, but height is generally below the fifth percentile. An especially severe (Jensen) type, which is associated with hypercalcemia and marked short stature, is caused by activating mutations in the parathyroid hormone/parathyroid hormonerelated protein receptor.85 Patients with spondylo-metaphyseal dysplasia have vertebral abnormalities as well as metaphyseal defects. Accordingly, thoracic deformity often occurs in these individuals. The medical history, physical findings, and unremarkable screening biochemical studies are usually sufficient to exclude rickets. Experienced radiologists generally provide a correct diagnosis.5 In milder or more unclear cases, histologic examination of nondecalcified bone sections (obtained after in vivo tetracycline administration) helps to exclude rickets or osteomalacia. Orthopedic management may include bracing of bowing deformities. Occasionally, osteotomy is necessary to prevent osteoarthritis or premature closure of growth plates.84 When the spine is involved, instability may occur from dysplastic changes of the vertebrae. If such abnormalities affect the cervical region, odontoid hypoplasia can be an important problem that may require neurosurgical intervention.

IDIOPATHIC JUVENILE OSTEOPOROSIS

Idiopathic juvenile osteoporosis is a poorly characterized, heterogeneous condition that almost certainly has a variety of causes.4 A few case reports and small clinical series indicate that the disorder typically presents during late childhood with fractureespecially in the lower limbs and with vertebral collapsebut afterwards improves spontaneously.86 Distinction from OI requires a careful physical examination and family history. Idiopathic juvenile osteoporosis is generally a diagnosis of exclusion; secondary causes of osteopenia in children must always be considered (see Chap. 70).4 No established therapy exists, but dietary deficiencies (e.g., suboptimal calcium intake) should be corrected, and patients should be cautioned against stooping to lift heavy objects or participating in potentially traumatic physical activity. Bisphosphonate therapy is being evaluated.

IDIOPATHIC OSTEOLYSIS
Unifocal or multifocal osteolytic lesions manifest in a variety of sporadic idiopathic or hereditary disorders.87 Acro-osteolysis occurs with polyvinyl chloride poisoning, hand trauma (e.g., use of vibrating instruments), systemic lupus erythematosus, and other conditions.87,88 Carpotarsal osteolysis (Fig. 66-16) is characterized by episodes of pain and swelling in the wrists and ankles during childhood that accompany osteolysis of predominantly the carpal and tarsal bones.89 The disorder occurs as an isolated problem when transmitted as an autosomal dominant trait; nephropathy occurs in some sporadic or autosomal recessive cases. In Gorham massive osteolysis, unifocal areas of osteolysis can cause considerable bone destruction.90

FIGURE 66-16. Carpotarsal osteolysis. Classic radiographic features include gradual disappearance of carpal and tarsal bones during childhood, leaving just a residual tapered (sucked candy) appearance to the metacarpals. In this 40-year-old man, the radiographic changes of the disease are early in the right hand, yet advanced in the left hand. (L, left; R, right.) (From Whyte MP, Murphy WA, Kleerekoper M, et al. Idiopathic multicentric osteolysis: report of an affected father and son. Arthritis Rheum 1978; 21:367.)

ECTOPIC CALCIFICATION
Many conditions are associated with extraskeletal deposition of calcium and phosphate.4 In some, the pathogenesis is an elevation in the circulating calcium-phosphate product (i.e., meta-static calcification). 91 In others, the precise mechanism is unknown, and either amorphous calcium (dystrophic calcification) or true bone is deposited.91 Metastatic calcification occurs with hypercalcemia of any cause (e.g., milk-alkali syndrome, hypervitaminosis D, tumoral calcinosis, sarcoidosis, hyperpara-thyroidism, and renal osteodystrophy). Typically, deposits occur in the kidneys, lungs, media of large arteries, fundus of the stomach, and periarticular regions. Hyperphosphatemia is also associated with ectopic calcification (e.g., hypoparathyroidism and pseudohypoparathyroidism), renal insufficiency, and massive cell destruction with liberation of inorganic phosphate (e.g., chemotherapy for leukemia). Dystrophic calcification, in which the calcium-phosphate product is normal, occurs in necrotic tissue in the calcinosis universalis and calcinosis circumscripta that accompany scleroderma, dermatomyositis, systemic lupus erythematosus, trauma, and metastatic malignancy, and as a primary cutaneous disease.4 Fasciitis is associated with ectopic bone formation (e.g., myositis ossificans progressiva), although other soft tissues can ossify, as in neurologic injury, burns, and trauma. Progressive osseous heteroplasia, an autosomal dominant disorder featuring extraskeletal ossification in subcutaneous and deep tissues, is caused by mutation in the GNASI gene that encodes the GSa stimulatory protein.91a

ISCHEMIC NECROSIS
Ischemic (avascular) necrosis is common92 (see Chap. 211). The condition has multiple causes: endocrine or metabolic (e.g., glucocorticoid therapy, Cushing disease, alcohol abuse, gout, osteomalacia, hyperparathyroidism, hypothyroidism), storage diseases (e.g., Gaucher disease), hemoglobinopathies, trauma, dysbaric disorders, irradiation, and pancreatitis.93 Some cases are idiopathic, and some are familial. The most common causes are prolonged exposure to glucocorticoids and alcohol abuse. Patients present with deep aching pain in the diseased region. Often, the head of the femur is affected. Microfracture of dead trabecular bone destroys the support for overlying cortical bone and articular cartilage. Radiographic studies typically reveal sclerosis of subcortical bone.5 Nuclear magnetic resonance imaging provides the greatest sensitivity and specificity to aid in early diagnosis.94 Treatment may require core decompression or joint arthroplasty or replacement with a prosthesis. CHAPTER REFERENCES
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. Beighton P. Inherited disorders of the skeleton, 2nd ed. New York: Churchill Livingstone, 1988. Royce PM, Steinmann BU, eds. Connective tissue and its heritable disorders, 2nd ed. New York: Wiley-Liss, 2001; in press. Beighton P, ed. Heritable disorders of connective tissue, 5th ed. St. Louis: MosbyYear Book, 1993. Favus MJ, ed. Primer on the metabolic bone diseases and disorders of mineral metabolism, 4th ed. Philadelphia: Lippincott Williams & Wilkins, 1999. Resnick D. Diagnosis of bone and joint disorders, 3rd ed. Philadelphia: WB Saunders, 1995. Whyte MP. Skeletal disorders characterized by osteosclerosis or hyperostosis. In: Avioli LV, Krane SM, eds. Metabolic bone disease and clinically related disorders, 2nd ed. San Diego: Academic Press, 1997. Taybi H, Lachman RS. Radiology of syndromes, metabolic disorders, and skeletal dysplasias, 4th ed. St. Louis: Mosby, 1996. Whyte MP. Osteopetrosis. In: Royce PM, Steinmann BU, eds. Connective tissue and its heritable disorders, 2nd ed. New York: Wiley-Liss, 2001; in press. Key LL Jr, Ries WL. Osteopetrosis. In: Bilezikian JP, Raisz LG, Rodan GA, eds. Principles of bone biology. San Diego: Academic Press, 1996. Heaney C, Shalev H, Elbedour K, et al. Human autosomal recessive osteopetrosis maps to 11q13, a position predicted by comparative mapping of the murine osteosclerosis ( oc) mutation. Hum Mol Genet 1998; 7:1407. Van Hul W, Bollerslev J, Gram J, et al. Localization of a gene for autosomal dominant osteopetrosis (Albers-Schnberg disease) to chromosome 1p21. Am J Hum Genet 1997; 61:363. Kahler SG, Burns JA, Aylsworth AS. A mild autosomal recessive form of osteopetrosis. Am J Med Genet 1984; 17:451. Whyte MP. Carbonic anhydrase II deficiency. Clin Orthop 1993; 294:52.

13a. Frattini A, Orchard PJ, Sobacchi C, et al. Defects in TCIRG1 subunit of the vacuolar proton pump are responsible for a subset of human autosomal recessive osteopetrosis. Nat Genet 2000; 25:343. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. Whyte MP. Recent advances in osteopetrosis. In: Cohn DV, Gennari C, Tashjian AH Jr, eds. Calcium regulating hormones and bone metabolism. New York: Elsevier Science, 1992. Athanasou NA, Sabokbar A. Human osteoclast ontogeny and pathological bone resorption. Histol Histopathol 1999; 14:635. Marks SC Jr. Osteopetrosis: multiple pathways for the interruption of osteoclast function. Appl Pathol 1987; 5:172. Marks SC Jr. Osteoclast biology: lessons from mammalian mutations. Am J Med Genet 1989; 34:43. Schneider GB, Key LL, Popoff SN. Osteopetrosis: therapeutic strategies. Endocrinologist 1998; 8:409. Gerritsen EJ, Vossen JM, van Loo IH, et al. Autosomal recessive osteopetrosis. Pediatrics 1994; 93:247. Lehman RA, Reeves JD, Wilson, WB, Wesenberg RL. Neurological complications of infantile osteopetrosis. Ann Neurol 1977; 2:378. Bollerslev J. Autosomal dominant osteopetrosis: bone metabolism and epidemiological, clinical, and hormonal aspects. Endocr Rev 1989; 10:45. Johnston CC Jr, Lavy N, Lord T, et al. Osteopetrosis: a clinical, genetic, metabolic, and morphologic study of the dominantly inherited, benign form. Medicine (Baltimore) 1968; 47:149. Reeves JD, August CS, Humbert JR, Weston WL. Host defense in infantile osteopetrosis. Pediatrics 1979; 64:202. Eapen M, Davies SM, Ramsay NK, Orchard PJ. Hematopoietic stem cell transplantation for osteopetrosis. Bone Marrow Transplant 1998; 22:941. Locatelli F, Beluffi G, Giorgiani G, et al. Transplantation of cord blood progenitor cells can promote bone resorption in autosomal recessive osteopetrosis. Bone Marrow Transplant 1997; 20:701. Kubo T, Tanaka H, Ono H, et al. Malignant osteopetrosis treated with high doses of 1a-hydroxyvitamin D 3 and interferon gamma. J Pediatr 1993; 123:264. Whyte MP, Murphy WA, Fallon MD, et al. Osteopetrosis, renal tubular acidosis, and basal ganglia calcification in three sisters. Am J Med 1980; 69:64. Orchard PJ, Dickerman JD, Mathews CHE, et al. Haploidentical bone marrow transplantation for osteopetrosis. Am J Pediatr Hematol Oncol 1987; 9:335. Key LL Jr, Ries WL, Rodriguiz RM, Hatcher HC. Recombinant human interferon gamma therapy for osteopetrosis. J Pediatr 1992; 121:119. Ozsoylu S. High dose intravenous methylprednisolone in treatment of recessive osteopetrosis. (Letter). Arch Dis Child 1987; 62:214. Whyte MP, Hamm LL 3rd, Sly WS. Transfusion of carbonic anhydrase-replete erythrocytes fails to correct the acidification defect in the syndrome of osteopetrosis, renal tubular acidosis, and cerebral calcification (carbonic anhydrase-II deficiency). J Bone Miner Res 1988; 3:385. Ogur G, Ogur E, Celasun B, et al. Prenatal diagnosis of autosomal recessive osteopetrosis, infantile type, by X-ray evaluation. Prenat Diagn 1995; 15:477. Sen C, Madazli R, Aksoy F, Ocak V. Antenatal diagnosis of lethal osteopetrosis. Ultrasound Obstet Gynecol 1995; 5:278. Strisciuglio P, Hu PY, Lim EJ, et al. Clinical and molecular heterogeneity in carbonic anhydrase II deficiency and prenatal diagnosis in an Italian family. J Pediatr 1998; 132:717. Gelb BD, Shi GP, Chapman HA, Desnick RJ. Pycnodysostosis, a lysosomal disease caused by cathepsin K deficiency. Science 1996; 273:1236. Soliman AT, Rajab A, AlSalmi I, et al. Defective growth hormone secretion in children with pycnodysostosis and improved linear growth after growth hormone treatment. Arch Dis Child 1996; 75:242. Fallon MD, Whyte MP, Murphy WA. Progressive diaphyseal dysplasia (Engelmann's disease). J Bone Joint Surg Am 1980; 62:465.

37a. Kinoshita A, Saito T, Tomita H, et al. Domain-specific mutations in TGFB1 result in Camurati-Engelmann disease. Nat Genet 2000; 26:19. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. Hundley JD, Wilson FC. Progressive diaphyseal dysplasia. Review of the literature and report of seven cases in one family. J Bone Joint Surg Am 1973; 55:461. Smith R, Walton RJ, Corner BD, Gordon IR. Clinical and biochemical studies in Engelmann's disease (progressive diaphyseal dysplasia). QJM 1977; 46:273. Kumar B, Murphy WA, Whyte MP. Progressive diaphyseal dysplasia (Engelmann disease): scintigraphic-radiographic-clinical correlations. Radiology 1981; 140:87. Naveh Y, Alon U, Kaftori JK, Berant M. Progressive diaphyseal dysplasia: evaluation of corticosteroid therapy. Pediatrics 1985; 75:321. De Rubin ZS, Ghiringhello G, Mansur JL. Clinical, humoral and scintigraphic assessment of a bisphosphonate as potential treatment of diaphy-seal dysplasia: Ribbing and Camurati-Engelmann diseases. Medicina 1997; 57(Suppl 1):56. Saul RA, Lee WH, Stevenson RE. Caffey's disease revisited: further evidence for autosomal dominant inheritance with incomplete penetrance. Am J Dis Child 1982; 136:55. Holtzman D. Infantile cortical hyperostosis of the scapula presenting as an ipsilateral Erb's palsy. J Pediatr 1972; 81:785. Van Buchem FSP, Prick JJG, Jasper HHJ. Hyperostosis corticalis generalisata familiaris (van Buchem's disease). Amsterdam: Excerpta, 1976. Eastman JR, Bixler D. Generalized cortical hyperostosis (van Buchem disease): nosologic considerations. Radiology 1977; 125:297. Balemans W, Van Den Ende J, Paes-Alves AF, et al. Localization of the gene for sclerosteosis to the van Buchem Disease-gene region on chromosome 17q12-q21. Am J Hum Genet 1999;

64:1661. 48. Krishnamachari KA. Skeletal fluorosis in humans: a review of recent progress in the understanding of the disease. Prog Food Nutr Sci 1986; 10:279. 49. Whyte MP, Murphy WA, Siegel BA. 99mTc-pyrophosphate bone imaging in osteopoikilosis, osteopathia striata, and melorheostosis. Radiology 1978; 127:439. 50. Uitto J, Starcher BC, Santa-Cruz DJ, et al. Biochemical and ultrastructural demonstration of elastin accumulation in the skin lesions of the Buschke-Ollendorff syndrome. J Invest Dermatol 1981; 76:284. 51. Beauvais P, Faure C, Montagne JP, et al. Leri's melorheostosis: three pediatric cases and a review of the literature. Pediatr Radiol 1977; 6:153. 52. Whyte MP, Murphy WA, Fallon MD, Hahn TJ. Mixed-sclerosing-bone-dystrophy: report of a case and review of the literature. Skeletal Radiol 1981; 6:95. 53. Pacifici R, Murphy WA, Teitelbaum SL, Whyte MP. Mixed-sclerosing-bone-dystrophy: 42-year follow-up of a case reported as osteopetrosis. Calcif Tissue Int 1986; 38:175. 54. Kaplan FS, Tabas JA, Gannon FH, et al. The histopathology of fibrodysplasia ossificans progressiva: an endochondral process. J Bone Joint Surg Am 1993; 75:220. 55. Connor JM, Evans DA. Fibrodysplasia ossificans progressiva. The clinical features and natural history of 34 patients. J Bone Joint Surg Br 1982; 64:76. 56. Smith R, Russell RG, Woods CG. Myositis ossificans progressiva. Clinical features of eight patients. J Bone Joint Surg A 1976; 58:48. 57. Moore SE, Jump AA, Smiley JD. Effect of warfarin sodium therapy on excretion of 4-carboxy-L-glutamic acid in scleroderma, dermatomyositis, and myositis ossificans progressiva. Arthritis Rheum 1986; 29:344. 58. Shafritz AB, Shore EM, Gannon FH, et al. Overexpression of an osteogenic morphogen in fibrodysplasia ossificans progressiva. N Engl J Med 1996; 335:555. 58a. Lucotte G, Semonin O, Lutz P. A de novo heterozygous deletion of 42 base-pairs in the noggin gene of a fibrodysplasia ossificans progressiva patient (Letter). Clin Genet 1999; 56:469. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. Whyte MP. Hereditary metabolic and dysplastic skeletal disorders. In: Coe FL, Favus MJ, eds. Disorders of bone and mineral metabolism. New York: Raven Press 1992:977. Harris WH, Dudley HR Jr, Barry RJ. The natural history of fibrosis dysplasia. J Bone Joint Surg Am 1962; 44:207. Johnson CB, Gilbert EF, Gottlieb LI. Malignant transformation of polyostotic fibrous dysplasia. South Med J 1979; 72:353. Harris RI. Polyostotic fibrous dysplasia with acromegaly. Am J Med 1985; 78:539. Lever EG, Pettingale KW. Albright's syndrome associated with a soft-tissue myxoma and hypophosphataemic osteomalacia: report of a case and review of the literature. J Bone Joint Surg Br 1983; 65:621. Shenker A, Weinstein LS, Moran A, et al. Severe endocrine and nonendocrine manifestations of the McCune-Albright syndrome associated with activating mutations of stimulatory G-protein GS. J Pediatr 1993; 123:509. Chapurlat RD, Meunier PJ. Fibrous dysplasia of bone. Baillire's Clin Rheumatol 2000; 14:385. Feuillan PP, Foster CM, Pescovitz OH, et al. Treatment of precocious puberty in the McCune-Albright syndrome with the aromatase inhibitor testolactone. N Engl J Med 1986; 315:1115. Albright JA, Millar EA. Osteogenesis imperfecta (symposium). Clin Orthop 1981; 159:1. Byers PH. Disorders of collagen biosynthesis and structure. In: Scriver CR, Beaudet AL, Sly WS, Valle D, eds. The metabolic and molecular basis of inherited disease, 7th ed. New York: McGraw-Hill, 1995:4029.

68a. Rauch F, Travers R, Parfitt AM, Glorieux FH. Static and dynamic bone histomorphometry in children with osteogenesis imperfecta. Bone 2000; 26:581. 69. Chines A, Petersen DJ, Schranck FW, Whyte MP. Hypercalciuria in children severely affected with osteogenesis imperfecta. J Pediatr 1991; 119:51. 70. Glorieux FH, Bishop NJ, Plotkin H, et al. Cyclic administration of pamidronate in children with severe osteogenesis imperfecta. N Engl J Med 1998; 339:947. 70a. Plotkin H, Rauch F, Bishop NJ, et al. Pamidronate treatment of severe osteogenesis imperfecta in children under 3 years of age. J Clin Endocrinol Metab 2000; 85:1846. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. 91. Brons JT, van der Harten HJ, Wladimiroff JW, et al. Prenatal ultrasonographic diagnosis of osteogenesis imperfecta. Am J Obstet Gynecol 1988; 159:176. Whyte MP. Hypophosphatasia. In: Scriver CR, Beaudet AL, Sly WS, Valle D, eds. The metabolic and molecular bases of inherited disease, 7th ed. New York: McGraw-Hill, 1995:4095. Fallon MD, Teitelbaum SL, Weinstein RS, et al. Hypophosphatasia: clinico-pathologic comparison of the infantile, childhood, and adult forms. Medicine (Baltimore) 1984; 63:12. Whyte MP, Teitelbaum SL, Murphy WA, et al. Adult hypophosphatasia: clinical, laboratory, and genetic investigation of a large kindred with review of the literature. Medicine (Baltimore) 1979; 58:329. Whyte MP. Hypophosphatasia. In: Econs MJ, ed. The genetics of osteoporosis and metabolic bone disease. Totowa, NJ: Humana Press, 1999; in press. Whyte MP, Mahuren JD, Vrabel LA, Coburn SP. Markedly increased circulating pyridoxal-5'-phosphate levels in hypophosphatasia: alkaline phosphatase acts in vitamin B6 metabolism. J Clin Invest 1985; 76:752. Whyte MP, Murphy WA, Fallon MD. Adult hypophosphatasia with chondrocalcinosis and arthropathy: variable penetrance of hypophos-phatasemia in a large Oklahoma kindred. Am J Med 1982; 72:631. Whyte MP, Magill HL, Fallon MD, Herrod HG. Infantile hypophosphatasia: normalization of circulating bone alkaline phosphatase activity followed by skeletal remineralization (evidence for an intact structural gene for tissue nonspecific alkaline phosphatase). J Pediatr 1986; 108:82. Fedde KN, Blair L, Terzic F, et al. Amelioration of the skeletal disease in hypophosphatasia by bone marrow transplantation using the alkaline phosphatase-knockout mouse model. (Abstract). Am J Hum Genet 1996; 59:A. Coe JD, Murphy WA, Whyte MP. Management of femoral fractures and pseudo-fractures in hypophosphatasia. J Bone Joint Surg Am 1986; 68:981. Brock DJ, Barron L. First-trimester prenatal diagnosis of hypophosphatasia: experience with 16 cases. Prenat Diagn 1991; 11:387. Whyte MP, Fallon MD, Murphy WA, Teitelbaum SL. Axial osteomalacia: clinical, laboratory and genetic investigation of an affected mother and son. Am J Med 1981; 71:1041. Lang R, Vignery AM, Jensen PS. Fibrogenesis imperfecta ossium with early onset: observations after 20 years of illness. Bone 1986; 7:237. Patel AC, McAlister WH, Whyte MP. Spondyloepimetaphyseal dysplasia: clinical and radiologic investigation of a large kindred manifesting autosomal dominant inheritance. Medicine 1993; 72:326. Kruse K, Schutz C. Calcium metabolism in the Jansen type of metaphyseal dysplasia. Eur J Pediatr 1993; 152:912. Evans RA, Dunstan CR, Hills E. Bone metabolism in idiopathic juvenile osteoporosis: a case report. Calcif Tissue Int 1983; 35:5. Whyte MP, Eddy MC, Podgornik MN, McAlister WH. Polycystic bone disease: a new, autosomal dominant disorder. J Bone Miner Res 1999; 14:1261. Destouet JM, Murphy WA. Acquired acroosteolysis and acronecrosis. Arthritis Rheum 1983; 26:1150. Whyte MP, Murphy WA, Kleerekoper M, et al. Idiopathic multicentric osteolysis. Arthritis Rheum 1978; 21:367. Bullough PG. Massive osteolysis. N Y State J Med 1971; 71:2267. Anderson HC. Calcific diseases: a concept. Arch Pathol Lab Med 1983; 107:341.

91a. Eddy MC, Jan de Beur SM, Yandow SM, et al. Deficiency of the a-subunit of the stimulatory G protein and severe extraskeletal ossification. J Bone Miner Res 2000; 15:2074. 92. Anonymous. Symposium on idiopathic osteonecrosis. Orthop Clin North Am 1985; 16:593. 93. Chang CC, Greenspan A, Gershwin ME. Osteonecrosis: current perspectives on pathogenesis and treatment. Semin Arthritis Rheum 1993; 23:47. 94. Totty WG, Murphy WA, Ganz WI, et al. Magnetic resonance imaging of the normal and ischemic femoral head. AJR Am J Roentgenol 1984; 143:1273.

CHAPTER 67 DISEASES OF ABNORMAL PHOSPHATE METABOLISM Principles and Practice of Endocrinology and Metabolism

CHAPTER 67 DISEASES OF ABNORMAL PHOSPHATE METABOLISM

MARC K. DREZNER Regulation of Phosphate Metabolism Phosphate Homeostasis: Intestine Phosphate Homeostasis: Kidneys Differential Diagnosis of an Abnormal Serum Phosphorus Level Hyperphosphatemia Reduced Renal Phosphate Excretion Increased Phosphate Load Hypophosphatemia Disturbance of Phosphate Intake/Gastrointestinal Absorption Renal Phosphate Loss Transcellular Shift of Phosphorus Combined Mechanisms Clinical Signs and Symptoms of Abnormal Serum Phosphate Levels Hyperphosphatemia Hypophosphatemia Treatment of Abnormal Serum Phosphate Levels Chapter References

Phosphorus is one of the most abundant constituents of all tissues, and disturbances in phosphate homeostasis can affect almost any organ system. Most phosphorus within the body is in bone (600700 g), whereas much of the remainder is distributed in soft tissue (100200 g). Less than 1% is in extracellular fluids. As the major intracellular anion, phosphorus plays a critical role in many aspects of cell function. These include (a) serving as a source of the high-energy phosphate in adenosine triphosphate (ATP); (b) providing an essential element of phospholipid in cell membranes; and (c) directly influencing a variety of enzymatic reactions (e.g., glycolysis) and protein functions (e.g., the oxygen-carrying capacity of hemoglobin by regulation of 2,3-diphosphoglycerate [2,3-DPG] synthesis). The plasma contains ~14 mg/dL phosphorus, of which 8 to 9 mg is lipid phosphorus, a trace is an anion of pyrophosphoric acid, and the remainder is inorganic phosphate (Pi). The Pi is present in the circulation as monohydrogen phosphate, which is divalent, and as dihydrogen phosphate, which is monovalent. At pH 7.4, the relative concentrations of monohydrogen to dihydrogen phosphate are 4:1. Although phosphorus is present as phosphate in biologic fluids, the concentration of elemental phosphorus is measured in routine serum assays and averages 2.5 to 4.5 mg/dL in normal adults. In children, normal levels are higher. The mechanism for this difference is not established but may be related to the higher levels of circulating growth hormone in growing children and the associated increases in the tubular reabsorption of phosphate.

REGULATION OF PHOSPHATE METABOLISM

The critical role that phosphorus plays in cell physiology has resulted in the development of elaborate mechanisms for extracting phosphates from the diet and the conservation of phosphate absorbed by the intestine. Such regulation maintains the plasma and extracellular fluid phosphorus within a relatively narrow range and centers on the intestine and kidney as the major organs of regulation. PHOSPHATE HOMEOSTASIS: INTESTINE The average dietary phosphate intake in humans, derived largely from dairy products, meat, and cereals, is 1200 to 1500 mg per day, two-fold to three-fold greater than the estimated minimum requirement. Approximately 60% to 70% of phosphate in the diet is absorbed, so that net absorption is proportional to intake. Absorption probably occurs throughout the small intestine; transport is greatest in the jejunum, less in the duodenum, and minimal in the ileum.1 The movement of phosphorus from the intestinal lumen to the blood requires (1) transport across the luminal brush border membrane; (2) movement transcellularly; and (3) transport across the basolateral plasma membrane of the epithelium. Phosphate transport at the luminal brush border membrane occurs by way of a saturable process against a lumen/intracellular electrochemical gradient at lumen phosphate concentrations below 2 mmol/L; above this concentration, transport predominantly proceeds by passive diffusion. The active transport of phosphate occurs by way of two independent carrier-mediated processes, a high-affinity and a low-affinity system. Both processes are dependent on sodium and potassium; however, the high-affinity transport system is enhanced by acid pH, whereas both alkaline pH and calcium stimulate the low-affinity system. The molecular basis for the sodium-dependent phosphate transport remains unknown. Likewise, little is known about the cellular events that mediate the movement of phosphorus from the luminal to the basolateral membrane. Nevertheless, available evidence suggests a role for the microtubule system in conveying phosphorus transcellularly and in extrusion of phosphorus at the basolateral membrane. Extrusion of phosphorus at the basolateral membrane of intestinal epithelial cells is passive, occurring via an anion-exchange mechanism, because the electrochemical gradient favors such movement. However, there have been reports of an ATP- and Na+ -dependent transport process at this locus. Calcitriol [1,25(OH)2D3] is the principal hormonal factor that influences gastrointestinal phosphate absorption.2,3 Facilitation of transport occurs by way of a calcium-dependent duodenal process and a calcium-independent jejunal system. The latter activity is modulated by calcitriol-induced transcription of messenger RNA. A major cause of defective phosphate absorption is calcitriol deficiency (secondary to a nutritional deficiency or caused by malabsorption). Alternatively, genetic and acquired disorders of vitamin D metabolism may decrease the availability of vitamin D (see Chap. 54, Chap. 63 and Chap. 70). Besides calcitriol deficiency, a number of additional factors may adversely influence the intestinal absorption of phosphate. Principal among these is the formation of nonabsorbable calcium, aluminum, or magnesium phosphate salts in the intestine. Moreover, phosphate absorption is reduced by more than one-half with advancing age in laboratory animals and probably in humans. PHOSPHATE HOMEOSTASIS: KIDNEYS The kidney is immediately responsive to changes in serum levels or dietary intake of phosphate.4,5 Altering either the filtered load or the tubular reabsorption of phosphate controls renal adaptation. The concentration of phosphate in the glomerular ultrafiltrate is essentially identical to that in plasma. Thus, the product of the serum phosphorus concentration and the glomerular filtration rate (GFR) is equal to the filtered load of phosphate. A change in the GFR may influence phosphate homeostasis if uncompensated by commensurate changes in tubular reabsorption. Normally, 80% to 90% of the filtered phosphate load is reabsorbed, predominantly in the proximal portions of the nephron.6,7 and 8 Sixty to seventy percent is reabsorbed in the proximal convoluted tubule and a smaller but significant fraction is reabsorbed in the pars recta. Along the proximal convoluted tubule, the transport is heterogeneous. In the most proximal portion, the S1 segment, phosphate reabsorption exceeds that of sodium. and water, whereas, more distally, phosphate reabsorption parallels that of fluid and sodium. Ample evidence also supports the existence of one or more distal tubular reabsorptive sites for phosphorus. However, conclusive proof for tubular secretion of phosphate in humans is lacking. In the proximal convoluted tubule, transepithelial phosphate transport is essentially unidirectional and involves uptake across the brush border membrane (BBM), translocation across the cell, and efflux at the basolateral membrane. The phosphate uptake at the apical cell surface proceeds by way of an active rate-limiting step in the reabsorptive process and is mediated by Na+ -dependent Pi transporters. The transporters are located in the BBM and depend on the basolateral membrane-associated Na+ ,K+-ATPase to maintain the Na+ gradient that drives the transport process. Recently, cDNAs encoding two classes of Na+-Pi cotransporters have been identified and designated NPT1 and NPT2. Although both transporters mediate high-affinity Na+-Pi cotransport, the pH profile and documented hormonal modulation of NPT2 indicate that the NPT2 transporter is the primary functional protein that regulates phosphate transport.9 In contrast to this regulated process, the exit of phosphate across the contraluminal membrane into the peritubular fluid is passive, going down a concentration gradient. The monovalent ion (H2 PO4) is preferentially transported, and reabsorptive capacity is saturated at a physiologic tubular fluid phosphate concentration of 2 mmol/L. Little is known regarding the

mechanisms of transport in the distal nephron. A transport maximum (TmP) or threshold that, when exceeded, results in quantitative excretion of filtered phosphate characterizes renal phosphate reabsorption in humans. In the fasting state, the phosphate concentration in the renal ultrafiltrate is generally less than the TmP; however, some of the filtered phosphate is excreted in the urine secondary to nephron heterogeneity with regard to GFR and reabsorption. The TmP can be determined in patients by phosphate infusions, but this method is cumbersome and nonphysiologic. Alternatively, a simplified method to accurately estimate the TmP has been developed.4 Using a nomogram, TmP is determined by measurements of serum and urine phosphorus (PO4) and creatinine (Cr) and calculation of the tubular reabsorption of phosphate (TRP):

Several factors modulate the renal TmP (Table 67-1). The best known and most important regulator is parathyroid hormone (PTH). PTH administration decreases renal phosphate reabsorption, causing major increases of urinary phosphate.10 Conversely, parathyroidectomy increases renal phosphate reabsorption and decreases phosphate excretion. The PTH-dependent changes result primarily from alteration of ion transport in the proximal convoluted tubule. Calcium-mediated, 1,25(OH)2D3-mediated, calcitonin-mediated, and estrogen-mediated changes in renal TmP have also been reported (see Chap. 206). In addition, dietary Pi intake is a major regulator of renal Pi handling. Pi deprivation elicits an increase, and Pi loading a decrease, in TmP/GFR, changes that are caused by an adaptive increase or decrease in BBM Na+ -Pi cotransport. In any case, a change in phosphate reabsorption and the TmP generally yields a comparable change in the serum phosphorus and eventually in the total phosphorus balance.

TABLE 67-1. Control of Renal Phosphate Reabsorption

DIFFERENTIAL DIAGNOSIS OF AN ABNORMAL SERUM PHOSPHORUS LEVEL

Frequently, a careful history and physical examination reveal the apparent cause of an abnormal serum phosphorus concentration. However, the variety of diseases, therapeutic agents, and physiologic states that affect phosphate homeostasis are numerous, and clinicians should make every effort to determine the mechanism underlying changes in the serum phosphorus level in each patient. Determining the cause for the abnormality permits a rational choice of appropriate therapy. HYPERPHOSPHATEMIA Two mechanisms generally underlie the genesis of hyperphosphatemia in humans. First and most commonly, reduced renal excretion of phosphate may cause phosphate retention. The reduction reflects either a reduced GFR or increased renal tubular reabsorption of phosphate (see Table 67-1). Second, hyperphosphatemia rarely ensues after an increase in phosphate intake or an acute increment of endogenous phosphate released into the extracellular fluid. REDUCED RENAL PHOSPHATE EXCRETION Renal Failure. The most common cause of hyperphosphatemia is chronic renal failure. The increased serum phosphate concentration results from a decline of the GFR.11 However, in early stages of renal failure, compensatory changes in the renal tubular reabsorption of phosphate, in part because of elevated circulating levels of PTH, limit expression of this abnormality. Indeed, only at a GFR of ~25 mL/min/l.73 m2 or less is hyperphosphatemia commonly seen. With advanced renal failure, however, serum phosphorus levels may increase up to a level of 11 or 12 mg/dL if dietary phosphate is high and phosphorus is freely mobilized from the skeleton by severe secondary hyperparathyroidism. Retention of phosphate in this disorder leads to the development of renal osteodystrophy, ectopic calcification (see later), depression of renal calcitriol synthesis, and impaired intestinal absorption of calcium. Hyperphosphatemia also lowers the serum ionized calcium and stimulates hypertrophy of the parathyroid glands and secretion of PTH. Hypoparathyroidism. Hyperphosphatemia characteristically complicates the hypoparathyroid states (idiopathic or postsurgical, or pseudohypoparathyroidism).12 The increased serum phosphate levels are secondary to enhanced renal reabsorption of phosphate. In idiopathic and postsurgical hypoparathyroidism, deficient PTH causes a loss of the hormonal effect to decrease the TmP. Conversely, the increased renal reabsorption of phosphate in the pseudohypoparathyroid disorders arises from renal resistance to PTH. The concurrence of clinically significant hypocalcemia and hyperphosphatemia usually leads to the diagnosis of hypoparathyroidism. Some studies have revealed that the causes of idiopathic hypoparathyroidism in some affected patients include (a) gain-of-function calcium-receptor mutations that result in autosomal-dominant and sporadic disease13; (b) an autoimmune polyglandular syndrome caused by a genetic abnormality at 21q22.314; (c) an X-linked recessive inherited form of the disorder15; and (d) autoimmune-targeting of the calcium-sensing receptor.13 In contrast, investigations of the familial parathyroid hormone-resistance syndromes indicate that (a) pseudohypoparathyroidism type 1a occurs because of paternally imprinted mutations of GNAS1 on chromosome 20q13.11, resulting in abnormalities of thea subunit of the G protein, which couples receptor binding to activation of adenylate cyclase16; and (b) pseudohypoparathyroidism type 1b similarly manifests secondarily to a paternally imprinted mutation of the stimulatory G protein that maps to 20q13.3.17 In all of these disorders, treatment of the hypocalcemia with oral calcium supplements and vitamin D preparations may reduce but not necessarily normalize the serum phosphate levels and the renal TmP. This effect may be the result of the elevation of the serum calcium concentration, or perhaps it is the result of a direct effect of 1,25(OH)2D3 on renal tubular phosphate handling. Thus, patients with hypoparathyroidism who are successfully treated may have normal calcium and phosphate levels. Tumoral Calcinosis. Tumoral calcinosis is a rare genetic disease characterized by periarticular cystic and solid tumorous calcifications (Fig. 67-1). Specific biochemical markers of the disorder include hyperphosphatemia and an elevated serum 1,25(OH)2D3 concentration. Using these criteria, evidence has been presented for autosomal-recessive inheritance of this syndrome. However, an abnormality of dentition, marked by short bulbous roots, pulp stones, and radicular dentin deposited in swirls, is a phenotypic marker of the disease that is variably expressed. Thus, this disorder may have multiple formes frustes that could complicate genetic analysis. Indeed, using the dental lesion as well as the more classic biochemical and clinical hallmarks of the disease, an autosomal-dominant pattern of transmission for tumoral calcinosis has been documented.18

FIGURE 67-1. Tumoral calcinosis. Radiograph of right femur and pelvis of a 9-year-old boy with hyperphosphatemic tumoral calcinosis. This tumor was his fourth, and it had to be removed surgically. Serum phosphate level was 7.92 mg/dL (normal, 4.56.5 mg/dL); 1,25 dihydroxyvitamin D was 80 pg/mL (normal, 1950 pg/mL).

An increase in capacity of renal tubular phosphate reabsorption causes the hyperphosphatemia that is typical in affected subjects. Hypocalcemia is not a consequence of this abnormality, however, and the serum PTH concentration is normal. Moreover, the phosphaturic and urinary cyclic adenosine monophosphate (cAMP) responses to PTH are not disturbed. Thus, the defect does not represent renal insensitivity to PTH, or hypoparathyroidism. Rather, the basis of the disease is probably a primary abnormality of the renal tubule that enhances phosphate reabsorption. Undoubtedly, the calcific tumors result from the elevated calcium-phosphorus product. The observation that long-term phosphorus depletion alone or in association with administration of acetazolamide, a phosphaturic agent, leads to resolution of the tumor masses supports this possibility.19,20 An acquired form of this disease is rarely seen in patients with end-stage renal failure. Affected patients manifest hyperphosphatemia in association with either (a) an inappropriately elevated calcitriol level for the degree of renal failure, hyperparathyroidism, or hyperphosphatemia21,22; or (b) long-term treatment with calcium carbonate, calcitriol, or high calcium-content dialysates. Calcific tumors again likely result from an elevated calciumphosphorus product. Indeed, complete remission of the tumors occurs on treatment with low-calcium dialysate23 or with vinpocetine, a mineral scavenger drug.24 Hyperthyroidism. Up to one-third of patients with thyrotoxicosis may have hyperphosphatemia. The elevation of the serum phosphorus is caused by increased bone resorption and calcium mobilization, consequent PTH suppression, and an increase in renal tubular reabsorption of phosphate. The frequent concurrence of an elevated serum osteocalcin and alka-line phosphatase activity, as well as increased urinary calcium and hydroxyproline, indicates that thyroid hormone effects on bone cause the hyperphosphatemia.25,26 The increased TmP has been confirmed by direct measurements and by reports of elevated tubular reabsorption of phosphate in hyperthyroidism that reverts to normal after successful treatment. Secondary hypoparathyroidism, which is a response to thyroid hormone associated hypercalcemia (see Chap. 42), is probably not the only explanation for the abnormal phosphate transport. Several studies suggest that thyroxine and triiodothyronine directly influence renal tubular phosphate reabsorption, but both enhanced and diminished phosphate transport have been reported. In any case, the hyperphosphatemia, when present, has no discernible effect on the clinical expression of the signs and symptoms of thyrotoxicosis. Acromegaly. Approximately two-thirds of patients with acromegaly have an elevated serum phosphate concentration that is secondary to the chronic effects of growth hormone on renal tubular reabsorption of phosphate27 (see Chap. 206). The serum phosphate level in affected subjects seldom exceeds 5.5 mg/dL but does correlate with disease activity. The mechanism underlying the increased phosphate reabsorption remains unknown. Although the administration of growth hormone to laboratory animals yields increased renal tubular phosphate reabsorption independent of alterations in the PTH concentration,28 somatotropin does not influence renal phosphate transport in a variety of in vitro systems. The widely distributed presence of insulin-like growth factor-I receptors in the kidney29 suggests that insulin-like growth factor-I probably mediates the effect of growth hormone on renal TmP (see Chap. 12). Bisphosphonate Therapy. The administration of disodium etidronate, a pyrophosphate analog, in doses generally higher than 5 mg/kg per day, can induce an increase in the serum phosphate concentration.30 The drug, which is used in the treatment of Paget disease and osteoporosis, causes a time-dependent and dose-dependent increment of the renal TmP and consequent phosphate retention that is largely responsible for the hyperphosphatemia. However, an inhibition of the normal intracellular translocation of phosphorus also contributes to the elevated serum levels. Although an elevation in serum phosphate levels occurs soon after the initiation of therapy, the maximum elevation of the phosphate concentration does not occur until after several days of treatment. The mechanism by which disodium etidronate alters the renal tubular reabsorption of phosphate is unclear. Drug therapy does not alter the serum PTH concentration. Moreover, the drug does not inhibit the urinary cAMP and phosphaturic response to infused PTH. A direct effect on renal phosphate transport, therefore, appears likely. INCREASED PHOSPHATE LOAD Vitamin D Intoxication. An increase of the Pi load from exogenous sources generally does not cause hyperphosphatemia because the kidney excretes the excessive phosphorus. However, an increased serum phosphate concentration may occur in vitamin D intoxication when the gastrointestinal absorption of phosphate is markedly enhanced. Increased phosphate mobilization from bone and a reduction of GFR, secondary to hypercalcemia or nephrocalcinosis, may also contribute to the evolution of the hyperphosphatemia. The chronic ingestion of large doses of vitamin D, in excess of 100,000 U per day, is generally required to cause intoxication. Suspected hypervitaminosis D may be investigated through competitive binding protein assays, which can document excessive amounts of vitamin D and its metabolites in the circulation (see Chap. 54 and Chap. 59). Rhabdomyolysis. Because muscle contains a large amount of phosphate, necrosis of muscle tissue may acutely increase the endogenous phosphate load and result in hyperphosphatemia. Such muscle necrosis (rhabdomyolysis) may complicate heatstroke, acute arterial occlusion, hyperosmolar nonketotic coma, trauma, toxic agents such as ethanol and heroin, and idiopathic paroxysmal myoglobinuria.31,32 Muscle biopsy often reveals myolytic denervation, and, as a consequence, acute renal failure caused by myoglobin excretion frequently complicates the clinical presentation and contributes to the hyperphosphatemia. However, an elevated serum phosphate concentration may precede evidence of renal failure or may occur in its complete absence when rhabdomyolysis is present. The diagnosis is confirmed by elevated serum creatine phosphokinase, uric acid, and lactate dehydrogenase concentrations and by the demonstration of heme-positive urine in the absence of red blood cells. Therapy is directed at the underlying disorder, with maintenance of the extracellular volume to avoid volume depletion and alkalinization of the urine to prevent uric acid accumulation and consequent acute tubular necrosis. Cytotoxic Therapy. Cytotoxic therapy often causes cell destruction and liberation of phosphorus into the circulation.33 The lysis of tumor cells begins within 1 to 2 days after treatment is initiated and is followed quickly by an elevation of the serum phosphate concentration. Hyperphosphatemia supervenes, however, only when the treated malignancies have a large tumor burden, rapid cell turnover, and substantial intracellular phosphorus content. Such malignancies include lymphoblastic leukemia, various types of lymphoma, and acute myeloproliferative syndromes, as well as solid tumors (e.g., small cell carcinoma, breast cancer, and neuroblastoma). Common risk factors for this syndrome include pretreatment renal insufficiency, elevated serum lactate dehydrogenase (LDH) concentration, and hyperuricemia. Additional biochemical abnormalities observed in affected patients include hyperkalemia and hypocalcemia. Malignant Hyperthermia. Malignant hyperthermia is a rare familial syndrome that is characterized by an abrupt rise in body temperature during the course of anesthesia.34 The disease appears to be autosomal dominant in transmission; an elevated serum creatine phosphokinase concentration is found in otherwise normal family members. Hyperphosphatemia results from shifts of phosphate from muscle cells to the extracellular pool. A high mortality rate accompanies the syndrome. HYPOPHOSPHATEMIA Routine measurement of the serum phosphate concentration in hospitalized patients frequently reveals hypophosphatemia (<2.5 mg/dL). The hypophosphatemia may be of no clinical significance or may reflect significant disease and phosphate depletion.35 Hypophosphatemia in humans may result from any one or a combination of three factors: (a) a decrease in either dietary intake or gastrointestinal absorption of Pi; (b) renal phosphate loss, owing to a decrease in the TmP (see Table 67-1); or (c) the transcellular shift of Pi from the extracellular to the intracellular space. The pathologic consequences of hypophosphatemia are variable and depend on the rapidity with which the serum phosphate concentration declines and the concurrence of such related abnormalities as defective vitamin D metabolism or inappropriate PTH secretion. DISTURBANCE OF PHOSPHATE INTAKE/GASTROINTESTINAL ABSORPTION Phosphate Deprivation. Hypophosphatemia and Pi depletion resulting from inadequate dietary intake are rare. With a decline in ingested phosphate, the renal TmP increases and urinary phosphate excretion decreases.36 In addition, gastrointestinal phosphate secretion gradually lessens. However, severe dietary deprivation (<100 mg/d) leads to a prolonged period of negative phosphate balance and total body depletion. Affected female patients may display hypophosphatemia (1.52.5 mg/dL); in interesting contrast, male patients generally do not manifest a decreased serum phosphate concentration in response to the deprivation. The reasons underlying this sex difference are not clear. Nevertheless, attempts to maintain phosphate homeostasis in both sexes include suppression of the serum PTH concentration and increased 1,25(OH)2D3 production. Such a phosphate deprivation syndrome occurs frequently in children with kwashiorkor. Affected subjects manifest severe hypophosphatemia (<1.0 mg/dL) that is often

life threatening. Total starvation does not cause hypophosphatemia. The catabolic effects of total food deprivation result in the release of phosphate from intracellular stores, which compensates for the negative phosphorus balance. However, refeeding of the starved person results in hypophosphatemia when phosphate deprivation is maintained. Hypophosphatemia also occurs in response to prolonged vomiting and gastric suction and is worsened by an associated hypochloremic alkalosis, which results in a urinary phosphate loss. Gastrointestinal Malabsorption. Gastrointestinal absorption of phosphorus may be decreased with the use of antacids that contain aluminum or magnesium; prolonged use of these drugs in large amounts has been associated with hypophosphatemia and a negative phosphorus balance.37 Actually, long-term reduction of the serum phosphorus concentration, owing to chronic, excessive use of antacids, leads to frank osteomalacia and myopathy. Mild to moderate hypophosphatemia may also occur secondary to gastrointestinal diseases that cause steatorrhea or rapid transit time (e.g., Crohn disease, postgastrectomy states, and intestinal fistulas). The decreased serum phosphate concentration is caused by vitamin D malabsorption or deficiency and by resultant secondary hyperparathyroidism and renal phosphate wasting. Affected patients variably manifest osteomalacia.38 However, the relationship between the metabolic bone disease, vitamin D deficiency, and hypophosphatemia is complex. Indeed, osteomalacia may occur in the absence of hypophosphatemia. RENAL PHOSPHATE LOSS Hyperparathyroidism. Approximately 30% of patients with hyperparathyroidism have hypophosphatemia. The decreased serum phosphate concentration results from a PTH-directed decrement of the renal TmP.39 The effects of PTH on the kidney are mediated by cAMP and the dual second messengers inositol triphosphate and diacylglycerol. Thus, nephrogenous cAMP is often elevated. The almost universal concurrence of hypercalcemia in primary hyperparathyroidism distinguishes hypophosphatemia caused by this disorder from virtually all other causes of hypophosphatemia. The presence and extent of hypophosphatemia reflects the extent of the hyperparathyroidism. Other factors, however, are important. The absence of a decreased serum phosphate level in more than one-half of the affected population, for example, may be the result of the effects of phosphate depletion, which in turn limits urinary phosphate excretion. X-linked Hypophosphatemic Rickets/Osteomalacia. The X-linked hypophosphatemic rickets/osteomalacia (XLH) syndrome is an X-linked dominant disorder characterized by hypophosphatemia, a low TmP, growth retardation, osteomalacia, and rickets in growing children (Fig. 67-2). This disease represents the prototypic genetic disorder, causing phosphopenic rickets/osteomalacia (Table 67-2; see Chap. 63 and Chap. 70). Studies that localized the gene locus for this disorder to the chromosome Xp22.1 led to construction of a YAC contig spanning the HYP gene region and ultimately made possible the cloning and identification of the disease gene as PHEXphosphate-regulating gene with homologies to endopeptidases located on the X chromosome.40 More recent investigations suggest that inactivating mutations of PHEX produce inadequate amounts of the endopeptidase, resulting in (a) ineffective or inadequate degradation/inactivation of a phosphate-transport inhibitory protein (phosphatonin); (b) circulation of excessive amounts of this protein; (c) consequent repressed expression of the sodium-dependent phosphate cotransporter; and (d) renal phosphate wasting and hypophosphatemia. In any case, the clinical expression of the disease is widely variable, ranging from an apparent asymptomatic defect to severe bone disease.41 The mildest abnormality is hypophosphatemia without clinically evident bone disease, and the most common clinical manifestation is short stature. Strikingly absent are any abnormalities in the calcium concentration and muscle weakness, both of which are common features of vitamin Ddeficiency osteomalacia.

FIGURE 67-2. X-linked hypophosphatemic rickets. This patient has X-linked hypophosphatemic rickets. Note that the skull has frontal bossing that has occurred from premature closure of the sutures. The legs are bowed despite previous tibial osteotomies to correct the bowing.

TABLE 67-2. Abnormalities Characteristic of the Rachitic/Osteomalacic Disorders

Several factors undoubtedly contribute to the short stature that commonly affects patients with XLH. Refractory hypophosphatemia is most likely a contributing cause. This view is supported by the normal growth realized during infancy in most affected children, in whom the onset of hypophosphatemia does not occur until 6 to 12 months of age. Moreover, the improved growth rate observed in many children when treated adequately (phosphorus and calcitriol) to normalize the The patient also has hypertrophy of the quadriceps muscles secondary to the serum phosphate level further supports the role of hypophosphatemia in the genesis of the growth abnormality.42 Likewise, the beneficial effects of growth hormone on the short stature of affected youths is associated with an increase in the serum phosphorus concentration secondary to the influence of this hormone on renal phosphate excretion (see earlier). The hypophosphatemia in XLH is associated with increased urinary phosphate excretion that arises from a decrease of the renal TmP. Until recently, whether this renal abnormality is primary or secondary to the elaboration of a humoral factor has been controversial. In this regard, the presence of a primary renal abnormality is supported by the observation that primary cultures of renal tubule cells from hyp-mice exhibit a persistent defect in renal Pi transport, which is probably caused by decreased expression of the Na+-phosphate co-transporter (NPT-2) mRNA and immunoreactive protein.43 In contrast, transfer of the defect in renal Pi transport to normal and/or parathyroidectomized normal mice parabiosed to hyp-mice has implicated a humoral factor in the pathogenesis of the disease.44 Current studies have provided compelling evidence that the defect in renal Pi transport in XLH is secondary to the effects of a circulating hormone or metabolic factor. Thus, immortalized cell cultures from the renal tubules of hyp- and gy-mice exhibit normal Na+-phosphate transport, suggesting that the paradoxical effects observed in primary cultures may represent the effects of impressed memory and not an intrinsic abnormality.45 Moreover, the report that cross-transplantation of kidneys in normal and hyp-mice results in neither transfer of the mutant phenotype nor its correction unequivocally established the humoral basis for XLH.46 Subsequent efforts that resulted in localization of the gene encoding the Na+-phosphate co-transporter to chromosome 5 further substantiated the conclusion that the renal defect in brush border membrane phosphate transport is not intrinsic to the kidney in XLH. Although these data establish the presence of a humoral abnormality in XLH, the identity of the putative factor, the spectrum of its activity, and the mechanism by which it functions have not been definitively elucidated. Regardless, preliminary reports suggest the production of a phosphaturic factor by hyp-mouse osteoblasts and marrow mesenchymal cells maintained in culture. These studies argue that a circulating factor, phosphatonin, may play an important role in the pathophysiologic cascade responsible for XLH. An ancillary abnormality in patients with XLH that affects the phenotypic expression of the disease is defective vitamin D metabolism.41 Several investigators have observed that untreated youths and adults have normal serum 1,25(OH)2D3 levels. However, an elevation of this active vitamin D metabolite would be anticipated in the setting of hypophosphatemia and phosphate depletion, factors that increase 1,25(OH)2D3 production in humans. The paradoxical occurrence of hypophosphatemia and normal serum 1,25(OH)2D3 levels is the result of abnormal regulation of renal 25(OH)D-1a-hydroxylase activity. Indeed, studies in hyp-mice have established that

defective regulation is confined to the enzyme localized in the proximal convoluted tubule, which is the site of the abnormal phosphate transport.47,48 Thus, the aberrant vitamin D metabolism would appear to be an acquired defect secondary to changes in the intracellular milieu that depend on the abnormal phosphate transport. Moreover, evidence indicates that increased renal catabolism of calcitriol may also contribute to the aberrant regulation of vitamin D metabolism.49 In any case, data indicate that the collective effects of hypophosphatemia and abnormal vitamin D metabolism underlie the phenotypic expression of XLH. Included in these data is the observation that differences in the manifestations of XLH and the syndrome of hereditary hypophosphatemic rickets with hypercalciuria (HHRH),50 two diseases marked by renal phosphate wasting, can be attributed to the normal vitamin D metabolism maintained in the latter disease (see Table 67-2). Thus, the elevated serum 1,25(OH)2D3 levels in patients with HHRH induce increased gastrointestinal absorption of calcium and phosphorus. Conversely, the defective vitamin D metabolism in patients with XLH abrogates the normal homeostatic response to phosphate depletion (increased gastrointestinal absorption) and alters the therapy required for healing of the bone disease. Successful treatment of XLH requires pharmacologic amounts of calcitriol as well as phosphate supplementation (Fig. 67-3). In fact, some studies indicate that the interplay between hypophosphatemia and vitamin D metabolism may be more complex, because decreased serum phosphate levels may impair binding of 1,25(OH)2D3 receptor complexes to nuclei and thereby promote vitamin D resistance.51

FIGURE 67-3. A, Bone biopsy specimen (Goldner stained) from an untreated patient with X-linked hypophosphatemic rickets viewed under high-power microscope. Characteristically, an excess of osteoid appears along mineralized trabecular bone surfaces. The thick osteoid seams are highlighted by the black arrowheads. B, Bone biopsy specimen (Goldner stained) from a patient with XLH after 9 months of treatment with calcitriol (1.5 g, bid) and oral phosphate supplements (2 g per day; 8 phosphorus supplement [K-Phos Neutral] tablets per day). There is notable resolution of the excess osteoid, and only normal amounts remain (arrowhead). Indeed, quantitative histomorphologic analysis of the specimen confirms complete healing of the osteomalacia, an event unprecedented when other forms of treatment are used.

Fanconi Syndrome. The Fanconi syndrome represents a group of diseases, both genetic and acquired, in which hypophosphatemia, owing to a reduced renal TmP, is associated with additional renal tubular defects, causing loss of glucose, bicarbonate, and amino acids. Additional features of the disorder include rickets, osteomalacia, short stature, and hypokalemia. Among the genetic diseases that cause the syndrome are cystinosis, tyrosinemia, galactosemia, Wilson disease, and fructose intolerance. Acquired forms of the disease have been described in association with heavy metal poisoning, hematologic malignancies, connective tissue diseases, and the use of outdated tetracycline. The disease is primary or idiopathic when the cause is unknown or when only the mode of inheritance is evident. The idiopathic disorder may be sporadic, autosomal-dominant, autosomal-recessive, or X-linked. The clinical presentation of the disease is variable, depending on the cause. However, rickets and osteomalacia are invariable consequences of the hypophosphatemia. In most patients, a decreased or inappropriately normal serum 1,25(OH)2D3 level contributes to the bone disease. In contrast, a small subgroup of affected subjects with inherited disease exhibits an appropriately elevated 1,25(OH)2D3 level. Phenotypic expression of the disorder in these patients is modified in a fashion similar to that in subjects with HHRH52 (see earlier). In any case, regardless of the underlying cause, osteomalacia associated with Fanconi syndrome appears to respond well to treatment with phosphate and vitamin D replacement. In fact, these patients do not necessarily appear to require 1,25(OH)3D3. Vitamin DDependent Rickets. Vitamin Ddependent rickets is an autosomal-recessive disorder that resembles vitamin D deficiency clinically and is often referred to as pseudovitamin D deficiency rickets53 (VDDR; see Chap. 63 and Chap. 70). Affected patients demonstrate, within the first 3 months of life, muscle weakness, hypocalcemia, and evidence of secondary hyperparathyroidism, including aminoaciduria, phosphaturia, and hypophosphatemia (see Table 67-2). With progression, patients develop classic signs of rickets and osteomalacia. Although a decreased serum phosphate concentration is common, this VDDR syndrome is generally considered a calciopenic form of rickets/osteomalacia because of the marked hypocalcemia. The type I disorder results from missense and null mutations in the 1a-hydroxylase gene, localized to chromosome 12q13.3, that abolish enzyme activity and limit production of the active vitamin D metabolite, 1,25(OH)2D.54 A physiologic dose of calcitriol (1 ag per day) generally promotes complete healing of the bone disease and resolution of the biochemical abnormalities, whereas a pharmacologic dose of vitamin D or 25(OH)D is required to achieve similar effects. In the majority of affected patients, therapy with vitamin D or its metabolites must be continued lifelong to prevent relapse. Another group of patients with similar features of VDDR but elevated serum 1,25(OH)2D3 levels has been described53 (see Table 67-2). These patients, with calciopenic rickets/osteomalacia, have variably associated abnormalities, including alopecia, ectodermal anomalies (e.g., multiple milia, epidermal cysts, and oligodontia), and immune dysfunction. This autosomal-recessive form of VDDR (type II) results from mutations in the DNA- and ligand-binding domains of the vitamin D receptor that cause decreased target-organ responsiveness to 1,25(OH)2D3 (Fig. 67-4). The role of the vitamin D receptor in the pathogenesis of this disorder has been confirmed in mice by targeted ablation of the DNA-binding domain of the receptor that causes hypocalcemia, hyperparathyroidism, and alopecia within the first month of life. Effective treatment of this disease likely depends on the nature of the underlying abnormality. Thus, patients with deficient affinity of 1,25(OH)2D3 to receptor and inadequate nuclear translocation respond to high-dose vitamin D or 1,25(OH)2D with complete clinical and biochemical remission. In contrast, patients with other forms of the disease generally remain refractory to treatment with vitamin D or its analogs. However, every patient should receive a 6-month trial of therapy with supplemental calcium (13 g per day) and vitamin D or its analogs [25(OH)D or 1,25(OH)2D], often in massive doses in severe cases. If the abnormalities of the syndrome do not normalize in response to this treatment, clinical remission may be achieved by administering high-dose oral calcium or long-term intracaval infusion of calcium.55

FIGURE 67-4. Model for potential cellular defects underlying vitamin Ddependent rickets type II. Sites of potential abnormalities underlying vitamin Ddependent rickets type II are indicated by the circled numbers. These defects include (1) a decreased number of cytosolic receptors for 1,25(OH)2D3; (2) altered affinity or stability of the cytosolic receptor; (3) abnormal translocation and/or transformation of the 1,25(OH) 2D3 receptor; (4, 5) an altered nuclear and/or chromatin binding site; (6) abnormal transcription; and/or (7, 8) defective protein synthesis or action, thereby precluding expression of 1,25(OH)2D3-specific effects such as calcium transport. [1,25, 1,25(OH)2D3; RP, receptor protein; HnRNA, heterogeneous RNA; ER, endoplasmic reticulum.]

Autosomal-Dominant Hypophosphatemic Rickets. The existence of autosomal-dominant hypophosphatemic rickets was first reported in 1979,56 and the autosomal-dominant transmission of this disease has since been confirmed. Affected patients exhibit hypophosphatemia as a result of renal phosphate wasting (without a diffuse tubular abnormality), lower-extremity bowing, rickets, and osteomalacia. Subgroups of these patients demonstrate delayed penetrance of clinically apparent disease and an increased tendency to fracture or apparent resolution of the disease postadolescence. Regardless, all patients with the disorder manifest

aberrant regulation of vitamin D metabolism, similar to that found in other phosphopenic rachitic diseases, and normal serum parathyroid hormone levels. The gene locus for this disorder has been identified as chromosome 12p13 in the 18-cM interval between the flanking markers D12S100 and D12S397.57 Tumor-Induced Osteomalacia. The tumor-induced osteomalacia syndrome is characterized by the remission of unexplained osteomalacia and rickets after resection of a coexisting tumor. The cases of ~100 patients with this disorder have been reported.41,58 Tumors associated with the syndrome have generally been of mesenchymal origin (e.g., giant cell tumors of bone, sarcomas, and hemangiomas).58a However, the frequent occurrence of Looser zones in the radiographs of moribund patients with carcinomas of epidermal and endodermal derivation indicates that the disease may be more widespread and prevalent than previously recognized. The observation of tumor-induced osteomalacia in association with breast carcinoma and prostate carcinoma supports this conclusion (see Chap. 63 and Chap. 219). Hypophosphatemia is the primary biochemical abnormality in affected patients, and it results from a decreased renal TmP. Moreover, affected subjects invariably exhibit a decreased serum 1,25(OH)2D3 concentration that apparently results from inhibition of renal 25(OH)D-1a-hydroxylase activity (see Table 67-2). These abnormalities result from the adverse effects of a tumor-secreted product on normal biochemical processes. Although a heat-sensitive protein of 8000 to 25,000 daltons has been isolated from a sclerosing hemangioma and is a likely tumor-secreted product,59 other studies, which document the presence in various disease states of additional phosphate-transport inhibitors, indicate that the tumor-secreted factors are heterogeneous. In affected patients, if tumor resection is possible, the hypophosphatemia [and abnormal serum 1,25(OH)2D3 levels] quickly normalize. If tumor resection is precluded by the size of the primary lesion or metastases, treatment with 1,25(OH)2D3 alone or with phosphate supplements restores the serum phosphate and often effects healing of the bone disease. Hereditary Hypophosphatemic Rickets with Hypercalciuria. Hereditary hypophosphatemic rickets with hypercalciuria is a rare, genetic disease marked by hypophosphatemia (secondary to a decreased renal TmP) and hypercalciuria.50 In contrast to the other diseases in which renal phosphate transport is abnormal, serum 1,25(OH)2D3 levels are elevated. The increased bio-active vitamin D causes enhanced gastrointestinal absorption of calcium and consequent hypercalciuria. In affected subjects, hypophosphatemia (and bone disease) often respond to phosphate supplements alone. TRANSCELLULAR SHIFT OF PHOSPHORUS In a large proportion of clinically important cases of hypophos-phatemia, a sudden shift of phosphorus from the extracellular to the intracellular compartment is responsible for the decline of the serum phosphate concentration. This ion movement occurs in response to naturally occurring disturbances and after the administration of certain compounds. Alkalosis. Alkalosis secondary to intense hyperventilation may depress serum phosphate levels to < 1 mg/dL.60 A similar degree of alkalemia, owing to excess bicarbonate, also causes hypophosphatemia, but of a much lesser magnitude (2.53.5 mg/dL). The disparity between the effects of a respiratory and metabolic alkalosis is related to the more pronounced intracellular alkalosis that occurs during hyperventilation. The phosphate shift to the intracellular compartment results from its use for glucose phosphorylation, which is a process stimulated by a pH-dependent activation of phos-phofructokinase. Glucose Administration. The administration of glucose and insulin often results in moderate hypophosphatemia.61 Endogenous or exogenous insulin increases the cellular uptake not only of glucose but also of phosphorus. The most responsive cells are those of the liver and skeletal muscle. The decline of the serum phosphate concentration generally does not exceed 0.5 mg/dL. A lesser decrease is manifest in patients with type 2 diabetes mellitus and insulin resistance or those with a disease causing a diminished skeletal mass. The administration of fructose and glycerol similarly reduces the serum phosphate concentration. In contrast to glucose, fructose administration may be associated with more pronounced hypophosphatemia; the more striking effect is caused by unregulated uptake by the liver. COMBINED MECHANISMS Special clinical situations occur in which more than one mechanism contributes to the development of hypophosphatemia. These disorders represent some of the more common and profound causes of a decreased serum phosphorus concentration. Alcoholism. Alcoholic patients frequently enter the hospital with hypophosphatemia. However, many do not exhibit a decreased serum phosphate concentration until several days have elapsed and refeeding has begun. The multiple factors that underlie the hypophosphatemia include poor dietary intake, use of phosphate binders to treat gastritis, excessive urinary loss of phosphorus, and shift of phosphorus from the extracellular to the intracellular compartment resulting from glucose administration or hyperventilation, which occurs in patients with cirrhosis or during alcohol withdrawal.61 Moreover, many alcoholic patients are hypomagnesemic, which potentiates renal-phosphate wasting by an unclear mechanism. Burns. Within several days after sustaining an extensive burn, patients often manifest severe hypophosphatemia. The initial insult induces a transient retention of salt and water. When the fluid is mobilized, significant urinary phosphorus loss ensues. Coupled with the shift of phosphorus to the intra-cellular compartment (which occurs secondary to hyperventilation), phosphate loss in cutaneous exudates, and the anabolic state, profound hypophosphatemia may result. Nutritional Recovery Syndrome. Refeeding of starved patients or maintaining nutritional support by parenteral nutrition or by tube feeding without adequate phosphorus supplementation may also cause hypophosphatemia.62 A prerequisite for the decreased serum phosphate concentration in affected patients is that their cells must be capable of an anabolic response. As new proteins are synthesized and glucose is transported intracellularly, phosphate demand depletes reserves. Several days are generally required after the initiation of refeeding to establish an anabolic condition. In patients receiving total parenteral nutrition, serum phosphate levels may be further depressed if sepsis supervenes and a respiratory alkalosis develops. Diabetic Ketoacidosis. Poor control of blood glucose levels and consequent glycosuria, polyuria, and ketoacidosis invariably cause renal phosphate wasting.61 The concomitant volume contraction may yield a normal serum phosphate concentration. However, with insulin therapy, the administration of fluids, and correction of the acidosis, serum and urine phosphate levels fall precipitously. The resultant hypophosphatemia may contribute to insulin resistance and slow the resolution of the ketoacidosis. Hence, the administration of phosphate may improve the capacity to metabolize glucose and facilitate recovery. Acute Phase Response Syndrome. Hypophosphatemia is common (11.4%) in acutely ill patients who have experienced severe trauma or infection and manifest fever with leukopenia or leukocytosis. Affected patients often have an associated hyperglycemia secondary to tissue injury and/or infection. Thus, the hypophosphatemia is most likely caused by shifts of the extracellular phosphate into cells. Alternatively, the high levels of inflammatory cytokines, which characterize early sepsis, may play a role in the genesis of the hypophosphatemia. With resolution of the acute phase of illness, the hypophosphatemia resolves.

CLINICAL SIGNS AND SYMPTOMS OF ABNORMAL SERUM PHOSPHATE LEVELS

A wide variety of diseases and syndromes with varying clinical manifestations have the characteristic biochemical abnormalities of hyperphosphatemia or hypophosphatemia. Further-more, a unique complex of disturbances often is directly related to the abnormal phosphate homeostasis. The recognition of these signs and symptoms may lead to appropriate biochemical testing, the diagnosis of an unsuspected disease, and initiation of lifesaving or curative treatment. HYPERPHOSPHATEMIA Hypocalcemia and consequent tetany are the most serious clinical sequelae of hyperphosphatemia.63 The decreased serum calcium concentration results from the deposition of calciumphosphate salts in soft tissue, a process that may lead to symptomatic ectopic calcification. The dystrophic calcification is frequently seen in acute and chronic renal failure, hypoparathyroidism, pseudohypopara-thyroidism, and tumoral calcinosis. Indeed, deposition of calciumphosphate complexes in the kidney may predispose a patient to acute renal failure. When the calciumphosphate product exceeds 70, the probability that soft-tissue calcification will occur increases sharply. In addition, local factors, such as tissue pH and injury (e.g., necrotic or hypoxic tissue), may predispose the patient to precipitation of the calciumphosphate salts. In chronic renal failure, calcification occurs in arteries, muscle tissue, periarticular spaces, myocardial conduction system, lungs, and kidney. Affected patients may also have ocular calcification, causing the red eye syndrome of uremia, and subcutaneous calcification, which also plays a role in uremic pruritus. Alternatively, a predisposition to calcification of periarticular surfaces of the hips, elbows, shoulders, and other large joints occurs in tumoral calcinosis. In some disease states, hyperphosphatemia may also play an important role in the development of secondary hyperparathyroidism.27 A decrement in the serum calcium concentration secondary to hyperphosphatemia stimulates the release of PTH. Furthermore, hyperphosphatemia decreases the activity of renal 25(OH)D-1a-hydroxylase. The consequent diminished production of 1,25(OH)2D3 impairs the gastrointestinal absorption of calcium and induces skeletal resistance to

PTH, which are influences that augment the development of hyperparathyroidism. Thus, hyperphosphatemia triggers a cascade of events that have an impact on calcium homeostasis at multiple sites. The prevention of secondary hyperparathyroidism, metabolic bone disease, and soft-tissue and vascular calcification in affected patients, therefore, depends on ultimately controlling the serum phosphate concentration. HYPOPHOSPHATEMIA A low serum phosphorus level is associated with symptoms only if there is concomitant phosphate depletion. The presence of phosphate deficiency, however, may cause widespread disturbances. This is not surprising, because severe hypophosphatemia causes two critical abnormalities that have an impact on virtually all organ systems. First, a deficiency of 2,3-diphosphoglycerate (2,3-DPG) occurs in red cells, which is associated with an increased affinity of hemoglobin for oxygen and, therefore, tissue hypoxia. Second, there is a decline of tissue ATP content and a concomitant decrease in the availability of energy-rich phosphate compounds that are essential for cell function.64,65 The major clinical syndromes resulting from these abnormalities include nervous system dysfunction, anorexia, nausea, vomiting, ileus, muscle weakness, cardiomyopathy, respiratory insufficiency, hemolytic anemia, and impaired leukocyte and platelet function. Additionally, phosphate deficiency causes osteomalacia and bone pain, which are clinical sequelae that are probably independent of the aforementioned abnormalities. Central nervous system dysfunction has been well characterized in severe hypophosphatemia, especially in patients receiving total parenteral nutrition for diseases causing severe weight loss. A sequence of symptoms compatible with a metabolic encephalopathy usually begins one or more weeks after the initiation of therapy with solutions that contain glucose and amino acids but lack adequate phosphorus supplementation to prevent hypophosphatemia. Irritability, muscle weakness, numbness, and paresthesia mark the onset of dysfunction, with progression to dysarthria, confusion, obtundation, coma, and death.66 These patients have a profoundly diminished red-cell 2,3-DPG content. Both biochemical abnormalities and clinical symptoms improve as patients receive phosphorus supplementation. Peripheral neuropathies, Guillain-Barrlike paralysis, hyporeflexia, intention tremor, and ballismus have also been described with hypophosphatemia and phosphate depletion. The effects of hypophosphatemia on muscle depend on the severity and chronicity of the deficiency. Chronic phosphorus deficiency results in a proximal myopathy with striking atrophy and weakness. Osteomalacia frequently accompanies the myopathy, so patients complain of pain in weight-bearing bones. Normal values for serum creatine phosphokinase and aldolase activities are characteristically present. Rhabdomyolysis does not occur with chronic phosphate depletion. In contrast, acute hypophosphatemia can lead to rhabdomyolysis with muscle weakness and pain. Most cases occur in chronic alcoholics or patients receiving total parenteral nutrition. In both groups of patients, muscle pain, swelling, and stiffness occur 3 to 8 days after the initiation of therapies that do not contain adequate amounts of phosphorus. Muscle paralysis and diaphragmatic failure may occur. Studies of muscle tissue from chronically phosphate-depleted dogs made acutely hypophosphatemic showed a decrement in cellular content of phosphorus, ATP, and adenosine diphosphate. Rhabdomyolysis occurred in these muscle fibers. The laboratory findings in patients with hypophosphatemic myopathy and with rhab-domyolysis include elevated serum creatine phosphokinase levels; however, serum phosphate levels may become normal if enough necrosis has occurred with subsequent phosphorus release. Also, renal failure and hypocalcemia can be associated with the syndrome. Myocardial performance can be abnormal at serum phosphate levels of 0.7 to 1.4 mg/dL. This occurs when ATP depletion causes impairment of the actinmyosin interaction, the calcium pump of the sarcoplasma, and the sodiumpotassium pump of the cell membrane.67 The net result is reduced stroke work and cardiac output, which may progress to congestive heart failure. These problems are reversible with phosphate replacement. Respiratory failure can occur as a result of failure of dia-phragmatic contraction in hypophosphatemic patients. When serum phosphate levels are raised, diaphragmatic contractility improves. The postulated mechanism for the respiratory failure is muscle weakness secondary to inadequate levels of ATP and decreased glycolysis as the result of phosphate depletion.68 Two disturbances of red-cell function may occur secondary to phosphorus deficiency. First, as intraerythrocyte ATP production is decreased, the erythrocyte cell membrane becomes rigid, which can cause hemolysis.69 This is rare and is usually seen in septic, uremic, acidotic, or alcoholic patients when serum phosphate levels are <0.5 mg/dL. Second, the limited production of 2,3-DPG causes a leftward shift of the oxyhemoglobin curve and impairs the release of oxygen to peripheral tissues. Leukocyte dysfunction, which complicates phosphate deficiency, includes decreased chemotaxis, phagocytosis, and bactericidal activity.70 These abnormalities increase the host susceptibility to infection. The mechanism by which hypophosphatemia impairs the various activities of the leukocyte probably is related to impairment of ATP synthesis. Decreased availability of energy impairs microtubules that regulate the mechanical properties of leukocytes and limit the rate of synthesis of organic phosphate compounds that are necessary for endocytosis. Abnormal platelet survival, causing profuse gastrointestinal bleeding and cutaneous bleeding, has also been described in association with phosphate depletion in animal studies. Despite these abnormalities, there is little evidence that hypophosphatemia is a primary cause of hemorrhage in humans. Perhaps the most consistent abnormalities associated with phosphate depletion are those on bone. Acute phosphate depletion induces dissolution of apatite crystal from the osseous matrix. This effect may be a result of the 1,25(OH)2D3 level, which is increased in response to phosphate depletion in both animals and humans. More prolonged hypophosphatemia leads to rickets and osteomalacia. This complication is a common feature of phosphate depletion. However, the ultimate cause is variable. Although simple phosphate depletion alone may underlie the genesis of the abnormal mineralization, in many disorders, the defect is secondary to phosphate depletion and commensurate 1,25(OH)2D3 deficiency. Thus, treatment of this complication may often require combination therapy, including phosphate supplements and calcitriol.

TREATMENT OF ABNORMAL SERUM PHOSPHATE LEVELS

Treatment of the myriad of diseases that characteristically display hyperphosphatemia or hypophosphatemia depends on determining the mechanism underlying their pathogenesis. The cause can almost always be ascertained by assessment of the clinical setting, determination of renal function, measurement of urinary phosphate excretion, and analysis of arterial carbon dioxide tension and pH. Therapy is aimed at correcting both the serum phosphate concentration and the associated complications. Theoretically, elevated serum phosphate levels may be reduced by decreasing the TmP, increasing the GFR, or diminishing the phosphate load (exogenous or endogenous). There are no generally available pharmacologic means of acutely altering the GFR or reducing the TmP. However, chronic use of drugs, such as acetazolamide, which decreases TmP and induces phosphaturia, is effective as ancillary treatment of disorders such as tumoral calcinosis. Nevertheless, regulation of hyperphosphatemia is most often achieved by reducing the renal phosphate load. In tumoral calcinosis and chronic renal failure, such an effect is obtained by restricting the dietary phosphate intake or by administering calcium carbonate or aluminum hydroxide. Alternative strategies for management of load-dependent hyperphosphatemia include the administration of intravenous calcium or intravenous glucose and insulin. The consequence of such intervention is sequestration of phosphate in bone or soft tissues. Dialysis can also be used for the acute management of load-dependent disorders or for the chronic maintenance of phosphate overload, such as that which complicates chronic renal failure. In the latter condition, phosphate binders are essential.71,72 The treatment of phosphate depletion depends on replacing body phosphorus stores. Preventive measures, however, preclude the onset of phosphate depletion. Thus, appropriate monitoring of patients taking large doses of antacids that contain aluminum and provision of phosphorus intravenously to patients with diabetic ketoacidosis preserves phosphate stores. Alternatively, the treatment of established phosphate depletion may require 2.5 to 4.0 g per day of phosphate, preferably administered orally in four equally divided doses. This can be done by giving phosphorus supplement tablets (K-Phos Neutral), which contain 250 mg elemental phosphorus per tablet. If oral therapy is not tolerated and the serum phosphate level shows a downward trend, approaching dangerous levels (<1.2 mg/dL), intravenous phosphate supplementation at a dose of 10 mg/kg per day may be administered. Such therapy should be discontinued when the serum phosphate concentration reaches values > 2 mg/dL. However, therapy is not required for many of the conditions resulting in phosphate depletion. Only when the consequences of severe depletion are manifest does treatment need to be initiated. In many of the chronic causes of hypophosphatemia, marked by a reduction in the TmP, correction of the hypophosphatemia is difficult. In these diseases, chronic phosphate supplementation with K-Phos Neutral alone or in combination with vitamin D or its metabolites increases the serum phosphorus (not always to normal) and improves the complications associated with hypophosphatemia. CHAPTER REFERENCES
1. Wilkinson R. Absorption of calcium, phosphorus and magnesium. In: Nordin BEC, ed. Calcium, phosphate and magnesium metabolism. New York: Longman, 1976:218. 2. Gennari C, Bernini M, Nordi P, Fusi L. Dissociation of absorptions of calcium and phosphate in different pathophysiological states in man. In: Massry SG, Maschio G, Ritz E, eds. Phosphate and

3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58.

mineral metabolism. New York: Plenum Press, 1984:195. Walling MW. Intestinal calcium and phosphate transport: differential responses to vitamin D3 metabolites. Am J Physiol 1977; 233:E488. Bijvoet OLM. Kidney function in calcium and phosphate metabolism. In: Avioli LV, Krane SM, eds. Metabolic bone disease, vol 1. New York: Academic Press, 1977:49. Trohler U, Bonjour J-P, Fleisch H. Inorganic: phosphate homeostasis: renal adaptation to the dietary intake in intact and thyroparathyroidectomized rats. J Clin Invest 1976; 57:264. Agus ZS. Renal handling of phosphate. In: Massry SG, Glassock RJ, eds. Textbook of nephrology. Baltimore: Williams & Wilkins, 1983:389. Cheng L, Jacktor B. Sodium gradient-dependent phosphate transport in renal brush border membrane vesicles. J Biol Chem 1981; 256:1556. Donsa TP, Kempson SA. Regulation of renal brush border membrane transport of phosphate. Miner Electrolyte Metab 1982; 7:113. Tenenhous HS. Cellular and molecular mechanisms of renal phosphate transport. J Bone Miner Res 1997; 12:159. Dennis VW, Stead WW, Myers JL. Renal handling of phosphate and calcium. Annu Rev Physiol 1979; 41:257. Slatopolsky E, Robson AM, Elkan I, Bricker NS. Control of phosphate excretion in uremic man. J Clin Invest 1968; 47:1865. Nagant De Deuxchaisnes C, Krane SM. Hypoparathyroidism. In: Avioli LV, Krane SM, eds. Metabolic bone diseases, vol 11. New York: Academic Press, 1978:183. De Luca F, Baron J. Molecular biology and clinical importance of the Ca 2+-sensing receptor. Curr Opin Pediatr 1998; 10:435. Pearce SH, Cheetham T, Imrie H, et al. A common and recurrent 1 3-bp deletion in the autoimmune regulator gene in British kindreds with autoimmune polyendocrinopathy type 1. Am J Hum Gen 1998; 63:1675. Trump D, Dixon PH, Mumm S, et al. Localisation of X linked recessive idiopathic hypoparathyroidism to a 1.5 Mb region on Xq26-q27. J Med Gen 1998; 35:905. Ahmed SF, Dixon PH, Bonthron DT, et al. GNAS 1 mutational analysis in pseudohypoparathyroidism. Clin Endocrinol (Oxf) 1998; 49:525. Jppner H, Schipani E, Bastepe M, et al. The gene responsible for pseudohypoparathyroidism type 1b is paternally imprinted and maps in four unrelated kindreds to chromosome 20q13.3. Proc Natl Acad Sci U S A 1998; 95:11798. Lyles KW, Burkes EJ, Ellis GJ, et al. Genetic transmission of tumoral calcinosis: autosomal dominant, with variable clinical impressivity. J Clin Endocrinol Metab 1985; 60:1093. Mozaffanan G, Lafferty FW, Pearson OH. Treatment of tumoral calcinosis with phosphorus deprivation. Ann Intern Med 1972; 77:741. Sinah TK, Allen DO, Queener SF, et al. Effects of acetazolamide on the renal excretion of phosphate in hyperparathyroidism and pseudohypoparathyroidism. J Lab Clin Med 1977; 89:1188. Quarles LD, Murphy G, Econs MJ, et al. Uremic tumoral calcinosis: preliminary observations suggesting an association with aberrant vitamin D homeostasis. Am J Kidney Dis 1991; 18:706. Tezelman S, Siperstein AE, Duh QV, et al. Tumoral calcinosis: controversies in the etiology and alternatives in the treatment. Arch Surg 1993; 128:737. Kuriyama S, Tomonari H, Nakayama M, et al. Successful treatment of tumoral calcinosis using CAPD combined with hemodialysis with low-calcium dialysate. Blood Purif 1998; 16:43. Ueyoshi A, Ota K. Clinical appraisal of vinpocetine for the removal of intractable tumoral calcinosis in hemodialysis patients with renal failure. J Int Med Res 1992; 20:435. Simionesciu L, Dumitriu L, Dumitriu V, et al. The serum osteocalcin levels in patients with thyroid disease. Endocrinologie 1988; 26:27. DeMenis E, DaRin G, Roiter I, et al. Bone turnover in overt and subclinical hyperthyroidism due to autonomous thyroid adenoma. Horm Res 1992; 37:217. Slatopolsky E, Rutherford WE, Rosenbaum R, et al. Hyperphosphatemia. Clin Nephrol 1977; 7:138. Corvilain J, Abramov M, Bergaus A. Effect of growth hormone on tubular transport of phosphate in normal and parathyroidectomized dogs. J Clin Invest 1964; 43:1608. Ohashi H, Rosen KM, Smith FE, et al. Characterization of type 1 IGF receptor and IGF 1 mRNA expression in cultured human and bovine glomerular cells. Regul Pept 1993; 48:9. Walton RJ, Russell RGG, Smith R. Changes in the renal and extrarenal handling of phosphate induced by disodium etidronate (EHDP) in man. Clin Sci Mol Med 1975; 49:45. Grossman RA, Hamilton RW, Morse BM, et al. Nontraumatic rhabdomyolysis and acute renal failure. N Engl J Med 1974; 291:807. Koffler A, Fnedler RM, Massry SG. Acute renal failure due to nontraumatic rhabdomyolysis. Ann Intern Med 1976; 85:25. Zusman J, Brown DM, Nesbitt ME. Hyperphosphatemia, hyperphosphaturia and hypocalcemia in acute lymphoblastic leukemia. N Engl J Med 1973; 289:1335. Denborough MA, Forster JFA, Hudson MC, et al. Biochemical changes in malignant hyperpyrexia. Lancet 1970; 1:1137. Halevy J, Bulvik S. Severe hypophosphatemia in hospitalized patients. Arch Intern Med 1988; 148:153. Levine BS, Cooks PW, Katz JA, et al. Early events during renal adaptation to low dietary phosphorus. Kidney Int 1984; 25:148. Latz M, Zisman E, Bartter FC. Evidence for a phosphorus-depletion syndrome in man. N Engl J Med 1968; 278:409. Baker LRI, Acknll P, Cattell WR, et al. Iatrogenic osteomalacia and myopathy due to phosphate depletion. BMJ 1974; 3:150. Haas HG, Dambacher MA, Buncaga J, Lauffenburger T. Renal effects of calcitonin and parathyroid extract in man. J Clin Invest 1971; 50:2689. Francis F, Henning S, Korn B, et al. A gene (PEX) with homologies to endopeptidases is mutated in patients with X-linked hypophosphatemic rickets. Nat Gen 1995; 11:130. Lobaugh B, Burch WM Jr, Drezner MK. Abnormalities of vitamin D metabolism and action in the vitamin D resistant rachitic and osteomalacic diseases. In: Kumar R, ed. Vitamin D: basic and clinical aspects. Boston: Martinus Nijhoff, 1984:665. Friedman NE, Lobaugh B, Drezner MK. Effects of calcitriol and phosphorus therapy on the growth of patients with X-linked hypophosphatemia. J Clin Endocrinol Metab 1993; 76:839. Bell CL, Tennenhouse HS, Scriver CR. Primary cultures of renal epithelial cells from X-linked hypophosphatemic (Hyp) mice express defects in phosphate transport and vitamin D metabolism. Am J Hum Genet 1988; 43:293. Meyer RA Jr, Meyer MH, Gray RW. Parabiosis suggests a humoral factor is involved in X-linked hypophosphatemia in mice. J Bone Miner Res 1989; 4:493. Nesbitt T, Econs MJ, Byun JK, et al. Phosphate transport in immortalized cell cultures from the renal proximal tubule of normal and hyp mice. Evidence that the Hyp gene locus product is an extrarenal factor. J Bone Miner Res 1995; 10:1327. Nesbitt T, Coffman TM, Griffiths R, et al. Crosstransplantation of kidneys in normal and hyp-mice: evidence that the hyp-mouse phenotype is unrelated to an intrinsic renal defect. J Clin Invest 1992; 89:1453. Nesbitt T, Drezner MK, Lobaugh B. Abnormal parathyroid hormone stimulation of 25-hydroxyvitamin D-1a-hydroxylase activity in the hypophosphatemic mouse: evidence for a generalized defect of vitamin D metabolism. J Clin Invest 1986; 77:181. Nesbitt T, Lobaugh B, Drezner MK. Calcitonin stimulation of renal 25 hydroxyvitamin D-1a hydroxylase activity in Hyp-mice: evidence that the regulation of calcitriol production is not universally abnormal in X-linked hypophosphatemia. J Clin Invest 1987; 79:15. Tenenhouse HS, Yip A, Jones G. Increased renal catabolism of 1,25-dihydroxyvitamin D3 in murine X-linked hypophosphatemic rickets. J Clin Invest 1988; 81:461. Tieder M, Modai D, Samuel R, et al. Hereditary hypophosphatemic rickets with hypercalciuria (HHRH): a new syndrome. In: Norman AW, Schaefer K, Grigoleit H-G, Herrath DV, eds. Vitamin D: a chemical, biochemical and clinical update. Berlin: Walter de Gruyter, 1985:1107. Yamamoto T, Seino Y, Tanaka H, et al. Effects of the administration of phosphate on nuclear 1,25-dihydroxyvitamin D3 uptake by duodenal mucosal cells of Hyp-mice. Endocrinology 1988; 122:576. Tieder M, Arie R, Modai D, et al. Elevated serum 1,25-dihydroxyvitamin D concentrations in siblings with primary Fanconi's syndrome. N Engl J Med 1988; 319:845. Lieberman UA, Eil C, Marx SJ. Hereditary hypocalcemic vitamin D resistant rickets (HHDR). In: Frame B, Potts JT, eds. Clinical disorders of bone and mineral metabolism. Amsterdam: Excerpta Medica, 1983:441. Fu GK, Lin D, Zhang MY, et al. Cloning of human 25-hydroxyvitamin D-1a-hydroxylase and mutations causing vitamin D-dependent rickets type 1. Mol Endocrinol 1997; 11:1961. Bliziotes M, Yergey AJ, Nanes MS, et al. Absent intestinal response to calciferols in hereditary resistance to 1,25-dihydroxyvitamin D: documentation and effective therapy with high dose intravenous calcium infusions. J Clin Endocrinol Metab 1988; 66:294. Harrison HE, Harrison HC. Rickets and osteomalacia. In: Harrison HE, Harrison HC. Disorders of calcium and phosphate metabolism. Philadelphia: WB Saunders, 1979;141. Econs MJ, McEnery PT, Lennon F, et al. Autosomal dominant hypophosphatemic rickets is linked to chromosome 12p13. J Clin Invest 1997; 100:2653. Drezner MK. Tumor-associated rickets and osteomalacia.. In: Favus MJ, ed. Primer on the metabolic bone diseases and disorders of mineral metabolism, ed 2. New York: Raven Press, 1993:282.

58a. Sandhu FA, Martuza RL. Craniofacial hemangiopericytoma associated with oncogenic osteomalacia. J Neurooncol 2000; 46:241. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. Cai Q, Hodgson MD, Kao PC, et al. Inhibition of renal phosphate transport by a tumor product in a patient with oncogenic osteomalacia. N Engl J Med 1994; 330:1645. Mostellar ME, Tuttle EP. Effects of alkalosis on plasma concentration and urinary excretion of inorganic phosphate in man. J Clin Invest 1964; 43:138. Knochel JP. The pathophysiology and clinical characteristics of severe hypophosphatemia. Arch Intern Med 1977; 137:203. Sheldon GF, Grzyb S. Phosphate depletion and repletion: relation to parenteral nutrition and oxygen transport. Ann Surg 1975; 182:683. Herbert LA, Lemann J, Peterson JR, Lennon EJ. Studies of the mechanism by which phosphate infusion lowers serum calcium concentration. J Clin Invest 1966; 45:1886. Duhm J. 2,3-DPG-induced displacements of the oxyhemoglobin dissociation curve of blood: mechanisms and consequences. Adv Exp Biol Med 1971; 37A:179. Travis SF, Sugerman HJ, Ruberg RL, et al. Alterations of red cell glycolytic intermediates and oxygen transport as a consequence of hypophosphatemia in patients receiving intravenous hyperalimentation. N Engl J Med 1977; 297:901. Parfitt AM, Kleerekoper M. Clinical disorders of calcium, phosphorus and magnesium metabolism. In: Maxwell MH, Kleeman CR, eds. Clinical disorders of fluid and electrolyte metabolism. New York: McGraw-Hill, 1980:947. O'Connor LR, Wheeler WS, Bethune JE. Effect of hypophosphatemia on myocardial performance in man. N Engl J Med 1977; 297:901. Newman JH, Neff TA, Ziporen P. Acute respiratory failure associated with hypophosphatemia. N Engl J Med 1977; 296:1101. Klock JC, Williams HE, Mentzer WK. Hemolytic anemia and somatic cell dysfunction in severe hypophosphatemia. Arch Intern Med 1974; 134:360. Craddock PR, Yawota Y, Van Santen L, et al. Acquired phagocyte dysfunction: a complication of the hypophosphatemia of parenteral hyperalimentation. N Engl J Med 1974; 290:1403. Block GA, Port FK. Re-evaluation of risks associated with hyperphosphatemia and hyperparathyroidism in dialysis patients: recommendations for a change in management. Am J Kidn Dis 2000; 35:1226. Chertow GM, Dillon MA, Amin N, Burke SK. Sevelamer with and without calcium and vitamin D: observations from a long-term open-label clinical trial. J Ren Nutr 2000; 10:125.

CHAPTER 68 MAGNESIUM METABOLISM Principles and Practice of Endocrinology and Metabolism

CHAPTER 68 MAGNESIUM METABOLISM

ROBERT K. RUDE Magnesium Deficiency Etiology of Magnesium Deficiency Gastrointestinal Disorders Renal Loss Endocrine and Metabolic Causes Manifestations of Magnesium Deficiency Signs and Symptoms Hypocalcemia Hypokalemia Diagnosis Therapy Magnesium Excess Etiology of Hypermagnesemia Manifestations Therapy Chapter References

Magnesium is the fourth most abundant cation in the body and the major intracellular divalent cation. The amount of magnesium in an adult human is ~2000 mEq (24 g), of which 60% is in the skeleton and 40% is intracellular. Less than 1% is found in the extracellular fluid compartment.1,2 and 3 Magnesium is essential for the function of many enzyme systems, including those that use adenosine triphosphate.4,5 Magnesium can be absorbed along the entire gastrointestinal tract, including both the small and large intestine, but it is absorbed most efficiently in the jejunum and ileum, through both a saturable transport system and passive diffusion.6,7 Fractional intestinal magnesium absorption appears to depend on the amount of magnesium in the diet; with an average magnesium intake, ~40% is absorbed.6,7 The major dietary sources of magnesium include vegetables and meats, but magnesium is ubiquitous in food. Some studies have suggested that 1,25-dihydroxyvitamin D [1,25(OH)2D] enhances intestinal magnesium absorption,8 although this may be secondary to the effect of this vitamin on calcium transport.9 The regulation of magnesium homeostasis occurs primarily in the kidney.10 Micropuncture studies have demonstrated that magnesium is absorbed along the proximal convoluted tubule (5% to 15%), the thick ascending limb of the loop of Henle (50% to 60%), and the distal tubule (10%).11 Although a true tubular maximum for magnesium has not been demonstrated in isolated micropuncture studies of the nephron, the whole kidney does demonstrate a maximal capacity to reabsorb magnesium.12,13 When the filtered magnesium load exceeds that amount filtered under normal circumstances (serum ultrafilterable magnesium concentration of ~1.3 mEq/L), the excess is excreted. Under conditions of magnesium deprivation, however, when the filtered load is below 1.3 mEq/L, magnesium is virtually completely reabsorbed. Despite efficient regulation of magnesium metabolism, no factor or hormone has yet been described that primarily regulates this system.

MAGNESIUM DEFICIENCY
ETIOLOGY OF MAGNESIUM DEFICIENCY Although magnesium is avidly conserved by the body, magnesium deficiency is a common occurrence. Up to 10% of the patients admitted to large city hospitals are hypomagnesemic.14 In one hospital, 65% of the patients in a medical intensive care unit were hypomagnesemic.15 Because magnesium is found in virtually all foods, dietary deprivation is a most unusual cause of hypomagnesemia in persons with normal caloric intake. Much more commonly, magnesium deficiency is caused by excessive losses from either the gastrointestinal tract or the kidney.16 Table 68-1 outlines causes of magnesium deficiency.

TABLE 68-1. Differential Diagnosis of Magnesium Deficiency

GASTROINTESTINAL DISORDERS Magnesium may be lost via the gastrointestinal tract, either by excessive loss of secreted fluids or impaired absorption of both dietary and endogenous magnesium.2,7 The magnesium concentration of upper intestinal tract fluids is ~0.5 mmol/L, and vomiting or nasogastric suction may contribute to magnesium depletion because of loss of these fluids. Disorders of the small bowel are frequently associated with magnesium deficiency.2,7 Intestinal mucosal damage from nontropical sprue, radiation injury from the treatment for Whipple disease, and intestinal lymphectasia may cause magnesium deficiency. Steatorrhea also may cause or potentiate magnesium malabsorption through formation of nonabsorbable magnesium-lipid salts. The resection or bypass of the ileum for obesity, enteritis, or vascular infarction often results in magnesium deficiency. The magnesium content of intestinal fluids can be as high as 15 mEq/L.17 Thus, excessive magnesium loss may occur in acute or chronic diarrheal states, including regional enteritis and ulcerative colitis, or in patients with intestinal or biliary fistulas. The diarrhea encountered in proteincalorie malnutrition may be one mechanism for magnesium depletion. Acute hemorrhagic or edematous pancreatitis can also produce hypomagnesemia. The reason is unclear, but it may be the result of saponification of magnesium in the necrotic peripancreatic fat, which is a mechanism similar to that proposed for the hypocalcemia of acute pancreatitis. A selective defect in intestinal magnesium absorption (primary intestinal hypomagnesemia) is thought to be one cause of neonatal hypomagnesemia. RENAL LOSS Urinary loss of magnesium is a common contributing cause of magnesium deficiency.2,13,18 The renal magnesium transport is influenced by the filtered sodium and calcium load. Thus, excess urinary excretion of sodium and calcium will increase magnesium clearance and lead to urinary magnesium losses. The administration of excessive sodium in parenteral fluids is a common factor in the pathogenesis of hypomagnesemia in hospitalized patients. Most disorders causing an increase in the serum calcium concentration will also lead to renal magnesium wasting, presumably because of activation of the Ca2+-sensing receptor in the thick ascending limb of Henle, resulting in a decrease in the transepithelial voltage and a subsequent decrease in calcium and magnesium reabsorption.11,19 An exception is familial hypocalciuric hypercalcemia caused by an inactivating mutation of the Ca2+-sensing receptor, and lithium ingestion, in which urinary magnesium excretion is decreased. Commonly, diabetes mellitus is associated with magnesium deficiency and exemplifies osmotic-induced magnesium deficiency.20 The acidosis resulting from diabetic ketoacidosis, starvation, or alcoholism may also lead to renal magnesium wasting.2 The association of alcoholism with magnesium depletion is well known. Although the reason for magnesium deficiency is usually multifactorial in these patients, a rising blood alcohol level has been shown to cause renal magnesium loss. A wide spectrum of therapeutic agents may be associated with renal magnesium wasting.21 Hypomagnesemia is observed with the use of loop diuretics, such as furosemide and ethacrynic acid, and with therapy with the aminoglycosides, cisplatin, cyclosporine, amphotericin B, and pentamidine.22 Cardiac glycosides increase

renal magnesium loss in dogs, but the pathophysiologic role they play in human magnesium depletion is unclear. Many renal glomerular and tubular diseases are associated with renal magnesium wasting.2 There may be other accompanying tubular abnormalities, and a reduced glomerular filtration rate (GFR) may or may not be present. A rare form of hypomagnesemia (primary renal hypomagnesemia) is characterized by renal magnesium wasting in the absence of any other renal abnormality. This disorder may occur in children or in adults and may be familial. ENDOCRINE AND METABOLIC CAUSES Other disorders may also be associated with a disturbance in magnesium metabolism.2,23 Phosphate depletion, which is a common problem in hospitalized patients, causes renal magnesium wasting. An increase in urinary magnesium excretion and occasional hypomagnesemia is seen also in primary hyperpara-thyroidism, treated hypoparathyroidism, and hyperthyroidism.2,24 In these situations, hypercalciuria probably promotes the renal magnesium losses. In hypothyroidism, low intracellular magnesium content is found in association with high serum magnesium levels. A tendency to low serum magnesium concentrations in primary aldosteronism may be related to renal magnesium wasting resulting from volume expansion. Marked hypomagnesemia may occasionally accompany the hungry bone syndrome, which is a phase of rapid bone mineral accretion most commonly seen in patients with preexisting hyper-parathyroid bone disease who undergo successful neck surgery for hyperparathyroidism. Hypomagnesemia caused by excessive lactation is a rare occurrence. Massive transfusion with citrated blood may result in hypomagnesemia as well as hypocalcemia. MANIFESTATIONS OF MAGNESIUM DEFICIENCY SIGNS AND SYMPTOMS Pure magnesium deficiency in the absence of hormonal, gastrointestinal, or other metabolic disorders is uncommon. Signs and symptoms of magnesium deficiency are shown in Table 68-2. Because magnesium deficiency is usually associated with other disease states, the signs and symptoms of the primary disease process may mask those of magnesium deficiency.25 Neuromuscular hyperirritability may be the most common presenting problem in magnesium-depleted patients.2,23,26 Latent tetany may be present and may be elicited by positive Chvostek and Trousseau signs. Some patients may have spontaneous carpopedal spasm. Generalized seizures have also been reported. These neurologic signs in part may be caused by coexisting hypocalcemia. However, tetany has also been documented in magnesium-deficient patients with normal serum calcium levels.27 Other neurologic findings in hypomagnesemia include athetoid and choreiform movements, nystagmus, ataxia, and vertigo. Psychiatric abnormalities have been reported. Muscular complaints, such as tremors, fasciculations, muscle wasting, and weakness, may also occur.

TABLE 68-2. Common Clinical Manifestations of Magnesium Deficiency

Cardiac arrhythmias are a serious complication of magnesium depletion.2,28,29 and 30 Atrial and ventricular arrhythmias, including premature ventricular contractions, ventricular tachycardia, torsades de pointes, and ventricular fibrillation have been reported. Arrhythmias after myocardial infarction have also been associated with hypomagnesemia, and magnesium therapy has been demonstrated to decrease arrhythmias and improve survival in patients with acute myocardial infarction in some,31,32 but not all studies.33 Other electrocardiographic abnormalities include prolonged PR interval, U waves, and an increased QT interval. Associated hypokalemia and intracellular potassium depletion may contribute to these electrocardiographic abnormalities. Arrhythmias in magnesium-deficient patients may be resistant to the usual antiarrhythmic therapy and may respond only to magnesium therapy.29,34 An increased susceptibility of magnesium-deficient patients to digitalis intoxication has also been observed.28 Evidence also suggests that magnesium modulates vascular tone and may play a role in essential hypertension.34a,35 Other less recognized complaints include dysphagia caused by esophageal spasm.2 Also, anemia, characterized by short red blood cell survival, reticulocytosis, spherocytosis, microcytosis, and erythroid hyperplasia, has been described.2 Platelet-dependent thrombosis may occur.35a Magnesium therapy has been suggested as a beneficial form of therapy in renal calcium stone formation.36,37 In vitro studies have demonstrated that magnesium can prevent the precipitation and growth of calcium oxalate and calcium phosphate crystals.36 Experimental magnesium deficiency in rats results in nephrocalcinosis, and some patients with idiopathic renal magnesium wasting also have nephrocalcinosis.38 Assessment of magnesium status in renal stone formers, however, has demonstrated normal serum magnesium concentrations, muscle magnesium contents, and magnesium tolerance testing results.38 Gastrointestinal magnesium absorption was also found to be normal. Because of the influence of magnesium on crystal formation, alterations in renal magnesium excretion may be a potential cause of renal stone formation. Stone formers, however, have been found to have urinary magnesium excretion that is similar to a control population in most but not all studies.38,39 The ratio of urinary magnesium to urinary calcium has also been used as an index of the propensity to form stones. Magnesium excretion for corresponding calcium is lower in stone-forming patients, and a reduced ratio has been reported to be associated with an increased frequency of stones.36,40 Studies have assessed the benefit of magnesium supplementation in stone disease. One study of 56 patients who received 400 mg magnesium per day for 2 years noted a decrease in the rate of stone episodes from 0.8 stones per year to 0.003 stones per year.36 A control group also had a decrease in stone episodes of 0.5 to 0.22 stones per year. This was ascribed to the expected interval-free period after a stone episode. Another study suggested that oral magnesium supplements decrease intestinal calcium absorption in patients with hyperabsorptive hypercalciuric stone disease.37 Magnesium therapy, therefore, may be of benefit in reducing the recurrence rate in renal calcium stone disease. HYPOCALCEMIA An important clinical complication of magnesium deficiency is hypocalcemia.2 The hypocalcemia is caused in part by the suppressive effect of magnesium depletion on parathyroid hormone (PTH) secretion. Under normal circumstances, an abrupt decrease in magnesium causes an increase in PTH secretion, similar to that induced by a decrease in the serum calcium concentration. Magnesium affects PTH secretion by binding to the calcium-sensing receptor.20 Chronic severe hypomagnesemia, however, impairs the secretion of PTH. Serum immunoreactive PTH (IPTH) concentrations are undetectable or inappropriately low in most hypocalcemic patients with magnesium depletion, although some patients will have IPTH concentrations that are higher than normal.41,42 The ability of low magnesium concentrations to impair PTH secretionas opposed to PTH synthesisis well illustrated by the rapid increase in serum PTH levels after an intravenous injection of magnesium.41 An example of the acute response to magnesium administration in three magnesium-deficient patients with varying basal serum IPTH concentrations is shown in Figure 68-1.

FIGURE 68-1. The effect of an intravenous injection of 10 mEq magnesium on the serum concentrations of calcium, magnesium, and immunoreactive parathyroid hormone (IPTH) in hypocalcemic magnesium-deficient patients with undetectable (), normal (OO), or elevated (DD) levels of IPTH. Shaded areas represent the range of normal for each assay. The broken line for the IPTH assay represents the level of detectability. The magnesium injection resulted in a marked rise in parathyroid hormone secretion within 1 minute in all three patients.

Along with impaired PTH secretion in hypomagnesemia, patients may demonstrate end-organ resistance to PTH. Both skeletal and renal resistance to PTH have been demonstrated in magnesium-deficient patients by decreased rises in the serum calcium concentration and in urinary cyclic adenosine mono-phosphate (cAMP) excretion in response to exogenously administered PTH.42 Although the defect in PTH secretion is restored within minutes of magnesium administration, the skeletal resistance is not reversed as quickly. Figure 68-2 shows a 5-day course of magnesium therapy in a typical patient with hypocalcemia and magnesium depletion. In this patient, the serum calcium concentration does not rise appreciably until after ~2 days and is normalized only after 5 days of therapy. Serum PTH concentrations remain high until the serum calcium concentration returns to normal. The degree of PTH resistance seems to depend on the severity of the magnesium depletion. The fall in serum PTH in normal subjects experimentally depleted of magnesium correlated with the fall in red blood cell free magnesium.43

FIGURE 68-2. Changes in the serum calcium, magnesium, and immunoreactive parathyroid hormone (IPTH) concentrations during 5 days of magnesium therapy in a patient with magnesium deficiency. The horizontal dashed line indicates the upper limit of normal for the serum IPTH concentration. Arrows indicate times of intramuscular injection of 200 mg elemental magnesium. Asterisk indicates an initial intravenous injection of 300 mg elemental magnesium. Magnesium therapy resulted in a rise in the serum calcium, magnesium, and IPTH. The serum calcium concentration had returned to normal by the fourth or fifth day of therapy. (From Rude RK, Oldham SB, Singer FR. Functional hypopar-athyroidism and parathyroid hormone end-organ resistance in human magnesium deficiency. Clin Endocrinol 1976; 5:209.)

The serum concentration of 1,25-dihydroxyvitamin D [1,25(OH)2D] has been found to be low in most hypocalcemic magnesium-deficient patients.44 The 1,25(OH)2 D level remains low for several days after initiation of magnesium therapy, during which time the serum calcium concentration returns to normal. The reason for the low 1,25(OH)2 D level is unclear, but it does not appear to be related to either the amount of vitamin Dbinding protein in the plasma or the affinity of 1,25(OH)2 D for the binding protein. Magnesium-deficient patients with hypocalcemia are also resistant to the effects of pharmacologic doses of vitamin D as high as 50,000 units per day.45 The overall effect of magnesium deficiency on calcium metabolism is illustrated in Figure 68-3.

FIGURE 68-3. Disturbance of calcium metabolism during magnesium deficiency. Hypocalcemia is caused by a decrease in parathyroid hormone secretion as well as renal and skeletal resistance to the action of this hormone. Low serum concentrations of 1,25-dihydroxyvitamin D [1,25(OH)2D] result in reduced intestinal calcium absorption.

HYPOKALEMIA Hypokalemia is a frequent and important complication of magnesium deficiency.2,46,47 and 48 In one study, hypomagnesemia was present in 42% of patients with hypokalemia.47 Such patients also have a marked reduction in intracellular potassium content. Experimental human magnesium deficiency has demonstrated that the abnormalities in potassium metabolism are caused by a reduction in the ability of the kidney to conserve potassium.49 Potassium therapy alone will neither increase the intracellular potassium stores nor correct the renal potassium leak. Only simultaneous magnesium therapy will reverse the renal defect and normalize both the serum potassium concentration and the intracellular potassium content.50 DIAGNOSIS The serum magnesium concentration is the most commonly available index for assessing the presence of magnesium deficiency. The concentration of magnesium in the serum is commonly expressed in milliequivalents per liter or milligrams per deciliter. Milliequivalents of magnesium may be converted to milligrams by using the formula

A low serum concentration (<1.5 mEq/L) usually indicates magnesium deficiency. However, the magnesium concentration in the extracellular fluid compartment does not always accurately represent total body magnesium stores, because only ~1% of body magnesium is in the extracellular fluid. Magnesium deficiency with clinical manifestations may exist in the presence of a normal serum magnesium concentration.2,15,28,42,51 Ion-specific electrodes have become available for determining ionized magnesium in serum or plasma.52 Results suggest that these may provide a better index of magnesium status than the total serum magnesium concentration. Differences between ionized-magnesium analyzers must be resolved, however, before they are used in routine patient care.53 The mean intracellular peripheral lymphocyte magnesium content is reduced in patients with hypomagnesemia and in nor-momagnesemic patients at high risk for magnesium deficiency and, therefore, may be more precise in reflecting total body magnesium stores.51,54 Clinically, however, considerable overlap with normal values limits the usefulness of this index in an individual patient. Perhaps the most accurate assessment of magnesium deficiency in use is a magnesium tolerance test that is based on the amount of magnesium retained after a dose is administered parenterally55 (Table 68-3). In one study, the retention of a low dose of magnesium (0.2 mEq/kg or 2.4 mg/kg lean body weight) given intravenously over 4 hours was determined in hypo-magnesemic patients, normomagnesemic patients at high risk for having magnesium deficiency, and normal controls by measuring the 24-hour urine magnesium excretion. Hypomagnesemic patients retained a mean of 78% 7% of the magnesium load. Normomagnesemic

patients at risk for magnesium depletion retained 43% 5% of the magnesium load, whereas normal controls retained only 15% 5% of the administered magnesium. A suggested protocol for this magnesium tolerance test is outlined in Table 68-2. For example, a 70-kg patient underwent the magnesium tolerance test. The urinary magnesium/creatinine ratio was 0.02 before the start of the magnesium infusion. The patient was given 168 mg magnesium (2.4 mg/kg) intravenously over 4 hours. A 24-hour urine collection beginning with the infusion contained 120 mg magnesium and 1200 mg creatinine:

TABLE 68-3. Suggested Protocol for Clinical Use of Magnesium Tolerance Test

The retention of 43% of the intravenous magnesium load supports the diagnosis of magnesium deficiency. THERAPY Patients with symptoms or complications of magnesium deficiency should be treated with magnesium salts.2,56 Mild degrees of depletion may be corrected with oral magnesium supplementation, or even diet alone, if the process that caused the magnesium deficiency has resolved. Importantly, the hypocalcemia, hypokalemia, and other aspects of magnesium deficiency require treatment with magnesium. Treatment with vitamin D, calcium, or potassium alone usually is ineffective. Moderate to severe magnesium depletion is usually treated by the parenteral administration of magnesium, especially if there are continuing magnesium losses from the intestine or kidney. An effective treatment regimen is the intramuscular administration of 2 g MgSO4 7H2O (16.2 mEq magnesium) as a 50% solution every 8 hours. Because these injections may be painful, a continuous intravenous infusion (48 mEq/24 h) may be better tolerated. This treatment usually yields a serum magnesium concentration of 1.7 to 2.1 mEq/L. Therapy should be continued for at least 5 days, or until the hypocalcemia and hypokalemia are corrected. Note that in the patient whose treatment course is shown in Figure 68-2, the serum calcium concentration did not become normal until the fifth day of therapy. Continuing magnesium losses may require longer therapy to maintain magnesium stores. If the patient is unable to eat, a maintenance dose of ~8 mEq should be added daily to the intravenous fluids once repletion is accomplished. If the patient has a reduced GFR, the dose of magnesium given should be reduced (one-half) to avoid magnesium intoxication. Serum magnesium concentrations should be determined daily when a patient with azotemia is given magnesium-replacement therapy. Occasionally, some patients may require long-term oral magnesium supplementation. Oral magnesium salts in the form of oxide, sulfate, lactate, hydroxide, chloride, or glycerophosphate can be used. An initial dose of ~300 mg elemental magnesium usually is well tolerated. The dose may be increased to the desired amount until side effects such as diarrhea occur. If the magnesium is given in small doses, three to four times a day, the cathartic effect of magnesium can be minimized.

MAGNESIUM EXCESS
Magnesium intoxication is an uncommon clinical problem, although mild to moderate elevations in the serum magnesium concentration are not unusual. In acute care hospitals, up to 12% of admissions have been found to have hypermagnesemia, usually in association with chronic renal failure.14 ETIOLOGY OF HYPERMAGNESEMIA Orally administered magnesium has rarely been reported to result in hypermagnesemia because excess magnesium is normally excreted rapidly in the urine. However, enemas and cathartics that contain magnesium57 have been reported to cause hypermagnesemia in patients with apparently normal renal function. A reduction in the GFR is the most common factor related to elevation in the serum magnesium concentration.2,13 As the GFR falls, so does the clearance of magnesium. Hypermagnesemia is commonly seen when the GFR is <30 mL per minute. The presence of frank magnesium intoxication in these patients is usually associated with the administration of antacids, enemas, or parenterally administered fluids that contain magnesium. Dialy-sates high in magnesium content have also been reported to result in hypermagnesemia. Hypermagnesemia may be seen in acute renal failure. Contributing factors may include azotemia, rhabdomyolysis, acidosis, and magnesium administration. The parenteral administration of magnesium in patients with normal renal function is uncommonly associated with magnesium intoxication. The treatment of pregnancy-induced hypertension (toxemia of pregnancy) with magnesium salts results in serum magnesium concentrations of 4 to 7 mEq/L as a therapeutic end point. Excessive magnesium administration in these patients could lead to symptomatic magnesium intoxication. Hypermagnesemia also has been reported to occur in the neonates of mothers treated with parenteral magnesium. Other uncommon diseases associated with an increase in the serum concentration of magnesium include familial hypocalci-uric hypercalcemia, lithium ingestion, hypothyroidism, milk-alkali syndrome, viral hepatitis, and Addison disease. MANIFESTATIONS Symptoms of magnesium intoxication are usually not apparent until the serum magnesium concentration exceeds at least 4 mEq/L.2,58 One biochemical clue may be the presence of unexplained hypocalcemia, because hypermagnesemia causes a fall in the serum calcium concentration. This may be the result of decreased PTH secretion59; however, a fall in the serum calcium concentration also occurs in treated hypoparathyroid patients, suggesting a peripheral effect of magnesium.60 Hypermag-nesemia may directly decrease the biologic effect of PTH on its end-organ.61 Hypermagnesemia has been reported to increase serum calcitonin levels despite the associated decrease in serum calcium levels. Neuromuscular symptoms are the most common presenting problem of magnesium intoxication.2,62 Disappearance of the deep tendon reflexes occurs when the serum magnesium concentration reaches 4 to 6 mEq/L. Somnolence may be observed at concentrations of 4 to 7 mEq/L or above. Impaired respiration and apnea caused by paralysis of the voluntary muscles occurs at serum magnesium levels of 10 mEq/L or higher. Excessive magnesium decreases impulse transmission across the neuromuscular junction and, therefore, has a curare-like effect.63 A moderate elevation in the serum magnesium concentration to 4 to 5 mEq/L may result in bradycardia and mild supine and orthostatic hypotension. Higher concentrations (510 mEq/L) are associated with electrocardiographic changes including an increase in the PR interval and an increase in the QRS and QT intervals. Complete heart block as well as cardiac arrest in systole may occur at levels higher than 15 mEq/L.63 Other nonspecific manifestations of magnesium intoxication include nausea, vomiting, and cutaneous flushing at serum magnesium concentrations of 3 to 9 mEq/L. THERAPY

Most cases of magnesium intoxication can be prevented.2 It should be anticipated in any patient receiving parenteral magnesium, and, especially if there is a reduction in renal function, monitoring of the serum magnesium should be performed daily. Magnesium therapy should be discontinued in patients with mild to moderate elevations of the serum magnesium level. Excess magnesium will be excreted by the kidney, and any symptoms or signs of magnesium intoxication will resolve. Patients with severe magnesium intoxication may be treated with intravenous calcium; calcium antagonizes the toxic effects.58,62 The usual dose is an infusion of 100 to 200 mg elemental calcium over 5 to 10 minutes. If the patient is in renal failure, peritoneal dialysis or hemodialysis will rapidly and effectively lower the serum magnesium concentration. CHAPTER REFERENCES
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. Elin R. Assessment of magnesium status. Clin Chem 1987; 33:1965. Rude RK. Magnesium disorders. In: Kokko JP, Tannen RL, eds. Fluids and electrolytes, 3rd ed. Philadelphia: WB Saunders, 1996:421. Teuboi S, Nakagaki H, Ishiguro K, et al. Magnesium distribution in human bone. Calcif Tissue Int 1994; 54:34. Frausto da Silva JJR, Williams RJP. The biological chemistry of magnesium: phosphate metabolism. In: Frasto Da Silva JR, Williams RJP. The biological chemistry of the elements. Oxford: Oxford University Press, 1991:241. Birch NJ. Magnesium and the cell. New York: Academic Press, 1993. Fine KD, Santa Ana CA, Porter JL, et al. Intestinal absorption of magnesium from food and supplements. J Clin Invest 1991; 88:396. Kayne LH, Lee DBN. Intestinal magnesium absorption. Miner Electrolyte Metab 1993; 19:210. Krejs GJ, Nicar MJ, Zerwekh JE, et al. Effect of 1,25-dihydroxyvitamin D 3 on calcium and magnesium absorption in the healthy human jejunum and ileum. Am J Med 1983; 75:973. Hodgkinson A, Marshall DH, Nordin BEC. Vitamin D and magnesium absorption in man. Clin Sci 1979; 57:121. Quamme GH. Magnesium homeostasis and renal magnesium handling. Miner Electrolyte Metab 1993; 19:218. Quamme GA. Renal magnesium handling: new insights in understanding old problems. Kidney Int 1997; 52:1180. Rude RK, Bethune JE, Singer FR. Renal tubular maximum for magnesium in normal, hypoparathyroid, and hyperthyroid man. J Clin Endocrinol Metab 1980; 51:1425. Massry SG. Pharmacology of magnesium. Ann Rev Pharmacol Toxicol 1977; 17:67. Wong ET, Rude RK, Singer FR. A high prevalence of hypomagnesemia in hospitalized patients. Am J Clin Pathol 1983; 79:348. Ryzen E, Wagers PW, Singer FR, Rude RK. Magnesium deficiency in a medical ICU population. Crit Care Med 1985; 13:19. Dorup I. Magnesium and potassium deficiency. Acta Physiol Scand 1994; 150(Suppl 618):2. Thoren L. Magnesium deficiency in gastrointestinal fluid loss. Acta Chir Scand Suppl 1963; 306:5. Sutton RAL, Domrongkitchaiporn S. Abnormal renal magnesium handling. Miner Electrolyte Metab 1993; 19:232. Brown EM, Hebert SC. A cloned Ca 2+-sensing receptor: a mediator of direct effects of extracellular Ca 2+ on renal function. J Am Soc Nephrol 1995; 6:1530. Tosiello L. Hypomagnesemia and diabetes mellitus. A review of clinical implications. Arch Intern Med 1996; 156:1143. Shah GM, Kirschenbaum MA. Renal magnesium wasting associated with therapeutic agents. Miner Electrolyte Metab 1991; 17:58. Shah GM, Alvarado P, Kirschenbaum MA. Symptomatic hypocalcemia and hypomagnesemia with renal magnesium wasting associated with pentamidine therapy in a patient with AIDS. Am J Med 1990; 89:380. Flink E. Magnesium deficiency: etiology and clinical spectrum. Acta Med Scand 1981; 647:125. Disashi T, Iwaoka T, Inoue J, et al. Magnesium metabolism in hyperthy-roidism. Endocrinol J 1996; 43:97. Kingston ME, Al-Sibai MB, Skooge WC. Clinical manifestations of hypo-magnesemia. Crit Care Med 1986; 14:950. Whang R, Hampton EM, Whang DD. Magnesium homeostasis and clinical disorders of magnesium deficiency. Ann Pharmacol 1994; 28:220. Wacker WEC, Moore FD, Ulmer DD, Vallee BL. Normocalcemic magnesium deficiency tetany. JAMA 1962; 180:161. Cohen L, Kitzes R. Magnesium sulfate and digitalistoxic arrhythmias. JAMA 1983; 249:2808. Ebel H. Role of magnesium in cardiac disease. J Clin Chem Clin Biochem 1983; 21:249. Hollifield JW. Magnesium depletion, diuretics, and arrythmias. Am J Med 1987; 82(Suppl 3A):30. Rasmussen HS, McNair P, Norregard P, et al. Intravenous magnesium in acute myocardial infarction. Lancet 1986; 1:234. Woods KL, Fletcher S. Long-term intravenous magnesium sulfate in suspected acute myocardial infarction: the second Leicester Intravenous Intervention TRIAL (LIMIT-2). Lancet 1994; 343:816. ISIS-4. A randomized factorial trial assessing early oral captopril, oral mononitrate, and intravenous magnesium sulphate in 58,050 patients with suspected acute myocardial infarction. Lancet 1995; 345:669. Boriss MN, Papa L. Magnesium: a discussion of its role in the treatment of ventricular dysrhythmia. Crit Care Med 1988; 16:292.

34a. Yang ZW, Wang J, Zheng T, et al. Low Mg2t induces contraction of cerebral arteries: roles of cytokine and mitogen-activated protein kinases. Am J Physiol Heart Circ Physiol 2000; 279:H185. 35. Nadler JL, Goodsow S, Rude RK. Evidence that prostacyclin mediates the vascular action of magnesium in man. Hypertension 1987; 9:379. 35a. Shechter M, Merz CN, Rude RK, et al. Low intracellular magnesium levels promote platelet-dependent thrombosis in patients with coronary artery disease. Am Heart J 2000; 140:212. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. Johansson G, Backman U, Danielson BG, et al. Biochemical and clinical effects of the prophylactic treatment of renal calcium stones with magnesium hydroxide. J Urol 1980; 124:770. De Swart PMJR, Sokole EB, Wilmink JM. The interrelationship of calcium and magnesium absorption in idiopathic hypercalciuria and renal calcium stone disease. J Urol 1998; 159:669. Johansson G, Backman U, Danielson BG, et al. Magnesium metabolism in renal stone disease. Invest Urol 1980; 18:93. Evans RA, Forbes MA, Sutton RAL, Watson L. Urinary excretion of calcium and magnesium in patients with calcium-containing renal stones. Lancet 1967; 2:958. Oreopoulos DG, Soyannwo MA, McGeown MG. Magnesium-calcium ratio in urine of patients with renal stones. Lancet 1968; 2:420. Rude RK, Oldham SB, Sharp CF, Singer FR. Parathyroid hormone secretion in magnesium deficiency. J Clin Endocrinol Metab 1978; 47:800. Rude RK, Oldham SB, Singer FR. Functional hypoparathyroidism and parathyroid hormone end-organ resistance in human magnesium deficiency. Clin Endocrinol (Oxf) 1976; 5:209. Fatemi S, Ryzen E, Flores JF, et al. Effect of experimental human magnesium depletion on PTH secretion and 1,25(OH) 2 vitamin D metabolism. J Clin Endocrinol Metab 1991; 72:1067. Rude RK, Adams JS, Ryzen E, et al. Low serum concentrations of 1,25-dihydroxyvitamin D in human magnesium deficiency. J Clin Endocrinol Metab 1985; 61:933. Medalle R, Waterhouse C, Hahn TJ. Vitamin D resistance in magnesium deficiency. Am J Clin Nutr 1976; 29:854. Ryan MP. Interrelationships of magnesium and potassium homeostasis. 1993; 19:290. Whang R, Oei TO, Arkawa JK, et al. Predictors of clinical hypomagnesemia. Arch Intern Med 1984; 144:1794. Chernow B, Bamberger S, Stoiko M, et al. Hypomagnesemia in patients in postoperative intensive care. Chest 1989; 95:391. Shils ME. Experimental human magnesium depletion. Medicine 1969; 48:61. Cronin RE, Knockel JP. Magnesium deficiency. Adv Intern Med 1983; 28:509. Ryzen E, Elkayam U, Rude RK. Low blood mononuclear cell magnesium in intensive cardiac care unit patients. Am Heart J 1986; 111:475. Altura BT, Shirey TL, Young CC, et al. A new method for the rapid determination of ionized Mg in whole blood, serum and plasma. Methods Find Exp Clin Pharmacol 1992; 14:297. Cecco SA, Hristova EN, Rehak NN, Elin RJ. Clinically important inter-method differences for physiologically abnormal ionized magnesium results. Am J Clin Pathol 1997; 108:564. Ryzen E, Nelson TA, Rude RK. Low blood mononuclear cell magnesium content and hypocalcemia in normomagnesemic patients. West J Med 1987; 147:549. Hebert P, Mehta N, Wang J, et al. Functional magnesium deficiency in critically ill patients identified using a magnesium-loading test. Crit Care Med 1997; 25:749. McLean RM. Magnesium and its therapeutic uses: a review. Am J Med 1994; 96:63. Weber CA, Santiago RM. Hypermagnesemia: a potential complication during treatment of theophylline intoxication with oral activated charcoal and magnesium-containing cathartics. Chest 1989; 95:56. Mordes JP, Wacker WEC. Excess magnesium. Pharmacol Rev 1978; 29:273. Cholst IN, Steinberg SF, Tropper PJ, et al. The influence of hypermagnesemia on serum calcium and parathyroid hormone levels in human subjects. N Engl J Med 1984; 310:1221. Jones KH, Fourman P. Effects of infusions of magnesium and of calcium in parathyroid insufficiency. Clin Sci 1966; 30:139. Slatopolsky E, Mercado A, Morrison A, et al. Inhibitory effects of hypermagnesemia on the renal action of parathyroid hormone. J Clin Invest 1976; 58:1273. Fishman RA. Neurological aspects of magnesium metabolism. Arch Neurol 1965; 12:562. Morisaki H, Yamamoto S, Morita Y, et al. Hypermagnesemia-induced cardiopulmonary arrest before induction of anesthesia for emergency cesarean section. J Clin Anesth 2000; 12:224.

CHAPTER 69 NEPHROLITHIASIS Principles and Practice of Endocrinology and Metabolism

CHAPTER 69 NEPHROLITHIASIS
MURRAY J. FAVUS AND FREDRIC L. COE Clinical Features of Stone Disease Types of Kidney Stones Calcium Stones Uric Acid Stones Cystine Stones Struvite Stones Epidemiology of Stones Symptoms and Signs of Stones Attacks of Colic Silent Obstruction Crystalluria Hematuria Nephrocalcinosis Surgical Treatment of Stones Cystoscopy Operative Procedures Lithotripsy Medical Treatment of Stones Prevention Attempts at Dissolution Calcium Stones Pathogenetic Factors Disorders of Calcium Stone Formation Benign Familial Idiopathic Hypercalciuria Primary Hyperparathyroidism Hypocalcemia and Hypercalciuria Hyperoxaluria Renal Tubular Acidosis Hyperuricosuria and Low Urine PH Hypocitricuria Abnormalities of Glycoprotein Crystal Growth-Inhibitor Protein Renal Stones in Patients Without any Known Abnormalities Uric Acid Stones Pathogenesis Low Urine PH Hyperuricosuria Treatment Alkali Allopurinol Cystine Stones Pathogenesis Treatment Struvite Stones Pathogenesis Treatment Role of Antibiotics Surgery Lithotripsy Evaluation Protocol for Patients with Nephrolithiasis Chapter References

CLINICAL FEATURES OF STONE DISEASE

TYPES OF KIDNEY STONES Kidney stones usually consist of calcium salts, uric acid, cystine,or struvite (magnesium ammonium phosphate). Each of the four types of stones has its own natural history, pathogenesis, and forms of treatment, but all stones share a set of general clinical features. Kidney stones are formed on the surfaces of the renal papillae and represent a form of pathologic soft-tissue mineralization. The stones are composed of crystals embedded in a protein matrix; often they contain several types of crystals. When the stones escape the renal surface, they act as foreign bodies in the urinary tract, causing obstruction, pain, bleeding, and a tendency to infection. CALCIUM STONES Calcium forms salts with oxalic acid and with phosphate. Because these salts are extremely insoluble, they frequently crystallize to form renal stones. The most common stone, accounting for 66% to 70% of the total stones in unselected series,1 consists of mainly calcium oxalate monohydrate or dihydrate crystals. These common stones are usually small, 1 mm to 1 cm in their maximal dimension, black, radiodense, and hard. Approximately 5% to 15% of the crystals in a calcium oxalate stone are calcium phosphate in the form of apatite. Less commonly, in 1% to 5% of patients, the stones are mainly calcium phosphate, either apatite or brushite (calcium hydrogen phosphate). The formation of these stones is favored when the urine pH is abnormally elevated. Approximately 10% of calcium oxalate stones contain uric acid, and patients who form such stones pose special treatment problems. URIC ACID STONES Uric acid contains two dissociable protons; the undissociated uric acid is sparingly soluble in human urine a 37 C (98 2 mg/L).2 When the urine is unduly acidic, or when uric acid excretion is abnormally high, stones can be produced. Uric acid stones are radiolucent and often so large that they fill the renal pelvis and even extend out into the branches of the calyces to form a branching staghorn stone. They are white or red because of adsorption of the pigment urochrome from urine. CYSTINE STONES Cystine, an essential sulfur-containing amino acid, is poorly soluble (e.g., urine at 37 C dissolves only 300 mg/L). Patients with a hereditary tubular defect for the reabsorption of filtered cystine excrete so much that stones form.2a Cystine stones are moderately radiopaque because of the sulfur, and, like uric acid stones, they often grow to be large. The stones are lemon-yellow, hard, and covered with sparkling, shiny crystals that resemble crystallized sugar. STRUVITE STONES Magnesium, ammonium, and phosphate ions spontaneously combine to form an insoluble triple salt called struvite. This occurs in urine only because of infection with bacteria, usually Proteus sp,3 that possess the enzyme urease. This enzyme hydrolyzes urea to ammonia and carbon dioxide, and these stones are often called infection stones. Struvite stones grow to be extremely large and frequently form staghorns. They are gray to yellow, friable, and radiopaque. Because of their size and their intimate association with urinary tract infection, struvite stones are frequently damaging to the kidneys. EPIDEMIOLOGY OF STONES

Calcium stones occur more frequently in men than in women and in the wealthy than in the poor. Overall, the incidence in the United States is ~0.5%, and these patients are clustered geographically in warm regions. Uric acid stones develop mainly in men, especially in those predisposed to gout. Cystine stones occur equally among men and women. Struvite stones occur mainly in women. There are no obvious geographic or socioeconomic predictors of cystine or struvite stones. SYMPTOMS AND SIGNS OF STONES ATTACKS OF COLIC As long as they are firmly attached to the renal papillae, stones are painless; but once they break loose, they tend to pass into the renal pelvis and urinary tract, obstructing the flow of urine and causing an acute attack of pain, called renal colic. The pain is the result of sudden dilation of the urinary tract from increased pressure. Typically, the pain begins suddenly on one side. It may start in the flank, lower back, or lower anterior abdomen. Although it is called colic, it increases steadily in intensity over about an hour and remains constant as long as the stone remains in one place. As the stone moves downward, the pain shifts with it. Stones at the junction of the ureter and the bladder cause urinary frequency and dysuria, often incorrectly ascribed to urinary tract infection. Pain from low-lying stones radiates to the ipsilateral testicle or vulva, mimicking genital disease. If the stone passes, the pain ends rapidly. Colic causes nausea and vomiting, and diarrhea is common. These latter symptoms may falsely suggest intestinal obstruction, biliary colic, or acute diverticulitis. Renal colic is thought to be among the worst pains that can be encountered, and its management usually requires analgesic medications, including narcotics. Hydration with intravenous fluid is needed if vomiting is protracted. Once an attack ends, by stone removal or passage, little or no further treatment is needed. SILENT OBSTRUCTION Large stones, even staghorn stones, may occlude the ureteropelvic junction so slowly that no pain occurs, even though hydronephrosis may damage or destroy the kidney. A mass in the flank from a hydronephrotic kidney, recurrent infection, unexplained azotemia (when obstruction is bilateral or involves a solitary functioning kidney), or the passage of stone fragments may call the problem to attention. Frequently, however, the stone is found in the course of an unrelated radiographic procedure. CRYSTALLURIA Calcium oxalate or phosphate crystalluria, or uric acid crystals, commonly cause episodes of colic that are unassociated with a stone but resemble a stone attack. Hematuria is common, although transient. The urine contains large amounts of crystals, but only briefly after the attack. HEMATURIA Crystals and stones are common causes of isolated hematuria, abnormal proteinuria, or infection. The other main causes are malignant or benign tumors of the kidneys and urinary tract, infections such as tuberculosis, cysts of the kidneys, and bleeding from the prostatic bed. Hematuria even occurs from stones that are anchored to the renal papillae and otherwise are asymptomatic. NEPHROCALCINOSIS Numerous small papillary stones that are seen on an abdominal radiograph are termed nephrocalcinosis. A better term is papillary nephrocalcinosis, which distinguishes this localized condition from generalized renal calcification caused by hypercalcemic states. Because of its dramatic radiographic appearance, nephrocalcinosis suggests unusual diseases but usually is a result of the routine causes of calcium stone disease and requires no special evaluation.4 Medullary sponge kidneys, a hereditary cystic renal disease, is associated with nephrocalcinosis,5 probably because the ectasia of terminal collecting ducts causes stasis and favors the deposition of small crystals in the renal papillary areas. SURGICAL TREATMENT OF STONES Stones need surgical intervention if they cause severe pain, bleeding, obstruction, or serious infection. Most attacks of colic can be managed medically for a few days to determine if an obstructing stone is gradually moving downward. Based on stone size alone, the likelihood of spontaneous passage for stones >6 mm is ~25%; for stones from 4 to 6 mm it is ~60%, and for stones <4 mm it is ~90%. Many stones in the kidneys do not cause symptoms and are not obstructing. Also, most episodes of colic terminate by spontaneous stone passage. When intervention is required, cystoscopy, surgery, or shock-wave lithotripsy are used. The frequency of these procedures varies with the cause of the calcium oxalate stone. Patients with hyperuricosuria and calcium oxalate stones are more likely to require stone-removal procedures or nephrectomy. CYSTOSCOPY The indication for cystoscopy is removal of a stone that is close to or at the ureterovesical junction. Another reason for cystoscopy is to perform retrograde pyelography to assess the patency of the urinary tract. Sometimes, a stone is in the bladder but cannot pass because of a narrow bladder outlet or because it gradually has grown too large. In these locations, the stone can be fragmented and removed. OPERATIVE PROCEDURES Stones lodged in the ureter may be removed by ureterolithotomy. Stones in the renal pelvis have been removed by pyelolithotomy, which is a major surgical procedure involving a flank incision and prolonged recovery. Most of these stones are now removed by lithotripsy. LITHOTRIPSY A form of stone therapy that has revolutionized the treatment of many stones is shock-wave lithotripsy. Shock waves reduce the stones to small fragments that are then passed with relative ease. In one form of treatment, the patient is lowered into a tank of water, either under general anesthesia or spinal anesthesia, and positioned so that the stone is centered at the secondary focus of a parabolic reflector. An electric spark generator is positioned such that the shock wave it produces in the water is at the primary focus of the reflector, and electrical discharges are produced. The mechanism is synchronized with the standard electrocardiogram so that the shock waves occur immediately after the QRS complex. The shock waves pass through the stone and disrupt it. Anywhere from 500 to 1500 or more discharges are needed to fragment stones of moderate size. The stone fragments move into the ureter and are subsequently passed. The procedure works well with stones in the renal pelvis or calyces. Kidney stones smaller than 2 cm but larger than 0.5 cm in diameter are best treated with lithotripsy. Stones greater than 2 cm or those larger than 1 cm in the lower poles of the kidney are best treated with percutaneous nephrolithotomy. Ureterolithotomy is used for removal of stones of any size located in the lower ureter below the pelvic brim. Transurethral ureterolithotripsy (TUL) and laser technology disintegrate ureteral stones and have largely supplanted the open surgical procedures. Most stones (80% to 90%) subjected to lithotripsy are passed, and there is only limited information on the fate of retention of small fragments or the recurrence of new stones.6 Infrequent but serious complications include bleeding into the renal parenchyma, spinal cord damage, and paraplegia.7 MEDICAL TREATMENT OF STONES PREVENTION The main goals of the internist are to identify the metabolic disturbance that causes stone formation in a given patient and, through specific treatment, to prevent the development of new stones. The search may involve considerable effort because most of the abnormalities are found in the chemical analysis of the urine; however, such studies are not widely available in hospital clinical chemistry laboratories. ATTEMPTS AT DISSOLUTION Uric acid stones and some cystine stones may be dissolved by altering the urine chemistry. Calcium stones and struvite stones are rarely dissolved, despite vigorous treatment.

CALCIUM STONES
PATHOGENETIC FACTORS The factors critical to calcium stone formation are supersaturation and nucleation. Supersaturation of a solution occurs when a solution contains a salt (e.g., calcium oxalate) at a concentration that exceeds its solubility.1,8 When such a solution comes into contact with a precipitated salt with which it is already supersaturated, the solid phase serves as a nucleation center for the aggregation and precipitation of the excess salt in solution. Supersaturation is defined and measured in terms of the particular salts present in concentrations exceeding their solubilities. A solution may be supersaturated with several salts, each to a different extent. The magnitude of the supersaturation provides the driving force for crystallization of the solid phase (e.g., stone formation from urine). The supersaturation of urine with calcium oxalate and calcium phosphate phases, such as brushite or apatite, is determined by (1) the concentrations of calcium, oxalate, and phosphate; (2) the concentrations of calcium ligands such as citrate (which chelates calcium, preventing the precipitation of calcium oxalate or phosphate salts); and (3) the pH (determines relative concentrations of hydrogen phosphate and phosphates, which form brushite and apatite, respectively). The actual measurement of supersaturation is difficult but useful, because the elevation of supersaturation corresponds to stone composition.9 Clinically, the excretion rates and concentrations of calcium and oxalate and the urine pH are measured to detect patients with hypercalciuria, hyperoxaluria, or abnormally alkaline urine. These measurements frequently represent the diagnostic end points of patient evaluation for supersaturation. Important to the formation of renal stones is a surface (nucleation center) within the supersaturated solution that can promote aggregation of ions to form crystals. Given a foreign surface to which ions can adhere to form clusters, stones can be formed, even if supersaturation is mild; therefore, the probability of stone formation can be increased simply by a foreign nucleation surface. Clinical examples of foreign, or heterogeneous, nucleation include stone crystallization on an injured urothelial surface and calcium oxalate stones grown on a crystal of uric acid induced by low urine pH and increased uric acid excretion. The uric acid crystals offer a nucleation surface on which calcium oxalate overgrowth is likely to occur.10 When the foreign surface strongly resembles the one to be formed, the ions aggregate on the surface in a strongly oriented manner, reflecting the alignment of the overgrowing and preformed crystals. This process is called epitaxial growth, and although it has been proposed as a mechanism for uric acid nucleation of calcium oxalate,10,11 it is unlikely to be a common occurrence. Urine contains substances that inhibit the growth and formation of calcium oxalate and phosphate stones. Citrate, which lowers the supersaturation of calcium salts by chelating calcium, can also reduce the growth of calcium oxalate and phosphate crystals.12,13 An anionic glycoprotein inhibi- tor, glycoprotein crystal growth-inhibitor (GCI) protein,14,15 which is in urine, dramatically inhibits calcium oxalate growth. Patients with stones often have low citrate levels and excrete an abnormal form of GCI.14 Urinary inorganic pyrophosphate inhibits calcium phosphate crystal growth.12,16 Magnesium may inhibit calcium oxalate crystal growth because it forms a soluble complex with oxalate. DISORDERS OF CALCIUM STONE FORMATION BENIGN FAMILIAL IDIOPATHIC HYPERCALCIURIA The abnormality that is found in 50% to 75% of patients with calcium stones is a normocalcemic hypercalciuria that is unexplained by any known systemic disease, such as sarcoidosis, hyperthy-roidism, rapidly progressive bone disease, vitamin D intoxication, immobilization, Paget disease, malignancy, or glucocorticoid excess.1,17 Called idiopathic hypercalciuria, it is more appropriately named benign familial hypercalciuria because it is found in the families of affected patients in a pattern that is compatible with an autosomal-dominant inheritance18,19,19a (Fig. 69-1). Familial hypercalciuria affects 2% to 4% of normal adults and accounts for ~40% of stone formers. In ~80% to 90% of patients, hypercalciuria is silent. The distribution of values for calcium excretion in the population is not bimodal, as expected for an inherited disorder, but non-Gaussian, with a long tail of high values. Therefore, the upper limit of normal for calcium excretion and the diagnosis of idiopathic hypercalciuria are arbitrary. An out-patient 24-hour urine collection is obtained; the patient eats ad libitum, including dairy products. Many investigators have determined calcium excretion under various dietary conditions that are important for studies of the pathogenesis of the disease. However, because changes in dietary calcium intake or the use of artificial diets change the urine chemistry, the authors believe that a free-choice diet is preferable when the diagnosis is in doubt. Hypercalciuria is determined by a 24-hour calcium excretion of above 300 mg for men and 250 mg for women, or either 4 mg/kg of body weight or 140 mg/g of urinary creatinine in either sex. Hypercalciuria increases urine supersaturation with calcium oxalate or brushite.20,21

FIGURE 69-1. Family pedigrees of nine probands with idiopathic hypercalciuria. Marginal hypercalciuria was present in four siblings, one each in families 1, 2, 3, and 5; in the mother of proband 4; in two aunts and one niece in family 5; and in one nephew in family 3. Altogether, hypercalciuria occurred in 11 of 24 siblings, 7 of 16 offspring, and 1 of 3 parents of the probands. (Squaresindicate males; circles indicate females; solid circles and squares, family members with hypercalciuria; S, stone disease; *, children younger than 20 years of age; arrows,probands; interrupted circles and squares,relatives who were not studied but who were included to complete the pedigree.) (From Coe FL, Parks JH, Moore ES. Familial idiopathic hypercalciuria. N Engl J Med 1979; 300:337.)

The excessive urinary calcium excretion arises from increased intestinal calcium absorption.1 The pathogenesis of benign familial hypercalciuria is probably heterogeneous, with some patients overproducing 1,25-dihydroxyvitamin D3 [1,25(OH)2D3], the biologically active form of vitamin D that stimulates intestinal calcium absorption.22 Target organ hyper-responsiveness to 1,25(OH)2D3 may also be present, and elevated and normal serum 1,25(OH)2D3 levels have been reported.1 The apparent discrepancies of 1,25(OH)2D3 measurements among familial hypercalciuric patients may reflect their sensitivity to dietary calcium intake.23 The stimulus for increased 1,25(OH)2D3 production remains unknown, but low serum phosphorus or excess parathyroid hormone is unlikely. If dietary calcium is reduced severely, calcium balance becomes more negative than in normal people given an equivalent diet1; urine calcium excretion remains high in many of these patients24 (Fig. 69-2). This observation suggests that, under these conditions, calcium is being mobilized from bone. Because of this adverse response to a low calcium diet, which also occurs in normal people who are made hypercalciuric but not hypercalcemic by exogenous 1,25(OH)2D3, a low calcium diet is not advisable for these patients.25

FIGURE 69-2. Calcium intake and excretion (CaE) (A) and serum parathyroid hormone (PTH) and 1,25(OH)2D3 (B) values in patients and normal subjects. Urine calcium excretion (UCa) and intake (CaI) are during low calcium diet. Serum values for 1,25(OH)2D3 and PTH during low calcium diet (Ca intake of 2 mg/kg body

weight [BW] per day for 10 days, solid circles) and free-choice diet (open circles) are both shown. Mean calcium intake of normal subjects (2.29 0.15 SEM in mg/kg BW of calcium in 24 hours) and patients (2.31 0.05) did not differ. Mean excretion rates during low calcium diet (1.18 0.11 vs. 2.87 0.11, p <.001) for normal subjects and patients, respectively, and values of intake minus excretion (CaI CaE, 1.14 0.12 vs. 0.58 0.11, p <.001) differed significantly from each other. Calcium excretion in normal subjects (1.78 0.16) and patients (4.38 0.15) during free-choice diet (not shown) differed from each other (p <.001) and from values during low calcium diet (p <.01, both normal subjects and patients). Normal subjects and patients are shown in order of ascending mean calcium excretion; numbers are for identification of individual normals and patients. [From Coe FL, Favus MJ, Crockett T, et al. Effects of low-calcium diet on urine calcium excretion, parathyroid function, and serum 1,25(OH)2D levels in patients with idiopathic hypercalciuria and normal patients. Am J Med 1982; 72:25.]

Therapies to prevent stone recurrence focus on the importance of the urine chemistry, specifically, the level of supersaturation with calcium oxalate and calcium phosphate (Table 69-1). Urine calcium is the major predictor of supersaturation and stone recurrence. Thiazide diuretic agents lower calcium excretion, diminish supersaturation with calcium oxalate monohydrate, and decrease new stone formation.17 These results have been documented in a placebo-controlled trial.26 When pre-treatment and posttreatment stone formation rates are compared, thiazide may reduce new stone formation by 90%.17 Thiazide reduces urine calcium by increasing distal tubular cal- cium reabsorption. The higher the pretreatment urine calcium level, the greater is the reduction of the urine calcium concentration with thiazide. Urine calcium values may decline by 200 mg per day in patients with idiopathic hypercalciuria. The maximally effective dose is determined by the fall in urine calcium. Although the dose of thiazide varies with the specific agent, a starting dose can be equivalent to the usual dose for the treatment of edema or hypertension. The effect of thiazide on calcium excretion should be evaluated after 6 weeks of treatment by analysis of calcium in a 24-hour urine collection. Possible adverse effects of thiazides include hypokalemia, increased serum cholesterol levels, and reduced glucose tolerance. Interestingly, thiazides may increase bone mass.27 Amiloride can reduce the potassium wasting, and it may increase the hypocalciuric effect.

TABLE 69-1. Urine Uric Acid Saturation in Calcium Oxalate Renal Stone Formers

Other treatments include orthophosphate, which increases the urine concentration of the crystal growth inhibitor, pyrophosphate. Magnesium therapy may reduce urine supersaturation of calcium oxalate, but this requires further study. Cellulose phosphate is a nonabsorbable substance that binds dietary calcium in the lumen of the intestinal tract, preventing absorption. Like dietary calcium restriction, cellulose phosphate may not be suitable for patients with idiopathic hypercalciuria who mobilize calcium from bone during low calcium intake.25 PRIMARY HYPERPARATHYROIDISM Approximately 5% of patients with calcium stones have primary hyperparathyroidism1,17 (see Chap. 58). The mechanism for stone production appears to be hypercalciuria, which is presumed to arise from a high filtered load of calcium because of the hypercalcemia. The increased urinary calcium excretion is associated with increased intestinal calcium absorption caused by increased serum 1,25(OH)2D3 levels.28 However, patients with primary hyperparathyroidism and stones have 1,25(OH)2D3 levels that are no different from those of patients without stones. The treatment for patients with stone disease and primary hyperparathy-roidism is parathyroidectomy. HYPOCALCEMIA AND HYPERCALCIURIA A phenotype of familial hypocalcemia with hypercalciuria has been described in six kindred.29 Treatment of the hypocalcemia with vitamin D and calcium resulted in nephrocalcinosis and renal impairment. Mutations of the calcium-sensing receptor gene lead to a gain of function in which lower levels of serum calcium are required to reduce parathyroid hormone secretion. HYPEROXALURIA Excessive urinary oxalate excretion may increase supersaturation, thereby increasing the risk for stone formation. The pathogenesis of hyperoxaluria is diverse and may be genetic or acquired.29a,29b It may be separated into two categories: metabolic overproduction of oxalate (hereditary, types I and II; ethylene glycol ingestion; methoxyflurane excess; and pyridoxine deficiency, which occurs in the experimental animal but perhaps not in humans), and gastrointestinal overabsorption of oxalate (oxalate overingestion; ileal resection as in Crohn disease; malabsorption syndromes such as celiac sprue or pancreatic insufficiency; smallbowel bypass surgery; and cellulose phosphate ingestion). Enteric Hyperoxaluria. Patients with malabsorption syndrome and a functioning colon are prone to increased absorption and urine excretion of oxalate and to severe stone disease.30 Usual clinical settings for enteric hyperoxaluria are intestinal bypass surgery for obesity, Crohn disease, ileal resection for any cause, and other forms of malabsorption.30 Steatorrhea is almost invariably present.31 The hyperabsorption of oxalate occurs in the colon; hence, patients with an ileostomy do not develop enteric hyperoxaluria. Urine oxalate excretion, normally 20 to 40 mg per day, frequently is 60 to 80 mg per day or more. The mechanism of oxalate overabsorption appears to be increased colon permeability to oxalate, resulting from delivery of undigested mediumchain fatty acids and bile salts to this site.32 The colon appears to possess a size- and charge-selective barrier to neutral sugars that reduces the permeability of oxalate below that predicted from the molecule's molecular radii. Ricinoleate, a medium-chain fatty acid normally absorbed in the small bowel, increases the absorption of oxalate and other small molecules by increasing the epithelial permeability of the colon. Hyperoxaluria greatly raises urine calcium oxalate supersaturation,33 causing stones. Tubulointerstitial nephropathy also is produced by the crystals; hence, progressive azotemia and defects of tubule function, such as renal tubular acidosis (RTA), may occur.34 The effects of hyperoxaluria are exaggerated by the loss of urine citrate that occurs in small-bowel malabsorption states; by low urine volumes caused by fluid loss from diarrhea; and by an acid urine pH from intestinal bicarbonate loss, which promotes uric acid crystallization and the formation of both uric acid and calcium oxalate stones. The treatment of enteric hyperoxaluria is directed mainly at reducing intestinal oxalate absorption. If intestinal bypass surgery has been performed, reanastomosis is indicated. A low oxalate diet is practical and useful to the extent that variety in the diet and good nutrition can be maintained. A low-fat diet helps by reducing the colonic delivery of undigested fatty acids; the dietary fat content should be as low as is consistent with good nutrition. Because these patients often have marginally adequate nutrition, each patient requires an individual assessment of the role of dietary modification. An additional therapeutic approach is oral calcium, which can reduce oxalate absorption, probably by precipitating calcium oxalate in the intestinal lumen.35 The dose of calcium preparation should be commenced at 1 g elemental calcium in four divided doses and gradually increased to 2 to 4 g in four divided doses as needed, depending on the urine oxalate excretion rate. Finally, the ion-exchange resin, cholestyramine, can be used in a dosage of 4 to 16 g in four daily divided doses. This resin binds oxalate, preventing its absorption, and also binds bile salts so they cannot injure the colonic epithelial surface.36 The doses of calcium and cholestyramine depend on patient tolerance. Calcium may cause constipation and abdominal discomfort, which may limit its usefulness. Cholestyramine has an unpleasant taste and causes abdominal pain and diarrhea in some patients. Urine calcium and oxalate levels must be monitored during treatment because calcium loading may raise urine calcium without lowering urine oxalate, thus increasing the risk of stones. Cholestyramine may interfere with the absorption of drugs, and it often causes a vitamin K depletion that must be offset by monthly supplements.

Low urine citrate and pH are best treated, if present, with oral citrate or bicarbonate, 2 to 4 mEq/kg, in four divided doses daily. Primary Hyperoxaluria. Rarely, stone formers have hyperoxaluria caused by one of several hereditary diseases.1,34,37 The urinary oxalate excretion is frequently above 100 mg per day, leading to supersaturation. Consequently, stones are numerous, tubulointerstitial nephropathy is progressive, and chronic renal failure may develop. Stones usually begin in childhood, and in most people, renal failure develops by 20 to 30 years of age. Unfortunately, the treatment is highly unsatisfactory. Pyridoxine, 400 mg per day, may reduce oxalate production. High fluid intake, above 3 L per day, reduces supersaturation. Orthophosphate, in a dosage of 1 to 2 g phosphate in four divided doses daily, may lower urine calcium and reduce crystallization.34 Despite treatment, stone disease and renal failure are frequently progressive. The missing enzyme in type I disease can be restored by liver transplantation.38 Dietary Hyperoxaluria. Some patients excrete excessive amounts of oxalate, in the range of 50 to 70 mg per day, because of dietary excesses of foods high in oxalate, such as spinach, rhubarb, nuts, chocolate, pepper, and tea.1 The extent of oxaluria reflects the dietary load, which is usually variable. Renal damage usually does not occur. The treatment is dietary modification. RENAL TUBULAR ACIDOSIS Few patients with calcium stones, perhaps 1%, develop stones as a result of disordered renal acidification. The stones are composed mainly of calcium phosphate, such as brushite or apatite. Many patients who have RTA acquired it from previous stone disease or other renal disease. A few patients have the hereditary form of RTA. Whether acquired or hereditary, the mechanisms of stone formation are the same. Renal tubular acidosis conventionally is divided into five types: type 1, gradient; type 2, proximal; type 3, mixed 1 and 2; type 4, acquired, associated with hyperkalemia and an acid urine; and type 5, resulting from chronic renal failure.39 Of these, it is type 1 that is associated with stones. In type 1 RTA, the renal tubules lose their capacity to lower urine pH normally, causing inadequate net acid excretion and systemic metabolic acidosis. The acidosis causes hypercalciuria and reduces urine citrate excretion.39,40 The alkaline urine pH increases the concentration of phosphate ion and of monohydrogen phosphate; consequently, apatite and brushite become further supersaturated.41 The result is calcium phosphate crystalluria, nephrocalcinosis, and stones. The stones frequently are large and numerous. The diagnosis of RTA is suggested in any patient who repeatedly forms calcium phosphate stones. The pH of the 24-hour urine is elevated above the normal mean of 6.1. It does not fall below 5.5 in the 6 hours after administration of oral ammonium chloride, 1.9 mEq/kg, given as a single dose in the morning in the postabsorptive state (available as 500-mg tablets; 1 g ammonium chloride = 18.7 mEq chloride). The presence or absence of metabolic acidosis does not critically affect the diagnosis because some stone formers who cannot lower urine pH properly nevertheless can excrete acid in the form of ammonium at a rate sufficient to prevent systemic acidosis. These patients have incomplete RTA. Because hereditary RTA is a dominant trait, family members should be screened for stone disease or metabolic acidosis (serum bicarbonate <24 mEq/L). The chronic metabolic acidosis of hereditary RTA may cause stunted growth and bone disease.42 The reduced growth rate can be restored by alkali treatment. Bone disease is not common, but when it occurs, it takes the form of rickets in children and osteomalacia in adults. In children, the acidosis is markedly worsened by intercurrent illness; they may develop severe dyspnea as a respiratory compensation to the systemic acidosis. The presentation may resemble pneumonia. The treatment of RTA is with alkali, 2 to 4 mEq/kg in four daily divided doses, in the form of bicarbonate or citrate. The alkali raises urine citrate levels and lowers those of urine cal- cium; the results of its use are excellent. 43 Patients with incomplete RTA may have acquired RTA as well as benign familial hypercalciuria. Along with alkali, thiazides may be required to further lower urinary calcium excretion. HYPERURICOSURIA AND LOW URINE PH Calcium Oxalate Uric Acid Stones. Approximately 10% of patients form stones that contain variable amounts of uric acid admixed with calcium oxalate.1 The urine of such patients most often is abnormally supersaturated with calcium oxalate and with undissociated uric acid. In contrast, urine from patients with calcium oxalate stones is not supersaturated with uric acid, and urine from patients with pure uric acid stones is not overly supersaturated with calcium oxalate. The supposedly dual supersaturation with calcium oxalate and uric acid leads to crystallization from both phases and the development of mixed stones. Because the urine is supersaturated with calcium oxalate monohydrate, uric acid crystals become a favorable surface for promoting calcium oxalate nucleation. Whenever the uric acid forms crystals, calcium oxalate formation is fostered. The actual stones show calcium oxalate and uric acid aligned along one dimension compatible with heterogeneous nucleation. The treatment of this type of mixed stone is directed both at the hyperuricosuria (above 800 mg for men or 750 mg for women) and at the urine pH that is below normal. Both the uri-cosuria and the low pH are caused by excessive dietary intake of meat, fish, and poultry44 and often can be corrected by dietary counseling. When dietary treatment is unsuccessful, allopurinol is an alternative.17 Calcium Oxalate Stones. The pattern of uric acid supersaturation combined with modest calcium oxalate supersaturation is also encountered in some patients whose stones predominantly are calcium oxalate. Uric acid may play a critical role in nucleation, but quantitatively it may be only a trivial component of the eventual mass of the stone. It appears that uric acid supersaturation fosters uric acid crystallization, which in turn promotes heterogeneous nucleation of calcium oxalate monohydrate.1 As in patients with mixed stones, the hyperuricosuria and mildly acidic urine pH found in patients with calcium oxalate stones can be treated by either changes in diet or allopu-rinol. Only the latter has been studied clinically.17 HYPOCITRICURIA Secondary Hypocitricuria. Low excretion of urinary citrate occurs whenever there is systemic acidosis. RTA and malab-sorption have already been mentioned in connection with hypocitricuria. Other causes are ileostomy, chronic colitis with loss of bicarbonate, and potassium depletion. In all these conditions, the replacement of alkali, either as bicarbonate or citrate, increases urinary citrate and lowers the urine calcium ion concentration. Angiotensin II converting-enzyme inhibitors decrease urine citrate, but the doses commonly used have not been associated with calcium nephrolithiasis.45 Primary Hypocitricuria. In the absence of systemic disease, some patients with stones excrete less citrate than normal.46 The term primary hypocitricuria is not yet used routinely for this condition, but it seems appropriate. Generally, women excrete more citrate than men. Stone formers excrete less citrate than normal people of the same sex. The disparity in urinary citrate excretion between normal people and stone formers is more dramatic in women than in men. The mechanism of low citrate excretion appears to be enhanced reabsorption by the renal tubules. Reasonable limits for normal urine citrate excretion for both sexes are given in Table 69-2. The normal range depends on local populations and must be determined by each laboratory. Low urinary citrate levels may occur in isolation or in association with hypercalciuria, hyperuricosuria, hyperoxaluria, or bowel disease. Whatever the cause, it should be treated by either potassium citrate or by potassium bicarbonate. Potassium lowers urine calcium and increases calcium balance.47 The initial dosage of alkali, in any form, should be 1 to 2 mEq/kg per day, in two or three divided doses, and adjusted as needed, according to the response of urine citrate excretion.

TABLE 69-2. Selected Urinary Measurements

ABNORMALITIES OF GLYCOPROTEIN CRYSTAL GROWTH-INHIBITOR PROTEIN Gross measurements of the degree to which urine inhibits the growth of calcium oxalate monohydrate have repeatedly shown less inhibition by the urine of patients with stones.48 Whether the reduction was caused by fewer inhibitor molecules or by abnormal molecules of poor inhibitory activity could not be determined, because the actual molecules were unknown. Now, one such inhibitor substance, the GCI protein, has been purified from urine and from a human renal cell line49; its properties in normal people and stone formers have been partly determined. Normal GCI contains g-carboxyglutamic acid (GLA), two to three residues per mole.14 It forms strong films at the airwater interface and has an apparent affinity for the calcium oxalate crystal surface of 100 nmol/L.14 It also exhibits microheterogeneity, as shown in four distinct peaks eluting from ion-exchange columns. All four peaks have the same molecular mass of 14,000 daltons and the same affinity, but they have decreasing amounts of GLA and decreasing airwater interfacial stability.14 When compared with that of normal people, the GCI from patients with calcium oxalate stones lacks GLA altogether and forms weak films at the airwater interface.14 More important, the third and fourth peaks of GCI, which comprise half of the total GCI in urine, have a much lower affinity for the crystal surface (1 mol/L), making it weaker as an inhibitor. The obvious implication of these results is that some inherited or acquired defect of GCI may diminish its effectiveness and expose some people to a greater than normal risk of calcium oxalate crystal formation. In principle, defective GCI permits stone formation to occur even with normal supersaturation, so that specific disorders of urine minerals need not occur in such patients. The validity of this hypothesis remains to be tested. RENAL STONES IN PATIENTS WITHOUT ANY KNOWN ABNORMALITIES Every stone program reports that some patients (~20%) have no disorder explaining their stones. The recent recognition of low urinary citrate and the new discovery of abnormal GCI proteins inevitably will decrease the number of these patients. The conventional criterion for hypercalciuria is the top fifth percentile of a normal population. However, lower urinary calcium excretion than these rates may confer a risk of stones. The use of more refined criteria and these newer measurements should decrease the percentage of patients in whom stone disease is not associated with any recognized abnormalities. Thiazide treatment of patients with calcium oxalate stones who have no discernible metabolic disorder reduces the rate of new stone formation by reducing urine calcium excretion and calcium oxalate supersaturation. Like patients with idiopathic hypercalciuria, normocalciuric stone formers may lose bone mass when fed a diet severely restricted in calcium. Neutral orthophosphate may also reduce the rate of calcium oxalate stone formation, but this regimen has not been adequately tested.

URIC ACID STONES

PATHOGENESIS LOW URINE PH The main cause of increases in uric acid supersaturation is a low urine pH (Fig. 69-3). The concentration of undissociated uric acid, which is the only form that crystallizes under most circumstances, is controlled by pH.2 Two other factors, low urine volume and hyperuricosuria, are less important. Reasons for a low urine pH include a hereditary predisposition (as in patients with clinical gout and familial uric acid nephrolithiasis), ileostomy, small-bowel malabsorption, colitis, chronic renal insufficiency, and diets with a high methionine content (so-called acid ash diets).50 The latter is common in calcium stone formers who eat large amounts of meat. Dehydration, from any cause, increases the urine uric acid concentration and increases supersaturation at any given pH.

FIGURE 69-3. Nomogram showing the undissociated uric acid concentration at values of urine pH and total uric acid concentration. The solubility limit for uric acid is shown by crosshatched bars (96 2 mg/L). (From Coe FL. Uric acid and calcium oxalate nephrolithiasis. Kidney Int 1983; 24:392.)

HYPERURICOSURIA The main causes of hyperuricosuria are diets high in purines and primary overproduction of uric acid. The latter occurs in gout and in rare hereditary disorders of purine metabolism (see Chap. 192). Figure 69-3 illustrates that rises in the urine pH and the urine volume easily prevent excessive supersaturation of uric acid up to excretion levels of 1000 mg per day. Thus, hyper-uricosuria is a much less important risk factor for uric acid stone disease than is the urine pH. Before hyperuricosuria is implicated, the urine should be studied on at least three occasions. All drugs that could produce a falsely high measurement of urine uric acid must be discontinued. The colorimetric method, which is not completely specific, yields falsely high values in the presence of ascorbic acid, salicylates, caffeine, theophylline, high concentrations of glucose, and certain chromogens retained in renal failure. The uricase method is much more specific and does not give falsely high readings, but it is not generally available. TREATMENT ALKALI The goal of treatment is to raise the average urine pH to ~6, as determined by a 24-hour collection. The urine pH should be kept above 6 throughout the day, as determined by monitoring the patient at each voiding. Generally, 2 to 4 mEq/kg alkali is needed in four divided doses. Sodium bicarbonate has a rapid rate of absorption and excretion in the urine, giving rise to sharp increases and decreases in pH. Citrate produces a more sustained increase of urine pH. The potassium salt of citrate may be preferable to sodium bicarbonate because it is more palatable. It is important to avoid alkaline pH values above 6.5, which foster crystallization of brushite and apatite. ALLOPURINOL Reduction of the uric acid excretion rate is advisable only when total uric acid excretion is above 1000 mg per day; allopurinol, 100 mg twice daily, usually is sufficient. The twice-daily dosage is helpful in achieving continuous control of the urine uric acid level.

CYSTINE STONES
PATHOGENESIS The only known cause of cystine stones is a hereditary defect of renal tubule transport that permits large amounts of cystine to pass into the urine. The defects, which include three separate varieties,51 involve not only cystine but also the other dibasic amino acidsornithine, lysine, and arginine. Only the cystinuria has consequences that are clinically important. One might think that cystine loss would cause an amino-acid deficiency state, but this is not a recognized problem. Stone formation primarily is caused by the low solubility of cystine: 300 mg/L urine. Normal people excrete < 60 mg per day of cystine, a level below supersaturation. Heterozygotes for

the so-called type 1 form of cystinuria excrete normal amounts of cystine; heterozygotes for types 2 and 3 excrete abnormal amounts, in the range of 80 to 300 mg per day, along with an excess of the other three dibasic amino acids. However, heterozygotes do not form stones. Homozygous cystinurics, of all three types, excrete more than 400 mg per day of cystine. Patients with cystinuria may have other causes of stones as well, including those associated with calcium salts. TREATMENT Treatment is based on high fluid intake to produce at least 3 L per day of urine. Urine output should be distributed evenly between the day and night. The amount of urine to be excreted can be calculated by dividing the total cystine excretion by the solubility of cystine. Any period of dehydration can cause fresh crystallization; hence, these patients should never be allowed to develop the sensation of thirst or symptoms of dehydration. Moreover, hydration is important because preformed stones can be dissolved by high fluid intake. Another therapeutic approach is to increase urine pH. Above pH 7.5, cystine solubility increases dramatically.1 Treatment with alkali is controversial. It often is used at a dosage of 2 mEq/kg per day, when fluids alone are insufficient. With a bedtime dose of acetazolamide, 250 to 500 mg, a high urine pH can be maintained and excessive sodium alkali salts excreted. The danger of alkali administration is calcium phosphate stone formation, especially in patients who also are hypercalciuric. When fluids are insufficient, D-penicillamine can be tried. This drug forms a soluble disulfide with cysteine that is in equilibrium with cystine. Cystine dissociates to cysteine as cysteine binds to the penicillamine, so the concentration of cystine falls. Cysteine penicillamine is significantly more soluble than cystine. The drug is effective in doses of 1 to 2 g per day in four divided doses, but it has many potential side effects: skin eruptions, fever, lymphaden-opathy, arthritis, leukopenia, thrombocytopenia, a form of pseudothrombocytopenia caused by clumping of platelets in vitro (that deceives cell-counting machines), loss of the senses of smell and taste, proteinuria with focal glomerulitis, cutaneous lupus, and, rarely, Goodpasture syndrome with pulmonary hemorrhage and hemorrhagic acute renal failure.51

STRUVITE STONES
PATHOGENESIS Some bacteria, especially Proteus, Pseudomonas, Klebsiella, and Aerobactersp, possess urease, which is an enzyme that catalyzes the hydrolysis of urea to CO2 and ammonia. By accepting protons to become NH4+, the pH of the urine is raised. Therefore, the concentration of PO43 increases, and the ion associates with NH4+ and magnesium ion to form the crystalline struvite stone.1,52 Because the amounts of urea, phosphate, and magnesium generally are large, struvite stones grow rapidly to form staghorns. The other product of urea hydrolysis, CO2, is readily converted to HCO3 in the tubules. The HCO3, in turn, becomes carbonate (CO 32) as it loses a proton to NH3. The CO32ion crystallizes with calcium to form calcium carbonate, which helps to increase the mass of stone material, and accounts for the almost universal finding of calcium salts in struvite stones.53 Many patients with struvite stone disease initially are metabolic stone formers with calcium or uric acid stones. Presumably, infection from stone passage or from instrumentation leads to treatment that selects for resistant, urease-possessing organisms that are responsible for struvite stone formation. Struvite stone formers should also be evaluated for metabolic causes of stone formation. TREATMENT ROLE OF ANTIBIOTICS Virtually all struvite stones, when cultured, are infected, even when the urine from the patient contains so few bacteria that cultures are sterile. Antibacterial treatment is directed against an infected foreign body as well as infected tissue. Although antibiotics are not successful in preventing stone growth, they may sterilize the urine and the stone surface. At best, the prolonged use of antibiotics may slow the growth of stones. Their principal use is to treat acute infections, which presumably represent an invasion of bacteria into the renal tissue or the urinary tract wall. Another use of antibiotics is after the removal or disruption of a stone, to sterilize the kidneys and urinary tract and any tiny stone fragments that are left behind. For both purposes, bactericidal agents should be used in full dose, for a full time-course, and selected to correspond to the sensitivity pattern of the infecting bacteria. SURGERY The indications for surgery or lithotripsy are, in order of frequency, pain, obstruction, and serious infection. Bleeding is rarely of sufficient magnitude to require treatment. LITHOTRIPSY Lithotripsy is not always effective for large stones, such as are common when struvite stones are present. The number of shock waves needed to disrupt a stone increases with its mass; after several thousand shocks, there is a risk of renal damage. Furthermore, the volume of debris from a large stone may obstruct the ureter, requiring antegrade or retrograde instrumentation. Percutaneous removal, or pyelolithotomy, may be preferable for large stones despite the known hazards of major surgery.54,55 and 56

EVALUATION PROTOCOL FOR PATIENTS WITH NEPHROLITHIASIS

A protocol for patient evaluation is presented in Table 69-3. Three 24-hour urine collections are obtained from outpatients eating their own ad libitum diets. Three corresponding blood samples are drawn between 8:00 and 9:00 a.m., in the postab-sorptive state. Hypercalcemia, hypercalciuria, hyperoxaluria, hyperuricosuria, low pH, and hypocitricuria all are specifically investigated. This information, along with a complete history, physical examination, and review of available radiographs, allows reasonably good classification into a particular type of stone disease. Appropriate treatment is then instituted.

TABLE 69-3. Protocol for Evaluation of Urine and Serum Chemistry Profiles in Kidney Stone Formers

CHAPTER REFERENCES
1. Coe FL, Favus MJ. Disorders of stone formation. In: Brenner BM, Rector FC Jr, eds. The kidney. Philadelphia: WB Saunders, 1986:1403. 2. Coe FL, Strauss AL, Tembe V, Dun SL. Uric acid saturation in calcium nephrolithiasis. Kidney Int 1980; 17:662. 2a. Guillen M, Corella D, Cabello ML, et al. Association between M467T and 114 C-->A gene and some phenotypical variants within the SLC3A1 traits in cystinuria patients from Spain. Hum Genet 2000; 106:314.

3. Griffith DP, Gibson JR, Clinton C, Musher DM. Acetohydroxamic acid: clinical studies of a urease inhibitor in patients with staghorn calculi. J Urol 1978; 119:9.

4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19.

Cotran RS, Rubin RH, Tolkoff-Rubin NE. Tubulointerstitial diseases. In: Brenner BM, Rector FC Jr, eds. The kidney. Philadelphia: WB Saunders, 1986:1143. Kuiper JJ. Medullary sponge kidney. In: Gardner KD Jr, ed. Cystic diseases of the kidney. New York: John Wiley & Sons, 1976:1168. Miles SG, Kaude JV, Newman RC, et al. Extracorporeal shock-wave litho-tripsy: prevalence of renal stones 321 months after treatment. AJR Am J Roentgenol 1988; 150:307. Williams CM, Kaude JV, Newman RC, et al. Extracorporeal shock-wave lithotripsy: long term complications. AJR Am J Roentgenol 1988; 150:311. Coe FL, Parks JH. New insights into the pathophysiology and treatment of nephrolithiasis: new research venues. J Bone Miner Res 1997; 12:522. Parks JH, Coward M, Coe FL. Correspondence between stone composition and urine supersaturation in nephrolithiasis. Kidney Int 1997; 51:894. Coe FL, Lawton RB, Goldstein RB, Tembe V. Sodium urate accelerates precipitation of calcium oxalate in vitro. Proc Soc Exp Biol Med 1975; 149:926. Nielson AE. Kinetics of precipitation. New York: Pergamon Press, 1964. Bisaz S, Felix R, Neuman WF, Fleisch H. Quantitative determination of inhibitors of calcium phosphate precipitation in whole urine. Miner Electrolyte Metab 1978; 1:74. Meyer JL, Smith LH. Growth of calcium oxalate crystal. Invest Urol 1975; 13:36. Nakagawa Y, Abram V, Parks JH, et al. Urine glycoprotein crystal growth inhibitors: evidence for a molecular abnormality in calcium oxalate neph-rolithiasis. J Clin Invest 1985; 76:1455. Nakagawa Y, Ahmed M, Hall SL, et al. Isolation from human calcium oxalate renal stones of nephrocalcin, a glycoprotein inhibitor of calcium oxalate growth. J Clin Invest 1987; 79:1782. Fleisch H, Bisaz S. Isolation from urine of pyrophosphate, a calcification inhibitor. Am J Physiol 1962; 203:671. Coe FL. Treated and untreated recurrent calcium nephrolithiasis in patients with idiopathic hypercalciuria, hyperuricosuria, or no metabolic disorder. Ann Intern Med 1977; 87:404. Coe FL, Parks JH, Moore ES. Familial idiopathic hypercalciuria. N Engl J Med 1979; 300:337. Favus MJ. Familial forms of hypercalciuria. J Urol 1989; 141:719.

19a. Alon US, Berenbom A. Idiopathic hypercalciuria of childhood: 4- to 11-year outcome. Pediatr Nephrol 2000; 14:1011. 20. Pak CYC, Holt K. Nucleation and growth of brushite and calcium oxalate in urine of stone formers. Metabolism 1976; 25:665. 21. Marshall RW, Cochran M, Robertson WG, et al. The relationship between the concentration of calcium salts in the urine and renal stone composition in patients with calcium containing stones. Clin Sci 1972; 43:433. 22. Insogna KL, Broadus AE, Dryer BE, et al. Elevated production rate for 1,25-dihydroxyvitamin D in patients with absorptive hypercalciuria. J Clin Endocrinol Metab 1985; 61:490. 23. Broadus AE, Insogna KL, Lang RAS, et al. Evidence for disordered control of 1,25-dihydroxyvitamin D in absorptive hypercalciuria. N Engl J Med 1984; 31:73. 24. Coe FL, Favus MJ, Crockett T, et al. Effects of low-calcium diet on urine calcium excretion, parathyroid function and serum 1,25(OH) 2D levels in patients with idiopathic hypercalciuria and normal patients. Am J Med 1982; 72:25. 25. Coe FL, Parks JH, Favus MJ. Diet and calcium: the end of an era? Ann Intern Med 1997; 126:553. 26. Laerum E, Larsen S. Thiazide prophylaxis of urolithiasis: a double-blind study in general practice. Acta Med Scand 1984; 215:383. 27. Coe FL, Parks JH, Bushinsky DA, et al. Chlorothalidone promotes mineral retention in patients with idiopathic hypercalciuria. Kidney Int 1988; 33:1140. 28. Kaplan RA, Haussler MR, Deftos LJ, et al. The role of 1,25-dihydroxyvitamin D in the mediation of hyperabsorption of calcium in primary hyper-parathyroidism and absorptive hypercalciuria. J Clin Invest 1977; 59:756. 29. Pearce SH, Williamson C, Kifor O, et al. A familial syndrome of hypocalcemia with hypercalciuria due to mutations in the calcium-sensing receptor. N Engl J Med 1996; 335:1115. 29a. Basmaison O, Rolland MO, Cochat P, Bozon D. Identification of 5 novel mutations in the AGXT gene. Hum Mutat 2000; 15:577. 29b. Mashour S, Turner JF Jr, Merrell R. Acute renal failure, oxalosis, and vitamin C supplementation. Chest 2000; 118:561. 30. Smith LH, Fromm H, Hofman AF. Acquired hyperoxaluria, nephrolithiasis and intestinal disease. N Engl J Med 1972; 286:1371. 31. Dobbins JW, Binder HJ. Importance of colon in enteric hyperoxaluria. N Engl J Med 1977; 296:298. 32. Kathpalia SC, Favus MJ, Coe FL. Evidence for size and charge permeability of rat ascending colon: effects of ricinoleate and bile salts on oxalic acid and neutral sugar transport. J Clin Invest 1984; 74:805. 33. Robertson WG, Peacock M, Nordin BEC. Calcium oxalate crystalluria and urine saturation in recurrent renal stone formers. Clin Sci 1971; 40:365. 34. Smith LH. Enteric hyperoxaluria and other hyperoxaluric states. In: Coe FL, Brenner BM, Stein JH, eds. Contemporary issues in nephrology, vol 4. New York: Churchill Livingstone, 1980:136. 35. Barilla DE, Notz C, Kennedy D, Pak CYC. Renal oxalate excretion following oral oxalate loads in patients with ileal disease and with renal and absorptive hypercalciurias: effect of calcium and magnesium. Am J Med 1978; 64:579. 36. Chadwick VS, Modha K, Dowling RH. Mechanism for hyperoxaluria in patients with ileal dysfunction. N Engl J Med 1973; 289:172. 37. Chlebeck PT, Milliner DS, Smith LH. Long-term prognosis in primary hyperoxaluria type II (L-glyceric aciduria.) Am J Kidney Dis 1994; 23:255. 38. Watts RWE, Morgan SH, Danpure CL, et al. Combined hepatic and renal transplantation in primary hyperoxaluria type I: clinical report of nine cases. Am J Med 1991; 90:179. 39. Alpern RJ, Warnock DG, Rector FC Jr. Renal acidification mechanisms. In: Brenner BM, Rector FC Jr, eds. The kidney. Philadelphia: WB Saunders, 1986:206. 40. Lemann J Jr, Lennon EJ, Goodman AD, et al. The net balance of acid in subjects given large loads of acid or alkali. J Clin Invest 1965; 44:507. 41. Robertson WG, Peacock M, Nordin BEC. Activity products in stone-forming and non-stone-forming urine. Clin Sci 1968; 34:579. 42. McSherry E, Morris RC Jr. Attainment and maintenance of normal stature with alkali therapy in infants and children with classic renal tubular acidosis. J Clin Invest 1978; 61:509. 43. Coe FL, Parks JH. Stone disease in hereditary distal renal tubular acidosis. Ann Intern Med 1980; 93:60. 44. Coe FL, Moran E, Kavalach AG. Dietary purine consumption by hyperuri-cosuric calcium oxalate kidney stone formers and normal subjects. J Chronic Dis 1976; 29:793. 45. Melnick JZ, Preisig PA, Haynes S, et al. Converting enzyme inhibition causes hypocitraturia independent of acidosis or hypokalemia. Kidney Int 1998; 54:1670. 46. Parks JH, Coe FL. Pathogenesis and treatment of calcium stones. Semin Nephrol 1996; 16:398. 47. Lemann J Jr, Gray RW, Pleuss JA. Potassium bicarbonate, but not sodium bicarbonate, reduces urinary calcium excretion and improves calcium balance in healthy men. Kidney Int 1989; 35:688. 48. Fleisch H. Inhibitors and promoters of stone formation. Kidney Int 1978; 13:361. 49. Nakagawa Y, Margolis HC, Yokoyama S, et al. Purification and characterization of a calcium oxalate monohydrate crystal growth inhibitor from human kidney tissue culture medium. J Biol Chem 1981; 256:3936. 50. Hall AP, Berry PE, Dawber TR, et al. Epidemiology of gout and hyperuricosuria: a long term population study. Am J Med 1967; 42:27. 51. Geeson MJ, Kobashi K, Griffith DP. Noncalcium nephrolithiasis. In: Coe FL, Favus MJ, eds. Disorders of bone and mineral metabolism. New York: Raven Press, 1992:801. 52. Elliot JS, Sharp RF, Lewis L. The solubility of struvite in urine. J Urol 1959; 80:169. 53. Kristensen C, Parks JH, Lindheimer M, Coe FL. Reduced glomerular filtration rate and hypercalciuria in primary struvite nephrolithiasis. Kidney Int 1987; 32:749. 54. Chaussy CG, Fuchs GJ. Current state and future developments of noninvasive treatment of human urinary stones with extracorporeal shock wave lithotripsy. J Urol 1989; 141:782. 55. Lingeman JE, Woods J, Toth PD, et al. Role of lithotripsy and its side effects. J Urol 1989; 141:793. 56. Lingeman JE, Siegel YI, Steele B. Management of lower pole nephrolithiasis: a critical analysis. J Urol 1994; 151:663.

CHAPTER 70 DISORDERS OF CALCIUM AND BONE METABOLISM IN INFANCY AND CHILDHOOD Principles and Practice of Endocrinology and Metabolism

CHAPTER 70 DISORDERS OF CALCIUM AND BONE METABOLISM IN INFANCY AND CHILDHOOD

THOMAS O. CARPENTER Abnormal Serum Calcium Neonatal Hypocalcemia Early Neonatal Hypocalcemia Late Neonatal Hypocalcemia Neonatal Hypercalcemia Primary Hyperparathyroidism in the Neonatal Period Idiopathic Hypercalcemia Rickets Vitamin DDeprivation and Calcium Deficiency Inherited Disorders of Vitamin D Metabolism and Activity X-Linked Hypophosphatemic Rickets Osteopenia in Childhood Major Childhood Osteopenia Immobilization Osteogenesis Imperfecta Idiopathic Juvenile Osteoporosis Turner Syndrome Homocystinuria Lysinuric Protein Intolerance Menkes Kinky Hair Syndrome Malnutrition Hypercortisolism Leukemia Osteopenia of Prematurity Toward a Definitive Diagnosis Chapter References

ABNORMAL SERUM CALCIUM

Hypocalcemia is the most common disorder of mineral metabolism in the neonatal period and is observed frequently in premature infants. Although less common, hypercalcemia may also occur in the neonatal period and, if severe, may be life threatening. NEONATAL HYPOCALCEMIA Except for the unusual cases of congenital hypoparathyroidism, neonatal hypocalcemia is usually transient, persisting for a few days to a few weeks.1,2 Neonatal hypocalcemia is categorized according to the time of its onset: early neonatal hypocalcemia and late neonatal hypocalcemia. EARLY NEONATAL HYPOCALCEMIA Early neonatal hypocalcemia occurs in the first 24 to 48 hours of life and is observed most frequently in premature infants, sick infants, and infants born of an abnormal labor or pregnancy. The condition can be explained as an exaggeration of the normal postnatal decrease in serum calcium; the fall in serum calcium is inversely proportional to the gestational age of the infant. The serum calcium remains low for a few to several days and then gradually increases, usually reaching normal levels by 1 to 2 weeks of age. The serum inorganic phosphate concentration is usually normal, although it may be elevated in asphyxiated infants and in infants born of diabetic mothers. It has been suggested that inappropriate parathyroid secretion, parathyroid hormone (PTH) resistance, and/or vitamin D metabolic abnormalities may contribute to the development of this condition, but no single conclusive mechanism has been confirmed. Many premature infants with early neonatal hypocalcemia are asymptomatic, but in others, tetany or convulsions may be present. Symptomatic infants obviously require calcium therapy, but there is no unanimity regarding treatment of hypocalcemic infants who are asymptomatic. The emergency treatment of neonatal hypocalcemia consists of the intravenous administration of 1 mL per minute 10% calcium gluconate, which should not exceed 2.0 mL/kg. This may be repeated three to four times in 24 hours to control the acute symptoms. After the acute symptoms have been controlled, 5.0 mL/kg 10% calcium gluconate may be given with intravenous fluids over a 24-hour period, or calcium supplements may be given orally if feedings are tolerated. Occasionally, hypomagnesemia is concomitantly identified; this can be treated with 0.1 to 0.2 mL/kg of a 50% solution of magnesium sulfate (MgSO47H2O). LATE NEONATAL HYPOCALCEMIA Late neonatal hypocalcemia appears at the end of the first week of life or later, often in full-term infants who have received a high phosphate load, such as that formerly encountered with feedings of evaporated cow's milk formula or from a phosphate enema.2a Infants with late neonatal hypocalcemia usually have clinical manifestations of tetany or convulsions. Hyperphosphatemia is a prominent feature, and the serum PTH level may be low, reflecting a state of functional hypoparathyroidism in the presence of hypocalcemia. This form of hypocalcemia is seen in infants born to hyperparathyroid mothers and in children with congenital heart disease in the postoperative period. The emergency treatment of acute tetany or convulsions secondary to hypocalcemia is the same as for early neonatal hypocalcemia. Dietary factors are of importance in the management of late neonatal hypocalcemia; the phosphate load should be diminished, with an increase of the calcium/phosphate ratio of milk feedings to 4:1. The author often uses Similac PM 60/40 (Ross Laboratories, Columbus, Ohio) in this setting. The serum calcium level usually increases when the infants are given such milk feedings; and, after several days to weeks, the serum PTH level gradually rises and the infants can tolerate higher phosphate loads. The pathogenesis of the transient hypoparathyroidism in these infants is unknown. NEONATAL HYPERCALCEMIA Neonatal hypercalcemia is found in association with several disorders of calcium metabolism.3 In the presence of mild to moderate hypercalcemia (11.013.0 mg/dL) the infant is often asymptomatic. Infants with more severe hypercalcemia have various clinical findings, including failure to thrive, poor feeding, hypotonia, vomiting, seizures, lethargy, polyuria, and hypertension. The medical management of acute symptomatic hypercalcemia consists of the administration of intravenous saline. Additionally, furosemide, in a dose of 1 mg/kg, is frequently given intravenously at 6- to 8-hour intervals. Intravenous infusion of pamidronate has also been useful in this setting. Specific long-term therapy depends on the specific hypercalcemic disorder. Disorders associated with neonatal hypercalcemia are listed in Table 70-1.

TABLE 70-1. Clinical Disorders Associated with Hypercalcemia in the Neonatal Period

PRIMARY HYPERPARATHYROIDISM IN THE NEONATAL PERIOD In infants with neonatal primary hyperparathyroidism, the symptoms of hypercalcemia usually develop during the early days of life. Severe hypercalcemia is usually present, with serum calcium levels that range between 15 and 30 mg/dL. Occasionally, however, the hypercalcemia is mild, with serum calcium values between 11 and 13 mg/dL. The serum inorganic phosphate level is often depressed, whereas the serum PTH level is elevated (see Chap. 58). Radiographic studies of bone are characteristic of primary hyperparathyroidism. Moreover, renal calcinosis may be present. Primary hyperparathyroidism may be sporadic, or it may be inherited in an autosomal-recessive fashion. An interesting association between neonatal severe primary hyperparathyroidism and familial hypocalciuric hypercalcemia (FHH) occurs4 (see Chap. 58). FHH, an autosomal-dominant trait, is manifest by modest asymptomatic hypercalcemia with relative hypocalciuria and either normal or somewhat elevated serum PTH levels. An inactivating mutation in the gene for the extracellular calcium-sensing receptor, which is present on parathyroid cells, acts in a dominant negative manner in FHH. A greater level of serum ionized calcium is required to suppress PTH secretion than would normally be necessary; thus, modest increases in serum calcium are maintained. Individuals who are homozygous for such a mutation effectively fail to suppress PTH secretion and manifest severe neonatal hyperparathyroidism.5 Severe neonatal primary hyperparathyroidism is considered a surgical emergency. The serum calcium levels often range between 15 and 30 mg/dL. Total parathyroidectomy may be necessary because of recurrence after subtotal parathyroidectomy.6 Parathyroidectomy and heterotopic autotransplantation have also been suggested; however, a current common practice is to leave a small portion of one gland in the neck, marked with a surgical clip for ease in identification should reexploration be necessary. In some infants, the disorder is milder, with low serum calcium concentrations (i.e., 1113 mg/dL). These infants may not require surgical intervention; their bony lesions may heal spontaneously, they may be asymptomatic in the presence of modest hypercalcemia in association with relative hypocalciuria, and the disorder may be self-limited.7 IDIOPATHIC HYPERCALCEMIA There are mild and severe forms of idiopathic infantile hypercalcemia. In the mild form, symptoms associated with hypercalcemia usually appear between 2 and 9 months of age. The infants may fail to thrive and, in a few, a cardiac murmur may be apparent. The facies frequently appears normal, and the prognosis for physical and mental development is usually good. In the severe form, the symptoms may date from birth; however, this is frequently difficult to document. Prenatal and postnatal growth failure are common, and a number of the phenotypic features of Williams syndrome (Fig. 70-1) are observed in some patients with the severe disorder. These features, in addition to hypercalcemia, include cardiovascular abnormalities (usually supravalvular aortic stenosis or peripheral pulmonic stenosis), late psychomotor development, selective mental deficiency, a characteristic unusual facies, and short stature. A deletion of the elastin gene is found in many cases of Williams syndrome.8

FIGURE 70-1. Idiopathic hypercalcemia (Williams syndrome) in a 3- year-old boy. Note the typical broad forehead, depressed bridge of the nose, anteverted nares, long philtrum (the vertical median groove extending from beneath the nose to the upper lip), the slight strabismus, the wide mouth with large lips, the drooping of the lower lip, and the large ears. The cheeks are full and dependent, and there is mandibular hypoplasia.

The serum calcium levels range between 12 and 19 mg/dL. The hypercalcemia usually subsides spontaneously by the age of 4 years. The prognosis for patients with the severe form is poor, and the mortality may be as high as 25% during the first 4 years of life. The pathogenesis of idiopathic hypercalcemia is uncertain. In one study, an exaggerated 25-hydroxyvitamin D [25(OH)D] response to vitamin D administration was observed, suggesting that the regulation of this metabolite may be abnormal.9 There are conflicting reports on the circulating levels of 1,25-dihydroxyvitamin D [1,25(OH)2D] in the severe form of idiopathic hypercalcemia.10,11 Treatment consists of placing the child on a low calcium, vitamin Dfree diet. In some resistant cases, short-term therapy with corticosteroids may be necessary. The author has used intravenous pamidronate to correct hypercalcemia in Williams syndrome. Although rare in the neonatal period, intoxication with vitamin D or vitamin A should be excluded in older infants with hypercalcemia. Another rare condition in this differential diagnosis includes subcutaneous fat necrosis, a self-limited disorder of infancy for which increased production of 1,25(OH)2D has been described.12

RICKETS
Rickets may be defined as a disorder in which there is a lag in the rate of mineralization of the matrix of bone and growth cartilage. Growth plate cartilage typically hypertrophies in an unorganized fashion, producing the characteristic widened ends of the long bones seen in this disorder. The formation of hydroxyapatite depends on adequate concentrations of extra-cellular calcium and phosphate; therefore, a deficiency of either or both of these minerals may cause rickets. There are more than 30 causes of rickets: One classification scheme that categorizes the various forms of rickets is shown in Table 70-2 (see Chap. 63).

TABLE 70-2. Causes of Rickets

VITAMIN DDEPRIVATION AND CALCIUM DEFICIENCY There has been a resurgence of vitamin Ddeprivation rickets associated with breast-feeding and special dietary practices, including macrobiotic and other vegetarian diets. Human breast milk may provide as little as 10 to 20 IU of vitamin D per day, and macrobiotic diets may provide comparable amounts of vitamin D. Marginal vitamin D stores are often present in the breast-feeding mother of a rachitic infant. Thus, infants who receive only human milk and children who receive macrobiotic diets are at high risk for developing rickets in the early months and years of life. The physical findings at the time of the rickets diagnosis may include flaring of the wrists and ankles, frontal bossing, rachitic rosary, Harrison grooves (a groove extending laterally from the xiphoid process that corresponds to the diaphragmatic attachment), long-bone fractures, bowed legs, leg tenderness, and large fontanelles (Fig. 70-2,Fig. 70-3 and Fig. 70-4). The biochemical findings include a low serum calcium or phosphate level, or both, and elevated serum alkaline phosphatase activity. In the presence of hypocalcemia, the circulating PTH level is increased, which, in turn, leads to generalized aminoaciduria. The serum 25(OH)D concentrations are depressed, whereas the serum 1,25(OH)2D levels may be either low, normal, or modestly elevated. Radiographic findings include widening of the space between the calcified plate at the end of the metaphysis and the center of ossification. Other changes at the cartilage-shaft junction include cupping, spreading, spur formation, fringing, and stippling.

FIGURE 70-2. Nutritional rickets (vitamin D deficiency) in a 3-year-old boy. Note the severe bowing of the lower extremities and the widened wrists and ankles.

FIGURE 70-3. Flaring of the wrists (arrows) of a 2-year-old boy with rickets.

FIGURE 70-4. Rachitic rosary (arrows) in a 5-year-old boy. This condition is caused by enlargement of the costochondral junctions. (N, nipple; A, axilla; S, scapula.)

The author usually prescribes 1000 to 2000 IU vitamin D per day as treatment for vitamin Ddeficient rickets. Others may prefer the administration of intermediate amounts of oral vitamin D (i.e., 800016,000 IU per day). After radiographic evidence of healing, the dosage is usually reduced to prophylactic amounts of 400 IU per day. Where compliance is an issue, the so-called stoss form of therapy is used, which consists of the administration of a single oral or intramuscular dose of 600,000 IU vitamin D. (Some prefer to divide the large oral dose into two or more doses given several hours apart on the same day.) Often, these children are referred to specialty clinics after the diagnosis is suspected and supplementation with some form of vitamin D has been initiated. These children often have low normal levels of 25(OH)D. The radiographic hallmark of partially treated vitamin Dresistant rickets is a thin dense line of opacity at the metaphyses of long bones, representing recent rapid mineralization at the edge of the growth plate. It is not infrequent for children with vitamin D deficiency to require supplemental calcium. Some of these children apparently have a low dietary intake of calcium, and others may manifest the hungry bone syndrome, in which mineralization is sufficiently rapid to effect a decrease in the serum calcium levels. For this reason, supplemental calcium (such as calcium glubionate or calcium carbonate) should be included in the therapy, so that the child receives a total daily intake of 30 to 50 mg/kg elemental calcium. Indeed, stores of vitamin D are subject to relatively rapid turnover in the setting of concomitant low dietary intake of calcium.13

INHERITED DISORDERS OF VITAMIN D METABOLISM AND ACTIVITY

Autosomal recessive 1a -hydroxylase deficiency (also known as hereditary pseudovitamin D deficiency rickets or vitamin Ddependent rickets type I) is an autosomal-recessive disorder in which the renal 25(OH)D 1a-hydroxylase that converts 25(OH)D to 1,25(OH)2D is dysfunctional. Individuals with this disorder have a mutation in the gene encoding a specific cytochrome P450 that donates electrons to the substrate, 25(OH)D.14 Thus, this enzyme component is essential to the hydroxylation reaction. Onset of symptoms is usually within 4 to 12 months of age, and the clinical course and biochemical features are similar to those of severe vitamin Ddeficiency rickets. In contrast to vitamin Ddeficient rickets, there is a history of adequate dietary intake of vitamin D or of sunlight exposure. In untreated patients, the serum 25(OH)D level is normal. The serum 1,25(OH)2D is low to low-normal and remains so in patients treated with pharmacologic doses of vitamin D or 25(OH)D. Patients with 1a-hydroxylase deficiency may respond to high dosages of vitamin D or 25(OH)D; however, treatment with 1,25(OH)2D is optimal, using initial dosages from 0.5 to 3.0 g per day and maintenance dosages ranging from 0.25 to 2.0 g per day. Hereditary vitamin D resistance (also known as vitamin D dependent rickets type II) is a rare inherited disorder transmitted as an autosomal-recessive trait. The

clinical, radiologic, and most of the biochemical features are similar to those observed in 1a-hydroxylase deficiency (Fig. 70-5). The major biochemical distinction is that circulating 1,25(OH)2D levels are high in patients with hereditary vitamin D resistance and low in patients with 1a-hydroxylase deficiency. A unique clinical feature in some kindreds with hereditary vitamin D resistance is total body and scalp alopecia. Other patients may demonstrate oligodontia (Fig. 70-6). Studies in affected individuals of several kindreds have revealed defects in the 1,25(OH)2D receptor, which is one of a large superfamily of hormone receptors, including glucocorticoid, mineralocorticoid, and thyroid hormone receptors. Mutations, which have been identified in the gene encoding this receptor, result in altered DNA binding, altered ligand binding, or interruption of complete synthesis of the receptor (see Chap. 54). In some families, no genetic defect has been identified.

FIGURE 70-5. Rachitic changes in the lower extremities of a 2-year-old girl with hereditary resistance to vitamin D. At presentation (A) severely deformed epiphyses, demineralization, and bilateral fractures of the femora and tibiae are evident. After 4 months of therapy with high-dose calcitriol (B), dramatic remodeling of the metaphyseal edges has occurred.

FIGURE 70-6. The limited eruption of teeth (oligodontia), as shown here, is characteristic of certain kindreds with hereditary resistance to vitamin D.

Patients with hereditary vitamin D resistance have been treated with megadoses of vitamin D or vitamin D metabolites with variable responses to such therapy. Parenteral administration of calcium has been shown to completely correct the skeletal abnormalities in severe forms of this disorder.15 X-LINKED HYPOPHOSPHATEMIC RICKETS X-Linked hypophosphatemic rickets16 (XLH) is inherited as an X-linked dominant trait and is characterized by hypophosphatemia and decreased renal tubular phosphate reabsorption. The inherited disorder may vary in degree, from severe to mild bone disease associated with hypophosphatemia, to hypophosphatemia alone without evidence of active or former rickets. The severity of bone disease does not correlate with the degree of hypophosphatemia. Children with XLH are usually seen in the second or third year of life with bowed legs and short stature. Controversy exists whether the mineralized tissue abnormalities are more severe in affected males compared with females.17,18 Hallmark laboratory findings include hypophosphatemia, normocalcemia, and elevated serum alkaline phosphatase activity. The tubular reabsorption rate and the maximum reabsorption rate of phosphate are decreased. The serum PTH level is usually normal but can be elevated in the untreated state. Hyperparathyroidism is not infrequently seen as a complication of therapy. In the untreated state, the serum 25(OH)D levels are normal. The serum 1,25(OH)2D levels are either normal or somewhat decreased. The normal circulating levels of this metabolite are inappropriately low in the presence of hypophosphatemia. Pathophysiologic features of XLH include a defect in renal tubular transport of phosphate and impaired generation of 1,25(OH)2D in response to both phosphate deprivation and PTH. The mutated gene19 (PHEX) encodes a protein with homology to known metallo-endopeptidases (including nephrolysin and endothelin-converting enzyme-1), although the role of such an enzyme in the pathophysiology of the disease has not been clearly established. Studies in an animal model, the hyp mouse, indicate that a circulating factor influences the renal phosphate transport.20,21 Whether the mutation directly affects the skeleton independent of serum phosphate levels is not entirely clear, but has been suggested by other experiments in hyp mice.22,23 Treatment consists of the administration of phosphate plus vitamin D or a vitamin D metabolite. Phosphate treatment alone, or in conjunction with vitamin D, induces healing of the rachitic lesions at the growth plate, but has little effect on the mineralization defect in trabecular bone. Furthermore, hyper-parathyroidism is likely to result from solitary phosphate therapy. In practice, treatment of XLH has employed relatively large amounts of phosphorus (14 g per day) in conjunction with 1,25(OH)2D(0.52.0g per day or 1040 ng/kg per day). This regimen has been demonstrated to improve the rachitic lesion at the growth plate and mineralization in trabecular bone.24 The administration of vitamin D or vitamin D metabolites counteracts the hypocalcemic effect of phosphate therapy, thereby serving to prevent secondary hyperparathyroidism. Complications of higher dosages of 1,25(OH)2D (averaging 68.210 ng/kg per day in one study) include hypercalcemia and hypercalciuria.25 Because of the propensity to develop secondary26 and, in some instances, tertiary hyperparathyroidism27 with large doses of phosphate, and because of concern for the complication of vitamin D intoxication, the author maintains patients on lower dosages of phosphate and 1,25(OH)2D than previous reports recommend. The average dose in our pre-pubertal patients is 0.75 g per day of elemental phosphorus, given in three to four divided doses; rarely does the author recommend more than 1.5 g per day of phosphorus. The average dose of 1,25(OH)2D used is 0.75 g per day, which is usually given in two divided doses. Furthermore, the author has found that supplementation of this regimen with 24,25(OH)2D at a dose of 10 g per day can aid in the correction of mild secondary hyperparathyroidism.28 The rachitic disease is exacerbated by the adolescent growth spurt; doses of 1,25(OH)2D may be transiently increased to as high as 2.0 g per day, but such high-dose therapy is rarely used for longer than a year and should be carefully monitored. Likewise, oral phosphorus during this phase may be transiently increased to 2 g per day, but careful monitoring is important. In the past, treatment has been limited to affected children and stopped after achievement of ultimate height. It now appears likely, however, that symptomatic adults may benefit from therapy.29 Nephrocalcinosis is common in treated patients with X-linked hypophosphatemia. Ultrasound examinations of the kidney should be performed occasionally to monitor this complication. Data suggest that the clinical significance of this lesion is very limited.30 In addition to X-linked transmission, both autosomal-recessive and autosomal-dominant transmission patterns of hypophosphatemic rickets have been reported. The gene associated with the autosomal-dominant form has been mapped to the short arm of chromosome 12.31 Acquired, sporadic forms of hypophosphatemic rickets occur later in childhood or in adulthood. Hypophosphatemic rickets, with a low serum 1,25(OH)2D concentration in some patients, is found in association with skeletal and soft tissue mesenchymal tumors; the term oncogenic rickets has been applied to these cases. It is possible that a mesenchymal tumor may be a pathogenetic factor in many cases of acquired hypophosphatemic rickets, but because the tumors are often benign and small, they may not be discovered (see Chap. 63 and Chap. 219). Phosphate-wasting rickets also occurs in linear sebaceous nevus syndrome,32 neurofibromatosis,33 and in the McCune-Albright syndrome. In this disorder, a somatic

mutation of the membrane guanyl nucleotide stimulatory protein results in constitutive stimulation of adenylate cyclase. The somatic nature of the mutation results in a random pattern of affected tissues (see Chap. 60).34 Should the renal tubule possess the mutation, a PTH-like inhibition of phosphate transport and renal phosphate wasting may result. Hereditary hypophosphatemic rickets with hypercalciuria is yet another form of phosphate-wasting rickets, which is distinguished from X-linked hypophosphatemia by an autosomal-recessive mode of transmission and a normal hypophosphatemia1a-hydroxylase axis.35 Such patients, therefore, have an appropriately elevated circulating 1,25(OH)2D level, which results in increased intestinal calcium absorption, a propensity to hypercalcemia, hypercalciuria, and reduced circulating PTH levels. Patients generally respond well to treatment with phosphate salts alone, and vitamin D supplementation provokes frank hypercalcemia. Urinary calcium excretion should be determined in children with phosphate-wasting rickets to correctly identify this disease variant.

OSTEOPENIA IN CHILDHOOD
Children occasionally come to the physician's attention because of osteopenia (radiographic evidence of reduced bone mass). Abnormalities in serum calcium, phosphorus, or alkaline phosphatase may confirm the suspicion of rickets. In less clear cases, however, differentiation between osteomalacia (defective mineralization of osteoid) and osteoporosis (decreased organic and calcified bone mass) must be sought (see Chap. 55 and Chap. 64). This discussion is concerned primarily with disorders in which osteopenia is a presenting feature (Table 70-3). There are several conditions in which osteopenia may be an associated finding and not the primary manifestation of the disease (e.g., thyrotoxicosis, acromegaly, hypogonadism [Klinefelter syndrome, Noonan syndrome, hyperprolactinemia], hyperparathyroidism, diabetes mellitus, the lactating adolescent, heparin therapy, mastocytosis, increased erythropoiesis, Down syndrome, Wilson disease, cystic fibrosis, familial dysautonomia, anorexia nervosa, and other severe chronic diseases).

TABLE 70-3. Osteoporotic Disorders with Osteopenia as Presenting Feature

It appears that racial differences in bone mass begin at an early age. Adequate but not excess calcium intake in children may have positive effects on bone mineral status in adulthood. MAJOR CHILDHOOD OSTEOPENIA IMMOBILIZATION Severe trauma, myelodysplasia, and juvenile rheumatoid arthritis are predisposing conditions for immobilization-related osteoporosis in children. OSTEOGENESIS IMPERFECTA Osteogenesis imperfecta, a heterogeneous group of diseases, has an estimated incidence of 1 per 15,000 live births. The osteopenia results from abnormal production of bone matrix secondary to defective collagen synthesis or assembly.36 Other tissues that contain collagen may be affected, as evidenced by joint laxity and thin skin. The sclerae may appear blue because of thinning, allowing the blue pigment of the choroid to be transmitted. Improper formation of dentin may occur, and teeth may be yellow or opalescent and chip easily. Sensorineural or conductive hearing loss may be present. The skeleton is variably affected. Wormian bones (islands of bone with a rich vascular supply) are frequently observed on skull radiographs. Biochemical findings are nonspecific; however, mild hypercalcemia and hypercalciuria may be present in younger patients. Markers of bone turnoversuch as serum alkaline phosphatase activity and osteocalcin as well as urinary excretion of resorptive markersmay also have higher than average values. Serum PTH and vitamin D metabolites are nearly always normal. Numerous mutations in genes encoding the a1 and a2 chains of type I collagen (bone collagen) have been described.36a Most kindreds are affected with unique mutations, and sporadic mutations are not uncommonly reported. Studies note variable severity associated with the mutations, being dependent on the region of the gene in which the mutation occurs; thus a regional model for site of mutation/severity has been proposed.37 Mutations in introns resulting in abnormal mRNA processing have been described; these result in abnormal ratios of chain production, not abnormal chains, per se. Prenatal diagnosis of osteogenesis imperfecta has been possible through genetic studies of chorionic villous samples or through high-resolution fetal ultrasound studies in severe cases. After years of disappointing results with various experimental therapies of the disease, trials with intravenous pamidronate have shown dramatic changes in the bone density and fracture rate over the short term.38 The bisphosphonate preparation is given intravenously in 3- to 4-hour daily infusions that are repeated for 3 successive days per cycle. The cycle is repeated at 4-month intervals. Therapy is sufficiently new that long-term effects on the skeleton have not yet been clearly established. Nevertheless, serious consideration for therapy should be undertaken in patients with severe disease. It is less clear whether mild forms of the disease warrant such an intervention. Finally, it seems likely that a number of unidentified mutations in the collagen gene could be responsible for disorders of childhood osteoporosis that have not classically been diagnosed as osteogenesis imperfecta. Related connective tissue disorders, particularly the Ehlers-Danlos syndrome and, rarely, Marfan syndrome, may also manifest osteopenia (see Chap. 66 and Chap. 189). IDIOPATHIC JUVENILE OSTEOPOROSIS In contrast to osteogenesis imperfecta, idiopathic juvenile osteoporosis usually becomes evident just before puberty. Such individuals may have pain in the weight-bearing skeleton, refusal to walk, or gait abnormalities. Radiographs can show generalized osteopenia, compression fractures of the thoracolumbar vertebrae, and metaphyseal compression fractures. Marked osteopenia in the metaphyseal region of new bone formation is characteristic. There are no consistent biochemical changes in this disorder. Increased hydroxyproline excretion and negative calcium balance may occur. The circulating PTH level has been reported as elevated relative to the serum calcium level. Serum 1,25(OH)2D levels may be decreased, normal, or increased. The active disease is almost always transient, but residual deformity is not uncommon. Rarely, patients have persistence of active osteoporosis through adolescence and into adulthood (see Chap. 64). Some patients have been shown to respond to bisphosphonate therapy.39 TURNER SYNDROME A number of skeletal abnormalities have been reported in Turner syndrome, including osteopenia. However, recent studies of bone density in girls with Turner syndrome have indicated that when bone mineral density is normalized to height or body mass rather than age, bone mass may actually be normal. Girls or older women with Turner syndrome may have an increased incidence of fractures. HOMOCYSTINURIA

Homocystinuria refers to several disorders of methionine metabolism resulting in increased blood and urinary homocystine levels (see Chap. 191). The prototype disorder, caused by cystathionine b-synthase deficiency, is commonly manifested by a dislocated lens and a consistent, marked osteoporosis that is most evident in the spine. Some of the skeletal abnormalities resemble those of Marfan syndrome, including kyphosis, scoliosis, pectus excavatum and carinatum, and arachnodactyly. Other features include mental retardation, seizures, malar flush, and vascular thromboses. The detection of homocysteine or homocystine in the urine by the cyanide-nitroprusside reaction is a useful diagnostic screening procedure. Management has included low methionine diets, large amounts of pyridoxine, and supportive management of complications. LYSINURIC PROTEIN INTOLERANCE Osteopenia is an almost constant feature of lysinuric protein intolerance, which is an autosomal-recessive disorder of aminoacid transport. Episodic vomiting, anorexia, variable hepatomegaly, protein aversion, or unexplained osteopenia may bring the child to the physician's attention. Defective renal, intestinal, and hepatocellular transport of dibasic amino acids (ornithine, lysine, and arginine) cause limited urea cycle activity and episodic hyperammonemia. Characteristically, decreased levels of circulating dibasic amino acids and massive urinary excretion of lysine occur. Generalized aminoaciduria also may be present. Very marked elevations in serum lactate dehydrogenase may occur, as well as elevations in serum transaminases. Histomorphometric analysis of bone in one patient confirmed an osteoporotic process. Restricted protein intake probably accounts for the osteoporosis. Citrulline therapy has improved general well-being as well as the osteoporosis. MENKES KINKY HAIR SYNDROME Menkes kinky hair syndrome is an X-linked disorder characterized by defective intestinal absorption and tissue accumulation of copper. Deficient copper-dependent enzymes result in abnormal hair, neuronal degeneration, and hypothermia. The osteoporosis is caused by defective collagen synthesis, resulting from the defective copper-dependent enzyme lysyl oxidase, which is required for normal cross-linking (also see Chap. 131). MALNUTRITION Celiac disease, proteincalorie malnutrition, and copper deficiency may result in osteoporosis. Copper deficiency is thought to cause osteoporosis secondary to defective cross-linking of collagen. HYPERCORTISOLISM Cushing syndrome, which is caused by an excess of endogenous glucocorticoids, may present as osteopenia, although other features of hypercortisolism are usually evident. The vitamin D metabolites 25-OHD and 1,25(OH)2D have been proposed as therapeutic agents for steroid-induced osteoporosis, although use of the latter in adults resulted in a significant incidence of hypercalcemia. Biphosphate therapy may prove to be useful in this setting. LEUKEMIA Leukemia may produce osteoporosis before abnormal cellular forms are evident in the peripheral circulation. OSTEOPENIA OF PREMATURITY Osteopenia of prematurity is common in low-birth-weight infants, in whom it is difficult to provide, ex utero, the mineral requirements necessary for the rapidly growing and developing skeleton. Rickets or osteomalacia may be a significant factor in the bone disease of these patients. Human milk fortifiers have been developed to provide adequate mineral to counter this problem but are not without complications (see section on hypercalcemia). TOWARD A DEFINITIVE DIAGNOSIS In the absence of evidence of primary osteomalacia or other systemic disease, determinations of serum copper, bicarbonate, uric acid, and serum and urine amino-acid levels and, occasionally, chromosomal analysis should be performed. A search for evidence of hyperadrenalism is undertaken if subtle cushingoid manifestations are present. If the child is severely affected or systemically ill, yet with no features of a specific disorder, a bone marrow aspiration may be necessary and could be performed during a bone biopsy procedure. Usually, no definitive diagnosis is evident, and one is left to distinguish between idiopathic juvenile osteoporosis and mild osteogenesis imperfecta. Analysis of collagen produced in skin samples obtained at biopsy may confirm a diagnosis of osteogenesis imperfecta. A definitive diagnosis is important in view of the potentially effective therapy with bisphosphonates described earlier. CHAPTER REFERENCES
1. Hillman LS, Haddad JG. Hypocalcemia and other abnormalities of mineral homeostasis during the neonatal period. In: Heath DA, Marx SJ, eds. Calcium disorders. Clinical endocrinology, Butterworths international medical reviews. London: Butterworths, 1982; 248. 2. Cole DEC, Carpenter TO, Goltzman D. Calcium homeostasis and disorders of bone and mineral metabolism. In: Collu R, ed. Pediatric endocrinology, comprehensive endocrinology series. New York: Raven Press, 1988; 509. 2a. Walton DM, Thomas DC, Aly HZ, Short BL. Morbid hypocalcemia associated with phosphate enema in a six-week-old infant. Pediatrics 2000; 106:E37.

3. Anast CS, David L. Human neonatal hypercalcemia. In: Holick MF, Anast CS, Gray TK, eds. Perinatal calcium and phosphorus metabolism. Amsterdam: Elsevier 1983; 363. 4. Marx SJ, Attie MF, Spiegel AM, et al. An association between severe primary hyperparathyroidism and familial hypocalciuric hypercalcemia in three kindreds. N Engl J Med 1982; 306:257. 5. Pollack MR, Chou Y-HW, Marx SJ, et al. Familial hypocalciuric hypocalcemia and neonatal severe hyperparathyroidism: effects of mutant gene dosage on phenotype. J Clin Invest 1994; 93:1108. 6. Cooper L, Wertheimer J, Levy R, et al. Severe primary hyperparathyroidism in a neonate who was the product of two related hypercalcemic parents managed by a parathyroidectomy and heterotopic autotransplantation. Pediatrics 1986; 78:263. 7. Harris SS, D'Ercole J. Neonatal hyperparathyroidism: the natural course in the absence of surgical intervention. Pediatrics 1989; 83:53. 8. Ewart AK, Morris CA, Atkinson D, et al. Hemizygosity at the elastin locus in a developmental disorder, Williams syndrome. Nat Genet 1993; 5:11. 9. Taylor AB, Stern PH, Bell NH. Abnormal regulation of circulating 25-hydroxyvitamin D in the Williams syndrome. N Engl J Med 1982; 306:972. 10. Goodyer PR, Frank A, Kaplan BS. Observations on the evolution and treatment of idiopathic infantile hypercalcemia. J Pediatr 1984; 105:771. 11. Garabedian M, Jacqz E, Guillozo H, et al. Elevated plasma 1,25-dihydroxy-vitamin D concentrations in infants with hypercalcemia and an elfin facies. N Engl J Med 1985; 312:948 12. Kruse K, Irle U, Uhlig R. Elevated 1,25-dihydroxyvitamin D serum concentrations in infants with subcutaneous fat necrosis. J Pediatr 1993; 122:460. 13. Clements MR, Johnson L, Fraser DR. A new mechanism for induced vitamin D deficiency in children. Nature 1987; 325:62. 14. Fu GK, Lin D, Zhang MY, et al. Cloning of human 25-hydroxyvitamin D-1a-hydroxylase and mutations causing vitamin Ddependent rickets type 1. Mol Endocrinol 1997; 11:1961. 15. Balsan S, Garabedian M, Larchet M, et al. Long-term nocturnal calcium infusions can cure rickets and promote normal mineralization in hereditary resistance to 1,25-dihydroxyvitamin D. J Clin Invest 1986; 77:1661. 16. Carpenter TO. New perspectives on the biology and treatment of X-linked hypophosphatemic rickets. Pediatr Clin North Am 1997; 44:443. 17. Whyte MP, Schranck FW, Armamento-Villareal R. X-linked hypophosphatemia: a search for gender, race, anticipation, or parent of origin effects on disease expression in children. J Clin Endocrinol Metab 1996; 81:4075. 18. Winters RW, Graham JB, Williams TF, et al. A genetic study of familial hypophosphatemia and vitamin D resistant rickets with a review of the literature. Medicine 1958; 37:97. 19. The HYP Consortium. A gene (PEX) with homologies to endopeptidases is mutated in patients with X-linked hypophosphatemia. Nat Genet 1995; 11:130. 20. Nesbitt T, Coffman TM, Griffiths R, et al. Cross transplantation of kidneys in normal and Hyp mice: evidence that the Hyp mouse phenotype is unrelated to an intrinsic renal defect. J Clin Invest 1992; 89:1453. 21. Meyer RA Jr, Meyer MH, Gray RW. Parabiosis suggests a humoral factor is involved in X-linked hypophosphatemia in mice. J Bone Miner Res 1989; 4:493. 22. Ecarot B, Glorieux FH, Desbarats M, et al. Effect of 1,25-dihydroxyvitamin D 3 treatment on bone formation by transplanted cells from normal and X-linked hypophosphatemic mice. J Bone Miner Res 1995; 10:424. 23. Carpenter TO, Gundberg CM. Osteocalcin abnormalities in Hyp mice reflect altered genetic expression and are not due to altered clearance, affinity for mineral, or ambient phosphorus levels. Endocrinology 1996; 137:5213. 24. Glorieux FH, Pierre JM, Pettifor JM, Delvin EE. Bone response to phosphate salts, ergocalciferol, and calcitriol in hypophosphatemic vitamin D resistant rickets. N Engl J Med 1980; 303:1023. 25. Harrell RM, Lyles KW, Harrelson JM, et al. Healing of bone disease in X-linked hypophosphatemic rickets/osteomalacia. J Clin Invest 1985; 75:1858. 26. Carpenter TO, Mitnick MA, Ellison A, et al. Nocturnal hyperparathyroidism: a frequent feature of X-linked hypophosphatemia. J Clin Endocrinol Metab 1994; 78:1383. 27. Rivkees SA, El-Hajj-Fuleihan G, Brown EM, Crawford JD. Tertiary hyper-parathyroidism during high phosphate therapy of familial hypophosphatemic rickets. J Clin Endocrinol Metab 1992; 75:1514. 28. Carpenter TO, Keller M, Schwartz D, et al. 24,25 Dihydroxyvitamin D supplementation corrects hyperparathyroidism and improves skeletal abnormalities in X-linked hypophosphatemic ricketsa clinical research center study. J Clin Endocrinol Metab 1996; 81:2381. 29. Sullivan W, Carpenter TO, Glorieux F, et al. A prospective trial of phosphate and 1,25 dihydroxyvitamin D3 therapy in symptomatic adults with X-linked hypophosphatemic rickets. J Clin Endocrinol Metab 1992; 75:879. 30. Eddy MC, McAlister WH, Whyte MP. X-linked hypophosphatemia: normal renal function despite medullary nephrocalcinosis 25 years after transient vitamin D 2-induced renal azotemia. Bone 1997; 21:515.

31. 32. 33. 34. 35. 36.

Econs MJ, McEnery PT, Lennon F, Speer MC. Autosomal dominant hypophosphatemic rickets is linked to chromosome 12p13. J Clin Invest 1997; 100:2653. O'Neill EM. Linear sebaceous naevus syndrome with oncogenic rickets and diffuse pulmonary angiomatosis. J R Soc Med 1993; 86:177. Konishi K, Nakamura M, Yamakawa H, et al. Case report: hypophosphatemic osteomalacia in von Recklinghausen neurofibromatosis. Am J Med Sci 1991; 301(5):322. Hahn SB, Lee SB, Kim DH. Albright's syndrome with hypophosphatemic rickets and hyperthyroidism: a case report. Yonsei Med J 1991; 32(2):179. Chen C, Carpenter TO, Steg N, et al. Hypercalciuric hypophosphatemic rickets: mineral balance, bone histomorphometry, and therapeutic implications of hypercalciuria. Pediatrics 1989; 84:276. Prockop DJ, Kivirikko KI. Collagens: molecular biology, diseases, and potentials for therapy. Ann Rev Biochem 1995; 64:403.

36a. Nuytinck L, Tutel T, Kayserili H, et al. Glycine to tryptophan substitution in type I collagen in a patient with OI type III: a unique collagen mutation. J Med Genet 2000; 37:371 37. Wang Q, Orrison BM, Marini JC. Two additional cases of osteogenesis imperfecta with substitutions for glycine in the alpha 2(I) collagen chain. A regional model relating mutation location in phenotype. J Biol Chem 1993; 268(33):25162. 38. Glorieux F, Bishop NJ, Plotkin H, et al. Cyclic administration of pamidronate in children with severe osteogenesis imperfecta. N Engl J Med 1998; 349:947. 39. Shaw NJ, Boivin CM, Crabtree NJ. Intravenous pamidronate in juvenile osteoporosis. Arch Dis Child 2000; 83:143.

CHAPTER 71 MORPHOLOGY OF THE ADRENAL CORTEX AND MEDULLA Principles and Practice of Endocrinology and Metabolism

CHAPTER 71 MORPHOLOGY OF THE ADRENAL CORTEX AND MEDULLA

DONNA M. ARAB O'BRIEN Embryology Anatomy Histopathology of the Adrenal Cortex Normal Histology Pathology Histopathology of the Adrenal Medulla Normal Histology Adrenal Medullary Hyperfunction Adrenal Medullary Hypofunction Chapter References

The adrenal glands are paired organs adjacent to the kidneys, containing a cortex and a medulla. These two divisions are best understood as functionally separate endocrine organs. This chapter reviews the embryology, anatomy, and histopathology of the adrenal cortex and medulla, emphasizing those aspects most relevant to clinical endocrinologists.1,2,3,4 and 5

EMBRYOLOGY
The adrenal cortex and medulla arise from separate embryologic tissues (Fig. 71-1). Beginning in the fourth week of gestation, cells destined to form the adrenal cortex develop from primitive mesoderm just medial to the urogenital ridge. These cells penetrate the overlying retroperitoneal mesenchyme and, over the next several weeks, form a gradually enlarging adrenal cortex. During the sixth week of gestation, the developing cortex is penetrated by nerve fibers, along which primitive medullary cells will eventually migrate. By the eighth week, the cortex has formed two distinct zones: a relatively large and centrally located fetal zone and a thin rim of definitive cortex, which will later form the adult adrenal cortex. Throughout this period, proliferation is largely confined to the outer definitive cortex, suggesting that the same cells give rise to both the fetal and definitive zones. About this time, the fetal adrenal circulation is established, with several adrenal arteries arising from the descending aorta, and a central vein draining each adrenal gland. Between the second and third month of gestation, the adrenal glands increase in weight from about 5 to 80 mg, are much larger than the adjacent kidney, and reach their greatest size relative to total fetal weight. Growth of the adrenal cortex after the 20th week of gestation is dependent on pituitary stimulation (anencephalic fetuses are born with an atrophic adrenal fetal zone). After birth, the fetal zone involutes and the definitive cortex evolves into the three zones of the adult adrenal cortex (Fig. 71-2).

FIGURE 71-1. Formation of the human adrenal (suprarenal) gland and sympathetic ganglia. A, Migration of sympathetic ganglion and chromaffin cells from the neural crest and neural tube. B, The formation of the sympathetic ganglia. C, The chromaffin cells of the future suprarenal medulla reaching the provisional cortex. (From Allan FD. Essentials of human embryology, 2nd ed. New York: Oxford University Press, 1969.)

FIGURE 71-2. Internal development of the human suprarenal gland. The fetal cortex (cross-hatched) decreases in volume from the fifth fetal month to its complete disappearance in the first year of postnatal life. Zonation of the permanent cortex begins with the zona glomerulosa, indicated in the eighth fetal month; complete zonation is established by the fourth year of life. (From Gray SW. Embryology for surgeons. Phila- delphia: WB Saunders, 1972:555; based on data from Sucheston ME, Cannon MS. Development of zonular patterns in the human adrenal gland. J Morphol 1968; 126:477.)

During fetal development, primitive adrenocortical cells may migrate widely. Accessory adrenocortical rests have been found adjacent to the celiac plexus, in the broad ligament, adjacent to ovarian or spermatic vessels, and around the kidney or uterus. Less common sites include the abdominal organs and, even more rarely, the lung, spinal nerves, or brain. These rests may be responsible for recurrent Cushing disease (see Chap. 75) after bilateral adrenalectomy and, rarely, for neoplastic transformation. Intratesticular or intraovarian adrenal rests represent a special problem because they may be difficult to distinguish from normal gonadal tissue. Such rests are rarely recognized in normal persons, but they may appear as gonadal enlargement in patients with Nelson syndrome or untreated congenital adrenal hyperplasia (CAH). The adrenal medulla arises from primitive sympathetic nervous system cells (sympathogonia) derived from the neuroectoderm. These cells differentiate into neuroblasts, which migrate ventrally from the neural crest to form paravertebral and preaortic sympathetic ganglia, and into pheochromoblasts, which form catecholamine-secreting (chromaffin) cells (see Fig. 71-1). In the fetus, chromaffin cells are found not only in the adrenal medulla but also throughout the sympathetic chain. Similarly, neuroblast cell clusters often are present in the fetal adrenal medulla and involute after birth. A rudimentary adrenal medulla may be seen after the sixth week of gestation, but chromaffin cells continue to migrate to this location throughout the third month. After birth, extraadrenal chromaffin tissue generally involutes, although it may persist near the origin of the celiac and mesenteric arteries. The paired organs of Zucker-kandl, located near the inferior mesenteric artery and prominent in the first year of life, represent such extraadrenal catecholaminesecreting tissue. The widespread prenatal occurrence of extraadrenal chromaffin cells accounts for the location of extraadrenal pheochromocytomas later in life.

ANATOMY
The adrenal glands are paired structures in the retroperitoneal space, lying anteromedial to the upper pole of the kidneys between the level of the 11th thoracic and the first lumbar vertebrae (Fig. 71-3). Although often abutting the kidneys, the adrenals in obese patients can be widely separated from the kidneys and other

retroperitoneal organs by fat.

FIGURE 71-3. Normal adrenal anatomy. Both adrenal glands are supplied by multiple small arteries. The right adrenal vein empties directly into the vena cava, whereas the left adrenal vein enters the left renal vein.

Normal adult adrenal glands average about 5.0 2.5 cm in length and width and are, at most, only about 0.6 cm thick. In cross-section, as seen with computerized imaging (see Chap. 88), the right gland is thickest centrally with two wings of tissue extending at each side, whereas the left gland often has a semi-lunar appearance. Each gland has a head, body, and tail. The head lies inferiorly, is the largest portion, and contains most of the medulla. The average weight of each gland is between 3 and 5 g, but this may increase by more than 50% during stress. The adrenals are supplied by numerous small arteries arising from the celiac, superior mesenteric, inferior phrenic, renal, and iliac arteries and from the aorta. These arteries anastomose over the surface of the glands, with numerous smaller unbranched arteries descending through the capsule (Fig. 71-4). Blood supplying the adrenal medulla first traverses the cortex through capillary sinusoids. Adrenal blood flow is differentially regulated in the cortex and the medulla.6 Nitric oxide is important for maintaining high basal levels of blood flow in both zones. Blood flow in the medulla is neurally regulated and correlates with catecholamine secretion. In the cortex, blood flow is not as closely linked to cortical secretory activity, although in many models, increased cortical blood flow is associated with increased adrenal steroid secretion.

FIGURE 71-4. The adrenal vasculature showing distinct arteries supplying medulla and cortex. Note transition of cortical capillaries into veins at the corticomedullary junction (CMJ). (From Breslow MJ. Regulation of adrenal medullary and cortical blood flow. Am J Physiol 1992; 262:H1317.)

The adrenal blood empties through a single central vein, which is surrounded by a cuff of normal adrenocortical tissue as it passes through the medulla. The adrenal vein enters the inferior vena cava on the right and the renal vein on the left. The acute angle between the right adrenal vein and the vena cava often makes catheterization of this vessel technically difficult.

HISTOPATHOLOGY OF THE ADRENAL CORTEX

NORMAL HISTOLOGY The adult adrenal cortex includes the outer 80% of the adrenal glands and a cuff of cortex surrounding the central vein. Three functional zones are present: glomerulosa, fasciculata, and reticularis (Fig. 71-5). The zona glomerulosa is responsible for aldosterone secretion and constitutes about 5% of the cortex. It is a discontinuous layer of nests of cells beneath the adrenal capsule. These cells are small, with a dense nucleus, a high nuclear/cytoplasmic ratio, and relatively low cytoplasmic lipid content. On electron microscopic examination, there are elongate mitochondria with abundant tubular cristae, a small amount of smooth endoplasmic reticulum (SER), and few lysosomes and microvilli.

FIGURE 71-5. Histologic features of the normal adrenal cortex. Zona glomerulosa cells (G) are distributed focally beneath the capsule (C). Clear cells of the zona fasciculata (F) are arranged in columns between the cortex (or glomerulosa) and the inner compact cells of the zona reticularis (R). A small area of adrenal medulla can be seen at the lower right (M). (From Neville AM, O'Hare MJ. The human adrenal cortex. Pathology and biologyan integrated approach. New York: Springer-Verlag, 1982:20.)

Functionally, the zonae fasciculata and reticularis form a unit, with both cell types having the capacity to secrete cortisol and androgens. Histologically, however, they are distinct. The zona fasciculata makes up about 70% of the adrenal cortex and consists of columns of cells extending from the inner reticularis zone to the glomerulosa (or to the capsule where the glomerulosa is absent). The cells are large, with a low nuclear/cytoplasmic ratio and abundant cytoplasmic lipid. During fixation, the lipid is removed, giving the cells a vacuolated appearance and hence the name clear cells. On electron microscopic examination, the cells form a continuum: from the outer zone, in which the cells have ovoid mitochondria with few internal vesicles, little SER, and few lysosomes, to the inner zone, in which the cells contain spherical mitochondria with numerous internal vesicles, more SER, and more lysosomes. The zona reticularis comprises the inner 25% of the adrenal cortex. It consists of anastomosing columns of cells that vary widely in size and contain densely granular cytoplasm and sparse lipid (hence, the name compact cells). The nuclear/cytoplasmic ratio is intermediate to that between cells of the glomerulosa and cells of the fasciculata. Electron microscopic examination reveals ovoid to elongate mitochondria with tubular to vesicular cristae, abundant SER, and numerous lysosomes. In

older persons, an increasing amount of lipofuscin may be seen, which imparts a darker color to these cells. The reason for functional zonation of the adrenal cortex has been a matter of considerable debate. Embryologic evidence suggests that all cortical cells arise from the same precursor. Although cells from each of the three zones initially have distinct steroid secretory patterns in vitro, long-term cultures demonstrate that the histologic and functional distinctions between these cells disappear.2 The blood supply to the adrenal glands flows inwardly from the capsule. This creates a gradient of increasing steroid hormone concentration from the outer to the inner cortex. Intraadrenal steroid concentrations may modulate the relative activities of the steroidogenic enzymes.7,8 and 9 Such hormone gradients, which likely increase with increasing adrenal size, may explain not only the functional zonation of the adrenal cortex but also the development of the reticularis zone during adrenarche. Electron microscopy has been useful in differentiating among cortical cell types, demonstrating transitional cells between the three zones of the adrenal cortex. It has not been useful, however, in elucidating pathologic changes in the adrenal cortex or in making the difficult distinction between benign and malignant neoplasms. PATHOLOGY ADRENAL NODULES: THE ADRENAL RESPONSE TO AGING With advancing age, the adrenal glands of most persons begin to show microscopic nodular changes within the cortex. These changes begin as nests of clear cells located peripherally in the cortex or in cortical tissue surrounding the central vein. Larger nodules compress the adjacent cortex and eventually distort the adrenal capsule. They may contain foci of compact cells and areas of fibrosis, hemorrhage, and cyst formation (Fig. 71-6). These yellow nodules have been recognized in 3% of normotensive research subjects and may reach 2 to 3 cm in diameter. Their incidence appears to increase not only with age but also with the presence of vascular damage, as is seen with hypertension and diabetes. Pigmented (black) nodules seen at autopsy appear to represent a variant of yellow nodules and contain compact (zona reticularislike) cells with increased amounts of lipofuscin.

FIGURE 71-6. Adrenal nodules. Multiple nodules are shown, separated by fibrovascular connective tissue (CT). Normal cortical architecture is totally disrupted. (Courtesy of Dr. Stephen Boudreau.)

These nodules are not neoplasms and, although they produce steroids in vitro, are not associated with adrenocortical hyperfunction. The unaffected cortex remains normal in appearance. Their main significance lies in their incidental detection on radiologic imaging. In the absence of clinical or biochemical evidence of hormonal hypersecretion, small asymptomatic adrenal nodules seen incidentally during abdominal imaging procedures are unlikely to represent clinically significant pathologic entities.10 ADRENOCORTICAL HYPERFUNCTION Response to Stress. Adrenal glands examined at autopsies after prolonged illnesses have mean weights of about 6 g and may weigh as much as 9 g. This appears to be the result of prolonged corticotropin (ACTH) stimulation. Histologically, within several hours of ACTH stimulation, clear (fasciculata) cells at the fasciculata-reticularis junction begin to lose their lipid content and take on the light-microscopic and ultrastructural appearance of compact (reticularis) cells (Fig. 71-7). Eventually, this lipid depletion may involve the entire fasciculata, and compact cells may extend from the medulla to the glomerulosa or adrenal capsule.

FIGURE 71-7. Effect of stress on the adrenal cortex. Columns of compact cells have totally replaced the clear cells of the normal zona fasciculata. (From Neville AM, O'Hare MJ. Histopathology of the adrenal cortex. J Clin Endocrinol Metab 1985; 14:791.)

Corticotropin-Dependent Cushing Syndrome. Corticotropin-dependent Cushing syndrome includes both pituitary hypersecretion of ACTH (Cushing disease) and paraneoplastic (ectopic) ACTH secretion from a benign or malignant neoplasm. The pathologic changes in the adrenal glands form a continuum from the mild stress-induced changes to the extreme changes resulting from severe, prolonged ACTH hypersecretion commonly seen with the paraneoplastic ACTH syndrome. Cushing disease accounts for about 60% of endogenous (noniatrogenic) Cushing syndrome in adults and ~35% in children (see Chap. 14, Chap. 75 and Chap. 83). The ACTH hypersecretion is generally mild but often prolonged. The adrenal glands are usually enlarged, each weighing 6 to 12 g. The cortex is widened, with a broadened zone of compact cells, suggesting a hyperplastic zona reticularis. Some of the reticularis may be replaced by lipid. The clear cells often are larger than normal, with increased lipid content. The zona glomerulosa is normal. Cortical micronodules, consisting of clusters of clear cells, also may be seen in the periphery or around the central vein (Fig. 71-8). The only difference between these micronodules and those seen with aging is that the remainder of the cortex is hyperplastic.

FIGURE 71-8. Bilateral nodular hyperplasia with cortical nodules. The adrenal glands from this patient with Cushing disease demonstrate a hyperplastic cortex (dark

areas beneath the light-colored capsule) containing multiple small light-colored nodules (arrows). (Courtesy of Dr. Stephen Boudreau.)

With prolonged ACTH stimulation, macronodular hyperplasia may occur. Clinically, this is seen as unilateral or bilateral nodules in bilaterally hyperplastic adrenal glands in patients with long-standing Cushing disease.11 Each nodule-containing gland usually weighs between 12 and 20 g, although much larger glands are possible. The nodules have macroscopic and microscopic features similar to those of the nodules seen with aging. The remainder of the cortex is hyperplastic. Although adrenal glands with macronodular hyperplasia are not capable of prolonged cortisol secretion without ACTH, there is evidence that such glands have limited autonomy compared with glands with bilateral hyperplasia that do not have nodules (see Chap. 75). Paraneoplastic ACTH secretion accounts for ~15% of endogenous Cushing syndrome in adults. The ACTH levels range from normal to greatly increased. Because of the natural history of the underlying tumor, the duration of disease is often brief (see Chap. 219). The adrenal glands frequently are markedly enlarged, weighing from 12 g to more than 20 g, but nodules usually are not present. Microscopically, columns of hypertrophic compact cells extend from the medulla to the glomerulosa or capsule. Foci of clear cells may be seen capping the compact cell columns. Corticotropin-Independent Cushing Syndrome. The category of ACTH-independent Cushing syndrome includes cortisol-secreting adrenocortical neoplasms, micronodular adrenal disease (which is seen in children and young adults), and drug-induced Cushing syndrome (which induces adrenocortical atrophy; see Chap. 75). Benign cortisol-secreting adenomas account for 10% to 15% of cases of endogenous Cushing syndrome. These tumors usually are small (<50 g). Functionally, they tend to present a pure picture of glucocorticoid excess, with little or no androgen production. Gross and microscopic appearances resemble adrenal nodules, with the important exception that the noninvolved adrenal cortex is atrophic with cortisol-secreting tumors. Most adenomas contain both clear and compact cells in various proportions, although clear cells generally predominate. Necrosis is uncommon in adrenal adenomas, presumably because of their small size. Cortisol-secreting adrenal carcinomas account for ~10% of cases of endogenous Cushing syndrome in adults but represent nearly 50% of these cases in children. Adrenocortical carcinomas are inefficient hormone producers, usually weigh more than 100 g, and may exceed 1000 g by the time they become clinically apparent. Adrenal carcinomas produce multiple hormones, and the mixed clinical presentation of Cushing syndrome and virilization, with markedly elevated serum levels of testosterone and dehydroepiandrosterone sulfate, is common. These characteristics often suggest to the clinician the malignant nature of an adrenal neoplasm, even when metastases are absent. Nevertheless, benign and malignant adrenocortical neoplasms may be difficult to distinguish. On gross examination, areas of necrosis, hemorrhage, or calcification are common with adrenal carcinoma, but these features appear to be related more to the size of the tumor than to its malignant potential. When satellite lesions, metastases, capsular penetration, or vascular invasion are present, the malignant nature of these lesions is confirmed. Often, however, none of these features is evident. Similarly, whereas compact cells often predominate in carcinomas, the cellular morphology and the degree of cellular atypia overlaps with benign tumors (Fig. 71-9). Electron microscopy also is not useful in making the distinction between benign and malignant lesions. Adrenal carcinomas metastasize most commonly to local lymph nodes, peritoneum, lung, and liver, and less commonly to bone, pleura, mediastinum, brain, and kidney.

FIGURE 71-9. Adrenal carcinoma. Marked cellular pleomorphism and atypia are seen in this clearly malignant neoplasm. (Courtesy of Dr. Stephen Boudreau.)

Micronodular adrenal disease is a rare form of Cushing syndrome seen almost exclusively in children and young adults. It occurs either as an isolated entity or in combination with other abnormalities, such as cardiac and cutaneous myxomas and a variety of pigmented skin lesions.12 The adrenal glands range from small to slightly enlarged for age. Although they have a diffusely nodular appearance on gross examination, the nodules are usually too small (<3 mm) to be apparent on radiographs. In infancy, nodular hyperplasia of the definitive and fetal zones may be seen, whereas in later childhood, the intra-cortical nodules consist largely of compact cells. The nonnodular adrenal cortex has been described as atrophic12,13 or normal1; however, ACTH levels are undetectable and the hypercortisolism is not suppressed by dexamethasone. Although the cause is unknown, these characteristics suggest that micronodular adrenal disease represents a primary adrenal disorder. Virilizing and Feminizing Adrenal Syndromes. The virilizing and feminizing adrenal syndromes include sex hormoneproducing adrenal neoplasms and CAH (see Chap. 75 and Chap. 77). Virilizing and feminizing adrenal tumors are rare. They are recognized most frequently in children and young adults, although they can occur at any age. They also may present with signs of hypercortisolism or hyperaldosteronism. As with cortisol-producing neoplasms, the larger tumors (>200 g) are more likely to be malignant, whereas those weighing <70 g usually are benign. Considerable overlap exists, however. The cells of small sex hormonesecreting neoplasms consist almost exclusively of compact cells, resembling those of normal zona reticularis. With increasing size, greater cellular atypia occurs, as well as a higher frequency of hemorrhage, necrosis, and calcification. Although certain tumors can be classified as benign or malignant based on extremes of cellular morphology, the degree of cellular and nuclear pleomorphism overlaps in benign and malignant sex hormoneproducing neoplasms. The pattern of androgen secretion by an adrenal neoplasm depends on the relative enzyme activities within the tumor. Adenomas often are relatively deficient in 17,20-lyase (desmolase) activity, resulting in diminished androgen secretion relative to cortisol. Adrenal carcinomas frequently have diminished 3b-hydroxysteroid dehydrogenase activity, resulting in diminished cortisol and mineralocorticoid secretion and enhanced secretion of androgen precursors. These patterns of enzyme activity explain the association of virilization with adrenal carcinomas.14 The CAHs are all characterized by defects in cortisol synthesis, causing ACTH hypersecretion. In untreated patients, the adrenal glands are variably enlarged, depending on the degree of enzyme deficiency and the age of the patient. In adults with 21-hydroxy-lase deficiency, individual glands weighing more than 30 g have been found. The adrenal glands are multinodular with diffuse cortical hyperplasia. The cortex is widened and convoluted, consisting almost entirely of compact cells. Hyperplasia of the zona glomerulosa usually is present in patients with coexistent mineralocorticoid deficiency. The adrenal abnormalities in some of the rarer, nonlethal forms of CAH (such as 17-hydroxylase deficiency) have not been described but are likely to be similar. Adrenal glands from patients with a deficiency of the cholesterol side chain cleavage enzyme, however, demonstrate cortical hyperplasia with exclusively clear, lipid-laden cells (congenital lipoid hyperplasia). This feature is consistent with the inability of the adrenal glands to mobilize cholesterol for the synthesis of steroid hormones. Primary Hyperaldosteronism. Although primary hyperaldosteronism includes several syndromes, pathologically these patients can be divided into those who harbor solitary adenomas (or, rarely, carcinoma) and those who have only diffuse hyperplasia. Complicating the pathologic presentation is that hypertension alone is associated with an increased frequency of nonfunctional adrenal nodule formation and that true functional adenomas also may be associated with diffuse (nodular or nonnodular) hyperplasia of the zona glomerulosa. Because patients without an adenoma do not benefit from surgical therapy, it is crucial that the differential diagnosis of hyperaldosteronism be made on biochemical grounds. Twenty percent to 50% of patients with primary hyperaldosteronism have bilateral hyperplasia as the cause of their mineralocorticoid excess (the variation in incidence depends, in part, on the criteria used to differentiate bilateral hyperplasia from low-renin essential hypertension; see Chap. 80). Most of the remaining patients have solitary adenomas. Multiple adenomas have been reported, but most of these cases, in retrospect, represented nodules. Malignant aldosterone-secreting neoplasms are rare and do not differ in size or behavior from other adrenocortical carcinomas. Aldosterone-secreting adenomas usually weigh <10 g, and fewer than 10% weigh >20 g. They have a characteristic golden yellow cut surface, which allows them to be differentiated macroscopically from other hormone-secreting adenomas. The cellular morphology of these adenomas also is distinct, with large and small clear cells, compact cells, and glomerulosa cells being present to varying degrees in the same adenoma (Fig. 71-10). Some investigators feel that this variability may be related to the influence of local microvasculature and corticosteroid concentrations on cellular structure, analogous to that which occurs in the normal adrenal cortex.1 Surprisingly, the uninvolved cortex in patients with aldosterone-secreting adenomas demonstrates hyperplasia of the zona glomerulosa. The glomerulosa may surround

the entire cortex and is increased in width. Moreover, clear-cell nodules as large as 3 cm in diameter, similar to those generally seen in patients with hypertension, may be present throughout the cortex. Removal of the adenoma results in correction of the hyperaldosteronism, which suggests that the morphologically hyperplastic zona glomerulosa does not contribute to the pathologic condition.

FIGURE 71-10. Aldosterone-secreting adenoma. Lipid-laden clear cells are arranged in nests and cords. A compressed rim of normal cortex (NC) can be seen at the top of the photograph. (Courtesy of Dr. Stephen Boudreau.)

The adrenal morphology in patients with hyperaldosteronism and bilateral hyperplasia (without an adenoma) does not differ from that described in patients with adenomas, except that the aldosterone-secreting tumor is not present. The adrenals usually are normal to moderately increased in size. Often, the cortex appears micronodular, although macronodules may be present and may cause confusion radiologically with aldosterone-secreting adenomas. Similar findings are noted in patients with secondary hyperaldosteronism. Hypertension generally is not corrected by bilateral adrenalectomy in these patients, which suggests that the morphologic changes are related to the hypertension. NONSECRETORY ADRENAL TUMORS Nonsecretory adrenal adenomas produce no symptoms and have the same clinical relevance as adrenal nodules. Nonfunctioning adrenal carcinomas are similar to hormone-producing adrenocortical carcinomas except that they are often detected even later. They usually secrete inactive steroid hormone precursors and thus are not truly nonfunctioning. They have a poor prognosis, similar to that of the cortisol-secreting adrenal carcinomas. Myelolipomas are uncommon benign adrenocortical tumors that usually occur in older persons.15 They generally are <2 cm in diameter but have been reported as large as 34 cm in diameter. Myelolipomas contain varying amounts of adipose and bone marrow cells. They present mainly in the differential diagnosis of incidental, nonsecretory adrenal masses, although larger ones may be associated with pain, abdominal fullness, or hematuria. ADRENOCORTICAL HYPOFUNCTION Secondary Adrenal Insufficiency (Corticotropin Deficiency). Secondary adrenal insufficiency includes hypopituitarism and drug-induced (iatrogenic) Cushing syndrome. Both conditions result in adrenal atrophy, the extent of which is dependent on the degree of ACTH deficiency. The adrenal glands are decreased in size. The cortex is thinned and consists of a normal zona glomerulosa; a decreased, but otherwise normal-appearing zona fasciculata; and an absent zona reticularis. When this condition is present at birth, the definitive zone is relatively normal but the fetal zone is greatly diminished in width. Primary Adrenal Insufficiency. The two principal causes of Addison disease are autoimmune adrenalitis and adrenal tuberculosis (see Chap. 76). Adrenal hemorrhage, acquired immunodeficiency syndrome, and metastatic cancer uncommonly cause sufficient adrenocortical damage to result in adrenal insufficiency. Systemic fungal infections, sarcoidosis, amyloidosis, neonatal and sex-linked adrenoleukodystrophy, and congenital adrenal hypoplasia are much less common causes.16,17 The adrenal glands in autoimmune adrenalitis are mildly to markedly reduced in size. The cortex is thinned and may be absent. The remaining cortical cells usually resemble those of the zona reticularis. Hypertrophied cells and micronodules also may be found. Often, a diffuse mononuclear cell infiltrate is present. The medulla is histologically normal. The adrenal glands in tuberculous adrenalitis are as much as fourfold enlarged. The adrenals may be adherent to surrounding tissues and, in acute cases, are replaced by a caseous semisolid exudate. The normal cellular architecture is destroyed, and caseous necrosis is present throughout the glands. A lymphocytic and giant-cell infiltrate usually is present in the acute stage. Tuberculous bacilli can be seen in about half the acute cases. Fibrosis and calcium deposition are present in varying amounts, depending on the duration of the destructive process (see Chap. 213). Clinically significant adrenal hemorrhage generally occurs in the setting of hypotension, severe illness, or systemic anticoagulation.18,19 and 20 Although adrenal hemorrhage is present in about 1% of subjects at autopsy, this is often microscopic and not associated with adrenal insufficiency. At the other extreme, extensive retroperitoneal bleeding may result from adrenal hemorrhage. Pathologically, the hemorrhage often appears to start in the medulla, with the hematoma extending outward into the cortex. Large hematomas cause capsular rupture and a retroperitoneal mass. Although areas of necrosis usually accompany the bleeding, extensive cortical destruction must occur before adrenal insufficiency results. Areas of calcification within the adrenal glands may be seen in patients who survive. Adrenal insufficiency has been described but is uncommon in patients with human immunodeficiency virus infection.21 Despite higher than normal plasma ACTH levels, basal and stimulated cortisol and aldosterone concentrations and plasma renin activity are normal in most patients with the acquired immunodeficiency syndrome.22 At autopsy, a necrotizing adrenalitis has been found to be associated with cytomegalovirus infection. The hemorrhage and necrosis are most prominent in the medulla, although the cortex also may be involved.

HISTOPATHOLOGY OF THE ADRENAL MEDULLA

NORMAL HISTOLOGY The medulla is located primarily in the head of the adrenal glands and accounts for ~10% of their normal adult weight. The chromaffin cell is the major cell type, named for the yellow-brown color imparted to its epinephrine-containing granules by chromatic salts. Chromaffin cells are rounded or polygonal, with a finely granular cytoplasm and an eccentric nucleus with prominent nucleoli. The cells are arranged in nests or cords. They are surrounded and directly innervated by preganglionic sympathetic nerve fibers and hence are analogous to norepinephrine-secreting postganglionic fibers in the remainder of the sympathetic nervous system. When specific immunostains are used, chromaffin cells have been found in the human adrenal cortex and cortical cells have been found in the adrenal medulla, which provides a cellular basis for potential intraadrenal interactions.23 The medullary blood supply is derived mainly from the capillary plexus draining the cortex. The resulting high intramedullary cortisol levels are thought to be important in inducing the enzyme that converts norepinephrine to epinephrine in the chromaffin cells. ADRENAL MEDULLARY HYPERFUNCTION Adrenal medullary hyperfunctional states include adrenal medullary hyperplasia, neuroblastoma, and pheochromocytoma. There is a close relationship between neural cells of the sympathetic ganglia and chromaffin cells. Small numbers of ganglion cells can be found in the adrenal medulla, and chromaffin cells occur in the sympathetic ganglia. Hence, neoplastic disorders affecting either cell type can arise throughout this system. The diagnosis of adrenal medullary hyperplasia depends on an increased volume of adrenal medulla in relation to the cortex. There is a diffuse nodular proliferation of normal medullary elements. This condition is seen mainly in families with multiple endocrine neoplasia type 2A and is considered a pre-neoplastic condition (see Chap. 188). Neuroblastomas, ganglioneuroblastomas, and ganglioneuromas arise from neuroblasts. These tumors form a continuum from least to most differentiated and from malignant to benign. Neuroblastoma usually is a tumor of infancy and childhood, with a median incidence at the age of 2 years. It originates in the paraganglia of the sympathetic nervous system or the adrenal medulla, and 60% arise in the abdomen. Grossly, it is a multinodular tumor with areas of hemorrhage, necrosis, and cystic degeneration.24 The cells are arranged in nests and contain small, dark-staining nuclei with little cytoplasm. Although the light-microscopic features are not necessarily

distinct from those of other childhood tumors, electron microscopic examination shows characteristic cytoplasmic neurosecretory granules. Neuroblastomas usually secrete catecholamines, their metabolites, or both. Characteristically, these tumors grow rapidly and metastasize early to local lymph nodes, the liver and other abdominal organs, and bone. Although spontaneous regression or differentiation into a more benign tumor may occur, mortality is high and treatment often includes combinations of palliative surgery, radiation, and chemotherapy.25 Ganglioneuroblastomas contain some mature ganglion cells and have a better prognosis. Ganglioneuromas are benign tumors arising from mature neuronal elements.26,27 Pheochromocytomas arise from chromaffin cells. As expected from their cellular origin, 95% or more are in the abdomen, with most being in the adrenal glands.28 Similar to chromaffin cells, however, they occasionally can be found anywhere along the sympathetic chain of ganglia from the base of the skull to the neck of the urinary bladder. Ten percent of patients with sporadic pheochromocytomas and 50% of those with familial pheochromocytomas have bilateral tumors. The pheochromocytoma that may occur in von Hippel-Lindau disease commonly is bilateral.29,30 Although only 5% to 10% are malignant, it is often impossible to distinguish benign from malignant neoplasms histologically. Pheochromocytomas are highly vascular tumors with local hemorrhage and cystic degeneration. Microscopically, they often have a chaotic pattern of pleomorphic elongated cells with prominent cytoplasmic granules (see Chap. 86). Although the light-microscopic features may not be diagnostic, electron microscopic examination demonstrates characteristic dense catecholamine secretory granules. ADRENAL MEDULLARY HYPOFUNCTION The adrenal medulla is affected by the same systemic diseases (tubercular and fungal infections, sarcoidosis, amyloidosis) as is the cortex and also is often the initial site of hemorrhage. Isolated adrenal medullary hypofunction is uncommon and usually occurs in the setting of diffuse autonomic insufficiency. Sparse amounts of medullary tissue may be found in some elderly persons, but no specific abnormality has been described. CHAPTER REFERENCES
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. Neville AM, O'Hare MJ. Histopathology of the adrenal cortex. J Clin Endocrinol Metab 1985; 14:791. Neville AM, O'Hare MJ. The human adrenal cortex. Pathology and biologyan integrated approach. New York: Springer-Verlag, 1982. Symington T. The adrenal cortex. In: Bloodworth JMB Jr, ed. Endocrine pathology. General and surgical. Baltimore: Williams & Wilkins, 1982:419. Gould VE, Sommers SC. Adrenal medulla and paraganglia. In: Bloodworth JMB Jr, ed. Endocrine pathology. General and surgical. Baltimore: Williams & Wilkins, 1982:473. Francis IR, Gross MD, Shapiro B, et al. Integrated imaging of adrenal disease. Radiology 1992; 184:1. Breslow MJ. Regulation of adrenal medullary and cortical blood flow. Am J Physiol 1992; 262:H1317. Hornsby PJ. Regulation of adrenocortical function by control of growth and structure. In: Anderson DC, Winter JSD, eds. Adrenal cortex: BIMR clinical endocrinology. Boston: Butterworth, 1985:1. Hyatt PJ. Functional significance of the adrenal zones. In: D'Agata R, Chrousos GP, eds. Recent advances in adrenal regulation and function. New York: Raven Press, 1987:35. Winter JSD. Functional changes in the adrenal gland during life. In: D'Agata R, Chrousos GP, eds. Recent advances in adrenal regulation and function. New York: Raven Press, 1987:51. Copeland PM. The incidentally discovered adrenal mass. Ann Intern Med 1983; 98:940. Doppman JL, Miller DL, Dryer AJ, et al. Macronodular adrenal hyperplasia in Cushing's disease. Radiology 1988; 166:347. Carney JA, Young WF Jr. Primary pigmented nodular adrenocortical disease and its associated conditions. The Endocrinologist 1992; 2:6. Ruder HJ, Loriaux DL, Lipsett MB. Severe osteopenia in young adults associated with Cushing's syndrome due to micronodular adrenal disease. J Clin Endocrinol Metab 1974; 39:1138. Sakai Y, Yanase T, Hara T, et al. Mechanism of abnormal production of adrenal androgens in patients with adrenocortical adenomas and carcinomas. J Clin Endocrinol Metab 1994; 78:36. Del Gaudio A, Solidoro G. Myelolipoma of the adrenal gland: report of two cases with a review of the literature. Surgery 1986; 90:293. Kelley RI, Datta NS, Dobyns WB, et al. Neonatal adrenoleukodystrophy: new cases, biochemical studies, and differentiation from Zellweger and related peroxisomal polydystrophy syndromes. Am J Med Genet 1986; 23:869. Schwartz RE, Stayer SA, Pasquariello CA, et al. Anesthesia for the patient with neonatal adrenoleukodystrophy. Can J Anaesth 1994; 41:56. Xarli VP, Steele AA, Davis PJ, et al. Adrenal hemorrhage in the adult. Medicine (Baltimore) 1978; 57:211. Rao RH, Vagnucci AH, Amico JA. Bilateral massive adrenal hemorrhage: early recognition and treatment. Ann Intern Med 1989; 110:227. Caron P, Chabanier MH, Cambus JP, et al. Definitive adrenal insufficiency due to bilateral adrenal hemorrhage and primary antiphospholipid syndrome. J Clin Endocrinol Metab 1998; 83:1437. Stolarczykk, Ruio SI, Smolyar D, et al. Twenty-four-hour urinary free cortisol in patients with acquired immunodeficiency syndrome. Metabolism 1998; 47:690. Verges B, Chavanet P, Degres J, et al. Adrenal function in HIV infected patients. Acta Endocrinol (Copenh) 1989; 121:633. Bornstein SR, Gonzalez-Hernandez JA, Erhart-Bornstein M, et al. Intimate contact of chromaffin and cortical cells within the human adrenal gland forms the basis for important intraadrenal interactions. J Clin Endocrinol Metab 1994; 78:225. Askin FB, Perlman EJ. Neuroblastoma and peripheral neuroectodermal tumors. Am J Clin Pathol 1998; 109(4 Suppl 1):S23. Katzenstein HM, Cohn SL. Advances in the diagnosis and treatment of neuroblastoma. Curr Opin Oncol 1998; 10:43. Schulman H, Laufer L, Barki Y, et al. Ganglioneuroma: an incidentaloma' of childhood. Eur Radiol 1998; 8:582. Fujiwara T, Kawamura M, Sasou S, Hiramori K. Results of surgery for a compound tumor consisting of pheochromocytoma and ganglioneuroblastoma in an adult; 5-year follow-up. Intern Med 2000; 39:58. Meideiros LJ, Wolf BC, Balogh K, Federman M. Adrenal pheochromocytoma: a clinicopathologic review of 60 cases. Hum Pathol 1985; 16:580. Chew SL, Dacie JE, Reznik RH, et al. Bilateral pheochromocytomas in von Hippel-Lindau disease. Q J Med 1994; 87:49. Couch V, Lindor HM, Karnes PS, Michels VV. von Hippel-Lindau disease.

CHAPTER 72 SYNTHESIS AND METABOLISM OF CORTICOSTEROIDS Principles and Practice of Endocrinology and Metabolism

CHAPTER 72 SYNTHESIS AND METABOLISM OF CORTICOSTEROIDS

PERRIN C. WHITE Structure and Nomenclature Biosynthesis Importation into Mitochondria Enzymes Pathways Origin of Cholesterol Secretory Products Regulation of Adrenal Corticosteroid Production Cortisol Secretion Aldosterone Secretion Androgen Secretion Distribution in the Circulation Glucocorticoids Aldosterone Adrenal Sex Steroids Alterations in Binding Proteins Metabolism Cortisol, Corticosterone, 11-Deoxycorticosterone, and 11-Deoxycortisol Aldosterone Chapter References

The adrenal glands are endocrinologically complex organs that are composed of two distinct endocrine tissues derived from different embryologic sources1 (see Chap. 71). The adrenal cortex (outer layer) constitutes 80% to 90% of the gland and is the source of the steroid hormones, whereas the adrenal medulla is the source of catecholamines. Although the adrenal cortex and medulla are in close proximity, they function independently. This chapter describes the biosynthesis, metabolism, mechanisms of action, and regulation of the steroid products of the adrenal cortex. Three major groups of hormones are produced by the adrenal cortex: mineralocorticoids, glucocorticoids, and sex steroids. Mineralocorticoids are produced primarily by the zona glomerulosa; glucocorticoids are produced by the zona fasciculata; and sex steroids originate primarily from the zona reticularis. The hormonal products of the adrenal cortex share cholesterol as a common precursor. Cholesterol is also the precursor for the gonadal steroids, vitamin D and derivatives, and the bile acids.

STRUCTURE AND NOMENCLATURE

Steroids have a common structure with 17-carbon atoms arranged in three six-membered rings and a fourth five-membered ring labeled A, B, C, and D, respectively (Fig. 72-1). Each of the 17 carbons is numbered in a standard way. Two additional carbons, numbered 18 and 19, may be attached at carbons 13 and 10, respectively. Carbon atoms 20 and 21 may be attached at the 17 position. These various additions yield three steroid families: the C18 estranes with an aromatic ring (estrogens); the C19 androstanes (androgens); and the C21 pregnanes (corticoids and progestins) (see Fig. 72-1). The steroid nucleus lies in a plane that can be modified by the addition of substituents either above or below (Fig. 72-2). The a-substituents occur below the plane (indicated by dotted lines in Fig. 72-2) and the b-substituents lie above the plane (indicated by solid lines). The A and B rings may be attached so that the substituents at positions 5 and 10 are in either the cis or trans orientations (see Fig. 72-2).

FIGURE 72-1. The steroid nucleus. The four rings are labeled A, B, C, D, and the carbons are numbered as shown. Examples are shown of steroids from each of the three structural categories of steroid hormones: the 21-carbon pregnane derivatives, the 19-carbon androstane derivatives, and the 18-carbon estrane derivatives. The names of the individual steroids shown are indicated in parentheses.

FIGURE 72-2. Cis-trans orientation of the steroid nucleus and location of the a- and b-substituents.

A multiplicity of trivial and systematic or biochemical names exist for each steroid (Table 72-1). Between 1930 and 1950, two groups under the direction of Reichstein and of Kendall isolated most of the naturally occurring steroids. Each group labeled steroids alphabetically in the order in which they were discovered, with the result that the same compound was sometimes given two different alphabetical designations. Thus, Kendall's compound A is not the same as Reichstein's compound A.

TABLE 72-1. Adrenal Steroidogenesis: Nomenclature

BIOSYNTHESIS
IMPORTATION INTO MITOCHONDRIA The rate-limiting step in steroid biosynthesis is importation of cholesterol from cellular stores to the matrix side of the mitochondria inner membrane where the cholesterol side-chain cleavage system (CYP11A, adrenodoxin, adrenodoxin reductase) is located. This is controlled by the steroidogenic acute regulatory protein (StAR),2,2a the synthesis of which is increased within minutes by trophic stimuli such as adrenocorticotropic hormone (ACTH) or, in the zona glomerulosa, by increased intracellular calcium. StAR is synthesized as a 37-kDa phosphoprotein that contains a mitochondrial importation signal peptide. However, importation into mitochondria is not necessary for StAR to stimulate steroidogenesis; to the contrary, the likelihood now appears that mitochondrial importation rapidly inactivates StAR.3 The mechanism by which StAR mediates cholesterol transport across the mitochondrial membrane is not yet known. StAR clearly is not the only protein that mediates cholesterol transfer across the mitochondrial membrane. Another protein that appears necessary (but not sufficient, at least in the adrenals and gonads) for this process is the peripheral benzodiazepine receptor, an 18-kDa protein in the mitochondrial outer membrane that is complexed with the mitochondrial voltage-dependent anion carrier in contact sites between the outer and inner mitochondrial membranes.4 This protein does not appear to be directly regulated by typical trophic stimuli but is stimulated by endozepines, peptide hormones also called diazepam-binding inhibitors. Endozepines may be regulated by ACTH to some extent, but not with a rapid time course. Thus far, whether or not a direct physical interaction exists between StAR and the peripheral benzodiazepine receptor is not clear. ENZYMES The enzymes required for adrenal steroid biosynthesis comprise two classes: cytochrome P450 enzymes and short chain dehydrogenases (Table 72-2).5 These enzymes are located in either the lipophilic membranes of the smooth endoplasmic reticulum or the mitochondrial inner membrane. As the successive biosynthetic reactions involved in steroidogenesis occur, the adrenal steroids shuttle between the mitochondria and the smooth endoplasmic reticulum (Fig. 72-3).

TABLE 72-2. Characteristics of Enzymes Involved in Adrenal Steroid Biosynthesis

FIGURE 72-3. Biosynthesis of cortisol from cholesterol. The functional names of the steroidogenic enzymes are listed in Table 72-2 and Figure 72-4.

FIGURE 72-4. The steroidogenic pathway and enzymatic steps.

CYTOCHROME P450 ENZYMES Cytochrome P450 (CYP) enzymes are responsible for most of the enzymatic conversions from cholesterol to biologically active steroid hormones.5,6 Five P450

enzymes are involved in cortisol and aldosterone synthesis (see Table 72-2). ThreeCYP11A (cholesterol desmolase, P450scc; the scc stands for side chain cleavage), CYP11B1 (11b-hydroxylase, P450c11), and CYP11B2 (aldosterone synthase, P450aldo)have been localized in mitochondria; twoCYP17 (17a-hydroxylase/17,20-lyase, P450c17) and CYP21 (21-hydroxylase, P450c21)are located in the endoplasmic reticulum (see Fig. 72-3). P450 enzymes are membrane-bound hemoproteins with molecular masses of ~50 kDa. Their name arises from their property of absorbing light at a peak wavelength of 450 nm. These enzymes are located in the membranes of the smooth endoplasmic reticulum and the mitochondrial inner membrane and are encoded by a large superfamily of genes.7 P450s are mixed function oxidases. They use molecular oxygen and reducing equivalents (i.e., electrons) provided by reduced nicotinamide-adenine dinucleotide phosphate (NADPH) to catalyze oxidative conversions of an extremely wide variety of mostly lipophilic substrates. The reducing equivalents are not accepted directly from NADPH but instead from accessory electron transport proteins. These accessory proteins are necessary because NADPH donates electrons in pairs, whereas P450s can only accept single electrons.8 Microsomal P450s use a single accessory protein, NADPH-dependent cytochrome P450 reductase. Mitochondrial P450s require two proteins; NADPH-dependent adrenodoxin reductase donates electrons to adrenodoxin, which in turn transfers them to the P450.9 Adrenodoxin (or ferredoxin) reductase, like cytochrome P450 reductase, is a flavoprotein. Adrenodoxin (or ferredoxin) contains nonheme iron complexed with sulfur. One gene encodes each accessory protein in mammals. The heart of the P450 catalytic site is a heme group that interacts with a highly conserved peptide of the P450. A completely conserved cysteine near the C terminus10 is the fifth ligand of the heme iron atom. The other axial ligand is either water or molecular oxygen. When oxygen is bound, the long axis of the oxygen molecule is parallel with the axis of the heme iron. According to one of the several proposed models of catalysis,11 the first step of the reaction is binding of substrate to oxidized (ferric, Fe3+) enzyme. One electron is donated from the accessory electron transport protein to the enzyme so that the iron is in the reduced (ferrous, Fe 2+) state. This complex binds molecular oxygen and then accepts a second electron from the accessory protein, leaving the bound oxygen molecule with a negative charge. The distal oxygen atom accepts two protons from a conserved threonine residue, possibly via a bound water molecule.12 The distal oxygen atom is then released as a water molecule, leaving the iron in the Fe3+ state. The remaining oxygen atom is highly reactive (the iron-oxygen complex is a ferryl moiety) and attacks the substrate, resulting in an hydroxylation. Cholesterol Desmolase (CYP11A, P450scc). The first enzymatic step in all steroid hormone biosynthesis is the conversion of cholesterol to pregnenolone by CYP11A (Fig. 72-4; see also Fig. 72-3). This enzyme catalyzes 20a- and 22R-hydroxylation followed by oxidative cleavage of the C2022 carbon bond of cholesterol to release the C2227 side chain.13 These reactions require a total of three oxygen molecules and six electrons. Like other mitochondrial cytochrome P450 enzymes, CYP11A receives these electrons from NADPH through the accessory proteins, adrenodoxin reductase and adrenodoxin. The three oxidations are normally performed in succession without release of hydroxylated intermediates from the enzyme. CYP11A is structurally related to the CYP11B (11b-hydroxylase) isozymes (see later). It is synthesized as a precursor protein. As is the case with other mitochondrial proteins encoded by nuclear genes, an amino-terminal peptide is required for transport to the mitochondrial inner membrane. This peptide of 39 residues is removed in the mitochondria to yield the mature protein. 17a-Hydroxylase/17,20-Lyase (CYP17, P450c17). CYP17 catalyzes conversion of pregnenolone to 17-hydroxypreg-nenolone. In rats, this enzyme also converts progesterone to 17-hydroxyprogesterone, but progesterone is not a good substrate for the human enzyme. CYP17 also catalyzes an oxidative cleavage of the 17,20 carbon-carbon bond, converting 17-hydroxypregnenolone and 17-hydroxyprogesterone to dehydro-epiandrosterone (DHEA) and androstenedione, respectively.14 A pair of electrons and a molecule of oxygen are required for each hydroxylation or lyase reaction. CYP17 is a microsomal cytochrome P450 and accepts these electrons from NADPH-dependent cytochrome P450 reductase. In the human adrenal cortex, 17a-hydroxylase activity is required for the synthesis of cortisol. The human enzyme preferentially uses pregnenolone rather than progesterone, and 3b-hydroxysteroid dehydrogenase then converts 17-hydroxy-pregnenolone to 17-hydroxyprogesterone. Because most rodents do not express CYP17 in the adrenal cortex, they secrete corticosterone as their primary glucocorticoid. Both 17a-hydroxylase and 17,20-lyase activities are required for androgen and estrogen biosynthesis. Because the same enzyme catalyzes production of cortisol in the adrenal cortex and androgens in the gonads, the relative level of 17,20-lyase activity must be independently regulated by a mechanism other than gene expression, or else excessive amounts of androgens would be synthesized in the adrenal cortex. Cytochrome b5 levels are likely to be one such mechanism, because interactions with cytochrome b5 increase 17,20-lyase activity, and mutations in cytochrome b5 interfere with androgen biosynthesis. Phosphorylation of specific serines and threonines in CYP17 may also be important.15 21-Hydroxylase (CYP21, P450c21). The enzyme 21-hydroxylase resides in the endoplasmic reticulum and is responsible for hydroxylating 17-hydroxyprogesterone to 11-deoxycortisol and progesterone to 11-deoxycorticosterone. The preferred substrate for the human enzyme is 17-hydroxyprogesterone. The structural gene encoding human CYP21 (CYP21, CYP21A2, or CYP21B) and a pseudogene (CYP21P, CYP21A1P, or CYP21A) are located in the HLA major histocompatibility complex on chromosome 6p21.3 ~30 kilobases (kb) apart, adjacent to and alternating with the C4B and C4A genes encoding the fourth component of serum complement. Both the CYP21 and C4 genes are transcribed in the same direction. CYP21 and CYP21P each contain 10 exons spaced over 3.1 kb. Their nucleotide sequences are 98% identical in exons and ~96% identical in introns. This tandem duplication is genetically unstable and frequent recombinations occur between CYP21 and CYP21P, causing congenital adrenal hyperplasia due to 21-hydroxylase deficiency16 (see Chap. 77). The two microsomal enzymes, CYP21 and CYP17, are 36% identical in amino-acid sequence and their genes have a similar intronexon organization, a finding which suggests that both genes evolved from a common ancestor.17 11 b-Hydroxylase (CYP11B1, P450c11) and Aldosterone Synthase (CYP11B2, P450aldo). Humans have distinct 11b-hydroxylase isozymes, CYP11B1 and CYP11B2, that are responsible for cortisol and aldosterone biosynthesis, respectively. In vitro, both isozymes can convert 11b-hydroxylate 11-deoxycorticosterone to corticosterone and 11-deoxycortisol to cortisol. CYP11B2 also has 18-hydroxylase and 18-oxidase activities, so that it can convert deoxycorticosterone to aldosterone. In contrast, the CYP11B1 isozyme does not synthesize detectable amounts of aldosterone from corticosterone or 18-hydroxycorticosterone.18 These isozymes are mitochondrial cytochrome P450 enzymes and are synthesized with a signal peptide that is cleaved in mitochondria. The sequences of the proteins are 93% identical. CYP11B1 and CYP11B2 are encoded by two genes on chromosome 8q21-q22. CYP11B2 is located to the left of CYP11B1 if the genes are pictured as being transcribed from left to right; the genes are located approximately 40 kb apart. The location of introns in each gene is identical to that seen in the CYP11A gene encoding cholesterol desmolase, and the predicted protein sequences of the CYP11B isozymes are each ~36% identical to that of CYP11A. CYP11B1 is expressed at high levels in normal adrenal glands. CYP11B2 is normally expressed at low levels, but its expression is dramatically increased in aldosterone-secreting tumors. Transcription of CYP11B1 is regulated mainly by ACTH, whereas CYP11B2 is regulated by angiotensin II and potassium levels (see later). Recombinations between these two genes can lead to abnormal regulation of CYP11B2 and cause hypertension, a condition termed glucocorticoid suppressible hyperaldosteronism (see Chap. 80). SHORT CHAIN DEHYDROGENASES Most short chain dehydrogenases catalyze reversible reactions.19 In the dehydrogenase direction, a hydride (i.e., a proton plus two electrons) is removed from the substrate and transferred to an electron acceptor that, depending on the enzyme, is oxidized nicotinamide-adenine dinucleotide (NAD+) or oxidized nicotinamide-adenine dinucleotide phosphate (NADP+). If, for example, the reaction involves a hydroxylated substrate, the reaction converts the hydroxyl to a keto group; NADH or NADPH and a proton are also produced. Oxoreductase reactions reverse this process. Unlike oxidations mediated by cytochrome P450 enzymes, no accessory protein is required for these reactions. Most short chain dehydrogenases contain 250 to 300 aminoacid residues. Some enzymes are found in cytosol, whereas others are located mainly in the endoplasmic reticulum. These enzymes have a conserved cofactor-binding domain near the N terminus and all have tyrosine and lysine residues near the C terminus that are crucial for catalysis.20 The lysine residue lowers the apparent pKa of the phenolic hydroxyl of tyrosine from 10.0, its value in solution, into the physiologic range. The deprotonated hydroxyl is then able to remove a proton from the hydroxyl group of the substrate, which facilitates transfer of a hydride from the substrate to the oxidized cofactor. The most important enzyme of this type in the adrenal cortex is 3b-hydroxysteroid dehydrogenase. This enzyme is only ~15% identical to most other short chain dehydrogenases, although it retains the essential structural features of this class of enzymes. Other enzymes of this type that are important in steroid metabolism include 11b-hydroxysteroid dehydrogenase, which interconverts cortisol and cortisone, and 17-ketosteroid reductase, which converts androstenedione and estrone to

testosterone and estradiol, respectively. 3b-Hydroxysteroid Dehydrogenase/D5-D4-Isomerase. Conversion of D5-3b-hydroxysteroids (pregnenolone, 17-hydroxypregnenolone, DHEA) to D4-3-ketosteroids (progesterone, 17-hydroxyprogesterone, androstenedione) is mediated by 3b-hydroxysteroid dehydrogenase (3b-HSD) in the endoplasmic reticulum. This enzyme uses NAD+ as an electron acceptor. In addition to 3b-hydroxysteroid dehydrogenase activity, the enzyme mediates a D5-D4-ene isomerase reaction that transfers a double bond from the B ring to the A ring of the steroid so that it is conjugated with the 3-keto group. Certain isozymes are also able to mediate 3-ketosteroid reduction of several substrates using NADH as an electron donor. Genomic blot analysis suggests that as many as six closely linked HSD3B genes exist in humans, but only two isozymes have actually been identified. HSD3B1 encodes the type I isozyme expressed in placenta, skin, and adipose tissue,21 whereas HSD3B2 encodes the type II isozyme expressed in the adrenals and gonads.22 The remaining genes apparently do not encode active enzymes. Any of the enzymatic conversions required for cortisol biosynthesis may be defective.5 Inherited disorders of cortisol biosynthesis are collectively termed congenital adrenal hyperplasia. These disorders are discussed in Chapter 77. 11b-Hydroxysteroid Dehydrogenase. The interconversion of cortisol and cortisone is mediated by two isozymes of 11b-hydroxysteroid dehydrogenase (11-HSD).23 The liver or 11-HSD1 isozyme is expressed at the highest levels in liver. In vivo, it acts primarily as a reductase and uses NADPH as an electron donor. It may act to buffer circulating levels of cortisol. This enzyme is responsible for the bioactivity of intrinsically inactive steroids like cortisone and prednisone when they are administered systemically. The kidney or 11-HSD2 isozyme is expressed mainly in mineralocorticoid target tissues, such as the kidney, colon and salivary glands, and also in placenta. It acts almost exclusively as a dehydrogenase, uses NAD+ as an electron acceptor, and has a much higher affinity for steroids than 11-HSD1 (10100 nmol/L versus 1 mol/L). The two isozymes are only 20% identical in amino-acid sequence. The 11-HSD2 isozyme plays a critical role in maintaining ligand specificity of the mineralocorticoid receptor, which can otherwise be activated by cortisol. Deficiency of this isozyme leads to a severe form of hypertension termed apparent mineralocorticoid excess. This is discussed in more detail in Chapter 82.

PATHWAYS
ORIGIN OF CHOLESTEROL Cholesterol, the common precursor of all adrenal steroids, is a 27-carbon compound that originates from three sources (Fig. 72-5). Most cholesterol used in steroidogenesis is derived from the lysosomal degradation of circulating low-density lipoproteins. Cholesterol-rich low-density lipoprotein binds to specific plasma membrane receptors on adrenocortical cells, is internalized by endocytosis, and is degraded to liberate free cholesterol. The cholesterol that is not used in steroidogenesis is esterified and stored within cytoplasmic vacuoles in the cortical cells. When additional steroid production is needed, the stored cholesterol esters are hydrolyzed by cytoplasmic enzymes to generate free cholesterol. Finally, adrenocortical cells produce some of their cholesterol by de novo synthesis from acetate.

FIGURE 72-5. Sources of the cholesterol used in adrenal steroidogenesis: (1) low-density lipoprotein (LDL) cholesterol, (2) release from storage vacuoles, and (3) de novo synthesis from acetate. (R, LDL receptor; CoA, coenzyme A.)

SECRETORY PRODUCTS The relative activities of the steroidogenic enzymes within a given steroid-secreting cell determine which secretory products that cell produces. For example, 17a-hydroxylase activity exists only in the zona fasciculata and reticularis. Therefore, cortisol and adrenal androgens cannot be synthesized by the zona glomerulosa. All of the precursors to the major adrenal biosynthetic products are measurable in the peripheral circulation. Because steroid hormones cannot be stored in steroidogenic cells, they are secreted immediately after their biosynthesis. Cortisol, 11-deoxy-cortisol, aldosterone, corticosterone, and 11-deoxycorticosterone are derived almost exclusively from adrenal whereas most other steroids are derived from a combination of adrenal and gonadal sources. Table 72-3 gives representative values for the production rates, basal plasma concentrations, and relative adrenal contribution to the blood production rate for the principal steroids of adrenal origin.

TABLE 72-3. Mean Blood Production Rate, Adrenal Contribution, and Normal Plasma Concentration of the Major Steroids in the Circulation

GLUCOCORTICOIDS Cortisol is the principal glucocorticoid in humans. It is synthesized primarily in the zona fasciculata with a small contribution from the zona reticularis. In humans, the five necessary transformations proceed predominantly as follows (see Fig. 72-3). Cholesterol is imported into mitochondria and converted to pregnenolone by cholesterol desmolase (CYP11A). Pregnenolone is transferred to the endoplasmic reticulum, where 17a-hydroxylase (CYP17) converts it to 17-hydroxypregnenolone. This is converted to 17-hydroxyprogesterone by 3b-hydroxysteroid dehydrogenase/D5-D4 isomerase in the endoplasmic reticulum. A further hydroxylation in the smooth endoplasmic reticulum by 21-hydroxylase (CYP21) converts 17-hydroxyprogesterone to 11-deoxycortisol. The latter is transferred back to the mitochondria and is converted to cortisol by 11b-hydroxylase (CYP11B1). This step is highly efficient, resulting in 98% conversion of 11-deoxycortisol to cortisol. The end product, cortisol, diffuses rapidly into the circulation without significant storage in adrenal cortex. MINERALOCORTICOIDS Aldosterone is the major mineralocorticoid. It is normally produced solely by the zona glomerulosa. The biosynthetic pathway of aldosterone initially parallels that of cortisol, except that 17-hydroxylation does not occur because of absent CYP17 in the zona glomerulosa. Thus, 11-deoxycorticosterone is produced in the endoplasmic reticulum rather than 11-deoxycortisol. The 11-deoxycorticosterone is transferred to the mitochondria, where a zone-specific 11b-hydroxylase isozyme (CYP11B2, aldosterone synthase) converts 11-deoxycorticosterone to corticosterone (and, to a lesser degree, 18-hydroxy, 11-deoxycorticosterone) and progressively oxidizes it to

18-hydroxycorticosterone and aldosterone. The secretion of 18-hydroxycorticosterone closely parallels that of aldosterone, and normally it is derived almost entirely from the zona glomerulosa. However, circulating levels of 11-deoxycorticosterone, 18-hydroxydeoxycorticosterone, and corticosterone originate primarily in the zona fasciculata due to the presence of a 17-deoxy pathway in that zone. Although aldosterone is the most potent endogenous mineralocorticoid, many steroids, including 11-deoxycorticosterone, 18-hydroxydeoxycorticosterone, corticosterone, and cortisol, possess a lesser degree of mineralocorticoid activity. Their metabolites, such as 19-nordeoxycorticosterone, also act as mineralocorticoids. Any of these steroids may contribute significant mineralocorticoid activity in pathologic states, such as adrenal tumor or congenital adrenal hyperplasia. However, only cortisol is thought to exert appreciable mineralocorticoid activity under normal circumstances. ANDROGENS The adrenal sex steroids are produced primarily in the zona reticularis, although the fasciculata is also capable of their production. The key enzyme in their production is CYP17, which has both 17a-hydroxylase and 17,20-lyase (oxidative cleavage of the C1720 bond) activities. This enzyme is not necessary for mineralocorticoid synthesis and is absent in the zona glomeru-losa. DHEA is the principal D5 steroid produced by the adrenal glands. It may also undergo sulfation by a sulfokinase to form DHEA sulfate (DHEAS).24 These two steroids have no intrinsic androgenic activity but can be converted to the potent androgen, testosterone. This conversion requires the enzymes 3b-hydroxysteroid dehydrogenase/D5-D4 isomerase and 17-ketosteroid reductase. Depending on the sequence of these reactions, conversion of DHEA to testosterone proceeds through either androstenedione or D5-androstenediol as an intermediate (see Fig. 72-4; see Chap. 114). In men, the testis produces such a large quantity of testosterone (~7000 g per day) that the contribution of the adrenal glands (~100 g per day) is insignificant. In women, however, the adrenal glands contribute ~50% of total circulating androgens, either directly or through the secretion of precursors such as DHEA and androstenedione that are converted peripherally to testosterone. ESTROGENS In both men and women, estrogens are derived either from the adrenal gland or the gonad. The aromatase enzyme converts androgens to estrogens: testosterone is converted to estradiol, and androstenedione to estrone.25 Aromatase activity is present both in the adrenal gland and in peripheral tissues such as fat. Therefore, circulating estrogens are produced by both adrenal and gonadal aromatization and by peripheral aromatization of adrenal and gonadal precursors. In the normal woman with intact ovarian function, the adrenal contribution to the circulating estrogens is relatively insignificant. However, in the absence of normal ovarian function, such as in postmenopausal or agonadal women, the adrenal gland may be the principal or only source of endogenous estrogen.

REGULATION OF ADRENAL CORTICOSTEROID PRODUCTION

CORTISOL SECRETION ADRENOCORTICOTROPIC HORMONE ACTH is a 39-amino-acid peptide that is produced in the anterior pituitary. It is synthesized as part of a larger molecular weight precursor peptide known as proopiomelanocortin (POMC). This precursor peptide is also the source of b-lipotropin (b-LPH). In addition, ACTH and b-LPH are cleaved further to yield a- and b-melanocyte-stimulating hormone (a-MSH and b-MSH, respectively), corticotropin-like intermediate lobe peptide (CLIP), g-LPH, b- and g-endorphin, and enkephalin (see Chap. 14). The POMC precursor peptide is found in a variety of extrahypothalamic tissues, including the gastrointestinal tract, numerous tumors, and the testis. It is secreted in small amounts from the anterior pituitary gland and does not bind significantly to the ACTH receptor. ACTH is the primary regulator of adrenal cortisol secretion. ACTH acts through a specific G proteincoupled receptor to increase levels of cAMP (cyclic or 3',5'-adenosine monophosphate).26 The cAMP has short-term effects (minutes to hours) on cholesterol transport into mitochondria2 but longer term effects (hours to days) on transcription of genes encoding the enzymes required to synthesize cortisol.27 The transcriptional effects occur, at least in part, through increased activity of protein kinase A, which phosphorylates transcriptional regulatory factors, not all of which have been identified. ACTH also influences the remainder of the steps in steroidogenesis as well as the uptake of cholesterol from the plasma lipoproteins and the maintenance of the size of the adrenal glands. In addition to these effects on the adrenal gland, it also stimulates melanocytes and results in hyperpigmentation when secreted in excess (see Chap. 14). CORTICOTROPIN-RELEASING HORMONE Corticotropin-releasing hormone (CRH) is the principal hypothalamic factor that stimulates the pituitary production of ACTH.28,29 Vasopressin also stimulates ACTH release, gizes with CRH, and is an important physiologic regulator of ACTH.30 CRH is produced in the paraventricular nuclei of the hypothalamus and is also found in other parts of the central nervous system (CNS) as well as in extra-CNS locations such as the peripheral leukocytes. Hypothalamic CRH is transported to the anterior pituitary cells by the hypophysial portal vessels. CRH activates ACTH secretion through a cAMP-dependent mechanism. CRH is secreted in a pulsatile fashion that results in the episodic secretion of ACTH and the diurnal variation of cortisol secretion. The diurnal variation of cortisol secretion also is discussed in Chapter 6 and Chapter 14. Numerous factors, such as metabolic, physical, or emotional stress, result in increased glucocorticoid secretion, which is mediated by hypothalamic secretion of CRH and vasopressin. HYPOTHALAMICPITUITARYADRENAL AXIS Cortisol is the primary regulator of resting activity of the hypothalamicpituitaryadrenal (HPA) axis by its negative feedback effects on ACTH and CRH (see Chap. 14). Furthermore, it may inhibit some of the higher cortical activities that lead to CRH stimulation. The negative feedback effects of cortisol are exerted at the level of both the hypothalamus and the pituitary. The magnitude of the cortisol response to each ACTH burst remains relatively constant, whereas the number of secretory bursts may vary. Therefore, it is the number of secretory periods of CRH and ACTH that determines the total daily cortisol secretion. The increased cortisol secretion observed during the early morning hours, and the increased secretion in response to stress, result primarily from increased CNS activity, possibly mediated by vasopressin, rather than by decreased sensitivity to the negative feedback effects of cortisol. For example, the diurnal variation in cortisol secretion is not abolished in the addisonian patient in whom little cortisol is present to suppress CRH and ACTH secretion. In addition, in major stress situations, the negative feedback effects of cortisol are overridden by increased CRH and ACTH activity, which results in an increase in cortisol secretion. ALDOSTERONE SECRETION The rate of aldosterone synthesis, which is normally 100- to 1000-fold less than that of cortisol synthesis, is regulated mainly by angiotensin II and potassium levels, with ACTH having only a short-term effect.31 Angiotensin II occupies a G proteincoupled receptor,32 activating phospholipase C. The latter protein hydrolyzes phosphatidylinositol bisphosphate to produce inositol triphosphate and diacylglycerol, which raise intracellular calcium levels and activate protein kinase C and CaM kinases. Similarly, increased levels of extracellular potassium depolarize the cell membrane and increase calcium influx through voltage-gated L -type calcium channels.33 Phosphorylation of as yet unidentified factors by CaM kinases increases transcription of the aldosterone synthase (CYP11B2) enzyme required for aldosterone synthesis.31 ANDROGEN SECRETION The mechanisms by which the adrenal steroids, DHEAS and androstenedione, are regulated are not completely understood. The compounds themselves have only weak androgenic activity, but because they serve as precursors for more active androgens, testosterone and dihydrotestosterone, they are often referred to as the adrenal androgens. Adrenarche is a maturational process in the adrenal gland that results in increased adrenal androgen secretion between the ages of 5 and 20 years.34 The process begins before the earliest signs of puberty and continues throughout the years when puberty is occurring. Histologically, it is associated with the appearance of the zona reticularis. Whereas ACTH stimulates adrenal androgen production acutely and clearly is the primary stimulus for cortisol release (see earlier), additional factors have been implicated in the stimulation of the adrenal androgens. These include a relative decrease in expression of 3-hydroxysteroid dehydrogenase in the zona reticularis,35 and possibly increases in 17,20-lyase activity due to phosphorylation or increased cytochrome b5 expression.15

DISTRIBUTION IN THE CIRCULATION

Substantial variability exists in the production rate of the various steroids by males and females (see Table 72-3). Steroid products are secreted into the circulation as free hormones, where they bind to plasma proteins36,37 (Table 72-4). Only the free hormone is biologically active. However, the amount of hormone that is potentially available to tissues is determined by the equilibrium of the free and bound fractions. Bound hormone may be freed from the binding protein as the free hormone is

metabolized or taken up into tissues. Thus, binding to plasma proteins increases the amount of hormone that can circulate in the blood and serves as a mechanism for transport of steroids in an inactive form. Binding makes the hormones more resistant to metabolism, thereby increasing the plasma half-life (t1/2).

TABLE 72-4. Steroid-Binding Proteins and Factors Affecting Their Circulating Levels

GLUCOCORTICOIDS Only 3% to 10% of circulating cortisol is in the free state. Approximately 80% to 90% is bound to a specific a2-globulin, known as corticosteroid-binding globulin (CBG) or transcortin.36,37 This 52-kDa glycoprotein binds cortisol with high affinity and specificity. It is produced primarily in the liver and has been found in all mammalian species. The association constant for a single molecule of cortisol is 3 107 L/mol. The CBG circulates in a concentration of ~3 mg/dL. The cortisol-binding capacity of CBG is ~20 g/dL, and it is increased in patients receiving oral estrogen or the adrenolytic drug mitotane (o,p'-DDD). The CBG also has high affinity for cortisone, corticosterone, 11-deoxycorticosterone, progesterone, and 17-hydroxyprogesterone. Between 5% and 10% of cortisol is bound to albumin. Albumin has a lower affinity for cortisol than CBG but a much higher capacity. The dissociation of cortisol from albumin is more rapid. Therefore, the steroid bound by albumin is more readily available to the tissues than that bound by CBG. A third binding protein, a1-acid glycoprotein, also binds cortisol, but it appears to be of minor importance. It binds principally to progesterone. ALDOSTERONE Although aldosterone does not have a specific binding protein, ~64% of the total circulating aldosterone is in a bound form. Approximately 17% is weakly bound to CBG. Aldosterone also binds to red blood cells, albumin (47%), and other plasma proteins. It dissociates quickly from the plasma proteins to which it binds, so that it is readily available to its target tissues. ADRENAL SEX STEROIDS The adrenal androgens DHEA and androstenedione are bound weakly to plasma proteins, predominantly albumin. DHEAS, however, is tightly bound to albumin, explaining its slow metabolic clearance rate and high, stable levels in plasma. The sex steroids testosterone and estradiol are bound by a distinct binding protein, known as sex hormonebinding globulin (SHBG) or testosterone-binding globulin (TeBG).36,37 The relative binding affinity to TeBG, in order of decreasing affinity, is dihydrotestosterone, testosterone, androstenediol, androstanediol, estradiol, and estrone.

ALTERATIONS IN BINDING PROTEINS

Generally, it is the total level of a hormone that is measured. Many factors can alter binding globulin concentrations in plasma38 (see Table 72-4). Because the free hormone determines biologic activity, knowledge of the conditions affecting the binding proteins is important in assessing total serum steroid levels. Sex steroidsin particular, estrogenincrease CBG levels. Therefore, in pregnant women or women taking oral contraceptives, higher circulating serum cortisol levels are expected, although free cortisol and cortisol activity should be normal. Thyroid hormone also stimulates CBG synthesis. Some families appear to have inherited elevations in CBG levels; elevations may also be seen in diabetes mellitus. Subnormal levels are seen in states of decreased protein production, such as in liver disease, or in states of protein loss (e.g., the nephrotic syndrome). Some patients have heritable varients in their CBG, resulting in low cortisol-binding affinity.38a Hypothyroidism, multiple myeloma, and obesity have all been associated with low circulating CBG levels. The circulating concentration of TeBG is affected by age, body weight, and circulating estrogen, testosterone, and thyroid hormone levels.

METABOLISM
The liver and the kidney are the principal organs involved in clearing the steroid hormones from the circulation. Although most tissues can metabolize steroids, the liver is the primary site of steroid hormone metabolism, and the kidney is the primary site of steroid hormone excretion. Unconjugated steroids that are filtered by the kidney are largely reabsorbed. Hepatic metabolism accomplishes two functions: a decrease in the biologic activity of the hormones; and an increase in their water solubility, because of conversion to a hydrophilic form that can be excreted in urine. Additional metabolism and excretion may take place in the gut, although most of the metabolized products are reabsorbed. Because plasma cortisol is bound with such high affinity, it is relatively protected from degradation. The plasma t1/2 of cortisol is 60 to 100 minutes. This contrasts with many other steroids, such as aldosterone, DHEA, androstenedione, testosterone, and estradiol, that have metabolic clearance rates of ~2000 L per day, corresponding to a t1/2 of under 20 minutes. Although these compounds also circulate in a bound form, they dissociate more rapidly from their binding proteins, making them more susceptible to degradation. The episodic secretion of these adrenal hormones, combined with this short t1/2, results in wide fluctuations in plasma steroid concentrations. By contrast, the steroid sulfates, such as DHEAS, which bind with high affinity to albumin, are cleared slowly from the circulation and have high stable plasma concentrations. The pathways involved in metabolism of the steroids yield more than 50 different metabolic products formed by approximately a dozen distinct enzymatic reactions. CORTISOL, CORTICOSTERONE, 11-DEOXYCORTICOSTERONE, AND 11-DEOXYCORTISOL Metabolized products of cortisol can be measured in the urine (Fig. 72-6). Unmetabolized cortisol comprises only ~0.1% of the total urinary cortisol metabolites. At least 90% of the cortisol and cortisone metabolites are excreted as the sulfate or the glucuronide conjugates. The metabolism of corticosterone, 11-deoxycorticosterone, and 11-deoxycortisol resembles that of cortisol. The major reactions involved in the metabolism of cortisol include (a) reduction of the C45 double bond, from either the a or b side of the A ring; this reaction is irreversible and produces dihydrocortisol, which is further rapidly metabolized by (b) reduction of the C3 ketone group; this reaction yields tetrahydrocortisol, which is rapidly (c) conjugated with glucuronic acid at the 3-hydroxyl position, yielding a highly water soluble form; (d) tetrahydrocortisol also undergoes reduction at the C20 position to yield cortol.39 Another important pathway of cortisol metabolism includes oxidation at the 11b-hydroxyl group to yield cortisone (see Chap. 73). Cortisone may then be metabolized by steps a, b, c, and d, as outlined for cortisol. The resulting products include tetrahydrocortisone, cortolone, and the conjugated products of each. In infants, 6b-hydroxylation is an important metabolic pathway. Moreover, 16a-hydroxylation, which is important for the conversion of estradiol to estriol, may be a minor pathway in the metabolism of cortisol. Finally, cleavage of the C17 side chain may take place, producing C19 metabolites that have a ketone group at the C17 position, such as 11b-hydroxyan-drostenedione and 11-ketoetiocholanolone, which are measured as 17-ketosteroids.

FIGURE 72-6. Major metabolic pathways for cortisol. Reduction of the double bond at C4 produces urinary metabolites with both the 5b and 5a orientations.

ALDOSTERONE Aldosterone has a plasma t1/2 of <15 minutes. Less than 0.5% of the aldosterone is secreted into the urine in the free state. It is metabolized in a fashion similar to cortisol to yield tetrahydroaldosterone and its conjugate, tetrahydroaldosterone glucuronide. It also is conjugated in the C18 position to yield aldosterone glucuronide.40 Urinary assays for aldosterone usually measure this metabolite. Further metabolic pathways for aldosterone, such as oxidation at the 11b hydroxyl group, play a smaller role. When aldosterone is in aqueous solution, the 11b hydroxyl group is in a hemiketal conformation with the adjacent 18-aldehyde group and, thus, is protected from oxidation. This fact allows aldosterone to occupy the mineralocorticoid receptor in target tissues, whereas cortisol is oxidized to cortisone by 11b-hydroxysteroid dehydrogenase. CHAPTER REFERENCES
1. Mesiano S, Jaffe RB. Developmental and functional biology of the primate fetal adrenal cortex. Endocr Rev 1997; 18:378. 2. Stocco DM, Clark BJ. Regulation of the acute production of steroids in steroidogenic cells. Endocr Rev 1996; 17:221. 2a.Stocco DM. The role of the STAR protein in steroidogenesis: challenges for the future. J Endocrinol 2000; 164:247. 3. Arakane F, Kallen CB, Watari H, et al. The mechanism of action of steroidogenic acute regulatory protein (StAR). StAR acts on the outside of mitochondria to stimulate steroidogenesis. J Biol Chem 1998; 273:16339. 4. Papadopoulos V. Structure and function of the peripheral-type benzodiazepine receptor in steroidogenic cells. Proc Soc Exp Biol Med 1998; 217:130. 5. White PC. Genetic diseases of steroid metabolism. Vitam Horm 1994; 49:131. 6. Nebert DW, Nelson DR, Coon MJ, et al. The P450 superfamily: update on new sequences, gene mapping, and recommended nomenclature [published erratum in DNA Cell Biol 1991 10(5):397]. DNA Cell Biol 1991; 10:1. 7. Schenkman JB, Greim H. Cytochrome P450. Berlin: Springer-Verlag, 1993. 8. Sevrioukova IF, Peterson JA. NADPH-P-450 reductase: structural and functional comparisons of the eukaryotic and prokaryotic isoforms. Biochimie 1995; 77:562. 9. Lambeth JD, Seybert DW, Lancaster JRJ, et al. Steroidogenic electron transport in adrenal cortex mitochondria. Mol Cell Biochem 1982; 45:13. 10. Nelson DR, Strobel HW. On the membrane topology of vertebrate cytochrome P-450 proteins. J Biol Chem 1988; 263:6038. 11. Rein H, Jung C. Metabolic reactions: mechanisms of substrate oxygenation. In: Schenkman JB, Greim H, eds. Cytochrome P450. Berlin: Springer-Verlag, 1993:105. 12. Ravichandran KG, Boddupalli SS, Hasemann CA, et al. Crystal structure of hemoprotein domain of P450BM-3, a prototype for microsomal P450's. Science 1993; 261:731. 13. Lambeth JD, Kitchen SE, Farooqui AA, et al. Cytochrome P450scc-substrate interactions. Studies of binding and catalytic activity using hydroxycholesterols. J Biol Chem 1982; 257:1876. 14. Yanase T, Simpson ER, Waterman MR. 17 alphahydroxylase/17,20-lyase deficiency: from clinical investigation to molecular definition. Endocr Rev 1991; 12:91. 15. Miller WL, Auchus RJ, Geller DH. The regulation of 17,20 lyase activity. Steroids 1997; 62:133. 16. Speiser PW, White PC. Congenital adrenal hyperplasia due to steroid 21-hydroxylase deficiency. Clin Endocrinol (Oxf) 1998; 49:411. 17. Picado-Leonard J, Miller WL. Cloning and sequence of the human gene for P450c17 (steroid 17 alpha-hydroxylase/17,20 lyase): similarity with the gene for P450c21. DNA 1987; 6:439. 18. White PC, Curnow KM, Pascoe L. Disorders of steroid 11 beta hydroxylase isozymes. Endocr Rev 1994; 15:421. 19. Persson B, Krook M, Jornvall H. Characteristics of short-chain alcohol dehydrogenases and related enzymes. Eur J Biochem 1991; 200:537. 20. Obeid J, White PC. Tyr-179 and Lys-183 are essential for enzymatic activity of 11 beta-hydroxysteroid dehydrogenase. Biochem Biophys Res Commun 1992; 188:222. 21. Lachance Y, Luu-The V, Labrie C, et al. Characterization of human 3beta-hydroxysteroid dehydrogenase/delta 5-delta 4-isomerase gene and its expression in mammalian cells [published erratum in J Biol Chem 1992; 267(5):3551]. J Biol Chem 1990; 265:20469. 22. Lachance Y, Luu-The V, Verreault H, et al. Structure of the human type II 3beta-hydroxysteroid dehydrogenase/delta 5-delta 4 isomerase (3 beta-HSD) gene: adrenal and gonadal specificity. DNA Cell Biol 1991; 10:701. 23. White PC, Mune T, Agarwal AK. 11b-hydroxysteroid dehydrogenase and the syndrome of apparent mineralocorticoid excess. Endocr Rev 1997; 18:135. 24. Longcope C. Dehydroepiandrosterone metabolism. J Endocrinol 1996; 150(Suppl):S125. 25. Simpson ER, Mahendroo MS, Means GD, et al. Aromatase cytochrome P450, the enzyme responsible for estrogen biosynthesis. Endocr Rev 1994; 15:342. 26. Mountjoy KG, Robbins LS, Mortrud MT, Cone RD. The cloning of a family of genes that encode the melanocortin receptors. Science 1992; 257:1248. 27. Waterman MR, Bischof LJ. Cytochromes P450 12: diversity of ACTH (cAMP)-dependent transcription of bovine steroid hydroxylase genes. FASEB J 1997; 11:419. 28. Orth DN. Cushing's syndrome [published erratum in N Engl J Med 1995; 332(22):1527]. N Engl J Med 1995; 332:791. 29. Itoi K, Seasholtz AF, Watson SJ. Cellular and extracellular regulatory mechanisms of hypothalamic corticotropin-releasing hormone neurons. Endocr J 1998; 45:13. 30. Scott LV, Dinan TG. Vasopressin and the regulation of hypothalamic-pituitary-adrenal axis function: implications for the pathophysiology of depression. Life Sci 1998; 62:1985. 31. Rainey WE, White PC. Functional adrenal zonation and regulation of aldosterone biosynthesis. Curr Opin Endocrinol Diabetes 1998; 5:175. 32. Matsusaka T, Ichikawa I. Biological functions of angiotensin and its receptors. Annu Rev Physiol 1997; 59:395. 33. Barrett PQ, Bollag WB, Isales CM, et al. Role of calcium in angiotensin II-mediated aldosterone secretion. Endocr Rev 1989; 10:496. 34. McKenna TJ, Fearon U, Clarke D, Cunningham SK. A critical review of the origin and control of adrenal androgens. Baillires Clin Obstet Gynaecol 1997; 11:229. 35. Gell JS, Carr BR, Sasano H, et al. Adrenarche results from development of a 3beta-hydroxysteroid dehydrogenase-deficient adrenal reticularis. J Clin Endocrinol Metab 1998; 83:3695. 36. Rosner W. The functions of corticosteroid-binding globulin and sex hormonebinding globulin: recent advances. Endocr Rev 1990; 11:80. 37. Hammond GL. Molecular properties of corticosteroid binding globulin and the sex-steroid binding proteins. Endocr Rev 1990; 11:65. 38. Hammond GL. Determinants of steroid hormone bioavailability. Biochem Soc Trans 1997; 25:577. 38a.Emptoz-Bonneton A, Cousin P, Seguehik, et al. Novel human corticoster-oid-binding globulin variant with low cortisol-binding affinity. J Clin Endocrinol Metab 2000; 85:361. 39. Brownie AC. The metabolism of adrenal cortical steroids. In: James VH, ed. The adrenal gland, 2nd ed. New York: Raven Press, 1992:209. 40. Morris DJ, Brem AS. Metabolic derivatives of aldosterone. Am J Physiol 1987; 252:F365.

CHAPTER 73 CORTICOSTEROID ACTION Principles and Practice of Endocrinology and Metabolism

CHAPTER 73 CORTICOSTEROID ACTION

PERRIN C. WHITE General Mechanisms of Action Actions of the Glucocorticoids Carbohydrate Metabolism Lipid Metabolism Protein Metabolism Immunologic Effects Effects on Skin Effects on Bone and Calcium Circulatory and Renal Effects Central Nervous System Effects Growth Effects on Other Hormones Actions of the Mineralocorticoids Actions of the Adrenal Androgens Chapter References

GENERAL MECHANISMS OF ACTION

The steroid hormones, vitamin D, retinoic acid, and the thyroid hormones all share a similar mechanism of action.1,2 These hormones diffuse through the target cell membrane and interact with a specific receptor protein for each hormone. The activated hormone-receptor complex binds to specific DNA sequences, the hormone-responsive elements (HREs), which are usually located in the 5' flanking region of each hormone-responsive gene. These complexes may also bind to other transcription factors. The binding of the hormone-receptor complex to these DNA sequences or transcription factors leads to selective increases or decreases in gene transcription. The altered protein levels that result from this change in transcription rate are responsible for the hormonal response seen in that particular tissue.3 At least six classes of steroid receptors exist, corresponding to the known bioactivities of the steroid hormones: glucocorticoid, mineralocorticoid, progestin, estrogen, androgen, and vitamin D. Additional orphan receptors of incompletely understood function are found that bind related compounds such as androstanes.4 Steroid receptors belong to a larger superfamily of nuclear transcriptional factors that includes the thyroid hormone and retinoic acid receptors. All of these receptors share a common structure that includes a carboxy-terminal ligand-binding domain and a midregion DNA-binding domain. The latter domain contains two zinc fingers, each of which consists of a loop of amino acids stabilized by four cysteine residues chelating a zinc ion.5 Unliganded steroid hormone receptors shuttle between the cytoplasm and the cell nucleus. Importation into the nucleus is an energy-dependent process. This process requires one or more nuclear localization signal sequences on the receptor, which consist of clusters of basic amino-acid residues located in or near the DNA-binding domain. When not occupied by ligand, the various hormone receptors differ in their propensity to be transported to the nucleus. For example, the estrogen receptor is predominantly located within the nucleus, whereas the unoccupied glucocorticoid and mineralocorticoid receptors are found mainly in the cytosol.6 The cytosolic glucocorticoid receptor, when not bound to its steroid ligand, forms a heterooligomer with two molecules of heat shock protein (HSP) 90 and one molecule each of HSP 70 and HSP 56 (immunophilin).7 Binding of ligand changes the conformation of the receptor and, thus, has several effects. HSP 90 is associated with the unliganded glucocorticoid receptor at the ligand-binding domain and dissociates from the receptor complex after glucocorticoid binds to the receptor. A dimerization region that overlaps the steroid-binding domain is exposed, promoting dimerization of the occupied receptor. Finally, a hormone-dependent nuclear localization signal located in a hinge between the DNA and steroid-binding domains is activated, which leads to increased importation of occupied receptors into the nucleus. The occupied receptors are then able to bind DNA and/or other transcription factors and modulate transcription of various genes.8,9 Glucocorticoids affect transcription of a wide variety of genes through several different mechanisms.8 First, the glucocorticoid-receptor complex can stimulate transcription by binding to specific glucocorticoid-responsive elements (GREs) in the 5' flanking region of glucocorticoid-responsive genes. GREs, like other specific hormone response elements, are often imperfect palindromes (in a palindrome, the two complementary strands of a DNA molecule, when read in opposite directions, have the identical sequence). Most often, GREs are variants of the sequence GGTACAnnnTGTTCT, where n is any nucleotide. The existence of two half-sites separated by three nucleotides suggests that glucocorticoid receptors interact with GREs as dimers, with one monomer binding to each half-site. However, many GREs consist of isolated half-sites or half-sites with variable spacing between them. Moreover, marked variations in sequence can be tolerated in one half-site. Thus, monomeric glucocorticoid receptors can also bind DNA, but the binding can apparently be stabilized by interactions with other bound receptor molecules or other transcription factors. Thus, binding of the monomeric receptor to one half-site markedly increases the ability of a second monomer to bind to the other half-site. The interaction of the glucocorticoid receptor and DNA has been studied in detail by x-ray crystallography and nuclear magnetic resonance techniques.5 The two zinc fingers form a single domain. Alpha helices adjacent to each finger on the carboxy-terminal side are oriented perpendicularly to each other; the first helix fits into the major groove of the DNA helix and makes direct contact with bases. The tips of both fingers contact the phosphate backbone, and the second finger also mediates DNA-dependent dimerization of the receptor. GREs cannot constitute the only DNA sequences mediating the transcriptional effects of glucocorticoids. GREs are indistinguishable in sequence from the elements binding mineralocorticoid, progestin, and androgen receptors, and these receptors are >90% identical in amino-acid sequence in their DNA-binding domains. However, the amino-terminal domains of these receptors are <15% identical in amino-acid sequence, and at least some interactions with other transcriptional factors are mediated by this domain.10 As a second type of effect, glucocorticoid receptors can inhibit or activate transcription by interacting with other transcription factors.8,9,11 In particular, they can regulate gene activity by repressing gene transcription mediated by AP-1 or NF-kB elements in the regulatory regions of some genes. These AP-1 and NF-kB sites bind cFos-cJun or RelA-p50 hetero-dimers, respectively. The ligand-bound glucocorticoid receptor monomer and/or dimer interacts with AP-1 or NF-kB and prevents them from exerting their transactivational effects on the genes they normally regulate. AP-1 and NF-kB serve as intra-cellular messenger systems for many growth factors and inflammatory cytokines, respectively. The profound antigrowth and antiinflammatory effects of glucocorticoids are exerted to a great extent via transrepression of these transcription factors. In addition, glucocorticoid receptors may modulate effects of the Stat4, Stat5, NF-1, Oct-1, SP-1, C/EBP, HNF3, and HNF4 transcription factors. Unlike glucocorticoids, mineralocorticoids do not appear to interfere with cFos-cJun or NF-kB binding. This functional difference may be localized to the amino-terminal domain of the receptor.10 Two new classes of nuclear proteins that influence the transacti-vational activity of nuclear receptors have been identified and collectively called coregulators.12,13 According to their ability to potentiate or diminish the activity of nuclear receptors, they are respectively called coactivators and corepressors. Known coregulators are large proteins with many functional domains. One could think of coactivators as bridges between the DNA-bound nuclear receptor and components of the transcription machinery, such as ancillary factors of DNA polymerase II, that stabilize and hence stimulate the activity of the initiation complex. In addition, coactivators have enzymatic activities that promote transcription, such as histone acetyl-transferase activity, which loosens the DNA double helix from the nucleosome and allows the polymerase complex to exert its activity.14 On the other hand, corepressors prevent the nuclear receptor from binding to DNA and/or transactivating their target genes and have enzymatic activities that impede transcription, such as histone deacetylase, which strengthens the interactions of the DNA with the nucleosome. Coregulators are expressed in a tissue-specific fashion and have varying degrees of specificity for particular nuclear receptors. Some of these proteins serve as coregulators of other transcription factors, such as AP-1, NF-kB, and the Stats, and hence serve as cross-points between different signal transduction systems in the cell. Several factors regulate tissue-specific effects of steroids at several levels both before and after the receptor. Most obviously, hormone receptors are widely but not ubiquitously expressed, and a particular class of steroid fails to have effects on cells that do not express the corresponding receptor. Of physiologic importance, enzymes may increase or decrease the affinity of steroids for their receptors and thus modulate their activity. For example, the mineralocorticoid receptor has identical affinities in vitro for cortisol and aldosterone, yet cortisol is a weak mineralocorticoid in vivo. This discrepancy may result from the action of 11b-hydroxysteroid dehydrogenase, which converts cortisol to cortisone. Cortisone is not a ligand for the receptor, whereas aldosterone is not a substrate for the enzyme. Pharmacologic

or genetic inhibition of this enzyme allows cortisol to occupy renal mineralocorticoid receptors and produce sodium retention and hypertension.15 Whereas different steroids may share bioactivities because of their ability to bind to the same receptor, a given steroid may exert diverse biologic effects in different tissues. The diversity of hormonal responses is determined by the different genes that are regulated by the hormone in different tissues. Glucocorticoids, for example, have primarily GRE-mediated metabolic effects in liver and mainly antiNF-kBmediated antiinflammatory properties in lymphoid tissue.16 In addition to the actions resulting from the binding of steroids to nuclear steroid receptors, some effects might be mediated through other mechanisms. Such effects often take place with extreme rapidity (milliseconds to minutes) and/or have been documented not to require protein synthesis, a sine qua non of the transcriptional effects mediated by nuclear-hormone receptors. These effects have been most extensively documented for 1,25-dihydroxyvitamin D3, progesterone, and aldosterone; they appear to involve second messengers systems including protein kinase C, intracellular calcium levels, nitric oxide, and tyrosine kinases.17 Thus far, however, no steroid-specific membrane receptors have been isolated or cloned. (Also see Chap. 4 and Chap. 54.)

ACTIONS OF THE GLUCOCORTICOIDS

Glucocorticoids are essential for survival. The term glucocorticoid refers to the glucose-regulating properties of these hormones. However, the glucocorticoids have multiple effects that include an important role in carbohydrate, lipid, and protein metabolism (Table 73-1). They also regulate immune, circulatory, and renal function. They influence growth, development, bone metabolism, and central nervous system (CNS) activity.

TABLE 73-1. Major Glucocorticoid Actions

In stress situations, glucocorticoid secretion can increase up to almost 10-fold.18,19 This increase is believed to enhance survival by increasing cardiac contractility, cardiac output, sensitivity to the pressor effects of the catecholamines and other pressor hormones, work capacity of the skeletal muscles, and capacity to mobilize energy through gluconeogenesis, proteolysis, and lipolysis. Persons with unrecognized adrenal insufficiency are at risk of life-threatening adrenal crisis if subjected to stress without glucocorticoid replacement.20 CARBOHYDRATE METABOLISM The daily secretion rate of cortisol varies little in the absence of stress. Cortisol interacts in a permissive fashion with many other hormones, including insulin, glucagon, catecholamines, and growth hormone, to achieve full homeostasis. For example, glucocorticoids are essential for normal epinephrine- or glucagon-stimulated lipolysis, gluconeogenesis, and glycogenolysis.21,22 Excess cortisol increases hepatic glycogen and glucose production and decreases glucose uptake and utilization in the peripheral tissues. These effects combine to cause hyperglycemia. This may lead to overt diabetes in persons who have a decreased capacity to produce insulin. By contrast, glucocorticoid deficiency decreases glucose production and hepatic glycogen content and may cause hypoglycemia. However, serum glucose levels may be normal in the chronically ill patient with Addison disease because of a compensatory decrease in insulin secretion. The primary action of the glucocorticoids on carbohydrate metabolism is to increase glucose production by increasing hepatic gluconeogenesis. Gluconeogenesis uses substrates derived from glycolysis, proteolysis, and lipolysis. Lactate is derived from glycolysis in muscle. Alanine is the primary substrate derived from proteolysis; fatty acids and glycerol are derived from lipolysis. In addition to inducing gluconeogenic enzymes, glucocorticoids stimulate glycolysis, proteolysis, and lipolysis, thus providing more substrate for gluconeogenesis. Glucocorticoids also increase cellular resistance to insulin, thereby decreasing entry of glucose into the cell. This inhibition of glucose uptake occurs in adipocytes, muscle cells, and fibroblasts. (Also see Chap. 75 and Chap. 139.) In addition to opposing insulin action, glucocorticoids may work in parallel with insulin to protect against long-term starvation by stimulating glycogen deposition and production in liver. Both hormones stimulate glycogen synthetase activity and decrease glycogen breakdown. LIPID METABOLISM Glucocorticoids increase free fatty acid levels by enhancing lipolysis, decreasing cellular glucose uptake, and decreasing glycerol production, which is necessary for reesterification of fatty acids. This increase in lipolysis is also stimulated through the permissive enhancement of the lipolytic action of other factors such as epinephrine. This action affects adipocytes differently according to their anatomic locations. In the patient with glucocorticoid excess, fat is lost in the extremities, but it is increased in the trunk (centripetal obesity), neck, and face (moon facies).23 This may involve effects on adipocyte differentiation.24 PROTEIN METABOLISM The glucocorticoids generally exert a catabolic/antianabolic effect on protein metabolism. This proteolysis in fat, skeletal muscle, bone, and lymphoid and connective tissue increases amino-acid substrates that can be used in gluconeogenesis. In muscle, the type II white glycolytic fibers are more affected than the type I fibers. Cardiac muscle and the diaphragm are almost entirely spared from this catabolic effect. IMMUNOLOGIC EFFECTS Glucocorticoids play a profound role in immune regulation.16,18 At high concentrations, they inhibit most immunologic and inflammatory responses. Although these effects may have beneficial aspects, they may also be detrimental to the host by inducing a state of immunosuppression that predisposes to infection. Glucocorticoids inhibit eicosanoid and glycolipid synthesis and the actions of bradykinin. They also block histamine and proinflammatory cytokine (tumor necrosis factor a, interleukin-1, and interleukin-6) secretion and effects.25 These actions inhibit vasoactive agents and diminish the inflammatory process. Glucocorticoids may cause lymphocytopenia with a relative T-cell depletion, monocytopenia, and eosinopenia. They do so at least in part by inducing cell cycle arrest in the G1 phase and by activating the apoptosis pathway through glucocorticoid receptormediated effects.26 In contrast, glucocorticoids increase circulating polymorphonuclear cell counts, mostly by preventing their egress from the circulation. Generally, glucocorticoids decrease diapedesis, chemotaxis, and phagocytosis of polymorphonuclear cells. Thus, the mobility of these cells is altered such that they do not arrive at the site of inflammation to mount an appropriate immune response. Some of these effects may be mediated by changes in levels of the cytokine migration inhibitory factor (MIF) from macrophages and T cells. Whereas physiologic levels of glucocorticoids promote release of MIF, pharmacologic doses inhibit MIF secretion.27 The suppressive effect of glucocorticoids is primarily exerted on T helper 1 cells and hence on cellular immunity, whereas the T helper 2 cells are spared, which effectively leads to a predominantly humoral immune response.16,28 Indeed, glucocorticoids enhance secondary anamnestic antibody responses, whereas they inhibit primary antibody responses. Pharmacologic doses of glucocorticoids may also decrease the size of the immunologic tissues (i.e., the spleen, thymus, and lymph nodes). In summary, high levels of glucocorticoids decrease inflammatory and cellular immune responses and increase susceptibility to certain bacterial, viral, fungal, and parasitic infections.

EFFECTS ON SKIN Glucocorticoids inhibit fibroblasts, which leads to increased bruising and poor wound healing through cutaneous atrophy. This effect explains the thinning of the skin that is seen in patients with Cushing syndrome29 (see Chap. 75). Immunosuppressive effects of glucocorticoids makes them effective for skin conditions such as psoriasis. EFFECTS ON BONE AND CALCIUM Glucocorticoids have the overall effect of decreasing serum calcium and have been used in emergency therapy for certain types of hypercalcemia (see Chap. 59). This hypocalcemic effect probably results from a decrease in the intestinal absorption of calcium and a decrease in the renal reabsorption of calcium and phosphorus. The serum calcium level, however, generally does not fall below normal because of the secondary increase in parathyroid hormone secretion. The most significant effect of long-term glucocorticoid excess on calcium and bone metabolism is osteoporosis.30 Glucocorticoids inhibit osteoblastic activity by decreasing the number and activity of osteoblasts.31 Glucocorticoids also decrease osteoclastic activity, but to a lesser extent, leading to low bone turnover with an overall negative balance. The tendency of glucocorticoids to lower serum calcium and phosphate levels causes secondary hyperparathyroidism. Together, these actions decrease bone accretion and cause a net loss of bone mineral. CIRCULATORY AND RENAL EFFECTS Glucocorticoids have a positive inotropic influence on the heart, increasing the left ventricular work index. Moreover, they have a permissive effect on the actions of epinephrine and norepinephrine on both the heart and the blood vessels. In the absence of glucocorticoids, decreased cardiac output and shock may develop; in states of glucocorticoid excess, hypertension is frequently observed. This may be due to activation of the mineralocorticoid receptor (see later) that occurs when renal 11b-hydroxysteroid dehydrogenase is saturated by excessive levels of glucocorticoids. CENTRAL NERVOUS SYSTEM EFFECTS Glucocorticoids readily penetrate the bloodbrain barrier and have direct effects on brain metabolism. They decrease CNS edema and are commonly used in therapy for increased intracranial pressure. Paradoxically, they also may contribute to the development of pseudotumor cerebri (increased intracranial pressure in the absence of a structural lesion). Their effects on mood and behavior are well recognized. They stimulate appetite and cause insomnia with a reduction in rapid eye movement sleep. An increase in irritability and emotional lability is seen, with an impairment of memory and ability to concentrate. Libido is decreasedan effect that may be secondary to both a direct glucocorticoid effect on behavior and glucocorticoid-induced inhibition of the reproductive system. Glucocorticoid excess and deficiency may both be associated with clinical depression. Furthermore, glucocorticoid excess may produce a psychosis in some patients. Mild to moderate glucocorticoid excess for a limited period of time often causes a feeling of euphoria or well-being. Therefore, patients may object to a decrease in glucocorticoid dosage. Patients who have primary psychiatric disorders, such as depression or anorexia nervosa, may have abolition of the normal circadian pattern of glucocorticoid secretion and an increase in glucocorticoid production. These abnormalities are reversible with remission of the psychiatric illness and have been referred to as states of pseudo-Cushing syndrome (see Chap. 201). Glucocorticoid effects in brain are mediated largely through interactions with two closely related receptors, sometimes referred to as type I and type II receptors. The type II receptor is the conventional glucocorticoid receptor. The type I receptor is identical to the mineralocorticoid receptor, but it has identical affinities for both glucocorticoids and mineralocorticoids in most areas of the brain in which it is expressed because of lack of concomitant expression of 11b-hydroxysteroid dehydrogenase15 (see later and Chap. 72); this enzyme oxidizes glucocorticoids to inactive compounds in mineralocorticoid target tissues. The type I receptor, which is expressed at highest levels in the limbic system (i.e., the hippocampus),32 has a 10-fold higher affinity for glucocorticoids than the type II receptor. Thus, glucocorticoid effects in the limbic system may be mediated predominantly through the type I receptor at normal levels and may activate the type II receptor mainly at elevated levels as seen under stress conditions.32,33 These receptors have divergent and in some cases opposing effects. For example, activation of the type I receptors reduces sensitivity of hippocampal neurons to the neurotrans-mitter serotonin whereas activation of the type II receptors increases it.34 Increased sensitivity of the hippocampus to serotonin may help explain the euphoria associated with high doses of glucocorticoids. In an analogous manner, glucocorticoids suppress release of corticotropin-releasing hormone (CRH) in the anterior hypothalamus but stimulate it in the central nucleus of the amygdala and lateral bed nucleus of the stria terminalis, where it may mediate fear and anxiety states.35 In addition, glucocorticoids and other steroids may have non-genomic effects by modulating activities of both g-amino butyric acid (GABA) and N-methyl-D-aspartate (NMDA) receptors.36 GROWTH In excess, glucocorticoids inhibit linear growth and skeletal maturation in children.37 This results primarily from the direct inhibitory effect of glucocorticoids on the epiphyses. This effect may in part be mediated by decreasing levels of insulin-like growth factor-I (IGF-I)31 and by increasing levels of IGF-binding protein-1 (IGFBP-1), which inhibits somatic growth by decreasing circulating levels of free IGF-I.38 Also, chronic glucocorticoid excess has been associated with inhibition of growth hormone secretion.39 Although excess glucocorticoids clearly impair growth, glucocorticoids are necessary for normal growth and development. In the fetus and neonate, they accelerate the differentiation and development of various tissues. Their actions include promoting the development of the hepatic and gastrointestinal systems as well as the production of surfactant in the fetal lung (see Chap. 202). Glucocorticoids are routinely given to pregnant women at risk for delivery of premature infants in an effort to accelerate these maturational processes. EFFECTS ON OTHER HORMONES Glucocorticoids have several effects on pituitary function.19,40 They primarily regulate adrenocorticotropic hormone (ACTH) secretion through glucocorticoid (type II) receptormediated effects at the hypothalamic and pituitary levels (see Chap. 72). In addition, they affect both thyroid and reproductive function. Thyroid Function. Glucocorticoids blunt the thyroid-stimulating hormone response to thyrotropin-releasing hormone stimulation. They decrease the peripheral conversion of thyroxine (T4) to triiodothyronine (T3) with a concomitant increase in reverse T3. A decrease in both thyroid-binding globulin and thyroid-binding prealbumin is seen. The sum of these effects is usually a low-normal total T4 and free T4 level, without clinical manifestations of hypothyroidism. Gonadal Function. The glucocorticoids inhibit gonadotropin secretion both in the basal state and in response to gonadotropin-releasing hormone. These actions cause a decrease in gonadal sex steroid production.41 The glucocorticoids also have direct inhibitory effects on the gonad, and they lead to a decrease in libido.42 Together, these actions impair reproductive function.

ACTIONS OF THE MINERALOCORTICOIDS

In order of decreasing potency, the mineralocorticoids include aldosterone, 11-deoxycorticosterone, 18-oxocortisol, corticosterone, and cortisol. As a class of hormones, they have more specific actions than the glucocorticoids. Their major function is to maintain intravascular volume by conserving sodium and eliminating potassium and hydrogen ions. They exert these actions in kidney, gut, and salivary and sweat glands. In addition, aldosterone may have distinct effects in other tissues. Mineralocorticoid receptors are found in the heart and vascular endothelium,43 and aldosterone increases myocardial fibrosis.44 Such receptors are also found in the brain (see earlier); although they are occupied predominantly by glucocorticoids in some regions of the brain, aldosterone has distinct effects in increasing salt hunger and blood pressure when administered into the cerebral ventricles.45 The distal convoluted tubules and collecting ducts of the kidney are the major sites of action of the mineralocorticoids.46 In the cortical collecting duct, mineralocorticoids induce reabsorption of sodium and secretion of potassium. In the medullary collecting duct, they act in a permissive fashion to allow vasopressin to increase osmotic water flux. Patients with mineralocorticoid deficiency may develop weight loss, hypotension, hyponatremia, and hyperkalemia, whereas patients with mineralocorticoid excess may

develop hypertension, hypokalemia, and metabolic alkalosis47 (see Chap. 79, Chap. 80 and Chap. 81). The mechanisms by which aldosterone affects sodium excretion are incompletely understood. Most effects of aldosterone are presumably due to changes in gene expression mediated by the mineralocorticoid receptor; indeed, levels of subunits of both the Na+/K+ adenosine triphosphatase and the epithelial sodium channel (ENaC) increase in response to aldosterone. However, the mineralocorticoid receptor has not been shown to bind to hormone-response elements or to influence the transcription of these or any other genes. Moreover, the effects of aldosterone on the activity of the ENaC are considerably more rapid than these effects on gene expression, suggesting that aldosterone also (or perhaps mainly) affects translocation of channels to the apical cell membrane or the percentage of time the channels stay open. Such effects might be mediated by post-translational modifications to the regulatory subunits of ENaC, including proteolysis, phosphorylation, or carboxymethylation.48 These changes might be mediated by the mineralocorticoid receptor or they might represent nongenomic effects, which have also been implicated in effects of aldosterone on cardiovascular function and leukocyte cell volume.17 Because the mineralocorticoid receptor actually has comparable affinities for cortisol and aldosterone, yet circulating levels of aldosterone are normally much lower than those of cortisol, aldosterone is able to have discrete bioeffects only because the 11b-hydroxysteroid dehydrogenase enzyme is able to oxidize glucocorticoids, but not aldosterone, to inactive steroids.15

ACTIONS OF THE ADRENAL ANDROGENS

Many actions of adrenal androgens are exerted through their conversion to active androgens or estrogens such as testosterone, dihydrotestosterone, estrone, and estradiol. In men, <2% of the biologically important androgens are derived from adrenal production, whereas in women ~50% of androgens are of adrenal origin. The adrenal contribution to circulating estrogen levels is mainly important in pathologic conditions, such as feminizing adrenal tumors. Adrenal androgens contribute to the physiologic development of pubic and axillary hair during normal puberty. They also play an important role in the pathophysiology of congenital adrenal hyperplasia, premature adrenarche, adrenal tumors, and Cushing syndrome (see Chap. 75, Chap. 77 and Chap. 83). In humans, circulating levels of dehydroepiandrosterone (DHEA) and dehydroepiandrosterone sulfate (DHEAS), the chief adrenal androgens, reach a peak in early adulthood and then decline.49 This has led to speculation that age-related physiologic changes might be reversed by DHEA administration, and beneficial effects have been proposed for insulin sensitivity, bone mineral density, muscle mass, cardiovascular risk, obesity, cancer risk, autoimmunity, and the central nervous system.50 Many of these effects have been observed in rodents, which normally do not synthesize significant amounts of these steroids, but the extent to which they exist in humans is controversial at this time. CHAPTER REFERENCES
1. Mangelsdorf DJ, Thummel C, Beato M, et al. The nuclear receptor superfamily: the second decade. Cell 1995; 83:835. 2. Luisi BF, Schwabe JW, Freedman LP. The steroid/nuclear receptors: from three-dimensional structure to complex function. Vitam Horm 1994; 49:1. 3. Katzenellenbogen JA, O'Malley BW, Katzenellenbogen BS. Tripartite steroid hormone receptor pharmacology: interaction with multiple effector sites as a basis for the cell- and promoter-specific action of these hormones. Mol Endocrinol 1996; 10:119. 4. Forman BM, Tzameli I, Choi HS, et al. Androstane metabolites bind to and deactivate the nuclear receptor CAR-beta. Nature 1998; 395:612. 5. Freedman LP. Anatomy of the steroid receptor zinc finger region. Endocr Rev 1992; 13:129. 6. Akner G, Wikstrom AC, Gustafsson JA. Subcellular distribution of the glucocorticoid receptor and evidence for its association with microtubules. J Steroid Biochem Mol Biol 1995; 52:1. 7. Pratt WB, Toft DO. Steroid receptor interactions with heat shock protein and immunophilin chaperones. Endocr Rev 1997; 18:306. 8. Bamberger CM, Schulte HM, Chrousos GP. Molecular determinants of glucocorticoid receptor function and tissue sensitivity to glucocorticoids. Endocr Rev 1996; 17:245. 9. Truss M, Beato M. Steroid hormone receptors: interaction with deoxyribonucleic acid and transcription factors. Endocr Rev 1993; 14:459. 10. Pearce D, Yamamoto KR. Mineralocorticoid and glucocorticoid receptor activities distinguished by nonreceptor factors at a composite response element. Science 1993; 259:1161. 11. McEwan IJ, Wright AP, Gustafsson JA. Mechanism of gene expression by the glucocorticoid receptor: role of protein-protein interactions. Bioessays 1997; 19:153. 12. Shibata H, Spencer TE, Onate SA, et al. Role of co-activators and co-repressors in the mechanism of steroid/thyroid receptor action. Recent Prog Horm Res 1997; 52:141. 13. Horwitz KB, Jackson TA, Bain DL, et al. Nuclear receptor coactivators and corepressors. Mol Endocrinol 1996; 10:1167. 14. Spencer TE, Jenster G, Burcin MM, Allis CD, Zhou J, Mizzen CA, et al. Steroid receptor coactivator-1 is a histone acetyltransferase. Nature 1997; 389:194. 15. White PC, Mune T, Agarwal AK. 11b-hydroxysteroid dehydrogenase and the syndrome of apparent mineralocorticoid excess. Endocr Rev 1997; 18:135. 16. Chrousos GP. The hypothalamic-pituitary-adrenal axis and immune-mediated inflammation. N Engl J Med 1995; 332:1351. 17. Wehling M. Specific, nongenomic actions of steroid hormones. Annu Rev Physiol 1997; 59:365. 18. McEwen BS, Biron CA, Brunson KW, et al. The role of adrenocorticoids as modulators of immune function in health and disease: neural, endocrine and immune interactions. Brain Res Brain Res Rev 1997; 23:79. 19. Chrousos GP. Stressors, stress, and neuroendocrine integration of the adaptive response. The 1997 Hans Selye Memorial Lecture. Ann N Y Acad Sci 1998; 851:311. 20. Lamberts SW, Bruining HA, de Jong FH. Corticosteroid therapy in severe illness. N Engl J Med 1997; 337:1285. 21. McMahon M, Gerich J, Rizza R. Effects of glucocorticoids on carbohydrate metabolism. Diabetes Metab Rev 1988; 4:17. 22. Pilkis SJ, Granner DK. Molecular physiology of the regulation of hepatic gluconeogenesis and glycolysis. Annu Rev Physiol 1992; 54:885. 23. Bjorntorp P. Body fat distribution, insulin resistance, and metabolic diseases. Nutrition 1997; 13:795. 24. MacDougald OA, Lane MD. Transcriptional regulation of gene expression during adipocyte differentiation. Annu Rev Biochem 1995; 64:345. 25. Almawi WY, Beyhum HN, Rahme AA, Rieder MJ. Regulation of cytokine and cytokine receptor expression by glucocorticoids. J Leukoc Biol 1996; 60:563. 26. Cidlowski JA, King KL, Evans-Storms RB, et al. The biochemistry and molecular biology of glucocorticoid-induced apoptosis in the immune system. Recent Prog Horm Res 1996; 51:457. 27. Bucala R. MIF rediscovered: cytokine, pituitary hormone, and glucocorticoid-induced regulator of the immune response. FASEB J 1996; 10:1607. 28. Norbiato G, Bevilacqua M, Vago T, Clerici M. Glucocorticoids and Th-1, Th-2 type cytokines in rheumatoid arthritis, osteoarthritis, asthma, atopic dermatitis and AIDS. Clin Exp Rheumatol 1997; 15:315. 29. Yanovski JA, Cutler GBJ. Glucocorticoid action and the clinical features of Cushing's syndrome. Endocrinol Metab Clin North Am 1994; 23:487. 30. Lane NE, Lukert B. The science and therapy of glucocorticoid-induced bone loss. Endocrinol Metab Clin North Am 1998; 27:465. 31. Delany AM, Dong Y, Canalis E. Mechanisms of glucocorticoid action in bone cells. J Cell Biochem 1994; 56:295. 32. de Kloet ER, Vreugdenhil E, Oitzl MS, Joels M. Brain corticosteroid receptor balance in health and disease. Endocr Rev 1998; 19:269. 33. Lupien SJ, McEwen BS. The acute effects of corticosteroids on cognition: integration of animal and human model studies. Brain Res Brain Res Rev 1997; 24:1. 34. Meijer OC, de Kloet ER. Corticosterone and serotonergic neurotransmission in the hippocampus: functional implications of central corticosteroid receptor diversity. Crit Rev Neurobiol 1998; 12:1. 35. Schulkin J, Gold PW, McEwen BS. Induction of corticotropin-releasing hormone gene expression by glucocorticoids: implication for understanding the states of fear and anxiety and allostatic load. Psychoneuroendocrinology 1998; 23:219. 36. Revelli A, Tesarik J, Massobrio M. Nongenomic effects of neurosteroids. Gynecol Endocrinol 1998; 12:61. 37. Allen DB. Growth suppression by glucocorticoid therapy. Endocrinol Metab Clin North Am 1996; 25:699. 38. Lee PD, Giudice LC, Conover CA, Powell DR. Insulin-like growth factor binding protein-1: recent findings and new directions. Proc Soc Exp Biol Med 1997; 216:319. 39. Bertherat J, Bluet-Pajot MT, Epelbaum J. Neuroendocrine regulation of growth hormone. Eur J Endocrinol 1995; 132:12. 40. Kononen J, Honkaniemi J, Gustafsson JA, Pelto-Huikko M. Glucocorticoid receptor colocalization with pituitary hormones in the rat pituitary gland. Mol Cell Endocrinol 1993; 93:97. 41. Chrousos GP, Torpy DJ, Gold PW. Interactions between the hypothalamic-pituitary-adrenal axis and the female reproductive system: clinical implications. Ann Intern Med 1998; 129:229. 42. Michael AE, Cooke BA. A working hypothesis for the regulation of steroidogenesis and germ cell development in the gonads by glucocorticoids and 11b-hydroxysteroid dehydrogenase (11bHSD). Mol Cell Endocrinol 1994; 100:55. 43. Lombes M, Alfaidy N, Eugene E, et al. Prerequisite for cardiac aldosterone action. Mineralocorticoid receptor and 11b-hydroxysteroid dehydrogenase in the human heart. Circulation 1995; 92:175. 44. Young M, Fullerton MJ, Dilley R, Funder JW. Mineralocorticoids, hypertension, and cardiac fibrosis. J Clin Invest 1994; 93:2578. 45. Gomez-Sanchez EP. Central hypertensive effects of aldosterone. Front Neuroendocrinol 1997; 18:440. 46. Rogerson FM, Fuller PJ. Mineralocorticoid action. Steroids 2000; 65:61. 47. White PC. Disorders of aldosterone biosynthesis and action. N Engl J Med 1994; 331:250. 48. Garty H, Palmer LG. Epithelial sodium channels: function, structure, and regulation. Physiol Rev 1997; 77:359. 49. Lamberts SW, van den Beld AW, van der Lely AJ. The endocrinology of aging. Science 1997; 278:419. 50. Baulieu EE, Robel P. Dehydroepiandrosterone (DHEA) and dehydroepiandrosterone sulfate (DHEAS) as neuroactive neurosteroids. Proc Natl Acad Sci U S A 1998; 95:4089.

CHAPTER 74 TESTS OF ADRENOCORTICAL FUNCTION Principles and Practice of Endocrinology and Metabolism

CHAPTER 74 TESTS OF ADRENOCORTICAL FUNCTION

D. LYNN LORIAUX Tests for Documenting Hypercortisolism Urinary Free Cortisol Porter-Silber Chromogens Plasma Cortisol Tests for the Differential Diagnosis of Hypercortisolism Dexamethasone Suppression Test Metyrapone Test Corticotropin-Releasing Hormone Test Tests for Adrenal Hypofunction Tests for Congenital Adrenal Hyperplasia Tests for Localization Petrosal Sinus Sampling Computed Tomography Magnetic Resonance Imaging Iodocholesterol Scan Chapter References

The diagnosis of adrenocortical disease depends on accurate and precise laboratory measurements. The optimal application of the available tests is an evolving process. This chapter concentrates on those tests that the author believes are most useful. The discussion focuses on three categories of test: tests used to document hypercortisolism or hypocortisolism; tests used for the differential diagnosis of hypercortisolism or hypocortisolism; and tests used to localize disease. Further details concerning the rationale and interpretation of adrenal function tests are found in Chapter 14, Chapter 17, Chapter 75, Chapter 76, Chapter 77, Chapter 78, Chapter 79, Chapter 80, Chapter 81, Chapter 82, Chapter 83, Chapter 237 and Chapter 241.

TESTS FOR DOCUMENTING HYPERCORTISOLISM

URINARY FREE CORTISOL Urinary free cortisol is cortisol in urine that is not conjugated to amino acids or glucuronic acid and, hence, is free. Urinary free cortisol retains its preferential solubility in organic solvents and can be extracted from urine with diethyl ether or dichloromethane. After extraction, the organic solvent is evaporated and the residual cortisol is measured by one of several techniques. The most popular is radioimmunoassay (RIA).1 The advantage of the measurement of urinary free cortisol over other techniques used to document hypercortisolism is sensitivity. Patients with hypercortisolism rarely fall into the normal range.2,3 and 4 The ranges of urinary free cortisol excretion for normal persons and persons subsequently proved to have hypercortisolism overlap by only 1% to 2%. This overlap is considerably greater with other methods. The physiologic mechanism underlying this high degree of specificity is probably the relationship between corticosteroid-binding globulin (CBG) binding capacity and the normal cortisol production rate. The binding capacity of CBG is nearly saturated at normal rates of cortisol production. At greater than normal rates of cortisol secretion, the unbound fraction of plasma cortisol increases disproportionately. Because urinary free cortisol is derived from the unbound or bioactive cortisol by glomerular filtration, the free cortisol concentration in the urine rises rapidly as the binding capacity of CBG is exceeded. Thus, the urinary free cortisol level directly reflects the bioactive plasma cortisol concentration. The most important variable affecting the excretion of urinary free cortisol is its reabsorption by the nephron. Because >95% of filtered cortisol is reabsorbed, small changes in the reabsorbed fraction can make large differences in the quantitative urinary excretion of the hormone. Fortunately, this is not a complicating variable in most cases of hypercortisolism. The normal values for urinary free cortisol, as with all tests, vary with the individual laboratory, but the normal basal excretion of urinary free cortisol typically is <100 g/d. PORTER-SILBER CHROMOGENS Porter-Silber chromogen measurement often is termed the 17-hydroxycorticosteroids test. Measurement of Porter-Silber chromogens and 17-hydroxycorticosteroids, however, are different tests. In the United States, the available test virtually always is the Porter-Silber chromogens test.5 In this procedure, the side chain of cortisol (carbons 17, 20, and 21) is measured. This side chain can be thought of as dihydroxyacetone. Dihydroxyacetone forms a colored adduct with phenylhydrazine that has an absorption maximum at 410 nm, which allows the concentration to be measured using spectrophotometry (Fig. 74-1). The measurement involves, first, a hydrolysis step to free the steroid of its conjugate and render it lipid soluble; second, an extraction with an organic solvent; and, third, a color reaction followed by spectrophotometric measurement. The main steroids measured by this test are metabolites of cortisol, not cortisol itself (Fig. 74-2).

FIGURE 74-1. Principle of measurement of Porter-Silber chromogens.

FIGURE 74-2. Urinary corticosteroids measured by Porter-Silber chromogen technique.

The complexity of the test offers several opportunities for artifact. Examples include druginduced interference with the hydrolysis step (e.g., by aspirin)6; decreased efficiency of organic extraction subsequent to increased polarity of the metabolites of cortisol, resulting from a druginduced alteration in hepatic metabolism (e.g., by phenytoin, phenobarbital)7,8; and interference with the color reaction by medications and their metabolites (e.g., some of the minor tranquilizers). Many other

unrecognized hazards can render the test unreliable. Because most of these hazards can be attributed to medications, the test always should be performed with the patient in a drugfree state. The normal range for Porter-Silber chromogens remains constant throughout life, at 4.5 2.5 mg/g creatinine per day.9 Approximately one-third of the daily production of cortisol finds its way into the Porter-Silber chromogen pool. A note of caution: Porter-Silber chromogen excretion has a diurnal variation, whereas creatinine excretion does not. Thus, inadequate urine collections cannot be corrected on the basis of the creatinine measurement. Collections that are deficient early in the day give artificially low values when corrected in this assay, and collections that are deficient late in the day give artificially high values. PLASMA CORTISOL Cortisol is secreted episodically. Normally, 8 to 10 bursts of cortisol are secreted each day. These bursts tend to cluster in the morning hours, which leads to a diurnal pattern in the concentration of plasma cortisol. 10 This feature of cortisol secretion makes the interpretation of a single plasma cortisol value hazardous. An additional confounding factor is that cortisol circulates bound to CBG. The concentration of this a-globulin modulates the concentration of plasma cortisol at any given production rate. Thus, factors that change plasma CBG concentration (e.g., estrogens) ultimately are reflected in the plasma concentration of cortisol.11 Plasma cortisol is measured almost exclusively by RIA. This technique involves the extraction of cortisol from plasma with an organic solvent and subsequent measurement based on a competition with radiolabeled cortisol for binding sites on antibodies with high affinity and specificity for cortisol. The technique is fast, inexpensive, and reliable. The physician should be aware, however, that, depending on the antibody used, molecules other than cortisol can compete for binding sites. The most important of these are 11-deoxycortisol (especially during the administration of metyrapone) and synthetic glucocorticoids such as prednisone and dexamethasone.1

TESTS FOR THE DIFFERENTIAL DIAGNOSIS OF HYPERCORTISOLISM

The traditional differential diagnostic tests for adrenal hyper-function, which focus on establishing the presence or absence of an intact feedback loop, are the dexamethasone suppression test and the metyrapone test. These tests are difficult to interpret and are quickly becoming obsolete. Because metyrapone is now hard to obtain, this test is included here mainly for historical interest. DEXAMETHASONE SUPPRESSION TEST As developed by Liddle and colleagues, the dexamethasone suppression test measures the response of the Porter-Silber chromogens to the exogenous administration of dexamethasone.12 Urine collections are made for 6 days. Basal, or control, collections are made on days 1 and 2 of the test. Dexamethasone is given at a dose of 2 mg per day on days 3 and 4, and 8 mg per day on days 5 and 6. At some point during the test, the excretion of Porter-Silber chromogens is reduced by half in 95% of patients with Cushing disease (Fig. 74-3). It is reduced by less than half in patients with other causes of Cushing syndrome. The urine free cortisol also can be used as an end point, but must decrease by 80% or more to indicate that suppression has occurred13 (see Chap. 75).

FIGURE 74-3. The pattern of Porter-Silber chromogen excretion in normal humans (lower area) and in patients with Cushing disease (upper area) in response to the administration of dexamethasone. Porter-Silber chromogens are reduced in both groups, but considerably less so in patients with Cushing disease.

METYRAPONE TEST Metyrapone is a competitive inhibitor of the 11-hydroxylase enzyme (Fig. 74-4). It blocks the biosynthesis of cortisol. This leads to a reduced plasma cortisol concentration. If the feedback axis is intact, adrenocorticotropic hormone (ACTH) secretion increases in an attempt to reestablish normal plasma cortisol concentrations. Because large amounts of 11-deoxy-cortisol are produced, 11-deoxycortisol and its metabolites raise the urinary concentration of Porter-Silber chromogens (Fig. 74-5). The traditional rule of thumb is that a doubling of the basal level of Porter-Silber chromogens indicates an intact feedback axis.14

FIGURE 74-4. Metyrapone blockade of biosynthesis of cortisol.

FIGURE 74-5. Principle of metyrapone testing. See text for explanation. (ACTH, adrenocorticotropic hormone.)

A similar pattern of changes can be seen in the plasma concentrations of cortisol and 11-deoxycortisol. During effective blockade with metyrapone, plasma cortisol concentrations fall to <5 g/dL. Plasma 11-deoxycortisol concentrations rise to over 50% of the baseline cortisol concentrations (always >7 g/dL) if the feedback axis is intact. Before the metyrapone test can be interpreted, it must be shown that a blocking dose of metyrapone has been delivered. Plasma cortisol concentrations at the detection limit of the cortisol assay are the most reliable guide for this.15 The two main variations of the metyrapone test are the standard 2-day test and the overnight test. The standard 2-day test is performed by giving 750 mg of metyrapone orally every 4 hours for 1 day. The urinary concentration of Porter-Silber chromogens is measured that day and again the next day. A doubling of the level of Porter-Silber chromogens is a rough guide to the normal response. Plasma cortisol, measured 4 hours after the last metyrapone dose, should be <5 g/dL, and plasma 11-deoxycortisol should range between 10 and 30 g/dL. The overnight metyrapone test is performed by giving metyrapone, 30 mg/kg orally, at midnight. Plasma cortisol values at 8:00 a.m. should be lower than 5 g/dL and 11-deoxycortisol levels should be >7 g/dL.16 The response to metyrapone can be altered by abnormalities of thyroid function.17 Several drugs, including estrogens and diphenylhydantoin, can invalidate the test.18,19 CORTICOTROPIN-RELEASING HORMONE TEST The corticotropin-releasing hormone (CRH) test is currently the test of choice for the differential diagnosis of Cushing syndrome.20 CRH, a 41amino-acid peptide of hypothalamic origin, plays an important role in the regulation of ACTH secretion (see Chap. 14). A single intravenous bolus of CRH stimulates the release of ACTH in normal persons (Fig. 74-6). Plasma cortisol concentrations >20 g/dL, however, prevent the release of ACTH from the normal pituitary gland. Thus, CRH fails to stimulate the release of ACTH in patients with Cushing syndrome if the cause is anything other than a pituitary tumor that is secreting ACTH. Alternatively, ACTH-secreting pituitary tumors are resistant to the feedback effects of cortisol; they retain the ability to respond to CRH, even at high plasma cortisol concentrations. A plasma ACTH concentration of >10 pg/mL after CRH administration defines ACTH-dependent hypercortisolism. Values of <10 pg/mL are characteristic of ACTH-independent hypercortisolism.21

FIGURE 74-6. Effect of ovine corticotropin-releasing hormone (oCRH) on levels of plasma adrenocorticotropic hormone (ACTH) and cortisol in normal persons.

The test is performed by administering ovine CRH, 1 g/kg, intravenously over 1 minute. Blood samples are drawn after 30, 45, and 60 minutes for the measurement of ACTH. A positive response is any value >10 pg/mL. The untoward effects of CRH are limited to a facial flush and sense of warmth, which occur in ~10% of patients. Larger doses can induce hypotension and should not be used.

TESTS FOR ADRENAL HYPOFUNCTION

The normal ranges of plasma cortisol and urinary free cortisol overlap the detection limit of most assays. In other words, a value of zero can be normal. Thus, reliable tests of adrenal insufficiency evaluate the ability of plasma cortisol to respond to an ACTH stimulus. The cosyntropin (Cortrosyn) test is the most convenient and reliable. This test is performed by administering an intravenous bolus of 250 g of synthetic b1-24 ACTH and measuring the plasma cortisol concentration 45 and 60 minutes later.22 A plasma cortisol concentration of 20 g/dL is the lower limit of the normal response. Anything less than this suggests impaired adrenal function.23 Localizing the cause of adrenal insufficiency to the adrenal gland itself or to the hypothalamic-pituitary unit can be accomplished efficiently by measuring plasma ACTH. High ACTH concentrations localize the defect to the adrenal gland; normal or low concentrations localize the defect to the hypothalamus or pituitary.24 The RIA for ACTH has been notoriously unreliable, but the commercially available two-site immunoradiometric assay measurements are of high quality.25

TESTS FOR CONGENITAL ADRENAL HYPERPLASIA

Diagnosis of the various syndromes of congenital adrenal hyperplasia depends on the measurement of the steroid biosynthetic intermediate before the blocked step (see Chap. 77). All the plasma marker steroids now can be measured by specific RIAs. The appropriate measurements include 17-hydroxy-progesterone (for 21-hydroxylase deficiency), 11-deoxycortisol (for 11-hydroxylase deficiency), 17-hydroxypregnenolone (for 3b-hydroxysteroid dehydrogenase deficiency), and progesterone or pregnenolone (for 17-hydroxylase deficiency).

TESTS FOR LOCALIZATION

Four tests are useful in localizing the site of disease in disorders of adrenal function: petrosal sinus sampling, computed tomography (CT), magnetic resonance imaging, and the iodocholesterol scan. PETROSAL SINUS SAMPLING The venous drainage of the pituitary gland can be sampled in the inferior petrosal sinuses (Fig. 74-7). The finding of an ACTH gradient between petrosal sinus blood and peripheral blood in a patient with hypercortisolism localizes the site of ACTH production to the sella or parasellar region.26,26a Because the venous drainage of the pituitary gland tends to lateralize (the right half drains into the right inferior petrosal sinus and the left half drains into the left sinus), the site of an ACTH-producing microadenoma usually can be localized within the pituitary gland. The usefulness of the test is increased by sampling during the administration of CRH: the central/peripheral ACTH gradients and the right/left gradient are enhanced by this addition. The average gradient in Cushing disease is ~15, with a lower limit of 1.7. The lower limit increases to 3 after CRH administration, which completely separates this population from those with nonpituitary causes of Cushing syndrome. Current data suggest that the test correctly identifies pituitary disease >95% of the time, which makes it the most powerful single test for differential diagnosis in Cushing syndrome.27 In previously untreated patients, the procedure correctly lateralizes the disease ~85% of the time. Untoward effects are largely confined to the catheterization process. A few patients have had transient signs and symptoms of brainstem dysfunction or other neurological findings.27a These have not led to permanent disability. Other untoward effects include bleeding, bruising, and minor infections at the venipuncture site.

FIGURE 74-7. Anatomy of venous draining of the pituitary gland.

COMPUTED TOMOGRAPHY CT has become the procedure of choice in localizing adrenal disease (see Chap. 88). Most adrenal cancers are large, and CT, for practical purposes, identifies all these, as well as nearly all the smaller adrenal adenomas. Other useful findings revealed by CT include adrenal hyperplasia, usually bilateral and symmetric, and an abnormal intraabdominal fat distribution that supports the clinical diagnosis of Cushing syndrome.28,29,30 and 31 MAGNETIC RESONANCE IMAGING Magnetic resonance imaging offers only a single advantage over CT in the differential diagnosis of adrenal disease: the classification of adrenal masses into probably benign versus probably malignant categories on the basis of the T2-weighted image.32 Normal adrenal glands are dark on T1-weighted magnetic resonance imaging scans. Benign adenomas remain dark when the scan is shifted to T2-weighting, but malignant lesions become bright. A hyperintense rim on fat-saturated spin-echo magnetic resonance imaging is often seen in adrenal adenomas and may aid in their differentiation from adrenal metastases.33 Adrenal cysts and pheochromocytomas also are bright, but usually present no problem in diagnosis when viewed in the context of other endocrine tests and the ultrasound examination for cystic structure (see Chap. 88). IODOCHOLESTEROL SCAN The iodocholesterol scan is an imaging procedure that depends on the function of the adrenal gland. It is rarely required in the differential diagnosis of hypercortisolism. Circulating cholesterol serves as the precursor for most of the steroids produced by the adrenal gland. Not surprisingly, therefore, circulating cholesterol is concentrated in the adrenal glands. Iodocholesterol, a g-emitter, can be used to examine this physiologic process. Normal and hyperactive glands image bilaterally. Adenomas causing Cushing syndrome image, but the contralateral side does not, because the resultant hypercortisolism suppresses ACTH and diminishes the function of the normal gland. Adrenal carcinomas causing Cushing syndrome fail to produce an image because of an inefficient cholesterol-concentrating mechanism. The contralateral normal side also fails to image because ACTH secretion is suppressed. Thus, if both sides are visualized, an ACTH-dependent process is implied. If one side is visualized, an adrenal adenoma is implied. If neither side is seen, adrenal carcinoma or factitious disease is suggested.34,35 and 36 CHAPTER REFERENCES
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. Ruder HJ, Guy RL, Lipsett MB. Radioimmunoassay for cortisol in plasma and urine. J Clin Endocrinol Metab 1972; 35:219. Murphy BEP. Clinical evaluation of urinary cortisol determinations by competitive protein binding radioassay. J Clin Endocrinol Metab 1968; 28:343. Hsu TH, Bledsoe T. Measurement of urinary free corticoids by competitive protein binding radioassay in hypoadrenal states. J Clin Endocrinol Metab 1970; 30:443. Beardwell CG, Burke CW, Cope CL. Urinary free cortisol measured by competitive protein binding. J Endocrinol 1968; 42:79. Silber RH, Porter CC. The determination of 17-21-dihydroxy, 20-ketosteroids in urine and plasma. J Biol Chem 1954; 210:923. Borushek S, Gold JJ. Commonly used medications that interfere with routine endocrine laboratory procedures. Clin Chem 1964; 10:41. Werk EE, MacBee T, Sholiton LJ. Effect of diphenylhydantoin on cortisol metabolism in man. J Clin Invest 1964; 43:1824. Burstein S, Klaiber EL. Phenobarbital-induced increase in 6-hydroxycortisol excretion: clue to its significance in human urine. J Clin Endocrinol Metab 1965; 25:293. Franks RC. Urinary 17-hydroxycorticosteroid and cortisol excretion in childhood. J Clin Endocrinol Metab 1973; 36:702. Krieger DT, Allen W, Rizzo F, Krieger HP. Characterization of the normal temporal pattern of plasma corticosteroid levels. J Clin Endocrinol Metab 1971; 32:266. Doe RP, Zimmerman HH, Flink EB. Significance of the concentration of nonprotein bound plasma cortisol in normal subjects, Cushing's syndrome, pregnancy, and during estrogen therapy. J Clin Endocrinol Metab 1960; 20:1484. Liddle GW, Island DP, Meador CK. Normal and abnormal regulation of corticotropin secretion in man. Recent Prog Horm Res 1962; 18:125. Flack MR, Loriaux DL, Nieman LK. Urinary free cortisol excretion: a better measure of response to the dexamethasone suppression test in Cushing's syndrome? (Abstract). New Orleans: Endocrine Society, 1988:1. Liddle GW, Estep HL, Kendall TW. Clinical application of a new test of pituitary reserve. J Clin Endocrinol Metab 1959; 19:875. Meikle AW, Jubiz W, Hutchings MD, et al. A simplified metyrapone test with determination of plasma 11-deoxycortisol (metyrapone test with plasma S). J Clin Endocrinol Metab 1969; 29:985. Jubiz W, Meikle AW, West CD, Tyler FH. A single dose metyrapone test. Arch Intern Med 1970; 125:472. Cushman P. Hypothalamic-pituitary-adrenal function in thyroid disorders: effects of metyrapone infusion on plasma corticosteroids. Metabolism 1968; 17:263. Meikle AW, Jubiz W, Matsukura S, et al. Effect of estrogen on the metabolism of metyrapone and release of ACTH. J Clin Endocrinol Metab 1970; 30:259. Jubiz W, Levinson RA, Meikle AW. Absorption and conjugation of metyrapone during diphenylhydantoin therapy: mechanism of abnormal response to oral metyrapone. Endocrinology 1970; 86:328. Chrousos GP, Schulte HM, Oldfield EH, et al. The corticotropin releasing factor stimulation test: an aid in the evaluation of patients with Cushing's syndrome. N Engl J Med 1984; 310:622. Nieman LK, Chrousos GP, Oldfield EH, et al. The corticotropin-releasing hormone stimulation test and the dexamethasone suppression test in the differential diagnosis of Cushing's syndrome. Ann Intern Med 1986; 105:862. Kehlet H, Binder C. Value of an ACTH test in assessing hypothalamic-pituitary-adrenocortical function in glucocorticoid treated patients. BMJ 1973; 2:147. Lindholm J, Kehlet H, Blichert-Toft M. Reliability of the 30-minute ACTH test in assessing hypothalamic-pituitary-adrenal function. J Clin Endo-crinol Metab 1978; 47:272. Schulte HM, Chrousos GP, Avgerinos P. The CRF test: a possible aid in the evaluation of patients with adrenal insufficiency. J Clin Endocrinol Metab 1984; 58:1064. Slyper AH, Findling JW. Use of a two-site immunoradiometric assay to resolve a factitious elevation of ACTH in primary pigmented nodular adrenocortical disease. J Pediatr Endocrinol 1994; 7:61. Oldfield EH, Chrousos GP, Schulte HM, et al. Preoperative localization of ACTH secreting pituitary microadenomas by bilateral and simultaneous inferior petrosal sinus sampling. N Engl J Med 1985; 312:100. Zovickian J, Oldfield EH, Doppman JL, et al. Usefulness of inferior petrosal sinus venous endocrine markers in Cushing's disease. J Neurosurg 1988; 68:205.

27a.Lefournier V, Gatta B, Martinie M, et al. One transient neurological complication (sixth nerve palsy) in 166 consecutive interior petrosal sinus samplings for the etiological diagnosis of Cushing's syndrome. J Clin Endocrinol Metab 1999; 84:3401. 28. 29. 30. 31. 32. 33. 34. 35. 36. Vincent JM, Morrison ID, Armstrong P, Reznek RH. The size of normal adrenal glands on computed tomography. Clin Radiol 1994; 49:453. Ganguly A, Pratt JH, Yune HY, et al. Detection of adrenal tumors by computerized tomographic scan in endocrine hypertension. Arch Intern Med 1979; 139:590. Kelestimur F, Unlu Y, Ozesmi M, Tolu I. A hormonal and radiological evaluation of adrenal gland in patients with acute or chronic pulmonary tuberculosis. Clin Radiol 1994; 49:453. White FE, White MC, Drury PL, et al. Value of computed tomography of the abdomen and chest in investigation of Cushing's syndrome. BMJ 1982; 284:771. Heinz-Peer G, Honigschabel S, Schneider B, et al. Characterization of adrenal masses using MR imaging with histopathologic correlation. AJR Am J Roentgenol 1999; 173:15. Ichikawa T, Ohtamo K, Uchiyama G, et al. Adrenal adenomas: characteristic hyperintense rim sign on fat-saturated spin-echo MR images. Radiology 1994; 193:247. Moses DC, Schteingart DS, Sturman MF, et al. Efficacy of radiocholesterol imaging of the adrenal glands in Cushing's syndrome. Surg Gynecol Obstet 1974; 139:201. Chatal JR, Charbonnel B, Le Mevel BP, et al. Uptake of 131 I-19-iodocholes-terol by an adrenal cortical carcinoma and its metastases. J Clin Endocrinol Metab 1976; 43:248. Beierwaltes WH, Sisson JC, Shapiro B. Diagnosis of adrenal tumors with radionuclide imaging. Spec Top Endocrinol Metab 1984; 6:1.

CHAPTER 75 CUSHING SYNDROME Principles and Practice of Endocrinology and Metabolism

CHAPTER 75 CUSHING SYNDROME

DAVID E. SCHTEINGART Pathophysiology Corticotropin-Dependent Cushing Syndrome Pituitary Corticotropin-Dependent Cushing Syndrome Ectopic Corticotropin Syndrome Ectopic Corticotropin-Releasing Hormone Syndrome Corticotropin-Independent Cushing Syndrome Primary Bilateral Micronodular Hyperplasia Adrenal Adenoma and Carcinoma Clinical Presentations Corticotropin-Dependent Cushing Disease Ectopic Corticotropin and Corticotropin-Releasing Hormone Syndromes Corticotropin-Independent Types of Cushing Syndrome Diagnosis Standard Diagnostic Evaluation Difficulties in Interpretation of the Diagnostic Tests Anatomic Localization Detection of Pituitary Lesions Detection of Adrenal Lesions Localization of Ectopic Corticotropin-Secreting Lesions PseudoCushing Syndrome Obesity Major Depressive Disorder Alcohol-Induced PseudoCushing Syndrome Differential Diagnosis of Cushing-Type Hypercortisolism and PseudoCushing Syndrome Corticotropin-Releasing Hormone in the Pathophysiology of PseudoCushing Syndrome Treatment of Cushing Disease Pituitary Surgery Pituitary Irradiation Adrenal Surgery Drug Therapy Treatment of Ectopic Adrenocorticotropin Syndrome Treatment of Corticotropin-Independent Cushing Syndrome Chapter References

Cushing syndrome is the clinical expression of the metabolic effects of persistent, inappropriate hypercortisolism. Since the disorder was first described by Harvey Cushing in 1932, great progress has been made in understanding its pathophysiology and the spectrum of its clinical presentation. The diagnosis can now be precisely established by biochemical means, by noninvasive imaging of the pituitary and adrenal lesions, and by selective identification of the source of corticotropin (ACTH or adrenocorticotropic hormone) in the ACTH-dependent types. The clinical manifestations of Cushing syndrome involve many organ systems and metabolic processes.1 Hypertension, obesity, diabetes, androgen-type hirsutism, and acne are commonly seen in these patients and are prevalent in patients who do not have primary cortisol excess. When present, the following findings increase the suspicion of Cushing syndrome: symptoms and signs of protein catabolism (e.g., ecchymoses, myopathy, osteopenia); truncal obesity; lanugal hirsutism; cutaneous lesions (i.e., wide, purple striae; tinea versicolor; verruca vulgaris); hyperpigmentation; and psychiatric manifestations (i.e., impairment of affect, cognition, and vegetative functions).2,3,4,5 and 6 In some instances, the diagnosis of Cushing syndrome can be strongly suspected on clinical grounds alone, but for most patients, the diagnosis is established by laboratory studies.7

PATHOPHYSIOLOGY
Cushing syndrome can be exogenous, resulting from the administration of glucocorticoids or ACTH, or endogenous, resulting from a primary increased secretion of cortisol or ACTH. An etiologic classification of Cushing syndrome is given in Figure 75-1.

FIGURE 75-1. Classification of Cushing syndrome. (ACTH, corticotropin; CRH, corticotropin-releasing hormone.)

The changes in ACTH and cortisol secretion observed in patients with ACTH-dependent and ACTH-independent Cushing syndrome are illustrated in Figure 75-2. In the ACTH-dependent types, ACTH is secreted by the pituitary gland or by an extrapituitary neoplasm capable of producing peptide hormones, including ACTH and corticotropin-releasing hormone (CRH). In pituitary-dependent Cushing disease, the feedback relationship between the pituitary and the adrenal glands is maintained but is abnormal. In the ectopic ACTH syndrome, pituitary ACTH release is suppressed by excessive production of cortisol, which is stimulated by an unsuppressible overproduction of ACTH by the ectopic source. In ectopic CRH syndrome, tumor CRH stimulates ACTH secretion by the pituitary corticotropes; this causes an excessive production of cortisol that is not sufficient to suppress ACTH secretion. In patients with ACTH-independent types of Cushing syndrome, pituitary ACTH secretion is suppressed by excessive cortisol production that originates in adrenocortical tumors (i.e., adenomas or carcinomas) or in autonomous nodular hyperplastic glands.

FIGURE 75-2. Corticotropin (ACTH) and cortisol levels in ACTH-dependent and ACTH-independent types of Cushing syndrome. (CRH, corticotropin-releasing

hormone.)

CORTICOTROPIN-DEPENDENT CUSHING SYNDROME

PITUITARY CORTICOTROPIN-DEPENDENT CUSHING SYNDROME Approximately 85% of patients with pituitary ACTH-dependent Cushing syndrome, also called Cushing disease, have pituitary microadenomas. The origin, nature, and biochemical characteristics of these microadenomas are not clear.8,9 In most cases, the microadenomas are located in the periphery of the gland and can be identified by imaging techniques or transsphenoidal pituitary exploration. Occasionally, they are located deep in the central wedge of the pituitary and can be missed on imaging or at surgery.8 The suggestion has been made that excessive ACTH secretion may be caused by neurohypothalamic stimulation, which results in corticotrope hyperplasia.10,11 However, fewer than 17% of cases of Cushing disease are associated with ACTH hypersecretion by nonneoplastic corticotrope cells.12 Whether pituitary ACTH-dependent Cushing disease develops from a primary pituitary disorder or is the result of corticotrope stimulation by CRH has been unclear. In the case of a pituitary adenoma, the nonadenomatous corticotropes are usually suppressed, and removal of the adenoma is followed by ACTH and cortisol insufficiency. In the rare cases of corticotrope hyperplasia, the possibility exists that these corticotropes have been stimulated from a hypothalamic source. Clonal analysis of the pituitary tissue obtained from patients with Cushing disease shows that ACTH-secreting adenomas have a monoclonal pattern consistent with a monoclonal proliferation of a genetically altered cell. Thus, a spontaneous somatic mutation in pituitary corticotropes is the primary pathogenetic mechanism in this disorder. In contrast, the corticotrope hyperplasia seen in patients with CRH-secreting bronchial carcinoids is polyclonal.13 Direct measurement of CRH levels in patients with Cushing disease secondary to a pituitary adenoma has not been performed. In the few patients in whom CRH levels have been measured in the spinal fluid, low levels have been found, suggesting suppressed CRH secretion. The manifestations of abnormal ACTH secretion in Cushing disease are a loss of normal negative feedback and a blunted circadian rhythm of ACTH and cortisol secretion. Although the relative sensitivity of ACTH release to corticosteroid suppression remains in most of these patients, as many as 40% exhibit resistance to suppression and resemble patients with the ectopic ACTH syndrome. Unusual abnormal patterns of release of ACTH and cortisol have been described, and episodes of hypersecretion have persisted for hours or weeks and are interrupted by periods of normal secretion.14,15 This periodic hormonogenesis has been found in patients with and without ACTH-secreting pituitary adenomas. ECTOPIC CORTICOTROPIN SYNDROME The association of Cushing syndrome with nonadrenal neoplasms has been recognized for decades and is called the ectopic ACTH syndrome.16 Typically, a rapidly developing malignancy causes the autonomous secretion of high levels of ACTH and cortisol, with clinical manifestations of Cushing syndrome and pigmentation. Approximately 50% of these ectopic ACTH-secreting tumors are in the lung, and the remainder occur in a variety of tissues. Table 75-1 describes the type and frequency of these tumors.

TABLE 75-1. Tumors Causing Ectopic Corticotropin Syndrome

Pancreatic islet cell tumors have been associated with Cushing syndrome in fewer than 5% of patients. In addition to ACTH, these tumors sometimes contain other hormones, such as insulin, gastrin, and glucagon, which are not produced in sufficient amounts to cause a metabolic syndrome. Less frequent causes of the ectopic ACTH syndrome include thymic carcinoids; gastric and appendicular carcinomas; pheochromocytomas; adrenal medullary paragangliomas; cystadenomas of the pancreas; epithelial carcinomas of the thymus; medullary thyroid carcinomas; prostatic carcinomas; carcinomas of the esophagus, ileum, and colon; ovarian tumors; squamous cell carcinoma of the lung, larynx, and cervix; and melanomas. Although the relationship between the more commonly occurring tumors and ectopic ACTH production has been well established, the production of ACTH by the less frequently occurring tumors has not been absolutely proved.17 Several mechanisms of ectopic elaboration of peptide hormones have been postulated. Neoplastic transformation of dormant stem cells results in expression by the tumor of a specific hormone gene that would not otherwise be expressed by the mature cells of that organ. An alternative hypothesis is that cells with endocrine potential migrate to various organs during embryogenesis, where they remain dormant. These cells have the capacity to initiate amine precursor uptake and decarboxylation (APUD) and synthesize a variety of peptides.18 This hypothesis does not account for the large number of ectopic hormones elaborated by non-APUD cells. An enhanced expression of genes normally encoding for proopiomelanocortin (POMC) in extrapituitary tissues is a more likely explanation for the development of ectopic hormone syndromes. These tissues include stomach, adrenal medulla, pancreas, brain, and placenta.19 The tumor cells in patients with the ectopic ACTH syndrome appear to have developed a more efficient mechanism for translation of messenger RNA (mRNA) to immunoreactive POMC-derived peptides or may have lost the mechanism for preventing the continuous expression of the POMC gene.20 Unique transcription factors may be constitutively expressed in tumor cells and cause unregulated secretion of POMC-derived peptides.21 An unusual cause of ectopic ACTH production is an ectopic pituitary adenoma. In some cases, normal pituitary tissue occurs in an ectopic location and not in continuity with intrasellar pituitary tissue. Occasionally, adenomas may develop in this ectopic pituitary tissue, causing excessive ACTH production.22 From the biochemical standpoint, these patients resemble those with pituitary ACTH-dependent or ectopic ACTH-dependent Cushing syndrome. The possibility of an ectopic pituitary adenoma in a patient presenting with an ACTH-dependent Cushing syndrome should be considered if the patient exhibits an atypical hormonal profile in the absence of an ectopic ACTH-secreting, extracranial neoplasm or in the absence of a pituitary abnormality at the time of transsphenoidal exploration. A computerized tomographic (CT) study of the parasellar region and the sphenoid sinus may help to identify such ectopically located pituitary adenomas. ECTOPIC CORTICOTROPIN-RELEASING HORMONE SYNDROME Ectopic CRH production can cause Cushing syndrome.23,24 Tumors with a capacity to secrete CRH ectopically include bronchial carcinoids, medullary thyroid carcinoma, and metastatic prostatic carcinoma. In patients with ectopic ACTH production, the pituitary ACTH content is low and only Crooke changes (i.e., hyaline degeneration) are observed in pituitary corticotropes. In contrast, patients with ectopic CRH secretion demonstrate corticotrope hyperplasia. The presence of an ectopic CRH-secreting tumor should be suspected in patients who have disease that behaves biochemically like pituitary ACTH-dependent disease but who have a clinical course of short duration and high plasma ACTH levels.

CORTICOTROPIN-INDEPENDENT CUSHING SYNDROME

Primary nodular hyperplasia, adrenocortical adenomas, and carcinomas as well as rare cases of ectopic cortisol production by tumor are found in 30% of patients presenting with Cushing syndrome.25 PRIMARY BILATERAL MICRONODULAR HYPERPLASIA Primary bilateral micronodular hyperplasia is characterized by the presence in the adrenal cortex of one or more yellow nodules visible to the naked eye and 2 to 3 cm

in diameter. Its pathogenesis is unknown. Whether this type of hyperplasia is of primary adrenal origin or the result of a combined mechanism by which the pituitary predominates initially but the adrenal eventually becomes autonomous, is not clear.26 Increased adrenocortical cell sensitivity to ACTH has also been suggested.27 Histologically, the micronodules consist of clear cells arranged in acini and cords, and are occasionally encapsulated and surrounded by simple or micronodular cortical hyperplasia. The frequency of mitotic figures is low. In one form of ACTH-independent micronodular adrenal disease, darkly pigmented micronodules are seen in the presence of atrophy of the perinodular adrenal tissue, disorganization of normal zonation of the cortex, and small glands28 (Fig. 75-3). Microscopically, the nodules are composed predominantly of large globular cortical cells with granular eosinophilic cytoplasm that often contains lipofuscin.

FIGURE 75-3. Appearance of an adrenal gland with pigmented micronodular hyperplasia removed from a patient with familial, corticotropin-independent Cushing syndrome. The patient was also found to have an atrial myxoma and a Sertoli cell testicular tumor.

A condition associated with pigmented micronodular hyperplasia is Carney complex, a form of multiple endocrine neoplasia. Patients with this condition have, in addition to the adrenal cortical disease, growth hormonesecreting pituitary tumors, Sertoli cell testicular tumors, atrial or ventricular myxomas, single or multiple mammary fibroadenomas, and cutaneous myxomas. A prominent feature in these patients is spotty skin pigmentation similar to that observed in several lentiginosis syndromes. The genetic defect(s) responsible for Carney complex map(s) to the short arm of chromosome 2 (2p16). This region has been found to exhibit cytogenic aberrations in atrial myxomas associated with the complex and has been characterized by microsatellite instability in human neoplasias.29 ACTH-independent macronodular adrenal cortical hyperplasia is a rare cause of Cushing syndrome in which the adrenal glands become very enlarged and occupied by multiple large cortical nodules. In some patients, the masses can be extremely large and weigh close to 100 g each. Histologically, the nodules are composed mostly of clear cells, and no cortical architecture has been observed in the internodular regions.30 Several interesting pathogenic mechanisms have been described in patients with ACTH-independent nodular adrenal cortical hyperplasia. In one case, the secretion of cortisol was stimulated by food. The patient had elevated postprandial cortisol levels that could not be suppressed with dexamethasone. Cortisol increased in response to oral glucose, a lipid-rich meal, or a protein-rich meal but not to the intravenous administration of glucose. In response to test meals, plasma glucose-dependent insulinotropic polypeptide (GIP) increased in parallel to the plasma cortisol. The ability of GIP to stimulate adrenal cortisol production could be demonstrated in cell suspensions of adrenal tissue from the patient. Nodular adrenal cortical hyperplasia and Cushing syndrome were postulated to be food dependent and to result from abnormal responsiveness of adrenal cells to the physiologic secretion of GIP. The suggestion was made that ectopic expression of GIP receptors on the adrenal cells mediate this disorder.31 A case of Cushing syndrome secondary to ACTH-independent bilateral adrenal hyperplasia was also described in which cortisol secretion responded to catecholamines acting through ectopic adrenal b receptors. The increased cortisol secretion was inhibited by b-blockade with propranolol. High-affinity binding sites compatible with b1 or b2 receptors were detected in the adrenal tissue from this patient but not from control subjects. The ectopic expression of b receptors in the adrenal cortex of this patient led to nodular hyperplasia, hypersecretion of cortisol and aldosterone, and feedback suppression of ACTH and angiotensin.32 ADRENAL ADENOMA AND CARCINOMA Adrenal adenomas are unilateral, well-circumscribed, brown to yellow tumors with a homogeneous appearance, weighing <30 g. Histologically, they have the characteristics of normal adrenal cortex, with compact and clear cells arranged in cords (Fig. 75-4). Adrenal carcinomas are large neoplasms with irregularly shaped nodules, exhibiting areas of necrosis or hemorrhage. Histologically, the cells are small, with nuclear pleomorphism and mitotic figures, and evidence of capsular and blood vessel invasion by tumor is found (Fig. 75-5).

FIGURE 75-4. A, This well-encapsulated, adrenal cortical adenoma, 2.6 cm in diameter, with homogeneous appearance, was removed from a 35-year-old woman with a 3-year history of Cushing syndrome. B, The histologic pattern was consistent with a benign adrenal adenoma, with clear and compact cells arranged in cords.

FIGURE 75-5. A, This 4.2 8.5 cm adrenal neoplasm with multiple, large nodules and evidence of capsular invasion was removed from a 50-year-old woman with a 9-month history of Cushing syndrome. B, Histologically, this tumor revealed angioinvasion and small cells with nuclear pleomorphism and numerous mitotic figures.

Adrenal cortical carcinoma may develop in patients with hereditary cancer syndromes. The Li-Fraumeni syndrome is characterized by cancers derived from all three germinal cell lines, and include sarcomas, brain tumors, leukemia, lymphoma, and adrenal cortical carcinoma, in addition to early-onset breast cancer. These tumors tend to occur in family members under the age of 45. Mutations in the p53 tumor-suppressor gene have been identified in patients with cancer associated with

Li-Fraumeni syndrome.33,34

CLINICAL PRESENTATIONS
The clinical manifestations of Cushing syndrome are determined by the increased abnormal secretion of cortisol, ACTH, and other adrenocortical steroids under ACTH control. However, the clinical presentation varies among the different types. CORTICOTROPIN-DEPENDENT CUSHING DISEASE Patients with pituitary ACTH-dependent disease (i.e., Cushing disease) usually have a history of disease that begins 2 to 3 years before a definitive diagnosis is made. In milder cases, symptoms may be present for 5 to 10 years. Earliest symptoms include weight gain, progressive changes in physical appearance, hypertension, and glucose intolerance. Patients may have been treated for these conditions with weight-reduction diets, antihypertensive drugs, and oral hypoglycemic agents. The response to diet is frequently unsuccessful. Those who do not experience actual weight gain notice redistribution of adipose tissue resulting in changes in appearance, with facial rounding, increased central adiposity, and thinning of upper and lower extremities. After 2 to 3 years of mild symptoms, additional symptoms or increased severity of the initial symptoms draw patients to the attention of the physician. Patients relate the appearance of striae, easy bruising, and increased body hair growth; irregular menses or amenorrhea may be present in women or gynecomastia in men. Proximal muscle weakness and atrophy are common, and some patients exhibit manifestations of a steroid myopathy. Other manifestations include peripheral edema, back pain, and loss of height if patients have developed severe osteoporosis and compression fractures of the vertebrae (Fig. 75-6A).

FIGURE 75-6. A, A 45-year-old man with a 2-year history of progressive muscle weakness and atrophy, truncal obesity, severe back pain, loss of height, and multiple ecchymoses. B, Truncal obesity with a very prominent cervicodorsal fat pad (i.e., buffalo hump) in a woman with Cushing syndrome. C, Atrophic striae over hips, breasts, and axillae in a 24-year-old woman with Cushing syndrome. D, A 28-year-old woman with Cushing syndrome. She had pigmented lesions of tinea versicolor over the anterior upper chest. E, Severe, intractable verruca vulgaris lesions over the fingers in a 32-year-old man with Cushing syndrome and a markedly depressed T-lymphocyte count.

The physical examination reveals plethora, a round face with preauricular fullness, a prominent upper lip (i.e., the Cupid bow), supraclavicular fossa fullness, and a cervicodorsal fat pad (i.e., buffalo hump) disproportionate to the degree of obesity (see Fig. 75-6B). Patients exhibit truncal obesity, with large breasts or gynecomastia and a protuberant abdomen. This protuberance is the result of an increased abdominal panniculus and relaxed abdominal muscles. The extremities are thin, with decreased muscle mass and strength. Patients may show pretibial edema. The skin is thin, with wide, atrophic purple striae present over the abdomen, chest, and upper thighs (see Fig. 75-6C). Other skin lesions include keratosis pilaris seen over the upper trunk, tinea versicolor over the anterior chest and back (see Fig. 75-6D), and verruca vulgaris, which occasionally may be large and present in multiple locations (see Fig. 75-6E). Hirsutism is often present as a mixture of fine, lanugal-type hair growth over the face and trunk, together with coarse, terminal-type hair over the face, trunk, and extremities. This latter type of hair is associated with increased androgen secretion. Increased skin pigmentation is infrequent in patients with pituitary ACTH-dependent Cushing disease, because ACTH levels are not sufficiently elevated to induce hyperpigmentation. When hyperpigmentation is present, it is associated with high ACTH levels and characterized by pigmentation of the hands, palmar creases, elbows and knees, gums, and oral mucosae. Another type of pigmentation is associated with insulin resistance and hyperinsulinemia; this type is localized to the neck, elbows, and dorsum of the hands and has the characteristics of acanthosis nigricans. In individuals with dark skin, a history of further darkening is indicative of development of hyperpigmentation. ECTOPIC CORTICOTROPIN AND CORTICOTROPIN-RELEASING HORMONE SYNDROMES The duration of symptoms is usually shorter in patients with the ectopic ACTH syndrome than in those with the pituitary ACTH-dependent type, because the source of ACTH is a malignant neoplasm, often one with a rapidly progressive course. Frequently, the symptoms and signs of Cushing syndrome appear 4 to 6 months before a diagnosis is made. Some patients exhibit minimal symptoms of Cushing syndrome because of the short duration of hypercortisolemia. In these patients, very high cortisol levels are often associated with manifestations of mineralocorticoid excess, including hypernatremia, hypokalemia, and metabolic alkalosis. Because the levels of ACTH in these patients are 5 to 10 times higher than in those with pituitary ACTH-dependent disease, hyperpigmentation of the skin and oral mucosa may be an early manifestation. Severe hypokalemia may also lead to significant muscle weakness. The manifestations of the underlying malignant neoplasm, including anorexia, weight loss, and focal symptoms of organ involvement, may occur with the manifestations of cortisol excess. Some patients with the ectopic ACTH syndrome have a more protracted disease course. These are patients with bronchial or mediastinal carcinoids that may be present for several years before detection and whose disease may share clinical and biochemical features with pituitary ACTH-dependent disease. CORTICOTROPIN-INDEPENDENT TYPES OF CUSHING SYNDROME Patients with benign adrenocortical adenomas usually have a long history of manifestations of cortisol excess before a definitive diagnosis is made. Because adrenocortical adenomas are often monotropic and purely cortisol-secreting tumors, the predominant clinical manifestations are indicative of the protein catabolic effects of cortisol unopposed by androgen excess. These include skin and muscle atrophy, easy bruising, and osteopenia. The patients have the typical fat distribution associated with hypercortisolism, but acne and coarse, terminal-type hair growth are unusual. Patients with adrenocortical carcinoma have a disease history of shorter duration, with symptoms appearing within 4 to 6 months of diagnosis. The severity of the clinical manifestations depends on the degree of hypercortisolemia. Patients with adrenocortical carcinoma may exhibit manifestations of androgen excess, including hirsutism, acne, scalp hair loss, and clitoromegaly. The androgen excess can moderate the severity of the catabolic effects of hypercortisolism. In particular, skin and muscle atrophy may not be as apparent. Because adrenocortical carcinomas are large, frequently >100 g, they may be palpable on the abdominal examination, and hepatomegaly may be noted if hepatic metastases are present. Although the discovery of an adrenocortical adenoma is usually made while a patient is being investigated for the possibility of Cushing syndrome, the adenoma occasionally is discovered incidentally during the investigation of abdominal complaints. Silent adrenal masses are found in 2% to 3% of patients who are undergoing CT scans for investigation of abdominal symptoms.34a Although these patients may not have obvious clinical manifestations of Cushing syndrome and their baseline urinary cortisol levels are normal, cortisol secretion is autonomous. Evidence of autonomous cortisol secretion includes a blunted circadian rhythm, suppressed ACTH levels, and a lack of response to dexamethasone. Adrenal scintigraphy with iodocholesterol shows increased uptake on the side of the tumor but suppression of uptake in the contralateral adrenal gland. Although these patients may have minimal manifestations of Cushing syndrome, these manifestations improve after surgical removal of the lesion. Patients with primary nodular adrenocortical hyperplasia are clinically indistinguishable from patients with benign adrenocortical adenomas or pituitary ACTH-dependent disease.

DIAGNOSIS

STANDARD DIAGNOSTIC EVALUATION A biochemical evaluation of Cushing syndrome is necessary for confirming the clinical diagnosis and for determining the presence of the syndrome in patients with an equivocal clinical presentation.35 Preliminary testing includes the measurement of urinary free cortisol and of the ability to suppress serum cortisol levels with a low dose of dexamethasone (Table 75-2). In patients with Cushing syndrome, the urinary free cortisol usually exceeds 90 g per day. Serum cortisol levels obtained 9 hours after the oral administration of 1 mg of dexamethasone at 23:00 hours fail to decrease below 10 g/dL, and plasma ACTH levels (assessed by immunoradiometric assay) fail to decrease to <9 pg/mL.36 In patients without Cushing syndrome, cortisol levels are suppressed below 5 g/dL. In patients whose levels fall between these two values, additional studies are required for confirmation of the diagnosis. After the preliminary diagnosis has been determined, a definitive diagnosis may require more extensive testing (Table 75-3). This includes the measurement of serum cortisol and plasma ACTH levels at various times of the day to determine the absence of a circadian rhythm.

TABLE 75-2. Outpatient Evaluation of Cushing Syndrome

TABLE 75-3. Definitive Evaluation of Cushing Syndrome

Patients with Cushing syndrome, regardless of cause or type, have high baseline urinary free cortisol and serum cortisol levels and lack normal circadian changes in level. Occasionally, the early morning cortisol levels may be indistinguishable from normal, but levels obtained between 22:00 hours and midnight are clearly higher than in normal persons. This testing, however, requires that the patient be hospitalized for blood drawing. A more convenient procedure is the measurement of salivary cortisol at 08:00 and 22:00 hours or midnight in a collection that can be obtained by the patient at home. Salivary cortisol testing measures free cortisol and results correlate well with plasma cortisol levels. Patients with Cushing syndrome have high evening salivary cortisol levels.37 In borderline or confusing cases, loss of circadian rhythm may be key to establishing the differential diagnosis between Cushing syndrome and normal cortisol secretion. The pituitary response to dexamethasone is measured to detect abnormal feedback regulation of ACTH and cortisol secretion. Dexamethasone is given in doses of 0.5 mg every 6 hours for 8 doses, followed by 2 mg every 6 hours for 8 doses. The response of plasma ACTH levels to dexamethasone depends on the type of Cushing syndrome. When an immunoradiometric assay is used (normal levels, 952 pg/mL), baseline levels are normal or high (50100 pg/mL) in ACTH-dependent types and low (<9 pg/mL) in ACTH-independent types. Although ACTH levels are frequently at the upper limit of normal in patients with pituitary causes, they range from 200 to 1000 pg/mL in patients with the ectopic ACTH syndrome. The lack of normal suppression of cortisol with low doses of dexamethasone is found in all types of Cushing syndrome, but most patients with pituitary ACTH-dependent Cushing syndrome (i.e., Cushing disease) shown suppression with high doses of dexamethasone to values below 50% of the baseline levels (Table 75-4). Fewer than 10% of patients with ectopic ACTH syndrome respond in this manner. In patients with ACTH-dependent Cushing syndrome, high ACTH levels and lack of response to dexamethasone should strongly raise the possibility of ectopic ACTH syndrome.

TABLE 75-4. Differential Diagnosis of Pituitary and Ectopic Corticotropin (ACTH)-Dependent Cushing Syndrome

The high-dose dexamethasone suppression test can be performed as a rapid single oral dose test and by intravenous administration in hospitalized patients unable to take oral medication.38,39 A single dose of 8 mg of dexamethasone is given at 23:00 hours, and the serum cortisol level is measured at 08:00 hours the next morning. A suppression of plasma cortisol levels to <50% of baseline values suggests a pituitary cause, and a lack of suppression below that limit indicates the existence of an adrenal tumor or of ectopic ACTH syndrome. The diagnostic efficacy or predictive power of the overnight high-dose dexamethasone suppression test (defined as the ratio between the number of cases in which the diagnosis is correctly predicted and the total number of cases) is at least 82.4%, compared with 84.6% for the classic 8-mg Liddle test. Because the two tests are comparable, the shorter test has the obvious advantage of speed and lower cost. The ACTH response to the intravenous administration of CRH is also a useful test in the differential diagnosis of pituitary and ectopic ACTH syndrome.40,41 After the administration of ovine CRH (1 g/kg), patients with Cushing disease usually show a further increase in the elevated levels of ACTH and cortisol. In contrast, most patients with the ectopic ACTH syndrome who also have high basal plasma concentrations of ACTH and cortisol fail to respond to CRH (see Chap. 74 and Chap. 241). Several tests have been suggested for establishing a diagnosis of Cushing disease in patients with borderline elevations of cortisol and negative pituitary imaging for a pituitary adenoma. These tests include the high-dose dexamethasone suppression test with revised criteria to improve sensitivity and specificity; the combined CRH stimulation/dexamethasone suppression test (CRH/DST), and a hydrocortisone suppression test in cases for which dexamethasone gives a false-negative result.42

In the CRH/DST, the CRH stimulation is performed 2 hours after completion of the low-dose dexamethasone suppression test. Normal individuals and patients with pseudo-Cushing states show a lower response to CRH than patients with pituitary ACTH-dependent Cushing syndrome. The separation appears to be valid even in patients with mild Cushing disease.43 The combined CRH/DST could be a potentially useful modification of the dexamethasone suppression test and the CRH stimulation test to distinguish patients with pseudo-Cushing from those with pituitary ACTH-dependent Cushing syndrome. Benign and malignant adrenocortical tumors can be differentiated biochemically. Malignant neoplasms often have partial enzymatic deficiencies in the steroid biosynthetic pathway. When this occurs, steroid biosynthetic intermediates or other metabolites are found in increased quantity in serum or urine. Impaired activities of 3b-hydroxysteroid dehydrogenase and 11b-hydroxylase are the most common. These deficiencies lead to increased serum levels of pregnenolone and 11-deoxy-cortisol. DIFFICULTIES IN INTERPRETATION OF THE DIAGNOSTIC TESTS The tests described produce a small but significant percentage of false-positive and false-negative results that contribute to difficulties in establishing a clear biochemical diagnosis of the disease.44 For example, the 08:00-hour plasma cortisol values are found to yield 9% false-positive and 60% false-negative results. A diurnal variation of cortisol occurs in 30% of patients with Cushing syndrome and is lacking in 18% of patients without Cushing syndrome. The overnight dexamethasone suppression test may be abnormal in 30% of hospitalized patients. In contrast, urinary free cortisol level has a very great sensitivity and specificity (uncorrected or corrected for creatinine) and is the most efficient screening method for Cushing syndrome in hospitalized patients. Several factors can lower the sensitivity and specificity of these tests. First, cortisol secretion is episodic or pulsatile and can result in widely variable levels. A normal diurnal variation is a pattern in which plasma cortisol levels from 16:00 hours to midnight are <75% of the 08:00-hour values. Values at 16:00 hours can show considerable overlap in normal individuals and patients with Cushing syndrome. The plasma cortisol value at 23:00 hours is probably superior, because only ~3% of patients with Cushing syndrome have been shown to have normal plasma cortisol levels at that time. Frequent sampling for serum cortisol is useful but requires hospitalization and may be relatively expensive. Second, 30% to 50% of obese people, particularly those with abdominal obesity and without Cushing syndrome, have increased cortisol secretion rates, but urinary free cortisol values in patients with Cushing syndrome are separated by at least 60 g per day from those in obese individuals. Third, a variety of drugs that cause induction of hepatic microsomal hydroxylating enzymes may increase the metabolism of dexamethasone and result in lower levels of these compounds than expected for the dose administered.45 Under these circumstances, the response to dexamethasone may be abnormal without the presence of Cushing syndrome. Patients with Cushing disease who show normal suppression with the low-dose dexamethasone test usually show a decrease in the clearance of dexamethasone, which produces plasma dexa-methasone levels that are higher than expected.46 The measurement of plasma dexamethasone together with cortisol can help exclude abnormalities in dexamethasone clearance. Rarely, patients with factitious Cushing syndrome, who have been taking pharmacologic doses of synthetic glucocorticoids, present with a paradoxical finding of the clinical manifestations of Cushing syndrome with suppressed ACTH and cortisol levels.

ANATOMIC LOCALIZATION
DETECTION OF PITUITARY LESIONS After the diagnosis of Cushing disease has been established on clinical and biochemical grounds, the presence of pituitary lesions should be further determined by magnetic resonance imaging (MRI) of the sella turcica. Because of the small size of most pituitary tumors associated with Cushing disease, the results of plain radiographic studies are often normal or misleading. Large microadenomas secreting ACTH are extremely rare at the time the initial diagnosis of the disease is made. However, small (<10 mm) pituitary microadenomas can be identified by MRI of the pituitary gland with gadolinium contrast using coronal cuts in approximately 70% of patients with Cushing disease. On T1-weighted images, the lesion usually appears as a nonenhancing abnormality surrounded by an enhancing normal pituitary gland (Fig. 75-7).

FIGURE 75-7. T1-weighted magnetic resonance image of the pituitary gland with gadolinium contrast in a patient with corticotropin-dependent Cushing syndrome (Cushing disease). The pituitary gland and the stalk (arrow) enhanced with contrast, and an area of decreased enhancement (arrowhead) consistent with an adenoma was outlined in the right side of the gland.

Petrosal sinus sampling with or without CRH stimulation can be used for localization of pituitary tumors.47 Through a percutaneous bilateral femoral approach, catheters are placed in the left and right inferior petrosal sinuses, and blood for ACTH is withdrawn simultaneously from both catheters and from a peripheral vein. To account for fluctuations in levels caused by the pulsatile nature of ACTH secretion, sequential samples (i.e., four samples drawn on both sides over a 20-second period at 2-to 5-minute intervals) are obtained. A plasma ACTH concentration gradient (i.e., petrosal sinus/peripheral vein ACTH of 1.6) verifies the pituitary source of ACTH. A right to left discrepancy may also help to lateralize ACTH secretion and aids in the preoperative localization of the lesion. A lateralization of the source of ACTH by this procedure can help in cases in which a tumor is not found at the time of surgery. Partial resection of the side of the pituitary where the highest ACTH levels are recorded can be accompanied by remission of the disease. Failure to find the pituitary microadenoma or to induce cure after resection of the side of the pituitary with the highest concentration of ACTH may be due to failure to remove the involved part of the pituitary gland completely or to the fact that confluent pituitary veins do not invariably drain into the ipsilateral inferior petrosal sinus, so that a microadenoma may be present in the opposite side of the pituitary gland.48 Another possibility is the presence of multiple microadenomas. The administration of CRH at the time of catheterization may enhance differences between the two sides; however, a simultaneous response by the uninvolved gland may blur the differences. To correct for unequal dilution by nonpituitary venous blood, other pituitary hormones, including prolactin, thyroid-stimulating hormone (TSH), human chorionic gonadotropin, and a subunit, have been measured simultaneously. The assumption was that an ACTH-secreting microadenoma would not cause unequal delivery of these hormones into the two inferior petrosal sinuses, but correction of the ACTH gradients by these hormones did not improve the discriminatory ability of the test.49 Catheterization of the inferior petrosal sinuses requires a great deal of technical expertise. To make this type of testing more accessible to neuroradiologists with limited experience in this procedure, the diagnostic accuracy of the less invasive sampling of the internal jugular veins has been compared with sampling of the inferior petrosal sinuses.50 Jugular vein sampling correctly identified ACTH-secreting adenomas in 80% of patients with proven Cushing disease. Administration of CRH was essential for diagnostic results in 65% of the patients. Negative results on jugular vein sampling should be confirmed by petrosal sinus sampling. DETECTION OF ADRENAL LESIONS Techniques for the anatomic localization of adrenal lesions include abdominal CT scans and MRI, ultrasonography, and adrenal scintigraphy with 6-b-(131I)-iodomethyl-19-norcholesterol.51 These techniques are noninvasive and provide good definition of structure and function of the adrenal glands. Adrenocortical masses can be anatomically defined by CT scanning, MRI, or ultrasonography (Fig. 75-8). CT can identify adrenal lesions as small as 1 cm in diameter. This level of resolution can be important in patients with adrenal carcinoma to detect involved lymph nodes and hepatic and pulmonary metastases. Hepatic

metastases are found in approximately 42% of patients, and pulmonary metastases are found in 45%.

FIGURE 75-8. A 13-cm left adrenal mass (A) with areas of necrosis was detected by computed tomographic scanning in a 72-year-old woman with abdominal pain. She also had pulmonary nodules consistent with metastatic adrenal carcinoma.

MRI is also helpful. Adrenal carcinomas are hypointense compared with liver in T1-weighted images. They are hyperintense compared with liver in T2-weighted images. The superior blood vessel identification and the multiplanar capacity of MRI may make it the modality of choice in evaluating the extent of disease and in planning surgical excision.52 Ultrasonography can help define the benign or malignant character of a lesion. Malignant lesions are heterogeneous in appearance, with focal or scattered echopenic or echogenic zones representing areas of tumor necrosis, hemorrhage, or calcification.53 Scintigraphy with iodocholesterol provides an image reflective of the structure of the adrenal gland and information about its function. Patients with bilateral adrenocortical hyperplasia demonstrate bilateral increased adrenal uptake of iodocholesterol. Patients with cortisol-secreting adrenocortical adenomas, which suppress pituitary ACTH secretion and the function of the contralateral gland, demonstrate unilateral concentration of the tracer. Patients with Cushing syndrome secondary to an adrenocortical carcinoma fail to show tracer uptake on either side. Carcinomas have relatively low functional capacity per gram of tissue and fail to concentrate iodocholesterol in quantities sufficient to produce an image.54 Occasionally, actively secreting carcinomas may image in a manner indistinguishable from that of a benign tumor. Patients with primary macronodular hyperplasia may show asymmetric uptake, with greater activity on the side of the predominant nodules. This finding is important in determining the presence of bilateral adrenocortical disease, because a CT scan or MRI may show a mass on one side, but surgical resection of that mass may not result in remission if the contralateral gland is also abnormal. Radiographic differences between patients with macronodular hyperplasia secondary to ACTH-secreting pituitary adenomas and primary micronodular hyperplasia of the adrenal glands have been described. In macronodular adrenocortical hyperplasia, the nodules are visible on CT or MRI, and in micronodular disease, the micronodules are seen microscopically, but the adrenal glands do not demonstrate focal masses.55 LOCALIZATION OF ECTOPIC CORTICOTROPIN-SECRETING LESIONS Ectopic sources of ACTH are localized using a variety of procedures, including CT of the chest, CT or MRI of the abdomen, thyroid scan, and selective venous catheterization and sampling in search of concentration gradients of ACTH (Fig. 75-9). Approximately 50% of these tumors are within the thorax (e.g., small cell carcinomas, bronchial carcinoids, thymomas), and thoracic CT scans usually can localize the lesion. Other ACTH-secreting tumors are found in the pancreas (islet cell tumors), the thyroid (medullary carcinoma), and the adrenal medulla (pheochromocytoma).

FIGURE 75-9. Results of selective venous catheterization and sampling in a 55-year-old man with ectopic corticotropin (ACTH)-dependent Cushing syndrome. ACTH levels (pg/mL) were measured in samples obtained throughout the venous system. A peak level in the left innominate vein led to the discovery of a malignant, ACTH-secreting thymoma.

Ectopic ACTH-secreting carcinoids may be small and slow growing, and many years may elapse before they become radiographically apparent.56 Carcinoid tumors that possess somatostatin receptors may take up the somatostatin analog octreotide. Use of indium-111 (111In) pentetreotide has allowed these tumors to be detected as areas of increased uptake at the exact location of a suspected thoracic lesion seen on CT.57 However, these scans are often negative when a lesion is not detected by CT. The finding of a pulmonary lesion may lead to a percutaneous biopsy for confirmation and eventual surgical excision of the lesion. The biopsy tissue can be analyzed by immunocy-tochemistry to demonstrate POMC-derived peptide secretory granules in the tumor cells or by extraction and analysis of the ACTH content in the tumor. The presence of POMC mRNA may also be a way of confirming that the tumor has the ability to produce ACTH.58,59 and 60

PSEUDOCUSHING SYNDROME
Hypercortisolism may exist in patients who do not have true Cushing syndrome. These patients experience obesity, depression, alcoholism, anorexia nervosa, bulimia, or acute stress. Because their clinical presentation and biochemical findings overlap with those of patients with Cushing syndrome, the differential diagnosis can be challenging. OBESITY Obese people, particularly those with abdominal or upper-body-segment obesity, share many of the metabolic, hormonal, and behavioral findings observed in Cushing syndrome. These include diabetes mellitus, hypertension, myocardial infarction, insomnia, and depression.61 Obesity is associated with increased cortisol production, as indicated by high cortisol secretion rates and levels of urinary 17-hydroxycorticosteroids, but normal or slightly diminished serum cortisol levels; these findings suggest an increased metabolic clearance rate for cortisol.62 Urinary free cortisol is increased in proportion to the waist-to-hip ratio; these individuals also exhibit increased cortisol responsiveness to ACTH stimulation and to physical and mental stress.63 Unlike in Cushing syndrome with its well-defined causes, in pseudoCushing syndrome the cause of the abnormality in adrenal function and its relationship to the complications of obesity have not been established. From the clinical standpoint, obese people do not usually exhibit signs of protein catabolism. Biochemically, they have a normal circadian rhythm of cortisol and ACTH secretion and show normal suppression with a low dose of dexamethasone.62

MAJOR DEPRESSIVE DISORDER Approximately 50% of patients with major depressive disorder (MDD) hypersecrete cortisol and show early escape from the normal feedback inhibition by dexamethasone. Conversely, neuropsychiatric symptoms such as irritability, decreased libido, insomnia, and depressed mood also are frequently seen in patients with Cushing syndrome. Deciding whether a patient with hypercortisolemia has primary depression or early Cushing syndrome may be difficult. Some patients who eventually show full-blown Cushing syndrome begin with only intermittent elevations of cortisol and symptoms of a major affective disorder. Although similar in many respects to MDD, the depression seen in Cushing syndrome does have distinguishing clinical characteristics that may aid in the differential diagnosis. Irritability is a prominent and consistent feature, as are symptoms of autonomic activation such as shaking, palpitations, and sweating. Depressed affect is often intermittent, and Cushing patients feel their best, not their worst, in the morning. Psycho-motor retardation is not so pronounced as to be clinically obvious, and most of these patients are not withdrawn, apathetic, or hopeless. Significant cognitive impairment, including disorder of memory, is a consistent and prominent clinical feature.64 Biochemically, patients with Cushing syndrome exhibit higher cortisol levels than those with MDD, and these levels do not return to normal with administration of antidepressants. Assessment of the ACTH response to CRH may be helpful in the differential diagnosis. Although a substantial overlap exists between the two groups, patients with Cushing syndrome show a response, and those with MDD fail to respond to CRH.65 ALCOHOL-INDUCED PSEUDOCUSHING SYNDROME As originally described in the literature, patients with alcohol-induced pseudoCushing syndrome present with moon facies, truncal obesity, easy bruising, and biochemical features consistent with hypercortisolism. They may fail to show normal suppression with the 1.0-mg overnight dexamethasone test but recover normal suppressibility within 3 weeks of alcohol withdrawal. Similarly, the initially elevated 24-hour urinary free cortisol level returns to normal within a few days.66,67 Doubt exists, however, as to whether chronic alcohol use leads to stimulation of the hypothalamicpituitaryadrenal axis. In cases in which cortisol levels have been found to be increased, patients had consumed large quantities of alcohol and appeared intoxicated, with nausea, vomiting, and considerable stress. Some of the phenotypic changes associated with chronic alcohol abusesuch as central adiposity, myopathy, skin atrophy, depression, and psychosismay be the result of ethanol itself rather than of cortisol. Bruises and rib fractures may result from falls, and diabetes may result from pancreatitis. Thus, whether alcohol-induced pseudo Cushing syndrome really exists is questionable.68

DIFFERENTIAL DIAGNOSIS OF CUSHING-TYPE HYPERCORTISOLISM AND PSEUDOCUSHING SYNDROME

The biochemical and phenotypic presentation of patients with mild hypercortisolism in Cushing syndrome is often indistinguishable from that seen in pseudo-Cushing states such as depression. These include elevated urine free cortisol and 17-hydroxycorticosteroid excretion and abnormalities in the normal suppression of plasma cortisol after the administration of a low (1-mg) dose of dexamethasone. In most cases of pseudo Cushing syndrome, the history and physical examination provide important clues as to whether a given patient has Cushing syndrome or one of these other clinical states. This is particularly true with regard to obesity, in which comorbid conditions exist that overlap with those of Cushing syndrome. In these cases, biochemical studies are necessary to establish a differential diagnosis. Several of the tests used in this differential diagnosis vary in the level of diagnostic accuracy. The low-dose dexamethasone suppression test has 74% specificity, 69% sensitivity, and 71% diagnostic accuracy using the standard criterion of suppression of urinary 17-hydroxycorticoids to <2.5 mg per day on the second day of suppression. For urinary free cortisol, a value >36 g per day (100 nmol per day) has 100% specificity, 56% sensitivity, and 71% diagnostic accuracy for Cushing syndrome. The CRH stimulation test without dexamethasone pretreatment has 100% specificity, 64% sensitivity, and 76% diagnostic accuracy. The diagnostic accuracy of the CRH/DST for the diagnosis of Cushing syndrome is greater than the accuracy of either the low-dose dexamethasone test or the CRH test alone.69 A plasma cortisol concentration of >1.4 g/dL (38 nmol/L) measured 15 min after the administration of CRH had 100% specificity, sensitivity, and diagnostic accuracy. The test has not been worked out for patients with severe melancholic depression, anorexia nervosa, cortisol resistance syndrome, or recent surgical stress. It also has not been defined for patients with periodic hormonogenesis and intermittent Cushing syndrome. In these patients, measurement of repeated 24-hour urine free cortisol levels may help detect times when an upsurge of ACTH and cortisol secretion occurs. The hypothesis behind the test is that at a dexamethasone dose sufficient to suppress normal cortisol production, patients with pseudo-Cushing states exhibit low basal plasma cortisol and ACTH levels and a diminished response to exogenous CRH. By contrast, patients with Cushing disease have higher basal cortisol and ACTH levels after the administration of dexamethasone and a greater peak response after CRH.69 CORTICOTROPIN-RELEASING HORMONE IN THE PATHOPHYSIOLOGY OF PSEUDOCUSHING SYNDROME CRH secretion may help to differentiate pseudoCushing syndrome and true Cushing syndrome. Although CRH secretion may be increased in patients with pseudoCushing syndrome, in most patients with ACTH-dependent or ACTH-independent Cushing syndrome, CRH secretion appears to be suppressed. When exogenous CRH is administered to patients with MDD, ambulatory alcoholics, and patients with anorexia nervosa, the ACTH response is diminished, whereas the response is increased in patients with pituitary ACTH-dependent disease. Unfortunately, a 25% overlap is found between the two groups.65

TREATMENT OF CUSHING DISEASE

Optimal treatment of Cushing disease depends on an accurate diagnosis of the underlying pathology. Four approaches are used in the management of pituitary ACTH-dependent Cushing disease: pituitary surgery, pituitary irradiation, adrenal surgery, and drug therapy.7,70,71 and 72 PITUITARY SURGERY The treatment of choice for pituitary ACTH-dependent Cushing disease is the microsurgical removal of microadenomas or macroadenomas. If a pituitary adenoma or microadenoma can be demonstrated by radiographic techniques, a transsphenoidal operation of the pituitary gland should be the preferred treatment. If a tumor is not detected by imaging techniques, a transsphenoidal exploration of the pituitary gland is still in order, because the tumor can be found in ~90% of patients.73,74 and 75 Even with considerable suprasellar extension, the tumor can be resected transsphenoidally. If a tumor invades the dura, total resection may be impossible, but good remission rates of 45% to 75% have been described for these cases. The microsurgical transsphenoidal selective resection of ACTH-secreting pituitary microadenomas is the most common treatment of Cushing syndrome and comes closest to the ideal form of treatment for this condition.76,77 and 78 Several reports have described a high cure rate with transsphenoidal surgical treatment of Cushing disease.73,74 and 75,79,80 The probability of finding pituitary pathology and of surgically correcting the disease is highest among patients with a typical endocrine testing pattern.81 Typical diagnostic criteria for Cushing disease consist of elevated basal urinary 17-hydroxycorticosteroids and free cortisol levels, cortisol secretion rates, and mean basal serum cortisol levels; a positive response to metyrapone (i.e., elevated ACTH levels associated with a rise in urinary 17-hydroxycorticosteroids); and abnormal suppression with low-dose dexamethasone but >50% suppression with high-dose dexamethasone. Patients with atypical diagnostic criteria have elevated basal levels but do not respond as described to metyrapone or to low or high doses of dexamethasone. Pituitary disease was found in 18 of 19 patients with typical preoperative endocrine test results but in only 6 of 11 patients with atypical test results. If a microadenoma can be identified and totally and discretely resected, then the remaining pituitary tissue remains functional, and patients can enjoy remission without loss of endocrine function.82 If a specific adenoma cannot be identified during surgery, the decision must be made as to whether to perform a partial or total hypophysectomy. If preoperative inferior petrosal sinus sampling has been carried out and is clearly lateralizing, an appropriate hemiresection of the hypophysis should be performed. If the endocrine studies strongly indicate a pituitary origin but the petrosal sinus sampling is not lateralizing and the patient does not wish to have children, a total hypophysectomy should be considered, but only after a lengthy preoperative discussion with the patient regarding this possibility. If the patient wishes to have children, alternative forms of therapy, including medical treatment or a total adrenalectomy, must be considered. Transsphenoidal surgery for Cushing disease seems to be a reasonably safe procedure, with a mortality rate of <1%. Main complications in these patients are anterior pituitary insufficiency, which occurs in 19%, and diabetes insipidus, which occurs in 18%. Overall incidence of cerebrospinal fluid fistulas is 4%. Other complications, including meningitis, carotid artery injuries, loss of vision, and hypothalamic injuries, occur in from 1% to 2% of patients. An inverse relationship is found between the experience of the surgeon and the likelihood of complications. A statistically significant decreased incidence of morbidity and death has been reported after 200 and even 500 transsphenoidal operations have been performed.83 With transsphenoidal surgery, permanent anterior or posterior pituitary hormone deficiencies are rare. Transient diabetes insipidus may occur during the early weeks after surgery. Permanent diabetes insipidus and cerebrospinal fluid rhinorrhea are uncommon complications with an initial procedure but may occur more commonly with repeated transsphenoidal surgery. Treatment failures are most common in patients with pituitary macroadenomas or in those in whom a distinct microadenoma has not been found. DELAYED RECURRENCE A delayed recurrence of Cushing disease after removal of a pituitary adenoma may be the result of regrowth of adenoma cells left behind in the peritumoral tissue during the first operation. Attempts are usually made to prevent this cause of recurrence by peritumoral edge resection after selective adenomectomy.74 The rates of

disease recurrence vary among series, which may reflect the evaluation criteria and the technical ability of the pituitary neurosurgeons.84 After selective adenoma resection, patients commonly experience transient secondary adrenal insufficiency (of 6 months to 2 years) requiring maintenance hydrocortisone replacement. During this period, the ACTH response to CRH is subnormal.85 If ACTH and cortisol do not fall to low levels, recurrence is more likely. Within 6 to 8 months after surgery, recovery of hypothalamic pituitary adrenal function and other tropic function usually takes place. The finding of transient deficiency of ACTH secretion after resection of a microadenoma favors the hypothesis of a pituitary origin of the disease. This contrasts with cases of CRH hypersecretion in which over-stimulation of the residual corticotropes would be expected after the resection of the adenomatous tissue, leading to persistent or recurrent disease. SURGICAL RESULTS IN CHILDREN Transsphenoidal microadenomectomy is also successful in children and adolescents with Cushing disease.86 With successful treatment, growth retardation is replaced by catch-up growth or resumption of the growth rate; hypogonadism is followed by pubertal maturation and normal pubertal levels of gonadal steroids; and the blunted TSH response to thyroid-releasing hormone returns to normal. Effective treatment of Cushing disease must be instituted early to prevent early fusion of the epiphysis and permanent stunting of growth. PITUITARY IRRADIATION When transsphenoidal surgery has failed or alternative forms of treatment are desired, pituitary irradiation is an option. The most widely used type of pituitary irradiation is high-voltage irradiation provided by cobalt-60 (60Co). The total recommended dose is 40 to 50 Gy, and favorable results can be achieved in ~50% of patients.70 The best responses are observed in patients with the juvenile form of the disease or in adults younger than 40 years. When successful, 60Co irradiation has several advantages. Remission occurs with preservation of pituitary and adrenal function; panhypopituitarism seldom develops; normal reproductive function is restored when the patient is in remission; corticosteroid replacement therapy is not needed; recurrence is rare; and normal cortisol secretion may be restored (i.e., circadian rhythm, normal suppressibility on dexamethasone).87 The major disadvantage is the slow therapeutic response to pituitary irradiation; 6 to 18 months may elapse before a clinical and biochemical remission is achieved. Treatment with 60Co irradiation alone is not adequate in patients who have severe Cushing syndrome. Symptoms may progress if the patient waits for remission, and severe, perhaps irreversible, complications may result. Heavy particle beam irradiation and Bragg peak proton irradiation therapy, with a rate of improvement or remission as high as 80%, appears to be more effective than 60Co irradiation in the treatment of Cushing syndrome.88,89 However, the prevalence of postirradiation panhypopituitarism is high. Implants of gold-198 or yttrium-90 yield complete or partial response in 77% of patients with Cushing disease, but the treatment is also complicated by hypopituitarism in 30% to 50% of the patients. Some reports suggest that significant disturbances of hypothalamic-pituitary function follow megavoltage therapy; these disturbances may progress to overt hypopituitarism.90 The side effects and the efficacy of therapy may be related to the dosage regimen, with the greatest responses being observed when the highest dose of irradiation is administered. A higher incidence of postirradiation side effects, including radiation necrosis of the brain, occurs with higher dosage.91 Because this complication may occur from 2 to 20 years after completion of radiation therapy, its possibility should be entertained when local neurologic symptoms develop in patients with a history of radiation therapy for pituitary disease. Another method of focally targeted radiation therapy is stereotactic radiosurgery or gamma knife therapy. These techniques have been applied to patients who experience recurrence of Cushing disease after transsphenoidal resection of pituitary adenomas and in whom residual tumor can be detected on MRI. Radiation doses of 20 to 35 Gy delivered to the center of the tumor have produced remission of Cushing syndrome and stabilization of tumor growth.92,93 ADRENAL SURGERY In patients with advanced Cushing disease in whom transsphenoidal surgery has failed, bilateral total adrenalectomy is the preferred treatment. The major disadvantage of adrenalectomy is that it fails to attack the cause underlying the hypersecretion of ACTH. A complication that may occur months or years later is Nelson syndrome, with hyperpigmentation and an ACTH-secreting pituitary macroadenoma becoming clinically apparent.94 These ACTH-secreting tumors may become locally invasive and difficult to control by surgery or radiation therapy. Rarely, these tumors undergo distant metastases, and discrete hepatic metastatic nodules have been found. In a series of 79 consecutive patients with Cushing disease treated with bilateral adrenalectomy, three early postoperative deaths occurred, but most patients did not fare worse than the general population.95 The most common cause of death among patients who had undergone adrenalectomy for Cushing disease was cardiovascular disease; in other cases, death was the result of the local effects of pituitary tumors. Laparoscopic adrenalectomy has become a common procedure for resection of benign adrenal cortical tumors. Bilateral adrenalectomy can also be performed by laparoscopic technique. The operating time is longer than with open adrenalectomy, but the duration of the procedure depends on the surgical instrumentation used and the experience of the surgeon. Laparoscopic surgery may take 3 to 4 hours to complete in contrast to 1 to 1.5 hours for the open procedure. Estimated blood losses are comparable. The main advantage of laparoscopic surgery is the short recovery time; most patients are able to resume oral intake and ambulation 1 to 4 days postoperatively and normal daily activities in 5 to 7 days.96 DRUG THERAPY INHIBITORS OF ADRENAL FUNCTION Various inhibitors of adrenal function have been used to suppress cortisol secretion in patients with Cushing syndrome. They include aminoglutethimide, ketoconazole, and mitotane. Aminoglutethimide. When initially used as an anticonvulsive drug, aminoglutethimide was observed to inhibit cortisol secretion. Subsequently, aminoglutethimide was shown to suppress cortisol secretion effectively in patients with adrenal cancer and ACTH-dependent Cushing disease.97,98 The mechanism of action of aminoglutethimide is the inhibition of cholesterol side-chain cleavage and blocking of the conversion of cholesterol to D5 -pregnenolone in the adrenal cortex. As a result, the synthesis of cortisol, aldosterone, and androgens is inhibited. Histologically, the adrenocortical cells appear loaded with cholesterol droplets as in the lipoidic form of congenital adrenal hyperplasia. Aminoglutethimide also inhibits the aromatization of androstenedione to estrone, an effect that has made it useful in the treatment of patients with carcinoma of the breast. The drug has been used in adults and in children in doses of 0.5 to 2 g daily.99,100 Cortisol levels fall gradually, and eventually patients may need glucocorticoid replacement. In persons with cortisol-secreting adrenocortical carcinoma, the effect of amino-glutethimide can be maintained for many months with regression of the clinical manifestations of Cushing syndrome.97 The drug is only transiently effective in patients with ACTH-dependent Cushing syndrome. The inhibitory effect of the drug is overcome by the high levels of ACTH and cortisol levels, which return to pretreatment levels within days of instituting therapy.98 The effect of aminoglutethimide is promptly reversed by interruption of therapy. Aminoglutethimide causes gastrointestinal symptoms (i.e., anorexia, nausea, vomiting) and neurologic side effects (i.e., lethargy, sedation, blurred vision) and can cause hypothyroidism in 5% of patients. A skin rash is frequently observed during the first 10 days of treatment; this usually subsides despite continuation of treatment. Headaches have also been observed when the larger doses are given. Ketoconazole. Ketoconazole is an imidazole derivative that inhibits the synthesis of ergosterol in fungi and cholesterol in mammalian cells by blocking demethylation of lanosterol. In addition to the effect on cholesterol synthesis, ketoconazole inhibits mitochondrial cytochrome P450dependent enzymes, such as 11b-hydroxylase and the enzymes needed for cholesterol side-chain cleavage. The drug was also found to inhibit H-dexamethasone binding to glucocorticoid receptors in hepatoma tissue culture cell cytosol and to block glucocorticoid action. Used in clinical practice as an antifungal medication, ketoconazole has been found to be a potent inhibitor of gonadal and adrenal steroidogenesis in vivo.101 In normal men, ketoconazole, administered in doses of 200 to 600 mg per day, is a potent inhibitor of testosterone production. Ketoconazole can also inhibit abnormal cortisol production in patients with adrenal adenoma and Cushing syndrome. These patients respond promptly, with disappearance of the clinical and metabolic manifestations of the disease within 4 to 6 weeks of treatment.101 The effect of ketoconazole appears to be persistent, because no escape from its suppressive effect has been reported for patients who receive the drug for prolonged periods. When patients are treated with ketoconazole, adrenal insufficiency is avoided by decreasing the dose sufficiently to maintain normal cortisol levels. The most frequent adverse reactions to this drug are nausea and vomiting, abdominal pain, and pruritus in 1% to 3% of patients. Hepatotoxicity, primarily of the

hepatocellular type, has been associated with its use. The patient should be monitored for this side effect by performing liver function tests such as those for alkaline phosphatase, serum aspartate aminotransferase (AST), serum alanine aminotransferase (ALT), and bilirubin, which should be measured before starting treatment and at intervals during treatment. Transient minor elevations in liver enzymes have occurred during treatment, but the drug should be discontinued if even minor liver function test abnormalities persist, if the abnormalities worsen, or if they are accompanied by symptoms of possible liver injury. Mitotane. Of all the pharmacologic agents described, mitotane is the only one that inhibits biosynthesis of corticosteroids and destroys adrenocortical cells secreting cortisol, producing a long-lasting effect. Mitotane acts on adrenocortical cell mitochondria, where it inhibits 11b-hydroxylase and cholesterol side-chain cleavage enzymes. As a result of this inhibition, the production of cortisol, aldosterone, and dehydroepiandro-sterone is suppressed. Mitotane appears to require metabolism for its action. Under the effect of mitochondrial P450 monooxygenases, the drug is probably transformed into an acylchloride, which covalently binds to important macromolecules in the cell mitochondria. The result is the destruction of the mitochondria with necrosis of the adrenal cortex. The zona reticularis of the adrenal cortex appears to be most sensitive to the action of mitotane, and the glomerulosa is the least sensitive. A combination of 60Co irradiation of the pituitary gland and selective suppression of cortisol secretion with mitotane has been used in the treatment of Cushing disease.102 Eighty percent of the patients treated with this drug plus pituitary irradiation had clinical and biochemical remission of their disease (Fig. 75-10). One-half of the patients treated by this combination of therapy had suppression of their elevated cortisol levels within 4 months of therapy. The initial decrease in cortisol secretion appeared to be a response to the administration of mitotane, because plasma ACTH levels were still high when indices of cortisol secretion returned to normal. In addition to suppressing cortisol secretion, this drug appears to have a par tial suppressive effect on ACTH. After suppression of cortisol secretion, minimal feedback increase in ACTH production occurred, and in 70% of treated patients, an actual decrease in ACTH levels was observed. This response contrasts with the increase in ACTH levels in patients treated by adrenalectomy or with other adrenal inhibitors.

FIGURE 75-10. A, A 24-year-old man with severe Cushing disease failed to respond to 60Co pituitary irradiation. B, After mitotane therapy for 9 months, the features of Cushing syndrome disappeared.

Mitotane is a selective inhibitor of the reticularis and fascicu lata zones of the adrenal cortex. Aldosterone secretion is usu ally spared, and the patients do not require mineralocorticoid replacement when they develop adrenal insufficiency. The side effects of therapy include anorexia, nausea, diarrhea, somno lence, pruritus, hypercholesterolemia, and hypouricemia. Ele vated alkaline phosphatase levels can also occur as a result of the hepatotoxic effects of mitotane. Hypouricemia, a consistent finding, appears to be caused by increased renal clearance of uric acid. Low thyroxine levels found in patients treated with mitotane are probably the result of interference with protein binding for this hormone. Mitotane competitively binds to thyroxine-binding globulin in a manner similar to that of phenytoin, another diphenyl compound. Other indices of thy roid function are in the normal range. All of the side effects described can be reversed by reducing the dose or by interrupt ing therapy. Mitotane also has well-known extraadrenal effects, derived from its effect as an inducer of microsomal hydroxylating enzymes. As such, it can alter cortisol metabolism and the met abolic degradation of synthetic substituted steroids, which are predominantly inactivated by hydroxylation. A consequence of this effect is an early fall in the urinary excretion of 17-hydroxy corticosteroids, which occurs independently of the suppressive effects on cortisol secretion. Because of this effect, measurement of urinary 17-hydroxycorticosteroid levels does not provide a reliable index of adrenal suppression by the drug. Assessment of changes in cortisol secretion rates, urinary free cortisol levels, and plasma cortisol levels is required to determine the bio chemical response. Combined therapy with pituitary irradiation and mitotane can be given as primary medical treatment of Cushing syndrome when surgery is contraindicated or as the next modality of therapy when transsphenoidal pituitary surgery fails to bring about remission of the disease. The dose used for the treatment of Cushing disease is 2 to 4 g daily. Therapy may be initiated with a dose of 500 mg twice daily and increased to 1 g four times daily, as needed to achieve adrenal-suppressive effects. The urinary free cortisol excretion should be monitored, and the dose should be titrated to maintain the urinary free cortisol excretion in the normal range. If adrenal insufficiency is suspected, hydrocortisone should be given orally. Because mitotane induces liver monooxygenases, which metabolize corticosteroids and other drugs, the adequate dosage of fludrocortisone or hydrocortisone may be higher than expected. As the effect of radiation therapy becomes apparent with suppression of ACTH secretion, treatment with the drug can be gradually discontinued. A compound of a different class, mifepristone or RU-486 [17-b-hydroxy-11-b-(4-dimethyl amino)-17-a-(1-propynyl) estra-4,9-dien-3-one] is a glucocorticoid antagonist that has been shown to effectively antagonize hypercortisolism causing Cushing syndrome.103 This compound is a 19-norsteroid with substitutions at positions C11 and C17; it antagonizes cortisol action competitively at the receptor level. With therapy, the somatic features of Cushing syndrome ameliorate, mean arterial blood pressure is normalized, and suicidal depression can resolve. All biochemical glucocorticoid-sensitive parameters normalize, and ACTH and cortisol levels, as expected, remain unchanged. Although this drug may be effective in patients with autonomous cortisol production, the blocking effect of the compound on cortisol receptors may increase the secretion of ACTH in patients with pituitary ACTH-dependent Cushing syndrome and cause a further increase in cortisol production. This additional cortisol production may be sufficient to overcome the antiglucocorticoid effect of the drug. Mifepristone is well tolerated and nontoxic in doses ranging between 10 and 20 mg/kg per day. It is still an investigational drug, and its use is limited to investigational protocols. TREATMENT OF ECTOPIC ADRENOCORTICOTROPIN SYNDROME Treatment of the ectopic ACTH syndrome involves the surgical resection of the primary tumor, followed by radiation therapy or chemotherapy, depending on the type of neoplasm producing the illness. In patients whose neoplasms cannot be resected, the use of adrenal inhibitors such as aminoglutethimide, metyrapone, and ketoconazole may ameliorate the clinical manifestations of the Cushing syndrome. However, the very high ACTH levels may overcome the suppressive effect of these drugs. Bilateral adrenalectomy is an alternative approach but is not a practical one for patients who have rapidly progressive metastatic disease. Because the underlying tumors may occasionally be slow growing (e.g., metastatic carcinoid, bronchial carcinoids), an adrenalectomy followed by replacement therapy with normal amounts of hydrocortisone may improve considerably the metabolic consequence of the Cushing syndrome. TREATMENT OF CORTICOTROPIN-INDEPENDENT CUSHING SYNDROME ADRENOCORTICAL ADENOMA Adrenocortical adenomas should be surgically removed.104 Because of suppression of the hypothalamicpituitaryadrenal axis in these patients, adrenal insufficiency occurs postoperatively. These patients require replacement therapy with physiologic doses of cortisol until recovery of the hypothalamic pituitaryadrenal axis takes place. This recovery may require 6 to 16 months. Occasionally, recovery of pituitary-adrenal function never occurs and patients require lifelong replacement therapy with cortisol. ADRENAL CARCINOMA Adrenal carcinomas causing Cushing syndrome are highly malignant neoplasms resulting in a shortened life expectancy.105,105a Their treatment has not been well standardized, and the prognosis has been poor regardless of therapy. Prognosis depends on the stage of the tumor at the time of diagnosis. Staging by size and extent (Table 75-5) is helpful in assessing the magnitude of the disease and probable life expectancy.106 Patients with stage I or stage II tumors may experience cure or long

disease-free periods. Patients with stage IV tumors have short life expectancy even with currently available chemotherapy.

TABLE 75-5. Staging of Adrenal Cortical Carcinoma Based on Size and Extent of Tumor Involvement

Several approaches to therapy have been used. One method is surgical excision of the primary tumor and of large neoplastic abdominal masses. Although temporary remission of the disease frequently occurs with this approach, recurrence and eventual death from metastatic disease is the rule.107 Another approach is radiation therapy and nonspecific chemotherapy.108 These have been generally effective only for palliation of local disease, because this neoplasm is generally resistant to these types of therapy. Agents that block corticosteroid hormone production by the tumor have also been used. Although effective in reversing the metabolic consequences of these tumors, these drugs do not alter the progression and eventual fatal outcome of the disease. Mitotane has been the only drug that has proven effective in patients with adrenocortical carcinoma.109 Most of the experience with mitotane comes from its use in patients with obvious metastases, but its effectiveness under those conditions has been disputed in the literature.107,110,111 and 112 Decreases in elevated urinary steroid levels, measurable disease response, and overall clinical response have been described.111 However, mean survival is short (8 months) when the drug is used after the appearance of metastatic disease. Isolated case reports have described impressive remissions and even cures of adrenocortical carcinoma after therapy with this drug.109,113,114 In general, survival appears to depend on the size of the primary lesion and the degree of local and distant extension of the neoplasm at the time of initial surgery. Criteria for staging adrenal cancer have been suggested. Patients with primary tumors <5 cm in diameter and without local or distant extension appear to have relatively good prognosis and longer survival. If mitotane therapy is to be effective, it should be instituted early, as adjuvant therapy, after the resection of the primary tumor and before local extension or distant metastasis has occurred.112,115 Adverse effects of mitotane therapy are found to be dose dependent; they are intolerable with doses higher than 6 g daily. Treatment is usually begun with doses of 1 g twice daily and gradually increased up to tolerance. The drug should be administered with fat-containing food, because its absorption and transport appear to be coupled to lipoproteins. The prominent early side effects of larger doses are anorexia and nausea. The discomfort with these side effects can be minimized by administering the largest dose at bedtime so that patients sleep through the most uncomfortable period. Side effects are usually reversed by interrupting therapy for several days and restarting the drug at a lower dose level. Because of possible teratogenic effects, patients on mitotane should be advised against pregnancy. CHAPTER REFERENCES
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. Gold EM. The Cushing's syndromes: changing views of diagnosis and treatment. Ann Intern Med 1979; 90:829. Kathol RG, Delahunt JW, Hannah L. Transition from bipolar affective disorder to intermittent Cushing's syndrome: case report. J Clin Psychiatry 1985; 46:194. Kelly WF, Checkley SA, Bender BA. Cushing's syndrome and depressiona prospective study of 26 patients. Br J Psychiatry 1983; 142:16. Reed K, Watkins M, Dobson H. Mania in Cushing's syndrome: case report. J Clin Psychiatry 1983; 44:416. Saad MF, Adams F, Mackay B, et al. Occult Cushing's disease presenting with acute psychosis. Am J Med 1984; 76:759. Starkman MN, Schteingart DE. Neuropsychiatric manifestations of patients with Cushing's syndrome. Arch Intern Med 1981; 141:215. Schteingart DE. The diagnosis and medical management of Cushing's syndrome. In: Thompson NW, Vinik AL, eds. Endocrine surgery update. New York: Grune & Stratton, 1983:87. Daughaday, WH. Cushing's disease and basophilic microadenomaseditorial retrospective. N Engl J Med 1984; 310:919. Lamberts SWJ, Delange SA, Stefanko SZ. Adrenocorticotropin-secreting pituitary adenomas originate from the anterior or intermediate lobe in Cushing's disease: differences in the regulation of hormone secretion. J Clin Endocrinol Metab 1982; 54:286. Lamberts SWJ, Stefanko SA, deLange SA, et al. Failure of clinical remission after transsphenoidal removal of a microadenoma in a patient with Cushing's disease: multiple hyperplastic and adenomatous cell nests in surrounding pituitary tissue. J Clin Endocrinol Metab 1980; 50:793. Young WF, Scheithauer BW, Gharib H, et al. Cushing's syndrome due to primary multinodular corticotrope hyperplasia. Mayo Clin Proc 1988; 63:256. McKeever PE, Koppelman MCS, Metcalf D, et al. Refractory Cushing's disease caused by multinodular ACTH-cell hyperplasia. J Neuropathol Exp Neurol 1982; 41:490. Biller BM, Alexander JM, Zervas NT, et al. Clonal origins of adrenocorti-cotropin-secreting pituitary tissue in Cushing's disease. J Clin Endocrinol Metab 1992; 75(5):1303. Glass AR, Zabadil AP, Halberg F, et al. Circadian rhythm of serum cortisol in Cushing's disease. J Clin Endocrinol Metab 1984; 59:161. Schteingart DE, McKenzie AK. Twelve hour cycles of ACTH and cortisol secretion in Cushing's disease. J Clin Endocrinol Metab 1980; 51:1195. Liddle GW, Island DP, Ney RI. Cushing's syndrome caused by recurrent malignant bronchial carcinoid. Arch Intern Med 1963; 111:471. Schteingart DE. Ectopic secretion of peptides of the proopiomelanocortin family. Endocrinol Metab Clin North Am 1991; 20:453. Melmed S, Rushakoff RJ. Ectopic pituitary and hypothalamic hormone syndromes. Endocrinol Metab Clin North Am 1987; 16:805. DeBold CR, Menefee JK, Nicholson WE, et al. Proopiomelanocortin gene is expressed in many normal tissues and in tumors not associated with ectopic adrenocorticotropin syndrome. Mol Endocrinol 1988; 2:862. White A, Clark AJ, Stewart MF. The synthesis of ACTH and related peptides by tumors. Baillires Clin Endocrinol Metab 1990; 4:1. Picon A, Leblond-Francillard M, Raffin-Sanson ML, et al. Functional analysis of the human proopiomelanocortin promoter in the small cell lung carcinoma cell line DMS-79. J Mol Endocrinol 1995; 15:187. Schteingart DE, Chandler WF, Lloyd RV, et al. Cushing's syndrome caused by an ectopic pituitary adenoma. Neurosurgery 1987; 21:223. Belsky JL, Cuello B, Swanson LW, et al. Cushing's syndrome due to ectopic production of corticotropin-releasing factor. J Clin Endocrinol Metab 1985; 60:496. Schteingart DE, Lloyd RV, Akil H, et al. Cushing's syndrome secondary to ectopic CRH-ACTH secretion. J Clin Endocrinol Metab 1986; 63:770. Marieb NJ, Spangler S, Kashgarian M, et al. Cushing's syndrome secondary to ectopic cortisol production by ovarian carcinoma. J Clin Endocrinol Metab 1983; 57:737. Smalls AGH, Pieters GFFM, Van Haelst UJG, et al. Micronodular adrenocortical hyperplasia in long standing Cushing's disease. J Clin Endocrinol Metab 1983; 58:25. Lamberts SWJ, Bons EG, Bruining HA. Different sensitivity to adrenocorticotropin of dispersed adrenocortical cells from patients with Cushing's disease with macronodular and diffuse adrenal hyperplasia. J Clin Endocrinol Metab 1984; 58:1106. Shenoy BV, Carpenter PC, Carney JA. Bilateral primary pigmented nodular adrenal cortical diseaserare cause of Cushing's syndrome. Am J Surg Pathol 1984; 8:335. Stratakis CA, Carney JA, Lin JP, et al. Carney complex. A familial multiple neoplasia and lentiginosis syndrome: analysis of 11 kindreds and linkage to the short arm of chromosome 2. J Clin Invest 1996; 97(3):699. Terzolo M, Boccuzzi A, Ali A, et al. Cushing's syndrome due to ACTH-independent bilateral adrenocortical hyperplasia. J Endocrinol Invest 1997; 20 (5):270. Lacroix A, Bolte E, Tremblay J, et al. Gastric inhibitory polypeptide-dependent cortisol hypersecretiona new cause of Cushing's syndrome. N Engl J Med 1992; 327(14):974. Lacroix A, Tremblay J, Russeau G, et al. Brief report: propranolol therapy for ectopic (beta)-adrenergic receptors in adrenal Cushing's syndrome. N Engl J Med 1997; 337(20):1429. Lynch HT, Lynch J, Conway T, et al. Hereditary breast cancer and family cancer syndromes. World J Surg 1994; 18(1):21. Tsunematsu Y, Watanabe S, Oka T. et al. Familiar aggregation of cancer from proband cases with childhood adrenal cortical carcinoma. Jap J Cancer Res 1991; 82(8):893.

34a. Schteingart DE. Management approaches to adrenal incidentalomas. A view from Ann Arbor, Michigan. Endocrinol Metab Clin North Am 2000; 29(1):127. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. Crapo L. Cushing's syndrome: a review of diagnostic tests. Metabolism 1971; 28:955. Kennedy L, Atkinson AB, Johnston H, et al. Serum cortisol concentrations during low dose dexamethasone suppression test to screen for Cushing's syndrome. BMJ 1984; 289:1188. Raff H, Raff JL, Findling JW. Late-night salivary cortisol as a screening test for Cushing's syndrome. J Clin Endocrinol Metab 1998; 83(8):2681. Bruno OD, Rossi MA, Contreras LN, et al. Nocturnal high dose dexa-methasone suppression test in the etiological diagnosis of Cushing's syndrome. Acta Endocrinol 1985; 109:158. Abou Samra AB, Dechaud H, Estour B, et al. Beta-lipotropin and cortisol responses to intravenous infusion dexamethasone suppression test in Cushing's syndrome and obesity. J Clin Endocrinol Metab 1985; 61:116. Chrousos GP, Schulte HM, Oldfield EH, et al. A corticotropin-releasing factor stimulation testan aid in the evaluation of patients with Cushing's syndrome. N Engl J Med 1984; 310:622. Orth DN, DeBold CR, DeCherney GS, et al. Pituitary microadenomas causing Cushing's disease respond to corticotropin-releasing factor. J Clin Endocrinol Metab 1982; 55:1017. Joebson RV, Hockings GI, Torpy PJ. New diagnostic tests for Cushing's syndrome. Clin Exp Physiol 1996; 23(67):579. Yanovski JA, Cutler GB, Chrousos GP, Nieman LK. The dexamethasone-suppressed, corticotropin-releasing hormone stimulation test differentiates mild Cushing's disease from normal physiology. J Clin Endocrinol Metab 1998; 83(2):348. Dunlap NE, Grizzle WE, Siegel AL. Cushing'sscreening methods in hospitalized patients. Arch Pathol Lab Med 1985; 109:222. Werk EE, MacGee J, Sholiton LJ. Effect of diphenylhydantoin on cortisol metabolism in man. J Clin Invest 1964; 43:1824. Kapcala LP, Hamilton SM, Meikle AW. Cushing's disease with normal suppression due to decreased dexamethasone clearance. Arch Intern Med 1984; 144:636. Oldfield EH, Chrousos GP, Schulte HM, et al. Preoperative lateralization of ACTH-secreting pituitary microadenomas by bilateral and simultaneous inferior petrosal venous sinus sampling. N Engl J Med 1985; 312:100. Snow RB, Patterson RH, Howith M, et al. Usefulness of preoperative inferior petrosal vein sampling in Cushing's disease. Surg Neurol 1988; 29:17. Zovickian J, Oldfield EH, Doppman JL, et al. Usefulness of inferior petrosal sinus venous endocrine markers in Cushing's disease. J Neurosurg 1988; 68:205. Doppman JL, Oldfield EH, Nieman LK. Bilateral sampling of internal jugular vein to distinguish between mechanisms of adrenocorticotropic hormone-dependent-Cushing syndrome. Ann Intern Med 1998; 128(1):33.

51. Moses DC, Schteingart DE, Sturman MF, et al. Efficacy of radiocholesterol imaging of the adrenal glands in Cushing's syndrome. Surg Obstet Gynecol 1974; 139:1. 52. Smith SM, Patel SK, Turner DA, Matalon TA. Magnetic resonance imaging of adrenal cortical carcinoma. Urol Radiol 1989; 11:1. 53. Hamper UM, Fishman EK, Hartman DS, et al. Primary adrenocortical carcinoma: sonographic evaluation with clinical and pathological correlation in 26 patients. AJR Am J Roentgenol 1987; 148:915. 54. Schteingart DE, Seabold JE, Gross MD, et al. Iodocholesterol adrenal tissue uptake and imaging in adrenal neoplasms. J Clin Endocrinol Metab 1981; 52:1157. 55. Doppman JL, Miller DL, Dwyer AJ, et al. Macronodular adrenal hyperplasia in Cushing's disease. Radiology 1988; 166:347. 56. DeStefano D, Lloyd RV, Schteingart DE. Cushing's syndrome produced by a bronchial carcinoid tumor. Hum Pathol 1984; 15:890. 57. Philipponneau M, Nocaudie M, Epelbaum J, et al. Somatostatin analogs for the localization and preoperative treatment of adrenocorticotropin-secreting bronchial carcinoid tumor. J Clin Endocrinol Metab 1994; 78:20. 58. Barbareschi M, Mariscotti C, Frigo B, et al. Mediastinal malignant carcinoid with Cushing's syndrome: immunohistochemical and ultrastructural study. Appl Pathol 1989; 7:161. 59. Coates PJ, Doniach I, Howlett TA, et al. Immunocytochemical study of 18 tumors causing ectopic Cushing's syndrome. J Clin Pathol 1986; 39:955. 60. Doppman JL, Loughlin T, Miller DL, et al. Identification of ACTH producing intrathoracic tumors by measuring ACTH levels in aspirated specimens. Radiology 1987; 163:501. 61. Lapidus L, Bengtsson C, Hllstrm T, Bjrntorp P. Obesity, adipose tissue distribution and health in women. Results from a population study in Gothenburg, Sweden. Appetite 1989; 12:25. 62. Schteingart DE, Gregerman RI, Conn JW. A comparison of the characteristics of increased adrenocortical function in obesity and in Cushing's syndrome. Metabolism 1963; 12:484. 63. Marin P, Darin N, Aneniva T, et al. Cortisol secretion in relationship to body fat distribution in premenopausal women. Metabolism 1992; 41:882. 64. Starkman MN. The HPA axis and psychopathology. Cushing's syndrome. Psychiatr Ann 1993; 23:691. 65. Gold PW, Loriaux DL, Roy A, et al. Responses to corticotropin-releasing hormone in the hypercortisolism of depression and Cushing's disease. N Engl J Med 1986; 314:1329. 66. Rees LH, Besser GM, Jeffcoate WF, et al. Alcohol-induced pseudo-Cushing's syndrome. Lancet 1977; 1:726. 67. Lambert SW, Klijn JGM, deJong FH, Birkenhager JC. Hormone secretion in alcohol-induced pseudo-Cushing's syndrome: differential diagnosis with Cushing's disease. JAMA 1979; 242:1640. 68. Jeffcoate W. Alcohol-induced pseudo-Cushing's syndrome. Lancet 1993; 341(8846):676. 69. Yanovski JA, Cutler GB, Chrousos GP, Nieman LK. Corticotropin-releasing hormone stimulation following low-dose dexamethasone administration: a new test to distinguish Cushing's syndrome from pseudo-Cushing's states. JAMA 1993; 269(17):2232. 70. Aron DC. Cushing's syndrome: current concepts in diagnosis and treatment. Compr Ther 1987; 13:37. 71. Jeffcoate WJ. Treating Cushing's disease. BMJ 1988; 296:227. 72. Orth D, Liddle GW. Results of treatment in 108 patients with Cushing's syndrome. N Engl J Med 1971; 285:243. 73. Boggan JE, Tyrrell JB, Wilson CB. Transsphenoidal microsurgical management of Cushing's diseasereport of 100 cases. J Neurosurg 1983; 59:195. 74. Nakane T, Kuwayama A, Watanabe M, et al. Long term results of transsphenoidal adenomectomy in patients with Cushing's disease. Neurosurgery 1987; 21:218. 75. Pelkonen R, Eistola P, Grahme B, et al. Treatment of pituitary Cushing's disease: results of adrenal and pituitary surgery. Acta Endocrinol Suppl 1983; 251:38. 76. Hardy J. Transsphenoidal surgery of hypersecreting pituitary tumors. In: Kohler PO, Ross GT, eds. Diagnosis and treatment of pituitary tumors. Amsterdam: Excerpta Medica, 1979:179. 77. Salassa RM, Laws ER Jr, Carpenter PC, et al. Transsphenoidal removal of pituitary microadenoma in Cushing's disease. Mayo Clin Proc 1978; 53:24. 78. Tyrrell JB, Brooks RM, Fitzgerald PA, et al. Cushing's disease: selective transsphenoidal resection of pituitary microadenomas. N Engl J Med 1978; 289:753. 79. Kuwayama A, Kageyama N. Current management of Cushing's disease. Part I. Contemp Neurosurg 1985; 7:1. 80. Swearingen B, Biller BM, Barker FG II, et al. Long-term mortality after transsphenoidal surgery for Cushing disease. Ann Intern Med 1999; 130 (10):821. 81. Chandler WF, Schteingart DE, Lloyd RV, et al. Surgical treatment of Cushing's disease. J Neurosurg 1987; 66:204. 82. Schnall AM, Brodkey JS, Kaufman B, et al. Pituitary function after removal of pituitary microadenomas in Cushing's disease. J Clin Endocrinol Metab 1978; 47:410. 83. Ciric I, Ragin A, Baumgartner C, Pierce D. Complications of transsphenoidal surgery: results of a national survey, review of the literature, and personal experience. Neurosurgery 1997; 40(2):225. 84. Burch WM. Diagnosis and treatment of Cushing's disease at Duke University (19771982). N C Med J 1983; 44:293. 85. Avgerinos PC, Chrousos GP, Nieman LK, et al. The corticotropin-releasing hormone test in the postoperative evaluation of patients with Cushing's syndrome. J Clin Endocrinol Metab 1987; 65:906. 86. Styne DM, Grumbach MM, Kaplan SL, et al. Treatment of Cushing's disease in childhood and adolescence by transphenoidal microadenomectomy. N Engl J Med 1984; 310:889. 87. Futterweit W, Krieger DT, Gabrilove JL. Adrenal cortical function studies in Cushing's syndrome due to nontumorous adrenocortical hyperfunction treated with pituitary irradiation. J Clin Endocrinol Metab 1962; 22:364. 88. Kjellberg RN, Kliman B. A system for therapy of pituitary tumors. In: Kohler PO, Ross GT, eds. Diagnosis and treatment of pituitary tumors. Amsterdam: Excerpta Medica, 1973:234. 89. Lawrence JH, Okerlund MD, Linfoot JE, et al. Heavy particle treatment of Cushing's disease. N Engl J Med 1971; 285:1263. 90. Sharpe GF, Kendall-Taylor P, Prescott RWG, et al. Pituitary function following megavoltage therapy for Cushing's disease: long-term follow-up. Clin Endocrinol 1985; 22:169. 91. Aristizabal S, Caldwell WL, Avila J, et al. Relationship of time/dose factors to tumor control and complications in the treatment of Cushing's disease by irradiation. Int J Radiat Oncol Biol Phys 1977; 2:47. 92. Seo Y, Fukuoka S, Takanashi M, et al. Gamma knife surgery for Cushing's disease. Surg Neurol 1995; 43(2):170. 93. Wolffenbuttel BH, Kitz K, Beuls EM. Beneficial gamma-knife radiosurgery in a patient with Nelson's syndrome. Clin Neurol Neurosurg 1998; 100(1):60. 94. Nelson DH, Meakin JW, Thorn GW. ACTH-producing pituitary tumors following adrenalectomy for Cushing's syndrome. Ann Intern Med 1970; 52:560. 95. Welbourn RB. Survival and cause of death after an adrenalectomy for Cushing's disease. Surgery 1985; 97:16. 96. Gagner M, Pomp A, Heniford BT, et al. Laparoscopic adrenalectomy: lessons learned from 100 consecutive procedures. Ann Surg 1997; 226(3):238. 97. Schteingart DE, Cash R, Conn JW. Aminoglutethimide and metastatic adrenal cancer. Maintained reversal (six months) of Cushing's syndrome. JAMA 1966; 198:1007. 98. Schteingart DE, Conn JW. Effects of aminoglutethimide upon adrenal function and cortisol metabolism in Cushing's syndrome. J Clin Endocrinol Metab 1967; 27:1657. 99. Misbin RI, Canary J, Willard D. Aminoglutethimide in the treatment of Cushing's syndrome. J Clin Pharmacol 1976; 16:645. 100. Zachman N, Gitzelmann RT, Zagalaka M, et al. Effect of aminoglutethimide on urinary cortisol and cortisol metabolites in adolescents with Cushing's syndrome. Clin Endocrinol 1977; 7:63. 101. Sonino N, Boscaro M, Merola G, et al. Prolonged treatment of Cushing's disease by ketoconazole. J Clin Endocrinol Metab 1985; 61:718. 102. Schteingart DE, Tsao HS, Taylor CI, et al. Sustained remission of Cushing's disease with mitotane and pituitary irradiation. Ann Intern Med 1980; 92:613. 103. Nieman LK, Chrousos GP, Kellner C, et al. Successful treatment of Cushing's syndrome with a glucocorticoid antagonist RU486. J Clin Endocrinol Metab 1985; 71:536. 104. Scott HW, Abumrad NN, Orth DN. Tumors of the adrenal cortex in Cushing's syndrome. Ann Surg 1985; 201:586. 105. Huvos AJ, Hajdu SI, Brosfield RD, et al. Adrenal cortical carcinoma: clinico-pathologic study of 34 cases. Cancer 1970; 25:354. 105a. Wajchenberg BL, Albegaria Pereira MA, Medonca BB, et al. Adrenocortical carcinoma : clinical and laboratory observations. Cancer 2000; 88:711. 106. 107. 108. 109. 110. 111. 112. 113. 114. 115. Hogan T. A clinical and pathological study of adrenocortical carcinoma; therapeutic implications. Cancer 1980; 45:2880. Hajjar RA, Hickey RC, Samaan NA. Adrenal cortical carcinoma: a study of 22 patients. Cancer 1975; 35:549. Percarpio B, Knowlton AH. Radiation therapy of adrenal cortical carcinoma. Acta Radiol Oncol Radiat Phys Biol 1976; 15:288. Becker D, Schumacher OP. o,p'-DDD therapy in invasive adrenocortical carcinoma. Ann Intern Med 1977; 82:677. Hoffman DL, Mattox VL. Treatment of adrenocortical carcinoma with o,p'-DDD. Med Clin North Am 1972; 50:999. Hutter AM Jr, Kayhoe DE. Adrenal cortical carcinoma: results of treatment with o,p'-DDD in 138 patients. Am J Med 1966; 41:581. Lubitz JA, Freeman L, Okun R. Mitotane use in inoperable adrenal cortical carcinoma. JAMA 1973; 223:1109. Downing V, Eule J, Huseby RA. Regression of an adrenal cortical carcinoma and its neovascular bed, following mitotane therapy: a case report. Cancer 1974; 34:1882. Jarabak J, Rice K. Metastatic adrenocortical carcinoma: prolonged regression with mitotane therapy. JAMA 1981; 246:1706. Schteingart DE, Motazedi A, Noonan RA, et al. The treatment of adrenal carcinoma. Arch Surg 1982; 117:1142.

CHAPTER 76 ADRENOCORTICAL INSUFFICIENCY Principles and Practice of Endocrinology and Metabolism

CHAPTER 76 ADRENOCORTICAL INSUFFICIENCY

D. LYNN LORIAUX Syndromes of Adrenal Insufficiency Primary Adrenal Insufficiency Secondary Adrenal Insufficiency Symptoms and Signs Laboratory Findings Diagnosis Acute Adrenal Insufficiency Chronic Adrenal Insufficiency Treatment Chapter References

SYNDROMES OF ADRENAL INSUFFICIENCY

The adrenal glands produce three classes of steroid hormone: glucocorticoid, mineralocorticoid, and adrenal androgen (actually, androgen precursor). The primary glucocorticoid in humans is hydrocortisone. This hormone is essential for life. The syndrome of adrenal insufficiency is the clinical manifestation of deficient production of this hormone. The central pathophysiologic alteration is cardiovascular: reduced cardiac output and decreased vascular tone with relative hypovolemia. This stimulates increased vasopressin secretion, which results in water retention and hyponatremia. If salt balance is not maintained, a situation that occurs if aldosterone also is deficient, hyponatremia is accompanied by hyperkalemia. Left untreated, this disorder leads to debility, prostration, coma, and death. The causes of adrenal insufficiency can be considered under two general headings: primary adrenal insufficiency (Addison disease), caused by destruction of the adrenal glands themselves, and secondary adrenal insufficiency, caused by disordered hypothalamic-pituitary function leading to a relative or complete deficiency of adrenocorticotropic hormone (ACTH).

PRIMARY ADRENAL INSUFFICIENCY

The two main causes of primary adrenal disease are tuberculosis and autoimmune adrenal destruction.1 The latter is a component of the polyendocrine deficiency syndrome2 (see Chap. 197). The causes of primary adrenal insufficiency are listed in Table 76-1.

TABLE 76-1. Etiology of Adrenal Insufficiency

Estimates are that 70% of the cases of primary adrenal insufficiency in the industrialized world are associated with the polyendocrine deficiency syndrome. This disorder is autoimmune. Findings that support this conclusion include a high prevalence of suppressor T-cell abnormalities, the presence of antibodies against endocrine organs, and an association with other disorders of known autoimmune etiology such as vitiligo, alopecia areata, celiac disease, autoimmune thyroiditis, and mucocutaneous candidiasis.3,4 The histologic picture in the adrenal glands reveals lymphocytic infiltration, fibrosis, a loss of cortical cells, and an intact adrenal medulla. The disorder has two recognizable subtypes (Table 76-2). Type 1 is an illness of childhood, with a mean age of onset of 12 years. Adrenal insufficiency, hypoparathyroidism, and mucocutaneous candidiasis are the most common manifestations. Associated conditions include hypogonadism, pernicious anemia, chronic active hepatitis, and alopecia. The disorder has a recessive pattern of inheritance, clustering across sibships without apparent vertical transmission (see Chap. 187).

TABLE 76-2. Fatures of the Autoimmune Polyglandular Syndromes

Type 2 is a disorder of young adults. The mean age of onset is in the mid-20s. Adrenal insufficiency, autoimmune thyroid disease, and diabetes mellitus are the most common manifestations. Associated immune disorders are rare. Susceptibility to this disorder seems to be inherited as a dominant trait in linkage disequilibrium with the HLA-B region of chromosome 6 (see Chap. 194). This form of the disease exhibits a vertical pattern of transmission.5 Worldwide, tuberculosis still is the most common cause of primary adrenal insufficiency6 (see Chap. 213). The adrenal glands can be replaced entirely by caseating granulomas. This phenomenon is virtually always accompanied by evidence of tuberculosis in other organ systems, especially the lungs and kidneys. Tuberculous adrenal glands are large and often contain calcium. This contrasts strikingly with the normal-sized or small, noncalcified glands found in patients with the polyendocrine deficiency syndrome. Fungi can colonize and destroy the adrenal cortex. All the commonly occurring fungi, except Candida, can cause adrenal insufficiency.7,8,9 and 10 Histoplasmosis is the most frequently occurring. The adrenal glands are involved in more than one-third of persons who die of histoplasmosis. This fungus is endemic in the Piedmont plateau of the Middle Atlantic states and in the Ohio and Tennessee Valley regions of the Mississippi watershed. The gross and microscopic anatomic findings in adrenal glands infected with fungi are similar to those seen in tuberculosis. Rare causes of adrenal insufficiency include amyloidosis,11 adrenal hemorrhage12 (which is especially common during sepsis and anticoagulation), congenital adrenal hypoplasia,13 and two demyelinating disordersadrenoleukodystrophy, or Brown-Schilder disease, and adrenomyeloneuropathy, or sudanophilic leukodystrophy.14,15 and 16 The demyelinating diseases are similar in that they have an X-linked inheritance pattern (and, hence, are diseases of males) and markedly increased plasma

C-26 fatty acid concentrations that are thought to reflect the biochemical abnormality responsible for the disorders. Both disorders can be diagnosed by showing an increased plasma concentration of C-26 fatty acids and by demonstrating the pathognomonic, multilamellated inclusion bodies in the steroidogenic cells of the adrenal glands and testes. The two diseases differ in that adrenoleukodystrophy, a disorder of childhood, is dominated by a central neuropathy featuring cortical blindness and coma. Adrenomyeloneuropathy, a disorder of young adults, is dominated by an ascending mixed motor and sensory neuropathy culminating in spastic paraparesis.16a Two rare causes of primary adrenal insufficiency include heparin-induced thrombosis and adrenal insufficiency associated with the antiphospholipid antibody syndrome.17,18 Metastatic cancer often is cited as a cause of adrenal insufficiency. Although the adrenal glands are a common site of metastasis, these lesions uncommonly result in adrenal insufficiency.19 When adrenal insufficiency does occur in association with malignancy, the malignancy usually is an infiltrative neoplasm, such as lymphoma or leukemia, although metastatic carcinoma of the lung or breast are occasional offenders. Finally, the acquired immunodeficiency syndrome has an association with adrenal insufficiency (see Chap. 213). Whether this is the result of a direct effect of the virus on the adrenal glands or of a secondary superinfection (i.e., Cryptococcus, Mycobacterium sp., or cytomegalovirus) remains unresolved.20,21 and 22

SECONDARY ADRENAL INSUFFICIENCY

Secondary adrenal insufficiency has three causes: adrenal suppression after exogenous glucocorticoid or ACTH administration, adrenal suppression after the correction of endogenous glucocorticoid hypersecretion, and abnormalities of the hypothalamus or pituitary gland leading to ACTH deficiency. The histologic appearance of the adrenal glands in secondary adrenal insufficiency varies from normal-appearing to simple atrophy of the cortex with a normal medulla. Adrenal suppression by exogenous glucocorticoids is the most common cause of secondary adrenal insufficiency.23,24 and 25 The frequent use of glucocorticoids to treat inflammatory disorders, such as dermatitis, arthritis, and hepatitis, is the underlying reason. Supraphysiologic doses of glucocorticoids given long enough suppress hypothalamic corticotropin-releasing hormone production and the ability of the anterior pituitary gland to respond to this hormone. The degree of adrenal suppression depends on three variables: dosage, schedule of administration, and duration of administration. Significant adrenal suppression is rarely seen with dosages of hydrocortisone (or its equivalent) of <15 mg/m 2 per day. A divided dosage schedule is more suppressive than is a once-a-day or once-every-other-day schedule. Finally, the longer the duration of administration, the greater the likelihood of suppression. Treatment periods of <14 days, for example, rarely lead to clinically important adrenal suppression, whereas treatment periods long enough to allow the emergence of the signs of Cushing syndrome usually are associated with clinically significant suppression of adrenal function. Secondary adrenal insufficiency can become manifest shortly after the cessation of corticosteroid therapy or months later in a stressful setting such as after an injury or a surgical procedure. The duration of impairment can be as long as 1 year after the correction of hypercortisolism.26 These findings have led to the conservative practice of replacing glucocorticoid before an anticipated stress in any patient who has received supraphysio-logic dosages of glucocorticoids within the past year. An alternative, and more satisfactory, practice is to test the functional capacity of the adrenal glands directly with ACTH and to base the need for glucocorticoid supplementation on the results (see Chap. 78 for an extensive discussion of exogenous corticosteroid administration and its complications).27 Any lesion of the hypothalamus or pituitary gland can lead to secondary adrenal insufficiency; examples include space-occupying lesions such as craniopharyngioma,28 pituitary adenoma,29 metastases from distant malignancies,30 sarcoidosis,31 and infections with fungi (Nocardia, Actinomyces) or the tubercle bacillus.32 Trauma to the stalk or its blood supply also can lead to adrenal insufficiency.33 A deficiency of ACTH in the absence of any of these underlying causes is rare (see Chap. 11 and Chap. 17).

SYMPTOMS AND SIGNS

The symptoms and signs of primary and secondary adrenal insufficiency can be categorized as related to the chronic and acute syndromes.34 Symptoms of the chronic syndrome include weakness, fatigue, anorexia, nausea, abdominal pain, and diarrhea. An occasional patient, particularly one with primary adrenal insufficiency, complains of orthostatic dizziness and, rarely, syncope. Signs include weight loss and, particularly in primary insufficiency, orthostatic hypotension. Symptoms and signs specific to primary adrenal insufficiency include salt craving and pigmentation of the skin and mucous membranes (Fig. 76-1,Fig. 76-2 and Fig. 76-3). Vitiligo and alopecia also can occur in the autoimmune form of primary adrenal insufficiency.35 Radiologic findings include large, often calcified, glands in patients with an infectious cause.20 The symptoms of the acute syndrome include muscle, joint, and abdominal pain, and postural hypotension. Associated signs include fever, hypotension, and clouded sensorium. The patient may appear dehydrated. No hyperpigmentation is seen in acute primary adrenal insufficiency, unless the acute insufficiency is superimposed on a prior chronic condition.

FIGURE 76-1. An example of the increased skin pigmentation most characteristic of chronic primary adrenal insufficiency, with accentuation in the creases of the fingers and palms (arrows).

FIGURE 76-2. Patient with autoimmune adrenal insufficiency. Note the darkening of sun-exposed areas of the face, neck, and arms; the dark nipples; and the darkened scars of the pretibial region. Other common sites of pigmentation in this disease are areas of pressure or irritation, such as the elbows, knees, and knuckles.

FIGURE 76-3. Patient with tuberculous Addison disease before (left) and 2 years after (right) commencement of cortisol therapy. The hyperpigmentation of the face and lips has disappeared. Note the lightening of the freckle (arrowhead).

LABORATORY FINDINGS
The complete blood count can show a normocytic, normochromic anemia; neutropenia and eosinophilia (rarely >10%); and a relative lymphocytosis. Common chemical abnormalities include metabolic acidosis and prerenal azotemia. If the patient has not been eating, hypoglycemia can occur. Hyponatremia is found in 90% of patients with chronic primary adrenal insufficiency. Hyperkalemia occurs in 65%.36 Hypercalcemia occurs, but is rare37 (see Chap. 59). The electrolyte abnormalities differ in primary and secondary adrenal insufficiency. The pattern of electrolyte alterations in primary adrenal insufficiency usually is dominated by deficient mineralocorticoid activity, leading to hyponatremia and hyperkalemia. The electrolyte abnormalities in secondary adrenal insufficiency generally are dominated by water retention caused by increased vasopressin secretion, leading to a generalized hemodilution. Potassium concentrations are normal.38

DIAGNOSIS
The diagnosis of adrenal insufficiency requires the presence of both clinical and chemical abnormalities compatible with the known manifestations of the disorder. ACUTE ADRENAL INSUFFICIENCY Acute adrenal insufficiency should be suspected when hypotension occurs in a patient with chronic adrenal insufficiency or in association with any of the known causes of adrenal insufficiency (see Table 76-1). When it is suspected, a blood sample should be drawn immediately for the measurement of plasma cortisol. Because plasma cortisol concentrations range between 20 and 120 g/dL during severe stress or shock in patients with normal adrenal function, a plasma cortisol concentration of <20 g/dL favors a diagnosis of adrenal insufficiency. Treatment should not wait for the result but should be initiated as soon as the blood sample for cortisol has been obtained. Initial treatment should consist of the intravenous administration of 100 mg of hydrocortisone in association with steps to maintain blood pressure. Hydrocortisone, 100 mg every 6 hours, should be administered until the crisis is past or the diagnosis is excluded. Steps designed to establish a definitive diagnosis should be instituted immediately. CHRONIC ADRENAL INSUFFICIENCY When chronic adrenal insufficiency is suspected, or when the diagnosis of acute adrenal insufficiency is being confirmed, ACTH should be given as a screening test. Synthetic ACTH (Cortrosyn), 250 g, is given as an intravenous injection over 1 minute. Blood is drawn at 45 and 60 minutes for the measurement of cortisol.39 The lower limit of the normal response range is 20 g/dL. (Note that the results of this test may be normal in patients with adrenal insufficiency that immediately follows the removal of an exogenous or endogenous ACTH source40; see Chap. 74) Two tests are useful for the differential diagnosis of chronic adrenal insufficiency: the measurement of plasma ACTH41 and computed tomography (CT) of the adrenal glands. High levels of plasma ACTH imply primary adrenal insufficiency; normal or low levels suggest secondary adrenal insufficiency.42 A finding of large adrenal glands by CT implies primary disease.43 Aldosterone deficiency also points to primary adrenal disease. This can be diagnosed best by testing the ability of the patient to retain salt.44 On a diet with 10 mEq of sodium, normal persons come into balance in 3 days. To the extent that a patient fails to achieve this balance, salt supplementation or fludrocortisone treatment is required (see Chap. 78). An elevated resting plasma renin level also is a rough guide to miner-alocorticoid deficiency. Patients who appear to have secondary adrenal insufficiency without an obvious cause, such as exogenous glucocorticoid or ACTH administration, must undergo careful examination of the pituitary gland and hypothalamus by CT scanning or magnetic resonance imaging to exclude space-occupying lesions of these organs.

TREATMENT
The treatment of adrenal insufficiency consists of replacing the missing hormones.45 The treatment of an adrenal crisis theoretically would require administration of ~200 mg of hydrocortisone daily. The standard regimen used in most centers provides 400 mg per day intravenously, approximately twice the amount calculated to be necessary. The basal production rate of cortisol has been thought to be 12 to 15 mg/m2 per day. Some studies have challenged this, however, suggesting that the production rate of cortisol may be lower by as much as half. Nevertheless, replacement of cortisol at the rate of 12 to 15 mg/m2 per day provides an adequate amount of bioactive cortisol at the cellular level. Although current practice favors a split dose of cortisol, the frequency of administration often can be reduced to once a day. In the author's experience, compliance is enhanced by a once-a-day regimen. The current standard of practice also recommends an increase in the rate of cortisol administration during stress. The definition of stress is vague, but conservative criteria include fever >38C (100F), surgical procedures or injuries, and gastroenteritis with associated vomiting and diarrhea. Some studies question the need for increased cortisol during such periods; until this is firmly demonstrated in humans, however, current guidelines should be followed.46 The generally accepted guideline is that the daily dose of glucocorticoid should be doubled during periods of minor stress such as low-grade fever, vomiting, and diarrhea. In such circumstances, if the patient is not eating, the corticosteroid must be given parenterally. During periods of major stress, such as intraabdominal surgical procedures or major trauma, the dosage of hydrocortisone should be increased to 200 mg per day. Once the stress has passed, usually by the second postoperative day, the dosage of glucocorticoid is reduced immediately and directly to the usual daily rate of 12 to 15 mg/m2 (see Chap. 78). The measurement of serum cortisol, serum ACTH, or urinary free cortisol, or of Porter-Silber chromogens is not a reliable guide to the appropriate maintenance dosage of glucocorticoid. Normal serum electrolyte levels, a good appetite, and the patient's feeling of well-being are the best guides to adequate replacement. The appearance of the signs of Cushing syndrome, such as hypertension, weight gain, facial rounding, or supracla-vicular puffiness, indicate overtreatment. Overtreatment also may result in diminished bone density.47 The production rate of aldosterone is ~100 g per day at all stages of life in salt-replete humans.48 Fludrocortisone is roughly equipotent with aldosterone but is available only as an oral preparation. In normal persons, mineralocorticoid activity is supplied by both aldosterone and cortisol in roughly equal proportions. Thus, if cortisol is used for replacement, fludrocor-tisone, 100 g per day, will supply the remaining complement of mineralocorticoid activity. If the glucocorticoid preparation used does not have significant mineralocorticoid activity (i.e., prednisone or dexamethasone, which are not recommended for replacement therapy), the dosage of fludrocortisone should be doubled. The serum potassium level is the best guide to the adequacy of mineralocorticoid replacement. CHAPTER REFERENCES
1. 2. 3. 4. 5. 6. 7. 8. O'Donnell WM. Changing pathogenesis of Addison's disease. Arch Intern Med 1950; 86:266. Loriaux DL. The polyendocrine deficiency syndromes. N Engl J Med 1985; 312:1568. Neufeld M, MacLaren N, Blizzard R. Autoimmune polyglandular syndromes. Pediatr Ann 1980; 9:154. Vibo R, Aavik E, Peterson P, et al. Autoantibodies to cytochrome P450 enzymes P450 sec, P450 c17, and P450 c21 in autoimmune polyglandular disease types I and II and in isolated Addison's disease. J Clin Endocrinol Metab 1994; 78:323. Eisenbarth G, Wilson P, Ward F, Lebovitz HE. HLA type and occurrence of disease in familial polyglandular failure. N Engl J Med 1978; 298:92. Irvine WJ, Barnes EW. Adrenocortical insufficiency. J Clin Endocrinol Metab 1972; 1:549. Crispell KR, Parson W, Hamlin J. Addison's disease associated with histoplasmosis. Am J Med 1956; 20:23. Eberle DE, Evans RB, Johnson RH. Disseminated North American blastomycosis: occurrence with clinical manifestations of adrenal insufficiency. JAMA 1977; 238:2629.

9. 10. 11. 12. 13. 14. 15. 16.

Rawson AJ, Collins LH, Grant JL. Histoplasmosis and torulosis as causes of adrenal insufficiency. Am J Med Sci 1948; 215:363. Forbus WD, Beilerbreurtje AM. Coccidioidomycosis: a study of 95 cases of the disseminated type with special reference to the pathogenesis of the disease. Mil Surg 1946; 99:653. Irvine WJ, Toft AD, Feede CM. Addison's disease. In: James VHT, ed. The adrenal gland. New York: Raven Press, 1979:131. O'Connell TX, Aston SJ. Acute adrenal hemorrhage complicating anticoagulant therapy. Surg Gynecol Obstet 1974; 139:355. Sperling MW, Wolfsen AR, Fisher DA. Congenital adrenal hypoplasia: an isolated defect of organogenesis. J Pediatr 1973; 82:444. Schaumberg H, Powers JW, Raine CS. Adrenoleukodystrophy: a clinical and pathological study of 17 cases. Arch Neurol 1975; 32:577. Griffen JW, Goren E, Schaumberg H, et al. Adrenomyeloneuropathy: a probable variant of adrenoleucodystrophy. Neurology 1977; 27:1107. Blevins LS Jr, Shankroff J, Moser HW, Ladanson PW. Elevated plasma adrenocorticotropin concentration as evidence of limited adrenocortical reserve in patients with adrenomyeloneuropathy. J Clin Endocrinol Metab 1994; 78:261.

16a. Powers JM, DeCiero DP, Ito M, et al. Adrenomyeloneuropathy: a neuropathologic review featuring its noninflammatory myelopathy. J Neuropathol Exp Neurol 2000; 59:89. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. Bleasd JF, et al. Acute adrenal insufficiency secondary to heparin-induced thrombocytopenia-thrombosis syndrome. Med J Aust 1952; 157:192. Caron P, Chabannier MH, Cambus JP, et al. Definitive adrenal insufficiency due to bilateral adrenal hemorrhage and primary anti-phospholipid syndrome. J Clin Endocrinol Metab 1998; 83:1437. Sheeler LR, Myers JM, Eversham JJ, Taylor HC. Adrenal insufficiency secondary to carcinoma metastatic to the adrenal gland. Cancer 1983; 52:1312. Tapper ML, Rotterdam HZ, Lerner CW, et al. Adrenal necrosis in the acquired immunodeficiency syndrome. Ann Intern Med 1984; 100:239. Greene LW, Cole W, Green JB. Adrenal insufficiency as a complication of the acquired immunodeficiency syndrome. Ann Intern Med 1984; 101:497. Aron DC. Endocrine complications of the acquired immunodeficiency syndrome. Arch Intern Med 1989; 149:330. Danowski TS, Bonessi JV, Sabek G. Probabilities of pituitary-adrenal responsiveness after steroid therapy. Ann Intern Med 1964; 101:11. Plager JE, Cushman P. Suppression of the pituitary ACTH response in man by administration of ACTH or cortisol. J Clin Endocrinol Metab 1962; 22:147. Jasani MK, Boyle TA, Dick WC, et al. Corticosteroid-induced hypothalamo-pituitary-adrenal axis suppression: prospective study using two regimens of corticosteroid therapy. Ann Rheum Dis 1968; 27:352. Graber AL, Ney RL, Nicholson WE. Natural history of pituitary-adrenal recovery following long term suppression with corticosteroids. J Clin Endocrinol Metab 1965; 25:11. Kehlet H, Binder C. Value of an ACTH test in assessing hypothalamic-pituitary-adrenocortical function in glucocorticoid treated patients. BMJ 1973; 2:147. Banna M. Craniopharyngioma: a review article based on 160 cases. Br J Radiol 1976; 49:206. Hankinson T, Banna M. Pituitary and parapituitary tumors. London: WB Saunders, 1976:51. Vita JA, Silverberg SJ, Goland RS, et al. Clinical clues to the cause of Addison's disease. Am J Med 1985; 78:461. Stuart CA, Neelon FA, Lebovitz HE. Hypothalamic insufficiency: the cause of hypopituitarism in sarcoidosis. Ann Intern Med 1978; 88:589. Tandon PN, Pathak SN. Tuberculosis of the central nervous system. In: Tropical neurology. London: Oxford University Press, 1973:37. Ortega FJV, Longridge NS. Fracture of the sella turcica. Injury 1975; 6:335. Thorn GW. Diagnosis and treatment of adrenal insufficiency. Springfield, IL: Charles C Thomas, 1951. Hertz KC, Gazze LA, Kirkpatrick CH, Katz SI. Autoimmune vitiligo: detection of antibodies to melanin producing cells. N Engl J Med 1977; 297:634. Lipsett MB, Pearson OH. Pathophysiology and treatment of adrenal crises. N Engl J Med 1956; 254:511. Jorgensen H. Hypercalcemia in adrenocortical insufficiency. Acta Med Scand 1973; 193:175. Pearson OH, Whitmore WF, West CD. Clinical and metabolic studies of bilateral adrenalectomy for advanced cancer in man. Surgery 1953; 34:543. Lindholm J, Kehlet H, Blichert-Toft M. Reliability of the 30 minute ACTH test in assessing hypothalamic-pituitary-adrenal function. J Clin Endocrinol Metab 1978; 47:272. Kehlet H, Lindholm J, Bjerre P. Value of the 30 min ACTH test in assessing hypothalamic-pituitary-adrenocortical function after pituitary surgery in Cushing's disease. Clin Endocrinol (Oxf) 1984; 20:349. Oelkers W, Diederich S, Bahr V. Diagnosis and therapy surveillance in Addison's disease: rapid adrenocorticotropin (ACTH) test and measurement of plasma ACTH, renin activity and aldosterone. J Clin Endocrinol Metab 1992; 75:259. Schulte HM, Chrousos GP, Avgerinos P. The corticotropin-releasing hormone stimulation test: a possible aid in the evaluation of patients with adrenal insufficiency. J Clin Endocrinol Metab 1984; 58:1064. Schultz CL, Haaga JR, Fletcher BD. Magnetic resonance imaging of the adrenal glands: a comparison with computed tomography. Am J Radiol 1984; 143:1235. Gill JR, Bell NH, Barrter FL. Impaired conservation of sodium and potassium in renal tubular acidosis. Clin Sci 1967; 33:577. Thorn GW, Lauler DP. Clinical therapeutics of adrenal disorders. Am J Med 1977; 53:673. Symreng T, Karlberg BE, Kagedal B, Schlidt B. Physiological cortisol substitution of long term steroid-treated patients undergoing major surgery. Br J Anaesth 1981; 53:949. Zelissen PM, Croughs RJ, van Rijk PP, Raymakers JA. Effect of glucocorticoid replacement therapy on bone mineral density in patients with Addison's disease. Ann Intern Med 1994; 120:207. New MI, Seaman MP, Peterson RE. A method for the simultaneous determination of the secretion rates of cortisol, 11-deoxycortisol, corticosterone, 11-deoxycorticosterone, and aldosterone. J Clin Endocrinol Metab 1969; 29:514.

CHAPTER 77 CONGENITAL ADRENAL HYPERPLASIA Principles and Practice of Endocrinology and Metabolism

CHAPTER 77 CONGENITAL ADRENAL HYPERPLASIA

PHYLLIS W. SPEISER Embryology Clinical Features 21-Hydroxylase (P450C21) Deficiency 11b-Hydroxylase (P450C11) Deficiency 17a-Hydroxylase/17,20-Lyase (P450C17) Deficiency 3b-Hydroxysteroid Dehydrogenase Deficiency 20,22-Desmolase (P450SCC) Deficiency Diagnosis Postnatal Diagnosis Prenatal Diagnosis Treatment Genetics 21-Hydroxylase Deficiency 11b-Hydroxylase Deficiency 17a-Hydroxylase/17,20-Lyase Deficiency 3b-Hydroxysteroid Dehydrogenase Deficiency Cholesterol Desmolase Deficiency Other Genetic Disorders Related to Congenital Adrenal Hyperplasia Conclusion Chapter References

Congenital adrenal hyperplasia (CAH) is a group of inherited diseases caused by defective activity in one of five enzymes that contribute to the synthesis of cortisol from cholesterol in the adrenal cortex (Fig. 77-1). Details of normal adrenal steroidogenesis are discussed in detail in Chapter 72. The term adrenal hyperplasia derives from the tendency to glandular enlargement under the influence of adrenocorticotropic hormone (ACTH) in an effort to compensate for inadequate cortisol synthesis. The alternate term adrenogenital syndrome refers to the common associated finding of ambiguous external genitalia due to incidental deficiency or excess production of adrenal androgens. Each particular enzyme deficiency produces characteristic alterations in the ratio of precursor hormone to product hormones. These hormonal imbalances are accompanied by clinically evident abnormalities, including abnormal development of the genitalia and pseudohermaphroditism, disturbances in sodium and potassium homeostasis, blood pressure dysregulation, and abnormal somatic growth1,2,2a (Table 77-1). The molecular genetic basis for all but one of the enzymatic deficiencies is known; disease-causing mutations have been identified for the genes encoding the respective steroidogenic enzymes.

FIGURE 77-1. The pathways of corticosteroid synthesis are diagrammed. Cholesterol is converted in several steps to aldosterone, cortisol, or sex steroids. Hormones marked by an asterisk (*) are produced largely outside the adrenal cortex. Deficiency of a given enzyme causes accumulation of hormonal precursors and a deficiency of products. (DHEA, dehydroepiandrosterone; HSD,hydroxysteroid dehydrogenase.)

TABLE 77-1. Clinical, Biochemical, and Genetic Characteristics of Congenital Adrenal Hyperplasia

The most prevalent form of CAH is caused by deficiency of the cytochrome P450 enzyme, 21-hydroxylase (>90% of cases), followed by a deficiency of 11b-hydroxylase, 17a-hydroxylase/17,20-lyase, and 3b-hydroxysteroid dehydrogenase. No mutations have been identified in the gene encoding cholesterol desmolase (also termed side-chain cleavage); this apparent enzyme deficiency is instead explained by defective steroidogenic acute regulatory protein. Deficiency of the enzymes that do not impair cortisol synthesis do not come under the rubric of CAH and are therefore not discussed here. In addition to a classic form with onset in prenatal life, nonclassic forms of the enzyme deficiencies with onset in childhood or young adult life are also described.

EMBRYOLOGY
The fetal gonads remain undifferentiated until about the seventh week of gestation (see Chap. 90). At that time, in normal 46,XY fetuses, the gene encoding the sex-determining factor of the Y chromosome (i.e., SRY) is transcribed and testicular differentiation begins. SRY initiates a complex cascade of events3 in which other genes are inhibited (e.g., DAX-1, important in ovarian determination) or activated (e.g., SOX9 and SF-1, important in Sertoli cell differentiation and indirectly involved in ovarian inhibition via antimllerian hormone, AMH). Testosterone secretion begins by ~8 weeks of fetal life. In normal 46,XX fetuses, the absence of high local concentrations of AMH allows differentiation of the ovaries, beginning at ~10 weeks' gestation (see Chap. 90). The internal genital duct structures are recognizable by 7 weeks. Mllerian ducts develop into the rostral third of the vagina, the uterus, and fallopian tubes. Wolffian ducts develop into epididymis, vas deferens, seminal vesicles, and ejaculatory ducts under the influence of adequate local quantities of androgen. Embryonal development of the gonads and internal ducts are generally unaffected by vagaries in sex steroid hormone synthesis. In females, the development of the external genitals is passive. The urogenital sinus differentiates into separate urethral and vaginal orifices in normal females, and the labioscrotal folds remain separated by the labia minora, which hood the clitoris. In males, under the influence of high circulating levels of androgen, the urethral and genital orifices fuse to form an elongated penile urethra, and the labioscrotal folds fuse to form the scrotum. The latter changes may also occur in 46,XX fetuses exposed to high androgen levels from fetal adrenal or maternal sources. Conversely, male external genitals may be hypoplastic if a defect exists in testosterone synthesis or action.

At ~7 weeks of gestation, the adrenal cortex differentiates from mesodermal tissues. Soon after, two adrenal zones form: a small peripheral adult cortex and a large inner fetal cortex. Steroid production by the fetal cortex begins in the latter half of the first trimester. Adrenal mass increases markedly during this time, and peaks at 15 weeks.4 General regulation of adrenal growth early in gestation is incompletely understood and is thought to be only partly attributable to ACTH. Other likely trophic hormones include transforming growth factor-b (TGF-b), basic fibroblast growth factor (bFGF), and insulin-like growth factor-II (IGF-II).5 Beyond 20 weeks of gestation, adrenal growth and steroidogenesis are almost exclusively responsive to ACTH. In utero administration of glucocorticoids suppresses fetal ACTH and, under most circumstances, inhibits adventitious adrenal steroid production in fetuses affected with a severe virilizing form of CAH (e.g., 21- or 11b-hydroxylase deficiency).

CLINICAL FEATURES
21-HYDROXYLASE (P450C21) DEFICIENCY Patients with this commonly diagnosed enzyme deficiency cannot adequately synthesize cortisol. Insufficient cortisol synthesis results in overproduction of adrenal androgens, which are synthesized independently of 21-hydroxylase. Androgens induce somatic growth with inappropriately rapid advancement of linear growth, early epiphyseal fusion of the long bones, and short stature. Other features include precocious development of sexual hair, apocrine body odor, and penile or clitoral enlargement. Reduced fertility may be observed in both sexes. Clinical features are outlined in Table 77-1. Females affected with the severe classic form of 21-hydroxylase deficiency are exposed to excess androgens pre-natally and are born with masculinized external genitalia (Fig. 77-2). If the disease goes undiagnosed and the infant is untreated, further virilization ensues (Fig. 77-3). Approximately 75% of patients with the classic form cannot synthesize aldosterone efficiently because of impaired 21-hydroxylation of progesterone; these salt-wasting individuals fail to conserve sodium normally and usually come to medical attention in the neonatal period with hyponatremia, hyperkalemia, and hypovolemic shock. These adrenal crises may prove fatal if proper medical care is not delivered. Patients with sufficient aldosterone production and no salt wasting who have signs of prenatal virilization and markedly increased production of hormonal precursors of 21-hydroxylase (e.g., 17-hydroxyprogesterone) are referred to as simple virilizers. Earlier confusion regarding the origins of these two classic phenotypes has been resolved to a large extent by the understanding of the molecular genetics of the disease, and allelic variation in the gene encoding active 21-hydroxylase (CYP21) appears to be responsible for most phenotypic variation, as is discussed later.

FIGURE 77-2. External genitalia of a 2-month-old female infant with 21-hydroxylase deficiency.

FIGURE 77-3. Variations in the differentiation of the external genitalia in congenital adrenal hyperplasia. In genetic females, adrenal androgen hypersecretion (enzymes 2, 4, and 5) is associated with various degrees of masculinization, leading to apparent male external genitalia. (From Grumbach MM, Ducharme J. The effects of androgens on fetal development: androgen-induced female pseudohermaphroditism. Fertil Steril 1960; 11:157.)

Patients affected with the milder, nonclassic form of 21-hydroxylase deficiency may have signs of postnatal androgen excess.6 Except for rare cases showing mild clitoromegaly, females with the nonclassic disorder are born with normal external genitalia. The syndrome of polycystic ovarian disease has often been confused with nonclassic CAH 21-hydroxylase deficiency in young women with hirsutism, oligomenorrhea, and diminished fertility. Precise clinical distinction between the classic simple virilizing disease and nonclassic disorder is sometimes difficult among males, because the hormonal reference standards for diagnosis represent a continuum. Moreover, because males do not manifest ambiguous genitalia as a sign of in utero androgen excess, the only other distinguishing clinical parameters are bone age and somatic growth pattern, which are nonpathognomonic. Phenotypic severity in nonclassic 21-hydroxylase deficiency varies greatly, and in some individuals the disease has been detected solely on the basis of hormonal or genetic testing in the course of family studies. Aldosterone synthesis is normal in patients with nonclassic 21-hydroxylase deficiency. Table 77-2 describes features distinguishing salt-wasting, simple virilizing, and nonclassic forms of 21-hydroxylase deficiency. Table 77-3 describes specific mutations.

TABLE 77-2. Phenotype in 21-Hydroxylase Deficiency

TABLE 77-3. Most Common Disease-Causing Mutations in CYP21

Neonatal screening for 21-hydroxylase deficiency measuring 17-hydroxyprogesterone levels in heel-stick blood has been effective in reducing neonatal morbidity and mortality.7 This has been particularly useful in males with salt-wasting disease in whom no obvious phenotypic clue to the diagnosis, such as ambiguous genitalia, is present. The radioimmunoassay of heel-stick blood on filter paper was first used on a wide scale among the Alaskan Yupik Eskimos, one of two geographically isolated and genetically homogeneous groups at high risk for 21-hydroxylase deficiency CAH.8 Subsequently, many other newborn screening programs were developed for 21-hydroxylase deficiency CAH.9,9a The worldwide incidence of 21-hydroxylase deficiency CAH based on newborn screening is 1 in 14,554 live births; approximately 75% of infants in whom the disease is detected in these programs manifest the salt-wasting phenotype.9 According to the Hardy-Weinberg law for populations at equilibrium, the heterozygote frequency for all classic 21-hydroxylase gene defects is 1 in 61 persons. A high frequency of nonclassic 21-hydroxylase deficiency has also been discerned.10 This disorder occurs most frequently among Ashkenazi Jews (1 in 27), but it is also common among other ethnic groups, such as Hispanics, residents of the former Yugoslavia, and Italians. Overall, in a mixed white population, the disease is estimated to occur in ~1 in 100 individuals. These disease frequencies were derived indirectly based on response to ACTH stimulation combined with HLA typing. Confirmation was obtained using the statistical method of commingling distributions.11 Nonclassic 21-hydroxylase deficiency is among the most frequent autosomal recessive disorders in humans. Clinical investigation to diagnose nonclassic 21-hydroxylase deficiency is warranted in any patient showing the signs of androgen excess described previously; particularly high-risk groups include Ashkenazi Jews, children with precocious pubarche, and girls or women with hirsutism and oligomenorrhea. 11b-HYDROXYLASE (P450C11) DEFICIENCY As in the case of 21-hydroxylase deficiency, in patients with 11b-hydroxylase deficiency accumulating precursor steroids are channeled into androgen pathways beginning in prenatal life, which causes genital ambiguity in affected newborn females. Male infants show no abnormality of external genitalia. Later signs of androgen excess are observed in both sexes affected with 11b-hydroxylase deficiency if the disease is not promptly recognized and treated.12 Patients with 11b-hydroxylase deficiency account for approximately 5% of all CAH cases. Although in the general population this enzyme defect is found in ~1 in 100,000 live births, the disease frequency is ~1 in 5000 to 7000 among Jews of Moroccan descent.13 No systematic screening programs have been initiated to detect forms of CAH other than 21-hydroxylase deficiency. Nonclassic variants of 11b-hydroxylase deficiency have also been described.12 Hormonal imbalances differentiate 21-hydroxylase from 11b-hydroxylase deficiency. In most cases, classic 21-hydroxylase deficiency is accompanied by deficient aldosterone synthesis and limited ability to conserve sodium. In contrast, in patients with 11b-hydroxylase deficiency, excessive production of the mineralocorticoid agonist deoxycorticosterone (DOC) or its metabolites results in sodium retention, hypokalemia, volume expansion, suppressed plasma renin activity (PRA), and hypertension. Hypertension, however, is not the sine qua non for diagnosis of 11b-hydroxylase deficiency and is often absent in young children. Nonclassic cases may show variable elevations of DOC and 11b-deoxycortisol (compound S) and have normal PRA levels. 17a-HYDROXYLASE/17,20-LYASE (P450C17) DEFICIENCY In 17a-hydroxylase/17,20-lyase deficiency, impaired production of glucocorticoids and sex steroids (C19/C18 compounds) causes failure to develop estrogenic sexual characteristics at puberty in genetic females and incomplete development of the external genitals in genetic males.14,15 Shunting of P450c17 precursor steroids into the 17-deoxy pathway produces mineralocorticoid excess, with hypokalemic alkalosis and hypertension similar to that in the 11b-hydroxylase deficiency. In rare cases, selective 17,20-lyase deficiency is detected. In such patients, cortisol and DOC levels are normal, but adrenal and gonadal C21 to C19 steroid conversion is impaired, so that normal sex steroid production is prevented. More than 120 cases have been reported of severe or complete 17a-hydroxylase deficiency, mostly in combination with 17,20-lyase deficiency.16 Patients from Canadian-Dutch Mennonite kindreds who share the same genotype have been reported.17 Relatively few genetic females have been detected. Partial deficiency of this enzymatic activity may be found in males with ambiguous genitalia.18 3b-HYDROXYSTEROID DEHYDROGENASE DEFICIENCY The enzyme 3b-hydroxysteroid dehydrogenase (3b-HSD) is responsible for conversion of D5 to D4 steroids. Deficiency of this enzyme results in inefficient cortisol synthesis; oversecretion of dehydroepiandrosterone (DHEA), which is only weakly androgenic; and oversecretion of pregnenolone, which is ineffective as a mineralocorticoid. Affected individuals typically have cortisol insufficiency and salt wasting. Genital ambiguity is also part of the syndrome. Although lack of potent androgens produces hypospadias in males, high levels of DHEA may cause clitoromegaly without urogenital sinus formation in females.19 In 3b-HSD and 17a-hydroxylase/17,20-lyase deficiencies, potent androgens are deficient in prenatal life, which predisposes males to gynecomastia. The precise frequency of severe defects in the adrenal 3b-HSD gene is unknown. A nonclassic form of 3b-HSD deficiency is diagnosed with variable frequency in children with precocious pubarche and females with hirsutism and oligomenorrhea,20,21 but the hormonal profile with ACTH stimulation is less robust a diagnostic tool than that described for 21-hydroxylase deficiency. Some investigators have suggested that ovarian hyperandrogenism may be confused with 3b-HSD deficiency.22 The physician must also exercise caution in the interpretation of ACTH stimulation tests performed in infants younger than 1 year of age, because 3b-HSD is normally deficient in fetal life and relatively inactive in early infancy. Molecular genetic investigation has not revealed mutations to explain cases of mild 3b-HSD deficiency. No marked variations in the ethnic incidence of this defect are known; several classic cases have been identified in consanguineous families. 20,22-DESMOLASE (P450SCC) DEFICIENCY Deficiency of 20,22-desmolase is also called lipoid adrenal hyper-plasia or apparent cholesterol desmolase deficiency. This enzyme catalyzes the initial reaction for all steroid production from cholesterol substrate. Apparent deficiency of side-chain cleavage or cholesterol desmolase is extremely rare, with only ~30 cases reported in the world's literature.23,24 Complete cholesterol desmolase deficiency would be expected to produce global adrenocortical insufficiency and death because of marked cortisol deficiency and severe salt wasting. Partial apparent defects in this enzyme result in pseudohermaphroditism in genetic males; lack of secondary sexual characteristics can be expected in genetic females. A single case report describes long-term follow-up of a patient diagnosed in the newborn period and successfully treated for 18 years. 25 Apparent cholesterol desmolase deficiency seems to occur with less severity and somewhat more frequently among the Japanese. Lipoid adrenal hyperplasia is now known to result from genetic defects in the gene encoding steroidogenic acute regulatory protein, rather than from actual defects in cholesterol desmolase.25a

DIAGNOSIS
POSTNATAL DIAGNOSIS The diagnosis of 21-hydroxylase deficiency may be confirmed by administering an intravenous bolus of ACTH and measuring the resultant elevation in blood levels of 17-hydroxyprogesterone.26 Usually, a panel of adrenal hormones is assayed before and after ACTH administration, but the most specific available marker for which testing is commercially available is 17-hydroxyprogesterone. Clinicians should be aware that cortisol stimulation is suboptimal after ACTH infusion in patients with

severe defects in adrenal steroid synthesis. If for any reason blood testing cannot be used or radioimmunoassays for 17-hydroxyprogesterone are unavailable, the examiner can measure 17-ketosteroids or pregnanetriol in a 24-hour urine collection. The latter steroid is the principal direct urinary metabolite of 17-hydroxyprogesterone. Ancillary tests used in the initial evaluation of infants with ambiguous genitalia include karyotype analysis, pelvic and abdominal ultrasonography, and sonogram of the urogenital orifices using radiopaque dyes. Patients with nonclassic 21-hydroxylase deficiency have 17-hydroxyprogesterone levels that exceed those seen in heterozygous carriers of an affected gene, but these levels are lower than those of patients with the classic form of the disorder.26 In the nonstimulated state, these patients may have near-normal serum hormone levels. A serum 17-hydroxyprogesterone level below 200 ng/dL effectively excludes this diagnosis if the sample is obtained in the early morning (i.e., by 8:00 a.m.). The diagnosis of 11b-hydroxylase deficiency is made by the measurement of elevated basal or ACTH-stimulated DOC and/or 11-deoxycortisol (i.e., compound S) in the serum or elevated levels of the tetrahydro-compounds (i.e., DOC and/or S) in a 24-hour urine collection.27 Another marker useful in pediatric diagnosis is 6a-hydroxytetrahydro-11-deoxycortisol, which can be measured by gas chromatography and mass spectrometry of urine.28 As in 21-hydroxylase deficiency, urinary 17-ketosteroids are usually elevated, reflecting increased shunting of 11b-hydroxylase hormonal precursors into the sex steroid pathway. PRA is usually low in older children and is accompanied by low levels of aldosterone. The diagnosis of 17a-hydroxylase/17,20-lyase deficiency is made by a finding of marked elevations of serum DOC and corticosterone (i.e., compound B) and the metabolites of these two steroids.14 Aldosterone is often low secondary to suppression of renin by excess DOC, as in the case of 11b-hydroxylase deficiency. The 17a-hydroxylase-deficient patients do not experience adrenal crisis despite inadequate cortisol synthesis. Overproduction of corticosterone provides adequate physiologic response to stress. Plasma ACTH levels are less elevated than in other conditions of impaired cortisol production. Gonadotropin production is extremely elevated in both sexes because of the absence of any sex steroid feedback; the gonads are atrophic. A high ratio of D5 to D4 steroids characterizes the 3b-HSD deficiency.29 Serum levels of 17-hydroxypregnenolone and DHEA are elevated before and after ACTH stimulation. Increased excretion of the D5 metabolites pregnanetriol and 16-pregnanetriol in the urine is also diagnostic for this enzyme disorder. PRENATAL DIAGNOSIS Prenatal testing for 21-hydroxylase deficiency has been used for two decades in pregnancies known to be at risk.30,31 Hormonal diagnosis is accomplished by finding elevated levels of 17-hydroxyprogesterone or 21-deoxycortisol in amniotic fluid.32,33 Genetic diagnosis was first performed by identifying HLA markers on fetal cells cultured from the amniotic fluid; the genes encoding HLA antigens are closely linked to CYP21.34,35 Problems encountered with these diagnostic techniques included false-negative 17-hydroxyprogesterone levels in nonsalt-losing cases and intra-HLA recombination.36 Early amniocentesis and chorionic villus sampling have permitted diagnostic studies to be performed at the end of the first trimester.37,38 DNA obtained from such procedures may be analyzed by molecular genetic techniques, such as allele-specific hybridization with oligonucleotide probes for the normal and mutant alleles of CYP21.39,40 In pregnancies known to have a 25% risk for 21-hydroxylase deficiency, blind prenatal treatment of the fetus may be initiated by administering dexamethasone to the mother beginning in the first trimester. 41,42 and 43 Deferral of therapy until a molecular genetic diagnosis is known could hamper the ability to prevent genital ambiguity.44 Although prenatal treatment usually ameliorates virilization of affected females, results of prenatal treatment have not been completely successful in this regard.45,46 Failure to produce normal female genitalia in 20% to 25%47 of affected girls through the use of prenatal dexamethasone therapy has been attributed to cessation of therapy in midgestation, to noncompliance, or to suboptimal dosing. Some treatment failures had no ready explanation.48 No fetus treated with low-dose dexamethasone has been born with a congenital malformation specifically attributable to dexamethasone therapy.49 The incidence of fetal deaths in treated pregnancies does not exceed that for the general population. Complications observed in a rodent model of in utero exposure to high-dose glucocorticoids included cleft palate, placental degeneration, intrauterine growth retardation, and unexplained fetal death.50 Other concerns for human fetal exposure have been raised based on observations of relatively late sequelae involving neurologic and vascular changes in lower mammals.51 Although at present no data exist to suggest that such sequelae occur in humans, the oldest prenatally treated children from CAH studies are only now reaching adolescence, and more long-term follow-up is required. The incidence of maternal complications has varied among investigations. Serious side effects, such as overt Cushing syndrome, massive weight gain, and hypertension, have been reported in ~1% of all treated pregnant women. Caution must be exercised in recommending prenatal therapy with dexamethasone, and women must be fully informed of these potential risks and nonuniformity of beneficial outcome to the affected female fetus. Despite these caveats, many parents of affected girls still choose prenatal medical treatment because of the severe psychological impact of ambiguous genitalia. A similar diagnostic and therapeutic approach is effective in cases of 11b-hydroxylase deficiency, in which affected female fetuses are also at risk for prenatal virilization.

TREATMENT
Patients with simple virilizing or salt-wasting classic 21-hydroxylase deficiency or those with 11b-hydroxylase, 17a-hydroxylase/17,20-lyase, and 3b-HSD deficiencies, as well as select symptomatic patients with nonclassic forms of these diseases, are treated with daily oral hydrocortisone or similar drugs. Treatment with glucocorticoids suppresses excessive secretion of ACTH, correcting the adrenal hormone imbalance. Patients with the salt-wasting form of CAH require additional supplementation with mineralocorticoids (e.g., fludrocortisone acetate [Florinef], 50200 g per day) and sodium chloride supplements (1 to 2 g/10 kg body weight). Older children and adults with simple virilizing disease who are treated adequately with glucocorticoids usually do not have a clinically apparent deficiency of aldosterone, nor is renin markedly elevated. Many pediatric endocrinologists empirically treat all CAH patients with fludrocortisone and sodium chloride despite the lack of signs of salt wasting. Prudence dictates following PRA in all patients as an index of the need for mineralocorticoid and salt supplements. Caution is advised to avoid development of hypertension consequent to excessive or unnecessary treatment with the latter regimen. Glucocorticoid treatment also leads to reduction of mineralocorticoid hormones in 11b-hydroxylase and 17a-hydroxylase deficiency, with amelioration of hypertension. In cases of longstanding hypertension, adjunctive antihypertensive drugs may be required to completely normalize blood pressure. The usual mode of treatment for CAH in childhood is with two to three divided daily doses of hydrocortisone totaling 10 to 20 mg/m2 per day (average dosage ~15 mg/m2 per day). Even this relatively low dosage may be supraphysiologic, because healthy children and adolescents secrete an average of ~7 g/m2 of cortisol daily.52,53 Experience indicates that once-daily doses of hydrocortisone, because of its relatively rapid metabolism, is therapeutically suboptimal over the long term. Treatment efficacy should be monitored with frequent measurements of serum 17-hydroxyprogesterone (good control is indicated by a level of <1000 ng/dL), androstenedione, and testosterone, in addition to PRA in young children,54 and assessment of growth, skeletal maturation, and pubertal status. If the response to maintenance hydrocortisone at the standard dosage is poor despite good medical compliance, a 2- to 4-day trial of dexamethasone (in a dosage of ~2030 g/kg per day to a maximum of 2 mg per day) may be more effective in suppressing the adrenal. Maintenance hydrocortisone may then be resumed, and adrenal hormone levels are monitored. If epiphyses are fused and linear growth is not at risk, the maintenance regimen may be changed to one of the longer-acting glucocorticoids, prednisone or dexamethasone. The greater potency of these drugs means that slight dosing errors may result in iatrogenic Cushing syndrome and growth retardation in children. Life-threatening stress, severe illness, or surgery demand parenteral therapy with high doses of hydrocortisone in any patient who is undergoing chronic treatment with exogenous glucocorticoids. Clinical trials have been initiated in classic CAH using androgen-receptor blockers and inhibitors of steroid synthesis in conjunction with glucocorticoids to ameliorate virilization and to prevent iatrogenic glucocorticoid-induced growth retardation.55 Preliminary results appear promising, but more data are required.56 Compliance with multidrug regimens may be difficult for many patients. Another potentially useful adjunct to standard therapy is the gonadotropin-releasing hormone analogs, which serve to delay the onset of gonadarche. Adrenalectomy has been performed in a limited number of severely affected patients refractory to adrenocortical suppression with standard medical therapies.57 This

approach remains experimental, and such patients must obviously be treated with lifelong physiologic replacement dosages of glucocorticoid and mineralocorticoid. The suggestion has been made that androgen-receptor blockade may be preferable to glucocorticoids as primary treatment of mild 21-hydroxylase deficiency.58 Therapeutic trials of the latter treatment have been prompted by the observation that, although menses usually resume regularity within 2 to 6 months after initiation of glucocorticoid therapy in young women with nonclassic 21-hydroxylase deficiency, hirsutism is quite refractory to this mode of treatment. Surgical therapy is recommended in cases of ambiguous genitalia. Most often, this involves clitoroplasty and vagino-plasty in virilized females. Improved surgical techniques now permit these procedures to be performed in a single-stage operation by experienced urologists.59 Given the recognition of disturbed gender identity and role among young women with CAH, providing early and continuing psychological counseling for them is extremely important.60,61 With improved medical, surgical, and psychological treatments, an improved psychosexual outcome can be achieved.

GENETICS
The autosomal recessive mode of inheritance of adrenal steroidogenic defects was recognized in the early 1950s.62,63 and 64 The linkage of 21-hydroxylase deficiency to the HLA complex on chromosome arm 6p was discovered in the late 1970s,65 and by the mid-1980s, further molecular genetic details had been unravelled.66 Other human adrenal steroidogenic defects also have been successfully subjected to molecular genetic analysis. 21-HYDROXYLASE DEFICIENCY The CYP21 (CYP21B) structural gene encoding steroid 21-hydroxylase and a pseudogene (CYP21P or CYP21A) are located 30 kilobases (kb) apart in the HLA complex on chromosome band 6p21.3.67,68 Although the two genes are 98% identical in nucleotide sequence, CYP21P has accumulated a number of mutations that render any gene product completely inactive. These include an 8-base-pair (bp) deletion in exon 3, a frame shift in exon 7, and a nonsense mutation in exon 8. Most mutations causing 21-hydroxylase deficiency are caused by apparent recombinations between CYP21 and CYP21P. Approximately 20% of these are unequal meiotic crossovers resulting in a deletion of a 30-kb DNA segment that includes the 3' end of the pseudogene and the greater portion of the active gene.69 This deleted haplotype is incapable of producing any active enzyme. About 80% of mutations result from apparent gene conversions that transfer small segments containing one or more deleterious mutations from CYP21P to CYP21. The most common of these, accounting for an additional ~25% of mutant haplotypes, is a single AG transition at nucleotide 656 that causes abnormal premessenger RNA (mRNA) splicing. Less commonly found are missense mutations, which cause changes in the protein's amino-acid sequence. Several large studies have examined the prevalence of individual mutations in an attempt to correlate specific mutations with particular clinical manifestations of the disease.70,71,72,73 and 74 These correlations are most reliably made in individuals who are homozygous or hemizygous (the other chromosome carries a deletion) for each mutation. Table 77-3 illustrates the mutations commonly found in classic and nonclassic forms of 21-hydroxylase deficiency, grouped into three categories according to the predicted level of enzymatic activity based on in vitro mutagenesis and expression.75,76,77 and 78 Group A, with total ablation of enzyme activity, is most often associated with salt-wasting disease; group B, with 2% normal activity, consists predominantly of patients with simple virilizing disease; and group C, with 20% to 60% of normal activity, is most often associated with the nonclassic disorder. These studies suggest that mutant CYP21 enzymes carrying discrete amino-acid substitutions identified in CAH patients exhibit in vivo activities that are generally consistent with in vitro predictions, and correlate with disease severity. Exceptions to this rule include patients with moderate or severe mutations and relatively mild disease. Because aldosterone is normally secreted at a rate 100 to 1000 times lower than that of cortisol, residual enzyme activity as low as 0.6% of normal activity, as seen in the Ile-172Asn nonconservative substitution, allows enough aldosterone synthesis to prevent symptoms of salt wasting, resulting in the simple virilizing phenotype. Factors outside the CYP21 locus may also influence development of the salt-wasting phenotype. Another point of interest in phenotype-genotype analysis is the wide range of clinical manifestations in patients carrying the group C nonclassic mutations (e.g., Val-281Leu and Pro-30Leu). These mutations are expected to reduce enzyme activity to 20% to 60% of normal, with 17-hydroxyprogesterone being the preferred substrate. An individual heterozygous for a deletion of CYP21 (i.e., ablation of all enzyme activity derived from one chromosome) is also expected to have ~50% of normal 21-hydroxylase activity, but such individuals have no signs of disease and have hormonal abnormalities detectable only with ACTH stimulation. This suggests that in vivo 21-hydroxylase activity in patients with nonclassic 21-hydroxylase deficiency is often <50% of normal. One plausible explanation for such differences in clinical manifestations of disease is fluctuating intraad-renal concentrations of progesterone, which at physiologic levels (24 mol/L)79 acts as a competitive inhibitor of the nonclassic mutant enzyme for its main substrate, 17-hydroxyprogesterone. Individuals carrying two nonclassic alleles may have closer to 20% net 21-hydroxylase activity as intraadrenal progesterone concentration increases. Other factors contributing to phenotypic variability might include pseudosubstrate inhibition of other steroidogenic enzymes by accumulated precursors of 21-hydroxylase. Study of naturally occurring mutations has also shed light on structure-function relationships for the 21-hydroxylase enzyme; for example, G291s is important for catalytic activity, and E196 deletion and R483P reduce enzyme half-life.80 11b-HYDROXYLASE DEFICIENCY Two human genes on chromosome band 8q21-q22 encode 11-hydroxylase (CYP11B) isozymes with predicted amino-acid sequences that are 93% identical.81,82 Each has 9 exons spaced over ~7 kb. CYP11B1, expressed at high levels in normal adrenal glands, is regulated by ACTH.81,83 CYP11B2, not readily detectable in Northern blots using normal adrenal RNA, is regulated primarily by angiotensin II, rather than by ACTH.81 Transcripts of CYP11B2 have been detected by hybridization to RNA from an aldosterone-secreting tumor or in normal adrenal mRNA by the more sensitive technique of reverse transcription coupled with the polymerase chain reaction.84,85 Defects in the CYP11B1 gene result in virilizing, hypertensive CAH, and defects in CYP11B2 cause a rare salt-wasting disease known as corticosterone methyl oxidase II (CMO II) deficiency. A third disease, glucocorticoid-suppressible aldosteronism, ensues when the regulatory region of CYP11B1 is transposed to a position at which it controls synthesis of CYP11B2, promoting glucocorticoid-suppressible and ACTH-stimulable aldosterone synthesis.86 Mutations in the CYP11B1 gene tend to cluster in exons 6, 7, and 8.87 The most common genetic alteration in Moroccan Jews with 11b-hydroxylase deficiency is Arg-448His.88 When introduced into CYP11B1 complementary DNA and expressed in cultured cells, this mutation abolishes normal enzymatic activity and is therefore consistent with the classic, severe virilizing phenotype observed in these patients. Blood pressure was not uniformly elevated in all patients carrying this mutation, and as observed in 21-hydroxylase deficiency, apparently other factors exist that modify phenotype. Numerous other genetic defects in 11b-hydroxylase deficient patients from other ethnic backgrounds have been described, including insertions and nonsense, missense, and splice mutations.89,90,91,92,93 and 94 Mild missense mutations have been identified in some patients with nonclassic 11b-hydroxylase deficiency.95 17a-HYDROXYLASE/17,20-LYASE DEFICIENCY The P450c17 structural gene (CYP17) spans 12.6 kb on chromosome 10,96 with an intronexon organization similar to that of CYP21. The same gene is expressed in both the adrenal and the testis.97 Molecular characterization of specific mutations in CYP17have been reported in a number of patients.16 In patients with classic, severe 17a-hydroxylase and 17,20-lyase deficiencies, these have included a point mutation creating a termination codon in the first exon, a 7-bp duplication in exon 2 that produces a frame shift, and a four-base duplication in exon 8. Homozygous deletion of 3 bp in exon 1 was detected in a patient with apparent selective compromise of 17,20-lyase activity, a phenotypic female with sexual infantilism. A genetic male with ambiguous genitalia was found to be a compound heterozygote with a termination codon introduced by a single base substitution in exon 4 on one chromosome and a nonconservative prolinethreonine substitution in exon 6 on the second chromosome. Another patient with male pseudohermaphroditism was a compound heterozygote for two null mutations, including a splice mutation deleting exon 2.98 3b-HYDROXYSTEROID DEHYDROGENASE DEFICIENCY Two homologous genes encoding 3b-HSD type I, expressed in placenta and skin, and type II, expressed in adrenal and gonads, have been identified on chromosome band 1p13.99,100,101 and 102 This is the only enzyme discussed here that is not encoded by a gene in the cytochrome P450 superfamily. Type II gene mutations have been described in patients with classic 3b-HSD deficiency.102a These include two separate point mutations introducing termination codons in exon 4, insertion of a single base causing a frame shift, and two separate amino-acid substitutions in highly conserved portions of the protein.103,104 and 105 Mutations have not been detected in either the type I or II gene to explain mild or nonclassic 3b-HSD deficiency.106 Thus, the latter syndromes appear to be attributable to other causes, perhaps not genetic.

CHOLESTEROL DESMOLASE DEFICIENCY The CYP11A gene encoding cholesterol desmolase encompasses 20 kb on chromosome 15.107 Mutations in this gene have not been identified in patients with lipoid adrenal hyperplasia,108 although in vitro studies suggest that the 20a-hydroxylase function is deficient in at least one patient with the syndrome. Mutations affecting another cellular component fundamental to early steroidogenesis, steroidogenic acute regulatory proteins (StAR), are responsible for this syndrome.109

OTHER GENETIC DISORDERS RELATED TO CONGENITAL ADRENAL HYPERPLASIA

Several other disorders involve adrenal steroid-synthesizing enzymes. Strictly speaking, the CMO deficiencies and other enzyme defects discussed below are not included among the adrenal hyperplasias, because they do not affect cortisol synthesis and cause no disturbance of the hypothalamicpituitaryadrenal axis. A distal block in aldosterone synthesis results from defects in the CYP11B2 gene encoding 18-hydroxylase (CMO I) and 18-oxidase (CMO II) and causes rare forms of recessively inherited salt wasting with hypotension and failure to thrive in infancy. The biochemical profile in cases of CMO II deficiency is notable for low levels of aldosterone, with high PRA, and high serum corticosterone and 18-hydroxycorticosterone. The latter steroid is found in low levels in cases of CMO I deficiency.110 Affected individuals are often clinically asymptomatic in later life. Defects in the CYP19 gene encoding cytochrome P450 aromatase (P450arom) prevent the normal synthesis of estrogens and create a relative abundance of androgens. Such defects have been identified as a novel cause of ambiguous genitalia in 46,XX neonates, of primary amenorrhea with hypergonadotropic hypogonadism in adolescent girls, or of virilism in the setting of polycystic ovaries in young women.111,112 In men the defect results in continued linear bone growth beyond puberty, delayed bone age, and failure of epiphyseal closure, findings which prove that estrogens are critical for epiphyseal fusion. These symptoms are alleviated by estrogen administration.113 Classic deficiency of 17-ketosteroid reductase (i.e., 17a-hydroxysteroid dehydrogenase) is a cause of ambiguous genitalia in 46,XY neonates, who have a tendency to virilize at puberty, probably by means of extragonadal 17a-HSD activity and enhanced 5a-reductase activity.114,115 Mild forms of this enzyme deficiency exist in young men with gynecomastia and in women with hirsutism and polycystic ovaries.116,117 Specific defects in the gene HSD17B3 encoding gonadal 17a-HSD have been identified in subjects with classic deficiency of this enzyme.118,119 The typical hormonal profile shows a markedly elevated ratio of androstenedione to testosterone. Other rare inherited disorders of steroidogenesis, such as glucocorticoid-suppressible hyperaldosteronism (i.e., glucocorticoid-remediable aldosteronism or dexamethasone-suppressible hyperaldosteronism), a regulatory defect of the CYP11B2gene,120 and apparent mineralocorticoid excess, a defect in the renal isozyme 11-hydroxysteroid dehydrogenase, are discussed in Chapter 80.

CONCLUSION
The adrenal hyperplasias have been extensively studied from the clinical and molecular genetic perspectives. The molecular bases for the various forms of CAH have been identified. Severe mutations in the genes encoding steroidogenic enzymes, such as deletions, frame shifts, and nonsense codons, result in gene products with no enzymatic activity. In contrast, milder mutations, such as conservative or nonconservative substitutions, may cause a lesser degree of enzyme impairment. The catalytic activity of such gene products may be differently affected for each of two different substrates. Interindividual variation also may occur in individuals with similar genotypes. Two apparent enzyme deficiencies, nonclassic 3b-HSD deficiency and cholesterol desmolase/lipoid CAH, are not due to defects in the structural genes encoding these respective enzymes. One practical result of molecular genetic characterization is the ability to perform accurate and early prenatal diagnosis. Current research efforts are focused on achieving regulation of these genes, understanding more about gene and enzyme structure-function relationships, identifying further clinical-genetic correlations, and optimizing treatment. CHAPTER REFERENCES
1. New MI, White PC, Pang S, et al. The adrenal hyperplasias. In: Scriver CR, Beaudet AL, Sly WS, Valle D, eds. The metabolic basis of inherited disease, 6th ed. New York: McGraw-Hill, 1989:1881. 2. White PC, New MI, Dupont B. Congenital adrenal hyperplasia. N Engl J Med 1987; 316:1519. 2a. 3. 4. 5. 6. 7. Speiser PW, White PC. Congenital adrenal hyperplasia. Endocr Rev 2000; in press.

Swain A, Lovell-Badge R. A molecular approach to sex determination in mammals. Acta Paediatr Suppl 1997; 423:46. Branchaud CL, Murphy BEP. Physiopathology of the fetal adrenal. In: Pasqualini JR, Scholler R, eds. Hormones and fetal pathophysiology. New York: Marcel Dekker, 1992:53. Estivariz FE, Lowry PJ, Jackson S. Control of adrenal growth. In: James VHT, ed. The adrenal gland, 2nd ed. New York: Raven Press, 1992:43. Kohn B, Levine LS, Pollack MS, et al. Late-onset steroid 21-hydroxylase deficiency: a variant of classical congenital adrenal hyperplasia. J Clin Endocrinol Metab 1982; 55:817. Pang S, Hotchkiss J, Drash AL, Levine LS, New MI. Microfilter paper method for 17a-progesterone radioimmunoassay: its application for rapid screening for congenital adrenal hyperplasia. J Clin Endocrinol Metab 1977; 45:1003. 8. Pang S, Murphey W, Levine LS, et al. A pilot newborn screening for congenital adrenal hyperplasia (CAH) in Alaska. Pediatr Res 1981; 15:512. 9. Pang SP, Wallace MA, Hofman L, et al. Worldwide experience in newborn screening for classical congenital adrenal hyperplasia due to 21-hydroxylase deficiency. Pediatrics 1988; 81:866. 9a. Therrell BL Jr, Berenbaum SA, Manter-Kapanke V, et al. Results of screening 1.9 million Texas newborns for 21-hydroxylase-deficient congenital adrenal hyperplasia. Pediatrics 1998; 101(4 Pt 1):583.

10. Speiser PW, Dupont B, Rubinstein P, et al. High frequency of nonclassical steroid 21-hydroxylase deficiency. Am J Hum Genet 1985; 37:650. 11. Sherman SL, Aston CE, Morton NE, et al. A segregation and linkage study of classical and nonclassical 21-hydroxylase deficiency. Am J Hum Genet 1988; 42:830. 12. Zachmann M, Tassinari D, Prader A. Clinical and biochemical variability of congenital adrenal hyperplasia due to 11 beta-hydroxylase deficiency. A study of 25 patients. J Clin Endocrinol Metab 1983; 56:222. 13. Rosler A, Leiberman E, Cohen T. High frequency of congenital adrenal hyperplasia (classic 11 beta-hydroxylase deficiency) among Jews from Morocco. Am J Med Genet 1992; 42:827. 14. Biglieri EG, Herron MA, Brust M. 17-Hydroxylation deficiency in man. J Clin Invest 1966; 45:1946. 15. New MI. Male pseudohermaphroditism due to 17a-hydroxylase deficiency. J Clin Invest 1970; 49:1930. 16. Yanase T, Simpson ER, Waterman MR. 17a-Hydroxylase/17,20-lyase deficiency: from clinical investigation to molecular definition. Endocr Rev 1991; 12:91. 17. Imai T, Toshihiko T, Waterman MR, et al. Canadian Mennonites and individuals residing in the Friesland region of the Netherlands share the same molecular basis of 17a-hydroxylase deficiency. Hum Genet 1992; 89:95. 18. Ahlgren R, Yanase T, Simpson ER, et al. Compound heterozygous mutations (Arg239stop, Pro342Thr) in the CYP17 (P45017a) gene lead to ambiguous external genitalia in a male patient with partial combined 17a-hydroxylase deficiency. J Clin Endocrinol Metab 1992; 74:667. 19. Bongiovanni AM. The adrenogenital syndrome with deficiency of 3b-hydroxysteroid dehydrogenase. J Clin Invest 1962; 41:2086. 20. Bongiovanni AM. Acquired adrenal hyperplasia: with special reference to 3b-hydroxysteroid dehydrogenase. Fertil Steril 1981; 35:599. 21. Zerah M, Schram P, New MI. The diagnosis and treatment of nonclassical 3b-HSD deficiency. Endocrinologist 1991; 1:75. 22. Ehrmann DA, Rosenfield RL. Hirsutismbeyond the steroidogenic block. N Engl J Med 1990; 323:909. 23. Prader A, Gurtner HP. Das Syndrom des Pseudohermaphroditismus masculinus bei kongenitaler Nebennierenrinden-Hyperplasia ohne Androgenberproduktion (adrenaler Pseudoherm masc). Helv Pediatr Acta 1955; 10:397. 24. Prader A, Siebenmann RE. Nebenniereninsuffizienz bei kongenitaler Lipoid-hyperplasie der Nebennieren. Helv Pediatr Acta 1957; 12:569. 25. Hauffa BP, Miller WL, Grumbach MM, et al. Congenital adrenal hyperplasia due to deficient cholesterol side-chain cleavage activity (20,22-desmolase) in a patient treated for 18 years. Clin Endocrinol 1985; 23:481. 25a. Stocco DM. The role of the StAR protein in steroidogenesis: challenges for the future. J Endocrinol 2000; 164(3):247. 26. New MI, Lorenzen F, Lerner AJ, et al. Genotyping steroid 21-hydroxylase deficiency: hormonal reference data. J Clin Endocrinol Metab 1983; 57:320. 27. Eberlein WR, Bongiovanni AM. Plasma and urinary corticosteroids in the hypertensive form of congenital adrenal hyperplasia. J Biol Chem 1956; 223:85. 28. Hughes IA, Arisaka O, Perry LA, et al. Early diagnosis of 11-beta-hydroxylase deficiency in two siblings confirmed by analysis of a novel steroid metabolite in newborn urine. Acta Endocrinol (Copenh) 1986; 111:349. 29. Bongiovanni AM. Urinary excretion of pregnanetriol and pregnenetriol in two forms of congenital adrenal hyperplasia. J Clin Invest 1971; 60:2751. 30. Jeffcoate TNA, Fleigner JRH, Russell SH, et al. Diagnosis of the adrenogenital syndrome before birth. Lancet 1965; 2:553. 31. Merkatz IR, New MI, Seaman MP. Prenatal diagnosis of adrenogenital syndrome by amniocentesis. J Pediatr 1969; 75:977. 32. Frasier SD, Thorneycroft IH, Weiss BA, Horton R. Elevated amniotic fluid concentration of 17a-hydroxyprogesterone in congenital adrenal hyperplasia. J Pediatr 1975; 86:310. 33. Gueux B, Fiet J, Couillin P, et al. Prenatal diagnosis of 21-hydroxylase deficiency congenital adrenal hyperplasia by simultaneous radioimmunoassay of 21-deoxycortisol and 17-hydroxyprogesterone in amniotic fluid. J Clin Endocrinol Metab 1988; 66:534. 34. Couillin P, Nicolas H, Boue J, Boue A. HLA typing of amniotic-fluid cells applied to prenatal diagnosis of congenital adrenal hyperplasia. Lancet 1979; 1:1076. 35. Pollack MS, Levine LS, Pang S, et al. Prenatal diagnosis of congenital adrenal hyperplasia (21-hydroxylase deficiency) by HLA typing. Lancet 1979; 1:1107. 36. Pang S, Pollack MS, Loo M, et al. Pitfalls of prenatal diagnosis of 21-hydroxylase deficiency congenital adrenal hyperplasia. J Clin Endocrinol Metab 1985; 61:89. 37. Odink RJH, Boue A, Jansen M. The value of chorion villus sampling in early detection of 21-hydroxylase deficiency. Pediatr Res 1988; 23:131A.

38. Shulman DI, Mueller OT, Gallardo LA, et al. Treatment of congenital adrenal hyperplasia in utero. Pediatr Res 1989; 25:93A. 39. Owerbach D, Draznin MB, Carpenter RJ, Greenberg F. Prenatal diagnosis of 21-hydroxylase deficiency congenital adrenal hyperplasia using the polymerase chain reaction. Hum Genet 1992; 89:109. 40. Speiser PW, White PC, Dupont J, et al. Prenatal diagnosis of congenital adrenal hyperplasia due to 21-hydroxylase deficiency by allele-specific hybridization and Southern blot. Hum Genet 1994; 93:424. 41. David M, Forest MG. Prenatal treatment of congenital adrenal hyperplasia resulting from 21-hydroxylase deficiency. J Pediatr 1984; 105:799. 42. Evans MI, Chrousos GP, Mann DW, et al. Pharmacologic suppression of the fetal adrenal gland in utero. JAMA 1985; 253:1015. 43. Forest MG, Betuel H, David M. Prenatal treatment of congenital adrenal hyperplasia with 21-hydroxylase deficiency: a multicenter study. Ann Endocrinol (Paris) 1987; 48:31. 44. Speiser PW, LaForgia N, Kato K, et al. First trimester prenatal treatment and molecular genetic diagnosis of congenital adrenal hyperplasia (21-hydroxylase deficiency). J Clin Endocrinol Metab 1990; 70:838. 45. Migeon CJ. Comments about the need for prenatal treatment of congenital adrenal hyperplasia due to 21-hydroxylase deficiency. (Editorial). J Clin Endocrinol Metab 1990; 70:836. 46. Prenatal treatment of congenital adrenal hyperplasia. (Editorial). Lancet 1990; 510. 47. Forest MG, Morel Y, David M. Prenatal diagnosis and treatment of congenital adrenal hyperplasia. Horm Res 1997; 48(suppl 2): 22. 48. Pang S, Pollack MS, Marshall RN, Immken L. Prenatal treatment of congenital adrenal hyperplasia due to 21-hydroxylase deficiency. N Engl J Med 1990; 322:111. 49. Lajic S, Wedell A, Bui TH, Ritzen EM, Holst M. Long-term somatic follow-up of prenatally treated children with congenital adrenal hyperplasia. J Clin Endocrinol Metab 1998; 83:3872. 50. Goldman AS, Sharpior BH, Katsumata M. Human foetal palatal corticoid receptors and teratogens for cleft palate. Nature 1978; 272:464. 51. Seckl JR, Miller WL. How safe is long-term prenatal glucocorticoid treatment? JAMA 1997; 277:1077. 52. New MI, Seaman MP. Secretion rates of cortisol and aldosterone in various forms of congenital adrenal hyperplasia. J Clin Endocrinol 1970; 30:361. 53. Linder BL, Esteban NV, Yergey AL, et al. Cortisol production rate in childhood and adolescence. J Pediatr 1990; 117:892. 54. Golden MP, Lippe BM, Kaplan SA, et al. Management of congenital adrenal hyperplasia using serum dehydroepiandrosterone sulfate and 17-hydroxyprogesterone concentrations. Pediatrics 1978; 61:867. 55. Cutler GB Jr, Laue L. Seminars in medicine of the Beth Israel Hospital, Boston: congenital adrenal hyperplasia due to 21-hydroxylase deficiency. N Engl J Med 1990; 323:1806. 56. Merke DP, Keil MF, Jones JV, et al. Flutamide, testolactone, and reduced hydrocortisone dose maintain normal growth velocity and bone maturation despite elevated androgen levels in children with congenital adrenal hyperplasia. J Clin Endocrinol Metab 2000; 85(3):1114. 57. Van Wyk JJ, Gunther DF, Ritzen EM, et al. The use of adrenalectomy as a treatment for congenital adrenal hyperplasia. J Clin Endocrinol Metab 1996; 81:3180. 58. Spritzer P, Billaud L, Thalabard J-C, et al. Cyproterone acetate versus hydrocortisone treatment in late-onset adrenal hyperplasia. J Clin Endocrinol Metab 1990; 70:642. 59. Donahoe PK, Gustafson ML. Early one-stage surgical reconstruction of the extremely high vagina in patients with congenital adrenal hyperplasia. J Pediatr Surg 1994; 29:352. 60. Mulaikal RM, Migeon CJ, Rock JA. Fertility rates in female patients with congenital adrenal hyperplasia due to 21-hydroxylase deficiency. N Engl J Med 1987; 316:178. 61. Kuhnle U, Bollinger M, Schwarz HP, Knorr D. Partnership and sexuality in adult female patients with congenital adrenal hyperplasia. First results of a cross-sectional quality-of-life evaluation. J Steroid Biochem Mol Biol 1993; 45:123. 62. Knudson AG Jr. Mixed adrenal disease of infancy. J Pediatr 1951; 39:408. 63. Bentinck RC, Hinman F Sr, Lisser H, et al. The familial congenital adrenal syndrome: report of two cases and review of the literature. Postgrad Med 1952; 11:301. 64. Childs B, Grumbach MM, Van Wyk JJ. Virilizing adrenal hyperplasia: a genetic and hormonal study. J Clin Invest 1956; 35:213. 65. Dupont B, Oberfield SE, Smithwick EM, et al. Close genetic linkage between HLA and congenital adrenal hyperplasia (21-hydroxylase deficiency). Lancet 1977; 2:1309. 66. White PC, New MI. Genetic basis of endocrine disease, 2: congenital adrenal hyperplasia due to 21-hydroxylase deficiency. J Clin Endocrinol Metab 1992; 74:6. 67. White PC, Grossberger D, Onufer BJ, et al. Two genes encoding steroid 21-hydroxylase are located near the genes encoding the fourth component of complement in man. Proc Natl Acad Sci U S A 1985; 82:1089. 68. Carroll MC, Campbell RD, Porter RR. The mapping of 21-hydroxylase genes adjacent to complement component C4 genes in HLA, the major histocompatibility complex in man. Proc Natl Acad Sci U S A 1985; 82:521. 69. White PC, Vitek A, Dupont B, New MI. Characterization of frequent deletions causing steroid 21-hydroxylase deficiency. Proc Natl Acad Sci U S A 1988; 85:4436. 70. Owerbach D, Crawford YM, Draznin MB. Direct analysis of CYP21B genes in 21-hydroxylase deficiency using polymerase chain reaction amplification. Mol Endocrinol 1990; 4:125. 71. Mornet E, Crete P, Kuttenn F, et al. Distribution of deletions and seven point mutations on CYP21B genes in three clinical forms of steroid 21-hydroxylase deficiency. Am J Hum Genet 1991; 48:79. 72. Higashi Y, Hiromasa T, Tanae A, et al. Effects of individual mutations in the P-450(C21) pseudogene on the P-450(C21) activity and their distribution in the patient genomes of congenital steroid 21-hydroxylase deficiency. J Biochem 1991; 109:638. 73. Speiser PW, Dupont J, Zhu D, et al. Disease expression and molecular genotype in congenital adrenal hyperplasia due to 21-hydroxylase deficiency. J Clin Invest 1992; 90:584. 74. Wedell A, Thiln A, Ritzn EM, et al. Mutational spectrum of the steroid 21-hydroxylase gene in Sweden: implications for genetic diagnosis and association with disease manifestation. J Clin Endocrinol Metab 1994; 78:1145. 75. Higashi Y, Tanae A, Inoue H, et al. Aberrant splicing and missense mutations cause steroid 21-hydroxylase [P-450(C21)] deficiency in humans: possible gene conversion products. Proc Natl Acad Sci U S A 1988; 85:7486. 76. Tusie-Luna MT, Traktman P, White PC. Determination of functional effects of mutations in the steroid 21-hydroxylase gene (CYP21) using recombinant vaccinia virus. J Biol Chem 1990; 265:20916. 77. Tusie-Luna MT, Speiser PW, Dumic M, et al. A mutation (Pro-30 to Leu) in CYP21 represents a potential nonclassic steroid 21-hydroxylase deficiency allele. Mol Endocrinol 1991; 5:685. 78. Higashi Y, Hiromasa T, Tanae A, et al. Effects of individual mutations in the P-450(C21) pseudogene on the P-450(C21) activity and their distribution in the patient genomes of congenital steroid 21-hydroxylase deficiency. J Biochem (Tokyo) 1991; 109:638. 79. Dickerman Z, Grant DR, Faiman C, Winter JSD. Intraadrenal steroid concentrations in man: zonal differences and developmental changes. J Clin Endocrinol Metab 1984; 59:1031. 80. Nikoshkov A, Lajic S, Vlamis-Gardikas A, et al. Naturally occurring mutants of human steroid 21-hydroxylase (P450c21) pinpoint residues important for enzyme activity and stability. J Biol Chem 1998; 273:6163. 81. Mornet E, Dupont J, Vitek A, White PC. Characterization of two genes encoding human steroid 11b-hydroxylase (P-450 11b). J Biol Chem 1989; 264:20961. 82. Chua SC, Szabo P, Vitek A, et al. Cloning of cDNA encoding steroid 11b-hydroxylase (P450c11). Proc Natl Acad Sci U S A 1987; 84:7193. 83. Kawamoto T, Mitsuuchi Y, Toda K, et al. Cloning of cDNA and genomic DNA for human cytochrome P-45011b. FEBS Lett 1990; 269:345. 84. Kawamoto T, Mitsuuchi Y, Ohnishi T, et al. Cloning and expression of a cDNA for human cytochrome P-450aldos as related to primary aldosteronism. Biochem Biophys Res Commun 1990; 173:309. 85. Curnow KM, Tusie-Luna MT, Pascoe L, et al. The product of the CYP11B2 gene is required for aldosterone biosynthesis in the human adrenal cortex. Mol Endocrinol 1991; 5:1513. 86. White PC, Pascoe L, Curnow KM, et al. Molecular biology of 11b-hydroxylase and 11b-hydroxysteroid dehydrogenase enzymes. J Steroid Biochem Mol Biol 1992; 43:827. 87. Curnow KM, Slutsker L, Vitek J, et al. Mutations in the CYP11B1 gene causing congenital adrenal hyperplasia and hypertension cluster in exons 6, 7, and 8. Proc Natl Acad Sci U S A 1993; 90:4552. 88. White PC, Dupont J, New MI, et al. A mutation in CYP11B1 (Arg-448His) associated with steroid 11b-hydroxylase deficiency in Jews of Moroccan origin. J Clin Invest 1991; 87:1664. 89. Naiki Y, Kawamoto T, Mitsuuchi Y, et al. A nonsense mutation (TGG116TAG[Stop]) in CYP11B1 causes steroid 11b-hydroxylase deficiency. J Clin Endocrinol Metab 1993; 77:1677. 90. Helmberg A, Ausserer B, Kofler R. Frame shift by insertion of 2 basepairs in codon 394 of CYP11B1 causes congenital adrenal hyperplasia due to steroid 11 beta-hydroxylase deficiency. J Clin Endocrinol Metab 1992; 75:1278. 91. Nakagawa Y, Yamada M, Ogawa H, Igarashi Y. Missense mutation in CYP11B1 (CGA[Arg-384]GGA[Gly]) causes steroid 11 beta-hydroxylase deficiency. Eur J Endocrinol 1995; 132:286. 92. Yang LX, Toda K, Miyahara K, et al. Classic steroid 11 beta-hydroxylase deficiency caused by a CG transversion in exon 7 of CYP11B1. Biochem Biophys Res Commun 1995; 216:723. 93. Geley S, Kapelari K, Johrer K, et al. CYP11B1 mutations causing congenital adrenal hyperplasia due to 11 beta-hydroxylase deficiency. J Clin Endocrinol Metab 1996; 81:2896. 94. Merke DP, Tajima T, Chhabra A, et al. Novel CYP11B1 mutations in congenital adrenal hyperplasia due to steroid 11 beta-hydroxylase deficiency. J Clin Endocrinol Metab 1998; 83:270. 95. Joehrer K, Geley S, Strasser-Wozak EM, et al. CYP11B1 mutations causing non-classic adrenal hyperplasia due to 11 beta-hydroxylase deficiency. Hum Mol Genet 1997; 6:1829. 96. Matteson KJ, Picado-Leonard J, Chung B-C, et al. Assignment of the gene for adrenal P450c17 (steroid 17a-hydroxylase/17,20-lyase) to human chromosome 10. J Clin Endocrinol Metab 1986; 63:789. 97. Chung B-C, Picado-Leonard J, Haniu M, et al. Cytochrome P450c17 (steroid 17a-hydroxylase/17,20-lyase): cloning of human adrenal and testis cDNAs indicates the same gene is expressed in both tissues. Proc Natl Acad Sci U S A 1987; 84:407. 98. Suzuki Y, Nagashima T, Nomura Y, et al. A. A new compound heterozygous mutation (W17X, 436 + 5GT) in the cytochrome P450c17 gene causes 17 alpha-hydroxylase/17,20-lyase deficiency. J Clin Endocrinol Metab 1998; 83:199. 99. Lachance Y, Luu-The V, Labrie C, et al. Characterization of human 3b-hydroxysteroid dehydrogenase/D 5-D 4 isomerase gene and its expression in mammalian cells. J Biol Chem 1990; 265:20469. 100. Lorence MC, Corbin CJ, Kamimura N, et al. Structural analysis of the gene encoding human 3b-hydroxysteroid dehydrogenase/D 5D4-isomerase. Mol Endocrinol 1990; 4:1850. 101. Lachance Y, Luu-The V, Verreault H, et al. Structure of the human type II 3b-hydroxysteroid dehydrogenase/D 5-D 4 isomerase (3b-HSD) gene: adrenal and gonadal specificity. DNA Cell Biol 1991; 10:701. 102. Berube D, Luu-The V, Lachance Y, et al. Assignment of the human 3b-HSD gene to the p13 band of chromosome 1. Cytogenet Cell Genet 1989; 52:199. 102a. McCartin S, Russell AJ, Fisher RA, et al. Phenotypic variability and origins of mutations, in the gene encoding 3b-hydroxysteroid dehydrogenase type II. Mol Endocrinol 2000; 24:75. 103. Rheaume E, Simard J, Morel Y, et al. Congenital adrenal hyperplasia due to point mutations in the type II 3b-hydroxysteroid dehydrogenase gene. Nature 1992; 1:239. 104. Chang YT, Kappy MS, Iwamoto K, et al. Mutations in the type II 3 beta-hydroxysteroid dehydrogenase (3 beta-HSD) gene in a patient with classic salt-wasting 3 beta-HSD deficiency congenital adrenal hyperplasia. Pediatr Res 1993; 34:698. 105. Simard J, Rheaume E, Sanchez R, et al. Molecular basis of congenital adrenal hyperplasia due to 3b-hydroxysteroid dehydrogenase deficiency. Mol Endocrinol 1993; 7:716. 106. Sakkal-Alkaddour H, Zhang L, Yang X, et al. Studies of 3 beta-hydroxy-steroid dehydrogenase genes in infants and children manifesting premature pubarche and increased adrenocorticotropin-stimulated delta 5-steroid levels. J Clin Endocrinol Metab 1996; 81:3961. 107. Chung BC, Matteson KJ, Voutilainen R, et al. Human cholesterol side-chain cleavage enzyme, P450scc: cDNA cloning, assignment of the gene to chromosome 15, and expression in the placenta. Proc Natl Acad Sci U S A 1986; 83:8962. 108. Lin D, Gitelman SE, Saenger P, Miller WL. Normal genes for the cholesterol side chain cleavage enzyme, P450ssc, in congenital lipoid adrenal hyper-plasia. J Clin Invest 1991; 88:1955. 109. Miller WL. Congenital lipoid adrenal hyperplasia: the human gene knockout for the steroidogenic acute regulatory protein. J Mol Endocrinol 1997; 19:227. 110. Ulick S, Wang JZ, Morton DH. The biochemical phenotypes of two inborn errors in the biosynthesis of aldosterone. J Clin Endocrinol Metab 1992; 74:1415. 111. Harada N, Ogawa H, Shozu M, Yamada K. Genetic studies to characterize the origin of the mutation in placental aromatase deficiency. Am J Hum Genet 1992; 51:666. 112. Ito Y, Fisher CR, Conte FA, et al. Molecular basis of aromatase deficiency in an adult female with sexual infantilism and polycystic ovaries. Proc Natl Acad Sci U S A 1993; 90:11673. 113. Simpson ER, Zhao Y, Agarwal VR, et al. Aromatase expression in health and disease. Recent Prog Horm Res 1997; 52:185. 114. Eckstein B, Cohen S, Farkas A, Rosler A. The nature of the defect in familial male pseudohermaphroditism in Arabs of Gaza. J Clin Endocrinol Metab 1989; 68:477. 115. Rosler A, Belanger A, Labrie F. Mechanisms of androgen production in male pseudohermaphroditism due to 17 beta-hydroxysteroid dehydrogenase deficiency. J Clin Endocrinol Metab 1992; 75:773. 116. Castro-Magana M, Angulo M, Uy J. Male hypogonadism with gynecomastia caused by late-onset deficiency of testicular 17-ketosteroid reductase. N Engl J Med 1993; 328:1297. 117. Pang S, Softness B, Sweeney WJ, New MI. Hirsutism, polycystic ovarian disease, and ovarian 17-ketosteroid reductase deficiency. N Engl J Med 1987; 316:1295. 118. Andersson S, Moghrabi N. Physiology and molecular genetics of 17 beta-hydroxysteroid dehydrogenases. Steroids 1997; 62:143. 119. Moghrabi N, Hughes IA, Dunaif A, Andersson S. Deleterious missense mutations and silent polymorphism in the human 17beta-hydroxysteroid dehydrogenase 3 gene (HSD17B3). J Clin Endocrinol Metab 1998; 83:2855. 120. Nyckoff JA, Seely EW, Hurwitz S, et al. Glucocorticoid-remediable aldosteronism and pregnancy. Hypertension 2000; 35:668.

CHAPTER 78 CORTICOSTEROID THERAPY Principles and Practice of Endocrinology and Metabolism

CHAPTER 78 CORTICOSTEROID THERAPY

LLOYD AXELROD Glucocorticoids Structure of Commonly used Glucocorticoids Pharmacodynamics Glucocorticoid Therapy in the Presence of Liver Disease Glucocorticoid Therapy and the Nephrotic Syndrome Glucocorticoid Therapy and Hyperthyroidism Glucocorticoids During Pregnancy Glucocorticoid Therapy and Age Drug Interactions Considerations Before Initiating the use of Glucocorticoids as Pharmacologic Agents Effects of Exogenous Glucocorticoids Suppression of the Hypothalamic-Pituitary-Adrenal System Withdrawal of Patients from Glucocorticoid Therapy Alternate-Day Glucocorticoid Therapy Daily Single-Dose Glucocorticoid Therapy Glucocorticoids or Corticotropin? Dosage Mineralocorticoids Pharmacology Drug Interactions Indications Dosage Chapter References

This chapter examines the risks associated with the use of glucocorticoids and of mineralocorticoids for various illnesses, and provides guidelines for the administration of these commonly prescribed substances.

GLUCOCORTICOIDS
STRUCTURE OF COMMONLY USED GLUCOCORTICOIDS Figure 78-1 indicates the structures of several commonly used glucocorticoids.1,2 Cortisol (hydrocortisone) is the principal circulating glucocorticoid in humans.

FIGURE 78-1. The structures of commonly used glucocorticoids. In the depiction of cortisol, the 21 carbon atoms of the glucocorticoid skeleton are indicated by numbers and the four rings are designated by letters. The arrows indicate the structural differences between cortisol and each of the other molecules. (From Axelrod L. Glucocorticoid therapy. Medicine [Baltimore] 1976; 55:39, and Axelrod L. Glucocorticoids. In: Kelley WN, Harris ED Jr, Ruddy S, Sledge CB, eds. Textbook of rheumatology, 4th ed. Philadelphia: WB Saunders, 1993:779.)

Glucocorticoid activity requires a hydroxyl group at carbon 11 of the steroid molecule. Cortisone and prednisone are 11-keto compounds. Consequently, they lack glucocorticoid activity until they are converted in vivo to cortisol and prednisolone, the corresponding 11-hydroxyl compounds.3,4 This conversion occurs predominantly in the liver. Thus, topical application of cortisone is ineffective in the treatment of dermatologic diseases that respond to topical application of cortisol.4 Similarly, the antiinflammatory action of cortisone delivered by intraarticular injection is minimal compared with the effect of cortisol administered in the same manner.3 Cortisone and prednisone are used only for systemic therapy. All glucocorticoid preparations marketed for topical or local use are 11-hydroxyl compounds, which obviates the need for biotransformation. PHARMACODYNAMICS HALF-LIFE, POTENCY, AND DURATION OF ACTION The important differences among the systemically used glucocorticoid compounds are duration of action, relative glucocorticoid potency, and relative mineralocorticoid potency (Table 78-1).1,2 The commonly used glucocorticoids are classified as short-acting, intermediate-acting, and long-acting on the basis of the duration of corticotropin (ACTH) suppression after a single dose, equivalent in antiinflammatory activity to 50 mg of prednisone (Table 78-1).5 The relative potencies of the glucocorticoids correlate with their affinities for the intracellular glucocorticoid receptor.6 The observed potency of a glucocorticoid, however, is determined not only by the intrinsic biologic potency, but also by the duration of action.6,7 Consequently, the relative potency of two glucocorticoids varies as a function of the time interval between the administration of the two steroids and the determination of the potency. In particular, failure to account for the duration of action may lead to a marked underestimation of the potency of dexamethasone.7

TABLE 78-1. Commonly Used Glucocorticoids

The correlation between the circulating half-life (T1/2) of a glucocorticoid and its potency is weak. The T1/2 of cortisol in the circulation is in the range of 80 to 115 minutes.1 The T1/2s of other commonly used agents are cortisone, 0.5 hours; prednisone, 3.4 to 3.8 hours; prednisolone, 2.1 to 3.5 hours; methyl-prednisolone, 1.3 to 3.1 hours; and dexamethasone 1.8 to 4.7 hours.1,7,8 Prednisolone and dexamethasone have comparable circulating T1/2s, but dexamethasone is clearly more potent. Similarly, the correlation between the circulating T1/2 of a glucocorticoid and its duration of action is poor. The many actions of glucocorticoids do not have an equal duration, and the duration of action may be a function of the dose. The duration of ACTH suppression is not simply a function of the level of antiinflammatory activity, because variations in the duration of ACTH suppression are achieved by doses of glucocorticoids with comparable antiinflammatory activity. The duration of ACTH suppression produced by an individual glucocorticoid, however, probably is dose related.5 In short, the slight differences in the circulating T 1/2s of the glucocorticoids contrast with their marked differences in potency and duration of ACTH suppression. Thus, the duration of action of a glucocorticoid is not determined by its presence in the circulation. This is consistent with the mechanism of action of steroid hormones. A steroid molecule binds to a specific intracellular receptor protein (see Chap. 4). This steroid-receptor complex modifies the process of transcription by which RNA is transcribed from the DNA template. This process alters the rate of synthesis of specific proteins. The steroid thereby modifies the phenotypic expression of the genetic information. Thus, the glucocorticoid continues to act inside the cell after it has disappeared from the circulation. Moreover, the events initiated by the glucocorticoid may continue to occur, or a product of these events (such as a specific protein) may be present after the disappearance of the glucocorticoid. BIOAVAILABILITY, ABSORPTION, AND BIOTRANSFORMATION Normally, a person's plasma cortisol level is much lower after the oral administration of cortisone than after an equal dose of cortisol.9 Consequently, although oral cortisone may be adequate replacement therapy in chronic adrenal insufficiency, the oral form of this agent should not be used when larger, pharmacologic effects are sought. Comparable plasma prednisolone levels are achieved in normal persons after equivalent oral doses of prednisone and prednisolone.8,10 After the administration of either of these corticosteroids, however, there is wide variation in individual prednisolone concentrations, which may reflect variability in absorption.8 In contrast to the marked rises that follow the intramuscular injection of hydrocortisone, plasma cortisol levels rise little or not at all after an intramuscular injection of cortisone acetate. When it is given intramuscularly, cortisone acetate does not provide adequate plasma cortisol levels and offers no advantage over hydrocortisone delivered by the same route. The explanation for the failure of intramuscular cortisone acetate to provide adequate plasma cortisol levels is unknown. It may reflect poor absorption from the site of injection. Alternatively, intramuscular cortisone acetate, which reaches the liver through the systemic circulation, may be metabolically inactivated before it can be converted to cortisol in the liver, in contrast to oral cortisone acetate, which reaches the liver through the portal circulation. PLASMA TRANSPORT PROTEINS In normal humans, circadian fluctuations occur in the capacity of corticosteroid-binding globulin (transcortin) to bind cortisol and prednisolone. Patients who have been treated with prednisone for a prolonged period have no diurnal variation in the binding capacity of corticosteroid-binding globulin for cortisol or prednisolone, and both capacities are reduced in comparison with normal persons. Thus, long-term glucocorticoid therapy not only alters the endogenous secretion of steroids, but also affects the transport of some glucocorticoids in the circulation. This may explain why the disappearance of prednisolone is more rapid in those persons who have previously received glucocorticoids. GLUCOCORTICOID THERAPY IN THE PRESENCE OF LIVER DISEASE Plasma cortisol levels are normal in patients with hepatic disease. Although the clearance of cortisol is reduced in patients with cirrhosis, the hypothalamic-pituitary-adrenal (HPA) homeostatic mechanism remains intact. Consequently, the decreased rate of metabolism is accompanied by decreased synthesis of cortisol (see Chap. 205). The conversion of prednisone to prednisolone is impaired in patients with active liver disease.11 This is largely offset by a decreased rate of elimination of prednisolone from the plasma in these patients.11 In patients with liver disease, the plasma availability of prednisolone is quite variable after oral doses of either prednisone or prednisolone.12 This is further complicated by the lower percentage of plasma prednisolone that is bound to protein in patients with active liver disease; the unbound fraction is inversely related to the serum albumin concentration. An increased frequency of prednisone side effects is observed at low serum albumin levels.12 Both these findings may reflect impaired hepatic function. Because the impairment of conversion of prednisone to prednisolone is quantitatively small in the presence of liver disease and is offset by a decreased rate of clearance of prednisolone, and because of the marked variability in plasma prednisolone levels after the administration of either corticosteroid, there is no clear mandate to use prednisolone rather than prednisone in patients with active liver disease or cirrhosis.8 If prednisone or prednisolone is used, however, a somewhat lower than usual dose should be given if the serum albumin level is low.8 GLUCOCORTICOID THERAPY AND THE NEPHROTIC SYNDROME When hypoalbuminemia is caused by the nephrotic syndrome, the fraction of prednisolone that is protein bound is decreased. The unbound fraction is inversely related to the serum albumin concentration. The unbound prednisolone concentration remains normal, however.13,14 Because the pharmacologic effect is determined by the unbound concentration, altered prednisolone kinetics do not explain the increased frequency of predniso-lone-related side effects in these patients. GLUCOCORTICOID THERAPY AND HYPERTHYROIDISM The bioavailability of prednisolone after an oral dose of prednisone is reduced in patients with hyperthyroidism because of decreased absorption of prednisone and increased hepatic clearance of prednisolone.15 GLUCOCORTICOIDS DURING PREGNANCY Glucocorticoid therapy is well tolerated in pregnancy.16 glucocorticoids cross the placenta, but there is no compelling evidence that this produces clinically significant HPA suppression or Cushing syndrome in neonates,16 although subnormal responsiveness to exogenous ACTH may occur. Similarly, there is no evidence that glucocorticoids increase the incidence of congenital defects in humans.16 Glucocorticoids do appear to decrease the birth weight of full-term infants; the long-term consequences of this are unknown. Because the concentrations of prednisone and prednisolone in breast milk are low, the administration of these drugs to the mother of a nursing infant is unlikely to produce deleterious effects in the infant. GLUCOCORTICOID THERAPY AND AGE The clearance of prednisolone and methylprednisolone decreases with age.17,18 Despite the higher prednisolone levels seen in elderly subjects compared with young subjects after comparable doses, endogenous plasma cortisol levels are suppressed to a lesser extent in the elderly.17 These findings may be associated with an increased incidence of side effects and suggest the need to use smaller doses in the elderly than in young patients. DRUG INTERACTIONS The concomitant use of medications can alter the effectiveness of glucocorticoids; the reverse also is true.19 EFFECTS OF OTHER MEDICATIONS ON GLUCOCORTICOIDS The metabolism of glucocorticoids is accelerated by substances that induce hepatic microsomal enzyme activity, such as phenytoin, barbiturates, and rifampin. The administration of these medications can increase the corticosteroid requirements of patients with adrenal insufficiency or lead to deterioration in the conditions of patients whose underlying disorders are well controlled by glucocorticoid therapy. These substances should be avoided in patients receiving corticosteroids. Diazepam does not alter the metabolism of glucocorticoids and is preferable to barbiturates in this setting. If drugs that induce hepatic microsomal enzyme activity must be used

in patients taking corticosteroids, an increase in the required dose of corticosteroids should be anticipated. Conversely, ketoconazole increases the bioavailability of large doses of prednisolone (0.8 mg/kg) because of inhibition of hepatic microsomal enzyme activity.20 Oral contraceptive use decreases the clearance of prednisone and increases its bioavailability.21 The bioavailability of prednisone is decreased by antacids in doses comparable to those used clinically.22 The bioavailability of prednisolone is not impaired by sucralfate, H2-receptor blockade, or cholestyramine. EFFECTS OF GLUCOCORTICOIDS ON OTHER MEDICATIONS The concurrent administration of a glucocorticoid and a salicylate may reduce the serum salicylate level. Conversely, reduction of the corticosteroid dose during the administration of a fixed dose of salicylate may lead to a higher and possibly toxic serum salicylate level. This interaction may reflect the induction of salicylate metabolism by glucocorticoids.23 Glucocorticoids may increase the required dose of insulin or oral hypoglycemic agents, antihypertensive drugs, or glaucoma medications. They also may alter the required dose of sedative-hypnotic or antidepressant therapy. Digitalis toxicity can result from hypokalemia caused by glucocorticoids, as from hypo-kalemia of any cause. Glucocorticoids can reverse the neuro-muscular blockade induced by pancuronium. CONSIDERATIONS BEFORE INITIATING THE USE OF GLUCOCORTICOIDS AS PHARMACOLOGIC AGENTS Cushing syndrome (see Chap. 75) is a life-threatening disorder. The 5-year mortality was higher than 50% at the beginning of the era of glucocorticoid and ACTH therapy.24 Infection and cardiovascular complications were frequent causes of death. High-dose exogenous glucocorticoid therapy is similarly hazardous. Table 78-2 summarizes the important questions to consider before initiating glucocorticoid therapy.25 These questions enable the physician to assess the potential risks that must be weighed against the possible benefits of treatment. The more severe the underlying disorder, the more readily can systemic glucocorticoid therapy be justified. Thus, corticosteroids are commonly used in patients with severe forms of systemic lupus erythematosus, sarcoidosis, active vasculitis, asthma, chronic active hepatitis, transplantation rejection, pemphigus, or diseases of comparable severity. Generally, systemic corticosteroids should not be administered to patients with mild rheumatoid arthritis or mild bronchial asthma; such patients should receive more conservative therapy first. Although these patients may experience symptomatic relief from glucocorticoids, it may prove difficult to withdraw the drugs. Consequently, they may unnecessarily experience Cushing syndrome and HPA suppression.

TABLE 78-2. Considerations before the Use of Glucocorticoids as Pharmacologic Agents

DURATION OF THERAPY The anticipated duration of glucocorticoid therapy is another critical issue. The use of glucocorticoids for 1 to 2 weeks for a condition such as poison ivy or allergic rhinitis is unlikely to be associated with serious side effects in the absence of a contraindication. An exception to this rule is a corticosteroid-induced psychosis. This complication may occur after only a few days of high-dose glucocorticoid therapy, even in patients with no previous history of psychiatric disease (see Chap. 201).26,27 Because the risk of so many complications is related to the dose and duration of therapy, the smallest possible dose should be prescribed for the shortest possible period. If hypoalbuminemia is present, the dose should be reduced. If long-term treatment is indicated, the use of an alternate-day schedule should be considered. LOCAL USE A local corticosteroid preparation should be used whenever possible because systemic effects are minimal when these substances are administered correctly. Examples include topical therapy in dermatologic disorders, corticosteroid aerosols in bronchial asthma and allergic rhinitis, and corticosteroid enemas in ulcerative proctitis. Systemic absorption of inhaled glucocorticoids leading to Cushing syndrome and HPA suppression is a rare occurrence when these agents are administered correctly at prescribed doses.28,29 The intraarticular injection of corticosteroids may be of value in carefully selected patients if strict aseptic techniques are used and if frequent injections are avoided. SELECTING A SYSTEMIC PREPARATION Agents with little or no mineralocorticoid activity should be used when a glucocorticoid is prescribed for pharmacologic purposes. If the dosage is to be tapered over a few days, a long-acting agent may be impractical. For alternate-day therapy, a short-acting agent that generally does not cause sodium retention (e.g., prednisone, prednisolone, or methylprednisolone) should be used. There is no indication for glucocorticoid conjugates designed to achieve a prolonged duration of action (several days or several weeks) after a single intramuscular injection. The bioavailability of such preparations cannot be regulated precisely, the duration of action cannot be estimated reliably, and it is not possible to taper the dosage rapidly in the event of an adverse reaction such as a corticosteroid-induced psychosis. The use of such preparations may cause HPA suppression more frequently than do comparable doses of the same glucocorticoid given orally. The use of supplemental medications to minimize the systemic corticosteroid dose and to reduce the side effects of systemic glucocorticoids should always be considered. In asthma, for example, treatment should include inhaled glucocorticoids and bronchodilators, such as b-adrenergic agonists and theophylline, and may include cromolyn. EFFECTS OF EXOGENOUS GLUCOCORTICOIDS ANTIINFLAMMATORY AND IMMUNOSUPPRESSIVE EFFECTS Endogenous glucocorticoids protect the organism from damage caused by its own defense reactions and the products of these reactions during stress.30,30a Consequently, the use of glucocorticoids as antiinflammatory and immunosuppressive agents represents an application of the physiologic effects of glucocorticoids to the treatment of disease.30 Glucocorticoids have many effects on inflammatory and immune responses, which are described in this section. Glucocorticoids inhibit synthesis of almost all known cytokines and of several cell surface molecules required for immune function.31,32 and 33 When an immune stimulus such as tumor necrosis factor binds to its receptor, nuclear factor kappa B (NF-kB) moves to the nucleus, where it activates many immunoregulatory genes. This activation of NF-kB involves the degradation of its cytoplasmic inhibitor IkBb and the translocation of NF-kB to the nucleus. Glucocorticoids are potent inhibitors of NF-kB activation. This inhibition is mediated by the induction of the IkBb inhibitory protein, which traps activated NF-kB in inactive cytoplasmic complexes.31,32 and 33 This reduction in NF-kB activity appears to explain the ability of glucocorticoids to inhibit the production of cytokines and cell surface molecules and to suppress the immune response. Influence on Blood Cells and on the Microvasculature. Glucocorticoid effects on inflammatory and immune phenomena include effects on leukocyte movement, leukocyte function, and humoral factors (Table 78-3). In general, glucocorticoids have a greater effect on leukocyte traffic than on function, and more effect on cellular

than on humoral processes.34,35 glucocorticoids alter the traffic of all the major leukocyte populations in the circulation (see Chap. 212).

TABLE 78-3. Effects of Glucocorticoids on Inflammatory and Immune Responses in Humans

Probably the most important antiinflammatory effect of glucocorticoids is the ability to inhibit the recruitment of neutrophils and monocyte-macrophages to an inflammatory site.35 Cortico-steroids modify the increased capillary and membrane permeability that occurs in an area of inflammation. By decreasing the dilation of the microvasculature and the increased capillary permeability that occur during an inflammatory response, the exudation of fluid and the formation of edema may be reduced, and the migration of leukocytes may be impaired.2,35,36 The decrease in the accumulation of inflammatory cells is also related to decreased adherence of inflammatory cells to the vascular endothelium. It is not possible to determine the relative contributions of the direct vascular effect, the effect on inflammatory cell adherence to the vascular wall, and the effect on chemotaxis to the reduction in inflammation caused by glucocorticoids. Glucocorticoids have multiple effects on leukocyte function.35 Corticosteroids suppress cutaneous delayed hypersensitivity responses. Monocyte-macrophage traffic and function are sensitive to glucocorticoids (see Table 78-3). Glucocorticoids in divided daily doses depress the bactericidal activity of monocytes. The sensitivity of monocytes to glucocorticoids may explain the effectiveness of these agents in many granulomatous diseases because the monocyte is the principal cell involved in granuloma formation.35 Although neutrophil traffic is sensitive to glucocorticoids, neutrophil function appears to be relatively resistant to these agents.35 Whereas most in vivo studies of neutrophil phagocytosis have found no evidence for impairment of phagocytosis or bacterial killing,35 other studies suggest that glucocorticoids induce a generalized phagocytic defect affecting both granulocytes and monocytes. Glucocorticoid therapy retards the disappearance of sensitized erythrocytes, platelets, and artificial particles from the circulation.35 This may account for the efficacy of glucocorticoids in the treatment of idiopathic thrombocytopenic purpura and autoimmune hemolytic anemia. Influence on Arachidonic Acid Derivatives. Glucocorticoids inhibit prostaglandin (PG) and leukotriene synthesis by inhibiting the release of arachidonic acid from phospholipids.37 This inhibition of arachidonic acid release appears to be mediated by the induction of lipocortins, a family of related proteins that inhibit phospholipase A2, which is an enzyme that liberates arachidonic acid from phospholipids (see Chap. 172).38,39 This mechanism is distinct from the mechanism of action of the nonsteroidal antiinflammatory agents, such as salicylates and indomethacin, which inhibit the cyclooxygenase that converts arachidonic acid to the cyclic endoperoxide intermediates in the PG synthetic pathway; in some tissues, glucocorticoids inhibit cyclooxygenase activity. Thus, the glucocorticoids and the nonsteroidal antiinflammatory agents exert their antiinflammatory effects at two distinct but adjacent loci in the synthetic pathway of arachidonic acid metabolism. Glucocorticoids and nonsteroidal antiinflammatory agents have different spectra of antiinflammatory effects. Some of the therapeutic effects of corticosteroids that are not produced by the non-steroidal agents may be related to the inhibition of leukotriene formation.37 SIDE EFFECTS The side effects of glucocorticoids include the diverse manifestations of Cushing syndrome and HPA suppression (Table 78-4).40 Iatrogenic Cushing syndrome differs from endogenous Cushing syndrome in several respects: hypertension, acne, menstrual disturbances, male erectile dysfunction, hirsutism or virilism, striae, purpura, and plethora are more common in endogenous Cushing syndrome; benign intracranial hypertension, glaucoma, posterior subcapsular cataract, pancreatitis, and aseptic necrosis of bone are virtually unique to iatrogenic Cushing syndrome; and obesity, psychiatric symptoms, and poor wound healing have nearly equal frequency in both.40,41 These differences may be explained as follows. When Cushing syndrome is caused by exogenous glucocorticoids, ACTH secretion is suppressed. In spontaneous, ACTH-dependent Cushing syndrome, the elevated ACTH output causes bilateral adrenal hyperplasia. In the former circumstance, the secretion of adrenocortical androgens and mineralocorticoids is not increased. Conversely, when ACTH output is elevated, the secretion of adrenal androgens and mineralocorticoids may be increased.1 The augmented secretion of adrenal androgens may account for the higher prevalence of virilism, acne, and menstrual irregularities in the endogenous form of Cushing syndrome, and the enhanced production of mineralocorticoids may explain the higher prevalence of hypertension.1

TABLE 78-4. Adverse Reactions to Glucocorticoids

Some of the complications that are virtually unique to iatrogenic Cushing syndrome arise after the prolonged use of large doses of glucocorticoids. Examples are benign intracranial hypertension, posterior subcapsular cataract, and aseptic necrosis of bone.1 Although the association of glucocorticoid therapy and peptic ulcer disease is controversial,42,43,44,45,46 and 47 glucocorticoids appear to increase the risk of peptic ulcer disease and also gastrointestinal hemorrhage (see Chap. 204).45,46 The magnitude of the association between glucocorticoid therapy and these complications is small and is related to the total dose and duration of therapy.42,45 The risk of peptic ulcer disease and related gastrointestinal problems is increased by the concurrent use of glucocorticoids and nonsteroidal antiinflammatory drugs.48,49 Glucocorticoid therapy, especially daily therapy, may suppress the immune response to skin tests for tuberculosis. When possible, tuberculin skin testing is advisable before the initiation of glucocorticoid therapy. Routine isoniazid prophylaxis probably is not indicated for corticosteroid-treated patients, even for those with positive tuberculin skin test results.50 At similar doses, some patients respond to and experience side effects of glucocorticoids more readily than do others. Variations in responsiveness to glucocorticoids may be a consequence of drug interactions or of variations in the severity of the underlying disease. Alterations in bioavailability probably do not account for variations in the therapeutic response to glucocorticoids. In patients who experience side effects, the metabolic clearance rate of prednisolone and the volume of distribution are lower10,51 than in those who do and the circulating T1/2 is longer51 not experience side effects. Impaired renal function may contribute to a decrease in the clearance of prednisolone and an increase in the prevalence of cushingoid features.52 Patients who have a cushingoid habitus while taking prednisone have higher endogenous

plasma cortisol levels than do those without this complication, perhaps because of resistance of the HPA axis to suppression by exogenous glucocorticoids.53 Variations in the effectiveness of corticosteroids may be the result of altered cellular responsiveness to the drugs.54,55,56 and 57 In patients with primary open-angle glaucoma, exogenous glucocorticoids produce a more pronounced rise of intraocular pressure54; a greater suppression of the 8:00 a.m. plasma cortisol level when dexamethasone, 0.25 mg, is administered the previous evening at 11:00 p.m.56; and greater suppression of phytohemagglutinin-induced lymphocyte transformation55,57 than in normal persons. Primary open-angle glaucoma is relatively common. These findings suggest that a distinct subpopulation of patients are hyperresponsive to glucocorticoids and that this sensitivity is genetically determined (see Chap. 215). PREVENTION OF SIDE EFFECTS Increasingly, the issues of concern to physicians and patients with respect to glucocorticoid therapy are not only HPA suppression but long-term complications such as glucocorticoid-induced osteoporosis and Pneumocystis carinii pneumonia. Of course, the risk of many complications can be reduced by the use of the lowest possible dose of a glucocorticoid for the shortest possible period, by the use of regional or topical rather than systemic steroids, and by the use of alternate-day corticosteroid therapy. In addition, pharmacologic interventions to prevent specific complications such as bone disease and P. carinii pneumonia are now widely used. Osteoporosis. The majority of patients who receive long-term glucocorticoid therapy will develop low bone mineral density. By some estimates, more than one-fourth of these patients will sustain osteoporotic fractures.58 The prevalence of vertebral fractures in asthmatic patients on glucocorticoid therapy for at least a year is 11%.58 Patients with rheumatoid arthritis who are treated with glucocorticoids have an increased incidence of fractures of the hips, ribs, spine, legs, ankles, and feet.58 Skeletal wasting occurs most rapidly during the first year of therapy. Trabecular bone is affected more than cortical bone. The effects on the skeleton are related to the cumulative dose and duration of treatment. 58 Alternate-day glucocorticoid therapy does not reduce the risk of osteopenia. Inhaled steroids have been associated with bone loss. The pathogenesis of glucocorticoid-induced osteoporosis involves several different mechanisms.58 Glucocorticoids decrease intestinal absorption of calcium and phosphate by vitamin D-independent mechanisms. Urinary calcium excretion is increased, possibly as a result of direct effects on renal tubular calcium reabsorption. These changes may lead to secondary hyperparathyroidism in at least some patients. Glucocorticoids reduce sex hormone production. This may be a direct effect by decreasing gonadal hormone release. It may also be indirect by reducing ACTH secretion and adrenal androgen production. Also, inhibition of luteinizing hormone secretion can result in decreased estrogen and testosterone production by the gonads. Glucocorticoids also have an inhibitory effect on the proliferation of osteoblasts, attachment of osteoblasts to matrix, and the synthesis of type I collagen and noncollagenous proteins by osteoblasts. The evaluation of a patient should emphasize medical risk factors for osteoporosis, including inadequate dietary calcium and vitamin D intake, alcohol consumption, smoking, menopause, and any history of infertility or impotence suggesting hypogonadism in males. Attention should also be devoted to the possible presence of thyrotoxicosis, overtreatment with thyroid medication, renal osteodystrophy, multiple myeloma, osteomalacia, or primary hyperparathyroidism. When appropriate, laboratory studies should be ordered for evaluation of these disorders. When glucocorticoid therapy will be administered for more than a few months, it is reasonable to obtain a baseline measurement of bone mineral density using dual energy x-ray absorptiometry. In general, all patients should receive calcium and vitamin D supplementation to correct any nutritional deficiency. Calcium therapy alone is associated with rapid rates of spinal bone loss and offers only partial protection from this loss. There is no evidence that the combination of calcium and vitamin D completely prevents bone loss caused by glucocorticoids.59 Calcitriol and bisphosphonates, specifically alendronate and etidronate, are effective in the prevention of bone loss.60,61,62 and 63 If calcitriol is used, careful follow-up determinations of serum levels is necessary. Hypogonadotropic men should receive testosterone therapy; hormone replacement therapy should be considered for postmenopausal women. Patients should be educated about the risks and the consequences of osteoporosis and the factors in their own lives that may contribute. Because glucocorticoids also affect muscle mass and function, patients should be advised about exercises for maintaining muscle strength. Pneumocystis carinii Pneumonia. Glucocorticoids predispose patients to infections of many varieties. Until recently, pro-phylaxis against infections for patients treated with glucocorticoids was limited to patients receiving transplantation of organs, who also receive other forms of immunosuppression. Currently, prophylaxis for patients with other disorders who are treated with glucocorticoids is being used, particularly for P. carinii pneumonia.64,65 In a series of 116 patients without acquired immunodeficiency syndrome (AIDS) who experienced a first episode of P. carinii pneumonia between 1985 and 1991, 105 (90.5%) had received glucocorticoids within 1 month before the diagnosis of P. carinii pneumonia was established.64 The median daily dose was equivalent to 30 mg prednisone; 25% of the patients had received as little as 16 mg daily. The median duration of glucocorticoid therapy was 12 weeks before the development of the pneumonia. In 25% of the patients, P. carinii pneumonia developed after 8 weeks or less of glucocorticoid therapy. However, the attack rate in patients with primary or metastatic central nervous system tumors who received glucocorticoid therapy was 1.3% and may be lower in other conditions.65 Also, prophylactic therapy may produce side effects. Some physicians recommend prophylaxis (e.g., with trimethoprim-sulfa, one double-strength tablet a day) for patients with impaired immune competence conferred by chemotherapy, transplantation, or an inflammatory disorder who have received prednisone 20 mg or more per day for more than 1 month. No controlled studies with such prophylaxis in steroid-treated patients are available. Among patients undergoing bone marrow or organ transplantation at the Mayo Clinic from 1989 to 1995, no cases of P. carinii pneumonia were detected in those who received adequate chemoprophylaxis. WITHDRAWAL FROM GLUCOCORTICOIDS The symptoms associated with glucocorticoid withdrawal include anorexia, myalgia, nausea, emesis, lethargy, headache, fever, desquamation, arthralgia, weight loss, and postural hypotension. Many of these symptoms can occur with normal plasma glucocorticoid levels and in patients with normal responsiveness to conventional tests of the HPA system. 66,67 These patients may have abnormal responses to a more sensitive test using 1 g of a-1-24 ACTH rather than the conventional 250-g dose.68,69 Because glucocorticoids inhibit PG production and because many of the features of the corticosteroid withdrawal syndrome can be produced by PGs such as PGE2 and PGI2, this syndrome may be caused by a sudden increase in PG production after the withdrawal of exogenous corticosteroids. The corticosteroid withdrawal syndrome may contribute to psychologic dependence on glucocorticoid treatment and to difficulties in withdrawing such therapy. SUPPRESSION OF THE HYPOTHALAMIC-PITUITARY-ADRENAL SYSTEM DEVELOPMENT OF HYPOTHALAMIC-PITUITARY-ADRENAL SUPPRESSION Few well-documented cases of acute adrenocortical insufficiency have been reported after prolonged glucocorticoid therapy and none have been reported after ACTH therapy.1 After the introduction of ACTH and glucocorticoids into clinical practice in the late 1940s, patients were described in whom shock was attributed to adrenocortical insufficiency induced by these agents, but biochemical evidence of adrenocortical insufficiency was not available to substantiate the diagnosis.1 Prolonged hypotension, or even an apparent response of hypotension to intravenous hydrocortisone, is not a reliable means of assessing adrenocortical function. It must be demonstrated simultaneously that the plasma cortisol level is lower than the values found in normal persons experiencing a comparable degree of stress. When testing for plasma cortisol levels became available in the early 1960s, three cases were described in which these criteria were met. The paucity of reports may reflect the fact that acute adrenocortical insufficiency after glucocorticoid therapy is uncommon in properly treated patients, and that physicians may be reluctant to report such events. The minimal duration of glucocorticoid therapy that can produce HPA suppression must be ascertained from studies of adrenocortical weight and adrenocortical responsiveness to provocative tests.1,2 Any patient who has received a glucocorticoid in dosages equivalent to 20 to 30 mg per day of prednisone for more than 5 days should be suspected of having HPA suppression.1,2 If the dosages are closer to but above the physiologic range, 1 month is probably the minimal interval.1,2 The stress of general anesthesia and surgery is not hazardous to patients who have received only replacement doses (no more than 25 mg hydrocortisone, 5 mg prednisone, 4 mg triamcinolone, or 0.75 mg dexamethasone), provided the corticosteroid is given early in the day. If doses of this size are given late in the day, suppression may occur as a result of inhibition of the diurnal surge of ACTH release. ASSESSMENT OF HYPOTHALAMIC-PITUITARY-ADRENAL FUNCTION When HPA suppression is suspected, the physician may wish to assess the integrity of the HPA system. A test of HPA reserve is indicated only when the result will modify therapy. In practice, this applies to patients who may need an increase in the cortico-steroid dosage to cover a stressful event (such as general anesthesia and surgery) and to patients in whom withdrawal of glucocorticoid therapy is contemplated. In the latter group, a test of the HPA axis usually is indicated only when the glucocorticoid dosage has been reduced to replacement levels, for example, 5 mg prednisone daily (or an equivalent dosage of another glucocorticoid). In stable

patients receiving prolonged glucocorticoid therapeutic regimens, frequent tests of HPA reserve function are not indicated. For example, it is not necessary to test before each reduction in dosage during tapering of the steroid regimen. The responsiveness of the HPA system may change as corticosteroid therapy continues, and repeated testing is costly. The short ACTH test is a useful guide to the presence or absence of HPA suppression in patients treated with glucocorticoids (Table 78-5). Although this test assesses directly only the adrenocortical response to ACTH, it is an effective measure of the integrity of the HPA axis. Because hypothalamic-pituitary function returns before adrenocortical function during recovery from HPA suppression, a normal adrenocortical response to ACTH in this setting implies that hypothalamic-pituitary function also is normal. This rationale is supported by direct observation. Thus, the maximal response of the plasma cortisol level to ACTH corresponds to the maximal plasma cortisol level observed during the induction of general anesthesia and surgery in patients who have received glucocorticoid therapy.1,2 A normal response to ACTH before surgery is unlikely to be followed by markedly impaired secretion of cortisol during anesthesia and surgery in corticosteroid-treated patients. An abnormal response to ACTH is a necessary but not a sufficient condition for the diagnosis of adrenal insufficiency in glucocorticoid-treated patients who undergo surgery; some patients with an abnormal response to ACTH tolerate surgery without glucocorticoid treatment.70 Moreover, hypotension in the operative or postoperative period in patients who have been treated previously with glucocorticoid therapy is often a result of other causes, such as volume depletion and reactions to anesthetic medication. The hypotension often responds to treatment of these factors.

TABLE 78-5. Assessment of Hypothalamic-Pituitary-Adrenal (HPA) Function in Patients Treated with Glucocorticoids

Other tests of HPA function generally are not indicated. The low-dose (1 g) short ACTH test is more sensitive than the conventional-dose ACTH test in patients who have been treated with glucocorticoids.68,68a The conventional dose of ACTH used in the short ACTH test (and other ACTH tests) produces circulating ACTH levels that are far above the physiologic range. These supraphysiologic levels may produce a normal plasma cortisol level in patients with partial adrenocortical insufficiency. Nevertheless, the low-dose short ACTH test has not yet replaced the conventional-dose short ACTH test. The lower limit of the normal range for the low-dose ACTH test has not yet been defined.69 Also, there are no commercially available preparations of ACTH available for direct use in the low-dose short ACTH test. The injection for the low-dose short ACTH test must be prepared by dilution, which is a source of inconvenience and possible error. Insulin-induced hypoglycemia may be hazardous (especially in patients with cardiac or neurologic disease), and the symptoms may be uncomfortable. This procedure is more time-consuming and more costly than the ACTH test because more cortisol values must be determined. The measurement of plasma cortisol levels before and after the administration of corticotropin-releasing hormone also has been recommended.71 This test also is longer and more expensive than the ACTH test and has not been compared to a physiologic stress such as anesthesia and surgery. It offers no clear advantage over the ACTH test. CORTICOTROPIN AND THE HYPOTHALAMIC-PITUITARY-ADRENAL SYSTEM Pharmacologic doses of ACTH cause elevated cortisol secretory rates and increased plasma cortisol levels. The elevated plasma cortisol levels might be expected to suppress ACTH release. Actually, there is no evidence of clinically significant hypothalamic-pituitary suppression in patients who have received ACTH therapy.1 The failure of ACTH to suppress HPA function is not explained by the dose of ACTH used, the frequency of injection, the time of administration, or the plasma cortisol pattern after ACTH administration. Alternatively, it is possible that the hyperplastic and overactive adrenal cortex that results from ACTH therapy compensates for hypothalamic or pituitary suppression. Although threshold adrenocortical sensitivity to ACTH is not changed in patients who have received daily ACTH therapy, there may be altered adrenocortical responsiveness to ACTH in the physiologic range. Moreover, the normal response of the plasma cortisol level in patients treated with ACTH may be preserved, at least in part, because ACTH treatment reduces the rate of ACTH secretion but not the total amount secreted, whereas glucocorticoids reduce both the rate of secretion and the total amount secreted.72 RECOVERY FROM HYPOTHALAMIC-PITUITARY-ADRENAL SUPPRESSION During the recovery from HPA suppression, hypothalamic-pituitary function returns before adrenocortical function.1,2,73 Twelve months must elapse after the withdrawal of large doses of glucocorticoids given for a prolonged period before HPA function, including responsiveness to stress, returns to normal.1,2,73 Conversely, recovery from HPA suppression that has been induced by a brief course of corticosteroids (i.e., 25 mg prednisone twice daily for 5 days) occurs within 5 days.74 Patients with mild suppression of the HPA axis (i.e., normal basal plasma and urine corticosteroid levels but diminished responses to ACTH and insulin-induced hypoglycemia) resume normal HPA function more rapidly than do those with severe depression of the HPA axis (i.e., low basal plasma and urine corticosteroid levels and diminished responses to ACTH and insulin-induced hypoglycemia). The time course of recovery correlates with the total duration of previous glucocorticoid therapy and the total previous corticosteroid dose. Nevertheless, in an individual patient, it is not possible to predict the duration of recovery from a course of glucocorticoid therapy at supraphysiologic doses lasting more than a few weeks. Consequently, persistence of HPA suppression should be suspected for 12 months after such treatment. The recovery interval after suppression of the contralateral adrenal cortex by the products of an adrenocortical tumor may exceed 12 months. The recovery from HPA suppression that is induced by exogenous glucocorticoids may be more rapid in children than in adults. WITHDRAWAL OF PATIENTS FROM GLUCOCORTICOID THERAPY RISKS OF WITHDRAWAL The decision to discontinue glucocorticoid therapy provokes apprehension among physicians. The deleterious consequences of such an action include precipitation of adrenocortical insufficiency, development of the corticosteroid withdrawal syndrome, or exacerbation of the underlying disease. Adrenocortical insufficiency after the withdrawal of glucocorticoids is justly feared. The likelihood of precipitating the underlying disease depends on the activity and natural history of the illness in question. When there is any possibility that the underlying illness will flare up, the glucocorticoid should be withdrawn gradually, over an interval of weeks to months, with frequent reassessment of the patient. TREATMENT OF PATIENTS WITH HYPOTHALAMIC-PITUITARY-ADRENAL SUPPRESSION No proven means exists for hastening a return to normal HPA function once inhibition has resulted from glucocorticoid therapy. The use of ACTH does not prevent or reverse the development of glucocorticoid-induced adrenal insufficiency. Conversion to an alternate-day schedule permits but does not accelerate recovery. In children, alternate-day glucocorticoid therapy actually may delay recovery. The recovery from corticosteroid-induced adrenal insufficiency is time dependent and spontaneous. The rate of recovery is determined not only by the doses given when the corticosteroids are being tapered, but also by the doses administered during the initial phase of treatment, before tapering is commenced. During the course of recovery, small doses of hydrocortisone (1020 mg) or prednisone (2.55.0 mg) given in the morning may alleviate the withdrawal symptoms. Recovery of HPA function still occurs when small doses of glucocorticoids are administered in the morning. The possibility cannot be excluded, however, that small doses of glucocorticoids given in the morning retard the rate of recovery from HPA suppression. ALTERNATE-DAY GLUCOCORTICOID THERAPY Alternate-day glucocorticoid therapy is defined as the administration of a short-acting glucocorticoid with no appreciable mineralocorticoid effect (i.e., prednisone, prednisolone, or methylprednisolone) once every 48 hours in the morning, at about 8:00 a.m. The purpose of this approach is to minimize the adverse effects of glucocorticoids while retaining the therapeutic benefits. The original basis for this schedule was the hypothesis that the antiinflammatory effects of glucocorticoids

persist longer than do the undesirable metabolic effects.75,76 and 77 This hypothesis is not supported by observations of the duration of corticosteroid effects. A second hypothesis emphasizes that intermittent rather than continuous administration produces a cyclic, although not diurnal, pattern of glucocorticoid levels in the circulation and within the target cells that simulates the normal diurnal cycle.34 This may prevent the development of Cushing syndrome and HPA suppression while providing therapeutic benefit. Because the full expression of a disease frequently occurs only when the level of inflammatory activity is elevated over a protracted period, the intermittent administration of a glucocorticoid may be sufficient to shorten the interval during which the disorder develops without interruption and thereby to prevent the level of disease activity from becoming apparent clinically (Fig. 78-2).34 The duration of action of the glucocorticoid is important here. The selection of prednisone, prednisolone, and methylprednisolone as the agents of choice for alternate-day therapy and of 48 hours as the appropriate interval between doses has an empiric basis. It has been found that intervals of 36, 24, and 12 hours were accompanied by adrenal suppression, and that an interval of 72 hours was therapeutically ineffective when prednisone (and, occasionally, triamcinolone) was used.77 An interval of 48 hours is optimal.

FIGURE 78-2. The effect of glucocorticoid administration on the activity of the underlying disease. A divided daily dosage schedule may be necessary initially in some disorders. When the disease is controlled, or from the start of therapy in certain diseases, alternate-day therapy may be effective. (From Fauci AS, Dale DC, Balow JE. Glucocorticosteroid therapy: mechanisms of action and clinical considerations. Ann Intern Med 1976; 84:304.)

ALTERNATE-DAY GLUCOCORTICOID THERAPY AND MANIFESTATIONS OF CUSHING SYNDROME An alternate-day regimen can prevent or ameliorate the manifestations of Cushing syndrome.1,2 The susceptibility to infections that characterizes Cushing syndrome may be alleviated. Patients have been described in whom refractory infections appeared to clear after conversion from daily to alternate-day regimens. In addition, there is a low frequency of infections in patients receiving alternate-day therapy. Children treated with alternate-day steroid therapy regain or retain tonsillar and peripheral lymphoid tissue. The available information strongly suggests that alternate-day regimens are associated with a lower incidence of infections than are daily regimens, but it does not firmly establish this point. Host defense mechanisms have been studied in patients receiving alternate-day therapy. Patients maintained on such schedules who have been studied on the days they do not take the medication have normal blood neutrophil and monocyte counts, normal cutaneous inflammatory responses, and normal neutrophil T1/2s. Patients receiving daily therapy, however, demonstrate neutrophilia, monocytopenia, decreased cutaneous neutrophil and monocyte inflammatory responses, and pro longation of the neutrophil T1/2. Patients studied on the days they do not receive treatment do not have the lymphocytopenia observed in patients who receive daily therapy. Monocyte cellular function is normal in patients receiving alternate-day treatment at 4 hours and at 24 hours after a dose. Intermittently normal leukocyte kinetics, preservation of delayed hypersensitivity, and preservation of monocyte cellular function may explain the apparently reduced susceptibility to infection of patients receiving alternate-day therapy.78,79 and 80 EFFECTS OF ALTERNATE-DAY GLUCOCORTICOID THERAPY ON HYPOTHALAMIC-PITUITARY-ADRENAL RESPONSIVENESS Patients receiving alternate-day glucocorticoid therapy may have some suppression of basal corticosteroid levels, but they have normal or nearly normal responsiveness to provocative tests such as the corticotropin-releasing hormone stimulation test, the ACTH stimulation test, insulin-induced hypoglycemia, and the metyrapone test.1,2,81 They have less suppression of HPA function than do patients receiving daily therapy. EFFECTS OF ALTERNATE-DAY THERAPY ON THE UNDERLYING DISEASE Alternate-day glucocorticoid therapy is as effective, or nearly as effective, in controlling diverse disorders as daily therapy in divided doses.1,2 This approach has provided apparent benefit in patients with the following disorders: childhood nephrotic syndrome, adult nephrotic syndrome, membranous nephropathy, renal transplantation, mesangiocapillary glomerulonephritis, lupus nephritis, ulcerative colitis, rheumatoid arthritis, acute rheumatic fever, myasthenia gravis, Duchenne muscular dystrophy, dermatomyositis, idiopathic polyneuropathy, asthma, Sjgren syndrome, sarcoidosis, alopecia areata and other chronic dermatoses, and pemphigus vulgaris. Prospective, controlled studies demonstrate the efficacy of alternate-day therapy in membranous nephropathy and renal transplantation. The role of alternate-day therapy in giant cell arteritis is controversial.82,83 and 84 USE OF ALTERNATE-DAY THERAPY Because alternate-day therapy can prevent or ameliorate the manifestations of Cushing syndrome, can avert or permit recovery from HPA suppression, and is as effective (or nearly as effective) as continuous therapy, patients for whom long-term glucocorticoid administration is indicated should be placed on such programs whenever possible. Nevertheless, physicians sometimes are reluctant to use alternate-day schedules, often because of an unsuccessful experience. Many efforts fail because of lack of familiarity with the indications for and use of such therapy. The benefits of alternate-day glucocorticoid therapy are demonstrable only when corticosteroids are used for a prolonged period. There is no reason to use an alternate-day schedule when the anticipated duration of therapy is less than several weeks. Alternate-day therapy may not be necessary or appropriate during the initial stages of therapy or during exacerbation of the underlying disease. Nevertheless, patients with many chronic disorders have been treated with an alternate-day regimen as initial therapy with apparent benefit.1,2 In patients with rheumatoid arthritis, it appears to be easier to establish treatment with alternate-day corticosteroids than to convert from daily therapy. Physicians treating recipients of renal transplants initially use daily therapy and then convert to an alternate-day schedule. Alternate-day therapy may be hazardous in the presence of adrenocortical insufficiency of any cause because patients are unprotected against glucocorticoid insufficiency during the last 12 hours of the 48-hour cycle. In patients who have been taking glucocorticoids for more than a brief period, or in those who may have adrenal insufficiency on another basis, the adequacy of HPA function should be determined before the initiation of an alternate-day program. It may be possible to surmount this obstacle by giving a small dose of a short-acting glucocorticoid (i.e., 10 mg hydrocortisone) in the afternoon of the second day; this approach has not been studied systematically. Alternate-day glucocorticoid therapy may fail to prevent or ameliorate the manifestations of Cushing syndrome or HPA suppression if a short-acting glucocorticoid is not used, or if it is used incorrectly. For example, the use of prednisone four times a day on alternate days may be less successful than the use of the same total dose once every 48 hours. An abrupt alteration from daily to alternate-day therapy should be avoided. First, the prolonged use of daily-dose glucocorticoids may have caused HPA suppression. In addition, patients with normal HPA function may experience withdrawal symptoms and have an exacerbation of the underlying disease. No program of conversion from continuous therapy to alternate-day therapy has been shown to be optimal. One approach is to reduce the frequency of drug administration until the total dose for each day is given in the morning, and then to increase the dosage gradually on the first day of each 2-day period and to decrease the dosage on the second day. Another approach is to double the dosage on the first day of each 2-day cycle, to give this as a single morning dose if possible, and then to taper the dosage gradually on the second day.85 It is not clear how often changes in dosage should be made with any approach. This depends on many variables, including the underlying disease involved, the duration of previous glucocorticoid therapy, the personality of the patient, and the physician's ability to use

adjunctive therapy. Nonetheless, the conversion should be made as quickly as the patient can tolerate it. If adrenal insufficiency, the corticosteroid withdrawal syndrome, or an exacerbation of the underlying disease develops, the previously effective regimen should be reinstituted and then tapered more gradually. Occasionally, it is necessary to resume full daily dosages temporarily. An absolute change of dosage represents a larger percentage change in dosage at small total daily doses than at large total daily doses. Changes in the dosage should be about 10 mg prednisone (or equivalent) at total daily doses of more than 30 mg, 5 mg at total doses of more than 20 mg, and 2.5 mg at lower doses. The interval between changes in the dosage may be as short as 1 day or as long as many weeks. Optimal results from alternate-day glucocorticoid therapy may not be achieved because of failure to use supplemental therapy for the underlying disorder. Conservative (nonglucocorticoid) therapy often is used until a glucocorticoid is initiated, at which time these less toxic therapeutic measures are ignored. Adjunctive therapeutic measures may facilitate the use of the lowest possible corticosteroid dose. With alternate-day therapy, these measures especially should be used during the end of the second day, when symptoms may be prominent. Supplemental therapy may be especially helpful in disorders in which patients are likely to experience symptoms of the disease on the day off therapy, such as asthma and rheumatoid arthritis. In illnesses in which disabling symptoms are less likely to appear on the alternate day, such as the childhood nephrotic syndrome, less difficulty may be encountered. Alternate-day therapy may fail because of failure to inform patients about the purposes of this regimen. Because glucocor-ticoids may induce euphoria, patients may be reluctant to accept modification of a schedule of frequent doses. A careful explanation about the risks of glucocorticoid excess, attuned to patients' intellectual and emotional ability to comprehend, enhances the prospects of success. DAILY SINGLE-DOSE GLUCOCORTICOID THERAPY Sometimes, alternate-day therapy fails because patients experience symptoms of the underlying disease during the last few hours of the second day. In these situations, single-dose glucocorticoid therapy may be of value.1,2 This regimen appears to be as effective as divided daily doses in controlling such underlying diseases as rheumatoid arthritis, systemic lupus erythematosus, polyarteritis, and proctocolitis. In giant cell arteritis, a daily dose in the morning is nearly as effective as daily therapy in divided doses.82 Daily single-dose therapy reduces the likelihood that HPA suppression will develop. The manifestations of Cushing syndrome, however, probably are not prevented or ameliorated by a daily single-dose regimen. GLUCOCORTICOIDS OR CORTICOTROPIN? Disorders that respond to glucocorticoid therapy also respond to ACTH therapy if the adrenal cortex is normal. There is no evidence, however, that ACTH is superior to glucocorticoids for the treatment of any disorder when comparable doses are used.1,2,86 Hydrocortisone and ACTH, given intravenously in pharmacologically equivalent doses (determined by plasma cortisol levels and urinary corticosteroid excretion rates), are equally effective in the treatment of inflammatory bowel disease.87 Similarly, there is no apparent difference in the effectiveness of prednisone and ACTH for the treatment of infantile spasms.88 Because ACTH does not appear to offer any therapeutic advantage, glucocorticoids are preferable for therapeutic purposes: they can be administered orally, the dose can be regulated precisely, their effectiveness does not depend on adreno-cortical responsiveness (an important consideration in patients who have been treated with glucocorticoids), and they produce a lower frequency of certain side effects such as acne, hypertension, and increased pigmentation.1,2 If alternate-day therapy cannot be used, ACTH might appear to be preferable because it does not suppress the HPA axis. This benefit usually is outweighed by the advantages of glucocorticoids and by the fact that daily injections of ACTH are not superior to single daily doses of short-acting glucocorticoids; in both cases, HPA suppression is unlikely to result, but Cushing syndrome is not prevented. In life-threatening situations, glucocorticoids are indicated because maximal blood levels are obtained immediately after intravenous administration, whereas with ACTH infusion, the plasma cortisol level rises to a plateau over several hours. The principal indication for ACTH continues to be the assessment of adrenocortical function. DOSAGE ANTIINFLAMMATORY OR IMMUNOSUPPRESSIVE THERAPY The glucocorticoid dosage required for antiinflammatory or immunosuppressive therapy is variable, and depends on the disease under treatment. In general, the dosage ranges from just above that needed for long-term replacement therapy up to 60 to 80 mg prednisone or its equivalent daily. Although much larger dosages sometimes are recommended for diseases such as asthma, systemic lupus erythematosus, and cerebral edema, controlled studies have not established the need for such large amounts of medication. The role of massive doses of cortico-steroids in asthma is controversial.89,90 Most studies report no advantage of high-dose therapy (e.g., more than 6080 mg prednisone per day). Many physicians use intravenous pulse therapy (e.g., 1 g per day of methylprednisolone intravenously for 3 consecutive days) for severe manifestations of systemic lupus erythematosus, rapidly progressive glomerulonephritis, or other entities. There are no controlled studies that compare the results of pulse therapy with 60 to 80 mg per day of prednisone, however. Thus, the superiority of pulse therapy has not been demonstrated.91,92 When alternate-day therapy is used, the dosage is variable and depends on the disease under treatment. It may range from just above that needed for long-term replacement therapy to 150 mg prednisone every other day. PERIOPERATIVE MANAGEMENT Traditional doses of glucocorticoids recommended for periop-erative coverage in patients treated with steroids (e.g., 100 mg hydrocortisone intravenously every 8 hours or 20 mg methyl-prednisolone intravenously every 8 hours on the day of surgery, with a gradual taper over subsequent days) are arbitrary and have no empirical basis.70 A study in cynomolgus monkeys explored the doses required to prevent postoperative hypotension.93 Bilateral adrenalectomies were performed in the experimental animals, and replacement doses of glucocorticoids were given for 4 months. The animals were then divided into three groups, given normal, one-tenth normal, or 10 times the normal replacement doses of glucocorticoids. A cholecystectomy was performed on each animal under these conditions. The animals that received one-tenth normal replacement doses had an increased mortality rate, decreased peripheral vascular resistance, and hypotension. The group that received normal replacement doses of glucocorticoids had no more hypotension or postoperative complications than did the group receiving 10 times the replacement dose. A double-blind study in patients provided similar results.94 The investigators studied patients who had taken at least 7.5 mg prednisone a day for several months and had an abnormal response to an ACTH test. All patients received their usual daily dose of prednisone on the day of surgery. One group of 12 patients received perioperative injections of saline. The other group of 6 patients received hydrocortisone in the saline. There was no significant difference in outcome between the groups in this small study. It appears that patients with secondary adrenal insufficiency resulting from glucocorticoid therapy do not experience hypotension or tachycardia when given only their usual daily dose of steroids for surgical procedures such as joint replacements and abdominal operations. Based on an analysis of the literature, an interdisciplinary group suggests the use of variable doses, depending on the magnitude of the surgical stress.70 For minor surgical stress(e.g., an inguinal herniorrhaphy), the glucocorticoid target dose would be 25 mg hydrocortisone or equivalent. For moderate surgical stress (e.g., a lower extremity revascularization or total joint replacement), the target would be 50 to 75 mg hydrocortisone or equivalent. This might constitute continuation of the patient's usual dose of prednisone (i.e., 10 mg a day) and 50 mg hydro-cortisone intravenously intraoperatively. For major surgical stress(e.g., esophagogastrectomy or cardiopulmonary bypass), the patient might receive his or her usual steroid dose (e.g., 40 mg prednisone or the parental equivalent preoperatively within 2 hours of surgery) and 50 mg hydrocortisone intravenously every 8 hours after the initial dose for the first 48 to 72 hours. Corticosteroid therapy should not be tapered inadvertently to a dosage below that known to control the underlying disease. In patients with primary adrenocortical insufficiency, hydro-cortisone has the advantage of having mineralocorticoid as well as glucocorticoid activity at high dosages. At dosages less than 100 mg per day, it is necessary to use a mineralocorticoid agent in addition to hydrocortisone; fludrocortisone can be used when patients can take oral medications. Parenteral mineralocorticoid preparations such as desoxycorticosterone acetate are no longer available commercially in the United States. DRUG INTERACTIONS If patients also must take a hepatic microsomal enzyme inducer, the metabolism of the glucocorticoid will be accelerated and larger daily dosages may be needed. The treatment of adrenocortical insufficiency is considered in Chapter 76.

MINERALOCORTICOIDS
PHARMACOLOGY Mineralocorticoids are 21-carbon steroids characterized by their effects on fluid and electrolyte balance. They promote renal sodium reabsorption and potassium excretion (see Chap. 79). Mineralocorticoid deficiency (see Chap. 81) causes hyponatremia, volume depletion, hypotension, hyperkalemia, and a hyperchloremic metabolic acidosis. Mineralocorticoid excess (see Chap. 80) is associated with the retention of sodium and water, hypertension, potassium depletion, hypokalemia, and

a metabolic alkalosis. The excessive secretion or administration of a mineralocorticoid causes sodium retention, with consequent fluid retention and weight gain. Patients retain several hundred milliequivalents of sodium and gain several kilograms of weight. If mineralocorticoid excess persists, mineralocorticoid escape occurs; further sodium retention and weight gain do not occur. During this escape phenomenon, urinary sodium excretion increases until patients come into balance, and urinary sodium excretion again reflects sodium intake. Thus, patients with mineralocorticoid excess often do not retain sufficient fluid for peripheral edema to develop (although it occasionally does) unless there is another cause for edema such as hypoalbuminemia or right ventricular failure. Therefore, the absence of edema does not exclude the possibility of mineralocorticoid excess. Aldosterone is the principal mineralocorticoid in humans. Desoxycorticosterone, corticosterone, and cortisol (hydrocortisone) also are secreted in amounts sufficient to cause salt retention in certain pathologic situations. AGENTS USED CLINICALLY The agents with mineralocorticoid action that are used clinically are hydrocortisone and fludrocortisone (9a-fluorohydro-cortisone, Florinef [Apothecon, a Bristol-Meyers Squibb Co., Princeton, NJ]; Fig. 78-3). When hydrocortisone is given in large dosages (e.g., 100 mg per day or more), a mineralocorticoid effect may be anticipated. Aldosterone is not used clinically, although it is a potent mineralocorticoid and is essentially devoid of glucocorticoid effect. It would be of limited value because of its brief duration of action.

FIGURE 78-3. Structure of the mineralocorticoid fludrocortisone. The arrow indicates the structural difference between this synthetic steroid and cortisol.

Fludrocortisone is available only for oral therapy. The presence of the fluorine atom in the 9a position enhances the mineralocorticoid potency of hydrocortisone. The enhanced mineralocorticoid potency of 9a-fluorohydrocortisone is explained by impaired renal conversion of this molecule to 9a-fluorocortisone by 11-hydroxy-steroid dehydrogenase, in contrast to the rapid conversion of cortisol to cortisone.95 At recommended dosages, fludrocortisone is an effective mineralocorticoid, but it is essentially free of glucocorticoid activity. Its duration of action is ~12 to 24 hours.96,97,98 and 99 If patients also need large dosages of a glucocorticoid, hydrocortisone can be used alone. A dosage of 100 mg per day or more provides mineralocorticoid activity.96,97 and 98 Experimentally, it was found that fludrocortisone, 1 mg per day orally, is equivalent in mineralocorticoid activity to aldosterone, 1 mg per day in four divided intramuscular doses.98,99 No mineralocorticoid by itself can increase the sodium stores of sodium-depleted patients. The effectiveness of a mineralocorticoid hormone depends on substrate availability; patients with mineralocorticoid deficiency not only need the hormone, they also need salt and water. Both hormonal therapy and proper fluid and electrolyte administration are necessary to achieve an optimal clinical response. DRUG INTERACTIONS Mineralocorticoid activity is antagonized by spironolactone; the latter has no effect in patients with hypoaldosteronism. Amiloride (Midamor), a potassium-sparing diuretic agent that acts even in the absence of aldosterone, can reduce the effects of a mineralocorticoid on sodium and potassium balance. Anything that promotes salt loss, such as a diuretic medication, impairs mineralocorticoid efficacy. For patients who ordinarily need a diuretic (e.g., for the treatment of hypertension or fluid retention), the desired effect may be achieved by modifying the dosage of mineralocorticoid, the intake of salt, or both. Medications that alter sweating or promote vomiting or diarrhea may affect salt balance and therefore change the effectiveness of a mineralocorticoid. INDICATIONS Mineralocorticoid therapy is indicated for primary adrenocortical insufficiency; isolated hypoaldosteronism; salt-losing forms of congenital adrenal hyperplasia; and chronic orthostatic hypotension caused by autonomic insufficiency (multiple systems atrophy [e.g., the Shy-Drager syndrome], idiopathic orthostatic hypotension, and diabetic autonomic neuropathy). DOSAGE The usual dosage of fludrocortisone is 0.1 mg daily, but may range from 0.1 mg on alternate days to 0.2 mg daily. Generally, the starting dosage is 0.1 mg daily, with adjustments made according to clinical response. Orthostatic vital signs are of great value in assessing the adequacy of mineralocorticoid replacement therapy. A marked rise in the heart rate or a fall in blood pressure on standing may precede other manifestations of mineralocorticoid deficiency. CHAPTER REFERENCES
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. Axelrod L. Glucocorticoid therapy. Medicine (Baltimore) 1976; 55:39. Axelrod L. Glucocorticoids. In: Kelley WN, Harris ED Jr, Ruddy S, Sledge CB, eds. Textbook of rheumatology, 4th ed. Philadelphia: WB Saunders, 1993. Hollander JL, Brown EM Jr, Jessar RA, Brown CY. Hydrocortisone and cortisone injected into arthritic joints. JAMA 1951; 147:1629. Robinson RCV, Robinson HM Jr. Topical treatment of dermatoses with steroids. South Med J 1956; 49:260. Harter JG. Corticosteroids: their physiologic use in allergic disease. NY State J Med 1966; 66:827. Ballard PL, Carter JP, Graham BS, Baxter JD. A radioreceptor assay for evaluation of the plasma glucocorticoid activity of natural and synthetic steroids in man. J Clin Endocrinol Metab 1975; 41:290. Meikle AW, Tyler FH. Potency and duration of action of glucocorticoids: effects of hydrocortisone, prednisone and dexamethasone on human pituitary-adrenal function. Am J Med 1977; 63:200. Pickup ME. Clinical pharmacokinetics of prednisone and prednisolone. Clin Pharmacokinet 1979; 4:111. Jenkins JS, Sampson PA. Conversion of cortisone to cortisol and prednisone to prednisolone. BMJ 1967; 2:205. Gambertoglio JG, Amend WJC Jr, Benet LZ. Pharmacokinetics and bio-availability of prednisone and prednisolone in healthy volunteers and patients: a review. J Pharmacokinet Biopharm 1980; 8:1. Davis M, Williams R, Chakraborty J, et al. Prednisone or prednisolone for the treatment of chronic active hepatitis? A comparison of plasma availability. Br J Clin Pharmacol 1978; 5:501. Lewis GP, Jusko WJ, Burke CW, et al. Prednisone side effects and serum-protein levels, a collaborative study. Lancet 1971; 2:778. Frey FJ, Frey BM. Altered prednisolone kinetics in patients with the nephrotic syndrome. Nephron 1982; 32:45. Gatti G, Perucca E, Frigo GM, et al. Pharmacokinetics of prednisone and its metabolite prednisolone in children with nephrotic syndrome during the active phase and in remission. Br J Clin Pharmacol 1984; 17:423. Frey FJ, Horber FF, Frey BM. Altered metabolism and decreased efficacy of prednisolone and prednisone in patients with hyperthyroidism. Clin Pharmacol Ther 1988; 44:510. Schatz M, Patterson R, Zeitz S, et al. Corticosteroid therapy for the pregnant asthmatic patient. JAMA 1975; 233:804. Stuck AE, Frey BM, Frey FJ. Kinetics of prednisolone and endogenous cortisol suppression in the elderly. Clin Pharmacol Ther 1988; 43:354. Tornatore KM, Logue G, Venuto RC, Davis PJ. Pharmacokinetics of methyl-prednisolone in elderly and young healthy males. J Am Geriatr Soc 1994; 42:1118. Jubiz W, Meikle AW. Alterations of glucocorticoid actions by other drugs and disease states. Drugs 1979; 18:113. Zrcher RM, Frey BM, Frey FJ. Impact of ketoconazole on the metabolism of prednisolone. Clin Pharmacol Ther 1989; 45:366. Legler UF, Benet LZ. Marked alterations in dose-dependent prednisolone kinetics in women taking oral contraceptives. Clin Pharmacol Ther 1986; 39:425. Uribe M, Casian C, Rojas S, et al. Decreased bioavailability of prednisone due to antacids in patients with chronic active liver disease and in healthy volunteers. Gastroenterology 1981; 80:661. Graham GG, Champion GD, Day RO, Paull PD. Patterns of plasma concentrations and urinary excretion of salicylate in rheumatoid arthritis. Clin Pharmacol Ther 1977; 22:410. Plotz CM, Knowlton AI, Ragan C. The natural history of Cushing's syndrome. Am J Med 1952; 13:597. Thorn GW. Clinical considerations in the use of corticosteroids. N Engl J Med 1966; 274:775. Boston Collaborative Drug Surveillance Program. Acute adverse reactions to prednisone in relation to dosage. Clin Pharmacol Ther 1972; 13:694.

27. 28. 29. 30.

Perry PJ, Tsuang MT, Hwang MH. Prednisolone psychosis: clinical observations. Drug Intell Clin Pharm 1984; 18:603. Hollman GA, Allen DB. Overt glucocorticoid excess due to inhaled corticosteroid therapy. Pediatrics 1988; 81:452. Stead RJ, Cooke NJ. Adverse effects of inhaled corticosteroids. BMJ 1989; 298:403. Munck A, Guyre PM. Glucocorticoid physiology and homeostasis in relation to anti-inflammatory actions. In: Schleimer RP, Claman HN, Oronsky A, eds. Anti-inflammatory steroid action: basic and clinical aspects. San Diego: Academic Press, 1989.

30a. Sapolsky RM, Romero LM, Munck AU. How do glucocorticoids influence stress responses? Integrating permissive, suppressive, stimulatory, and preparative actions. Endocr Rev 2000; 27:55. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. Scheinman RI, Cogswell PC, Lofquist AK, Baldwin AS Jr. Role of transcriptional activation of IkBa in mediation of immunosuppression by glucocorticoids. Science 1995; 270:283. Auphan N, DiDonato JA, Rosette C, et al. Immunosuppression by glucocorticoids: inhibition of NF-kB activity through induction of IkB synthesis. Science 1995; 270:286. Marx J. How the glucocorticoids suppress immunity. Science 1995; 270:232. Fauci AS, Dale DC, Balow JE. Glucocorticosteroid therapy: mechanisms of action and clinical considerations. Ann Intern Med 1976; 84:304. Parrillo JE, Fauci AS. Mechanisms of glucocorticoid action on immune processes. Annu Rev Pharmacol Toxicol 1979; 19:179. Cupps TR, Fauci AS. Corticosteroid-mediated immunoregulation in man. Immunol Rev 1982; 65:133. Samuelsson B. Leukotrienes: mediators of immediate hypersensitivity reactions and inflammation. Science 1983; 220:568. DiRosa M, Flower RJ, Hirata F, et al. Anti-phospholipase proteins. Prostaglandins 1984; 28:441. Parente L, DiRosa M, Flower RJ, et al. Relationship between the anti-phospholipase and anti-inflammatory effects of glucocorticoid-induced proteins. Eur J Pharmacol 1984; 99:233. Axelrod L. Side effects of glucocorticoid therapy. In: Schleimer RP, Claman H, Oronsky A, eds. Antiinflammatory steroid action: basic and clinical aspects. San Diego: Academic Press, 1989. Ragan C. Corticotropin, cortisone and related steroids in clinical medicine: practical considerations. Bull NY Acad Med 1953; 29:355. Conn HO, Blitzer BL. Nonassociation of adrenocorticosteroid therapy and peptic ulcer. N Engl J Med 1976; 294:473. Langman MJS, Cooke AR. Gastric and duodenal ulcer and their associated diseases. Lancet 1976; 1:680. Jick H, Porter J. Drug-induced gastrointestinal bleeding. Lancet 1978; 2:87. Messer J, Reitman D, Sacks HS, et al. Association of adrenocorticosteroid therapy and pepticulcer disease. N Engl J Med 1983; 309:21. Spiro HM. Is the steroid ulcer a myth? N Engl J Med 1983; 309:45. Conn HO, Poynard T. Adrenocorticosteroid administration and peptic ulcer: a critical analysis. J Chronic Dis 1985; 38:457. Piper JM, Ray WA, Daugherty JR, Griffin MR. Corticosteroid use and peptic ulcer disease: role of nonsteroidal anti-inflammatory drugs. Ann Intern Med 1991; 114:735. Gabriel SE, Jaakkimainen L, Bombardier C. Risk for serious gastrointestinal complications related to use of nonsteroidal anti-inflammatory drugs. A meta-analysis. Ann Intern Med 1991; 115:787. Schatz M, Patterson R, Kloner R, Falk J. The prevalence of tuberculosis and positive tuberculin skin tests in a steroid-treated asthmatic population. Ann Intern Med 1976; 84:261. Kozower M, Veatch L, Kaplan MM. Decreased clearance of prednisolone, a factor in the development of corticosteroid side effects. J Clin Endocrinol Metab 1974; 38:407. Bergrem H, Jervell J, Flatmark A. Prednisolone pharmacokinetics in cushingoid and non-cushingoid kidney transplant patients. Kidney Int 1985; 27:459. Frey FJ, Amend WJC Jr, Lozada F, et al. Endogenous hydrocortisone, a possible factor contributing to the genesis of cushingoid habitus in patients on prednisone. J Clin Endocrinol Metab 1981; 53:1076. Becker B. Intraocular pressure response to topical corticosteroids. Invest Ophthalmol 1965; 4:198. Bigger JF, Palmberg PF, Becker B. Increased cellular sensitivity to glucocorticoids in primary open angle glaucoma. Invest Ophthalmol 1972; 11:832. Becker B, Podos SM, Asseff CF, Cooper DG. Plasma cortisol suppression in glaucoma. Am J Ophthalmol 1973; 75:73. Becker B, Shin DH, Palmberg PF, Waltman SR. HLA antigens and corticosteroid response. Science 1976; 194:1427. American College of Rheumatology Task Force on Osteoporosis Guidelines. Recommendations for the prevention and treatment of glucocorticoid-induced osteoporosis. Arthritis Rheum 1996; 39:1791. Sambrook PN. Calcium and vitamin D therapy in corticosteroid bone loss: what is the evidence? J Rheumatol 1996; 23:963. Sambrook P, Birmingham J, Kelly P, et al. Prevention of corticosteroid osteoporosis. A comparison of calcium, calcitriol, and calcitonin. N Engl J Med 1993; 328:1747. Adachi JD, Bensen WG, Brown J, et al. Intermittent etidronate therapy to prevent corticosteroid-induced osteoporosis. N Engl J Med 1997; 337:382. Reid IR. Preventing glucocorticoid-induced osteoporosis. N Engl J Med 1997; 337:420. Saag KG, Emkey R, Schnitzer TJ, et al. Alendronate for the prevention and treatment of glucocorticoid-induced osteoporosis. N Engl J Med 1998; 339:292. Yale SH, Limper AH. Pneumocystis carinii pneumonia in patients without acquired immunodeficiency syndrome: associated illnesses and prior corticosteroid therapy. Mayo Clin Proc 1996; 71:5. Sepkowitz KA. Pneumocystis carinii pneumonia without acquired immuno-deficiency syndrome: who should receive prophylaxis? Mayo Clin Proc 1996:71:102. Amatruda TT Jr, Hollingsworth DR, D'Esopo ND, et al. A study of the mechanism of the steroid withdrawal syndrome: evidence for integrity of the hypo-thalamic-pituitary-adrenal system. J Clin Endocrinol Metab 1960; 20:339. Amatruda TT Jr, Hurst MM, D'Esopo ND. Certain endocrine and metabolic facets of the steroid withdrawal syndrome. J Clin Endocrinol Metab 1965; 25:1207. Dickstein G, Shechner C, Nicholson WE, et al. Adrenocorticotropin stimulation test: effects of basal cortisol level, time of day, and suggested new sensitive low-dose test. J Clin Endocrinol Metab 1991; 72:773.

68a. Henzen C, Suter A, Lerch E, et al. Suppression and recovery of adrenal response after short-term, high-dose glucocorticoid treatment. Lancet 2000; 355:542. 69. Streeten DHP. Shortcomings in the low-dose (1 g) ACTH test for the diagnosis of ACTH deficiency states. J Clin Endocrinol Metab 1999; 84:835. 70. Salem M, Tainsh RE Jr, Bromberg J, et al. Perioperative glucocorticoid coverage. A reassessment 42 years after emergence of a problem. Ann Surg 1994; 219:416. 71. Schlaghecke R, Kornely E, Santen RT, Ridderskamp P. The effect of long-term glucocorticoid therapy on pituitary-adrenal responses to exogenous corticotropin-releasing hormone. N Engl J Med 1992; 326:226. 72. Daly JR, Fletcher MR, Glass D, et al. Comparison of effects of long-term corticotrophin and corticosteroid treatment on responses of plasma growth hormone, ACTH, and corticosteroid to hypoglycaemia. BMJ 1974; 2:521. 73. Graber AL, Ney RL, Nicholson WE, et al. Natural history of pituitary-adrenal recovery following long-term suppression with corticosteroids. J Clin Endocrinol Metab 1965; 25:11. 74. Streck WF, Lockwood DH. Pituitary adrenal recovery following short-term suppression with corticosteroids. Am J Med 1979; 66:910. 75. Haugen HN, Reddy WJ, Harter JG. Intermittent steroid therapy in bronchial asthma. Nord Med 1960; 63:15. 76. Reichling GH, Kligman AM. Alternate-day corticosteroid therapy. Arch Dermatol 1961; 83:980. 77. Harter JG, Reddy WJ, Thorn GW. Studies on an intermittent corticosteroid dosage regimen. N Engl J Med 1963; 269:591. 78. MacGregor RR, Sheagren JN, Lipsett MB, Wolff SM. Alternate-day prednisone therapy: evaluation of delayed hypersensitivity responses, control of disease and steroid side effects. N Engl J Med 1969; 280:1427. 79. Dale DC, Fauci AS, Wolff SM. Alternate-day prednisone: leukocyte kinetics and susceptibility to infections. N Engl J Med 1974; 291:1154. 80. Fauci AS, Dale DC. Alternate-day therapy and human lymphocyte sub-populations. J Clin Invest 1975; 55:22. 81. Schrmeyer TH, Tsokos GC, Avgerinos PC, et al. Pituitary-adrenal responsiveness to corticotropin-releasing hormone in patients receiving chronic, alternate day glucocorticoid therapy. J Clin Endocrinol Metab 1985; 61:22. 82. Hunder GG, Sheps SG, Allen GL, Joyce JW. Daily and alternate day corticosteroid regimens in treatment of giant cell arteritis: comparison in a prospective study. Ann Intern Med 1975; 82:613. 83. Abruzzo JL. Alternate-day prednisone therapy. Ann Intern Med 1975; 82:714. 84. Bengtsson B-A, Malmvall B-E. An alternate-day corticosteroid regimen in maintenance therapy of giant cell arteritis. Acta Med Scand 1981; 209:347. 85. Fauci AS. Alternate-day corticosteroid therapy. Am J Med 1978; 64:729. 86. Allander E. ACTH or corticosteroids? A critical review of results and possibilities in the treatment of severe chronic disease. Acta Rheum Scand 1969; 15:277. 87. Kaplan HP, Portnoy B, Binder HJ, et al. A controlled evaluation of intravenous adrenocorticotropic hormone and hydrocortisone in the treatment of acute colitis. Gastroenterology 1975; 69:91. 88. Hrachovy RA, Frost JD Jr, Kellaway P, Zion TE. Double-blind study of ACTH vs prednisone therapy in infantile spasms. J Pediatr 1983; 103:641. 89. Editorial. Steroids in acute severe asthma. Lancet 1992; 340:1384. 90. McFadden ER Jr. Dosages of corticosteroids in asthma. Am Rev Respir Dis 1993; 147:1306. 91. Elenbaas J. Steroid pulse therapy in systemic lupus erythematosus. Drug Intell Clin Pharm 1983; 17:342. 92. Kurki P, ed. High dose intravenous corticosteroid therapy of systemic lupus erythematosus and primary crescenteric rapidly progressive glomerulonephritis. Proceedings of a symposium. Scand J Rheumatol 1984; Suppl 54:1. 93. Udelsman R, Ramp J, Gallucci WT, et al. Adaptation during surgical stress: a reevaluation of the role of glucocorticoids. J Clin Invest 1986; 77:1377. 94. Glowniak JV, Loriaux DL. A double-blind study of perioperative steroid requirements in secondary adrenal insufficiency. Surgery 1997; 121:123. 95. Oelkers W, Buchen S, Diederich S, et al. Impaired renal 11b-oxidation of 9a-fluorocortisol: an explanation for its mineralocorticoid potency. J Clin Endocrinol Metab 1994; 78:928. 96. Goldfein A, Laidlaw JC, Haydar NA, et al. Fluorohydrocortisone and chlorohydrocortisone, highly potent derivatives of compound F. N Engl J Med 1955; 252:415. 97. Renold AE, Haydar NA, Reddy WJ, et al. Biological effects of fluorinated derivatives of hydrocortisone and progesterone in man. Ann NY Acad Sci 1955; 61:582. 98. Thorn GW, Renold AE, Morse WI, et al. Highly potent adrenal cortical steroids; structure and biological activity. Ann Intern Med 1955; 43:979. 99. Thorn GW, Sheppard RH, Morse WI, et al. Comparative action of aldosterone and 9-alpha-fluorohydrocortisone in man. Ann NY Acad Sci 1955; 61:609.

CHAPTER 79 RENIN-ANGIOTENSIN SYSTEM AND ALDOSTERONE Principles and Practice of Endocrinology and Metabolism

CHAPTER 79 RENIN-ANGIOTENSIN SYSTEM AND ALDOSTERONE

DALILA B. CORRY AND MICHAEL L. TUCK The Renin-Angiotensin System Angiotensinogen Renin Synthesis and Release Angiotensin I-Converting Enzyme Angiotensin Metabolism Tissue Renin-Angiotensin System Angiotensin Receptors Actions of Angiotensin II Aldosterone Synthesis Mineralocorticoid Receptors Aldosterone Action Control of Aldosterone Secretion Chapter References

THE RENIN-ANGIOTENSIN SYSTEM

ANGIOTENSINOGEN Angiotensinogen (AGT), also termed renin substrate, is the precursor for the angiotensin peptides, which include angiotensin (A)-I, II, III, and IV, and A 17. Levels of this peptide, which can be rate-limiting for the renin-angiotensin system, are produced mainly in the liver, where its precursor preproangiotensinogen is synthesized and glycosylated in the hepatocytes; nonetheless, there is evidence of production in other tissues as well (e.g., the heart, blood vessels, kidney, and adipocytes). In the circulation, AGT, with a half-life of 16 hours, is cleaved by renin (~10%) and/or other enzymes to release A-I (Fig. 79-1). In many tissues not expressing renin, AGT can be cleaved by enzymes other than renin (e.g., cathepsin G, tonin, and chymase). The extent to which AGT is glycosylated may influence the kinetics of renin in the circulation. Indeed, this has been hypothesized in a proposed, separate, brain-AGT system.

FIGURE 79-1. Angiotensinogen and angiotensin I and II pathways and their role in vascular and fluid homeostasis via angiotensin II receptors (AT 1 and AT 2). (CAGE, chymotrypsin-like angiotensin-generating enzyme.)

There is one copy of the AGT gene in the mammalian genome, which is ~11,800 base pairs (bp) in length.1,2 This gene has 5 exons that encode for the protein, separated by 4 introns. Exon 1 codes for the 5'-nontranslated region, whereas exon 2 contains the signal peptide and coding regions. AGT is secreted constitutively from hepatocytes; however, several factors (e.g., glucocorticoids, estrogens, thyroid hormone, insulin, and A-II) may exert a positive feedback.3,4 RELATIONSHIP TO HYPERTENSION A high molecular-weight form of AGT, which is released during pregnancy, may play a role in pregnancy-induced hypertension. In adipose tissue, there is a form of AGT that is increased by insulin and decreased by 3-adrenergic blockade and that possibly contributes to obesity-related hypertension. Moreover, AGT may play a role in certain other forms of hypertension, as observed in glucocorticoid excess states (e.g., Cushing syndrome) and thyroid disorders.1,2,3 and 4 Interestingly, antisense nucleotide sequences that have been used to block AGT mRNA can reduce blood pressure. In spontaneously hypertensive rats, central nervous system (CNS) administration of antisense sequences against the mRNA encoding AGT lowers blood pressure.5 Furthermore, rats made transgenic with human AGT develop hypertension because AGT is expressed in the blood vessel wall.6,7 Thus, there is ample documentation for a role of AGT in experimental hypertension. Further documentation for a role of AGT in hypertension is as follows: Epidemiologic studies have shown a relationship between plasma AGT and human hypertension.8,9 Polymorphisms of the AGT gene have been linked to familial hypertension, renal disease, and cardiovascular risk factors.9 Increased levels of AGT are associated with essential hypertension, and an M235T polymorphism in the AGT gene is associated with nephropathy in type 2 diabetics.10 The AGT M235T molecular variantthreonine substituted for methionine at amino acid 235is associated by linkage analysis with essential hypertension, especially in whites.11,12 Subjects bearing the 235 allele have higher levels of AGT (i.e., a 20% increase in homozygous subjects [TT] and a 10% increase in heterozygous subjects [MT]) compared with homozygous (MM) individuals. This linkage of AGT variants to hypertension is population-dependent (i.e., strong in whites, weak in Mexican Americans13 and the Chinese14). AGT mutations probably play a significant but modest role in blood pressure variation. All of these findings suggest that the renin-AGT reaction kinetics are increased by AGT variants, leading to more A-II production, which in turn may increase vascular resistance and growth. In turn, vascular injury induces AGT gene expression in the vascular media and neointima, suggesting that the renin-angiotensin system participates in myointimal proliferation.15 Finally, gene knock-out mice for AGT develop hypotension with polyuria when challenged with a high-salt diet.16 RENIN SYNTHESIS AND RELEASE Renin, an aspartyl proteolytic enzyme, catalyzes the rate-limiting cleavage of AGT (between the leucine at position 10 and the valine at position 11) to form the decapeptide A-I, which is further converted by angiotensin-converting enzyme I (ACE-I) to the octapeptide, A-II. It is the plasma renin activity that is accepted as an index of the renin-angiotensin system, because it is difficult to measure other components (e.g., A-II). Renin synthesis begins with the formation of preprorenin in the juxtaglomerular (JG) cells of the kidney.17,18 This precursor is transported into the rough endoplasmic reticulum (Fig. 79-2), whre it is cleaved to form a 23-amino-acid inactive form (prorenin), which is passed through the Golgi apparatus, glycosylated, and deposited in lysosomal granules.18 Here it is cleaved by cathepsin B to form renin, which can be secreted in esponse to various stimuli.18 Basally, there is a low rate of renin release, and this is increased several-fold in response to stimuli.

FIGURE 79-2. Renin synthesis, activation, and release.

Prorenin, which is not regulated by the factors that control renin release (i.e., pressure and volume changes), is released into the circulation at levels several-fold higher than those of bioactive renin.19 Whereas the role of circulating prorenin is unknown, it is found in tissue sites such as the adrenal glands, pituitary, and submandibular glands as well as in the kidney, and in spontaneously hypertensive rats, which are stroke-prone, prorenin may raise blood pressure.20 Several local factors, which are under genetic control, also contribute to further enhance the renin response to salt restriction and to other stimuli. The vascular tissue-derived JG cells are situated in the afferent arteriole near the glomerulus and the macula densa. With salt depletion, there is a recruitment of new cells facilitating a continued renin release.21 Additionally, the expression of the renin gene can be increased by sodium restriction.22 Regulation of renin release is multifactorial (i.e., by renal baroreceptors, JG cells, the macula densa, renal nerves, and humoral factors). The renal baroreceptor is an intrarenal vascular receptor in the afferent arteriole that stimulates renin secretion in response to reduced renal perfusion pressure. It is attenuated when renal perfusion is elevated (Fig. 79-3). This is the most powerful mechanism for renin release; it is exemplified in the Goldblatt models of hypertension, in which a unilateral stenotic lesion of the renal artery causes hypoperfusion. The hypoperfusion induces increases in renin, and, hence, results in a renin-dependent form of hypertension.

FIGURE 79-3. Control of renin release by negative feedback.

The macula densa is a modified group of cells in the distal tubules near the end of the loop of Henle and adjacent to the afferent arteriole and the JG cells. These cells sense distal tubular Na+ delivery through Cl and act through the Na+ ,K+,Cl cotransporter pathway. Diuretics, which stimulate the cotransporter pathway, enhance renin release. Other factors (e.g., adenosine, prostaglandin E2, nitric oxide, and the b1-adrenergic system) also may mediate the macula densa renin-response pathway.17,23 The renal nerves (i.e., the b1-adrenergic system) are stimulated through mechanoreceptors in the heart, pressor receptors in the aorta, and chemoreceptors in the vagal nerve. Central sympathetic stimulation increases renin secretion and is the major acute pathway for stress and posture; thus, upright posture increases plasma renin activity two- to four-fold. Cyclic adenosine monophosphate (cAMP) is the main intra-cellular signal pathway for renin release, which is induced by b1-adrenergic agonists and by prostaglandins.24 Increased intra-cellular calcium provides another signal pathway; renin secretion is suppressed after rises in cytosolic calcium. Electrical depolarization of the JG cells (by increased A-II) opens calcium channels, and the high intracellular calcium inhibits renin release. Hyperpolarization of the JG cells with a resultant decrease of intracellular calcium has the opposite effect.24 Other inhibitors of renin release include adenosine A1, a-adrenergic agonists, thromboxane, and endothelin.24 Atrial natriuretic peptides (ANPs) and nitric oxide inhibit renin release through stimulation of cyclic guanosine monophosphate (cGMP) and guanylate cyclase. Phospholipase C-regulated changes in intra-cellular calcium also affect renin secretion.24 Mice and rats that are transgenic for the renin message TGR(mREN2)27 (a model representing a precisely defined defect for the renin-angiotensin system) demonstrate that the renin gene is an important candidate gene for hypertension.25 Thus, TGR(mREN2)27 rats, harboring the murine REN-2 gene, develop severe hypertension despite low levels of renin. The high expression of the transgene in several tissues should cause high tissue levels of renin and A-II. In this model, the high blood pressure and suppressed plasma renin activity in the face of an increased tissue renin expression suggests that an extrarenal renin-angiotensin system can mediate the hypertension.

ANGIOTENSIN I-CONVERTING ENZYME

Angiotensin I-converting enzyme (ACE) converts the inactive decapeptide angiotensin I to the active octapeptide angiotensin II (Fig. 79-4). The enzyme is not specific for angiotensin because it cleaves bradykinin, luteinizing hormone-releasing hormone, the enkephalins, and substance P.26,27 It is a more efficient kininase than it is a converting enzyme.27 ACE activates the vaso-constrictor A-II and metabolizes the vasodilator bradykinin (see Fig. 79-4). ACE inhibitors block angiotensin formation and bradykinin degradation. The beneficial effects of ACE inhibitors in the heart and kidney may be a result of both A-II reduction and bradykinin accumulation.

FIGURE 79-4. Metabolism of vasoactive peptides by angiotensin-Iconverting enzyme (ACE).

ACE is a metalloenzyme that requires zinc as a cofactor and also needs chloride ions to cleave most substrates.28 Human ACE has a molecular mass of 150 kDa to 180 kDa, of which 146.6 kDa is protein and the remainder is carbohydrate.28 Most ACE is bound to plasma membranes of endothelial cells within vascular tissue. ACE is inserted into the membrane by a 17-amino acid hydrophobic region near the carboxyl terminus (C-terminus hydrophobic anchor peptide).ACE can be released from the plasma membrane by proteolytic cleavage; therefore, some soluble ACE can be detected in blood, urine, edema fluid, amniotic fluid, cerebrospinal fluid, lymph, seminal plasma, and prostate.28,29 The highest concentration of ACE is seen in the blood vessels of the lung, retina, and brain.29 The human kidney also contains a very large amount of this enzyme.28 The brush border of the proximal tubules (where A-II has major functions in electrolyte transport) is rich in ACE. It is also abundant in the choroid plexus, the placenta, and the epithelial lining of the small intestine, and seems to be highly concentrated in some areas of the brain (e.g., the subfornical organ, area postrema, substantia nigra, and locus coeruleus). The primary structure of ACE consists of two active centers that are located within two homologous domains.29,30 Most tissues have both domains, but testicular tissue has only one domain, perhaps representing an early form of the enzyme that has not been duplicated.30 In endothelial cells, ACE contains two functional sites (requiring two zinc ions) and two inhibitor-binding sites. ACE has in vitro activity for a broad range of substrates besides A-I (e.g., N -acetyl-seryl-aspartyl-lysyl-proline [Ac-SDKP], which is a regulatory factor for hematopoiesis).26,27,31 In vivo, plasma levels of Ac-SDKP increase sharply after oral administration of an ACE inhibitor.31 ACE inhibition, which is the treatment of choice for the erythrocytosis that occurs after renal transplantation, lowers the erythrocyte count independently of any effect on erythropoietin. Thus, ACE appears to have a role in the regulation of hematopoiesis. The human ACE gene displays an insertion (I) / deletion (D) polymorphism, which accounts for at least half of the variability in serum ACE levels.32 One such I/D polymorphism, which has been found in the noncoding region, corresponds to the presence or absence of a 287-bp sequence in intron 16.32 Individuals who are homozygous for the insertion polymorphism (II) have lower levels of circulating ACE than do individuals with the deletion (DD) genotype.33 The D allele may be associated with an increased risk for coronary heart disease and diabetic nephropathy. Individuals with diabetic nephropathy who are DD homozygous may respond less well to ACE inhibitors with respect to reducing blood pressure and proteinuria.33 These individuals are also more salt sensitive, such that on a high salt diet, ACE inhibition is less effective. Furthermore, a deletion variant of the ACE D allele is associated with an increased risk of diabetic nephropathy and hastened progression of IgA nephropathy.34 This deletion in ACE does not appear to affect outcome of renal transplants.35 A link between the D allele and plasminogen activator inhibitor-l levels in diabetes and microangiopathy suggests that an adverse consequence of thrombotic mechanisms can be activated with ACE polymorphism.36 Other A-IIforming serine proteinases such as chymase have been identified in the human cardiac left ventricle and in vascular, renal, and other tissues.37 These enzymes serve as alternative routes in the conversion of A-I to A-II.37 These non-ACE enzymes may explain the incomplete blockade of A-II formation during ACE-inhibition therapy.

ANGIOTENSIN METABOLISM
Of the several active angiotensins, A-II is the most important. The enzymes that are most active in degrading angiotensin are the aminopeptidases (Fig. 79-5). Glutamyl aminopeptidase cleaves A-II to form A-III (A28), and arginyl aminopeptidase cleaves A-III to A-IV (A38)38 Although A38 doeees not have major pressor activity, it does increase cGMP levels and may have central effects.38 The amino-terminal heptapeptide, A17 [des-phe8], which is formed directly from A-I by several tissue endopeptidases or by ACE,39 can stimulate the release of arginine vasopressin (AVP), vasodilator prostaglandins, bradykinin, and nitric oxide.39,40 Most angiotensins have a very short duration of action and undergo enzymatic degradation and endocytosis. The AT1 receptor can mediate intracellular transport of A-II and the ligand-receptor can also be internalized. The acute inhibition of ACE decreases levels of A-II, but chronic inhibition therapy may not suppress A-II, although it does raise levels of A17 (a competitive inhibitor of A-II); this phenomenon possibly explains the long-term antihypertensive effects of ACE-I.39,40

FIGURE 79-5. Enzymes that convert angiotensin I into other angiotensin peptide products.

TISSUE RENIN-ANGIOTENSIN SYSTEM

Local renin-angiotensin systems in multiple tissues can produce A-II independently of the circulating renin-angiotensin system (Table 79-1). Their function of these systems is probably to permit regional modulation of blood flow and also to assist in the action of growth factors. Most contractile cells (e.g., vascular smooth muscle, heart cells, mesangial cells, and sperm tails) have a well-defined local renin-angiotensin system. A-II is formed in these cells either by renin or through other proteases such as chymase, cathepsin, or tonin.41,42 and 43

TABLE 79-1. Differences in Circulating and Tissue Renin-Angiotensin System Control and Action

Kidney Renin-Angiotensin System. In addition to being the major site for release of systemic renin, the kidney has a local renin-angiotensin system that is involved in renal growth.44 ACE-I administration during pregnancy causes severe fetal renal abnormalities. The tubular epithelial cells contain ACE, and local A-II is found in proximal tubules and mesangial cells.45 The intrarenal A-II constricts afferent and efferent arterioles and directly increases Na+ reabsorption. Locally, A-II is also involved in renal pathologic conditions (e.g., the development of nephrosclerosis). Cardiac Renin-Angiotensin System. Evidence exists for a role for an intracardiac renin-angiotensin system in both normal and failing hearts.41,42 There is a direct

expression of AGT and the A-II receptor (AT1) in cardiomyocytes. The exact source of the renin in cardiac tissue is controversial, because levels of mRNA are low.43 The heart can also express A-I, which may be converted by intracardiac ACE to A-II.43 Probably, cardiac renin-angiotensin system activity is regulated by the amount of AGT, ACE, or renin-binding protein in the cardiomyocytes. Consequently, local A-II can regulate vascular tone, contractility, growth, and cardiac hypertrophy. Vascular Renin-Angiotensin System. Vascular tissue has an active tissue renin-angiotensin system. Both ACE mRNA and product are found in the endothelium, and also, to a lesser degree, in vascular smooth muscle cells.43 AGT, which is also expressed in cultured vascular smooth muscle cells, is stimulated by insulin and insulin-like growth factor-I (IGF-I).46 The specific origin of the renin in the vasculature is more controversial, because the mRNA is low. Nonetheless, in situ conversion of A-I to A-II does occur in the vasculature; thus, there is an active ACE, which probably has a regulatory role.43 Insulin and IGF-Iinduced vascular growth and hypertrophy are mediated in part by a local renin-angiotensin system and attenuated by ACE-I and AT1-receptor blockers.46 Ovarian Renin-Angiotensin System. The ovarian cells produce high concentrations of both renin and prorenin, and all components of the renin-angiotensin system are found in follicular fluid.47 The granulosa and theca cells also contain renin and angiotensin. Because prorenin is secreted throughout pregnancy, it may have functional significance during this time. A possible role for A-II as a reproductive hormone is being investigated.48 Testicular Renin-Angiotensin System. A-I, -II, and -III are found in the testicular tissue, and AT1 receptors are present in Leydig cells.48,49 The structure of testicular ACE is unique; it has only one of the two binding domains. Perhaps, testicular ACE and A-II participate in spermatogenesis and may play a role in male fertility.49 Adipose Renin-Angiotensin System. Adipose tissue is a rich source for AGT.50 Locally formed A-II from adipose tissue may cause vasoconstriction and hypertension, especially as body fat content increases. Skin Renin-Angiotensin System. The human skin fibroblast expresses A-II and AGT mRNA. Local injury elevates local A-II concentrations, and it then participates in wound healing through growth factors. Adrenal Renin-Angiotensin System. Some of the highest levels of A-II are found in the adrenal cortex, within the adrenal zona glomerulosa and fasciculata.51 Local A-II is thought to play an important role in aldosterone and corticosteroid production. Renin is known to be produced in adrenal tissue and acts independently of plasma renin. Brain Renin-Angiotensin System. The bloodbrain barrier prevents any blood-borne components of the renin-angiotensin system from entering the brain; thus, this organ has a local renin-angiotensin system. The expression of renin mRNA is low in the brain, but some cells (e.g., glial cells) have marked expression of AGT.52 However, the expression and production of ACE is widespread throughout the brain.52 Moreover, AT1receptors are in the hypothalamus and AT2 receptors are in the locus coeruleus and in the inferior olivary region. In addition, receptors for A-IV and A17 are also found in the brain. injection of A-II into the brain causes changes in drinking behavior and elevations of blood pressure. The anterior pituitary gland has high levels of renin and A-II (located mainly in the gonadotropes); thus, the renin-angio-tensin system may contribute to the regulation of estrogen production in females and testosterone production in males. Interestingly, A-II inhibits prolactin release. Importantly, the human brain (especially the pineal gland) is rich in chymase, which can convert A-I to A-II.53

ANGIOTENSIN RECEPTORS
A-II binds and acts on its receptors at the cell surface. Several receptor subtypes have been found, including the AT1, AT2, and AT454, 55 and 56 AT1Receptor Subtype. Almost all actions of A-II are mediated through the AT1 receptor, including its effects on vasoconstriction and aldosterone production and cell growth (Fig. 79-6).The AT1 receptor is part of the superfamily of peptide hormone receptors with seven membrane-spanning regions linked to G proteins (Fig. 79-7). AT1 activates phospholipase C, which in turn hydrolyzes phosphoinositide to inositol triphosphate54 These changes increase the intracel-(IP3) and diacylglycerol. These changes increase the intracellular calcium levels that in turn activate protein kinases. Also, AT1 receptors appear to lower levels of intracellular cAMP, an effect that is most evident in the adrenal gland on the control of aldosterone secretion.54

FIGURE 79-6. Three major pathways of the angiotensin II (AT1) receptor.

FIGURE 79-7. Model of angiotensin II (AT1) receptor. (From Hunyady L, Balla T, Catt KJ. The ligand binding sites of the angiotensin AT1 receptor. Trends Pharmacol Sci 1996; 17:135.)

Although the AT1 receptor contains no intrinsic protein kinase activity, numerous studies indicate that A-II activates both nonreceptor-type and receptor-type tyrosine kinases.56 Thus, A-II binding to the AT1 receptor stimulates a large variety of regulated cellular events, including activation of phospholipases, second messenger generation, and proteinphosphorylation. The enzymes that may be coupled to the AT1 receptor include adenylate cyclase and phospholipases C, D, and A.54 AT2 Receptor Subtype. Table 79-2 contrasts the actions of the A-II AT 1 and AT2 Receptor Subtypes. Attention has been directed to the AT2 receptor subtype, which is found in fetal tissue, where it is responsible for the growth and remodeling of organs.57 It is likely that the actions of the AT2 receptor may antagonize those of the AT1 receptor (Table 79-2). Using the gene transfer approach, it has been shown that the AT2 receptor participates in vascular smooth muscle growth through apoptosis.57

AT2 receptors have been identified in the adult brain, in the adrenal medulla, in the gastrointestinal tract, and in other tissues. AT2 receptors also facilitate intestinal sodium reabsorption. It has been demonstrated that an A-II receptor antagonist can correct endothelial dysfunction in human essential hypertension.

TABLE 79-2. Actions of the Angiotensin AT1 and AT2 Receptors

ACTIONS OF ANGIOTENSIN II
Renal Sodium and Water Retention. The best known function of A-II is to support the circulation under conditions of reduced intravascular volume. However, there are >50 known physiologic actions. These include vasoconstriction, salt retention through aldosterone, increased thirst and secretion of antidiuretic hormone resulting in retention of water, increased cardiac output, and stimulation of sympathetic nervous system activity (Table 79-3). Also, there are direct effects of A-II on early proximal tubule Na+ reabsorption that are mediated in part through the Na+, H+ antiporter; this accounts for 40% to 50% of sodium and water reabsorption.58 In AGT gene-null mutant mice, chronic volume depletion causes a sharp decrease in glomerular filtration rate (GFR) and a marked concentration defect, which is insensitive to AVP.59 Furthermore, the renin-angiotensin system is essential for two fundamental homeostatic functions in the kidney, stabilizing the GFR and maintaining urine diluting and concentrating capacity.

TABLE 79-3. Angiotensin II Actions in Various Target Tissues

Systemic Vasoconstriction. A-II produces arteriolar vaso-constriction and elevates vascular resistance and blood pressure. The signal transduction for this A-IImediated vasoconstriction includes an increase of intracellular calcium and formation of protein kinase C.60 The facilitated release of norepinephrine61 and endothelin-I62 may also contribute to the vasoconstriction that is induced by A-II. A-II stimulates the release of arachidonic acidproducts in the vasculature (e.g., lipoxygenase products and P450 epoxides), which have vasoconstrictive properties.63 Regulation of Gomerular Filtration Rate. A-II regulates the GFR and renal blood flow by constricting the efferent and afferent glomerular arterioles, and the interlobular artery.64 A-II differentially increases efferent resistance to levels that are two to three times those of the afferent resistance; this elevates glomerular capillary pressure to maintain the single nephron GFR. In AGT gene-null mutant mice, chronic volume depletion causes a sharp decrease in GFR and a marked concentration defect, which is insensitive to AVP.59 These physiologic changes are paralleled by changes in renal structure at the glomerular level as well as at the renal papilla and underscore the key role of angiotensins in the development of the mammalian kidney. As mentioned previously, A-II can also stimulate prostaglandins (e.g., thromboxane) that mediate renal hemodynamics.65 A-II controls the GFR by constricting the mesangium and altering tubuloglomerular feedback. In sodium-replete women, A-II infusion elicits a blunted renal microcirculatory response compared with the response in sodium-replete men; this demonstrates that sex steroids modulate the effects of A-II.66 A-II also has nonhemodynamic effects on the kidney (e.g., stimulation of cytokines in diabetic nephropathy).67 Cardiac Effects. Cardiac Effects. The cardiac effects of the tissue angiotensin system are multiple.68 A-II is involved in cardiac remodeling and, through the AT1 receptor, promotes extracellular production of matrix as well as a fibrous tissue response and increasing collagen levels. In addition, the vasoconstriction caused by A-II and oxidative stress further contribute to ischemia and left ventricular hypertrophy with a resultant greater stiffness of the myocardial wall.69 Elevations in plasma renin activity are associated with increased risk for myocardial infarction.70 This risk may be related to renin-angiotensin-systeminduced changes in the fibrinolytic system, because A-II inhibits fibrinolysis (thereby promoting clot formation) and stimulates excess production of plasminogen activator inhibitor.71 Also, the ACE-induced breakdown of bradykinin leads to lower tissue type plasminogen activator levels and decreases nitric oxide concentrations. A low-salt diet increases plasminogen activator inhibitor activity by increasing the renin-angiotensin system and raising the aldosterone levels.72 These findings help to explain why blockade of the renin-angiotensin system by ACE-I or by AT1 receptor antagonists protect against thrombotic cardiac events and reduce the risk of recurrent myocardial infarctions. Brain Effects. There are several actions of the angiotensins in the brain, an organ in which all three receptor subtypes exist.52 In addition, A-II exerts a central pressor activity and causes blunting of the baroreceptor; it also stimulates dipsogenic behavior and the synthesis of AVP. Other Effects. Other actions of A-II include platelet adhesion, glycogenolysis, and inhibition of prolactin release. A-II also has interactions with prostaglandins, nitric oxide, and endothelin. More diverse effects include contraction of the uterus, blockade of the effects of glucagon, stimulation of cate-cholamines, stimulation of endothelial cell growth, stimulation of ovulation, and regulation of steroidogenesis. Moreover, A-II selectively stimulates transforming growth factor-b and stimulates mRNA in the heart and kidneys (thereby contributing to cardiac and renal fibrosis), and it suppresses clusterin, a glyco-protein that is expressed in response to tissue injury.73 Functional analysis of A-II and other components of the renin-angiotensin system can now be achieved by transgenic and gene-targeting technology.74,75 The direct transfection of the human renin gene using gene-transfer techniques in hepatocytes in vivo elevates blood pressure. In vivo transfection of antisense oligodeoxynucleotides for AGT decreases plasma AGT levels and blood pressure in spontaneously hypertensive rats.74 Transfection of antisense to AGT or AT1 receptors decreases blood pressure in spontaneously hypertensive rats.76 Vascular injury is associated with abundant expression of AGT and ACE. This process of vascular hypertrophy can be duplicated by transfecting ACE vector locally into intact rat carotid arteries.74 In vivo transfer of antisense ACE oligodeoxynucleotides inhibits neointimal formation in balloon-injured rat carotid arteries.74 Antisense technology has been used to examine the role of the renin-angiotensin system in the central nervous system (see earlier).76 Conceivably, a gene therapy for hypertension could use an antisense inhibition with adeno-associated viral vector delivery to target the A-II type 1 receptor messenger ribonucleic acid.76 Prolonged reductions in blood pressure in hypertension may be achieved with a single dose of adeno-associated viral vectors to deliver antisense oligodeoxynucleotides to inhibit

AT1 receptors.76

ALDOSTERONE
Aldosterone is the most potent of several mineralocorticoid steroid hormones that act on the epithelium, especially in the kidney, to produce Na+ reabsorption and K+ excretion.77 Other steroids that have weaker mineralocorticoid effects include deoxycorticosterone (DOC), 18-oxycortisol, 19-nor-deoxycorti-costerone (19-nor-DOC), and 18-hydroxydeoxycorticosterone (18-OH-DOC). SYNTHESIS The adrenal gland is partitioned into different zones, which contain specific enzymes that produce either glucocorticoids (i.e., cortisol), mineralocorticoids (i.e., aldosterone), or weak adrenal androgens.78,79 Aldosterone is synthesized in the outer region of the adrenal gland (zona glomerulosa). Cortisol is synthesized in the inner regions in the zona fasciculata and reticularis. The same adrenal enzymes share most of the early synthetic steps of these steroids, but they differ markedly in their terminal enzymes. The zona glomerulosa is unique in lacking 17a-hydroxylase, the essential enzyme for cortisol formation. However, the zona glomerulosa is rich in the enzyme aldosterone synthase, which is required for the terminal steps in aldosterone synthesis.78,79 and 80 Aldosterone synthase is a single multifunctional P450 enzyme on the inner mitochondrial membrane that converts corticosterone to aldosterone. It catalyzes three successive hydroxylation steps: (a) conversion of deoxycorticosterone to corticosterone; (b) addition of a hydroxyl group to corticosterone (18-OH-corticosterone); and (c) conversion of 18-OH-corticosterone to an aldehyde, with consequent formation of aldosterone. Two cytochrome P450 enzymes have been found, including 11b-hydroxylase and aldosterone synthase.78 Their genes are located in close proximity (40 kb) on chromosome 8 (8q21-q22) and display 95% homology. In the familial disorder glucocorticoid-remediable hyperaldosteronism, there is hypertension and variable hypokalemia as a result of the formation of a chimeric gene that makes aldosterone synthesis ACTH-dependent78(see Chap. 80). MINERALOCORTICOID RECEPTORS The mineralocorticoid receptor (steroid receptor type 1 [SR1]) binds both aldosterone and cortisol with equal affinity, whereas the cortisol receptor (steroid receptor type 2 [SR2]) binds only cortisol with high affinity.81,82 Both receptors are part of the steroid/vitamin D/retinoic acid superfamily of transcription regulators. Both receptors have high amino acid homology and may overlap the actions of progesterone and androgen receptors. In humans, there are two isoforms of the SR1; the a and b isoforms.83,84 The SR1s are found mainly on epithelial cells of the renal tubule, the parotid gland, and the colon. SR1s have been found in vascular smooth muscle, in the heart, and in areas of the brain. These receptors can undergo down-regulation, as seen with sodium loading in which the b isoform expression is decreased.84 In contrast, SR2s are found in most cells of the body. Glucocorticoids and mineralocorticoids bind to SR1 with equal affinity, and glucocorticoid levels are 1000-fold higher than those of mineralocorticoids; thus, one would expect the receptor to be saturated with cortisol. However, specificity is conferred by the enzyme 11b-hydroxysteroid dehydrogenase (HSD),78,79,85 which oxidizes cortisol to inactive corticosterone, thus allowing aldosterone to bind to its receptor. Thus, coexpression of 11b-HSD and SR1 in tissues appears to be a mechanism by which specificity is conferred to aldosterone action. There are multiple forms of this enzyme in the mammalian kidney,86 including a high density form in the principal and intercalated cells.87 In the syndrome of apparent mineralocorticoid excess, the renal isoform for 11b-HSD2 is missing, thus allowing the normally high concentrations of cortisol in the kidney to bind to SR1.78 This effect creates an excess Na+ reabsorption with resultant hypertension and hypokalemia (see Chap. 80). The active ingredient of licorice, glycyrrhetinic acid, also inhibits 11b-HSD2 causing an, acquired syndrome of mineralocorticoid excess in humans. However, several questions remain regarding mineral corticoid receptors.87a ALDOSTERONE ACTION The major site of aldosterone action is on the distal nephron, where it increases Na+ reabsorption and K+ and H+ secretion. Aldosterone initiates its action by diffusing across the plasma membrane of the cell and binding avidly to specific receptors, which are located mostly in the cytoplasm.82 In the nucleus, the aldosterone-receptor complex acts on chromatin to increase mRNA and ribosomal RNA transcription.87,88 The binding of steroid to the carboxyl terminal of the receptor causes release of heat shock proteins, after which the receptor changes conformation to allow dimerization and binding to the regulatory end of specific target genes (hormone-responsive elements).82 Mineralocorticoids affect electrolyte balance by inducing and activating proteins in epithelial cells (e.g., aldosterone-induced protein may act as a regulatory unit for the luminal membrane Na+ channel to allow luminal Na+ to enter the cell).89,90 Another aldosterone-induced protein may be a subunit of the Na+, K+ pump. This strengthens the argument that aldosterone not only has an indirect effect through ionic changes but also has a direct action on the Na+, K+ pump.88 The major sites for aldosterone action are in the connecting segment and in the collecting tubules.78 Aldosterone promotes NaCl reabsorption and K+ secretion in the connecting segments and in the principal cells in the cortical collecting tubules. In addition, it may act on Na+ in the papillary (inner medullary) collecting tubule and is thought to increase ionic transport by increasing the number of open Na+ and K+ channels in the luminal membrane.89, 90 Aldosterone also stimulates the Na+, K+ pump in the basolateral membrane. Na+ then diffuses into the tubular cells from which the Na+ is removed by the Na+, K+ pump, thus creating an electrogenic negative potential difference within the lumen. K+ is also secreted from the cell as Clmoves in to maintain electroneutrality. Another mechanism for K+ efflux is through the Na+, K+ pump. It is believed that aldosterone mediates Na+ channel movement by the methylation of channel proteins that, in turn, activate silent or inoperative Na+ channels.90 Because amiloride blocks luminal Na+ permeability, it is likely that increasing luminal permeability is one of the primary actions of aldosterone.91 Intracellular calcium changes may serve as second messenger for this response by increasing the number of open Na+ channels.92 The intercalated cells in the renal cortex and the tubular cells of the outer medullary region seem to be the sites for aldosterone action on H+ secretion.93 These cells do not activate Na+ reabsorption; therefore, aldosterone-mediated H+ secretion and Na+ reabsorption occur in different regions of the kidney. However, there is a modest Na+/H+ exchange process mediated by aldosterone in the principal cells. Although aldosterone acts almost exclusively on the kidney, there are other epithelial sites (e.g., the colon, sweat and salivary glands) where the hormone reduces the Na+ and raises the K+ content of secretions.78 In very end-stage renal failure, the colonic secretion of K+ can become an important route of elimination. Aldosterone may be implicated in the fibrosis of left ventricular hypertrophy and congestive heart failure.94,95,96 and 97 In fact, the heart could be capable of de novo synthesis. The hormone may have effects on vascular remodeling and collagen formation as well as modifying endothelial cell function, and the cardiac angiotensin AT1 receptor may be one of its targets. 94,95,96 and 97 A local or tissue aldosterone, which has been found in vascular smooth muscle, may contribute to A-IIinduced vascular hypertrophy.96 Thus, hypertension, cardiac fibrosis, and hypertrophy by nonepithelial cells may, in part, be consequences of high aldosterone levels, in either the primary or the secondary forms of hyperaldosteronism.95,96 Evidence suggests that many of the effects of aldosterone on blood pressure and the resultant pathologic changes could be mediated through specific miner-alocorticoid receptors in the AV3V region of the brain and in the heart.98 Some of these nonepithelial effects are blocked either by aldosterone-receptor antagonists (e.g., spironolactone) or by selective aldosterone-receptor agonists (e.g., eplerenone). In the Randomized Aldactone Evaluation Study (RALES) trial, the addition of low-dose spironolactone (25 mg) to standard therapy for congestive heart failure further reduced total mortality and/or the hospitalization rate for heart failure by 31%.99 Hyperaldosteronism occurs in renal failure despite normal renin levels. Furthermore, aldosterone appears to contribute to the progression of renal disease by causing fibrosis and glomerular damage, as seen in the rat remnant kidney model.100 More details of aldosterone action may be seen in a review.100a CONTROL OF ALDOSTERONE SECRETION The Renin-Angiotensin System. A-II is the major stimulator of aldosterone secretion. The proximal site of action of A-II is in enhancing the conversion of cholesterol to pregnenolone, but it also can modulate the distal regulatory site influencing the conversion of corticosterone to aldosterone by acting on aldosterone synthase.80,101,102 In general, the renin-angiotensin system is the major mediator of the volume-induced changes in aldosterone. Thus, volume and/or salt depletion will stimulate renin release, producing more A-II to act on the adrenal gland. The resulting Na+ retention then restores the volume to normal and shuts off renin release. The converse effect is seen with volume overload or with a high-salt diet, in which the renin-angiotensin system and aldosterone secretion are suppressed, allowing excess Na+ to be excreted. With volume expansion, there also is release of the ANPs and of dopamine, which, as inhibitors of aldosterone secretion, may act in concert with a low A-II level to limit the release of this mineralocorticoid.

Plasma Potassium Concentration. There is a direct relationship between increases in plasma K+ and increases in aldosterone secretion.103 Studies of K+ infusion in humans show that a 0.1- to 0.2-mEq/L change in plasma K+ can alter aldosterone levels, suggesting that the zona glomerulosa is exquisitely sensitive to K+ .104 In the feedback cycle, an increasing plasma K+ level stimulates aldosterone, which in turn increases K+ secretion and restores it to normal. This mechanism protects against excessive K+ in the body; it represents the single pathway for handling K+ overload. K+ has a direct effect on the zona glomerulosa cells to stimulate aldosterone synthesis, thereby activating the conversion of cholesterol to pregnenolone.105 Although both A-II and K+ act on the zona glomerulosa through different and independent effects, their cumulative action can be synergistic.106 Thus, A-II is more effective in releasing aldosterone when the subject is on a high K+ diet. This effect may be a result of adrenal renin, because in vitro studies show that K+ enhances adrenal renin and adrenal A-II. Both young and elderly subjects have similar basal serum potassium levels and similar responses to a K+ infusion; however, during a K+ infusion, hyperkalemia is more pronounced in elderly subjects because of a blunted adrenal response.107 Corticotropin. As with cortisol release, ACTH, acting through adenylate cyclase, acutely stimulates aldosterone secretion in a dose-dependent manner; however, this effect is not sustained.108 Thus, ACTH is not considered an important chronic regulator of aldosterone. ACTH initially stimulates and then suppresses the aldosterone-synthase gene.108 ACTH also directs steroid biosynthesis through the enzyme 17a-hydroxylase, which diverts synthesis away from the mineralocorticoid pathway and into the glucocorticoid and androgen pathways. Other Factors. The plasma Na+ concentration is a weak regulator of aldosterone; hyponatremia increases aldosterone levels, and hypernatremia possibly has the opposite effect.109 In humans, extreme changes in plasma Na+ are required to alter aldosterone secretion independently from more powerful stimuli (e.g., A-II and K+ ). Hyponatremia usually involves water retention, which (by suppression of A-II) overrides any direct action of Na+ on adrenal gland aldosterone synthesis.110 There is a modest effect of systemic acidosis on adrenal aldosterone release; this increases the excretion of acid and reduces systemic acidemia.111 The effect of aldosterone is thought to be through the H+ -ATPase pumps on the luminal membrane of the collecting tubules. Aldosterone escape occurs when, after several days of administration of this hormone, there is a loss of normal sodium- and fluid-retaining capacity, and a spontaneous diuresis occurs. This escape mechanism, which is caused by increases in ANPs and by a rise in renal perfusion pressure that limits Na+ retention, is observed in primary hyperaldosteronism and perhaps in the inherited forms of mineralocorticoid hypertension, conditions in which edema is not seen, despite a volume retention state.112 CHAPTER REFERENCES
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. Corvol P, Jeunemaitre X. Molecular genetics of human hypertension: role of angiotensinogen. Endocr Rev 1997; 18(5):662. Kim HS, Krege JH, Kluckman KD, et al. Genetic control of blood pressure and the angiotensinogen locus. Proc Natl Acad Sci U S A 1995; 92:2735. Brasier AR, Li J. Mechanisms for inducible control of angiotensinogen gene transcription. Hypertension 1996; 27:465. Klett CP, Printz MP, Bader M, et al. Angiotensinogen messenger RNA stabilization by angiotensin II. J Hypertens 1996; 14(Suppl 1):S25. Wielbo D, Sernia C, Gyurko R, Phillips MJ. Antisense inhibition of hypertension in the spontaneously hypertensive rat. Hypertension 1995; 25:314. Naftilan AJ, Zuo WM, Ingelfinger J, et al. Localization and differential regulation of angiotensinogen mRNA expression in the vessel wall. J Clin Invest 1991; 88:1300. Schunkert H, Ingelfinger JR, Hirsch AT, et al. Evidence for tissue specific activation of renal angiotensinogen mRNA expression in chronic stable experimental heart failure. J Clin Invest 1992; 90:1523. lnoue I, Nakajima T, Williams CS, et al. A nucleotide substitution in the promoter of human angiotensinogen is associated with essential hypertension. J Clin Invest 1997; 99:1786. Corvol P, Persu A, Gimenez-Roqueplo AP, Jeunemaitre X. Seven lessons from two candidate genes in human essential hypertension: angiotensinogen and epithelial sodium channel. Hypertension 1999; 33:1324. Freire MBS, Ji L, Onuma T, et al. Gender-specific association of M235T polymorphism in angiotensinogen gene and diabetic nephropathy in NIDDM. Hypertension 1998; 31:896. Kunz R, Kreutze R, Beige J, et al. Association between the angiotensinogen 235T variant and essential hypertension in whites: a systematic review and methodological appraisal. Hypertension 1997; 30:1331. Staessen JA, Kuznetsova T, Wang JG, et al. M235T angiotensinogen polymorphism and cardiovascular renal risk. J Hypertens 1999; 17:9. Atwood LD, Kammerer CM, Samollow PB. Linkage of essential hypertension to the angiotensinogen locus in Mexican Americans. Hypertension 1997; 30:326. Niu T, Xu X, Rogus J, et al. Angiotensinogen gene and hypertension in Chinese. J Clin Invest 1998; 101:188. Rakugi H, Jacob HJ, Krieger JE, et al. Vascular injury induces angiotensinogen gene expression in the media and neointima. Circulation 1993; 87:283. Umemura S, Kihara M, Sumida Y, et al. Endocrinological abnormalities in angiotensinogen-gene knockout mice: studies of hormonal responses to dietary salt loading. J Hypertens 1998; 16:285. Navar LG, Inscho EW, Majid SA, et al. Paracrine regulation of the renal microcirculation. Physiol Rev 1996; 76:425. Skott O, Jensen BL. Cellular and intrarenal control of renin secretion. Clin Sci 1993; 84:1. Toffelmire EB, Slater K, Corvol P, et al. Response of prorenin and active renin to chronic and acute alterations of renin secretion in normal humans. Studies using a direct immunoradiometric assay J Clin Invest 1989; 83:679. Hosoi M, Kim S, Takada T, et al. Effects of prorenin on blood pressure and plasma renin concentrations in stroke-prone spontaneously hypertensive rats. Am J Physiol 1992; 262:E234. Gomez RA, Chevalier RL, Everett AD. Recruitment of renin gene-expressing cells in adult rat kidney. Am J Physiol 1990; 259:F660. Sigmund CD, Gross KW. Structure, expression, and regulation of the murine ren genes. Hypertension 1991; 18:446. Kopp U, DiBona GF. Interaction of renal b1-adrenoreceptors and prosta-glandins in reflex renin release. Am J Physiol 1983; 244:F418. Wagner C, Jensen BL, Kramer BK, Kurtz A. Control of the renal renin system by local factors. Kidney Int 1998; 54(Suppl 67):S78. Lee MA, Bohm M, Paul M, et al. Physiologic characterization of the hypertensive transgenic rat TGR(mREN2)27. Am J Physiol 1996; 270(Endocrinol Metab 33):E919. Erdos EG, Skidgel RA. Metabolism of bradykinin by peptidases in health and disease. In: Farmer SG, ed. The kinin system: handbook of immuno-pharmacology. London: Academic Press, 1997:112. Linz W, Wiemer G, Gohlke P, et al. Contribution of the kinins to the cardiovascular actions of angiotensin-converting enzyme inhibitors. Pharmacol Rev 1995; 47:25. Ramchandran R, Kasturi S, Douglas J, et al. Metalloprotease-mediated cleavage secretion of pulmonary ACE by vascular endothelial and kidney epithelial cells. Am J Physiol 1996; 271:H744. Soubrier F, Wei L, Hubert C, et al. Molecular biology of the angiotensin I-converting enzyme, II: structure function: gene polymorphism and clinical implications. J Hypertens l993; 11:599. Wei I, Clauser F, Athenc-Gelas F, Corval P. The two homologous domains of human angiotensin I-converting enzyme interact differently with competitive inhibitors. J Biol Chem 1992; 267:13398. Azizi M, Rousseau A, Ezan E, et al. Acute angiotensin-converting enzyme inhibition increases the plasma level of the natural stem cell regulator N-acetylseryl-lysyl-proline. J Clin Invest 1996; 97:839. Rigat B, Hubert C, Alhenc-Gelas F, et al. An insertion/deletion polymorphism in the angiotensin I-converting enzyme gene accounting for half the variance of serum enzyme levels. J Clin Invest 1990; 86:1343. Van der Kliej FGH, Schmidt A, Navis GJ, et al. Angiotensin converting enzyme insertion/deletion polymorphism and short-term renal response to ACE inhibition: role of sodium status. Kidney Int 1998; 53(Suppl 63):S23. Marre M, Bernadet P, Gallois Y, et al. Relationship between angiotensin I converting enzyme gene polymorphism, plasma levels, and diabetic retinal and renal complications. Diabetes 1994; 43:384. Biege J, Scherer S, Weber A, et al. Angiotensin-converting enzyme genotype and renal allograft survival. J Am Soc Nephrol 1997; 8:1319. Kimura H, Geiyo F, Suzuki S, et al. Polymorphisms of angiotensin converting enzyme and plasminogen activator inhibitor-1 genes in diabetes and microangiopathy. Kidney Int 1998; 54:1659. Urata H, Nishimura H, Ganten D. Chymase-dependent angiotensin II forming system in humans. Am J Hypertens 1996; 9:277. Harding JW, Wright JW, Swanson GN, et al. AT4 receptors: specificity and distribution. Kidney Int 1994; 46:1510. Chappel MC, Pirro NT, Sykes A, Ferrario CM. Metabolism of angiotensin (1-7) by angiotensin converting enzyme. Hypertension 1998; 31:362. Ferrario CM, Chappel MC, Tallant EA, et al. Counterregulatory actions of angiotensin (1-7). Hypertension 1997; 30:535. Unger T, Gohlke P. Tissue renin-angiotensin systems in the heart and vasculature: possible involvement in the cardiovascular actions of converting enzyme inhibitors. Am J Cardiol 1990; 65:31. Dostal DE, Baker KM. Evidence for a role of an intracardiac renin-angiotensin system in normal and failing hearts. Trends Cardiovasc Med 1993; 3:67. Martin P, Wagner J, Dzau VJ. Gene expression of the renin-angiotensin system in human tissues: quantitative analysis by polymerase chain reaction. J Clin Invest 1993; 91:2058. Wang DH, Du Y, Zhao H, et al. Regulation of angiotensin I receptor and its gene expression: role in renal growth. J Am Soc Nephrol 1997; 8:193. Yanagawa N. Potential role of local luminal angiotensin II in proximal tubule sodium transport. Kidney Int Suppl 1991; 32:S33. Kamide K, Hori M, Tuck ML, et al. Insulin-mediated growth in aortic smooth muscle and the vascular renin-angiotensin system. Hypertension 1998; 32:482. Nemeth G, Pepperell JR, Yamada Y, et al. The basis and evidence of a role for the ovarian renin-angiotensin system in health and disease. J Soc Gynecol Investig 1994; 1:118. Speth RC, Daubert DL, Grove KL. Angiotensin II: a reproductive hormone too? Regul Pept 1999; 79(l):25. Hageman JR, Moyer JS, Bachman ES, et al. Angiotensin-converting enzyme and male fertility. Proc Natl Acad Sci U S A 1998; 95:2552. Jones BH, Standbridge MK, Taylor JW, et al. Angiotensinogen gene expression in adipose tissue, analysis of obese models and hormonal and nutritional control. Am J Physiol 1997; 273(Regulatory Integrative Comp Physiol 42):236. Kifor I, Moore TJ, Fallo F, et al. Potassium-stimulated angiotensin release from superfused adrenal capsules and enzymatically digested cells of the zona glomerulosa. Endocrinology 1991; 129:823. Phillips MI, Speakman EA, Kimura B. Levels of angiotensin and molecular biology of the tissue renin angiotensin system. Regul Pept 1993; 43:1. Baltatu O, Nishimura H, Hoffman S, et al. High levels of human chymase expression in the pineal and pituitary gland. Brain Res 1997; 759:269. Griendling KK, Lassegue B, Alexander RW. The vascular angiotensin (AT1) receptor. Thromb Haemost 1993; 70(1):188. Coffman TM. Gene targeting in physiological investigations: studies of the renin-angiotensin system. Am J Physiol 1998; 274 (Renal Physiol):F999. Sadoshima J. Versatility of the angiotensin II type 1 receptor. Circ Res 1998; 82:1352.

56a. Schiffrin EL, Park JB, Intengan HD, Touyz RM. Correction of arterial structure and endothelial dysfunction in human essential hypertension by the angiotensin receptor antagonist losartan. Circulation 2000; 101(14):1653. 57. 58. 59. 60. 61. 62. 63. 64. Horiuchi M, Akishita M, Dzau VJ. Recent progress in angiotensin II type 2 receptor research in the cardiovascular system. Hypertension 1999; 33:613. Cogan MG. Angiotensin II: a powerful controller of sodium transport in the early proximal tubule. Hypertension 1990; 15:451. Okubo S, Numura F, Matsusaka T, et al. Angiotensinogen gene null-mutant lacks homeostatic regulation of glomerular filtration and tubular reabsorption. Kidney Int 1998; 53:617. Scholz H, Kurtz A. Role of protein kinase C in renal vasoconstriction caused by angiotensin II. Am J Physiol 1990; 259:C421. Clemson B, Gaul L, Gubin SS, et al. Prejunctional angiotensin II receptors. Facilitation of norepinephrine release in the human forearm. J Clin Invest 1994; 93:684. Moreau P, d'Uscio LV, Shaw S, et al. Angiotensin II increases tissue endothelin and induces vascular hypertrophy: reversal by ET(A)-receptor antagonist. Circulation 1997; 96:1593. Golub MS, Hori MT, Tuck ML. Arachidonic acid metabolites in the vascutature. In: Sowers JR, ed. Endocrinology of the vasculature. Totowa, NJ: Humana Press, 1996:357. Ichikawa I, Harris RC. Angiotensin actions in the kidney: renewed insight into the old hormone. Kidney Int 1991; 40:583.

65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87.

Wilcox CS, Welch WJ, Snellen H. Thromboxane mediates renal hemodynamic response to infused angiotensin II. Kidney Int 1991; 40:1090. Miller JA, Anacata LA, Cattran DC. Impact of gender on the response to angiotensin II. Kidney Int 1999; 55:278. Wolf G, Ziyadeh FN. The role of angiotensin II in diabetic nephropathy: emphasis on nonhemodynamic mechanisms. Am J Kidney Dis 1997; 29:153. Dzau VJ, Re R. Tissue angiotensin system in cardiovascular medicine: a paradigm shift. Circulation 1994; 89:493. Gibbons GH. The pathophysiology of hypertension. The importance of angiotensin II in cardiovascular remodeling. Am J Hypertens 1998; 11:177S. Alderman MH, Madhaven S, Ooi WL, et al. Association of renin-sodium profile with risk of myocardial infarction in patients with hypertension. N Engl J Med 1991; 324:1098. Vaughan DE. The renin-angiotensin system and fibrinolysis. Am J Cardiol 1997; 79(5A):12. Brown NJ, Agirbasli MA, Williams GH, et al. Effect of activation and inhibition of the renin-angiotensin system on plasma PAI-1. Hypertension 1998; 32:965. . Hwan K, Thornhill B, Wolstenholme T, Chevalier RL. Tissue-specific regulation of growth factors and clusterin by angiotensin II. Am J Hypertens 1998; 11:715. Morishita R, Ogihara T. In vivo evaluation of in vitro hypothesis using gain or loss of function: functional analysis of the renin-angiotensin system. Am J Hypertens 1998; 11:507. Tomita N, Morishita R, Higaki J, et al. Transient decrease in high blood pressure by in vitro transfer of antisense oligodeoxynucleotides against angiotensinogen. Hypertension 1995; 26:131. Phillips MI. Antisense inhibition and adeno-associated viral vector delivery for reducing hypertension. Hypertension 1997; 29:177. Marver D, Kokko JP. Renal target sites and the mechanism of action of aldosterone. Min Elect Metab 1983; 9:1. White PC. Mechanisms of disease. Disorders of aldosterone biosynthesis and action. N Engl J Med 1994; 331:250. Connell JM, Fraser R. Adrenal cortical synthesis and hypertension. J Hypertens 1991; 9:97. Holland OB, Carr B. Modulation of aldosterone synthase messenger ribonucleic acid levels by dietary sodium and potassium and by adrenocorticotropin. Endocrinology 1993; 132:2666. Arriza JL, Weinberger C, Cerelli C, et al. Cloning of human mineralocorticoid receptor complementary DNA: structural and functional kinship with the glucocorticoid receptor. Science 1987; 237:268. Funder JW. Mineralocorticoids, glucocorticoids, receptors and response elements. Science 1993; 259:1132. Bastl CP, Hayslett. The cellular action of aldsterone in target epithelia. Kidney Int 1992; 42:250. Zennaro M-C, Farman N, Bonvalet J-P, et al. Tissue-specific expression of alpha and beta messenger ribonucleic acid isoforms of the human mineral-ocorticoid receptor in normal and pathological states. J Clin Endocrinol Metab 1997; 82:1345. Funder JW. Enzymes and the regulation of sodium balance. Kidney Int 1992; 37(Suppl):S114. Kenouch S, Coutry N, Farman N, Bonvalet J-P. Multiple patterns of 11b-hydroxysteroid dehydrogenase catalytic activity along the mammalian nephron. Kidney Int 1992; 42:56. Naray-Fejes-Toth A, Rusvai E, Fejes-Toth G. Mineralocorticoid receptors and 11b-hydroxysteroid dehydrogenase activity in renal principal and intercalated cells. Am J Physiol 1994; 266:F76.

87a. Funder JW. Aldosterone and mineralocorticoid receptors: orphan questions. Kidney Int 2000; 57(4):1358. 88. 89. 90. 91. 92. Horisberger JD, Rossier BC. Aldosterone regulation of gene transcription leading to control of ion transport. Hypertension 1992; 19:221. Szerlip H, Palevsky P, Cox M, Blazer-Yost B. Relationship of the aldoster-one-induced protein GP70 to the conductive channel. J Am Soc Nephrol 1991; 2:1108. Schafer JA, Hawk CT. Regulation of Na+ channels in the cortical collecting duct by AVP and mineralocorticoids. Kidney Int 1992; 41:255. Hayhurst RA, O'Neil RG. Time-dependent actions of aldosterone and amiloride-blockable sodium channels in A6 epithelia. Am J Physiol 1988; 254:F689. Petzel D, Ganz MB, Nestler EJ, et al. Correlates of aldosterone-induced increases in Ca i 2+ and Isc suggest that Cai 2+ is the second messenger for stimulation of apical membrane conductance. J Clin Invest 1992; 89:150. 93. Hays SR. Mineratocorticoid modulation of apical and basolateral membrane H+/OH/HCO3, transport processes in rabbit inner strip of outer medullary collecting duct. J Clin Invest 1992; 90:180. 94. Brilla CG, Weber KT. Mineralocorticoid excess, dietary medium and myocardial fibrosis. J Lab Clin Med 1992; 120:893. 95. Young M, Fullerton M, Dilley P, Funder J. Mineralocorticoids, hypertension and cardiac fibrosis. J Clin Invest 1994; 93:2578. 95a. Falkenstein E, Christ M, Feuring M, Wehling M. Specific nongenomic actions of aldosterone. Kidney Int 2000; 57(4):1390. Hakeyama H, Miramuri I, Fujita T, et al. Vascular aldosterone, biosynthesis and a link to angiotensin II-induced hypertrophy of vascular smooth muscle. J Biol Chem 1994; 269:24316. Robert V, Heymes C, Silvestre J-B, et al. Angiotensin AT1 receptor subtype as a cardiac target of aldosterone; role in aldosterone-salt-induced fibrosis. Hypertension 1999; 33:981. Gomez Sanchez EP. What is the role of the central nervous system in mineralocorticoid hypertension? Am J Hypertension 1991; 4:374. The RALES Investigators. Effectiveness of spironolactone added to an angiotensin converting enzyme inhibitor and a loop diuretic for severe chronic congestive heart failure (The Randomized Aldactone Evaluation Study [RALES]). Am J Cardiol 1996; 78:902. 100. Ibrahim HN, Rosenberg ME, Greene EL, et al. Aldosterone is a major factor in the progression of renal disease. Kidney Int 1997; 52(Suppl 63):S115. 100a. Farman N, Verrey F. Forefronts in nephrology news in aldosterone action. Kidney Int 2000; 57:1239. 101. Brann DW, Hendry LB, Mahesh VB. Emerging diversities in the mechanism of action of steroid hormones. J Steroid Biochem Molec Biol 1995; 52:113. 102. Adler GK, Chen R, Menachery AI, et al. Sodium restriction increases aldosterone synthesis by increased late pathway, not early pathway, messenger ribonucleic acid levels and enzyme activity in normal rats. Endocrinology 1993; 133:2235. 103. Young DB. Quantitative analysis of aldosterone's role in potassium regulation. Am J Physiol 1988; 255:F811. 104. Himathongan T, Dluhy R, Williams GH. Potassium-aldosterone-renin interrelationships. J Clin Endocrinol Metab 1975; 41:153. 105. Kifor I, Moore TJ, Fallo F, et al. Potassium-stimulated angiotensin release from superfused adrenal capsules and enzymatically digested cells of the zona glomerulosa. Endocrinology 1991; 129:823. 106. Young DB, Smith MJ Jr, Jackson TE, Scott RE. Multiplicative interaction between angiotensin II and K concentration in stimulation of aldosterone. Am J Physiol 1984; 247:E328. 107. Mulkerrin E, Epstein FH, Clark BA. Aldosterone responses to hyperkalemia in healthy elderly humans. J Am Soc Nephrol 1995; 6:1459. 108. Braley LM, Adler GK, Mortensen RM, et al. Dose effect of adrenocorticotropin on aldosterone and cortisol biosynthesis in cultured bovine adrenal glomerulosa cells. In vitro correlate of hyperreninemic hypoaldosterone. Endocrinology 1992; 131:187. 109. Merrill DC, Ebert TJ, Skelton MM, Cowley AW Jr. Effect of plasma sodium on aldosterone during angiotensin II stimulation in normal humans. Hypertension 1989; 14:164. 110. Taylor RE, Glass GT, Radke KJ, Schneider EG. Specificity of effect of osmolality on aldosterone secretion. Am J Physiol 1987; 252:E118. 111. Jones GV, Wall BM, Williams HH, et al. Modulation of plasma aldosterone by physiologic changes in hydrogen ion concentration. Am J Physiol 1992; 262:R269. 112. White PC. Inherited forms of mineralocorticoid hypertension. Hypertension 1996; 28:927. 96. 97. 98. 99.

CHAPTER 80 HYPERALDOSTERONISM Principles and Practice of Endocrinology and Metabolism

CHAPTER 80 HYPERALDOSTERONISM
JOHN R. GILL, JR. Disorders of Mineralocorticoid Excess Physiology of aldosterone Synthesis Secretion Actions Primary Aldosteronism Aldosterone-Producing Adenomas Clinical Features and Pathophysiology Diagnosis and Differential Diagnosis Treatment Idiopathic Hyperaldosteronism Clinical Features and Pathophysiology Diagnosis and Differential Diagnosis Treatment Dexamethasone-Suppressible Hyperaldosteronism Clinical Features and Pathophysiology Diagnosis and Differential Diagnosis Treatment Adrenocortical Carcinoma Clinical Features and Pathophysiology Diagnosis and Treatment Syndromes of Real or Apparent Mineralocorticoid Excess not Caused by Aldosterone Clinical Features and Pathophysiology Diagnosis and Differential Diagnosis Treatment Secondary Hyperaldosteronism Bartter Syndrome Clinical Features and Pathophysiology Diagnosis and Differential Diagnosis Treatment Chapter References

DISORDERS OF MINERALOCORTICOID EXCESS

Mineralocorticoid excess, whether primary or secondary, and disorders that mimic it are commonly manifested clinically as hypokalemia, which usually but not always is accompanied by alkalosis. Thus, a useful strategy for diagnosing hypokalemia is to approach the evaluation as a differential diagnosis of disorders of mineralocorticoid excess. Clinical features that accompany hypokalemia and may be used to determine the specific disorders that need to be considered are illustrated in Table 80-1 and in the flow chart in Figure 80-1. This basic framework is expanded in the course of this review to include the disorders appropriate to each of the four listed patterns of abnormalities of renin and aldosterone. Ideally, evaluation of the renin and aldosterone systems should be done in the absence of medication during a standardized sodium intake (e.g., 100 mEq per day of sodium as sodium chloride). If this is not possible, then at least the use of medications that affect the production of renin and aldosterone (i.e., diuretics and angiotensin converting enzyme inhibitors) should be discontinued and the blood pressure should be controlled with a calcium channel blocking agent. The level of sodium intake immediately preceding the sampling of blood for plasma renin activity and plasma aldosteroneshould be estimated by collecting a 24-hour aliquot of urine for the determination of sodium excretion. The blood for the plasma renin activity and plasma aldosterone evaluations should be drawn after overnight bedrest (in the hospital) or after 30 minutes of bedrest (in the clinic), and then drawn again after 2 hours of standing and walking.

TABLE 80-1. Disorders Characterized by Hypokalemia and Alkalosis

FIGURE 80-1. Flow chart for evaluation of patients with hypokalemic alkalosis. (PRA, plasma renin activity; ALDO, aldosterone.) (See Table 80-1 for continuation of evaluation.)

PHYSIOLOGY OF ALDOSTERONE
SYNTHESIS The zona glomerulosa cells of the adrenal cortex synthesize aldosterone from corticosterone in a two-step, mixed function oxidation reaction that is catalyzed by the enzyme corticosterone methyl oxidase1 (Fig. 80-2). Normally, the adrenal glands secrete 60 to 200 g aldosterone each day, depending on the state of sodium chloride and water balance of the body; the steroid circulates in the blood either free or loosely bound to albumin. Although some of the circulating aldosterone may be excreted in the urine in the free form, most of it is metabolized by the liver to tetrahy-droaldosterone and by the kidneys to aldosterone-18-glucuronide. Measurement of the excretion of the 18-glucuronide metabolite has been used as an index of aldosterone production; values obtained by this method range from 1 to 15 g per day in normal persons receiving an unrestricted sodium chloride intake.

FIGURE 80-2. Proposed mechanism and structure of intermediates in the conversion of corticosterone to aldosterone. (From Ulick S. Diagnosis and nomenclature of the disorders of the terminal portion of the aldosterone biosynthetic pathway. J Clin Endocrinol Metab 1976; 43:92.)

SECRETION Aldosterone biogenesis is regulated primarily by the renin-angiotensin system, with potassium, atrial natriuretic hormone (ANH), anddopamine also making important contributions. Corticotropin (ACTH) also may stimulate aldosterone production, but its effects are self-limited except in certain disease states. In addition, the factors that determine aldosterone secretion have direct effects on the kidney and vasculature and, along with aldosterone, contribute to the maintenance of the volume of extracellular fluid, the concentration of extracellular potassium, and, ultimately, blood pressure. Thus, when the intake of sodium chloride is restricted and blood volume is contracted, an increase in the formation of angiotensin II stimulates the secretion of aldosterone. Angiotensin II and aldosterone, in turn, both increase the reabsorption of sodium by the renal tubule to curtail sodium loss and restore extracellular volume. Conversely, when sodium chloride intake is increased and blood volume is expanded, the formation of angiotensin II decreases, ANH is secreted by the heart, and the formation of dopamine by the adrenal glands and kidneys increases. ANH and dopamine inhibit aldosterone biogenesis; they also increase sodium excretion through direct effects on the kidney. Potassium exerts differential effects on aldosterone biosynthesis. Hyperkalemia, secondary to potassium administration or retention, stimulates aldosterone secretion, whereas hypokalemia, secondary to potassium depletion, inhibits it. A more detailed discussion of the renin-angiotensin system and its control of aldosterone secretion is presented in Chapter 79 and Chapter 183; ANH is discussed in Chapter 178. ACTIONS Aldosterone exerts its biologic actions by crossing the plasma membrane of target cells and binding to its receptor in the cytosol.2 The aldosterone-receptor complex then migrates to the nucleus, where it binds, thereby initiating transcription, translation, and synthesis of proteins responsible for expression of the action of aldosterone on the functions of that target cell (see Chap. 4). Collecting tubule cells of the kidney and epithelial cells of other transporting tissues (i.e., sweat, salivary gland, and intestine) are target tissues in which the physiologic effects of aldosterone are most discernible, although effects on vascular3 and other tissues may occur. In epithelial cells for which the physiologic actions of aldosterone have been extensively characterized, the steroid facilitates passage of sodium across the apical membrane into the cell and accelerates extrusion of sodium and uptake of potassium by Na+,K+-adenosine triphosphatase located in the basolateral membrane.4 In the renal collecting tubule, these actions of aldosterone increase the reabsorption of sodium and the secretion of potassium, leading to retention of sodium by the body and loss of potassium (see Chap. 206). As the foregoing discussion indicates, the renin-angiotensin-aldosteronesystems regulate the volume of extracellular fluid and contribute to the maintenance of the extracellular potassium concentration. As a result of disease, the production of aldosterone may become supernormal and cease to be responsive to the physiologic stimuli that regulate it. The overproduction of aldosterone may result from overactivity of the renin-angiotensin system (secondary hyperaldosteronism) or from other abnormalities (primary hyperaldosteronism). The consequences of continual overproduction of aldosterone, or hyperaldosteronism, depend in part on the disorder with which it is associated, and may include excessive sodium chloride and water retention, potassium loss, alkalosis, and hypertension.

PRIMARY ALDOSTERONISM
Autonomous overproduction of aldosterone occurs in ~2% of patients with hypertension and was recognized initially as being associated with either a unilateral adenoma (aldosterone-producing adenoma, or APA) or bilateral hyperplasia of the zona glomerulosa. Because patients with bilateral adrenal disease tended to have a clinical presentation similar to that of patients with an APA (hypertension and hypokalemia) and, like the patients with unilateral disease, they showed suppression of the renin-angiotensin system, they also were assumed to have a primary adrenal disorder. Subsequent observations have not proved this to be true; they suggest, instead, that bilateral hyperplasia of the zona glomerulosa probably is not a primary adrenal disorder in most cases and may have more than one cause. Unilateral adrenalectomy usually cures or ameliorates the hypertension in patients with an APA, whereas subtotal or total adrenalectomy, with few exceptions, has little effect on the blood pressure of patients with bilateral hyperplasia. In the case of those few patients who have primary adrenal hyperplasia, unilateral adrenalectomy may be curative.5 In some patients with bilateral hyperplasia, the hyper-aldosteronism was dependent on ACTH (so-called glucocorticoid-remediable hyperaldosteronism). The hypothesis that a trophic hormone, possibly of pituitary origin, may be responsible for bilateral hyperplasia not caused by ACTH has received considerable support, although it has not been definitively proved. As a result of advances in knowledge, the inclusive designation of primary hyperaldosteronism for those patients with hyperal-dosteronism associated with both adenoma and bilateral hyperplasia has tended to be replaced by the more specific designations of aldosterone-producing adenoma, idiopathic hyperaldosteronism, primary adrenal hyperplasia, and glucocorticoid-remediable hyper-aldosteronism. Rarely, adrenocortical carcinoma may overproduce aldosterone. These five disorders present as hypokalemic alkalosis, hypertension, low or suppressed plasma renin activity, and high aldosterone concentrations (see Table 80-1).

ALDOSTERONE-PRODUCING ADENOMAS
CLINICAL FEATURES AND PATHOPHYSIOLOGY Aldosterone-producing adrenocortical adenomas usually are small (0.52.5 cm), unilateral, solitary, and associated with hypoplastic zona glomerulosa. Occasionally, however, these lesions may be multiple and associated with hyperplasia of the zona glomerulosa. This benign tumor of the adrenal cortex occurs most frequently in the third through the fifth decades of life but also has been observed in prepubertal children6 and older persons (range, 1366 years). Aldosterone-producing adenomas occur equally in men and women and do not appear to have a racial predisposition; they tend to develop more frequently in the left adrenal gland than in the right, but this difference is small. (The occurrence of primary aldosteronism resulting from APA in two or more members of the same family has been reported,7 as well as the familial occurrence of idiopathic aldo-steronism. In one of the families, one member had bilateral nodular hyperplasia and another had APA. This familial form of primary aldosteronism was labeled familial hyperaldosteronism type II to distinguish it from glucocorticoid-remediable hyperaldosteronism. The clinical features of familial hyperal-dosteronism type II do not differ from those of sporadic patients with APA or idiopathic hyperaldosteronism.) The symptoms reported most frequently by patients with an APA are headache, easy fatigability, and weakness, although high blood pressure may be the only symptom in many cases. Hypokalemia, suppressed plasma renin activity, and high plasma and urinary aldosterone values are the findings usually associated with APA, but they may not always be demonstrable.8 Aldosterone production by an APA is autonomous and is affected only minimally by angiotensin II because of the decreased formation of this peptide as well as the decreased responsiveness of the adenoma. An APA is unresponsive to the infusion of dopamine, although endogenous dopamine may exert tonic inhibitory effects.9 ANH, levels of which appear to be elevated in patients with this disorder,10 does not inhibit aldosterone biogenesis by APA cells in vitro, as it does in normal glomerulosa cells.11 Therefore, this hormone may have little effect on aldosterone secretion by an APA in vivo. In contrast, changes in serum potassium or ACTH levels may exert important regulatory effects on aldosterone secretion in patients with this tumor: aldosterone production may be blunted by the potassium loss and the hypokalemia that it induces. The circadian rhythm of aldosterone in patients with APA tends to mirror that of ACTH. As a consequence of sustained overproduction of aldosterone, retention of sodium chloride and water occurs, but this usually is limited; escape occurs when extracellular fluid has been expanded by ~500 mL.12 Initially, the renal excretion of potassium increases. As hypokalemia develops, potassium excretion decreases to reflect intake of this cation. Hydrogen ion excretion, which increases principally because of an increase in urinary ammonium, is responsible for the development and

maintenance of metabolic alkalosis. Urinary calcium and magnesium levels are high; presumably, this is a reflection of the effects of expansion of extracellular fluid on the tubular reabsorption of these ions. The urinary concentrating ability is impaired because of potassium depletion. Although sodium and water retention may contribute to the increase in blood pressure that occurs, the precise mechanisms responsible for an increase in peripheral resistance and the associated hypertension are unclear. Receptors for aldosterone have been found in blood vessels and in the brain, and the effects of aldosterone on ion transport in these tissues may be important mediators of the increase in blood pressure. Preliminary studies in dogs indicate that an amount of aldosterone that is too small to exert an effect when infused peripherally increases blood pressure when it is infused into the third ventricle of the brain.13 DIAGNOSIS AND DIFFERENTIAL DIAGNOSIS Patients with hypertension who have borderline or low serum potassium concentrations, low stimulated plasma renin activity (<2 ng/mL per hour), and borderline or high plasma aldosterone concentrations (>14 ng/dL) should be evaluated further for autonomous overproductionof aldosterone. The evaluation should be done before antihypertensive treatment is instituted or after it has been discontinued for 2 weeks. Determining aldosterone suppressibility, by combining a high sodium intake and treatment with a mineralocorticoid such as 9-afluorohydrocortisone for 3 or more days, or by intravenous saline infusion, is a useful first step. Of these two interventions, the infusion of 2 L normal saline over 4 hours with the measurement of plasma aldosterone levels before and at the conclusion of the infusion is simpler and may be readily adapted to the evaluation of outpatients.14 At the conclusion of the infusion, the plasma aldosterone level normally is 5 ng/dL or less; in patients with hyperaldosteronism, it usually exceeds 10 ng/dL. Determination of the plasma aldosterone to plasma renin activity ratio also has been used to detect aldosteronism. Single determinations of this ratio tend to be of limited use because values from patients with hyperaldosteronism frequently are the same as those from patients with normal adrenal function. A solution to the problem of overlap that appears to improve specificity considerably is to collect blood samples every 30 minutes for 6 hours, so that an integrated plasma aldosterone/plasma renin activity ratio can be determined.15 A somewhat simpler but probably equally effective alternative is to collect blood 90 minutes after the administration of 25 to 50 mg captopril, which is a converting enzyme inhibitor, for determination of the plasma aldosterone/plasma renin activity ratio16 (Fig. 80-3). Both procedures are safe and do not require dietary preparation or hospitalization. It has been suggested that all patients with hypertension undergo captopril screening tests,17 because determination of the serum potassium concentration and plasma renin activity may be inadequate measures for screening the hypertensive population for hyperaldosteronism.8 The normogram in Figure 80-3 indicates the degree of separation of patients with aldosterone excess from those with essential hypertension.16 Subsequent evaluation of the captopril test indicates a false-negative rate of 6.3% and a false-positive rate of 0.6% in patients with primary hyperaldosteronism.18 The problem of identifying those patients with hyperaldosteronism who harbor an adenoma has been greatly facilitated by technologic advances. Computed tomography (CT) appears to have the capacity to discern an adrenal mass in ~80% of patients with an aldosterone-producing adenoma.19,20 Magnetic resonance imaging also has been used to evaluate the adrenal glands in patients with hyperaldosteronism; apart from confirming the findings of CT, however, it does not appear to offer any advantage over CT. These new techniques for imaging adrenal masses have replaced the more cumbersome and time-consuming procedure of131 Iiodocholesterol imaging in most centers (see Chap. 88). A limitation of all the imaging techniques, however, is that an adenoma may be poorly visualized or not detected at all. These problems result from the small size of some adenomas, which may be less than 1 cm in diameter. The problem of an inconclusive CT scan may be partially resolved by determining the plasma 18-hydroxycorticosterone level after overnight bedrest.21 In a series of 50 patients with aldosteronism, all but 1 of the 24 patients with an adenoma had 18-hydroxycorticosterone values that exceeded 50 ng/dL.8 Thus, a plasma 18-hydroxycorticosterone value >50 ng/dL in a patient with hyperaldosteronism suggests that the diagnosis is more likely to be an aldosterone-producing adenoma than bilateral hyperplasia, even though the adrenal glands may appear normal (Fig. 80-4). If the plasma aldosterone level were to show a paradoxic fall paralleling the circadian fall in plasma cortisol during 2 hours of ambulation after early morning sampling, this also would suggest that the diagnosis is more likely to be an aldosterone-producing adenoma than bilateral hyperplasia.22 So many patients with an adenoma, like those with bilateral hyperplasia, show a rise rather than a fall in the plasma aldosterone level that the usefulness of the procedure is limited.8

FIGURE 80-3. Values for the plasma aldosterone (ALDO)/plasma renin activity (PRA) ratio plotted as a function of plasma aldosterone for normal and hypertensive control subjects (hatched area), patients with aldosterone-producing adenoma (APA), and patients with idiopathic hyperaldosteronism (IHA) 2 hours after the administrationof 25 mg captopril. The plasma aldosterone value decreased to <15 ng/dLin normal persons, but remained higher than this in most with IHA and in all with APA. The ALDO/PRA ratio was <50 in normal persons and >50 in most with IHA and all with APA. (From Lyons DF, Kem DC, Brown RD, et al. Single dose captopril as a diagnostic test for primary aldosteronism. J ClinEndocrinol Metab 1983; 57:892.)

FIGURE 80-4. Plasma 18-hydroxycorticosterone (18OHB) and plasma aldosterone (ALDO)concentrations at bedrest and after 4 hours of ambulation in patients with aldosterone-producing adenomas (adenoma) or idiopathic hyperaldosteronism (hyperplasia). Note the wide separation in resting values for 18OHB between patients with adenoma and patients with hyperplasia and the different responses of the two groups to ambulation. (From Biglieri EG, Schambelan M. The significance of elevated levels of plasma 18-hydroxycorticosterone in patients with primary aldosteronism. J Clin Endocrinol Metab 1979; 49:87.)

The most reliable means of assessing adrenal glands in patients with hyperaldosteronism is bilateral catheterization of adrenal veins to sample for corticosteroids.23 The aldosterone/cortisol ratio in samples of venous blood from both adrenal glands and a peripheral site, before and after ACTH stimulation, provides a good comparative characterization of adrenal function and an extremely reliable means of distinguishing between bilateral and unilateral overproduction of aldosterone. Table 80-2 shows the results of adrenal vein sampling in a patient with an aldosterone-producing adenoma. Determination of the cortisol level provides important confirmation of roentgenographic evidence that the site of sampling is actually an adrenal vein; it also is used to correct values of adrenal venous aldosterone for dilution by blood of extraadrenal origin. In the illustrated patient, the findings show the importance of stimulation with ACTH to ensure that an adenoma is functioning, rather than quiescent, at the time of sampling and also to demonstrate suppression of aldosterone production by the contralateral gland. Although adrenal vein sampling is an invasive procedure, it is associated with minimal morbidity. When used as just described, it has an accuracy >90% in the lateralization of an adenoma, making the distinction between an APA and those disorders associated with bilateral overproduction of aldosterone. The observations that APA may be associated with other abnormalities seen on CT scans such as ipsilateral and contralateral nonfunctioning nodules and thickening of an adrenal limb indicate the important role of adrenal vein sampling in reaching a correct diagnosis.20 Only occasionally will a patient with idiopathic aldosteronism have a gradient between the two adrenal glands with an aldosterone/cortisol ratio that approaches 4 to 1, the lower limit of the range of the ratio of APA to normal adrenal.24

TABLE 80-2. Adrenal Venous Sampling

TREATMENT The preferred therapy for patients with an aldosterone-producing adenoma is operative removal of the adrenal gland containing the tumor.25 Adenomectomy alone is not sufficient.25a This can be done laparoscopically with minimal morbidity26 (see Chap. 89). The cure rate (correction of hypertension as well as aldosteronism) ranges from 60% to 75%. Because the prevalence of a family history of hypertension in patients with an aldosterone-producing adenoma may range from 40% to 60%, this, along with the duration of the hypertension, may account for the persistence of an elevated blood pressure despite the correction of hyperaldosteronism. Most of the patients who remain hypertensive after the correction of hyperaldosteronism become more responsive to antihypertensive medication. Spironolactone is probably the single most effective drug for treating patients with an APA. It acts as a competitive inhibitor, competing with aldosterone for its cytosol receptor, thereby antagonizing the action of the mineralocorticoid in target tissue. Spironolactone also is both an antiandrogen and a progestagen, and this explains many of its distressing side effects;2,4 decreased libido, mastodynia, and gynecomastia may occur in 50% or more ofmen, and menometrorrhagia and mastodynia may occur in an equally large number of women taking the drug.27 The problem of side effects may preclude the long-term use of spironolactone, particularly in younger men and women, and is more likely to occur when the dosage exceeds 100 mg per day. The usual dosage of spironolactone ranges from 50 mg once daily to 100 mg twice daily. Drugs such as amiloride and triamterene also may oppose the effect of aldosterone on the renal tubule, but these agents act by inhibiting sodium reabsorption and potassium secretion through a direct effect on the tubule cell, not by competing with aldosterone for its receptor. This may explain why triamterene and amiloride tend to be less effective than spironolactone as antihypertensive agents in patients with hyperaldosteronism. Calcium channel blocking drugs, such as nifedipine, also may be useful therapeutic agents in patients with hyperaldo-steronism. In addition to its antihypertensive action, nifedipine may decrease aldosterone production.28 Thus, the combined use of nifedipine and a potassium-sparing diuretic may serve as an alternative to spironolactone in the medical treatment of patients with an APA.

IDIOPATHIC HYPERALDOSTERONISM
CLINICAL FEATURES AND PATHOPHYSIOLOGY Idiopathic bilateral adrenocortical hyperplasia of the zona glomerulosa (idiopathic hyperaldosteronism) is characterized by features similar to those associated with APAs, although they are less pronounced in many patients.29 The reported prevalence of idiopathic hyperaldosteronism has ranged from a high of 70% of those patients with primary hyperaldosteronism14 to a low of 8%.20 The true prevalence probably is somewhere between these two extremes and may be ~30%. Although the etiology of idiopathic aldosteronism has not been established, the belief that a circulating stimulatory factor is responsible for hyperfunction of the zona glomerulosa is generally accepted. Extracts of urine from normal persons contain a protein fraction that selectively stimulates the production of aldosterone and produces hypertension when injected into rats.30 Subsequent studies of one of these extracts have shown that (a) it is a peptide of pituitary origin, (b) it is increased in patients with idiopathic hyperaldosteronism but not in those with an APA, and (c) it is not suppressed by dexamethasone.31 The specific peptide has not been identified. In other studies, plasma concentrations of g-melanotropin32 and of b-endorphin33 were increased in patients with idiopathic hyperaldosteronism but not in those with an aldosterone-producing adenoma. These two peptides are fragments of a larger peptide, pro-opiomelanocortin, which is formed in the pituitary; they may stimulate the secretion of aldosterone or increase the sensitivity of the adrenal gland to angiotensin II, or both. The report of an enlargement of the intermediate lobe of the pituitary in a patient with idiopathic hyperaldosteronism,34 together with the observations that central serotoninergic stimulation of aldosterone secretion may occur35 and that cypro-heptadine, which is an inhibitor of central serotonin release, may decrease plasma aldosterone in patients with idiopathic hyperaldosteronism,36 is consistent with a possible role for the pituitary in the pathogenesis of idiopathic hyperaldosteronism. It should be noted, however, that a pathogenetic schema for idiopathic hyperaldosteronism has to account for the hypertension as well as the hyperaldosteronism because adrenalectomy rarely lowers blood pressure. Therefore, a putative aldoster-one-stimulating factor of pituitary or other origin, orchanges proximal to its release, must be responsible for the hypertension in idiopathic hyperaldosteronism, unless the hypertension is a separate process. The effects of the putative aldosterone-stimulating factors on blood pressure are not known. As in an APA, the overproduction of aldosterone in idiopathic hyperal-dosteronism causes increased sodium reabsorption and potassium secretion by the renal tubule, expansion of the volume of extracellular fluid, suppression of plasma renin activity, and hypokalemia. DIAGNOSIS AND DIFFERENTIAL DIAGNOSIS Although the features that characterize prolonged aldosterone excess may be as prominent in patients with idiopathic hyperal-dosteronism as in patients with an APA, they usually are less pronounced and more subtle.29 (It has been suggested that approximately half the patients with hypertension and suppressed plasma renin activity who show a decrease in plasma aldosterone to values between 5 and 10 ng/dL in response to saline infusion may have idiopathic hyperaldosteronism, even though baseline plasma aldosterone and serum potassium values may be normal.14 When these diagnostic criteria are used, it may be difficult to distinguish between idiopathic hyperaldo-steronism and low-renin essential hypertension. Such a sharp distinction between idiopathic hyperaldosteronism and low-renin essential hypertension may not be made with certainty until the pathogenesis of both disorders is more clearly delineated.) More important distinctions are those between idiopathic hyperaldosteronism and an APA, and between idiopathic hyperaldosteronism and primary bilateral adrenal hyperplasia, because therapy for these disorders is different. The response to saline infusion may be helpful because the aldosterone/cortisol ratio increases in an APA but remains unchanged or decreases, with a value of 2.2 or less, in idiopathic hyperaldosteronism37 (Fig.80-5). A value for plasma 18-hydroxycorticosterone less than 50 ng/dL at 8:00 a.m. after overnight bedrest is the usual finding in idiopathic hyperaldosteronism.8,21 In patients with an APA, as well as in most patients with primary hyperplasia, values for 18-hydroxycorticosterone are >50 ng/dL and may exceed 100 ng/dL.5,21,38 An increase in plasma aldosterone levels associated with a decrease in plasma cortisol levels after 2 hours of ambulation in the morning favors a diagnosis of idiopathic hyperaldosteronism but also may occur in association with an APA.22 An increase in plasma aldosterone levels is more difficult to interpret than is a paradoxic decrease. Because of the number of false-positive and false-negative results, the test, as originally described, is not sufficiently dependable to be a reliable tool.8 A refinement in the interpretation of the postural stimulation test (i.e., correction of the percentage increase in aldosterone by subtraction of the percentage increase in cortisol with standing) and acceptance of an increase in levels of plasma aldosterone with standing of <30% as a positive response for an adenoma has improved the predictive value of the test.39 The response of plasma aldosterone to a single dose of a converting enzyme inhibitor, such as captopril, is more helpful in discriminating between hyperaldosteronism and low-renin essential hypertension than in separating patients with idiopathic hyperaldosteronism from those with an APA.16 Interestingly, others have observed a decrease in aldosterone levels and blood pressure in patients with idiopathic hyperaldosteronism who are given larger doses of converting enzyme inhibitors for longer periods.40 In contrast to the results obtained with a single dose of a converting enzyme inhibitor, testing with saralasin, which is an angiotensin II antagonist, correctly identified 15 patients with aldosteronism. After infusing saralasin, 10 g/kg per minute for 30 minutes, plasma aldosterone levels increased in eight patients with idiopathic hyperaldosteronism but did not change in six patients with an adenoma.41 Most important, thesemethods of evaluation assess adrenal function indirectly. Therefore, the more concordance there is among the results of several tests, the more confidence one can have in the diagnosis. The best procedure for resolving uncertainty and making a more definitive diagnosis is adrenal vein catheterization. This procedure provides a direct assessment of adrenal biology; hence, bilateral supernormal secretion of aldosterone is readily distinguished from supernormal secretion of aldosterone by one adrenal gland with suppression of secretion from the contralateral gland. The results of adrenal vein catheterization in a representative patient with idiopathic hyperaldosteronism are shown in Table 80-2.

FIGURE 80-5. The plasma aldosterone/plasma cortisol ratio before (control) and after the administration of 1250 mL of a normal saline infusion (SI) over 2 hours from 8 a.m. to 10 a.m. Note consistent increases and higher values for the ratio in patients with aldosterone-producing adenoma (APA) than in those with idiopathic hyperaldosteronism (IHA). (From Arteaga E, Klein R, Biglieri EG. Use of the saline infusion test to diagnose the cause of primary aldosteronism. Am J Med 1985; 79:722.)

Once a tentative diagnosis of idiopathic hyperaldosteronism has been made, it is necessary to test for the less common disorder of dexamethasone-suppressible hyperaldosteronism, which also is characterized by bilateral adrenal hyperfunction. Because increased aldosterone production in dexamethasone-suppressible hyperaldosteronism is ACTH dependent, treatment with dexamethasone, 0.5 mg every 6 hours for 3 days, decreases plasma and urinary aldosterone activity to extremely low values42 (<2.0 ng/dL and 2 g per day, respectively; Fig. 80-6). Although treatment with glucocorticoids also decreases plasma and urinary aldosterone values in idiopathic hyperaldosteronism, the change occurs more slowly and is less pronounced (Fig. 80-7). If the results of dexamethasone suppression suggest that a patient may have dexamethasone-suppressible hyperaldosteronism, determination of urinary 18-hydroxycortisol and 18-oxocortisol levels and of the response to ACTH should confirm the diagnosis. Identification of the gene mutations that cause dexametha-sone-suppressible hyperaldosteronism makes it possible to screen for the disorder directly by genetic testing.43

FIGURE 80-6. Response of mean blood pressure (BP), serum potassium (K+), ambulatory plasma renin activity (UP PRA), ambulatory plasma aldosterone concentration (UP plasma ALDO), urine ALDO, urine K+, urine sodium (Na+), and body weight (BW), to dexamethasone (DEX) in a patient with dexamethasone-suppressible hyperaldosteronism. Note the abrupt decrease in urine ALDO with DEX. (From Gill JR Jr, Bartter FC. Overproduction of sodium-retaining steroids by the zona glomerulosa is adrenocorticotropin-dependent and mediates hypertension in dexamethasone-suppressible aldosteronism. J Clin Endocrinol Metab 1981; 53:331.)

FIGURE 80-7. Response of mean blood pressure (BP), serum potassium (K+), ambulatory plasma renin activity (UP PRA), urine aldosterone (ALDO), urine sodium (Na+), and urine K+ to dexamethasone (DEX) in a patient with idiopathic hyperaldosteronism.

TREATMENT Because adrenalectomy does not correct hypertension in idiopathic hyperaldosteronism,44 except in rare patients with primary bilateral hyperplasia,5,38 most patients must be treated medically. A calcium channel blocking drug, such as nifedipine, is a good antihypertensive agent in this condition because it decreases plasma aldosterone levels, presumably by inhibiting secretion, in addition to reducing blood pressure.28 Enalapril, a converting enzyme inhibitor, also has been reported to decrease aldosterone production and to reduce blood pressure in a small series of patients with idiopathic hyperaldosteronism, and may prove to be a useful therapeutic agent.40 If hypokalemia persists during treatment with nifedipine or enalapril, it usually can be managed by the addition of a potassium-sparing diuretic, such as triamterene or amiloride. Unusual patients with bilateral hyperplasia who have high plasma aldosterone levels with positive postural stimulation test results and high 18-hydroxycorticosterone values should be given a trial of spironolactone.5,38,40 If treatment with spirono-lactone produces a sustained decrease in blood pressure, then either unilateral or bilateral adrenalectomy may be curative.5,38

DEXAMETHASONE-SUPPRESSIBLE HYPERALDOSTERONISM
CLINICAL FEATURES AND PATHOPHYSIOLOGY Dexamethasone-suppressible hyperaldosteronism (glucocorticoid-remediable aldosteronism) is a rare familial disorder inherited as an autosomal-dominant trait that exhibits the clinical features of hyperaldosteronism outlined previously.45 As a result of an abnormality in the adrenal glands of patients with dexametha-sone-suppressible hyperaldosteronism, ACTH causes a sustained overproduction of aldosterone (Fig. 80-8). This is in contrast to the usual response of aldosterone secretion to the prolonged administrationof ACTH: after an initial rise, plasma aldosterone levels fall back to baseline values on the second day and show little change during the subsequent days of ACTH treatment (Fig. 80-9). Patients with dexamethasone-suppressible hyperaldosteronism exhibit gene duplication in which there is an unequal crossing over and fusion of the 5' regulatory region of 11b-hydroxylase with the coding region of aldosterone synthase to form a chimeric gene.46 These chimeric gene formations probably have occurred because of the close proximity of the genes for 11b-hydroxylase and aldosterone synthase on chromosome 8 and their almost identical (95%) nucleotide sequences. These chimeric genes result in expression of aldosterone synthase under the regulatory control of ACTH in adrenal fasciculata cells. This explains why aldosterone is regulated by ACTH in these patients and why ACTH produces a sustained overproduction of aldosterone (see Fig. 80-8) instead of the normal transitory rise with a fall back to baseline values for the remainder of ACTH administration (see Fig. 80-9). The

presence of an ACTH-regulated aldosterone synthase in adrenal fasciculata cells also explains why these patients form and excrete in their urine supernormal amounts of 18-hydroxycortisol and 18-oxocortisol47 in addition to aldosterone. The production of aldosterone by fasciculata cells leads to suppression of the production of aldosterone by the zona glomerulosa. Because 18-oxocortisol has mineralocorticoid activity,48 it may represent the mineralocorticoid activity seen in the radioreceptor assay that is not accounted for by summation of known mineralocorticoids.49 An increase in 18-oxocortisol as well as in aldosterone during treatment with ACTH may explain the greater rise in blood pressure seen in patients with dexametha-sone-suppressible hyperaldosteronism when ACTH is given than when high doses of aldosterone are given for a similar period.50

FIGURE 80-8. Response of mean blood pressure (BP), serum potassium (K+), plasma aldosterone (plasma ALDO), urine aldosterone (urine ALDO), urine K+, urine sodium (Na+), and body weight (BW) to adrenocorticotropin (ACTH) in a patient with dexamethasone-suppressible hyperaldosteronism. Note the sustained increase in ALDO during ACTH administration. (From Gill JR Jr, Bartter FC. Overproduction of sodium-retaining steroids by the zona glomerulosa is adrenocorticotropin-dependent and mediates hypertension in dexamethasone-suppressible aldosteronism. J Clin Endocrinol Metab 1981; 53:331.)

FIGURE 80-9. Response of mean blood pressure (BP), serum potassium (K+), and urine aldosterone (ALDO) to adrenocorticotropin (ACTH) in a patient with idiopathic hyperaldosteronism. Note transient response of ALDO to ACTH.

DIAGNOSIS AND DIFFERENTIAL DIAGNOSIS The unique clinical features of this subset of hyperaldosteronism are the rapid and complete suppression of plasma and urinary aldosterone levels in 24 to 48 hours by dexamethasone (see Fig. 80-6) and the sustained increase in aldosterone production that accompanies treatment with ACTH42 (see Fig. 80-8). These features readily distinguish dexamethasone-suppressible hyperaldosteronism from idiopathic hyperaldosteronism, the disorder it most closely resembles. Improvement in hypertension and in potassium wasting has been observed during pregnancy and may be a consequence of antagonism of overproduced mineralocorticoids by progesterone.42 Once a diagnosis has been made, other members of the family should be screened by dexamethasone administration, by determination of urinary 18-hydroxycortisol and 18-oxocortisol levels, or by genetic testing. Although patients with dexamethasone-suppressible hyperaldosteronism usually have hypokalemia as well as hypertension, genetic screening has revealed that affected persons may have normal potassium levels with borderline elevation of the blood pressure.42,51 TREATMENT Dexamethasone, given in a dose that replaces the patient's output of glucocorticoids, corrects the overproduction of aldosterone, the hypokalemia, and the hyporeninemia, and reduces the blood pressure to normal levels (see Fig. 80-6). This treatment has the disadvantage of suppressing the hypothalamic-pituitary axis and blunting the maximal release of ACTH. Alternatively, patients with dexamethasone-suppressible hyperaldosteronism may be treated with a potassium-sparing diuretic, such as amiloride or triamterene, which is equally effective in correcting hypokalemia and decreasing blood pressure but does not affect the hypothalamic-pituitary axis. Spironolactone also is efficacious, but it tends to produce more side effects.

ADRENOCORTICAL CARCINOMA
CLINICAL FEATURES AND PATHOPHYSIOLOGY Adrenocortical carcinoma is an extremely rare cause of hyperal-dosteronism and accounts for fewer than 2% of aldosterone-producing tumors.52,52a Although aldosterone may be the only adrenal steroid overproduced early in the course of the disease, hypersecretion of cortisol usually occurs at some point. Occasionally, overproduction of adrenal androgens may occur. The circadian rhythm of plasma aldosterone levels, commonly seen in patients with an adenoma, is absent in patients with adrenal carcinoma, presumably because ACTH has little effect on aldosterone secretion by malignant tumor cells. DIAGNOSIS AND TREATMENT The clinical presentation of adrenal carcinoma may be similar to that ofadrenal adenoma, and the diagnosis may depend on the histologic demonstration of invasion of the capsule and blood vessels by tumor cells.52 More commonly, the large size of the adrenal mass and its involvement of the kidney identifies it as a malignant tumor before surgery. Although metastases may not be demonstrable at the time the primary tumor is removed, they appear subsequently, with a course from diagnosis to death that may range from 2 to 13 years.52 Treatment with mitotane (o,p'-DDD) appears to have little effect on the progression of the disease. Spironolactone may be variably effective in controlling renal potassium loss.

SYNDROMES OF REAL OR APPARENT MINERALOCORTICOID EXCESS NOT CAUSED BY ALDOSTERONE

CLINICAL FEATURES AND PATHOPHYSIOLOGY Occasionally, patients may have features of mineralocorticoid excess (hypokalemic alkalosis, suppressed plasma renin activity, and hypertension) but plasma aldosterone concentrations that are low or undetectable (see Table 80-1). Some of these patients have increased production of other mineralocorticoids, such as deoxycorticosterone or corticosterone, as a result of 11b-hydroxylase deficiency or 17-hydroxylase deficiency, respectively (Table 80-3; see Chap. 77); others do not overproduce any of the known adrenal steroids. Still others accumulate an abnormal amount of cortisol in renal tubule cells because of 11b-hydroxysteroid dehydrogenase deficiency that leads to increased occupancy of mineralocorticoid receptors by cortisol.53

TABLE 80-3. Hormonal Profile of Congenital Enzyme Deficiencies Causing Hypokalemia, Alkalosis, and Hypertension

In Liddle syndrome, increased renal tubular reabsorption of sodium and secretion of potassium presumably are independent of stimulation by a mineralocorticoid because they are unaffected by treatment with spironolactone.54 The cause of Liddle syndrome is mutations in the bsubunit of the epithelial sodium channel.55 It now appears that most and possibly all of the described patients with apparent mineralocorticoid excess, some of them siblings, who were thought to be overproducing an unidentified steroid because they responded to spironolactone, have a deficiency of 11b-hydroxysteroid dehydrogenase.56,57 As a result of this abnormality, those kidney and liver cells that lack the enzyme are unable to convert cortisol to cortisone. The manifestations of this block in cortisol metabolism are a prolonged half-life of plasma cortisol, suppression of ACTH, an accumulation of cortisol in renal tubule cells, and increased occupancy of mineralocorticoid receptors in those cells by cortisol. The last of these events is thought to mediate the changes of mineralocorticoid excess in these patients.53 No defects have been found in the gene encoding for 11b-hydroxysteroid dehydrogenase.58 Studies suggest that the syndrome of apparent mineralocorticoid excess produced by the ingestion of licorice may be the result of inhibition of 11b-hydroxysteroid dehydrogenase by glycyrrhizic acid and its metabolite, glycyrrhetinic acid.53,59 Normal volunteers fed glycyrrhetinic acid show a prolongation of the plasma cortisol half-life, with a tendency for the urinary metabolites of cortisol to increase and those of cortisone to decrease.60 Because licorice compounds have no effect in adrenalectomized animals, they presumably induce sodium retention and potassium secretion by increasing the cortisol available to bind to mineralocorticoid receptors in renal tubule cells. DIAGNOSIS AND DIFFERENTIAL DIAGNOSIS The syndrome of apparent mineralocorticoid excess caused by 11b-hydroxysteroid dehydrogenase deficiency can be readily diagnosed by determination of the urinary tetrahydrocortisol/tetrahydrocortisone ratio.56,57 Because plasma cortisol levels are normal, and because deoxycorticosterone and corticosterone levels, similar to those of aldosterone, may be low or undetectable, these patients are readily distinguished from those with the adrenogenital syndromes, who have 11b-hydroxylase deficiency or 17-hydroxylase deficiency. In those patients with 11b-hydroxylase deficiency, boys also show precocious puberty and girls show virilization, and those with 17-hydroxylase deficiency are, for the most part, phenotypic females with sexual infantilism. Furthermore, hydrocortisone, which improves the adrenogenital syndromes, worsens the syndrome of 11b-hydroxysteroid dehy-drogenase deficiency. Spironolactone produces improvement in patients with 11b-hydroxysteroid dehydrogenase deficiency as well as those with 11b-hydroxylase and 17-hydroxylase deficiency. This distinguishes these disorders from Liddle syndrome, which does not respond to the mineralocorticoid antagonist. TREATMENT Treatment with spironolactone decreases the blood pressure and corrects other abnormal features of apparent mineralocorticoid excess resulting from 11b-hydroxysteroid dehydrogenase deficiency.57 The effectiveness of dexamethasone, which has been reported to produce improvement in one patient but appeared to be ineffective in another, is uncertain. In patients with 11b-hydroxylase deficiency and 17-hydroxylase deficiency, treatment with hydrocortisone restores the overproduction of mineralocorticoids to normal levels. In Liddle syndrome, a potassium-sparing diuretic such as amiloride (510 mg daily) or triamterene (50100mg twice daily) is the drug of choice.54 Infants or extremely young children may require smaller dosages.

SECONDARY HYPERALDOSTERONISM
A variety of disorders of the kidney, heart, liver, and gastrointestinal tract may lead to hyperreninemia and, in turn, to hyperaldosteronism and hypokalemia. Because the blood pressure may be either normal or high, the disorders may be grouped on the basis of blood pressure. Disorders that may give rise to hypertension with secondary hyperaldosteronism are renal artery stenosis (see Chap. 82), renin-secreting tumor (see Chap. 219), malignant hypertension, and chronic renal disease (see Table 80-1). Occasionally, the renal disorder may not be apparent clinically, and the hypertension and hypokalemia may be mistaken for primary aldosteronism. Determination of plasma renin activity after overnight bedrest and after ambulation for 2 hours should demonstrate whether this hormone is suppressed or stimulated and make the distinction between primary and secondary hyperaldosteronism. In chronic renal disease with renal insufficiency, plasma renin activity may be increased or suppressed. If it is suppressed, aldosterone also tends to be suppressed, except when the potassium concentration is high. Patients with hyperreninemia, hyperaldosteronism, and hypertension should be evaluated carefully for renal artery stenosis or renin-secreting tumor because these causes of hypertension may be curable.61,62 Sampling of renal venous blood for the determination of plasma renin activity usually indicates whether the increase in renin release is bilateral or unilateral and, if it is unilateral, which kidney is the source.61 Arteriography then can be performed to identify the nature of the disorder and confirm its location. The development of percutaneous transluminal balloon dilation angioplasty for the treatment of renal artery stenosis has revolutionized the management of this disorder and stimulated a search for more effective ways to identify those patients with hypertension who have renal artery stenosis. The captopril test, in which the response of plasma renin activity and blood pressure to 50 mg captopril is determined over 60 minutes, reportedly has a high degree of sensitivity and specificity in untreated patients without renal parenchymal disease.63 Secondary hyperaldosteronism with normal blood pressure may occur in chronic renal disease, renal tubular acidosis, cardiac failure, cirrhosis of the liver, disorders of the gastrointestinal tract, and various renal tubulopathies (see Table 80-1). The mechanism that brings about an increase in renin release in each of these disorders is an alteration in cardiovascular function, which may be a decrease in either actual or effective circulating blood volume. Consideration of all these disorders in detail is beyond the scope of this review. The tubulopathies are discussed in some depth, but only those aspects of the other disorders that are responsible for secondary aldosteronism are noted. Renal tubular acidosis and, at times, chronic renal disease are characterized by an impairment in the renal tubular secretion of hydrogen ions, with hypokalemia and acidosis. The defect in hydrogen ion secretion results in a loss of sodium and potassium, with contraction of the volume of extracellular fluid and stimulation of the renin-angiotensin-aldosterone system.64 The increase in aldosterone secretion may intensify the renal potassium loss. In hepatic cirrhosis and cardiac failure, inadequate arterial filling appears to be the basis for increased renin release.65 The loss of potassium from the gastrointestinal tract or from the kidneys as a result of a disorder listed in Table 80-1 may cause severe potassium depletion (Fig. 80-10) and lead to hyperreninemia and hyperaldosteronism. Experimental evidence indicates that potassium depletion may decrease peripheral arterial resistance and increase pressor resistance of the arteries to angiotensin II, presumably by stimulating the synthesis of vasodilator prostaglandins, such as prostacyclin, in vascular tissue, because treatment with a prostaglandin synthetase inhibitor corrects the abnormalities.66,67 In hypokalemic animals, a dose of an angiotensin II antagonist that had no effect before the induction of potassium depletion produces a profound decrease in blood pressure.66 This suggests that hyperreninemia is a compensatory adjustment that opposes vasodilator stimuli and helps to maintain blood pressure in the hypokalemic state. The overproduction of prostaglandins characterizes the severe potassium depletion that is produced by gastrointestinal disorders and renal tubulopathies,68 and probably is the basis for the hyperreninemia, hyperaldosteronism, and pressor resistance associated with these disorders.

FIGURE 80-10. Severe potassium depletion caused by diuretics in a 58-year-old man. The serum potassium was 2.8 mEq/L. Note the prominent U waves (arrow), especially in lead V2. Although all investigators are not in accord, most feel that severe hypokalemia may progress to potentially lethal ventricular arrhythmias. An electrocardiogram such as seen in this patient often is an initial clue to the presence of hypokalemia and may occur in many of the conditions discussed in this chapter. (Courtesy of Dr. Steven Singh.)

BARTTER SYNDROME
CLINICAL FEATURES AND PATHOPHYSIOLOGY Bartter syndrome includes a number of disorders of tubular transport that may be familial or acquired as a result of a disease such as cystinosis.69,70 The syndrome may present in childhood as a failure to thrive, weakness, salt craving, polyuria, and polydipsia, or in adulthood as an incidental finding with little associated symptomatology. In the latter instance, the syndrome may have been present since childhood in a mild form that was not detected. Serum chloride and potassium levels are low and serum carbon dioxide content is normal or increased. The serum magnesium concentration may be normal or low, and creatinine clearance may be normal or decreased. Physiologic studies suggested that an impairment of chloride transport in the diluting segment of the nephron, presumably the thick ascending limb of the loop of Henle, was the cause of the syndrome in some patients.71 The defect in chloride reabsorption would impair the ability of the tubule not only to dilute, but also to concentrate tubular fluid, because it results in a decrease in medullary hypertonicity. A defect in chloride transport by the thick ascending limb also would decrease potassium reabsorption in this segment and stimulate its secretion distally by increasing the volume of tubular fluid delivered to the distal nephron, in which potassium secretion is flow-dependent.72 Molecular biology studies indicate that such patients may have a mutation in the gene encoding a renal chloride channel, CLCNKB.73 This channel represents a major pathway through which chloride travels from cells of the thick ascending limb across their basolateral membrane into the interstitium. Hypercalciuria may or may not be an associated finding. The Na+-K+ -2Cl- cotransporter also mediates the reabsorption of chloride, in addition to sodium and potassium, in the thick ascending limb.74 Mutations in the gene NKCC2 that encodes for this cotransporter produce a syndrome, apparent in the immediate postnatal period, characterized by failure to thrive, marked polyuria, and polydipsia.75 The delivery is frequently premature and complicated by polyhydramnios. Hypercalciuria is present and usually complicated by nephrocalcinosis and, presumably, is the basis for the impairment in urinary concentrating ability. Diluting ability appears to remain intact. Mutations in the gene ROMK that encodes for a potassium channel, which are also located in the luminal membrane of cells of the thick ascending limb, produce a syndrome clinically similar to that produced by mutations in NKCC2.76 ROMK encodes for an ATP-sensitive potassium channel, which recycles potassium that has been reabsorbed from tubular fluid by thick ascending limb cells (from the cells back into the tubular lumen). This process of potassium recycling ensures sufficient luminal potassium for Na+-K+ -2Cl transporter to function. Thus, an impairment in function of this channel has physiologic consequences similar to those produced by mutations in NKCC2. Mutations in the gene encoding for the thiazide-sensitive sodium-chloride cotransporter, TSC, located in the cells of the distal convoluted tubule produce a variant of Bartter syndrome also referred to as Gitelman syndrome.77 These patients differ from those with mutations in CLCNKB, NKCC2, and ROMK in that they have hypocalciuria and hypermagnesiuria with hypomagnesemia. Urinary diluting and concentrating abilities are usually intact. The impaired function that results from mutations in TSC and its associated phenotype appears to account for the majority of patients diagnosed as Bartter syndrome. The transport abnormalities previously discussed, by impairing sodium chloride reabsorption in the thick ascending limb or distal convoluted tubule, increase the delivery of these ions distally to the cortical collecting tubule, where increased sodium reabsorption leads to increased potassium secretion. Why impaired sodium chloride reabsorption in the thick ascending limb increases excretion of calcium but not that of magnesium, and why impaired sodium chloride reabsorption in the distal convoluted tubule decreases calcium excretion but increases that of magnesium is not yet understood. Renal potassium loss and consequent potassium depletion stimulate prostaglandin overproduction and, in turn, the abnormalities that once were thought to be unique to Bartter syndrome but now are recognized as the pathophysiologic results of potassium depletion. Although it is not possible to demonstrate a decrease in peripheral resistance in Bartter syndrome, as was done in experimental potassium depletion in dogs, evidence for an increase in vasodilator stimuli may be inferred from the following observations: 1. A high level of urinary prostaglandins, especially 6-keto-PGF1a, which is a metabolite of prostacyclin, consistent with an increase in vasodilator prostaglandins in vascular tissue.78 2. Increased pressor resistance to angiotensin II that is corrected only partially when plasma angiotensin II is restored to normal by converting enzyme inhibition.79 3. Blood pressure that is normal despite hyperreninemia and increased sympathoadrenal activity and that decreases markedly when an antagonist of angiotensin II is given.80,81 4. Correction of prostaglandin overproduction by treatment with a prostaglandin synthetase inhibitor corrects the hyperreninemia, the increased sympathoadrenal activity, the pressor resistance to angiotensin II and to norepinephrine, and the hypotensive response to an angiotensin II inhibitor.80,81,82 The vasodilative stimuli of prostaglandins and the increase in plasma bradykinin associated with them are opposed by compensatory increases in activity of the renin-angiotensin and sympathoadrenal systems that stabilize blood pressure, as shown in Figure 80-11.

FIGURE 80-11. Bartter syndrome. Increased synthesis of prostaglandins (PG), presumably prostacyclin, by vascular tissue is stimulated by low body potassium levels (K+) and leads to vasodilatation and an increase in pressor resistance. An increase in bradykinin, associated with the increase in PGs, augments the vasodilatation. Increases in activity of the renin-angiotensin and sympathoadrenal systems oppose vasodilatation and maintain blood pressure. (From Gill JR Jr. The role of chloride transport in the thick ascending limb in the pathogenesis of Bartter syndrome. Klin Wochenschr 1982; 60:1212.)

DIAGNOSIS AND DIFFERENTIAL DIAGNOSIS The clinical features of Bartter syndrome (hypokalemic alkalosis, hyperreninemia, and hyperaldosteronism with normal blood pressure) are not unique to this disorder because they may be caused by potassium loss from the gastrointestinal tract, as well as from the kidney. The clinical history may indicate the source of potassium loss when it is familial chloride diarrhea83 or a similar problem, but when the loss is caused by covert vomiting or laxative abuse, the physician must resort to other strategies to make the diagnosis. The measurement of daily urinary sodium, chloride, and potassium levels while patients are consuming constant diets containing 100 mEq per day of sodium and chloride may be helpful. If the potassium loss is caused by a tubulopathy, genetic evaluation may indicate the specific autosomal-recessive genetic trait. The presence of such a trait is suggested by stable values for urinary sodium, chloride, and potassium that approximate oral intake. If potassium loss is from the gastrointestinal tract, the amount and pattern of urinary electrolytes will reflect this. Overall, excretion will be considerably lower than intake and, if vomiting is the cause, the excretion of chloride will be disproportionately lower than that of sodium. Although the urinary potassium level usually also is decreased, it may be inappropriately high if alkalosis and hypokalemia are severe enough to impair potassium reabsorption by the thick ascending limb of the loop of Henle.84 Distinguishing between covert diuretic abuse and Bartter syndrome can be difficult. In addition to the clinical features of potassium depletion, diuretics such as furosemide and thiazides also decrease sodium chloride reabsorption by the distal tubule, thereby mimicking the transport abnormalities of Bartter syndrome. Screening the urine for diuretics may be helpful in reaching the correct diagnosis. To establish the diagnosis of the specific transport abnormality, the phenotype must be characterized by measurement of serum magnesium, urinary magnesium, and urinary calcium. Determination of maximal fractional free water clearance (CH2O/CIN) and concentrating ability provide additional information. The phenotypic patterns associated with the transport abnormalities demonstrated to date are shown in Table 80-4.

TABLE 80-4. A Comparison of Phenotypic Findings Caused by Mutations in Genes That Encode for a Renal Tubular Cotransporter or Channel That Has Been Identified in Patients Diagnosed with Bartter Syndrome

TREATMENT Supplementation with potassium chloride (120160 mEq per day of potassium) and, if hypomagnesemia is present, magnesium chloride (3050 mEq per day of magnesium) may be all that is required in some patients.85 If symptoms persist, a potassium-sparing diuretic such as amiloride or triamterene may be added to the regimen. Although prostaglandin synthetase inhibition may correct the overproduction of prostaglandins and, in turn, the prostaglandin-dependent abnormalities, it does not prevent potassium loss. Generally, the side effects of prostaglandin synthetase inhibitors outweigh the benefits, except in those patients with hypercalciuria in whom they may ameliorate renal calcium loss in addition to correcting hyperreninemia and hyperaldosteronism.85,86 CHAPTER REFERENCES
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. Ulick S. Diagnosis and nomenclature of the disorders of the terminal portion of the aldosterone biosynthetic pathway. J Clin Endocrinol Metab 1976; 43:92. Fanestil DD, Kipnowski J. Molecular action of aldosterone. Klin Wochenschr 1982; 60:1180. Kornel L, Kanamarlapudi N, Travers T, et al. Studies on high affinity binding of mineralo- and glucocorticoids in rabbit aorta cytosol. J Steroid Biochem 1982; 16:245. Sweiry JH, Binder HJ. Characterization of aldosterone-induced potassium secretion in rat distal colon. J Clin Invest 1989; 83:844. Irony I, Kater CE, Biglieri EG, Shackleton CHL. Correctable subsets of primary aldosteronism, primary adrenal hyperplasia and renin responsive adenoma. Am J Hypertens 1990; 3:576. Etker S, Enger EA. Aldosterone-producing adrenal adenoma in children. Eur J Pediatr Surg 1992; 2:370. Stowasser M, Gordon RD, Tunny TJ, et al. Familial hyperaldosteronism type II: five families with a new variety of primary aldosteronism. Clin Exp Pharmacol Physiol 1992; 19:319. Bravo EL, Tarazi RC, Dunstan HP, et al. The changing clinical spectrum of primary aldosteronism. Am J Med 1983; 74:641. Holland OB, Thomas C, Brown H, et al. Aldosterone suppression with dopamine infusion in low-renin hypertension. J Clin Invest 1983; 72:754. Yamaji T, Ishibashi M, Sekihara H, et al. Plasma levels of atrial natriuretic peptide in primary aldosteronism and essential hypertension. J Clin Endocrinol Metab 1986; 63:815. Higuchi K, Nawata H, Kato KI, et al. Lack of inhibitory effect of alpha-human atrial natriuretic polypeptide on aldosteronogenesis in aldosterone producing adenoma. J Clin Endocrinol Metab 1986; 63:192. Biglieri EG, Forsham PH. Studies on the expanded extracellular fluid and the response to various stimuli in primary aldosteronism. Am J Med 1961; 30:564. Kageyama Y, Bravo EL. Neurohumoral and hemodynamic responses to chronic intracerebroventricular infusion of aldosterone in conscious dogs.(Abstract.) Hypertension 1987; 10:373. Holland OB, Brown H, Lavon K, et al. Further evaluation of saline infusion for the diagnosis of primary aldosteronism. Hypertension 1984; 6:717. Zadik Z, Levin PA, Hamilton BP, Kowarski AA. Detection of primary aldosterone by the 6-hour integrated aldosterone-renin ratio. Hypertension 1986; 8:285. Lyons DF, Kem DC, Brown RD, et al. Single dose captopril as a diagnostic test for primary aldosteronism. J Clin Endocrinol Metab 1983; 57:892. Naomi S, Iwaoka T, Omeda T, et al. Clinical evaluation of the captopril screening test for primary aldosteronism. Jpn Heart J 1985; 26:549. Iwaoka T, Umeda T, Naomi S, et al. The usefulness of the captopril test as a simultaneous screening for primary aldosteronism and renovascular hypertension. Am J Hypertens 1993; 6:899. Vetter H, Fischer M, Galanski M, et al. Primary aldosteronism: diagnosis and noninvasive lateralization procedures. Cardiology 1985; 72(Suppl 1):57. Doppman JL, Gill JR Jr, Miller DL, et al. Distinction between hyperaldosteronism due to bilateral hyperplasia and unilateral aldosteronoma: reliability of CT. Radiology 1992; 184:677. Biglieri EG, Schambelan M. The significance of elevated levels of plasma 18-hydroxycorticosterone in patients with primary aldosteronism. J Clin Endocrinol Metab 1979; 49:87. Ganguly A, Dowdy A, Luetscher JA, Melada GA. Anomalous postural response of plasma aldosterone concentration in patients with aldosterone producing adrenal adenoma. J Clin Endocrinol Metab 1973; 36:401. Dunnick NR, Doppman JL, Gill JR Jr, et al. Localization of functional adrenal tumors by computed tomography and venous sampling. Radiology 1982; 142:429. Young WF Jr, Stanson AW, Grant CS, et al. Primary aldosteronism: adrenal venous sampling. Surgery 1996; 120:913. Bravo EL. Primary aldosteronism. Issues in diagnosis and management. Endocrinol Metab Clin North Am 1994; 23:271.

25a.Calvo-Romero JM, Ramos-Salado JL. Recurrence of adrenal aldosterone-producing adenoma. Postgrad Med J 2000; 76:160. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. Walther MM. Laparoscopic surgery for adrenal disease. Updates Principles and Practice of Oncology 1997; 11:Number 12. Loriaux DL, Menard R, Taylor A, et al. Spironolactone and endocrine dysfunction. Ann Intern Med 1976; 85:630. Nadler JL, Hsueh W, Horton R. Therapeutic effect of calcium channel blockade in primary aldosteronism. J Clin Endocrinol Metab1985; 60:896. Ferris JR, Beavers DG, Brown JJ, et al. Clinical, biochemical and pathological features of low-renin (primary) hyperaldosteronism. Am Heart J 1978; 95:375. Sen S, Bravo EL, Bumpus FM. Isolation of a hypertension-producing compound from normal human urine. Circ Res 1977; 40(Suppl 1):1. Carey RM, Sen S. Recent progress in the control of aldosterone secretion. Recent Prog Horm Res 1986; 42:251. Griffing GT, Berelowitz B, Hudson M, et al. Plasma immunoreactive gamma-melanotropin in patients with idiopathic hyperaldosteronism, aldosterone-producing adenomas, and essential hypertension. J Clin Invest 1985; 76:163. Griffing GT, McIntosh T, Berelowitz B, et al. Plasma b-endorphin levels in primary aldosteronism. J Clin Endocrinol Metab 1985; 60:315. Franco-Saenz R, Mulrow PJ, Kitai K. Idiopathic aldosteronism: a possible disease of the intermediate lobe of the pituitary. JAMA 1984; 251:2555. Shenker Y, Gross MD, Grekin RJ. Central serotonergic stimulation of aldosterone secretion. J Clin Invest 1985; 76:1485. Gross MD, Grekin RJ, Gniadek TC, Villareal JZ. Suppression of aldosterone by cyproheptadine in idiopathic aldosteronism. N Engl J Med 1981; 305:181. Arteaga E, Klein R, Biglieri EG. Use of the saline infusion test to diagnose the cause of primary aldosteronism. Am J Med 1985; 79:722. Banks WA, Kastin AJ, Biglieri EG, Ruiz AE. Primary adrenal hyperplasia: a new subset of primary hyperaldosteronism. J Clin Endocrinol Metab 1984; 58:783. Fontes RG, Kater CE, Biglieri EG, Irony I. Reassessment of the predictive value of the postural stimulation test in primary aldosteronism. Am J Hypertens 1991; 4:786. Griffing GT, Melby JC. The therapeutic effect of a new angiotensin-converting enzyme inhibitor, enalapril maleate, in idiopathic hyperaldosteronism. J Clin Hypertens 1985; 3:265. Brown RD, Kem DC, Hogan MJ, Hegstad RL. Evaluation of a test using saralasin to differentiate primary aldosteronism due to an aldosterone-producing adenoma from idiopathic hyperaldosteronism. Metabolism 1984; 33:734. Gill JR Jr, Bartter FC. Overproduction of sodium-retaining steroids by the zona glomerulosa is adrenocorticotropin-dependent and mediates hypertension in dexamethasone-suppressible aldosteronism. J Clin Endocrinol Metab 1981; 53:331. Liddle GW, Bledsoe I, Coppage WS. A familial renal disorder simulating primary aldosteronism but with negligible aldosterone secretion. Trans Assoc Am Physicians 1963; 76:199. George JM, Wright L, Bell NH, Bartter FC. The syndrome of primary aldosteronism. Am J Med 1970; 43:343. Fallo F, Kuhnle Y, Bossaro M, Sonino N. Abnormality of aldosterone and cortisol late pathways in glucocorticoid-remediable aldosteronism. J Clin Endocrinol Metab 1994; 79:772. Lifton RP, Dluhy RG, Powers M, et al. Hereditary hypertension caused by chimaeric gene duplications and ectopic expression of aldosterone synthase. Nature Genetics 1992; 2:66. Ulick S, Chan CK, Gill JR Jr, et al. Defective fasciculata zone function as the mechanism of glucocorticoid-remediable aldosteronism. J Clin Endocrinol Metab 1990; 71:1151. Ulick S, Land M, Chu MD. 18-Oxocortisol, a naturally occurring mineralo-corticoid agonist. Endocrinology 1983; 113:2320.

49. Speiser PW, Martin KO, Kao-Lo G, New MI. Excess mineralocorticoid receptor activity in patients with dexamethasone-suppressible hyperaldosteronism is under adrenocorticotropin control. J Clin Endocrinol Metab 1985; 6:129. 50. New MI, Peterson RE, Saenger P, Levine LS. Evidence for an unidentified ACTH induced steroid hormone causing hypertension. J Clin Endocrinol Metab 1976; 43:1283. 51. Rich GM, Ulick S, Cook S, et al. Glucocorticoid-remediable aldosteronism in a large kindred: clinical spectrum and diagnosis using a characteristic biochemical phenotype. Ann Intern Med 1992; 116:813. 52. Arteaga E, Biglieri EG, Kater CE, et al. Aldosterone-producing adrenocortical carcinoma: preoperative recognition and course in three cases. Ann Intern Med 1984; 101:316. 52a.Yoshimoto T, Naruse M, Ito Y, et al. Adrenocortical carcinoma manifesting pure primary aldosteronism: a case report and analysis of steroidogenic enzymes. J Endocrinol Invest 2000; 23:112. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. Edwards CR, Walker BR, Benediktsson R, Secki JR. Congenital and acquired syndromes of apparent mineralocorticoid excess. J Steroid Biochem Mol Biol 1993; 45:1. Liddle GW, Bledsoe I, Coppage WS. A familial renal disorder simulating primary aldosteronism but with negligible aldosterone secretion. Trans Assoc Am Physicians 1963; 76:199. Shimkets RA, Warnock DG, Bositis CM, et al. Liddles syndrome: Heritable human hypertension caused by mutations in the bsubunits of the epithelial sodium channel. Cell 1994; 79:407. Monder C, Shackleton CHL, Bradlow HC, et al. The syndrome of apparent mineralocorticoid excess: its association with 11b-dehydrogenase and 5b-reductase deficiency and some consequences for corticosteroid metabolism. J Clin Endocrinol Metab 1986; 63:550. DiMartino-Nardi J, Stoner E, Martin K, et al. New findings in apparent mineralocorticoid excess. J Clin Endocrinol Metab 1987; 27:49. Nikkila H, Tannin GM, New MI, et al. Defects in the HSD11 gene encoding 11 beta-hydroxysteroid dehydrogenase are not found in patients with apparent mineralocorticoid excess of 11-oxoreductase deficiency. J Clin Endocrinol Metab 1993; 77:687. Walker BR, Edwards CR. Licorice-induced hypertension and syndromes of apparent mineralocorticoid excess. Endocrinol Metab Clin North Am 1994; 23:359. Stewart PM, Valentino R, Wallace AM, et al. Mineralocorticoid activity of liquorice: 11-beta-hydroxysteroid dehydrogenase deficiency comes of age. Lancet 1987; 2:821. Pickering TG, Sos TA, Vaughan ED Jr, Laragh JH. Differing patterns of renal vein renin in patients with renovascular hypertension and their role in predicting the response to angioplasty of one or both renal arteries. Nephron 1986; 44(Suppl 1):8. Robertson JIS. Renin-secreting tumour. Contrib Nephrol 1984; 43:153. Muller FB, Sealey JE, Case DB, et al. The captopril test for identifying renovascular disease in hypertensive patients. Am J Med 1986; 80:633. Gill JR Jr, Bell NH, Bartter FC. Impaired conservation of sodium and potassium in renal tubular acidosis and its correction by buffer anions. Clin Sci 1967; 33:577. Gill JR Jr. Edema. Annu Rev Med 1970; 21:269. Galvez OG, Bay WH, Roberts BW, Ferris TF. The hemodynamic effects of potassium deficiency in the dog. Circ Res 1977; 40(Suppl 1):1. Gullner H-G, Graf AK, Gill JR Jr, Mitchell MD. Hypokalemia stimulates prostacyclin synthesis in the rat. Clin Sci 1983; 65:43. Gill JR Jr. Prostaglandins in Bartter's syndrome and in potassium-deficient disorders that mimic it. Miner Electrolyte Metab 1981; 6:76. Godard C, Vallotton MD, Broyer M, Roger P. A study of the inhibition of the renin-angiotensin system in renal potassium wasting syndromes, including Bartter's syndrome. Helv Paediatr Acta 1972; 27:495. Gitelman HJ. Unresolved issues in the pathogenesis of Bartter's syndrome and its variants. Curr Opin Nephrol Hypertens 1994; 3:471. Gill JR Jr, Bartter FC. Evidence for a prostaglandin-independent defect in chloride reabsorption in the loop of Henle as a proximal cause of Bartter's syndrome. Am J Med 1978; 65:766. Wright FC, Giebisch G. Renal potassium transport: contributions of individual nephron segments and populations. Am J Physiol 1978; 235:F515. Simon DB, Bundra RS, Mansfield TA, et al. Mutations in the chloride channel gene, CLCNKB, cause Bartter's syndrome type III. Nature Genet 1997; 17:171. Simon DB, Karet FE, Hamdan JM, et al. Bartter's syndrome, hypokalaemic alkalosis with hypercalciuria is caused by mutations in the Na-K-2Cl cotransporter NKCC2. Nature Genet 1996; 13:183. Seyberth HW, Rascher W, Schweer H, et al. Congenital hypokalemia with hypercalciuria in preterm infants: a hyperprostaglandinuric tubular syndrome different from Bartter's syndrome. J Pediatr 1985; 107:694. Simon DB, Karet FE, Rodriguez-Soriano J, et al. Genetic heterogeneity of Bartter's syndrome revealed by mutations in the K+ channel ROMK. Nature Genet 1996; 14:152. Simon DB, Nelson-Williams C, Bia MJ, et al. Gitelman's variant of Bartter's syndrome, inherited hypokalemic alkalosis, is caused by mutations in the thiazide-sensitive Na-Cl cotransporter. Nature Genet 1996; 12:24. Gullner H-G, Bartter FC, Cerletti C, et al. Prostacyclin overproduction in Bartter's syndrome. Lancet 1979; 2:767. Fujita T, Ando K, Sato Y, et al. Independent roles of prostaglandins and the renin and angiotensin system in abnormal vascular reactivity in Bartter's syndrome. Am J Med 1982; 73:71. Sasaki H, Okumura M, Asano T, et al. Responses to angiotensin II antagonist before and after treatment with indomethacin in Bartter's syndrome. BMJ 1977; 2:975. Gullner H-G, Gill JR Jr, Bartter FC, et al. Correction of increased sympathoadrenal activity in Bartter's syndrome by inhibition of prostaglandin synthesis. J Clin Endocrinol Metab 1980; 50:857. Bartter FC, Gill JR Jr, Frolich JL, et al. Prostaglandins are overproduced by the kidneys and mediate hyperreninemia in Bartter's syndrome. Trans Assoc Am Physicians 1976; 89:77. Pearson AJG, Sladen GE, Edmonds CJ, et al. The pathophysiology of congenital chloridorrhoea. Q J Med 1973; 167:453. Garella S, Chazan JA, Cohen JJ. Saline-resistant metabolic alkalosis or chloride-wasting nephropathy. Ann Intern Med 1970; 73:31. Gill JR Jr. Bartter's syndrome. In: Krieger DT, Bardin CW, eds. Current therapy in endocrinology and metabolism 19881989. St Louis: CV Mosby, 1988:153. Mourani CC, Sanjad SA, Akatcherian CY. Bartter syndrome in a neonate: early treatment with indomethacin. Pediatr Nephrol 2000; 14:143.

CHAPTER 81 HYPOALDOSTERONISM Principles and Practice of Endocrinology and Metabolism

CHAPTER 81 HYPOALDOSTERONISM
JAMES C. MELBY Isolated Hypoaldosteronism Primary Deficiency Inborn Errors Failure of Adrenal Glomerulosa Function Secondary Deficiency Syndrome of Hyporeninemic Hypoaldosteronism Hypoaldosteronism after Adrenalectomy for Aldosteronoma Pharmacologic Inhibition of Aldosterone Mineralocorticoid Resistance Pseudohypoaldosteronism Type I Amiloride-Sensitive Epithelial Sodium Channel Liddle Syndrome (Pseudoaldosteronism) Pseudohypoaldosteronism Type IMultiorgan The Two Types of Pseudohypoaldosteronism Type I Pseudohypoaldosteronism Type IRenal Treatment Pseudohypoaldosteronism Type II Chapter References

ISOLATED HYPOALDOSTERONISM
Isolated hypoaldosteronism, a selective deficiency of aldosterone secretion without alteration in cortisol production, results in a persistent hyperkalemia, which may be associated with profound muscle weakness and cardiac arrhythmias. Isolated hypoaldosteronism can result from inborn errors in aldosterone biosynthesis, failure of the zona glomerulosa owing to autoimmune adrenal disease in association with critical illness, altered function of the renin-angiotensin system in the syndrome of hyporeninemic hypoaldosteronism, unilateral adrenalectomy for an aldosterone-producing adenoma, or pharmacologic inhibition of aldosterone. Mineralocorticoid resistance occurs when there is a lack of response to aldosterone despite the presence of this hormone.1 PRIMARY DEFICIENCY INBORN ERRORS Inborn errors in the oxidation of corticosterone to form aldosterone have been described as corticosterone methyl oxidase type I deficiency (CMO I) and corticosterone methyl oxidase type II deficiency (CMO II). Corticosterone is at first hydroxylated and oxidized at the 18th position to yield aldosterone. This sequence of enzymatic events is seen in Figure 81-1. CMO I is also referred to as 18-hydroxylase; CMO II is also referred to as aldosterone synthase or aldosterone oxidase. CMO II deficiency is more common than CMO I deficiency. Some of the CMO I and CMO II enzymes have activity residing in the isozyme of steroid 11b-hydroxylase; this isozyme activity is limited to the zona glomerulosa. 18-Hydroxylase and 18-oxidase activities are required for the production of aldosterone. Mutations in the genes that encode the isozyme result in defective synthesis of aldosterone.

FIGURE 81-1. Terminal steps in the biosynthesis of aldosterone. Corticosterone methyl oxidase (CMO) type I and II deficiencies exhibit defects in the hydroxylase-oxidase reactions. In CMO type I deficiency, levels of the 18-hydroxylated product, 18-OH-corticosterone, are reduced. In CMO type II deficiency, 18-OH-corticosterone production is markedly increased. In both cases, aldosterone deficiency results.

Corticosterone Methyl Oxidase Type I Deficiency (18-Hydroxylase Deficiency). CMO I deficiency is extremely rare. Biochemically, it is characterized by a marked overproduction of corticosterone by the zona glomerulosa without a corresponding increase in 18-hydroxycorticosterone and a virtual absence of aldosterone. Scrutiny of some case reports casts doubt on whether they represent type I or type II deficiency because 18-hydroxycorticosterone measurement was not available. However, clinical results for a North American kindred in Pennsylvania, although attributed to CMO II deficiency, probably represent the type I variant.2,3 Corticosterone Methyl Oxidase Type II Deficiency (Aldosterone Synthase Deficiency). CMO II deficiency is inherited as an autosomal-recessive trait. This deficiency is rare but has been observed with an increased frequency among Jews of Iranian origin. The biomolecular studies have been extensive.4 In both CMO I and CMO II deficiencies, the severity of the clinical manifestations is inversely related to the age at diagnosis: it becomes less severe as the child ages. CMO II deficiency, if recognized clinically, has its onset between 1 week and 3 months of age and is characterized by severe dehydration, vomiting, and failure to grow. Hyponatremia, hyperkalemia, and metabolic acidosis are uniformly present. The plasma renin activity (PRA) is elevated, and plasma aldosterone levels are low. On the other hand, plasma 18-hydroxycorticosterone levels are markedly increased, and the 18-hydroxycorticosterone/aldosterone ratio in plasma exceeds 5. Also, the ratio of the urinary metabolite of 18-hydroxycorticosterone, 18-hydroxytetrahydroaldosterone, to the metabolite of aldosterone, tetrahydroaldosterone, also exceeds 5. In older children, adolescents, and adults, the abnormal steroid pattern described may be present and may persist throughout life without clinical manifestations. Mineralocorticoid (fludrocortisone) is given during infancy and early childhood, but this therapy does not have to be continued in most cases. Moreover, spontaneous normalization of growth can occur in patients who are untreated. It is not understood why aldosterone deficiency is so much more threatening in infancy than in adult life. It is particularly puzzling that isolated hypoaldosteronism caused by low renin secretion in aging patients has clinical significance, whereas patients with asymptomatic inherited hypoaldosteronism, resulting from either CMO I or CMO II deficiency, exhibit none of the manifestations of hyporeninemic hypoaldosteronism. These same observations are true of patients with pseudohypoaldosteronism (PHA). FAILURE OF ADRENAL GLOMERULOSA FUNCTION Autoimmume Adrenal Failure. As autoimmune adrenal failure evolves, selective aldosterone deficiency may emerge in the presence of preservation of zona fasciculata cell function. Although glucocorticoid responsiveness to corticotropin (ACTH), metyrapone, or insulin-induced hypoglycemia may be normal, PRA is elevated in the presence of a low or undetectable plasma corticotropin level. This is accompanied by mild metabolic acidosis and, occasionally, hyponatremia. During late stages of the disease, progression to panadrenal insufficiency may occur. A period of up to 1 year may separate the onset of the mineralocorticoid and glucocorticoid deficiencies.5 In selective aldosterone deficiency caused by autoimmune disease, antiadrenal antibodies can be detected. Mucocutaneous candidiasis and hypoparathyroidism may

occur concurrently,6 a form of multiple autoimmune endocrinopathy. A patient with idiopathic hemochromatosis, weakness, postural hypotension, and loss of libido was found to have mild glucose intolerance and low gonadotropins in addition to normokalemia, hyponatremia, and a modest elevation of urea. PRA was elevated, and aldosterone levels were suppressed. Cortisol response to ACTH and urinary 17-hydroxysteroid were normal. The patient failed to conserve sodium on a sodium-restricted diet. This suggests that isolated mineralocorticoid deficiency can occur secondary to glomerulosa cell failure caused by iron deposition.7 Hypoaldosteronism in Ill Patients. Hyperreninemic hypoaldosteronism can occur in critically ill patients, such as those in septic states, or in hemodynamically compromised subjects. Most of these patients have prolonged illness and hypotension of long duration, with or without hyperkalemia. The cortisol secretion is elevated, commensurate with the level of the stressful state. Because aldosterone, corticosterone, and 18-hydroxycorticosterone (but not cortisol) secretions become suppressed within 48 to 96 hours of continuous ACTH stimulation of the adrenal gland,8 prolonged ACTH secretion secondary to stress may impair 11b-hydroxylase and 18b-hydroxylase enzymes and may be the underlying mechanism of this syndrome. However, these patients have an increased plasma 18-hydroxycorticosterone/aldosterone ratio, and their aldosterone response to angiotensin II (A-II) infusion is impaired; this suggests that a selective inhibition of CMO II may play a role in the development of this disease. Because hypoxia is associated with increased renin activity and cortisol secretion and diminished aldosterone secretion,9 insufficiency of CMO II activity may be the result of hypoxia or other circulating factors affecting the zona glomerulosa. Autopsy studies in adrenal glands from patients with this syndrome have revealed atrophy or necrosis involving most of the layers of the adrenal gland, including the zona glomerulosa. However, an isolated necrosis of the zona glomerulosa has not been found; this is probably because blood flows from the outer cortex toward the medulla, and hypoper-fusion would affect the inner corticomedullary region first.10 Hyperreninemic hypoaldosteronism may be caused by the release of various cytokines in chronic illness. Tumor necrosis factor (TNF) and interleukin-1 (IL-1), which act both pyrogenically and catabolically, have been found to be elevated in body fluids during severe infection, and TNF and IL-1 are known to inhibit the stimulatory actions of A-II and ACTH on aldoster-one secretion. However, the action of potassium on aldosterone release is not altered. Although 12-hydroxyeicosanoic acid stimulates aldosterone and is a second messenger of A-II, TNF inhibits A-II stimulation of 12-hydroxyeicosanoic acid release.11 It is plausible that high circulating levels of atrial natriuretic hormone (ANH) during illness contribute to the aldosterone level in these patients. ANH is a powerful suppresser of aldosterone secretion both in vitro and in vivo. It suppresses aldosterone secretion in humans despite upright posture or A-II infusion.12 In cultured adrenal cells, it also blunts the aldosterone secretion that ordinarily occurs after stimulation by potassium, A-II, and ACTH.13 The ANH may be stimulated by subclinical volume expansion resulting from mildly reduced renal function, congestive heart failure, atrial arrhythmias, or other subclinical cardiac disease associated with atrial distension. In addition, many critically ill patients are on medications that may interfere with the renin-angiotensin-aldosterone axis (see later in this chapter) and consequently contribute to the reduced aldosterone production. Because no major clinical complications are reported with this form of hypoaldosteronism, it seldom requires treatment other than avoidance of drugs that may exacerbate the hypoaldosteronism. SECONDARY DEFICIENCY SYNDROME OF HYPORENINEMIC HYPOALDOSTERONISM The syndrome of hyporeninemic hypoaldosteronism (SHH), also referred to as distal renal tubular acidosis type 4, is common; it usually occurs in middle-aged and elderly patients (median age, 68 years) and in men more often than women. Diabetes mellitus occurs in less than half the patients; chronic renal insufficiency is present in 80% of patients. The condition is frequent in patients with tubulointerstitial forms of renal disease, but it has been described in virtually every type of renal abnormality.14 Fifty percent to 70% of patients with unexplained hyperkalemia and renal disease and a glomerular filtration rate that is sufficient to sustain normokalemia are found to have SHH.15 Most patients with this syndrome have low PRA and aldosterone that cannot be stimulated by appropriate maneuvers. Hyperchloremic metabolic acidosis occurs in less than 70% of cases, and mild to moderate hyponatremia is found in half of the patients. Importantly, hyperkalemia is observed in all subjects.16 These patients exhibit decreased fractional excretion of potassium in relation to glomerular filtration rate and a reduced response to kaliuretic stimuli (including sodium bicarbonate, sodium sulfate, and diuretics) as well as to intravenous potassium chloride. The hyperkalemia is out of proportion to the degree of renal insufficiency. Proposed mechanisms include hyporeninemia caused by a damaged juxtaglomerular apparatus, sympathetic insufficiency, altered renal prostaglandin production, or impaired conversion of prorenin to renin.17 Low renin production does not appear to be the sole factor because the PRA may be inappropriately normal in some patients. Certain cases may be explained by sodium retention leading to volume expansion and thus to a secondary suppression of renin and aldosterone.18 The leading causes of interstitial nephritis, in which hyperkalemia can occur early and before chronic renal failure, are anatomic genitourinary abnormalities, analgesic abuse with aspirin or phenacetin, hyperuricemia, nephrocalcinosis, nephrolithiasis, and sickle cell disease. Diabetic patients are predisposed to hyperkalemia because of insulin deficiency and hyperglycemia. Interestingly, both insulin deficiency and hyperglycemia can independently produce a maldistribution of total body potassium. Hyperglycemia results in extracellular hyperosmolality, producing an extracellular flux of potassium. Furthermore, insulin deficiency prevents the cellular uptake of potassium, presumably related to the metabolic actions of this hormone. Autonomic insufficiency, which can be a complication of diabetes, results in hyporeninemic hypoaldosteronism, and the degree of autonomic neuropathy correlates with the duration of hyperglycemia.19 Immunoglobulin M monoclonal gammopathy has been associated with nodular glomerulosclerosis, a concentrating defect, and hyporeninemic hypoaldosteronism. The hypoaldosteronism is associated with decreasing renal function, suggesting that k light-chain nephropathy is a cause of this syndrome.20 Patients with the acquired immunodeficiency syndrome may have persistent hyperkalemia secondary to either adrenal insufficiency or, less frequently, hyporeninemic hypoaldosteronism. The hypoaldosterone patients usually have adequate aldosterone stimulation, suggesting that inadequate renin is the cause of the hypoaldosteronism.21 Treatment. No ideal medical therapy has been established for SHH. Most patients with mild selective hypoaldosteronism require no therapy. With preventive measures and education of patients, therapy may be avoided. The decision to treat SHH and the selection of specific therapeutic agents depend on a number of factors, including the degree of hyperkalemia, the extent of renal insufficiency, the presence of diabetes mellitus, the level of blood pressure, and the status of sodium balance. Once the diagnosis of SHH has been made, factors that precipitate or perpetuate suppression of renin biogenesis, aldosterone biogenesis, or both should be avoided (see Chap. 79 and Chap. 183). Reducing the extracellular potassium load can be the single most effective preventive measure in controlling hyperkalemia in SHH. Reducing dietary intake of potassium is helpful. Low sodium foods and use of salt substitutes, which often contain potassium as the alternative cation (such as low-salt milk, which contains 60 mEq/L potassium), should be avoided. Examples of foods high in potassium include dried fruits (30 mEq/cup), meat (60 mEq/lb), and decaffeinated coffee (4 mEq/cup). Other sources of potassium include transfusions of bank blood (30 mEq/L) and high-dose penicillin (1.7 mEq/106 U). The long-term control of glucose homeostasis in diabetes mellitus may reduce the risk of developing SHH. Perhaps autonomic insufficiency is avoidable in well-controlled diabetics. Because many medications can interfere at multiple points in the renin-aldosterone axis, avoidance of these drugs can be of major importance. b-Adrenergic receptor blockers, prostaglandin synthetase inhibitors, and potassium-sparing diuretics should be avoided in patients with known SHH and in diabetic patients with latent hypoaldosteronism. Calcium-channel blockers, antidopaminergic agents, and drugs that impair adrenal function should be used with caution. Patients on angiotensin-converting enzyme inhibitors should be carefully monitored for hyperkalemia. Prolonged administration of heparin should be avoided because it can worsen the hypoaldosteronism and it has been associated with lethal hyperkalemia. In severe SHH, fludrocortisone acetate is used in dosages of 0.1 to 1.0 mg per day, which is equivalent to 200 to 2000 g of aldosterone daily. Ninety percent of patients become normokalemic on fludrocortisone, which, however, carries the risk of salt retention and hypertension and edema. Diuretics may be the cardinal therapy for patients with SHH and coexisting diseases associated with sodium retention. Older patients with hypertension, mild renal impairment, and congestive heart failure respond better to diuretic therapy than to mineralocorticoid replacement. Because kaliuresis is the goal of diuretic therapy, the choice of a potent kaliuretic drug is of the utmost importance. Chlorthalidone and hydrochlorothiazide, a slightly less kaliuretic agent, meet this requirement. The loop diuretics, such as furosemide and ethacrynic acid, are less potent kaliuretic agents and induce a greater degree of natriuresis. Another potential benefit of diuretic therapy is stimulation of residual renin release in patients with hyporeninism caused by autonomic insufficiency. Sodium bicarbonate cannot be recommended as routine therapy in the treatment of hyporeninemic hypoaldosteronism because it is hazardous in elderly patients with coexisting renal impairment, congestive heart failure, and hypertension.

Sodium polystyrene sulfonate (Kayexalate), a cation exchange resin, removes ~1 mEq/g potassium by exchanging sodium for potassium in a ratio of 1.0 to 1.5. Thus, this drug increases sodium load and may be contraindicated in patients unable to tolerate an increase of this cation. Calcium-exchange resins are under investigation and may become available as an alternative therapy. HYPOALDOSTERONISM AFTER ADRENALECTOMY FOR ALDOSTERONOMA Hypoaldosteronism from chronic volume expansion was reported in postadrenalectomy patients after surgical excision of an aldosterone-producing adenoma.22 These patients develop severe hyperkalemia and hypotension lasting several days to several weeks after surgery. Similar forms of hypoaldosteronism due to volume expansion were also reported in patients chronically ingesting sodium bicarbonate (baking soda) for the treatment of gastrointestinal symptoms. PHARMACOLOGIC INHIBITION OF ALDOSTERONE Pharmacologic agents such as cyclosporine, heparin sodium, and calcium-channel blockers specifically inhibit aldosterone biogenesis by the zona glomerulosa of the adrenal gland23 (Table 81-1). Cyclosporine A produces hypoaldosteronism by dual effects on the adrenal cortex; first, an acute blockade of the A-IIinduced aldosterone production, and second, an inhibition of growth and steroidogenic capacity of adrenocortical cells. The latter effect may be caused by an impairment of protein synthesis. Polysulfated glycosaminoglycans, such as heparin sodium, impair aldosterone biosynthesis from the zona glomerulosa. With prolonged administration, heparin sodium can produce significant hypoaldosteronism with severe hyperkalemia because of a direct toxic effect on the zona glomerulosa, evidenced by a hyperreninemic hypoaldosteronism and zona glomerulosa atrophy.24 The least toxic dose of heparin sodium is unknown, but a dose of 20,000 U per day for 5 days has been observed to reduce aldosterone secretion. This is an uncommon cause of hypoaldosteronism, but it may result in fatal hyperkalemia. This effect appears to be caused by chlorbutanol, which is used as a preservative with heparin, rather than by the heparin itself. Heparin-induced aldosterone deficiency has been used therapeutically in some patients with chronic glomerulonephritis and initial hyperaldosteronism.25 Calcium-channel blockers inhibit aldosterone biosynthesis and, under certain clinical conditions, may cause hypoaldosteronism by lowering aldosterone secretion through inhibition of calcium influx.

TABLE 81-1. Pharmacologic Agents That Induce Direct or Indirect Inhibition of Aldosterone or of Aldosterone Effects

b-Blockers and prostaglandin synthetase inhibitors are frequent causes of hyporeninemic hypoaldosteronism. Juxtaglomerular cells, which synthesize and secrete renin, contain b-adrenergic receptors; either intrinsic neuronal or extrinsic adrenergic stimuli trigger these receptors, resulting in an immediate release of renin. As a result, b-blocking adrenergic drugs interfere with renin secretion and may cause hypoaldosteronism. Prostaglandin synthetase inhibitors, which specifically inhibit cyclooxygenase, block renin release and can result in severe hyperkalemia caused by hyporeninemic hypoaldosteronism. Prostaglandin E2 (PGE2) directly stimulates renin release, probably by a direct action on the juxtaglomerular apparatus. Furosemide-induced renin release is blunted by indomethacin and other prostaglandin synthetase inhibitors.26 Angiotensin-converting enzyme inhibitors and potassium-sparing diuretics may contribute to the hypoaldosteronism and hyperkalemia of various conditions. Angiotensin-converting enzyme inhibitors act by inactivating angiotensin-converting enzyme, which interrupts the renin-aldosterone axis and results in iatrogenic hypoaldosteronism.27 Spironolactone has two effects: it is a mineralocorticoid receptor antagonist, and it inhibits aldosterone biosynthesis, presumably by competing with corticosteroid biogenesis.28 Triamterene produces potassium retention by a direct action on nonaldosterone-mediated distal tubular exchange sites.29 Amiloride acts on the luminal surfaces of epithelial membranes to block sodium channels, resulting in less sodium resorption but diminished potassium secretion. Drugs that impair adrenal function are increasingly used for the hormonal treatment of breast cancer and medical management of Cushing syndrome; these agents can cause hypoaldosteronism. Aminoglutethimide, metyrapone, and trilostane block various enzymatic steps in the synthesis of mineralocorticoids, glucocorticoids, and adrenal sex steroids. Lower doses of these drugs may not be associated with hyperkalemia because secretion of aldosterone precursors, such as deoxycorticosterone, may confer significant mineralocorticoid activity.27 Drugs affecting the dopaminergic system produce significant alterations in aldosterone secretion. It is believed that aldosterone is under tonic dopamine inhibition; thus, the administration of dopaminergic agonists, such as bromocriptine, may impair aldosterone secretion in certain physiologic situations.27

MINERALOCORTICOID RESISTANCE
Mineralocorticoid resistance implies a lack of response to aldosterone despite its presence. Aldosterone binds to intracellular mineralocorticoid receptors, which interact with DNA. To influence gene transcription and subsequent synthesis of protein such as Na+ /K+-ATPase on the basolateral surface of renal epithelial cells and the amiloride-sensitive epithelial sodium channel (ENaC) on the apical membrane,1 ENaC-mediated entry of sodium into the cell represents the rate-limiting step for the reabsorption of sodium. PSEUDOHYPOALDOSTERONISM TYPE I Pseudohypoaldosteronism type I (PHA I) is a rare inherited salt-wasting disorder that was first described in 1958 as a defective renal-tubular response to mineralocorticoid in infancy.30 Patients present in the neonatal period with dehydration, hyponatremia, hypokalemia, metabolic acidosis, and failure to thrive despite normal glomerular filtration and normal renal and adrenal function.5 When patients fail to respond to mineralocorticoid therapy, PHA I should be considered the underlying disorder. Diagnosis includes an elevated plasma aldosterone level and increased plasma renin activity. PHA I can be inherited either as a recessive or dominant trait. Neonates affected with PHA I have a primary defect that affects the renal reabsorption of sodium.30a Further analysis of PHA I has revealed that it can be divided into two distinct disorders with unique physiologic and genetic characteristicsthe renal form of PHA I and a multiorgan form of PHA I. The renal form of PHA I follows mendelian inheritance and is transmitted in an autosomal-dominant manner. This disease is limited to a mineralocorticoid resistance only in the kidneys. Interestingly, the patient's condition spontaneously improves within the first several years of his or her life, thus allowing discontinuation of therapy. The multiorgan or generalized form of PHA I also follows mendelian inheritance but is transmitted as an autosomal-recessive trait.31 There is a high incidence of consanguinity in the families; the parents of these patients usually have physiologically normal levels of aldosterone and renin.31 The two main characteristics that distinguish this form of PHA I from the renal form are (a) the patient has a multiorgan disorder such that mineralocorticoid resistance is not limited to the kidneys (it can be seen in the kidney, sweat and salivary glands, and the colonic mucosa), and (b) this condition does not spontaneously improve with age31,31a; hence, it is considered to be more severe. Because sodium reabsorption is coupled to potassium and hydrogen ion secretion, patients often exhibit decreased potassium- and hydrogenion secretion with decreased sodium reabsorption; hence, potassium and hydrogen ions accumulate in the body, ultimately causing hyperkalemia and metabolic acidosis. Moreover, a decrease in vascular volume is detected by vessels in the juxtaglomerular region, leading to increased secretion of renin. Elevated plasma renin levels will ultimately cause an increase in aldosterone secretion through the indirect impact of angiotensin II. Originally, it had been thought that PHA I, analogous to a steroid-hormone resistance syndrome, might result from a defect in the mineralocorticoid receptor, either because of an absence or deficiency of receptors or because of an abnormality in the structure of the receptor (such as, perhaps, a defect in the aldosterone-binding

domain). The mineralocorticoid-receptor gene was mapped to chromosome 4 by somatic cell hybridization and regionally localized to 4q31.1-31.2 by fluorescent in situ hybridization. Although evidence supporting a defect in the gene coding for the mineralocorticoid receptor was found in some studies, most studies have been inconsistent.32 Although the involvement of the mineralocorticoid receptor in PHA I was not completely ruled out, it has been hypothesized that PHA I may be a heterogeneous condition resulting from a disorder either at the prereceptor level or at the postreceptor level. At the prereceptor level, there may be some factor that competes for the aldosterone-binding site or causes the cleavage of the mineralocorticoid receptor. At the postreceptor level, it was hypothesized that PHA I might result from a defect in one or both of the aldosterone-induced proteins: the Na+/K+-ATPase and/or the ENaC.31,31a AMILORIDE-SENSITIVE EPITHELIAL SODIUM CHANNEL The ENaC is a highly selective sodium channel found at the apical surface of salt-reabsorbing tight epithelia of tissues, including distal nephrons, the distal colon, salivary and sweat glands, lung, and taste buds.33 As described earlier, it plays a critical role in the control of sodium balance, extracellular fluid volume, and blood pressure, because the ENaC-mediated entry of sodium into the cell in these epithelia represents the rate-limiting step for the movement of sodium from the mucosal side to the serosal side.34 These channels allow the transport of sodium into the cell by diffusion without coupling to the flows of other solutes and without the direct input of metabolic energy.33 ENaCs are often referred to as amiloride-sensitive because of their high sensitivity to the potassium-sparing diuretic amiloride and its analogs. These channels are directly stimulated by aldosterone and inhibited by amiloride.33 ENaCs are composed of three subunits: a, b, and g, which are 35% homologous at the amino-acid level and are conserved throughout evolution.35 Moreover, the three subunits are similar in structure and share the following characteristics: short intracellular amino-acid carboxyl termini, two transmembrane-spanning domains, and a large extracellular loop. The genetic locus of the gene coding for the a subunit of ENaC has been traced to chromosome 12, and the genes coding for the b and g subunits have been traced to chromosome 16.35 The a subunit has been regionally localized to 12p13.1pter, and the a and g subunits have been regionally localized to 16p12.213.11.31 Of the three, the a subunit is thought to be the most important for the proper functioning of ENaC. LIDDLE SYNDROME (PSEUDOALDOSTERONISM) (In 1994, evidence was found explaining the underlying cause of Liddle syndrome, a disorder for which the physiologic symptoms are opposite to those of PHA I.36,37 This condition, also known as pseudoaldosteronism, is an autosomal-dominant form of hypertension characterized by volume expansion, hypokalemia, and alkalosis. Using genetic analysis, the illness was traced to two distinct mutations in the genes coding for the subunits of the ENaC. It was found that it is caused by mutations in the genes coding for the b and g subunits of ENaC, which cause truncation of the cytoplasmic carboxyl termini that, in turn, lead to a constitutive activation of ENaC activity.36 It is now believed that these mutations increase the number of sodium channels in the apical membrane by deleting a critical consensus motif.) PSEUDOHYPOALDOSTERONISM TYPE IMULTIORGAN Recognizing that PHA I may result from a loss of ENaC function, researchers have attempted to prove that mutations coding for the b, g, and a subunits of ENaC (similar but opposite to those seen in Liddle syndrome) may render the ENaC nonfunctional and unresponsive to mineralocorticoids. Three mutations have been found that result in a loss of ENaC function in neonates with the multiorgan type of PHA I.31 These mutations better define the differences between the two forms of PHA I. Two of the mutations involve the subunit, and one mutation involves the b subunit.31 On the a subunit, a two base-pair deletion at codon 168 introduces a frameshift mutation and thus disrupts the protein before the first transmembrane domain. The other mutation of the a subunit is a single-base substitution at codon R508 that changes a cytosine base to thymine. This truncates the a subunit before the second transmembrane domain by introducing a premature termination codon in the extracellular domain. The result is normal production of the first transmembrane domain, but no production of a second domain or of the intracytoplasmic carboxyl termini, and an incomplete production of the extracellular domain. THE TWO TYPES OF PSEUDOHYPOALDOSTERONISM TYPE I The a subunit is the most important of the three subunits and is required for normal ENaC activity.34 Thus, it is reasonable to assume that an a subunit not containing all its domains may lead to a nonfunctional channel protein. Both of these mutations lead to a loss of ENaC activity because both transmembrane domains are required for normal channel activity. On the b subunit, a point mutation causes the substitution of serine for glycine at amino acid 37. The glycine at position 37 on the b subunit ENaC lies within a segment preceding the first transmembrane spanning region that is homologous among all members of the ENaC gene family. This mutation on the b subunit has been shown to diminish ENaC activity but does not lead to a complete loss of activity; after this mutation, ENaC activity is still higher than it would be in the absence of a b subunit.36a The gating of ENaC involves a complex mechanism that remains poorly understood; however, it has been suggested that the highly conserved 20 amino-acid segment containing the conserved glycine may control the normal gating of ENaC such that the glycine-to-serine mutation on the b subunit would alter the open probability of ENaC.34 The glycine-to-serine mutation may impair the normal functioning of the ENaC gating domain and shift the equilibrium between two different gating modes, thus favoring a mode with short open times and long closed times. Moreover, in addition to being involved in the gating of ENaC, glycine may be part of a functional domain of ENaC that is regulated directly or indirectly by aldosterone.34 As stated previously, ENaC is stimulated by the presence of aldosterone; thus, the previous hypothesis would help to explain why a PHA I patient with this mutation on the b subunit fails to respond to the body's elevated levels of aldosterone and to exogenously administered mineralocorticoids.8 Two additional mutations on the g subunit of ENaC further elucidate the cause of the autosomal-recessive form of PHA I.36 One mutation involves a replacement of a highly conserved amino acid triplet Lys-Tyr-Ser by Asn in the extracellular loop immediately adjacent to the transmembrane domain. The other mutation is a 300-nucleotide deletion in the extracellular loop that introduces a stop codon before the second transmembrane domain; both of these mutations are thought to cause either a complete or a partial loss of ENaC activity. PSEUDOHYPOALDOSTERONISM TYPE IRENAL Once the genetic basis and pathophysiology of the autosomal-recessive or multiorgan form of PHA I had been defined, and studies had shown that ENaC was probably not involved in the pathophysiology of the autosomal-dominant or renal form of PHA I, attention turned to the mineralocorticoid receptor. Although some studies have found evidence supportive of a role for the mineralocorticoid receptor as a candidate for PHA I, other studies have failed to find functional mutations in the gene coding for the mineralocorticoid receptor. Retrospectively, it is noteworthy that all but one of these studies investigated patients with the autosomal-recessive form of PHA I, which has since been linked to the ENaC. Two frameshift mutations have been identified that may help to explain the underlying cause of the renal form of PHA I37,38; sequence analysis revealed two single base-pair deletions on exon 2, one at codon 335 and the other at codon 459. Both of these deletions result in frameshifts that lead to a gene product lacking the entire DNA- and hormone-binding domains, as well as the dimerization motif.37 Improvement of Renal Pseudohypoaldosteronism Type I with Age. A plausible hypothesis for the improvement of the renal form of PHA I with age is that the very important early role that aldosterone plays in salt homeostasis diminishes later in life because of a transition to a higher salt intake diet. Breast milk is very low in sodium, and aldosterone levels are elevated in infancy to increase sodium reabsorption. Only those infants in whom salt homeostasis is stressed by an intercurrent illness with volume depletion actually will manifest clinically recognizable PHA I.38 Nevertheless, the phenomenon of improvement with age requires further study. TREATMENT PHA I patients are resistant to mineralocorticoid therapy; therefore, standard treatment includes supplementation with sodium chloride (28 g per day) and cation exchange resins. This usually corrects the patient's biochemical imbalance. However, if a patient shows signs of severe hyperkalemia, peritoneal dialysis may be necessary. For those PHA I patients in whom hypercalciuria is found, the recommended course of treatment usually involves treatment either with indomethacin or with hydrochlo-rothiazide. Indomethacin is thought to act by causing a reduction in the glomerular filtration rate and/or an inhibition of the effect of prostaglandin E2 on renal tubules.39 Indomethacin reduces polyuria, sodium loss, and the hypercalciuria. (Hydrochlorothiazide, which is a potassium-losing diuretic that sometimes is

administered to diminish hyperkalemia, also has been shown to reduce hypercalciuria in PHA I patients.39) In patients with the autosomal-dominant or renal form of PHA I, the signs and symptoms decrease with age; nevertheless, salt supplementation is required for the first 2 to 3 years of life.14 In patients with the autosomal-recessive or multi-organ type of PHA I, however, resistance to therapy with sodium chloride or drugs that decrease serum potassium concentrations often occurs and may even lead to death in infancy of hyperkalemia.40 These patients often require very high amounts of salt in their diet (as much as 45 g NaCl per day).41 Carbenoxolone (CBX), a derivative of glycyrrhetinic acid in licorice, is moderately successful in helping to reduce the high-salt requirement in the diets of renal PHA I patients (however, it is not useful in treating multiorgan PHA I patients).41 CBX acts by inhibiting 11b-hydroxysteroid dehydrogenase (11b-HSD) activity. This enzyme, which is present at high levels in aldosterone-responsive target tissues, exerts a mineralocorticoid effect mainly in patients with the renal form of PHA I. Normally, 11b-HSD has an antimineralocorticoid effect, because it converts cortisol into its inactive form, cortisone. However, by inhibiting this enzyme, CBX allows unmetabolized cortisol to bind to and activate miner-alocorticoid receptors in a manner similar to that of aldosterone.41 PSEUDOHYPOALDOSTERONISM TYPE II Pseudohypoaldosteronism type II (PHA II), also known as familial hyperkalemia and hypertension or Gordon syndrome, is a non-salt-wasting disorder characterized by hyperkalemia despite normal renal glomerular filtration and hypertension.42 Hyperchloremia, metabolic acidosis, and suppressed plasma renin activity variably occur. Linkage analysis reveals autosomal-dominant transmission and locus heterogeneity of this trait to chromosomes 1q31-q42 and 17p11-q21.15. Although the pathogenesis has not been completely resolved, clinical studies suggest possible primary defects including (a) a recessive renal-sodium reabsorption proximal to the site of aldosterone action, (b) a generalized membrane defect impairing movement of potassium ions into cells, or (c) an increased renal-chloride reabsorption that in turn would impair potassium secretion. Treatment with a combination of furosemide, desmopressin (DDAVP), and sodium bicarbonate has been effective in controlling the hyperkalemia, hypertension, hyperchloremia, and hyperchloremic metabolic acidosis.42 CHAPTER REFERENCES
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. White PC. Disorders of aldosterone biosynthesis and action. N Engl J Med 1994; 331:250. Veldhuis JD, Kulin HE, Santen RJ, et al. Inborn error in the terminal step of aldosterone biosynthesis. N Engl J Med 1980; 303:117. Ulick S, Wang J, Morton D. The biochemical phenotypes of two inborn errors in the biosynthesis of aldosterone. J Clin Endocrinol Metab 1992; 72:1415. White PC, Pascoe L. Disorders of steroid 11b-hydroxylase isozymes. Trends Endocrinol Metab 1992; 3:229. Kokko JP. Primary acquired hypoaldosteronism. Kidney Int 1985; 27:690. Marieb NJ, Melby JC, Lyall SS. Isolated hypoaldosteronism associated with idiopathic hypoparathyroidism. Arch Intern Med 1974; 134:424. Thomas JP. Aldosterone deficiency in a patient with idiopathic haemochromatosis. Clin Endocrinol (Oxf) 1984; 21:271. Aguilera G, Fujita K, Catt KJ. Mechanism of inhibition of aldosterone secretion by adrenocorticotropin. Endocrinology 1981; 108:522. Slater JDH, Tuffey RE, Williams ES, et al. Control of aldosterone secretion during acclimatization to hypoxia in man. Clin Sci 1969; 37:327. Davenport MW, Zipser RD. Association of hypotension with hyperreninemic hypoaldosteronism in critically ill patient. Arch Intern Med 1983; 143:735. Antonipillai I, Wang Y, Horton R. Tumor necrosis factor and interleukin-1 may regulate renin secretion. Endocrinology 1990; 126:273. Clark BA, Brown RS, Epstein FH. Effect of atrial natriuretic peptide on potassium-stimulated aldosterone secretion: potential relevance to hypoal-dosteronism in man. J Clin Endocrinol Metab 1992; 75:399. Rocco S, Opocher G, D'Agostino D, et al. Lack of aldosterone inhibition by atrial natriuretic factor in primary aldosteronism: in vitro studies. J Endocrinol Invest 1989; 12:13. Schambelan M, Sebastian A, Biglieri EG. Prevalence, pathogenesis and functional significance of aldosterone deficiency in hyperkalemic patients with chronic renal insufficiency. Kidney Int 1980; 17:89. DeFronzo RA. Hyperkalemia and hyporeninemic hypoaldosteronism. Kidney Int 1980; 17:118. Batlle DC. Hyperkalemic hyperchloremic metabolic acidosis associated with selective aldosterone deficiency and distal renal tubular acidosis. Semin Nephrol 1981; 1:260. Melby JC, Griffing GT. Isolated hypoaldosteronism. In, Kassier JP, ed: Current therapy in internal medicine. Philadelphia: BC Decker, 1991:1328. Holland OB. Hypoaldosteronism: disease or normal response? N Engl J Med 1991; 324:488. Jost-Vu E, Horton R, Antonipillai I. Altered regulation of renin secretion by insulin-like growth factors and angiotensin II in diabetic rats. Diabetes 1992; 41(9):1100. Nakamoto Y, Imai H, Hamanaka S, et al. IgM monoclonal gammopathy accompanied by nodular glomerulosclerosis urine concentrating defect and hyporeninemic hypoaldosteronism. Am J Nephrol 1985; 5:53. Kalin MF, Poretsky L, Seres DS, Zumoff B. Hyporeninemic hypoaldosteronism associated with acquired immune deficiency syndrome. Am J Med 1987; 82:1035. Biglieri EG, Slaton PE Jr, Silen WS, et al. Post-operative studies of adrenal function in primary aldosteronism. J Clin Endocrinol Metab 1966; 26:553. Sebastian A, Schambelan M, Lindenfeld S, Morris RC Jr. Amelioration of metabolic acidosis with fludrocortisone therapy in hyporeninemic hypoaldosteronism. N Engl J Med 1977; 297:576. Rimmer JM, Horn JF, Gennari FJ. Hyperkalemia as a complication of drug therapy. Arch Intern Med 1987; 147:867. Kutyrina IM, Nikshova TA, Tareyeva IE. Effects of heparin-induced aldosterone deficiency on renal function in patients with chronic glomerulonephritis. Nephrol Dial Transplant 1987; 2:219. Kutyrina IM, Androsova SO, Tareyeva IE. Indomethacin-induced hyporen-inaemic hypoaldosteronism. Lancet 1979; 1:785. Ponce SP, Jennings AE, Madias NE, Harrington JT. Drug induced hyperkalemia. Medicine 1985; 64:357. Corvol P, Claire M, Oblin ME, Greeving K, Rossier B. Mechanism of the antimineralocorticoid effects of spironolactones. Kidney Int 1981; 20:1. Benos DJ. Amiloride: a molecular probe of sodium transport in tissues and cells. Am J Physiol 1982; 242:C131. Corvol P. Editorial: The enigma of pseudohypoaldosteronism. J Clin Endocrinol Metab 1994; 79(1):25.

30a.Bonny O, Hummler E. Dysfunction of epithelial sodium transport: from human to mouse. Kidney Int 2000; 57:1313. 31. Chang SS, Grunder S, Hanukoglu A, et al. Mutations in subunits of the epithelial sodium channel cause salt wasting with hyberkalemic acidosis, pseudohypoaldosteronism type I. Nat Genet 1996; 12(3):248. 31a.Grunder S, Jaeger NF, Gautschi I, et al. Identification of a highly conserved segment at the N-terminus of the epithelial Nat channel alpha subunit involved in gating. Pflugers Arch 1999; 438:709. 32. Chung E, Hanukoglu A, Rees M, et al. Exclusion of the locus for autosomal recessive pseudohypoaldosteronism type I from the mineralocorticoid receptor gene region on human chromosome 4q by linkage analysis. J Clin Endocrinol Metab 1995; 80(11):3341. 33. Garty H, Palmer LG. Epithelial sodium channels: function, structure, and regulation. Physiol Rev 1997; 77(2):359. 34. Grunder S, Firsov D, Chang SS, et al. A mutation causing pseudohypoal-dosteronism type I identifies a conserved glycine that is involved in the gating of the epithelial sodium channel. EMBO J 1997; 16(5):899. 35. Strautnieks SS, Thompson RJ, Gardiner RM, Chung E. A novel splice-site mutation in the subunit of the epithelial sodium channel gene in three pseudohypoaldosteronism type I families. Nat Genet 1996; 13(2):248. 36. Strautnieks SS, Thompson RJ, Hanukoglu A, et al. Localization of pseudo-hypoaldosteronism genes to chromosome 16p12.2-13.11 and 12p13.1-pter by homozygosity mapping. Hum Molec Genet 1996; 5(2):293. 36a.Bonny O, Chraibi A, Loffing J, et al. Functional expression of a pseudohy-poaldosteronism type I mutated epithelial Na+ channel lacking the pore-forming region of its alpha subunit. J Clin Invest 1999; 104(7):967. Snyder PM. Liddle's syndrome mutations disrupt cAMP-mediated translocations of the epithelial Na + channel to the cell surface. J Clin Invest 2000; 105:45. Geller DS, Rodriguez-Soriano J, Vallo Boado A, et al. Mutations in the mineralocorticoid receptor gene cause autosomal dominant pseudohypoaldosteronism type I. Nat Genet 1998; 19(3):279. Stone RC, Vale P, Rosa FC. Effect of hydrochlorothiazide in pseudohypoal-dosteronism with hypercalciura and severe hyperkalemia. Pediatr Nephrol 1996; 10(4):501. White PC: Mechanisms of disease: disorders of aldosterone biosynthesis and action. N Engl J Med 1994; 331(4):250. Hanukoglu A, Joy O, Steinitz M, et al. Pseudohypoaldosteronism due to renal and multisystem resistance to mineralocorticoids respond differently to carbenoxolone. J Steroid Biochem Molec Biol 1997; 60(1-2):105. 42. Mansfield TA, Simon DB, Farfel Z, et al. Multilocus linkage of familial hyperkalemia and hypertension, pseudohypoaldosteronism type II, to chromosomes 1q31-42 and 17p11-q21. Nat Genet 1997; 16(2):202. 37. 38. 39. 40. 41.

CHAPTER 82 ENDOCRINE ASPECTS OF HYPERTENSION Principles and Practice of Endocrinology and Metabolism

CHAPTER 82 ENDOCRINE ASPECTS OF HYPERTENSION

DALILA B. CORRY AND MICHAEL L. TUCK Spectrum of Hypertension Renovascular Hypertension Pathogenesis Clinical Clues to Screen for Renovascular Hypertension Laboratory Evaluation Screening Tests Treatment of Renovascular Hypertension Renin-Producing Tumors Mineralocorticoid Hypertension Hypertension in other Endocrine Disorders Cushing Syndrome Diabetes Mellitus and Hypertension Insulin and Hypertension Obesity and Hypertension Hypercalcemia and Hypertension Acromegaly and Hypertension Thyroid Disease and Hypertension Endocrine Systems in Essential Hypertension Renin-Angiotensin System Sympathetic Nervous System In Hypertension Sodium and Hypertension Membrane Sodium Transport Sodium Pump Inhibitors Endogenous Natriuretic Peptides Kinins in Hypertension Nitric Oxide in Hypertension Endothelin in Hypertension Chapter References

SPECTRUM OF HYPERTENSION
Hypertension is an extremely common disorder, affecting 15% to 20% of the population. More than 95% of hypertension is of unknown etiology, termed primary or essential hypertension. There are several causes of endocrine hypertension (Table 82-1), with the most common forms being renal artery stenosis, primary aldosteronism, and pheochromocytoma. In the Hypertension Detection Follow-Up Program, 0.18% of 15,000 participants had secondary hypertension.1 Because endocrine hypertension is rare, it is recommended by the Joint National Committee on Hypertension VI (JNC VI) that the new-onset hypertensive patient not have extensive testing for endocrine hypertension.2 Thus, only if the history and physical examination suggest endocrine hypertension is it necessary to perform screening tests. Pheochromocytoma can be detected with high sensitivity and specificity using urine and plasma catecholamine measurements. Primary aldosteronism is suspected by low serum potassium and plasma renin activity (PRA); however, as screening tests, these measurements are not very sensitive or specific. The screening tests for renovascular hypertension are improving, but most testing procedures are either invasive or indirect and can be costly.

TABLE 82-1. Endocrine Causes of Hypertension

Secondary hypertension should be suspected in severe and resistant hypertension, new-onset hypertension in the young (<20 years) or the older patient (>55 years), hypertension with spontaneous hypokalemia, and episodic hypertension with sweating, tachycardia, and headaches. Important physical findings include the detection of an abdominal bruit (suggestive of renovascular hypertension); a radial-femoral pulse delay (suggestive of aortic coarctation); or central obesity, striae, and bruising (suggestive of Cushing syndrome). Other chapters deal in depth with aldosteronism, Cushing syndrome, congenital adrenal hyperplasia, and pheochromocytoma (see Chap. 75, Chap. 77, Chap. 80, and Chap. 86). This chapter deals with selected topics. RENOVASCULAR HYPERTENSION Renovascular hypertension is the most common cause of secondary hypertension (1% of cases of mild to moderate hypertension and up to 10% to 40% of subjects with acute, severe, or refractory hypertension). 3 There is still much debate over the method and extent of evaluation in such patients.3,4,5,6,7 and 8 Clinically, diffuse atherosclerosis with severe hypertension can be associated with a high incidence.9,10 Acute elevation of serum creatinine is another presentation, either spontaneously or after initiation of an angiotensin-converting enzyme (ACE) inhibitor.5 Asymmetric renal size is a clue, as are recurrent episodes of acute (flash) pulmonary edema and unexplained congestive heart failure.11 Hypertension results from obstructive lesions of the main renal artery or its segmental branches, but it takes 70% stenosis to lead to hypoperfusion, reduced renal blood flow, and ischemia. Thus, the primary anatomic stenotic lesion may or may not be related to the blood pressure (BP), as it takes significant stenosis to activate the renin-angiotensin system, the main mediator of hypertension in renovascular hypertension. Most renovascular hypertension results from either atherosclerotic plaques or fibromuscular dysplasia. Of 2442 hypertensive patients included in the Renovascular Cooperative Study,3 63% had atherosclerotic lesions and 32% had fibromuscular dysplasia. Fibromuscular dysplasia occurs mostly in children and young adults, especially in women, whereas atherosclerotic plaques are seen in men over 45 who have a history of cigarette smoking. The most common form of fibrous renal artery disease is medial fibroplasia; often it is bilateral (Fig. 82-1).

FIGURE 82-1. A, Angiographic picture of bilateral renal artery stenosis with multiple stenotic lesions. (A, aorta.) B, Anatomic cross section of renal artery from patient with fibromuscular dysplasia.

The syndrome of renovascular hypertension requires a broad definition (i.e., any secondary elevation of BP by any condition that interferes with the arterial circulation to kidney tissue).12,13 Trauma and subcapsular hematoma can exert pressure on the kidney, causing renal ischemia without stenosis of the renal arteries. Other causes include renal transplantation, necrotizing vasculitis, intravenous drug abuse, malignant hypertension, renal artery emboli, coarctation or dissection of the aorta, and renal artery aneurysms.13 PATHOGENESIS The two main types of renovascular hypertension are unilateral and bilateral renal artery stenosis. The renin-angiotensin system is the major hormone system in the pathogenesis of renovascular hypertension, with the critical initiating event being reduced renal perfusion pressure14,15,16,17 and 18 (see Chap. 79 and Chap. 183). There is increased renin and angiotensin II (A-II) production by the stenotic kidney, although PRA levels often are normal to only moderately elevated, reflecting renin suppression in the non-stenotic kidney. In experimental one-kidney, one-clip renovascular hypertension, the renin levels increase for 6 to 10 days after renal artery clipping and then decline, yet the hypertension persists. The long-term increase in BP is related to increases in vascular resistance; the early enhancement of the renin-angiotensin system triggers other factors, sustaining the hypertension. (Volume expansion may play a critical role in long-term BP elevations.) In contrast, in the two-kidney, one-clip model of renovascular hypertension, an elevated PRA persists throughout the chronic maintenance phase of hypertension. In unilateral renal artery stenosis, the uninvolved kidney should protect against fluid and electrolyte imbalance and hypertension; the rise in BP should increase sodium excretion in the nonstenotic kidney. However, the nonstenotic kidney does not adapt appropriately; there is a paradoxical rise in local A-II, despite a suppressed PRA15,16 (Fig. 82-2). This discordant response leads to mediation by A-II of vasoconstriction and sodium retention with a blunted pressure natriuresis response curve in the nonstenotic kidney.

FIGURE 82-2. Changes in hemodynamics and the renin-angiotensin-aldosterone system in the kidneys and circulation in unilateral renal artery stenosis. Intrarenal changes begin in the stenotic kidney, with subsequent changes in the nonstenotic kidney. (A-II, angiotensin II; BP, blood pressure.)

In bilateral renovascular hypertension, the hormonal, neural, and volume responses are different from those in unilateral disease. There is an initial rise in the renin-angiotensin system, but with time, there is normalization of the renin-angiotensin system activity, and hypertension in the steady-state condition becomes more dependent on volume-mediated factors. Other amplifying factors contribute to the maintenance of hypertension in renovascular hypertension; there may be a neurogenic phase, as enhanced sympathetic nervous system activity is found in the early stages. 19 Levels of several vasoactive eicosanoids are high in animals with renovascular hypertension.14,16 A-IIinduced activation of the lipoxygenase pathway of arachidonic acid has pressor functions and has been implicated in the sustained phase. Lipoxygenase inhibitors prevent the development of experimental renovascular hypertension.19 A-II in the kidney is taken up and incorporated by renal cells through the A-IIAT1 receptor16; this uptake might serve to prolong A-II action. Circulating levels of aldosterone are normal to high in renovascular disease, depending on the levels of renin. In marked renovascular hypertension, renin and aldosterone levels are high, and hypokalemia is found.20 CLINICAL CLUES TO SCREEN FOR RENOVASCULAR HYPERTENSION Abrupt onset of severe hypertension in a young (younger than 20 years) or older individual (older than 55 years). Prepubertal hypertension. Hypertension that is refractory to triple-drug therapy. Malignant or accelerated hypertension associated with grade III to IV retinopathy. Sudden flank pain due to renal embolic infarction. Unexplained azotemia. Detection of an upper flank or abdominal bruit that radiates to the flanks. (The bruit may be systolic, systolic-diastolic, or continuous, and is high pitched and difficult to discern.) Bruit in a hypertensive young woman suggests fibromuscular disease. Hypertension associated with diffuse atherosclerotic vascular disease, especially in a smoker. Unexplained deterioration in renal function with the use of an ACE inhibitor. Paradoxical worsening of hypertension with a diuretic. Spontaneous hypokalemia.

Recurrent flash pulmonary edema. Some patients have no such histories and are indistinguishable from those with essential hypertension.3,4 LABORATORY EVALUATION Routine Laboratory Tests. Measurements of electrolytes, renal function tests, and urinalysis often give normal results in renovascular hypertension.2 Hypokalemia is found in those cases in which secondary hyperaldosteronism is found. In unilateral stenosis of the renal artery, an impairment of renal function occurs infrequently, whereas in bilateral renal artery stenosis, serum creatinine and urea nitrogen may be increased. In embolic renovascular disease, hematuria may be detected. SCREENING TESTS Intravenous Pyelography. The findings on rapid-sequence pyelography that have the greatest significance are a unilateral decrease in renal size (>1.5 cm disparity in pole-to-pole diameter) and a unilateral delay in the appearance of contrast medium in the 1- to 5-minute exposures.3 False-negative results often occur in cases of renovascular hypertension because of bilateral renal artery disease.3,21 The hypertensive urogram has low sensitivity (a negative test does not exclude renovascular hypertension) and specificity and, importantly, has the risk of nephrotoxicity because of contrast agents. Plasma Renin Activity. Although hypoperfusion in the stenotic kidney activates the renin-angiotensin system, renin secretion in the nonstenotic kidney is often normal or low, so that the baseline or unstimulated circulating PRA levels are normal in most cases. Thus, when measured in peripheral venous samples, PRA is normal in 20% to 50% of cases of documented renovascular hypertension.7,13,20 The interpretation is confounded by the fact that 16% to 20% of subjects with essential hypertension have PRA values that are moderately elevated. However, in renovascular hypertension there are exaggerated PRA responses to procedures that stimulate the renin-angiotensin system, such as upright posture, sodium restriction, and the administration of angiotensin-converting enzyme inhibitors (ACEI).22 ACEI interrupt the conversion of A-I to A-II; the negative feedback effect of A-II on renin release is diminished, resulting in an increase in secretion. The administration of the ACEI captopril to patients with renovascular hypertension produces a marked increase in PRA compared to responses in other forms of hypertension.20,22,23 and 24 In the usual test procedure, captopril, 25 to 50 mg, is given orally, and PRA is obtained after 1 hour. The test has 75% to 100% sensitivity and 60% to 90% specificity.22,24 Furthermore, it is safe and inexpensive, and can be done on an outpatient basis or simultaneously with captopril renography. The utility of the test is hampered by the need for strict standardization, the need to stop antihypertensive medications, its reduced accuracy in renal failure, and low sensitivity and poor predictive value compared to a renogram. Renal vein renin levels, sampled by percutaneous catheterization, determine the functional significance of the stenotic lesion and predict the BP response to therapy.3 Renin production is increased in the stenotic kidney and suppressed in the contralateral kidney; asymmetry of renin production yields a renal vein renin ratio of 1.5:1 in renovascular hypertension. This procedure has predictive value for surgical curability, but use has declined because it is invasive, and up to 60% of patients do not lateralize but are nevertheless improved by surgery. 25 Radionuclide Renography. The renogram relies on a disparity of renal function and of perfusion between the stenotic and nonstenotic kidneys. Isotopic compounds such as technetium-99mdiethylenetetraminepentaacetic acid (DTPA) accurately measure glomerular filtration rate, and compounds such as 99m Tc-mercapto acetyl triglycine (99mTc-MAG3) measure both glomerular filtration rate and tubular secretion and are good in renal insufficiency.7 The addition of computer quantitation has improved their accuracy.3,4,5,6 and 7,13,26 Patients with renovascular hypertension show decreased uptake, delayed peak values, and prolonged excretion of these agents. Because of the low-risk and noninvasive nature of renography, it has many advantages. However, renography used alone has a high incidence of false-negative and false-positive results. Captopril administration during renography has improved its sensitivity and specificity.5,6 and 7,22,24,26,27 Captopril (2550 mg), which is given 1 hour before the isotope injection, reduces A-IImediated vasoconstriction in the efferent arterioles and lowers glomerular pressure and filtration rate. In the stenotic kidney, captopril reduces renal perfusion more than in the nonstenotic kidney, thereby revealing the asymmetric reduction in renal function in renovascular hypertension. Scintiphotographs and time-activity curves are analyzed to assess renal perfusion, function, and size. In high-risk populations, sensitivity and specificity exceed 90% for high-grade stenosis. The predictive value of the ACE-inhibitor renogram is less in low-risk populations and in bilateral disease. Duplex Doppler Ultrasonography. Ultrasonography has been improved by the use of low-frequency transducers (B-mode imaging) for better visualization of the renal arteries, and Doppler for the measurement of differential blood flow velocities.28,29,30 and 31 It has the advantage of determining both the anatomic and the functional significance of a stenotic lesion. A renal-aortic flow velocity ratio of 3.5 suggests renovascular hypertension. Duplex Doppler ultrasonography can detect unilateral and bilateral disease, can detect recurrent stenosis in previously treated patients, and can be enhanced with ACE inhibition30 and color coding.31 It has a 99% positive predictive value and a 97% negative predictive value.29 (Disadvantages: time-consuming and highly operator-dependent.) Magnetic Resonance Angiography. Magnetic resonance (MR) angiography is promising and noninvasive, and may become the method of choice.32,33,34,35 and 36 In comparison with arteriography, MR angiography showed 100% sensitivity and 96% specificity for finding stenosis of the main renal arteries.32,33,34,35 and 36 A sensitivity of 100% and a specificity of 71% were found for lesions of the proximal renal arteries with 50% to 75% stenosis.34 Significant advances in techniques include breath-holding MR angiography and paramagnetic contrast material. Thus, the procedure may be able to visualize accessory arteries.34 Phase-contrast MR angiography may enable determination of the hemodynamic significance of a stenotic lesion.35 The procedure cannot be used in individuals with metallic implants such as a pacemaker or a clip for an aneurysm. Spiral (Helical) Computed Tomographic Scan and Computed Tomographic Angiography. This offers great promise as the best noninvasive screening test.37 In a comparison of spiral computed tomography (CT) to arteriography in subjects investigated for renovascular hypertension, the sensitivity was 98% and the specificity 94%.28 However, renal insufficiency reduces the accuracy of this procedure due to diminishing renal blood flow. It identifies lesions of the main renal arteries but has more difficulty in finding branch lesions. Although it is minimally invasive, the risk of radiocontrast-induced nephrotoxicity still exists. Angiography. Despite the increase in types of procedures and technical advances in detecting renovascular hypertension, renal angiography by the percutaneous, transfemoral route remains the gold standard for detecting renal artery stenosis, as well as its location and pathology. Atherosclerotic lesions most often involve the proximal segment of the renal artery and have a circular configuration. By contrast, fibromuscular disease is in the more distal portions of the renal artery and is either localized or diffuse and bilateral. Angiography is costly and invasive and has the risk of nephrotoxicity. Because angiography does not always predict the functional significance of renal artery stenosis or its treatment outcome, many centers apply simultaneous angioplasty, with the BP response to percutaneous renal angioplasty or stenting serving as the indicator of functional significance. Digital subtraction angiography avoids some of the complications of percutaneous catheterization, because computer enhancement of the renal area allows the injection of smaller amounts of contrast material. 3 Injection has been in a peripheral vein, but the specificity and sensitivity of the procedure are both 90% or less when compared to arterial procedures. Because 150 to 200 mL of dye is injected, there is still risk for nephrotoxicity. An intraaortic injection site provides better visualization. The procedure can be performed safely in azotemic patients, and it has been useful in hypertension after renal transplantation.4 Thus, the renal arteriogram best establishes the presence of renal artery stenosis but does not distinguish functional from nonfunctional lesions. Tests such as the intravenous pyelogram and radionucleotide renogram lack sensitivity for detecting renovascular hypertension and, if used alone, cannot consistently diagnose renovascular hypertension. The newer tests (e.g., captopril renography, MR angiography, duplex ultra-sonography, and spiral CT) show great promise. TREATMENT OF RENOVASCULAR HYPERTENSION The goals of therapy of renovascular hypertension are the control of BP and the preservation of renal function. The three therapies for treatment of renovascular hypertension are medical, percutaneous transluminal renal angioplasty, and surgery. Each approach has advantages and disadvantages, and the selection also depends on cause, severity, and age. Few trials have compared the three therapies. Often no single test can help in this decision, and experience and clinical judgment are paramount. There are several differences in the approach to unilateral versus bilateral renal artery stenosis. Unilateral Renal Artery Stenosis

SURGERY. Often, surgery is reserved for patients for whom hypertension cannot be controlled on medical therapy or for those who fail angioplasty and in whom renal function is deteriorating. However, some believe that surgery should be considered in patients <50 to 60 years of age, as the outcome is more definitive and spares the chronic use of medication.38,39,40 and 41 Surgery is generally more effective than angioplasty, especially in patients with atherosclerosis. One indication for surgical revascularization is preservation of renal function, based on the emerging evidence of deteriorating renal function in atherosclerotic ischemic renal disease.41,42 and 43 Revascularization procedures include bypass using natural or synthetic arteries; natural arteries provide a better outcome.38,39,40 and 41 The surgical cure rate for hypertension due to unilateral atherosclerotic lesions may be up to 95%, especially if the disease is of short duration. Less favorable results are seen in older subjects with long-standing hypertension, probably due to either underlying essential hypertension or the onset of intrarenal disease. Mortality from surgery is ~2.5% in experienced centers; it is higher in older patients, especially those with diffuse atherosclerosis and heart failure.40 PERCUTANEOUS TRANSLUMINAL ANGIOPLASTY. The technical success and cure rate for percutaneous transluminal angioplasty are rapidly improving; this procedure and stenting are now treatments of choice.44,45,46,47,48,49 and 50 Patients with fibromuscular dysplasia have the best outcome (<10% restenosis). Best results are obtained in lesions of the main renal artery that are only partly occluded.44,45,46 and 47 In atherosclerotic lesions the results are not as good. Cure rates of only 10% to 20% are recorded; ~60% will improve but not be cured, and 15% incur no benefit. A failure can be predicted in the first 48 hours if BP fails to fall.46 Technical complications include renal artery thrombosis and perforation. The main overall advantage of percutaneous transluminal angioplasty may be the ability to decrease the number of medications needed to control BP without the risks of surgery.46,47 In transplant renal artery stenosis, angioplasty is first-line therapy, with a cure rate of 70% at 5 years; surgery is less successful because of extensive scar tissue.3 Ostial stenoses are less likely to respond to angioplasty (recurrence rate as high as 47% for atherosclerotic lesions). Here, the outcome of angioplasty is better with intravascular stents.48,49 and 50 Stents are placed most often after unsuccessful angioplasty and offset the effect of elastic recoil (success rate from 65% to 70%; restenosis in only 13%).48,49 MEDICAL THERAPY. In high-risk individuals with coexisting carotid or coronary artery disease in which surgical or angiographic risk is high, medical therapy is often the sole modality.51,52 and 53 However, the natural course of renovascular hypertension, especially due to atherosclerotic disease, shows that many stenotic lesions progress over time and that renal damage can occur in untreated lesions.52 Thus, it remains uncertain whether drug therapyby only controlling BPaffects the progression of renovascular hypertension. Because renovascular hypertension is a renin-dependent form of hypertension, ACEI and A-IIreceptor blockers should show selective efficacy.51 In unilateral renovascular hypertension, ACEI may be as effective as monotherapy.51 However, ACE inhibitors can produce rapid loss of renal function especially in bilateral renovascular hypertension or in renovascular hypertension with a solitary kidney, owing to the almost complete dependency of renal function on A-II.51 Usually, the use of ACEI in unilateral renovascular hypertension does not permanently reduce glomerular filtration and is safe and effective.51 Some clinicians monitor serum creatinine, renal size, and periodic ultrasonography during such therapy. Calcium antagonists are effective antihypertensive agents in renovascular hypertension and may induce less renal impairment than ACE inhibitors.3 However, in critical, high-grade stenosis, any agent that lowers BP excessively can impair renal function. A-IIreceptor blockers act by directly blocking the AT1 receptor without affecting other peptides (e.g., bradykinin) as do the ACEI. Bilateral Renal Artery Stenosis. Bilateral renal artery stenosis or unilateral stenosis in a solitary kidney usually involves more acute, severe, and difficult-to-treat hypertension, accompanied often by renal insufficiency. In bilateral renovascular hypertension, treatment should be guided toward lowering BP and preserving renal function. Patients who have bilateral disease with complete stenosis of one renal artery are at the greatest risk. Doppler ultrasound follow-up in bilateral renovascular hypertension shows that by 1 year, there often is a decrease in kidney size.54 To preserve renal function, the use of surgery or angioplasty should be considered in bilateral disease. Percutaneous transluminal angioplasty is generally less successful in bilateral renovascular hypertension and is associated with more complications (atheroemboli).42,55 It is limited to people who are at high surgical risk or those who cannot be controlled by medication. Intravascular stents may expand the success of angioplasty in this disease. Medical therapy in bilateral disease is best handled with ACEI or a calcium-channel blocker with the addition of a diuretic based on the volume expansion and volume-dependent hypertension.51 One can almost expect a hemodynamic decline in glomerular filtration rate following ACEI in bilateral renovascular hypertension, so careful monitoring is essential. If possible, the surgical correction of bilateral renal artery stenosis should be pursued, because BP and renal function can improve after revascularization.40,56 Optimal therapy for bilateral disease has not been established: Surgery is the treatment of choice in the young with bilateral severe stenosis. In older, high-risk subjects with atherosclerosis, medical therapy with either an ACEI or a calcium-channel blocker is indicated. RENIN-PRODUCING TUMORS Renin-producing tumors, a rare form of endocrine hypertension that occurs in young persons, manifest severe hypertension and hypokalemia.56,57 The PRA is among the highest recorded in hypertensive syndromes and can exceed 50 ng/mL/h.56 The hypokalemia, often <2.0 mEq/L, is due to intense secondary hyperaldosteronism. The combined high PRA and aldosterone levels distinguish this disorder from primary aldosteronism. Two categories of renin-producing tumors cause this syndrome: renal juxtaglomerular cell tumors and a variety of extrarenal tumors, such as Wilms and ovarian tumors.56,57 The renin originating from tumor production resembles that found in normal kidneys.58 Extrarenal renin-secreting tumors have higher concentrations of prorenin than do tumors of renal origin.57,59 Most cases of extrarenal renin-secreting tumors occur in women, and many are located in the reproductive tract.59 In these patients, the high prorenin level serves as a marker for the tumor. One presentation is the hyponatremic hypertensive syndrome and massive proteinuria seen in patients with renin-producing leiomyosarcomas.60 The ACEI captopril is effective in the control of BP due to renin tumors.61 In renal renin-secreting tumors, renal vein PRA measurements can localize the tumor. They are small, so that radiologic visualization, including urography, magnetic resonance imaging (MRI), ultrasonography, CT, and angiography may sometimes be unsuccessful57,62; even exploratory surgery may fail to find the lesion. In these patients, nephrectomy or selective tumor resection is curative, and ACEI are the medical treatment of choice.28 An A-IIreceptor blocker may be equally effective or superior by also blocking the antigensin type 2 (AT2) receptor.63 MINERALOCORTICOID HYPERTENSION Several mineralocorticoids, such as aldosterone and deoxycorticosterone (DOC), produce hypertensive syndromes.64,65 and 66 For more details of aldosterone function see review.66a Primary aldosteronism is the best example of mineralocorticoid hypertension. The mechanisms underlying hypertension include sodium retention, extracellular fluid expansion, high cardiac output, increased sympathetic nervous system activity, and structural changes in blood vessels. Hyperaldosteronism is reviewed elsewhere (see Chap. 80), as is congenital adrenal hyperplasia (see Chap. 77). Table 82-2 summarizes the levels of BP, serum potassium, PRA, aldosterone, cortisol, DOC, and corticotropin (ACTH) in several glucocorticoid and mineralocorticoid disorders associated with hypertension. Glucocorticoid resistance results in a mineralocorticoid form of hypertension accompanied by hypokalemia and suppressed renin and aldosterone (see Table 82-2). High levels of plasma cortisol and a paucity of stigmata of Cushing syndrome suggest the diagnosis.67 The insensitivity to glucocorticoids is caused by inherited glucocorticoid receptor mutations.67 An insensitivity to cortisol at the hypothalamic-pituitary feedback for ACTH increases ACTH production, which, in turn, stimulates DOC and corticosterone formation; the increased mineralocorticoid activity then causes volume expansion, hypertension, and the suppression of renin and aldosterone. An adrenal androgen excess can also be found, resulting in hirsutism in women and precocious pseudopuberty in children. The demonstration of a functionally abnormal glucocorticoid receptor in mononuclear cells confirms the diagnosis.67 Mutations occur in the glucocorticoid receptorligand-binding domain and at a splice site.67 High doses of dexa-methasone reduce the ACTH, and thereby correct the BP and serum potassium abnormalities. A form of cortisol resistance has also been described in the acquired immunodeficiency syndrome.67,68

TABLE 82-2. Biochemical Features of Hypertensive Syndromes Involving Mineralocorticoid and Glucocorticoid Excess

HYPERTENSION IN OTHER ENDOCRINE DISORDERS CUSHING SYNDROME Hypertension is common in endogenous Cushing syndrome (up to 80%) but occurs less frequently with exogenous glucocorticoid therapy.69 The cardiovascular mortality and morbidity are much higher in Cushing syndrome,70 attributed to the effects of cortisol on BP, on atherosclerosis, and on the heart. The hypertension is thought to directly result from cortisol excess through at least three mechanisms.71 Glucocorticoid excess increases the hepatic synthesis of angiotensinogen, the substrate for the production of the vasoconstrictor A-II. This mechanism might suggest a renin-dependent form of hypertension, yet renin levels in Cushing syndrome often are normal to decreased. Glucocorticoid administration to normal subjects will enhance vascular reactivity to pressor hormones such as phenylephrine; indeed, many attribute its hypertensive action to this mechanism. Although cortisol in excess binds to mineralocorticoid receptors, most cases of Cushing syndrome resulting from pituitary or adrenal adenoma do not have strong evidence for mineralocorticoid excess such as hypokalemia and suppressed renin. A more pronounced mineralocorticoid effect is seen in Cushing syndrome due to paraneoplastic ACTH production. In these disorders, the high levels of cortisol and DOC, but not aldosterone, cause a mineralocorticoid effect, leading to hypokalemia and hypertension71 (see Chap. 219). High circulating ACTH levels in these ectopic syndromes preferentially stimulate adrenal secretion of DOC over aldosterone; the excess DOC causes volume expansion and the suppression of renin and aldosterone. One should be alerted to this scenario in an established or suspected cancer patientwith unexplained hypokalemia and metabolic alkalosiswho may or may not have hypertension. Another mechanism for hypertension in paraneoplastic ACTH syndromes is one in which the renal form of the enzyme 11b-hydroxysteroid dehydrogenase (11b-HSD) is deficient or overwhelmed by the high circulating cortisol levels.72 This defect leads to the accumulation of cortisol in the kidneys, where it binds in excess to the mineralocorticoid receptor. Diagnosis is made by demonstrating a high ratio of the urine metabolites of cortisol/cortisone. These findings are similar to those seen in the genetic or acquired (licorice) forms of apparent mineralocorticoid excess.66 In some cases of adrenal carcinoma, there are partial enzyme blocks leading to the accumulation of steroid precursors (e.g., compound DOC), which can have mineralocorticoid effects. DIABETES MELLITUS AND HYPERTENSION The frequency of hypertension in diabetes mellitus is high, reaching an incidence rate of about 50%.73,74 Hypertension in the diabetic greatly enhances the risk of macrovascular atherosclerotic complications and the microvascular complications of nephropathy and retinopathy. It is also known that lowering BP slows the progression of diabetic nephropathy75 and diabetic retinopathy.76 In type 1 diabetes mellitus, hypertension appears after several years and is seen almost uniquely in those 40% of type 1 diabetics who will develop diabetic nephropathy. The BP begins to rise at about 3 years after the onset of microalbuminuria, and hypertension is eventually seen in 85% of them. Thus, renal mechanisms explain most of the pathophysiology of hypertension in type 1 diabetes mellitus. In type 2 diabetics, 40% to 70% develop hypertension. Here, the hypertension is due to aging, obesity, insulin resistance, hyperinsulinemia, atherosclerosis, and renal disease; it may be present at the time of diagnosis or even have preceded the onset of diabetes mellitus. However, a significant number of type 2 diabetics have microalbuminuria, suggesting that renal disease can be a contributor. Nevertheless, obesity probably is the major cause of their hypertension; genetic factors also contribute, as there is an association between the A2 allele of the glycogen synthase gene and insulin resistance.77 Unique mechanisms for the cause of hypertension in diabetes mellitus relate to the metabolic abnormalities; both type 1 and type 2 diabetics have increased vascular reactivity and sodium retention.78,79 Norepinephrine, A-II, exercise, and mental stress in diabetic subjects lead to exaggerated BP responses.78,79 Sodium retention and volume expansion in the diabetic are due to increased reabsorption through tubular glucose-sodium cotransport80; this causes a higher incidence of salt-sensitive hypertension.81 High levels of insulin in diabetes contribute to the hypertension. In type 2 diabetics, insulin resistance causes hyperinsulinemia; also, insulin excess may occur in type 1 diabetes due to exogenous therapy.74 Insulin increases sodium reabsorption and sympathetic nervous system activity and has direct effects on vascular reactivity.73,74 After starting insulin in type 2 diabetics, there is an increase in mean values of BP from 132/81 to 148/89 mm Hg.82 In human diabetes, as well as in experimental models, values for plasma renin activity often are reduced or completely suppressed,83 whereas in certain tissues (e.g., the kidney) the expression of renin and its products is elevated.84 Individuals with long-standing diabetes and renal impairment can have the syndrome of hyporeninemic hypoaldosteronism, which is detected by finding low PRA and aldosterone levels and hyperkalemia. The low PRA in diabetic patients is due to diabetic nephropathy, destruction of the juxtaglomerular apparatus, defective conversion of prorenin to active renin, chronic volume expansion, and decreased neural control of renin release.83,84 The efficacy of ACEI in treatment of hypertension and target-organ protection in diabetes mellitus occurs despite low circulating levels of renin and is thought to be due to effects on the tissue renin-angiotensin system84 and increases in bradykinin.85 Endothelial dysfunction, which is common in diabetes, may further contribute to the hypertension86; in type 1 diabetics, ACEI therapy improves endothelial function.87 INSULIN AND HYPERTENSION There is a positive correlation between insulin resistance, hyperinsulinemia, and BP.88,89 and 90 In high-risk patients with syndrome X, insulin resistance and hyperinsulinemia partly contribute to the hypertension as well as the dyslipidemia, atherosclerosis, and cardiovascular disease.88,89 and 90 Insulin resistance and the clustering of other risk factors can be found in nondiabetic subjects who are obese or lean but often have impaired glucose tolerance. Approximately 50% of patients with essential hypertension have the insulin-resistance syndrome.90 There has been disagreement over whether insulin resistance and hyperinsulinemia have the same risk impact on BP in African-Americans. A follow-up study demonstrated that in both African-Americans and whites, baseline insulin and the insulin/glucose ratio are associated with a higher risk of hypertension.91 There also may be elevations of steroid hormones (e.g., androgens), since circulating androgen levels are directly related to the degree of abdominal fat.92 Moreover, insulin resistance occurs with androgen excess in the polycystic ovary syndrome.93 There is a genetic contribution to the insulin-resistance syndrome, as normotensive individuals from hypertensive families have higher insulin levels and more insulin resistance.94,95 In a study to identify the genetic locus for diabetes and insulin resistance that might be linked to BP, linkage analysis was performed in 48 families with type 2 diabetes.96 Although no linkage to genes for the components of the renin-angiotensin system were found, one region of the lipoprotein lipase gene correlated with systolic BP. Insulin alters BP-regulating mechanisms, such as sympathetic nervous system activity, sodium handling, and vascular tone.73,74,86 In normal persons, the infusion of insulin increases plasma norepinephrine levels, regional sympathetic neuronal burst activity, and sodium reabsorption.73,74 Despite these changes, insulin infusion in normal subjects causes vasodilatation and increased blood flow, suggesting that insulin acts directly on blood vessels as a vasodilatora finding that might not support its role in hypertension.97 This vasodilatory effect is blunted in obese individuals and even more so in type 2 diabetics.97 This implies that a resistance to insulin's normal vasodilating effect could lead to hypertension by shifting the vasomotor balance away from vasodilatation toward vasocon-striction. The decrease in forearm blood flow could also decrease glucose delivery to skeletal muscles, thus contributing to insulin resistance.97 The vasodilatory effect of insulin is blocked by inhibitors of the nitric oxide pathway.97,98 In insulin-resistance states, there may be simultaneous reductions in insulin-mediated glucose uptake and insulin-mediated vasodilatation

through abnormal endothelial function and reduced nitric oxide accumulation.97 Insulin also has effects on the vascular renin-angiotensin system: There is an up-regulation of the A-II type 1 receptor through posttranscriptional changes99 and increases in angiotensinogen levels.100 Both insulin and insulin-like growth factor-I (IGF-I) increase vascular growth, an effect that is blocked by inhibitors of the renin-angiotensin system.100 This suggests both a prohy-pertensive effect and a growth effect directly on blood vessels, mediated through the vascular renin-angiotensin system. Not all studies show a positive effect of insulin on BP. In dogs, chronic insulin infusion has no effect on BP.101 In essential hypertensive patients with insulin resistance, insulin administration for 2 weeks actually decreased BP by a small amount.102 In addition, not all epidemiologic reports have found a correlation between insulin and BP.89 Perhaps, in order for insulin to raise BP, other prohypertensive factors must be present. OBESITY AND HYPERTENSION There is a positive relationship between body weight and BP; weight loss evokes a hypotensive effect that may be substantial.103,104,105 and 106 Hypertensive individuals have a greater prevalence of obesity than do normotensives. Obesity carries a much greater risk of left ventricular hypertrophy that is partly independent of the BP.107 The association of obesity with other risks, such as dyslipidemia and insulin resistance, further enhances total cardiovascular risk. There is a major role of fat distribution in cardiovascular risk in obesity. Central obesity with intraabdominal fat excess represents a high-risk form of obesity.103,104 Increased intraabdominal fat is associated with hypertension, insulin resistance, hyperin-sulinemia, dyslipidemia, accelerated atherosclerosis, coagulation defects, and cardiovascular disease. There is an outpouring of free fatty acids that lead to increased Apo B and hepatic lipase, producing a typical lipid profile including increased triglycerides and small dense low-density lipoprotein (LDL) and reduced high-density lipoprotein (HDL). Clinically, abdominal obesity is diagnosed by measuring the waist/hip circumference ratio (abnormal >0.95 in men and >0.85 in women). It is visceral or subcutaneous truncal fat rather than intraperitoneal fat that is high risk; this type of fat and its distribution are also predictors of a low HDL2 cholesterol,108 insulin resistance,109 and diabetes.110 Intraabdominal fat is detected more accurately by CT scan. Obese hypertensive patients have a unique hemodynamic profile: Blood volume and cardiac output are increased, but systemic vascular resistance is normal to low.107 There are differences between normotensive and hypertensive obese subjects: Normotensive obese subjects have high cardiac output but a systemic vascular resistance that is below the levels found in normotensive lean subjects. Hypertensive obese subjects have a higher systemic vascular resistance than do normotensive obese subjects.107 In obesity, norepinephrine and insulin levels are high.105 Because insulin increases norepinephrine release, it is possible that the caloric excess causes hyperinsulinemia and stimulation of the sympathetic nervous system and BP.111 Techniques to assess sympathetic nervous system activity in humans (plasma and urinary catecholamine levels, norepinephrine spillover, and microneurography) have provided important data on obesity and its relation to hypertension. Some studies have found that plasma norepinephrine levels are variably high in obesity. During weight loss there is an even more consistent relationship between reductions in norepinephrine and BP.112 Regional norepinephrine spillover in obesity is high in the kidney and could lead to sodium retention.113 Using microneurography techniques (which determine muscle sympathetic outflow), marked elevations were found in obese normotensive young men,114 but it is unclear if sympathetic activity is higher in hypertensive obese subjects. Thus, human obesity is characterized by abnormalities in sympathetic cardiovascular control, but its role in hypertension is still undetermined. Other variables may alter sympathetic nervous system activity and hypertension in obesity. Sleep apnea syndrome occurs in a substantial subgroup of obese patients, and activation of the sympathetic nervous system by repeated bouts of hypoxia could affect BP. The muscle sympathetic activity has been found to be high only in those obese subjects who have obstructive sleep apnea; when these patients are factored out, there are no demonstrable abnormalities in sympathetic nervous system activity in obesity.115 Leptin produced in adipose tissue acts on central receptors to decrease appetite and increase energy expenditure. Like insulin, leptin has cardiovascular properties that include increasing sympathetic nervous system activity, altering sodium excretion, and possibly affecting BP.116 Also, there is a reported association of a polymorphism in the b3-adrenergic receptor gene with features of the insulin resistance syndrome.117 The hypotensive response to weight loss is accompanied by reductions in plasma volume, insulin, norepinephrine, and renin.106,112 Often a successful BP response is attained before ideal body weight is achieved. Interestingly, sodium balance does not alter the effect of weight loss on BP.106 Based on the results of multiple weight loss studies, it can be estimated that for every kilogram of weight loss there is a 0.5 to 1.0 mm Hg fall in mean BP. Reversal of obesity is difficult to achieve, and anti-hypertensive agents are often needed to control the hypertension.118 HYPERCALCEMIA AND HYPERTENSION Estimates of hypertension in primary hyperparathyroidism have ranged from 50% to 70%, and as many as 35% of patients with chronic hypercalcemia of other etiologies are hypertensive119 (see Chap. 58 and Chap. 59). Hypertension is also seen in the syndrome of parathyroid hormone (PTH) resistance (pseudohypoparathyroidism) and in secondary hyperparathyroidism due to severe renal failure. Hypertension in primary hyperparathyroidism is sometimes associated with renal insufficiency, but other factors, such as hypercalcemia; PTH and its analog, PTH-related protein; and phosphorus and magnesium deficiency contribute to the hypertension. Calcium is a major mediator of vascular contractility; calcium infusion can increase BP120 and potentiate A-II and norepinephrine. The positive relationship between serum calcium and BP in hypertensive subjects suggests a cause-effect relationship.120 PTH may be involved in raising BP, but on acute administration it acts as a vasodilator and induces a fall in BP. However, more chronic exposure to PTH may raise BP. Approximately half of patients with pseudohypoparathyroidism have hypertension, but their serum calcium is low and their PTH is high.121 Perhaps the high PTH sustains the hypertension, as normalization of calcium has no effect on BP.121 1,25-dihydroxycholecal-ciferol is high in primary hyperparathyroidism and can exert mild effects on calcium mobilization in blood vessels. PRA is high in hyperparathyroidism, suggesting that the renin-angiotensin system participates in raising the BP122; in some, parathyroid surgery normalizes PRA, aldosterone, and the BP.123 However, in renal insufficiency, hypertension persists or is only partially corrected.123 Hypertension, per se, is not an indication for parathyroid surgery in patients with primary hyperparathyroidism. Despite the heightened renin-angiotensin system in primary hyperparathyroidism, the BP response to ACEI is limited, and there is no information on the best anti-hypertensive agent in this disorder. Of note, hypocalcemia also can raise BP,124 and, in such circumstances, calcium supplementation may show beneficial effects on systolic BP in patients with essential hypertension.125 In studies of normocalcemic patients with essential hypertension, calcium supplementation produces a mild, but statistically significant, reduction in BP.125 These studies have been disputed, but it appears that a subgroup of hypertensive patients do have BP sensitivity to calcium. An ionic hypothesis of cardiovascular disease is based on the findings of a negative correlation between ionic (free) calcium and magnesium, and BP.126 ACROMEGALY AND HYPERTENSION Hypertension is very common in acromegaly. In the largest published series, hypertension was found in up to 51% of 500 patients with acromegaly127 (see Chap. 12). Human growth hormone promotes sodium reabsorption, and the extracellular fluid volume is increased in acromegaly; this is accompanied by suppressed renin levels.128,129 Also, there is evidence for overactivity of the sympathetic nervous system in acromegaly. An increase of left ventricular mass is frequently found in active disease, accompanied by an increased stroke volume and a decreased afterload. Thus, increased cardiac output in acromegaly could be responsible for the rise in BP.130 Moreover, some have postulated a specific cardiomyopathy of acromegaly that could accompany the hypertension. The successful treatment of acromegaly often normalizes or improves BP and is beneficial to the other cardiac abnormalities, especially if the disease is treated early. THYROID DISEASE AND HYPERTENSION Thyroid hormone affects all regulatory aspects of the cardiovascular system.131,132,133 and 134 Thyroid hormone excess increases heart rate, cardiac output, stroke volume, and systolic BP; however, it decreases peripheral vascular resistance and diastolic BP. Cardiovascular changes with thyroid hormone are attributed to increased adrenergic nervous system activity through the b-adrenergic receptor.133,134 Because thyroid hormones are structurally related to catecholamines, they may be taken up and released at the neural synapse.133,134 Thyroid hormone stimulates the renin-angiotensin system mainly through increasing the hepatic production of

angiotensinogen (renin substrate).134 Hypothyroidism and Hypertension. The hypertension, which occurs in some patients with hypothyroidism, mainly involves elevations in diastolic BP; the incidence of the hypertension is reported to be 15% to 30%, but reports have varied widely from 0% up to 50% of subjects.135 Overall, hypothyroidism may account for 1% to 3% of patients who are diagnosed as having essential hypertension. Hypothyroidism is common in older individuals in whom the independent incidence of hypertension is high. However, comparison studies have shown an increased prevalence of hypertension in hypothyroid individuals even when age is accounted for.135 Sympathetic nervous system activity is increased in hypothyroidism, as documented by high circulating levels of norepinephrine.131,132,133 and 134 This enhanced sympathetic nervous system activity is predominantly a-adrenergic in origin and may raise total peripheral resistance and decrease cardiac outputhemodynamic findings that have been verified in hypertensive, hypothyroid patients. In general, PRA in hypertensive, hypothyroid patients is low.135 In about 30% of these patients, thyroid hormone replacement alone will reduce BP, although most of the patients will require antihypertensive therapy. Hyperthyroidism and Hypertension. Approximately 30% of subjects with thyrotoxicosis will have elevated BP that is predominantly an elevation in systolic BP.131,132,134 The high systolic pressure and pulse pressure are best explained by increased cardiac output. These cardiac adaptations to thyroid hormone excess are mediated through the adrenergic nervous system; b-adrenergic blockade reduces the cardiac output and pulse pressure.133 Nevertheless, circulating catecholamine levels are normal to reduced in hyperthyroidism.133 In hyperthyroid rats, there is an increased density of b-adrenergic receptors in the cardiovascular tissue, indicating that thyroid hormone may increase adrenergic activity through these receptors.133 Although PRA levels are elevated in hyperthyroidism, the increased renin-angiotensin system activity is not thought to contribute to the hypertension.131,132 Successful treatment of hyperthyroidism, especially in younger subjects, usually corrects the hemodynamic abnormalities and reduces the systolic BP (see Chap. 42). ENDOCRINE SYSTEMS IN ESSENTIAL HYPERTENSION In essential hypertension, there is evidence for the participation of several hormone systems, e.g., the renin-angiotensin system, serum aldosterone, and the sympathetic nervous system. Also, natriuretic hormones (e.g., the atrial natriuretic peptides [ANPs]), ouabain-like factors, endogenous vasodilator substances (e.g., the eicosanoids [prostaglandins, lipoxygenases, P450 epoxides]), and the kallikrein-kinin system have been implicated in hypertension. Many of these systems act in an autocrine-paracrine fashion; new additions to this list include endothelial factors such as nitric oxide and endothelin. However, the clinical use of measurements of these systems in the evaluation of hypertensive subjects is limited to special study situations. On the other hand, therapy with agonists or antagonists of these systems may prove to be promising. RENIN-ANGIOTENSIN SYSTEM In hypertension, the renin-angiotensin system activity, almost always measured as PRA, varies widely. There probably is a continuum of levels of PRA in hypertension; however, some authors, who believe that there are subsets, profile hypertensive subjects as having low, normal, and high PRA.136,137 Procedures to test for the release of PRA include sodium restriction, upright posture, diuretic administration, or correlation of PRA with 24-hour sodium excretion. There is an inverse relationship between PRA and sodium excretion; the concept of renin profiling in hypertension may enable a more definitive classification of hypertension. In support of the renin-profiling procedure is the increased cardiovascular risk associated with high circulating levels of PRA.137 There also may be differences in BP response in renin subgroups. Low-Renin Hypertension. Approximately 30% of subjects with essential hypertension have low PRA levels, and a higher incidence is found in certain ethnic groups such as African-Americans.136 Normally, low PRA is also thought to be associated with older age, but PRA may not be suppressed in the hypertensive individuals who survive to an older age.138 In patients with low-renin hypertension, there is an increased incidence of salt sensitivity, and these persons are at lower risk for myocardial infarction.137 In addition, individuals with low-renin hypertension have a superior antihypertensive response to diuretics and calcium-channel blockers.139 However, the magnitude of difference of BP response to thiazide diuretics between renin groups is small.140 The mechanism for low renin in essential hypertension is probably multifactorial, including mineralocorticoid excess, alterations in the tissue renin-angiotensin system, disordered calcium metabolism, and/or low sympathetic function. The best example of low-renin hypertension is primary aldosteronism in which volume expansion suppresses PRA levels. Other genetic disorders of mineralocorticoid excess (including glucocorticoid remediable aldosteronism,141 the apparent mineralocorticoid excess syndrome,142 and Liddle syndrome143) exhibit a suppressed PRA accompanied by hypertension and hypokalemia. Initial studies of the mechanisms for low-renin hypertension focused on aldosterone and other mineralocorticoids. Several candidate steroids (18-hydroxycorticosterone, 18-hydroxy-DOC, DOC, and 19-norDOC) have turned out not to be the cause.144 In addition, hypokalemia rarely occurs in low-renin essential hypertension, and aldosterone levels are normal. The exact cause of low-renin essential hypertension remains unknown; however, these subjects might have abnormal calcium and magnesium metabolism; they display a positive relationship between diastolic BP and cytosolic calcium and an inverse relationship between systolic and diastolic BP and intracellular free magnesium.145 High-Renin Hypertension. Approximately 10% to 20% of patients with essential hypertension have elevated levels of PRA.136 It is uncertain if these moderate increments represent a distinct pathophysiologic abnormality, but these individuals are at greater risk for cardiovascular disease.137 One explanation for elevated PRA in the hypertensive subject is nephron heterogeneity, with discordant renin secretion and sodium excretion; this causes a hypertensive vasoconstriction-volume relationship.146 The renin dependency of the BP is supported by showing a greater hypotensive response to blockade of the renin-angiotensin system (ACEI, A-IIreceptor blockers) in high-renin subjects. However, interpretation of these findings is difficult, as the antihypertensive responses related to ACEI correlate only minimally with baseline PRA.147 Specific secondary causes of high-renin hypertension include renovascular hypertension, malignant hypertension, and renin-producing tumors. Normal-Renin Hypertension. Normal PRA values are found in ~60% of subjects with hypertension.136 An abnormal tissue response to A-II has been reported in subjects with normal-renin hypertension (termed nonmodulating essential hypertension).148 The nonmodulation trait may be an intermediate phenotype that is inherited and is associated with a single nucleotide polymorphism (SNP) in the coding region of the angiotensinogen gene at codon 235.149 The phenotype occurs twice as frequently in men.149 Up to 85% of subjects have a family history of hypertension.149 In these subjects, the aldosterone response to A-II infusion is subnormal on a low-sodium diet, and the renal blood flow response to A-II is subnormal on a high-sodium intake (Fig. 82-3). Thus, sodium fails to normally modulate the tissue response to A-II in both the adrenal glands and the kidneys. Because administration of an ACE inhibitor completely corrects the abnormal responses, tissue A-II defects may account for nonmodulation.149 Nonmodulation could partially explain the impaired renal sodium handling and salt sensitivity that are found in hypertension. The diagnosis of nonmodulation is difficult, requiring sodium balance and A-II infusion. However, several tests, such as angiotensinogen gene expression and the red blood cell sodium, lithium (Na+, Li+) counter-transport assay could be markers for nonmodulation.149

FIGURE 82-3. Changes in paraaminohippuric acid (PAH) clearance (renal blood flow) in normotensive and essential hypertensive subjects on high-sodium diet in response to angiotensin II. Nonmodulators fail to change renal blood flow with high salt and angiotensin II levels. (From Shoback D, Williams GH, Moore TJ, et al. Defect in the sodium-modulated tissue responsiveness to angiotensin II in essential hypertension. J Clin Invest 1983; 72:2115.)

Tissue Renin-Angiotensin System. Many of the conclusions from profiling of circulating PRA must now be interpreted in light of the presence of a tissue (local) renin-angiotensin system in blood vessels, heart, kidney, adrenal glands, and other target organs.150 It is possible that much of the cause of essential hypertension resides in the tissue renin-angiotensin system. For example, in the kidney, A-II levels can be 1000-fold higher than in the circulation, especially in certain areas such as the glomerular and peritubular sites.151 In transgenic rats in which a mouse renin gene is inserted, there is hypertension, but circulating levels of renin and A-II and renal renin content are low at the same time that vascular A-II is high.152 Treatment with ACEI and removal of the adrenals corrects the BP, suggesting a tissue mediation of

the hypertension in what was otherwise a low-renin state.152 Genetics of the Renin-Angiotensin System. Genetic aspects of the renin-angiotensin system in hypertension are being intensely studied. Polymorphisms of the angiotensinogen gene on chromosome 1 have been linked to essential hypertension, and there is a 33% excess of shared angiotensinogen alleles in male sibling pairs with diastolic hypertension.153 This mutation might account for 6% of hypertension in young white subjects. One mutation, the T235 variant of the angiotensinogen gene, is associated with elevated levels of angiotensinogen. Another mutation, involving an A for G substitution in the 6 position of the promoter region, occurs on the angiotensinogen gene; in a large study, subjects with the AA genotype had better BP responses to sodium and weight reduction.154 Thus, genotyping for angiotensinogen might potentially help to predict BP responses to therapy. ACE gene polymorphism has also been noted in hypertension and vascular disease. The alleles found for ACE include insertion (I) and deletion (D); the DD genotype is associated with coronary heart disease,155 carotid artery disease,156 and atherosclerosis. These studies show that polymorphism of ACE is strongly associated with cardiovascular complications. Insulin resistance and angina pectoris also relate to converting-enzyme gene polymorphism.157 Other studies have shown that the DD allele is associated with higher ACE levels and the II genotype with lower ACE levels in the renal disease of type 1 diabetic patients.158 The diabetic group has less reduction in BP and proteinuria on treatment with an ACEI if the subjects are on a high-salt diet.159 SYMPATHETIC NERVOUS SYSTEM IN HYPERTENSION Increased sympathetic nervous system activity (see Chap. 85) occurs in essential hypertension.160 In studies measuring plasma norepinephrine, young hypertensive subjects had higher values than did normotensive controls.161 Hypertensive subjects also have increased renal and cardiac norepinephrine spillover, indicating regional sympathetic excess.162 In studies using microneurography to measure regional sympathetic nerve activity, the recorded activity in skeletal muscle is high in essential hypertension.163 Inhibitors of sympathetic activity lower BP proportionately to the baseline level of plasma norepinephrine. Stress, isometric exercise, and tilt table maneuvers all produce exaggerated norepinephrine responses in hypertension, and hyperactive BP responses occur with norepinephrine infusion. Thus, the combination of exaggerated BP responses and high norepinephrine levels could provide a substantial neurogenic stimulus for hypertension. Certain subgroups may be more prone to develop neurogenic hypertension, including young hypertensive subjects, very obese individuals, and those defined as having a hyperdynamic circulation.160 There may be two phases of sympathetic nervous system activity participation in the evolution of hypertension. In the early phase there is high cardiac output that is initiated by an increase in sympathetic nervous system activity.160 This early hyperdynamic phase of hypertension then converts to a more long-term phase of chronic BP elevation in which cardiac output returns toward normal and systemic vascular resistance rises.164 With time high sympathetic nervous system activity and BP also induce structural and hypertrophic changes in blood vessels that further sustain the hypertension.165 In humans, there is support for the concept of an inappropriately high sympathetic nervous system activity, both early and late, in the course of essential hypertension, implying that factprs that would normally or physiologically suppress the sympathetic nervous system are not operative.164 In studies to support this notion, drugs that lower BP cause an excessive response of norepinephrine, suggesting an unmasking of sympathetic nervous system responses. There may be a central disinhibition failure due to abnormal central function in the cells of the rostral ventricular medulla in neurogenic forms of hypertension.166 Thus, values considered in the normal range in a person with established hypertension may be abnormal or inappropriately high. Much of the neurogenic contribution to hypertension may be at a regional or organ-specific level. Increased organ-specific radiotracer spillover has been demonstrated in the heart and kidneys of essential hypertensive subjects.162 Similar regional increases in sympathetic burst activity or firing rates are found in borderline and established hypertensive individuals.163 A mutually reinforcing effect of the sympathetic nervous system and renin-angiotensin system may sustain high BP, since adrenergic receptors mediate renin release and A-II can activate the sympathetic nervous system.167 Both of these systems also have parallel growth-promoting properties that probably mediate the degenerative changes in chronic hypertension. Dopamine is an indispensable catecholamine product that not only acts as a neurotransmitter in neuronal tissues but also has autocrine and paracrine functions in nonneuronal tissues. When stimulating specific receptors located in the blood vessels or along the nephron, dopamine can regulate renal hemodynamics as well as sodium and water transport, and in the adrenal gland inhibits aldosterone secretion. In humans with essential hypertension, the response to dopamine is impaired at the receptor level. Renal dopamine modulates sodium excretion during sodium loading, mediated through the renal dopamine D3-receptor (as shown in mice with inactive D3-receptors).168 Both BP and the renin-angiotensin system are higher in these mice. The D3-receptor may be an important regulator of renal renin release and BP.168 A dopamine agonist (e.g., fenoldopam and the powerful dopamine D2-receptor agonist carmoxirole) can lower BP in very severe hypertension and has been used in hypertensive crises.169 Interestingly, a reduced urinary free dopamine response to salt loading is observed in essential hypertension, indicating a deficiency of renal dopamine that could blunt normal natriuretic mechanisms.170 SODIUM AND HYPERTENSION A positive relationship between sodium intake and hypertension is documented in epidemiologic studies,171 but increased salt intake does not uniformly raise BP in all persons. There may be individual differences in BP sensitivity to salt.171,172 Some hypertensive subjects may be salt sensitive (defined as an increase in mean BP of 10 mm Hg in response to a high-salt diet). Salt-sensitive hypertension is more common in older subjects, in African-Americans, and in diabetic subjects.171,172 Factors that control sodium (e.g., membrane transport systems, ANPs, and sodium pump inhibitors) are abnormal in hypertension. Other factors that regulate sodium (e.g., aldosterone, dopamine, prostaglandins, and bradykinin) also play a role in the relationship of salt to BP. MEMBRANE SODIUM TRANSPORT A major function of most cells is to pump in an outward direction Na+ in exchange for extracellular K+ by the action of the Na+ ,K+-adenosine triphosphatase (ATPase) pump. The pump then establishes a gradient of sodium across cell membranes that can serve as energy for other passive ion transport systems. Red and white blood cells from subjects with essential hypertension show a high Na+ content, a finding that matches the observations of high Na+ content in blood vessels and other target tissue from hypertensive animal models.173,174 Disturbances in several membrane ionic transport pathways are described in human hypertension.173,174 Measurement of these transport systems in humans uses peripheral cells (erythrocytes, leukocytes, platelets). The transport pathways are discussed in the following sections. Na+,K+ Pump. The Na+,K+-ATPase or Na+,K+ pump mediates the active transport of Na+ and K+, and is central for acute and chronic regulation of cell ions. Factors such as insulin, cate-cholamines, and b2-agonists increase intracellular uptake of K+ through the pump. Compounds such as digitalis increase cell Na+ by direct inhibition of the pump; the activity can be measured by ATPase enzyme activity, ion fluxes, ouabain binding, and so forth. In skeletal muscle, the activity of the pump is increased by training, thyroid hormones, and glucocorticoids. The pump is down-regulated in hypothyroidism, cardiac failure, muscle disease, K+ deficiency, and some forms of hypertension. Insulin stimulates K+ uptake through the pump in 3T3 fibroblasts by means of phosphatidylinositol 3-kinase and protein kinase C.175 Several studies found a defective pump activity in human and animal models of hypertension that could lead to a high intracellular Na+.176 This defect could produce hypertension by partial membrane depolarization and activation of voltage-dependent Ca2+ channels (with increased influx of Ca2+), thus promoting blood vessel contraction. Another proposal is that high intracellular Na+ could activate the Na+,Ca2+ exchanger that would increase intracellular Ca2+.176 Other studies find high pump activity in blood vessels, especially in animal metabolic models such as the fructose-fed rat.177 Na+,Li+ Countertransport. An elevated erythrocyte Na+,Li+ countertransport has been noted in numerous studies of subjects with essential hypertension.173,174 The Na+,Li+-countertransport system is genetically transmitted by a single gene determinant and may be a candidate marker for genetic hypertension. Na+,Li+-countertransport kinetics are abnormal in relatives of hypertensive patients who have high countertransport.178 One question has been the usefulness of this assay as a predictor of future hypertension. In a prospective study of normotensive middle-aged men, those who eventually developed hypertension had high Na+,Li+ countertransport, suggesting the assay as a predictor of subsequent hypertension.179 Enhanced erythrocyte Na+,Li+ countertransport is found in hypertensive patients who have hyperlipidemia.180 High countertransport is also a marker for metabolic correlates [higher fasting blood glucose, greater glucose, and insulin responses to an oral glucose tolerance test (OGGT)] in hypertensive subjects.181 In nonmodulating essential hypertension, in people who fail to change renal blood flow in response to high salt intake, countertransport is high, and the assay may be a marker for nonmodulation and salt-sensitive hypertension.182 The pathophysiologic consequences of an abnormal countertransport are more difficult to understand. The system is quantitatively a minor transport pathway, and its functional significance remains to be determined. The Na+,Li+-countertransport system is also a risk marker for hypertension and renal complications in diabetes mellitus. A high countertransport appears to be an inherited-risk marker for diabetic nephropathy. The assay may identify those type 2 diabetics who will develop microalbuminuria.183 The major kinetic abnormality of the

Na+,Li+ countertransporter in diabetes with renal disease is a raised affinity for extracellular sodium.184 Studies show a membrane defect in countertransport in diabetic nephropathy that explains the abnormal kinetics, including an increased Vmax and Vmax/Km(So) ratio, which reflects higher ion association for the transport system.185 The abnormal kinetic parameters can be completely corrected by thiol group alkylation with N-ethylmaleimide. Similar to the findings in essential hypertension, high countertransport values are correlated with insulin resistance and BP in patients with type 2 diabetes.186 Na+,H+ Exchange. Elevated Na+,H+-exchanger activity is found in various cell types from patients with essential hypertension and is a recognized intermediate phenotype for this disorder.187 The phenotype of an increased maximal transport capacity is preserved in Epstein-Barr virus immortalized lymphoblasts from hypertensive patients, suggesting strong genetic control.188,189 Both lymphocytes190 and erythrocytes191 from patients with essential hypertension show an enhanced Na+,H+ -exchanger activity. Higher Na+,H+ exchanger in peripheral cells is also noted in normotensive subjects with a family history of hypertension.187 Very similar findings for elevated Na+,H+ exchanger are noted in vascular myocytes from the spontaneously hypertensive rat.192,193 There is a family of Na+,H+ -exchanger isoforms of which the most studied in hypertension is the NHE-1 isoform. The NHE-1 isoform and, in some tissues, the NHE-3 isoform have been reported to be elevated in human and experimental hypertension in several cells, including erythrocytes, platelets, and fibroblasts.187 However, the gene for the NHE-1 isoform of the exchanger is not the cause of hypertension as shown in genetic linkage studies.191 Although this transporter is involved in growth regulation, the enhanced proliferation pattern seen in vascular tissue in hypertension is mostly independent of ion transport.194 The mechanism of Na+,H+ abnormality in hypertension is unclear. The physiologic role of this transport system is to respond to intracellular acid loads, and intracellular pH does appear to be slightly lower in hypertension.192 Exchanger abundance is identical in hypertensive versus normotensive subjects, indicating that the increased activity in the hypertensive group is due to elevated turnover of the exchanger.188,195 Na+,H+ exchanger phosphorylation is elevated in quiescent cells.188 Its dysfunction could represent a defect in signal transduction as it responds to several agonists including A-II.188 It has been further noted that A-II stimulates p90rsk in vascular smooth muscle cells, leading to the hypothesis of a potential Na+ ,H+-exchanger kinase.195 Metabolic acidosis, high salt intake, and circulating hormones (e.g., insulin) also regulate Na+,H+ exchange.187 Although increased Na+,H+ exchange is high in the erythrocytes of patients with primary aldosteronism, aldosterone directly added to cells in vitro does not affect this system. 196 ACEI corrects the Na+ ,H+-exchanger overactivity in essential hypertension.197 The addition of chelerythrine, a blocker of protein kinase C, also reduces Na+,H+ -exchanger activity in spontaneously hypertensive rats.198 The addition of an amiloride derivative, which inhibits the Na+,H+ exchanger, increases BP in spontaneously hypertensive rats.199 The Na+,H+ exchanger also exhibits altered kinetics in diabetic nephropathy.200,201 This phenotype persists in immortalized lymphoblasts201 and in skin fibroblasts202 from diabetic subjects with nephropathy. Similarly to hypertension, the mechanism in diabetes is due to posttranslational factors.201,202 Na+,K+ ,(2Cl) Cotransport. This transport pathway is inhibited by loop diuretics (furosemide, bumetanide). Initial studies found low erythrocyte cotransport in essential hypertension, but later reports noted high activity, probably representing a different hypertensive population.203 Ethnic and geographic variations in cotransport may make it a poor marker for essential hypertension. In a metaanalysis from the literature, it was found that essential hypertension, family history of hypertension, gender, and antihypertensive medications were the main determinants of cotransport.204 Indirect evidence has shown a circulating cotransport inhibitor of unknown nature.205 In summary, various membrane cation-transport defects are thought to contribute to the high cell sodium content and increased Na+ reabsorption seen in hypertension. In addition, there is evidence for increased passive inward Na+ leak in hypertension. These abnormal membrane ion transport pathways suggest a more complex picture for sodium handling than can be explained by a circulating pump inhibitor, natriuretic hormones, and many of the other factors that control sodium homeostasis. SODIUM PUMP INHIBITORS Enhanced sodium excretion in response to volume expansion is associated with increased blood levels of a factor(s) that inhibits the Na+ ,K+-ATPase pump.206 Reduced Na+ excretion causes volume expansion and signals a circulating pump inhibitor to block sodium reabsorption and correct volume overload. This pump inhibitor, however, could act on other tissue Na+,K+-ATPase pump sites. For example, in the blood vessels, there would be an increase of intracellular Na+ and Ca2+ and a resetting of vascular tone.206 Increased intracellular Na+ and Ca2+ and reduced Na+ ,K+-ATPase pump activity are found in experimental and human hypertension. The isolation and structural identification of a circulating compound that inhibits the Na+,K+ -ATPase pump has been difficult. One group has identified and characterized a ouabain-like compound from human plasma and adrenal glands that inhibits Na+,K+ -ATPase pump activity.207 The physiologic effects of this endogenous ouabain include stimulation of vascular contraction and control of intracellular calcium stores.207 The endogenous ouabain is a novel steroid counterpart that is isomeric with the plant glycoside ouabain.207 The primary source of endogenous ouabain is in the adrenal zona glomerulosa where ACTH and A-II AT2 receptors stimulate its release. The signal pathway for endogenous ouabain release is different than that for aldosterone. A binding site for endogenous ouabain has been found on the sodium pump. It mediates long-term BP control through neuronal pathways, yielding a slow pressor effect that could contribute to the induction of hypertension. ENDOGENOUS NATRIURETIC PEPTIDES There is a family of endogenous natriuretic peptides, including ANP, brain natriuretic peptide (BNP), and C-type natriuretic peptide (CNP).208,209 ANP, which was first isolated from the heart, has numerous actions, including stimulation of natriuresis and vasodilation, and inhibition of the renin-angiotensin system, aldosterone, the sympathetic nervous system, and the endothelin system.208 ANP is synthesized in atrial myocytes, implying an endocrine function in the heart (see Chap. 178). ANP and BNP are released by atrial stretch (intravascular volume, atrial pressure). BNP has properties similar to those of ANP, but is synthesized and stored in the central nervous system and in atrial cells.208 In hypertension, the levels of BNP may be greater than the levels of ANP; thus, it is debated which peptide is more important in guarding against BP elevations. CNP is produced in the endothelium and is a potent vasoactive peptide that dilates both veins and arteries, but it has no natriuretic properties. CNP is not a circulating peptide and acts in a paracrine fashion. Much of the function of these ANPs is determined by their receptors. Two of these receptors, ANPR-A and ANPR-B, are linked to guanyl cyclases. ANP and BNP bind to A receptors in the endothelium of blood vessels, whereas CNP binds to B receptors.208 The third is the ANPR-C receptor, which is a clearance receptor and plays an active role in determining availability of the peptides by controlling their rate of removal from the circulation.208 In essential hypertension, the levels of ANP and BNP are elevated, but it has been argued that these values are actually relatively low for the level of BP.208 The concept of a natriuretic peptide deficiency state in essential hypertension is appealing, but unproved. The most promising therapeutic approach to raise natriuretic peptide levels and lower BP is through the regulation of the ANPR-C clearance receptors and use of neutral endopeptidase inhibition (ompatrilat) to potentiate their bioaction.210 These agents also have ACE inhibitor properties and accumulate kinins.210a Cardiac natriuretic peptide levels may also be predictors of mortality in heart failure.211 KININS IN HYPERTENSION The products of the kallikrein-kinin system, the kinins, are natriuretic and diuretic and have vasodilating properties, through stimulation of nitric oxide and prostaglandins212,213 (see Chap. 170). Kinins are produced from kininogens through cleavage by the enzyme kallikrein to form the two main kinins, bradykinin and lysyl-bradykinin (kallidin). Kinins act as paracrine and autocrine hormones through two receptors, the B1-and B2-receptors. The B2-receptors, which belong to the family linked to G proteins, are the ones that mediate most known regulating effects on blood flow to meet metabolic demands. Renal kinins play a role in the regulation of the microcirculation and of water and sodium metabolism; the natriuretic effect is mediated in part by prostaglandins. Neutral endopeptidase inhibitors are natriuretic through the blocking of kinin hydrolysis leading to their accumulation. ACE is one of the main peptidases that hydrolyze kinins; ACE inhibitors block the breakdown of kinins. Tissue and urinary kinins increase with ACEI application. Although kinin accumulation potentiates the hypotensive action of ACE inhibitors, the administration of a kinin antagonist only blocks the acute but not chronic hypotensive effect of ACEI. Kinin release after ACEI may mediate the cardioprotection of ACEI during ischemic episodes and in the prevention of remodeling.214,215 A reduction in the kallikrein-kinin system has been found in animal models of genetic hypertension; in children, it is a good marker of genetic hypertension.212,213 A restriction fragment length polymorphism for the kallikrein gene family is linked to BP in the spontaneously hypertensive rat.212 The administration of kinin inhibitors raises BP, whereas in animals, which are transgenic for kinin expression, there is a prolonged hypotensive response. The bradykinin B2-receptor gene knockout mouse also develops hypertension on a high-salt diet. NITRIC OXIDE IN HYPERTENSION

Nitric oxide (NO) is formed from endogenous arginine by nitric oxide synthase (NOS) or from exogenous nitrovasodilators such as nitroglycerin (see Chap. 179). Cells and tissues oxidize via a five-electron oxidation of the guanido nitrogen of arginine to form citrulline and NO.216 This reaction is catalyzed by the enzyme NOS, which exists in three isoforms (neuronal NOS, inducible NOS, and endothelial eNOS).216,217 Studies have found eNOS to be membrane associated in the Golgi apparatus, in plas-malemmal membranes, and in caveolae that contain several key signal-transduction complexes.218 The enzymes are dependent on several cofactors, including reduced nicotinamide adenine dinucleotide phosphate (NADP), flavin, adenine dinucleotide, and tetrahydrobiopterin; they use heme as the prosthetic group. In addition, intracellular calcium and calmodulin can activate NOS and NO production, but there are also several calcium-independent pathways.219 The actions of agents such as acetylcholine, histamine, thrombin, and insulin are mediated through NO. Certain cytokines and endotoxins also activate NO in vascular smooth muscle, liver, and peripheral cells. Shear stress induces NO production through activation of heat-shock proteins such as Hsp90.220 The major signal-transduction pathway for NO is through guanylate cyclase, forming cyclic guanosine monophosphate (GMP).221 NO binds to the heme-containing enzyme guanylate cyclase to convert guanosine triphosphate (GTP) to cGMP. This reaction activates protein kinase G with phosphorylation of numerous proteins that mediate vascular relaxation and have antigrowth effects. Numerous hormones and drugs regulate NO through effects on NOS activity and at other regulatory sites.222 These factors can also act on the NO-cGMP complex to alter signal transduction. Several agents such as phosphodiesterase inhibitors, scavenger agents (hemoglobin), agents that alter cellular calcium, NOS-substrate inhibitors, and hormone and receptor antagonists will alter NO. The NO-cGMP complex mediates smooth muscle vasodilatation in blood vessels and other hollow organs. The system is also operative in platelets, nerves, macrophages, and other target organs. NO contributes to the regulation of insulin secretion and action, platelet function, neurotransmission, memory, penile erection, cytotoxicity induced by macrophages, and atherogenesis.222 NO can operate through cGMP-independent pathways having molecular effects on iron-containing enzymes, free radicals, proteins, and DNA synthesis. The role of NO in hypertensive vascular disease is complex.223 With the preponderance of data showing that NO is crucial in vascular homeostasis, it should be a major contributor to hypertension. A logical hypothesis is that decreased eNOS expression should be seen in hypertension. There have been eNOS polymorphisms reported in essential hypertension such as a calcium repeat polymorphism in intron 13,224 and a G to T substitution in exon 7 of the eNOS gene resulting in a Glu298Asp conversion.225 However, these altered genes seem to have a weak effect on phenotype or cardiovascular risk. A depressed response of blood flow to acetylcholine, which is eNOS sensitive, is taken as an indicator of endothelial dysfunction and decreased bioactive NO. Attenuated vasodilator responses are seen in hypertension and coronary heart disease.223 Paradoxically, NO responses can be increased in both genetic and primary hypertension, and the response may vary in different vascular beds.226 In the spontaneous hypertensive rat, eNOS is up-regulated in the aorta but normal in other vascular beds. Oxidative stress may be another important determinant of endothelial function. Increased production of superoxide radicals (O2) may account for the variable findings of NO activity in hypertension, as overproduction of the NO scavenger O2 (rather than a change in eNOS production) may be important in endothelial dysfunction. An increased vascular O2 has been associated with hypercholesterolemia and animal models of hypertension.226 In summary, reduced production or bioavailability of NO should promote cellular events in blood vessels that promote vasoconstriction, inhibit vasodilatation, and activate structural damage in vessels. Blockade of NO synthase with inhibitors will raise BP in humans and animal models.223 Administration of arginine will cause a mild reduction in BP both in humans and in animal models.223 ENDOTHELIN IN HYPERTENSION Endothelin-1 (ET-1) is a potent 21-amino acid vasoconstrictor peptide produced by the endothelium. There are three mammalian endothelins (i.e., ET-1, ET-2, and ET-3), which were first found in the vascular endothelium but exist in many other cells.227 They regulate cardiovascular function and other non-cardiovascular actions such as airway smooth muscles, the digestive tract, endocrine glands, the renal system, and the nervous system. Endothelial cells produce preproendothelin, from which is derived proendothelin, which is converted to big endothelin by endothelin-converting enzyme, a neutral endopeptidase.227 Big endothelin (39 amino acids) then is processed into endothelin (22 amino acids). Vascular endothelial cells (VEC) release predominantly ET-1 in response to low shear stress, A-II, vaso-pressin, catecholamines, and transforming growth factor-b. ET-1 acts on the ETA and ETB receptors to induce contraction, proliferation, and cell hypertrophy.227 Through the ETB receptor, ET-1 may release NO and prostacyclin, thus having the capacity to be both a vasoconstrictor and a vasodilator. This dual activity depends on the receptor predominance in any vascular bed. Coronary arteries lack ETB receptors, so ET-1 is a vasoconstrictor, whereas in cardiomyocytes ETB receptors predominate. In the kidney, predominantly ETA receptors are found in blood vessels and mesangial cells. In the distal tubule, the ETA receptor regulates sodium excretion. The endothelin system plays an important role in development. Several models of gene disruption produce malformations in the upper airways and aortic arch and abnormalities in pigment and megacolon. These models do not exhibit major changes in BP. The endothelin system seems to be activated mostly in severe salt-dependent hypertension such as the deoxycorticosterone acetate, salt-sensitive rat and the Dahl salt-sensitive rat. These models overexpress ET-1 in blood vessels. Endothelin antagonist therapy reverses hypertrophic remodeling in small blood vessels.228 In human hypertension, administration of endothelin-receptor antagonists has a mild BP-lowering effect, but the effect of a combined ETA/ETB antagonist is equal to that of an ACE inhibitor.229,230 Endothelins may be involved in renal failureinduced hypertension, in cyclosporine-induced hypertension, in erythropoietin-induced hypertension, in pheochromocytoma-induced hypertension, and in pregnancy-induced hypertension.229 In general, although the exact role of endothelins in hypertension is not yet clear, the receptor antagonists may have a promising future in the control and prevention of cardiovascular disease.229 CHAPTER REFERENCES
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. Hypertension Detection and Follow-Up Program Cooperative Group. The hypertension detection and follow-up program. Circ Res 1977; 40(Suppl 1):106. The Sixth Report of the Joint National Committee on Detection, Evaluation and Treatment of High Blood Pressure (JNC VI). Arch Intern Med 1997; 157:2413. Working Group on Renovascular Hypertension. Detection, evaluation and treatment. Arch Intern Med 1987; 147:820. Mann SJ, Pickering TG. Detection of renovascular hypertension: state of the art. Ann Intern Med 1992; 117:845. Connolly JO, Higgens RM, Walters HL, et al. Presentation, clinical features and outcome in different patterns of atherosclerotic renovascular disease. Q J Med 1994; 87:413. Canzanello VJ, Textor SC. Noninvasive diagnosis of renovascular disease. Mayo Clin Proc 1994; 69:1172. Ram CVS. Renovascular hypertension. Curr Opin Nephrol Hypertens 1997; 6:575. McGrath BP, Clarke K. Renal artery stenosis: current diagnosis and treatment. Med J Aust 1993; 158:343. Rimmer JM, Gennari FJ. Atherosclerotic renal vascular disease and progressive renal failure. Ann Intern Med 1993; 118:712. Choudhri AH, Cleland JG, Rowlands PL. Unsuspected renal artery stenosis in peripheral vascular disease. Br Med J 1990; 301:1197. Pickering TG, Devereux RB, James GD. Recurrent pulmonary edema in hypertension due to bilateral renal artery stenosis: treatment by angioplasty or surgical revascularization. Lancet 1988; 2:551. Svetkey LP, Kadir S, Dunnick NR, et al. Similar prevalence of renovascular hypertension in selected blacks and whites. Hypertension 1991; 17:678. Pohl MA. Renal artery stenosis, renal vascular hypertension and ischemic nephropathy. In: Schrier RW, Gottschalk CW, eds. Disease of the kidney, 6th ed. Boston: Little Brown and Company, 1997:1367. Martinez-Maldonado M. Pathophysiology of renovascular hypertension. Hypertension 1991; 17:707. Mitchell KD, Braam B, Navar LG. Hypertensinogenic mechanisms mediated by renal actions of renin-angiotensin system. Hypertension 1992; 19(Suppl 1):I-18-1. Navar LG. The kidney in blood pressure regulation and development of hypertension. Med Clin North Am 1997; 8:1165. Mitchell KD, Navar LG. Intrarenal actions of angiotensin II in pathogenesis of experimental hypertension. In: Laragh JH, Brenner BM, eds. Hypertension: pathophysiology, diagnosis, and management. New York: Raven Press, 1995:1437. Guan S, Fox MJ, Mitchell KD, Navar LG. Angiotensin and angiotensin-converting enzyme tissue levels in two-kidney, one-clip hypertensive rats. Hypertension 1992; 20:763. Romero JC, Feldstein AE, Rodriguez-Porcel MG, et al. New insights into the pathophysiology of renovascular hypertension. Mayo Clin Proc 1997 72:251. Corry DB, Tuck ML. Secondary aldosteronism. Endocrinol Metab Clin North Am 1995; 24:511. Thornbury JR, Stanley JC, Fryback DG. Hypertensive urogram: a nondiscriminatory test for renovascular hypertension. AJR 1982; 138:43. Wilcox CS. Use of angiotensin-converting-enzyme inhibitors for diagnosing renovascular hypertension. Kidney Int 1993; 44:1379. Elliot WJ, Martin WB, Murphy MB. Comparison of two noninvasive tests for renovascular hypertension. Arch Intern Med 1993; 153:755. Postma CT, Van Der Steen PH, Hoefnagels WH, et al. The captopril test in the detection of renovascular disease in hypertensive patients. Arch Intern Med 1990; 150:625. Roubidoux MA, Dunnick NR, Klotman PE, et al. Renal vein renins: inability to predict response to revascularization in patients with hypertension. Radiology 1991; 178:819. Nally JV, Black HR. State of the art review: captopril renography patho-physiological considerations and clinical observations. Semin Nucl Med 1992; 22:85. Mann SJ, Pickering TG, Sos TA, et al. Captopril renography in the diagnosis of renal artery stenosis: accuracy and limitations. Am J Med 1991; 90:30. Halpern EJ, Deane CR, Needleman L, et al. Normal renal artery spectral Doppler wave form. A closer look. Radiology 1995; 196:667. Olin JW, Piedmonte MR, Young JR, et al. The utility of duplex ultrasound scanning of the renal arteries for diagnosing significant renal artery stenosis. Ann Intern Med 1995; 196:667. Rene PC, Oliva VL, Bui BT, et al. Renal artery stenosis: evaluation of Doppler US after inhibition of angiotensin converting enzyme with captopril. Radiology 1995; 196:675. Riehl J, Schmitt H, Bongatz, et al. Renal artery stenosis: Evaluation with colour duplex ultrasonography. Nephrol Dial Transplant 1997; 12:1608.

32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66.

Debatin JF, Spritzer CE, Grist TM, et al. Imaging of the renal arteries: value of MR angiography. AJR 1991; 157:981. Postma CT, Joosten FB, Rosenbusch G, Thien T. Magnetic resonance angiography has a high reliability in the detection of renal artery stenosis. Am J Hypertens 1997; 10:957. Rieumont MJ, Kaufman JA, Geller SC, et al. Evaluation of renal artery stenosis with dynamic gadolinium-enhanced MR angiography. AJR 1997; 169:39. DeHaan MW, Kouwenhoven M, Thelissen GRP, et al. Renovascular disease in patients with hypertension: detection with systolic and diastolic gating in three-dimensional, phase-contrast MR angiography. Radiology 1996; 198:449. Olbricht CJ, Arlet IP. Magnetic resonance angiographythe procedure of choice to diagnose renal artery stenosis? Nephrol Dial Transplant 1998; 13:1620. Olbricht CJ, Paul K, Prokop M, et al. Minimally invasive diagnosis of renal artery stenosis by spiral computed tomography angiography. Kidney Int 1995; 48:1332. Weibull H, Bergqvist D, Bergentz SE, et al. Percutaneous transluminal angioplasty versus reconstruction of atherosclerotic renal artery stenosis. Prospective randomized study. J Vasc Surg 1993; 18:841. Lawrie GM, Morris GC, Glaeser DH, DeBakey MI. Renovascular reconstruction: factors affecting long-term prognosis in 919 patients followed up to 31 years. Am J Cardiol 1989; 63:1085. Hansen KJ, Starr SM, Sands E, et al. Contemporary surgical management of renovascular disease. J Vasc Surg 1992; 16:310. Textor SC. Revascularization in atherosclerotic renal artery disease. Kidney Dis 1998; 53:799. Caps MT, Zierler RE, Polissar NL, et al. Risk of atrophy in kidneys with atherosclerotic renal artery stenosis. Kidney Int 1998; 53:735. Greco BA, Breyer JA. Atherosclerotic ischemic renal disease. Am J Kidney Dis 1997; 29:167. Derkx FH, Schalecamp MA. Renal artery stenosis and hypertension. Lancet 1994; 344:237. Ramsey LE, Weller PC. Blood pressure response to percutaneous angioplasty. an overview of published series. Br Med J 1990; 300:569. Bonelli FS, McKusick MA, Textor SC. Renal artery angioplasty: technical results and clinical outcome in 320 patients. Mayo Clin Proc 1995; 70:104. Plouin PF, Chatelleir G, Darne B, et al. Blood pressure outcome of angioplasty in atherosclerotic renal artery stenosis. A randomized trial. Hypertension 1998; 31:823. Van de Ven PJ, Beutler JJ, Kaatee R. Transluminal vascular stent for ostial atherosclerotic renal artery stenosis. Lancet 1995; 346:672. Blum U, Krumme B, Flugel P, et al. Treatment of ostial-artery stenosis with vascular endoprosthesis after unsuccessful balloon angioplasty. N Engl J Med 1997; 336:459. Shannon HM, Gillespie IN, Moss JG. Salvage of the solitary kidney by insertion of a renal artery stent. AJR 1998; 171:459. Hollenberg NK. The treatment of renovascular hypertension: surgery, angioplasty, and medical therapy with converting enzyme inhibitors. Am J Kidney Dis 1987; 10(Suppl 1):52. Pohl MA, Novick AC. Natural history of atherosclerotic and fibrous renal artery disease: clinical implications. Am J Kidney Dis 1985; 5(4):A120. Dean RH. Comparison of medical and surgical treatment of renovascular hypertension. Nephron 1986; 44(Suppl 1):101. Stradness DE Jr. Natural history of renal artery stenosis. Am J Kidney Dis 1994; 24:630. Marshall FI, Hagen S, Mahaffy RG, et al. Percutaneous transluminal angioplasty for theromatous renal artery stenosisblood pressure response and discriminant analysis of outcome predictors. Q J Med 1990; 75:483. Corvol P, Pinet F, Galen FX, et al. Seven lessons from seven renin-secreting tumors. Kidney Int 1988; 34(Suppl 25):S38. Mimran A. Renin-secreting tumors. In: Swales J, ed. Textbook of hypertension. Oxford, UK: Blackwell Science, 1994:858. Bruneval P, Fournier JG, Soubrier F, et al. Detection and localization of renin messenger mRNA in human pathologic tissues using in situ hybridization. Am J Pathol 1988; 131:320. Anderson PW, Macaulay L, Do YS, et al. Extrarenal renin-secreting tumors: insights into hypertension and ovarian renin production. Medicine 1990; 68:257. Misiani R, Sonzogni A, Poletti EM, et al. Hyponatremic hypertensive syndrome and massive proteinuria in a patient with renin-producing leiomyosarcoma. Am J Kidney Dis 1994; 24:83. Aurell M, Rudin A, Tisell LE, et al. Captopril effect on hypertension in patients with renin-producing tumor. Lancet 1979; 2:149. Kawai M, Sahashi K, Yamase H, et al. Renin producing adrenal tumor: report of a case. Surg Today 1998; 28:974. Carey RM, Wang ZQ, Siragy HM. Role of the angiotensin type 2 receptor in the regulation of blood pressure and renal function. Hypertension 2000; 35(1 Pt 2):155. Lifton RP, Dluhy RG, Powers M, et al. Glucocorticoid remediable hyperal-dosteronism: chimeric gene from cross-over between 11b-hydroxylase and aldosterone synthase genes. Nature 1992; 355:262. Snyder PM, Price MP, McDonald FJ, et al. Mechanism by which Liddle's syndrome mutations increase activity of a human epithelial Na + channel. Cell 1995; 83:969. White PC, Speiser PW. Steroid 11b-hydroxylase deficiency and related disorders. Endocrinol Metab Clin North Am 1994; 23:325.

66a.Farman N, Verrey F. Forefronts in nephrology news in aldosterone action. Kidney Int 2000; 57:1239. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. 103. 104. 105. 106. 107. 108. 109. 110. 111. 112. 113. 114. 115. 116. 117. 118. 119. 120. 121. 122. 123. 124. 125. 126. 127. 128. 129. 130. 131. 132. 133. 134. 135. 136. Malchoff DM, Brufsky A, Reardon G, et al. A mutation of the glucocorticoid receptor in primary cortisol resistance. J Clin Invest 1993; 91:1918. Norbiato G, Bevilacqua M, Vago T, et al. Glucocorticoid resistance and the immune function in the immunodeficiency syndrome. Ann NY Acad Sci 1998; 840:835. Gomez-Sanchez CE. Cushing's syndrome and hypertension. Hypertension 1986; 8:256. Whitworth JA. Studies on the mechanisms of glucocorticoid hypertension in humans. Blood Pressure 1994; 3:24. Danese RD, Aron DC. Cushing's syndrome and hypertension. In: Bravo EL, ed. Endocrine hypertension. Endocrinol Metab Clin North Am 1994; 23:299. Parks LL, Turney MK, Gaitan D, Kovacs WJ. Expression of 11b-hydroxy-steroid dehydrogenase type 2 in an ACTH-producing small cell lung cancer. J Steroid Biochem Mol Biol 1998; 67(4):341. Sowers JR, Epstein M. Diabetes mellitus and associated hypertension, vascular disease and nephropathy. An update. Hypertension 1995; 26(Pt 1): 869. Corry DB, Tuck ML. Insulin and glucoregulatory hormones: implications for antihypertensive therapy. In: Epstein M, ed. Calcium antagonists in clinical medicine. Philadelphia: Hanley and Belfus, 1998:217. Mogensen CE. The kidney and hypertension. Boston: Kluwer Academic Publishers, 1998:1. Chaturvedi N, Sjolie A-K, Stephenson JM, et al. Effect of lisinopril on progression of retinopathy in normotensive people with type 1 diabetes. Lancet 1998; 351:28. Groop LC, Kankuri M, Schalin-Jantti, et al. Association between polymorphism of the glycogen synthase gene and noninsulin-dependent diabetes. Kidney Int 1993; 328:10. Weidmann P, Boehlen LM, de Courten M. Pathogenesis and treatment of hypertension associated with diabetes mellitus. Am Heart J 1993; 125:1498. Stern N, Tuck ML. Pathogenesis of hypertension in diabetes. In LeRoith D, Taylor S, Olefsky J, eds. Diabetes mellitus: a fundamental and clinical text. Philadelphia: LippincottRaven Publishers, 1996:780. Nosandi R, Sambataro M, Thomaseth K, et al. Role of hyperglycemia and insulin resistance in determining sodium retention in noninsulin-dependent diabetes. Kidney Int 1993; 44:139. Tuck ML, Corry DB, Trujillo A. Salt-sensitive blood pressure and exaggerated vascular reactivity in the hypertension of diabetes mellitus. Am J Med 1990; 88:210. Randeree HA, Omar HA, Mortala AA, Seedat MA. Effect of insulin on blood pressure in NIDDM patients with secondary failure. Diabetes Care 1992; 15:1258. Trujillo A, Egenna P, Barrett J, Tuck M. Renin regulation in type II diabetes mellitus: influence of dietary sodium. Hypertension 1989; 13:200. Hseuh WA. Effect of renin-angiotensin system in the vascular disease of type II diabetes mellitus. Am J Med 1992; 92(Suppl):13S Tomiyama H, Kushiro T, Abeta A, et al. Kinins contribute to the improvement of insulin sensitivity during treatment with angiotensin converting enzyme inhibitor. Hypertension 1994; 23:450. Steinberg HO, Brechtel G, Johnson A, et al. Insulin-mediated skeletal muscle vasodilation is nitric oxide dependent. A novel mechanism for insulin resistance. J Clin Invest 1995; 96:786 O'Driscoll G, Green D, Rankin J, et al. Improvement in endothelial function by angiotensin converting enzyme in insulin-dependent diabetes mellitus. J Clin Invest 1997; 100:678. Reaven GM, Lithell H, Landsberg L. Hypertension and associated metabolic abnormalitiesthe role of insulin resistance and the sympathoadrenal system. N Engl J Med 1996; 334:374. Muller DC, Elahi D, Pratley RE, et al. An epidemiological test of the hyperin-sulinemia-hypertension hypothesis. J Clin Endocrinol Metab 1993; 76:544. Lithell HO. Hyperinsulinemia, insulin resistance and the treatment of hypertension. Am J Hypertens 1996; 9:154. He J, Klag MJ, Caballero B, et al. Plasma insulin levels and incidence of hypertension in African Americans and whites. Arch Intern Med 1999; 159:498. Tchernof A, Prud'homme D, Despres J-P, et al. Relation of steroid hormones to glucose tolerance and plasma insulin levels in men. Diabetes Care 1995; 18:292. Mantzorose CS, Georgiadis EI, Young P, et al. Relative androgenicity, blood pressure levels, and cardiovascular risk factors in young healthy women. Am J Hypertens 1995; 8:606. Masuo K, Mikami H, Ogihara T, Tuck ML. Differences in insulin and sympathetic responses to glucose ingestion due to family history of hypertension. Am J Hypertens 1996; 9:739. Hausberg M, Sinkey CA, Mark AL, et al. Sympathetic nerve activity and insulin sensitivity in normotensive offspring of hypertensive parents. Am J Hypertens 1998; 11:1312. Wu DA, Bu X, Warden CH, et al. Quantitative trait locus mapping of human blood pressure to a genetic region at or near the lipoprotein lipase gene locus on chromosome 8q22. J Clin Invest 1996; 97:2111. Steinberg HO, Chaker H, Leaming R, et al. Obesity/insulin resistance is associated with endothelial dysfunction. Implications for the syndrome of insulin resistance. J Clin Invest 1996; 97:2601. Zeng G, Quon MJ. Insulin-stimulated production of nitric oxide is inhibited by wotmannin. Direct measurement in vascular endothelial cells. J Clin Invest 1996; 98:894. Nickenig G, Roling J, Strehlow K, et al. Insulin induces upregulation of vascular AT1 receptor gene expression by postranscriptional mechanisms. Circulation 1998; 98:2453. Kamide K, Hori MT, Zhu J-H, Tuck ML. Insulin-mediated growth in aortic smooth muscle and the vascular renin-angiotensin system. Hypertension 1998; 32:482. Hall JE, Coleman TG, Mizelle HL, Smith MJ. Chronic hyperinsulinemia and blood pressure regulation. Am J Physiol 1990; 258:F722. Heise T, Magnusson K, Heinemann L, Sawicki PT. Insulin resistance and the effect of insulin on blood pressure in essential hypertension. Hypertension 1998; 32:243. Kaplan NM. Obesity, insulin and hypertension. Cardiovasc Risk Factors 1994; 4:133. Peiris AN, Sothman MS, Hoffman RG, et al. Adiposity, fat distribution and cardiovascular risk. Ann Intern Med 1989; 110:867. Tuck ML. Obesity, the sympathetic nervous system, and essential hypertension. Hypertension 1992; 19(Suppl 1):I-67. Reisin E. Sodium and obesity in the pathogenesis of hypertension. Am J Hypertens 1990; 3:164. Schmieder RE, Messerli FH. Does obesity influence early target organ damage in hypertensive patients? Circulation 1993; 87:1482. Ostlund RE, Staten M, Kohrt WM, et al. The ratio of waist-to-hip circumference, plasma insulin level, and glucose intolerance as independent predictors of the HDL2 cholesterol level in older adults. N Engl J Med 1990; 322:229. Abate N, Garg A, Peshock RM, et al. Relationship of generalized and regional adiposity to insulin sensitivity in man. J Clin Invest 1995; 96:88. Chan JM, Rimm EB, Colditz GA, et al. Obesity, fat distribution and weight gain as risk factors for clinical diabetes in men. Diabetes Care 1994; 17:961. Landsberg L. Hyperinsulinemia: possible role in obesity-induced hypertension. Hypertension 1992; 19(Suppl 1):I-61. Maxwell MH, Heber D, Waks AU, Tuck ML. Role of insulin and norepi-nephrine in the hypertension of obesity. Am J Hypertens 1994; 7:402. Vaz M, Jennings G, Turner A, et al. Regional sympathetic and oxygen consumption in obese normotensive human subjects. Circulation 1997; 6:3423. Grassi G, Seravalle G, Cattaneo BM, et al. Sympathetic activation in obese normotensive subjects. Hypertension 1995; 25[Pt I]:560. Narkiewiez K, van de Borne PJ, Cooley RL, et al. Sympathetic activity in obese subjects with and without obstructive sleep apnea. Circulation 1998; 98:772. Haynes WG, Morgan DA, Walsh SA, et al. Receptor-mediated regional sympathetic nerve activation by leptin. J Clin Invest 1997; 100:270 Wilden E, Lehto M, Kanninen T, et al. Association of a polymorphism in the b3-adrenergic receptor gene with features of the insulin resistance syndrome in Finns. N Engl J Med 1995; 333:348. Reisin E, Mathew R, Falkner B, et al. Lisinopril versus hydrochlorothiazide in obese hypertensive patients. A multicenter placebo-controlled trial. Hypertension 1997; 30[Pt I]:140. Lind L, Hvarfner A, Palmer M, et al. Hypertension in primary hyperpar-athyroidism in relation to histopathology. Eur J Surg 1991; 157:457. Brickman AS, Nyby MD, vonHungen K, Tuck ML. Calciotropic hormones, platelet calcium and blood pressure in essential hypertension. Hypertension 1990; 16:515. Brickman AS, Stern N, Sowers JR. Hypertension in pseudohypoparathy-roidism. Am J Med 1988; 85:785. Gennari C, Nami R, Gonelli A. Hypertension and primary hyperparathyroidism: the role of adrenergic and renin-angiotensin-aldosterone systems. Miner Electrolyte Metab 1995; 21:77. Sancho JJ, Rouco J, Riera-Vidal R, Sitges-Serra A. Long-term effects of para-thyroidectomy for primary hyperparathyroidism on arterial hypertension. World J Surg 1992; 16:732. Zawada E Jr, Brickman AS, Maxwell MH, Tuck ML. Hypertension associated with hyperparathyroidism is not responsive to angiotensin blockade. J Clin Endocrinol Metab 1980; 50:912. Sowers JR, Standley PR, Tuck ML, Ram JL. Calcium and calcium regulatory hormones in hypertension. In: Laragh JH, Brenner BM, eds. Hypertension: pathophysiology, diagnosis and management, 2nd ed. New York: Raven Press, 1995:1155. Resnick L. The cellular ionic basis of hypertension and allied clinical conditions. Prog Cardiovasc Dis 1999; 42(1):1. Ezzat S, Forster MJ, Berchtold P, et al. Acromegaly: clinical and biochemical features in 500 patients. Medicine 1994; 73:233. Kratz C, Benker G, Weber F, et al. Acromegaly and hypertension: prevalence and relationship to the renin-angiotensin-aldosterone system. Klin Wochenschr 1990; 68:583. Connell JMC, Davies DL. Endocrine hypertension: thyroid disease and acromegaly. In: Swales JD, ed. Textbook of hypertension. Oxford, UK: Blackwell Science, 1994:959. Lopez-Velasco R, Escobar-Morreale HF, Vega B, et al. Cardiac involvement in acromegaly: specific myocardiopathy or consequence of systemic hypertension? J Clin Endocrinol Metab 1997; 82:1047. Akpununu BE, Mulrow PJ, Hoffman EA. Secondary hypertension: evaluation and treatment: thyrotoxicosis and hypertension. Dis Mon 1996; 42:689 and 42:696. Saito I, Saruta T. Hypertension and thyroid disorders. Endocrinol Metab Clin North Am 1994; 23:379. Levy GS, Klein I. Catecholamine-thyroid hormone interactions and cardiovascular manifestations of hyperthyroidism. Am J Med 1990; 88:642. Klein I, Ojama K. Cardiovascular manifestations of endocrine disease. J Clin Endocrinol Metab 1992; 75:339. Streeten DHP, Anderson GH, Howland T, et al. Effects of thyroid function on blood pressure: recognition of hypothyroid hypertension. Hypertension 1988; 11:78. Laragh JH, Sealey JE. The renin-angiotensin-aldosterone system and the renal regulation of sodium and potassium and blood pressure homeostasis. In: Windhager EE, ed. Handbook of renal physiology. New York: Oxford University Press, 1992:1409.

137. 138. 139. 140. 141. 142. 143. 144. 145. 146. 147. 148. 149. 150. 151. 152. 153. 154. 155. 156. 157. 158. 159. 160. 161. 162. 163. 164. 165. 166. 167. 168. 169. 170. 171. 172. 173. 174. 175. 176. 177. 178. 179. 180. 181. 182. 183. 184. 185. 186. 187. 188. 189. 190. 191. 192. 193. 194. 195. 196. 197. 198. 199. 200. 201. 202. 203. 204. 205. 206. 207. 208. 209. 210.

Alderman MH, Madhavan S, Ooi WL, et al. Association of the renin-sodium profile with risk of myocardial infarction in patients with hypertension. N Engl J Med 1991; 324:1098. Trenkwalder P, James GD, Laragh JH, Sealey JE. Plasma renin activity and plasma prorenin are not suppressed in hypertensives surviving to old age. Am J Hypertens 1996; 9:621. Blaufox MD, Lee HB, Davis B, et al. Renin predicts the blood pressure response to nonpharmacologic and pharmacologic therapy. JAMA 1992; 267:1221. Wyndham RN, Gimenez L, Walker WG, et al. Influence of renin levels on the treatment of essential hypertension with thiazide diuretics. Arch Intern Med 1987; 147:1021. Rich GM, Ulick S, Cook S, et al. Glucocorticoid-remediable aldosteronism in a large kindred: clinical spectrum and diagnosis using a characteristic biochemical phenotype. Ann Intern Med 1992; 116:813. Soro A, Ingram MC, Tonolo G, et al. Evidence of coexisting changes in 11b-dehydrogenase and 5b-reductase activity in subjects with untreated essential hypertension. Hypertension 1995; 25:67. Snyder PM, Price MP, McDonald FJ, et al. Mechanism by which Liddle's syndrome mutations increase activity of a human epithelial Na+ channel. Cell 1995; 83:969. Griffing GT, Dale SL, Holbrook MM, Melby JC. The regulation of urinary free 19-nordeoxycorticosterone and its relation to systemic arterial blood pressure in normotensive and hypertensive subjects. J Clin Endocrinol Metab 1983; 56:99. Resnick LM, Muller FB, Laragh JH. Calcium-regulating hormones in essential hypertension. Relation to plasma renin activity and sodium metabolism. Ann Intern Med 1986; 105:649. Sealey JE, Blumenfeld JD, Bell GM, et al. On the renal basis for essential hypertension: nephron heterogeneity with discordant renin secretion and sodium excretion causing a hypertensive vasoconstriction-volume relationship. J Hypertens 1988; 6:763. Williams GH. Converting-enzyme inhibitors in the treatment of hypertension. N Engl J Med 1988; 319:1517. Redgrave J, Rabinowe S, Hollenberg NK, Williams GH. Correction of abnormal renal blood flow response to angiotensin II by converting enzyme inhibitor. J Clin Invest 1985; 75:1285. Williams GH, Fisher ND, Hunt SC, et al. Effects of gender and genotype on the phenotypic expression of nonmodulating essential hypertension. Kidney Int 2000; 57(4):1404. Johnston CI, Franz Volhard Lecture. Renin-angiotensin system: a dual tissue and hormonal system for cardiovascular control. J Hypertens 1992; 10(Suppl 7):S13. Seikaly MG, Arant BS, Seney FD. Endogenous angiotensin concentration in specific intrarenal fluid compartments of the rat. J Clin Invest 1990; 86:1352. Bachman S, Peters J, Engler E, et al. Transgenic rats carrying the mouse renin gene-morphologic characterization of a low-renin hypertension model. Kidney Int 1992; 41:24. Jeunemaitre X, Soubrier F, Kotelevtsev YV, et al. Molecular basis of human hypertension: role of angiotensinogen. Cell 1992; 71:169. Hunt SC, Cook NR, Oberman A, et al. Angiotensinogen genotype, sodium retention, weight loss and prevention of hypertension: trials of hypertension prevention. Hypertension 1998; 32:393. Lindpianter K, Pfeffer MA, Kreutz R, et al. A prospective evaluation of an angiotensin-converting enzyme gene polymorphism and the risk of ischemic heart disease. N Engl J Med 1995; 332:706. Kauma H, Piavansalo M, Savolainen MJ, et al. Association between angiotensin converting enzyme gene polymorphism and carotid atherosclerosis. J Hypertens 1996; 14:1183. Takezako T, Saku K, Zhang B, et al. Angiotensin I converting enzyme gene polymorphism and insulin resistance in patients with angina pectoris. Am J Hypertens 1999; 12:291. Marre M, Bernadet P, Gallois Y, et al. Relationships between angiotensin I converting enzyme gene polymorphism, plasma levels, and diabetic retinal and renal complications. Diabetes 1994; 43:384. Fujisawa T, Ikegami H, Kawaguchi Y, et al. Meta-analysis of association of insertion/deletion polymorphism of angiotensin I-converting gene with diabetic nephropathy and retinopathy. Diabetologia 1998; 41:47. Julius S. Changing role of the autonomic nervous system in human hypertension. J Hypertens 1990; 9(Suppl 7):S59. Goldstein DS, Kopin IJ. The autonomic nervous system and catecholamines in normal blood pressure control and in hypertension. In: Laragh JH, Brenner BM, eds. Hypertension: Pathophysiology, Diagnosis and Management. New York: Raven Press, 1990:711. Esler M, Jennings G, Lambert G, et al. Overflow of catecholamine neurotransmitters to the circulation: source, fate, and function. Physiol Rev 1990; 70:963. Mark AL. The sympathetic nervous system in hypertension: a potential long-term regulator of arterial pressure. J Hypertens 1996; 14(Suppl):S159. Izzo JL. Sympathoadrenal activity, catecholamines and the pathogenesis of vas-culopathic hypertensive target-organ damage. Am J Hypertens 1989; 2:3055. Brooks VL, Osborn JW. Hormonal-sympathetic interactions in the long-term of arterial pressure: an hypothesis. Am J Physiol 1995; 268:R1343. McCarty R, Gold P. Catecholamine, stress and disease: a psychobiologic perspective. Psycosom Med 1996; 58:590. Ichihara A, Inscho EW, Imig JD, et al. Role of renal nerves in afferent arteriolar reactivity in angiotensin-induced hypertension. Hypertension 1997; 29:442. Ascio LP, Ladiness C, Fuchs S, et al. Distribution of the dopamine D3 receptor gene produces renin-dependent hypertension. J Clin Invest 1998; 102:493. Pa J, Eisner GM. Renal dopamine receptors in health and hypertension. Pharmacol Ther 1998; 80:149. Gill JR, Grossman E, Goldstein DS. High urinary dopa excretion and low urinary dopamine: dopa ratio in salt-sensitive hypertension. Hypertension 1991; 18:614. Weinberger MH. Salt sensitivity of blood pressure in humans. Hypertension 1996; 27:481. Lijnen P. Alterations in sodium metabolism as an etiological model for hypertension. Cardiovasc Drugs 1995; 9:377. Corry DB, Tuck ML. Altered sodium and potassium metabolism in the pathogenesis of hypertension. In: Maxwell MH, Kleeman C, Narrins R, eds. Disorders of fluid and electrolyte metabolism, 5th ed. New York: McGraw-Hill, 1993:1245. Bobick A, Neylon CB, Little PJ. Disturbances of vascular smooth muscle cation transport and the pathogenesis of hypertension. In: Swales JD, ed. Textbook of hypertension. Oxford, UK: Blackwell Science, 1994:175. Sweeney G, Somwar R, Ramlal T, et al. Insulin stimulation of K+ uptake in 3T3-L1 fibroblasts involves phosphatidylinositol 3-kinase and protein kinase C-zeta. Diabetologia 1998; 41:1199. Blaustein MP. Physiologic effects of endogenous ouabain: control of intracellular calcium stores and cell responsiveness. Am J Physiol 1993; 264:C1367. Berger ME, Ormsby BL, Bunnag P, et al. Increased functional Na( +)-K( +) pump activity in the vasculature of fructose-fed hyperinsulinemic and hypertensive rats. Hypertens Rev 1998; 21(2):73. Rutherford PA, Thomas TH, Wilkinson R. Na-Li countertransport kinetics in the relatives of hypertensive patients with abnormal Na-Li countertransport activity. Biochem Mol Med 1997; 62(1):106. Strazzulo P, Siani A, Cappuccio FP, et al. Red blood cell sodium-lithium countertransport and risk of future hypertension: the Olivetti Prospective Heart Study. Hypertension 1998; 31(6):1284. Van Norren K, Thien T, Berden JH, et al. Relevance of erythrocyte Na+/Li + countertransport measurement in essential hypertension, hyperlipidemia and diabetic nephropathy: a critical review. Eur J Clin Invest 1998; 28(5):339. Ragone E, Strazullo P, Siani A, et al. Ethnic differences in red blood cell sodium/lithium countertransport and metabolic correlates of hypertension: an international collaborative study. Am J Hypertens 1998; 11(8; Pt 1): 935. Sanchez RA, Gimenez MI, Migliorini M, et al. Erythrocyte sodium-lithium countertransport in non-modulating offspring and essential hypertensive individuals: response to enalapril. Hypertension 1997; 30:99. Monciotti CG, Semplicini A, Morocutti A, et al. Elevated sodium-lithium countertransport activity in erythrocytes is predictive of the development of microalbuminuria in IDDM. Diabetologia 1997; 40(6):654. Carr SJ, Moore D, Sikand K, Norman RI. Raised affinity for extracellular sodium of the sodium-lithium countertransporter is associated with a family history of hypertension and uraemia in patients with renal disease. Clin Sci 1997; 92(5):497. Jones SC, Thomas TH, Marshall SM. Thiol group modulation of sodium-lithium countertransport kinetics in diabetic nephropathy. Diabetologia 1997; 40:1079. Giordano M, Castellino P, Solini A, et al. Na +/Li + and Na+/H+ counter-transport activity in hypertensive noninsulin-dependent diabetic patients: role of insulin resistance and antihypertensive treatment. Metabolism 1997; 46:1326. Siffert W, Dusing R. Sodium-proton exchange and primary hypertension: an update. Hypertension 1995; 26:649. Ng LL, Sweeney FP, Siczkowski M, Davies JE, et al. Na( +)-H(+) antiporter phenotype, abundance, and phosphorylation of immortalized lymphoblasts from humans with hypertension. Hypertension 1995; 25:971. Rosskopf D, Schroder KJ, Siffert W. Role of sodium-hydrogen exchange in the proliferation of immortalised lymphoblasts from patients with essential hypertension and normotensive subjects. Cardiovasc Res 1995; 29:254. Garciandia A, Lopez R, Tisaire J, et al. Enhanced Na(+)-H(+) exchanger activity and NHE-1 mRNA expression in lymphocytes from patients with essential hypertension. Hypertension 1995; 25(3):356. Diez J, Alonso A, Garciandia A, et al. Association of increased erythrocyte Na +/H+ exchanger with renal Na + retention in patients with essential hypertension. Am J Hypertens 1995; 8(2):124. LaPointe MS, Ye M, Moe OW, et al. Na+/H+ antiporter (NHE-1 isoform) in cultured vascular smooth muscle from the spontaneously hypertensive rat. Kidney Int 1995; 47(1):78. Siczkowski M, Davies JE, Ng LL. Na( +)-H(+) exchanger isoform 1 phosphorylation in normal Wistar-Kyoto and spontaneously hypertensive rats. Circ Res 1995; 76:825. Siffert W, Rosskopf D, Moritz A, et al. Enhanced G protein activation in immortalized lymphoblasts from patients with essential hypertension. J Clin Invest 1995; 96:759. Takahashi E, Abe J, Berk BC. Angiotensin II stimulates p90rsk in vascular smooth muscle cells. A potential Na( +)-K( +) exchanger kinase. Circ Res 1997; 81(2):268. Koren W, Postnov IY, Postnov YV. Increased Na(+)-H(+) exchange in red blood cells of patients with primary aldosteronism. Hypertension 1997; 29(2):587. Fortuno A, Tisaire J, Lopez R, et al. Angiotensin converting enzyme inhibition corrects Na +,H+ exchanger overactivity in essential hypertension. Am J Hypertens 1997; 10(1):84. Gende OA. Chelerythrine inhibits Na(+)-H(+) exchange in platelets from spontaneously hypertensive rats. Hypertension 1996; 28(6):1013. Chen YF, Yang RH, Meng QC, et al. Sodium-proton (Na +/H+) exchange inhibition increases blood pressure in spontaneously hypertensive rats. Am J Med Sci 1994; 308:145. Trevisan R, Viberti G. Sodium-hydrogen antiporter: its possible role in the genesis of diabetic nephropathy. Nephrol Dialysis Transplant 1997; 12(4):643. Ng LL, Davies JE, Siczkowski M, et al. Abnormal Na +/H+ antiporter phenotype and turnover of immortalized lymphoblasts from type 1 diabetic patients with nephropathy. J Clin Invest 1994; 93(6):2750. Siczkowski M, Davies JE, Sweeney FP, et al. Na+/K+ exchanger isoform-1 abundance in skin fibroblasts of type I diabetic patients with nephropathy. Metabolism 1995; 44:791. Tuck ML, Corry DB, Maxwell M, et al. Erythrocyte Na+,K+ cotransport and Na+,K+ pump in black and Caucasian hypertensive patients: a kinetic analysis. Hypertension 1987; 10:204. Tepper T, Sluiter WJ, Huisman RM, deZeeuw D. Co-transport measurement in essential hypertension: useful diagnostic tool or failure? A meta-analysis of 17 years of literature. Pares I, de la Sierra A, Coca MM, et al. Detection of a circulating inhibitor of the Na( +)-K( +)-Cl( ) cotransport system in plasma and urine after high salt intake. Am J Hypertens 1995; 8:965. Blaustein MP. The physiological effects of endogenous ouabain: control of cell responsiveness. Am J Physiol 1993; 264:C1367. Hamlyn JM, Hamilton BP, Manunta P. Endogenous ouabain, sodium balance and blood pressure: a review and an hypothesis. J Hypertens 1996; 14:151. Wilkens MR, Redondo J, Brown LA. The natriuretic peptide family. Lancet 1997; 349:1307 Koller KJ, Goeddel DV. Molecular biology of the natriuretic peptides and their receptors. Circulation 1992; 86:1081. Burnett JC Jr. Vasopeptidase inhibition: a new concept in blood pressure management. J Hypertens 1999; 17(Suppl 1):S37.

210a.Schriger JA, Grantham JA, Kullo IJ, et al. Vascular actions of brain natri-uretic peptide: modulation by atherosclerosis and neutral endopeptidase inhibition. J Am Coll Cardiol 2000; 35:796. 211. 212. 213. 214. 215. 216. 217. 218. 219. 220. 221. 222. 223. 224. 225. Lainchbury JG, Espinar EA, Frampton CM, et al. Cardiac natriuretic peptides as predictors of mortality. J Intern Med 1997; 241:257. Bhoola KD, Figueroa CD, Worthy K. Bioregulation of kinins: kallikreins, kininogens, and kinases. Pharmacol Rev 1992; 44:1. Carretero OA, Scili AG. The kallikrein-kinin system. In: Fozzard HA, ed. The heart and cardiovascular system. New York: Scientific Foundations, Raven Press, 1991:1851. Liu LH, Yang XP, Sharov VG, et al. Effects of angiotensin-converting enzyme inhibitors and angiotensin II type 1 receptor antagonists in rats with heart failure: role of kinins and angiotensin type 2 receptor. J Clin Invest 1997; 99:1926. Yang XP, Liu YH, Scicli GM, et al. Role of kinins in the cardioprotective effect of preconditioning: study of myocardial ischemia/reperfusion injury in B 2 kinin receptor knockout mice and kininogen-deficient rats. Hypertension 1997; 30:735. Dattilo JB, Makhoul RG. The role of nitric oxide in vascular biology and pathobiology. Ann Vasc Surg 1997; 11:307. Ignarro L, Murad F. Nitric oxide biochemistry, molecular biology, and therapeutic implications. In: JT August, MW Anders, F Murad, ed. Advances in Pharmacology. New York: Academic Press, 1995. Garcia-Cardena G, Martasek P, Masters BS, et al. Dissecting the interaction between nitric oxide synthase (NOS) and caveolin: functional significance of the NOS caveolin binding domain in vivo. J Biol Chem 1997; 272:25437. Busse R, Fleming I. Nitric oxide, nitric oxide synthase and hypertensive vascular disease. Curr Hypertension Rep 1999; 1:88. Garcia-Cardena G, Fan R, Shah V, et al. Dynamic activation of endothelial nitric oxide by Hsp90. Nature 1998; 292:821. Murad F. The 1996 Albert Lasker Medical Research Awards. Signal transduction using nitric oxide and cyclic guanosine monophosphate. JAMA 1996; 276:1189. Haller H. Endothelial function: general considerations. Drugs 1997; 53(Suppl 1):1. Ferro CJ, Webb DJ. Endothelial dysfunction and hypertension. Drugs 1997; 53 (Suppl 1):30. Nakayama T, Soma M, Takahashi Y, et al. Association analysis of CA repeat polymorphism of the endothelial nitric oxide synthase gene with essential hypertension in Japanese. Clin Genet 1997; 51:26. Miyamoto Y, Saito Y, Kajiyama N, et al. Endothelial nitric oxide synthase gene expression is positively associated with essential hypertension. Hypertension 1998; 32:3.

226. Bauersachs J, Bouloumie A, Mulsch A, et al. Vasodilator dysfunction in aged spontaneously hypertensive rats; changes in NO synthase III and soluble guanylyl expression and in superoxide anion production. Cardiovasc Res 1998; 37:772. 227. Luscher TF, Oemar BS, Boulanger CM, Hahn AWA. Molecular and cellular biology of endothelin and its receptors, Pts 1 and 2. J Hypertens 1993; 11:7 and 11:121. 228. Schiffrin EL, Intengan HD, Thibault G, Touyz RM. Clinical significance of endothelin in cardiovascular disease. Curr Opin Cardiol 1997; 12:354. 229. Moreau P, Rabelink TJ. Endothelin and its antagonists in hypertension: can we foresee the future? Curr Hypertens Rep 1999; 1:69. 230. Krum H, Viskoper RV, Lacourciere Y, et al. The effects of an endothelin receptor antagonist, bosantin, on blood pressure in patients with essential hypertension. N Engl J Med 1998; 338:784.

CHAPTER 83 ADRENOCORTICAL DISORDERS IN INFANCY AND CHILDHOOD Principles and Practice of Endocrinology and Metabolism

CHAPTER 83 ADRENOCORTICAL DISORDERS IN INFANCY AND CHILDHOOD

ROBERT L. ROSENFIELD AND KE-NAN QIN General Principles Adrenocortical Insufficiency Primary Form Etiology Clinical Characteristics Diagnosis Secondary Form Etiology and Clinical Characteristics Diagnosis Treatment Corticosteroid Coverage for Stress Hypercortisolism Clinical Characteristics Diagnosis Treatment Withdrawal from Corticosteroid Excess in Children Precocious Pseudopuberty Etiology and Clinical Characteristics Diagnosis Treatment Syndromes Involving Mineralocorticoid Excess Varieties of Mineralocorticoid Excess Primary Hyperaldosteronism Excess of Non-Aldosterone Corticoids Pseudoaldosteronism Secondary Hyperaldosteronism Bartter Syndrome Differential Diagnosis and Treatment Chapter References

GENERAL PRINCIPLES
Alo-ng with the well-known perturbations of fluid, electrolyte, and glucose homeostasis, adrenocortical diseases that occur in children also cause disturbed growth. The proper management of children with these disorders requires careful documentation of height and weight at regular intervals. Physicians who are relatively unfamiliar with pediatric patients may assume that the fluid and electrolyte requirements of infants and children are similar to those of adults. Actually, infants are hypermetabolic, as compared to adults, because of the relatively large size of their high-energy-consuming organs (i.e., the brain, heart, liver, and kidneys) as compared to their somatic size. Water requirements change in proportion to caloric requirements. One milliliter of water is required for each kilocalorie of energy expenditure. Water and calorie requirements are approximately constant throughout life relative to surface area (1500 kcal/m2 per day). Surface area can be calculated as the square root of (cm kg/3600). Normal daily sodium and potassium maintenance requirements are ~2 mEq/dL and 1 mEq/dL water, respectively. When calculating electrolyte replacement, it must be remembered that the exchangeable fluid compartment is relatively larger in children than in adults, ranging from 60% of body weight in small children to 40% in adults. Because of the rapid rate of turnover of children's body fluids, the electrolyte concentrations of parenteral fluids should be distributed evenly throughout the day to prevent shifts in the tonicity of body fluids. Infants and children are intolerant of prolonged fasting because of their high metabolic rate and functionally immature gluconeogenic enzyme systems. When fasted, the normal young child will become hypoglycemic within as few as 20 hours. To prevent glycogenolysis, infants and young children need 6 to 8 mg/kg per minute of glucose.1 The production of glucocorticoids and the excretion of their metabolites are essentially constant with respect to surface area and lean body mass throughout life. The daily secretion rate of cortisol approximates 6 to 8 mg/m2.2 Urinary cortisol excretion ranges from 15 to 70 g/m2 or g/g creatinine. Urinary 17-hydroxycorticoid excretion (measured as Porter-Silber chromogens) is a poor index of cortisol secretion in the neonate due to deficient glucuronyl-transferase activity in the first month of life. Plasma cortisol concentrations vary from 4 to 25 g/dL, as in adults. The circadian rhythm of adrenocortical secretion is established by 3 months of age.3 The aldosterone secretion rate is constant throughout life at ~80 g per day.4 Thus, in infants it is several times more than that of adults in terms of surface area. Plasma aldosterone levels and plasma renin activity (PRA) are higher in infants than subsequently; the level of these two constituents gradually declines during childhood (Table 83-1), but absolute levels vary among laboratories.5,6 Plasma dehydroepiandrosterone (DHEA) and DHEA sulfate (DHEAS) levels are at adult levels at birth as a consequence of the function of the fetal zone of the adrenal gland, fall during early infancy, and rise with adrenarche7,8 (Table 83-2).

TABLE 83-1. Typical Normal Plasma Values in Infancy and Childhood for Renin-Aldosterone Axis

TABLE 83-2. Basal Corticosteroids: Typical Plasma Values in Normal Infants, Children, and Adults

ADRENOCORTICAL INSUFFICIENCY
PRIMARY FORM ETIOLOGY Congenital Adrenal Hyperplasia and Hypoplasia. The most prevalent cause of adrenocortical insufficiency in infancy is congenital adrenal hyperplasia (Fig. 83-1). This condition is discussed extensively in Chap. 77. Further details are emphasized in the sections Treatment and Precocious Pseudopuberty.

FIGURE 83-1. Congenital adrenal hyperplasia in three newborn infants, all with 21-hydroxylase deficiency. The female infants depicted in A and B have different degrees of ambiguity of the external genitalia. A, Clitoromegaly with labial enlargement. B, Severe clitoral hypertrophy and nearly complete labial fusion. Note the scrotal appearance of the labia. C, Male infant with precocious development of the external genitalia and rapid somatic growth (infant Hercules). (Courtesy of Dr. Wellington Hung.)

By contrast, congenital adrenal hypoplasia is a rare cause of adrenal insufficiency in infancy; it has been reported to occur as an apparently isolated defect, with or without a familial tendency.9 The autosomal recessive form is associated with miniature, adult-type, adrenal architecture and may be secondary to isolated adrenocorticotropic hormone (ACTH) deficiency. The X-linked form is due to DAX-1 gene mutations and is associated with cytomegaly of fetal-like adrenocortical cells and congenital gonadotropin deficiency.10 Mental retardation and other features may occur, depending on the extent of contiguous chromosomal deletion. Neonatal Adrenal Hemorrhage. Neonatal adrenal hemorrhage usually presents with the symptoms of shock, severe jaundice, and an abdominal mass during the first week of life.11 It is associated with birth trauma and is thought to result from severe hypoxemia. Ultrasonographic examination, repeated at 3- to 5-day intervals, may reveal a lesion that changes from solid to cystic as the hemorrhagic mass liquefies, and adrenal calcifications may be noted after 4 weeks of age. Children who survive should be tested for chronic adrenal insufficiency. Destruction of the Adrenal Cortex. Beyond the newborn period, adrenal insufficiency usually is attributable to destruction of the adrenal cortex. Addison disease usually is caused by autoimmune endocrinopathy that may be associated with diseases affecting other glands (see Chap. 76 and Chap. 197). The major autoantigens are the steroidogenic enzymes, particularly 21-hydroxylase (cytochrome P450c21),12 but their presence does not necessarily predict adrenal failure.13 There are two varieties of autoimmune polyglandular deficiency syndrome. Type I usually is seen in early life, with a mean age at onset of 12 years. Although Addison disease in this syndrome is commonly associated with hypoparathyroidism and mucocutaneous candidiasis, the incidence of associated type 1 diabetes mellitus and thyroid disease is low.14 Pernicious anemia and other autoimmune disorders are seen occasionally. The disorder is an autosomal recessive disease due to mutation of an autoimmune regulator gene on chromosome 21q22.3.15 Type II autoimmune polyglandular deficiency has a mean age at onset of 30 years. This form of the disease is associated with insulin-dependent diabetes mellitus, thyroid deficiency, or both, and has an autosomal dominant transmission pattern linked with the human leukocyte antigen (HLA)-B8 allele. Both type I and type II can be associated with vitiligo and primary gonadal failure. Acute adrenal insufficiency can complicate septic shock. The classic association of acute adrenal hemorrhage and purpura with fulminating meningococcemia (Waterhouse-Friderichsen syndrome)16 (Fig. 83-2; see Chap. 213). It resembles the generalized Shwartzman phenomenon.

FIGURE 83-2. Seven-year-old boy with Waterhouse-Friderichsen syndrome secondary to severe menin-gococcemia. The child subsequently died. Note the massive hemorrhaging into the skin. (Courtesy of Dr. Wellington Hung.)

Progressive degenerative or granulomatous diseases such as tuberculosis or histoplasmosis may involve the adrenal cortex and cause adrenal failure. Lysosomal acid lipase deficiency, or Wolman disease, is a disorder that occurs in infancy. It is caused by abnormal storage of triglycerides and cholesterol esters that cannot be degraded by the adrenals, spleen, liver, bone marrow, brain, or lymph nodes. The form of transmission is autosomal recessive.17 Patients present in the early weeks of life with vomiting, diarrhea, failure to thrive, hepatosplenomegaly, and adrenal calcification. Invariably, the patient progresses to death by a few months of age despite corticosteroid treatment. Adrenoleukodystrophy is a rare, fatal, or autosomal recessive demyelinating disorder of the brain that presents in childhood with blindness, quadriparesis, or dementia, typically before adrenal failure.18,19 Adrenomyeloneuropathy, a more slowly progressive form, may present in boys as adrenal failure. These disorders result from accumulation of very long-chain fatty acids (C-24 and longer). Most forms are X-chromosomelinked, due to mutations of a gene encoding a peroxisomal integral membrane protein that is located on Xq28. Neonatal adrenoleukodystrophy is an autosomal recessive disease caused by defects in peroxisome assembly; it results from mutations of PEX genes.19 Isolated Glucocorticoid Insufficiency. Glucocorticoid deficiency in the presence of normal mineralocorticoid secretion can result from primary adrenal disease or be secondary to defects in the secretion or mode of action of ACTH. In the early stages of autoimmune Addison disease, aldosterone secretion is sometimes intact. This must be distinguished from ACTH deficiency, which has been reported in the polyglandular autoimmune syndrome.20 Familial isolated glucocorticoid deficiency appears to be caused by a heterogeneous group of autosomal recessive disorders. These include degenerative disorders of

the zona fasciculata and defects in the ACTH receptor, or ACTH action.21 ACTH resistance may result from degeneration of the zona fasciculata-reticularis in association with achalasia of the esophagus and alacrima (triple-A syndrome).21 Approximately 15% of patients later develop mineralocorticoid deficiency. This disorder maps to a critical region on the chromosome 12q13 locus. Patients with isolated ACTH receptor defect average 2 standard deviations (SD) taller than average. Isolated Mineralocorticoid Deficiency. Isolated aldosterone deficiency may result from defects in the terminal portion of the aldosterone biosynthetic pathway or from hyporeninemia.22,23 Defective corticosterone 18-hydroxylation causes corticosterone overproduction. Defective 18-dehydrogenation (oxidation) of 18-hydroxycorticosterone is characterized biochemically by corticosterone and 18-hydroxycorticosterone overproduction (see Chap. 81). These enzyme activities, as well as 11b-hydroxylase, are all located on a single enzyme, cytochrome P450c11. Symptoms are reported to lessen as patients grow older. Isolated hypoaldosteronism resulting from defects in the renin-angiotensin system may be familial in infancy23a (also see Chap. 81). Pseudohypoaldosteronism type I is a syndrome of renal sodium chloride loss accompanied by hyperkalemia and metabolic acidosis that occurs as the result of an end-organ defect in the distal renal tubule that causes a failure of response to endogenous aldosterone. PRA, plasma aldosterone concentration, and urinary aldosterone excretion are usually elevated. The autosomal dominant form is a mild disease that remits with age and is usually due to mineralocorticoid receptor mutations.24 The severe form that persists into adulthood is usually caused by autosomal recessive inactivating mutations in genes encoding subunits of the amiloride-sensitive epithelial sodium channel.25 Pseudohypoaldosteronism type II is a very different form of aldosterone resistance. Hyperkalemia and hyperchloremic acidosis (type IV renal tubular acidosis) are accompanied by low or normal renin and aldosterone levels. It may result from a familial disorder in which hypertension is associated with an isolated increase in reabsorption of chloride by the renal tubule, a defect that can be corrected by hydrochlorothiazide.26 This is an autosomal dominant disorder that has been mapped to two different loci, chromosomes 1q3142 and 17p11q21.27 Either type of pseudohypoaldosteronism can also be acquired as the result of obstructive uropathy, sickle cell or lead nephropathy, amyloidosis or urinary tract infections.28,29 CLINICAL CHARACTERISTICS The characteristic, early symptoms of glucocorticoid deficiencyanorexia, weakness, and fatiguemay not be easy to elicit in the infant or young child. Abdominal pain, vomiting, and diarrhea are the nonspecific symptoms. In congenital forms, there may be a history of neonatal cholestatic jaundice due to hepatitis.21 Failure to thrive (particularly cachexia) may be the presenting complaint. Symptoms of hypoglycemia associated with ketosis often are the most prominent features (see Chap. 161). Anorexia, listlessness or irritability, tachypnea, tachycardia, emesis, and hypotonia may result from acute hypoglycemia. Hunger and excessive perspiration are unusual symptoms in young children. Hyperpig-mentation indicates ACTH excess. Mineralocorticoid deficiency and related salt-wasting disorders cause sodium diuresis, which produces polyuria and enuresis, as well as hyponatremia, hyperkalemia, metabolic acidosis, aciduria, and dehydration. They can present as a salt-losing crisis or as failure to thrive. DIAGNOSIS Plasma cortisol levels are seldom diagnostic of adrenal failure unless a low level is accompanied by a clearly elevated plasma ACTH level. Confirmation of primary adrenal insufficiency requires an ACTH test, which is usually performed with synthetic ACTH-(1-24) (cosyntropin), 0.15 mg/m2, with sampling at 0 and 60 minutes. A peak plasma cortisol level 18 g/dL, or an aldosterone increment 5 ng/dL above the control level, is indicative of normal adrenocortical reserve.30,31 Measurements of the 24-hour urine content of electrolytes and pH are useful in the assessment of hyponatremia and hyperkalemia.28 A high urinary sodium level distinguishes aldosterone deficiency from prerenal sodium-losing states (e.g., cystic fibrosis, in which sodium may be lost through perspiration, or chronic sodium-losing diarrhea). Low urinary pH (<5.5) distinguishes aldosterone deficiency from renal tubular disorders that cause hyperkalemic acidosis. If aldosterone or PRA levels are indeterminate, measurements should be repeated either 3 hours after challenge with furosemide (30 mg/m2 every 6 hours 3) or after a low-salt diet (7 mEq/m2 per day) for 5 days. SECONDARY FORM ETIOLOGY AND CLINICAL CHARACTERISTICS The most frequent cause of isolated glucocorticoid deficiency is withdrawal from glucocorticoid excess (Cushing syndrome), the cause of which may be iatrogenic or spontaneous. The degree of hypothalamic-pituitary-adrenal (HPA) axis suppression is dependent on the dose, duration, and frequency of corticosteroid therapy. Recovery from high doses administered for long periods may be delayed for as long as 9 months after the discontinuation of therapy.32 The data suggest successive recovery of the pituitary, adrenal, and hypothalamic portions of the axis, in that order. The most common organic cause of hypothalamic corticotropin-releasing hormone (CRH) or ACTH insufficiency is intracranial tumor or its treatment, or both. The sluggish HPA axis of extremely low birth weight newborns may predispose to transient glucocorticoid deficiency.33 Neonatal ACTH deficit also occurs as part of the congenital hypopituitarism syndrome of hypoglycemia and cholestatic jaundice, with micropenis in boys.34,35 Autoimmune hypophysitis involving corticotropes may occur as part of the polyglandular deficiency syndrome.20 Isolated ACTH deficiency may be congenital, or due to a proopiomelanocortin (POMC) mutation.36,37 The POMC mutations yield a syndrome, resulting from the lack of POMC-derived ligands for the melanocortin receptors: ACTH deficiency (ACTH receptor, also termed melanocortin 2 receptor); red hair pigmentation due to a-melanocytestimulating hormone deficiency (melanocortin 1 receptor); and severe early-onset obesity due to a-melanocytestimulating hormone deficiency (melanocortin 4 receptor). However, most cases of isolated ACTH deficiency probably have an autoimmune basis.36 The clinical manifestations of ACTH deficiency are the same as those of isolated glucocorticoid deficiency, except for the absence of hyperpigmentation. Dilutional hyponatremia may occur due to an altered osmotic threshold for vasopressin release. Adrenal medullary insufficiency may complicate management of hypoglycemia.38 DIAGNOSIS Plasma DHEAS concentrations are useful in screening post-adrenarchal patients for adrenal failure.39 The very low dose 1 g/1.73 m2 ACTH stimulation test should replace the standard 150 g/m2 ACTH stimulation test in screening the HPA axis for secondary hypocortisolism when the DHEAS level is indeterminate.40,41 A cortisol level <18 mg/dL (500 nmol/L) after 30 minutes signifies impaired adrenocortical reserve. The technique of delivering the low dose of ACTH is critical. A definitive test should be performed if this test results in a borderline value and the diagnosis is questioned.41 The cortisol response to insulin-induced hypoglycemia has been the best single standard test for assessing the function of the entire HPA axis.30 Insulin (0.050.15 U/kg) is administered by intravenous (iv) bolus under very close supervision to allow the study to be aborted instantaneously with iv administration of glucose should an emergency arise. Complications of hypoglycemia are unlikely in the absence of a history of syncope or seizures. A plasma cortisol level 18 g/dL in response to a 40% reduction in blood glucose is considered to be indicative of normal HPA reserve. Insulin-induced hypoglycemia normally acts on the hypothalamus to trigger sequential ACTH and cortisol secretion through stimulation of CRH, vasopressin, and epinephrine.42 It is not established, however, whether the outcome of this test corresponds with the capacity of the HPA axis to respond to stress in children with hypothalamic disease. The metyrapone test specifically checks the integrity of the negative feedback system; therefore, the results do not necessarily correspond to those for insulin tolerance testing. Metyrapone is obtained from Novartis Pharmaceutical by special order. The dose is 15 mg/kg (maximum 0.5 g) orally every 4 hours for six doses, commencing at 8:00 a.m. Normally, plasma 11-deoxy-cortisol levels increase to 9.0 g/dL or more 4 hours after the sixth dose. The place for CRH testing in the diagnosis of secondary adrenal insufficiency remains to be established.43 It is clear, however, that subnormal ACTH and cortisol responses indicate decreased ACTH secretion. The adequacy of replacement doses of corticosteroids is hard to judge in patients on drugs that accelerate their metabolism, such as phenobarbital or phenytoin. Measurement of plasma cortisol levels 1 to 3 hours after the oral administration of 13 mg/m2 may be helpful.44 TREATMENT In managing acute adrenal crisis, it is important to remember that affected children have incurred large losses of salt and water. Indeed, a newborn infant may excrete more than half a liter of isotonic urine daily, and will have accumulated a substantial salt deficit before the hyponatremic crisis occurs. Water, salt, and corticosteroids should be replaced intravenously. Intake and output should be strictly monitored, with special attention given to 24-hour sodium excretion. Fluid therapy should be

initiated with 0.45% saline and 5% dextrose administered at twice the average maintenance rate (i.e., 200 mL/kg per day). Once serum electrolytes have been measured, an increase to 0.675% saline and 5% dextrose may be indicated to correct the calculated sodium deficit. Normal saline (20 mL/kg, administered by rapid iv infusion) or colloid may be required initially for the treatment of shock. Corticosteroid treatment may not have a perceptible effect on the patient's clinical course for several hours. Glucocorticoid therapy should be initiated with cortisol, 50 mg/m2 iv immediately and 15 mg/m2 iv every 6 hours intramuscularly (im). When the crisis is past (within ~2448 hours), oral feedings can begin, and iv hydrocortisone can be discontinued. Fludrocortisone (Florinef), 0.1 mg per day, should be started as soon as oral administration is possible. Feeding initially should include 1 g sodium chloride daily. This is administered as 15 mL saline with the first ounce of each of an infant's six feedings. After a few days, the sodium dosage should be individualized, keeping in mind that eyelid, presacral, or labioscrotal edema will be the first sign of overtreatment in an infant. With the typical uncomplicated clinical course, infants usually are thriving and have normal serum electrolyte levels by the sixth day of hospitalization. Maintenance glucocorticoid therapy should be initiated with 15 (1020) mg/m2 per day oral hydrocortisone. This dose is twofold greater than the blood production rate to account for splanchnic metabolism. For infants, the dose of hydrocortisone suspension (Cortef) should be measured accurately in a syringe after vigorous shaking. Three divided doses usually are required to maintain low plasma ACTH concentrations. This is especially important in patients with congenital adrenal hyperplasia, to minimize hyperandrogenemia. In long-term therapy, the critical factor in avoiding glucocorticoid overdose is the careful monitoring of linear growth, since a subnormal growth rate is the earliest indicator of glucocorticoid excess. Patients with ACTH deficiency may require glucocorticoid replacement only during times of stressful illness; use of cortisol over 15 mg/m2 for asymptomatic patients may increase the growth hormone requirement in hypopituitary patients. Patients should be monitored every 4 to 12 weeks during the period of rapid growth (birth to 1 year) to determine the lowest maintenance glucocorticoid dose that will normalize symptoms and the plasma ACTH concentration or, in the case of congenital adrenal hyperplasia, the appropriate plasma steroids. In older children, regular follow-up at 3- to 6-month intervals is advisable. Occasionally, it is difficult to normalize some patients on standard therapeutic regimens.45 Such patients are best treated with a longer-acting glucocorticoid, such as prednisone twice a day, or dexamethasone at bedtime. Prednisone causes fewer striae than does dexamethasone, and it is possible to adjust the dose of tablets more precisely. Prednisone is ~3 times and dex-amethasone 80 times more potent in suppressing growth and ACTH than is cortisol.46 Because of associated adrenal medullary insufficiency, recovery from hypoglycemia may not be normalized by the usual replacement doses of cortisol. This is because of the lack of high local adrenomedullary concentrations of cortisol that seem to facilitate the N-methylation required for epinephrine synthesis.38 A dose of fludrocortisone acetate, 0.05 to 0.1 mg daily, is typically appropriate as mineralocorticoid replacement from infancy onward. The most sensitive index of the adequacy of salt and mineralocorticoid therapy is PRA (see Table 83-1). Underdose interferes with the control of androgens because angiotensin stimulates adrenal fasciculata-reticularis cells.47 Overdosage is indicated by edema, hypertension, or hypokalemia. Spironolactone, a competitive antagonist of aldosterone, can be used as an adjunct to the treatment of mineralocorticoid excess forms of congenital adrenal hyperplasia. Families should be given written instructions about emergency treatment, and the child should wear a bracelet (e.g., Medic Alert) that specifies the diagnosis in case of emergency. CORTICOSTEROID COVERAGE FOR STRESS The goal of therapy during physiologic stress is to provide exogenous cortisol in a sufficient amount to mimic the known normal response. Excessive doses of corticosteroid are undesirable, since the drug may interfere with electrolyte homeostasis, retard healing, and increase the patient's susceptibility to infection. If surgery is required, the following treatment regimen is suggested. Cortisol, 15 mg/m2, is administered orally on the evening before surgery, and cortisol, 15 mg/m2, is administered iv immediately preoperatively, and parenterally every 6 hours thereafter until recovery from the acute effects of surgery has occurred. During the course of any acute, stressful illness, such as diarrhea or an upper respiratory infection, the cortisol dose should be increased to two or three times maintenance. Such treatment is seldom required for more than 3 days. If medication cannot be taken orally or is vomited repeatedly, cortisol, 100 mg/m2, should be administered im daily to prevent adrenocortical crisis. Varicella is not dangerous to patients who are receiving replacement hydrocortisone therapy provided the dose of glucocorticoid does not exceed two-fold maintenance.48

HYPERCORTISOLISM
CLINICAL CHARACTERISTICS Cushing syndrome is the clinical manifestation of excess glucocorticoid. Along with the well-known manifestations in adults, affected children have attenuated linear growth.49 Indeed, growth arrest may be the only clear clinical sign of Cushing syndrome. Interestingly, often these children tend to be highly responsible and obsessive about schoolwork, housework, and hobbies. Multiple factors50,51 and 52 contribute to linear growth arrest. These range from direct inhibition of cell proliferation to inhibition of growth hormone release. The abnormal glucose tolerance associated with hypercortisolism is attributable to increased gluconeogenesis. Ecchymosis secondary to increased capillary fragility, hypertrichosis resembling that seen in states of malnutrition, and muscle weakness and fatigue all appear to be related to increased protein catabolism. The hypertension is caused by increased vascular reactivity to vasoconstrictive factors,53 increased plasma renin substrate,53 cortisol overload,54 and increased mineralocorticoid production.55 Hyperpigmentation is seen in those patients with elevated plasma ACTH levels. Concomitant androgen excess is responsible for hirsutism or virilization, the latter of which suggests a tumor. The most common cause of Cushing syndrome in children is pharmacologic glucocorticoid therapy. Attenuation of linear growth occurs at all doses that exceed the normal replacement range of 10 to 20 mg/m 2 per day cortisol, or the equivalent.52 The syndrome can occur after use of topical corticoids, depending on the steroid and the dose. Some individuals are hyperre-active to glucocorticoids, and become cushingoid on normal replacement doses.56 High doses of some synthetic progestins can cause cushingoid state.56a Cushing disease is hyperadrenocortisolism secondary to excessive secretion of pituitary ACTH, which results in bilateral adrenal hyperplasia (see Chap. 75). In children, as in adults, it is usually caused by a pituitary microadenoma (Fig. 83-3). It may have an autoimmune basis. 57 Hyperfunctioning adrenocortical tumors occur three times more frequently in girls than in boys. Cushing syndrome secondary to ACTH production by nonen-docrine tumors is very rare in the pediatric population. ACTH-independent, primary bilateral adrenal hyperplasia may be caused by McCune-Albright syndrome variants.58 Adrenocortical micronodular dysplasia causing Cushing syndrome is primarily a disorder of young girls. It may be familial, associated with multiple endocrine neoplasia-type I,59 or myxoid and other tumors,60,61 or it can occur as an autoimmune disease.49 In some cases, hormonogenesis is periodic, and in others it is estrogen-61 or food-dependent.62

FIGURE 83-3. Nine-year-old girl with Cushing disease. Later, the condition was treated successfully by transsphenoidal adenectomy. (Courtesy of Dr. Wellington Hung.)

DIAGNOSIS The diagnostic criteria for Cushing syndrome in children are similar to those in adults. Generally, after determining that cortisol production is excessive, the next step in the workup is the two-step dexamethasone suppression test using successive 2-day courses of dexamethasone, 1 mg/m2 and 2 mg/m2 per day in four divided doses.49 Plasma cortisol levels should decrease to <2 g/dL, 24-hour urinary 17-hydroxycorticoids should fall to <2 mg/g creatinine, and urinary free cortisol should fall at least 90%. CRH testing shows promise in yielding the same information in less time. Imaging studies should be performed as in adults. The diagnosis in some cases may require petrosal sinus sampling during the CRH test. Periodic hormonogenesis may present a diagnostic dilemma.63 The differential diagnosis of high cortisol levels includes familial cortisol resistance, which may be due to a defect in the receptor for glucocorticoids64,65 or a receptor coactivator.65a TREATMENT The potential for further growth and pubertal development is important to consider in reaching a decision about optimal therapy for children with Cushing disease. A trial of cyproheptadine hydrochloride may bring about remission.66 However, in children and adolescents, transsphenoidal microadenomectomy, performed by an experienced surgeon, usually appears to be the treatment of choice.67 Transsphenoidal surgery often corrects hypercortisolism, even if no adenoma is detected; cases of Cushing disease without definite pituitary adenoma have been reported in which removal of a portion of the central mucous zone of the pituitary gland was associated with remission. Following surgery, subsequent normal growth is expected, and few sequelae are reported. However, a few patients may require repeat transsphenoidal surgery. The presence of autoantibodies to corticotropes may be a marker indicating a poor prognosis.57 Pituitary irradiation of patients with Cushing disease, using ~4000 rad, has resulted in a 75% remission rate; however, this mode of treatment is also associated with a 50% incidence of growth hormone deficiency.68 Bilateral adrenalectomy not only causes permanent hypoadrenalism, but also causes the development of Nelson syndrome in 27% of patients.69 In the postoperative period, patients must be treated to prevent secondary adrenal insufficiency. After cure of Cushing syndrome, catch-up growth often occurs.70 Catch-up growth is poor in some cases with a long duration and high intensity of glucocorticoid exposure. The cause is unclear; perhaps, puberty causes chondrocytes to differentiate before recovery from inhibition of stem cell proliferation is completed. Catch-up growth is most likely to occur when glucocorticoid excess is relieved well before puberty. WITHDRAWAL FROM CORTICOSTEROID EXCESS IN CHILDREN Children in whom long-term pharmacologic glucocorticoid therapy is being withdrawn or in whom the HPA axis has been suppressed by Cushing syndrome require extended and gradually tapering therapeutic regimens. The dosage of corticosteroid can usually be lowered to a maintenance level (1020 mg/m2 per day of cortisol equivalents fairly rapidly; i.e., halving the dose at 1- to 2-week intervals over a few weeks). This will permit linear growth and prevent adrenal insufficiency. During this time, recrudescence of the underlying disease being treated, corticosteroid dependence,71 or pseudotumor cerebri72 may occur, necessitating a return to a higher corticosteroid dose followed by a slower taper. The use of alternate-morning administration of prednisone in such situations may minimize corticosteroid side effects and permit growth.73,74 After the patient is stabilized on maintenance glucocorticoids, it is appropriate to wean the patient from the corticosteroid over as long a period as the patient has been receiving therapy, up to 9 months.32 The symptoms of too-rapid weaning are often vague, and are characteristic of isolated glucocorticoid deficiency. Of course, intercurrent, acute, stressful illness requires supplementation for a few days. The return of normal function of the HPA axis is indicated by an 8:00 a.m. plasma cortisol level of >10 g/dL and a level >18 g/dL in response to a very low dose ACTH challenge.40,41

PRECOCIOUS PSEUDOPUBERTY
ETIOLOGY AND CLINICAL CHARACTERISTICS Adrenarche is an adrenal puberty during which the adrenal cortex acquires the ability to secrete 17-ketosteroids (see Table 83-2) and DHEA responsiveness to ACTH increases75,76 and 77 (Table 83-3). It normally commences slowly in midchildhood, eventually accounting for most DHEAS production, and reaches a magnitude sufficient to bring about the growth of pubic hair at an average age of ~12 years.

TABLE 83-3. ACTH-Stimulated Corticosteroids: Typical Normal Plasma Values in Children and Adults*

Adrenarche is independent of true (gonadotropin-mediated) puberty. It is related to the development of the zona reticularis. DHEA synthesis increases because this zone has low 3b-hydroxysteroid dehydrogenase activity and seemingly high 17,20-lyase activity; DHEAS is formed from DHEA because of the high sulfokinase activity of this zone. The existence of a putative pituitary adrenocortical androgen-stimulating factor, which is responsible for adrenarche, has been argued. However, ACTH is necessary and sufficient to stimulate adrenal androgens and is the only hormone established to do so. The authors' concept is that an adrenarche factor need be responsible only for bringing about the development of the zona reticularis or a shift in its pattern of response to ACTH. Premature pubarche (sexual pubic hair as an isolated phenomenon before 8.5 years of age in girls or 9.5 years in boys) usually is a benign condition. It has been termed premature adrenarche when the plasma DHEAS and testosterone levels are elevated for age, but appropriate for the pubic hair stage. Premature adrenarche is more common in girls than boys. Neither a growth spurt nor clitoromegaly occurs. There are no signs of gonadal maturation, and the bone age is not advanced to an abnormal extent. It has been thought to be a normal phenomenon happening early. However, the androgen level, particularly that of androstenedione, in childhood correlates with the emergence of a polycystic ovary syndrome (PCOS)like form of functional ovarian hyperandrogenism after menarche.78 This suggests that some cases are due to dysregulation of androgen production, which seems to underlie the primary functional adrenal hyperandrogenism (formerly termed exaggerated adrenarche) that accounts for most cases of postmenarcheal adrenal androgen excess, as well as the frequently associated ovarian androgen excess.77 Although such patients were originally thought to have a mild 3b-hydroxysteroid dehydrogenase deficiency, this is now known to be unlikely unless 17-hydroxypregnenolone responses to ACTH are 7 or more SD above average.79 Frankly virilizing adrenal disease is usually caused by congenital adrenal hyperplasia (see Chap. 77). Nonclassic congenital adrenal hyperplasia may closely resemble premature adrenarche; affected girls do not have the genital ambiguity of classic congenital adrenal hyperplasia. Rare causes of functional adrenal hyperandrogenism and premature pubarche are congenitally excessive cortisol metabolism80 and resistance to cortisol.65 Adrenal tumors may be responsible for virilism. In such cases, virilization is more often attributable to adrenal carcinoma than to adrenal adenoma. In patients with adrenal carcinoma, there is a high incidence of associated urinary tract anomalies, hemihy-pertrophy, neurofibromatosis, and familial tumors.81,82 Tumor size and histologic grade are the major prognostic factors.83 Feminizing adrenal tumors are rare in children, but both adenomas and carcinomas have been reported.84 These tumors cause abnormal advancement of bone age and sexual precocity in girls, and gynecomastia with advanced bone age in boys. Simultaneous feminization and androgenization occur in those tumors that produce

androgens and estrogens. Urinary free cortisol and plasma DHEAS and estrogen levels typically are markedly elevated. DIAGNOSIS Pseudopuberty is distinguished clinically from true puberty by the absence of gonadal development. This is indicated by the absence of breast enlargement in affected girls and by the absence of testicular enlargement in affected boys. Virilization is indicated by clitoromegaly in girls, an abnormally advanced bone age, and a plasma testosterone concentration greater than the adult female range. Virilizing disorders can usually be distinguished from premature adrenarche (see Table 83-2 and Table 83-3) and from one another by their characteristic plasma steroid patterns. Classic congenital adrenal hyperplasia (see Chap. 77) is characterized by very high levels of 17-hydroxyprogesterone (typically >2000 ng/dL), with only modest elevation of DHEAS. Nonclassic congenital adrenal hyperplasia may have plasma androgen levels in