Download the full version of the ebook at

https://ebookultra.com

Textbook of Neuro Oncology 1st Edition

Mitchel S. Berger

https://ebookultra.com/download/textbook-of-neuro-
oncology-1st-edition-mitchel-s-berger/

Explore and download more ebook at https://ebookultra.com

 Recommended digital products (PDF, EPUB, MOBI) that
you can download immediately if you are interested.

Neuro oncology First Edition Robert A. Gross(Eds.)

https://ebookultra.com/download/neuro-oncology-first-edition-robert-a-
grosseds/

ebookultra.com

Textbook of Surgical Oncology 1st Edition J. Poston Graeme

https://ebookultra.com/download/textbook-of-surgical-oncology-1st-
edition-j-poston-graeme/

ebookultra.com

Neuro Oncology Part II 1st Edition Wolfgang Grisold And

Riccardo Soffietti (Eds.)

https://ebookultra.com/download/neuro-oncology-part-ii-1st-edition-
wolfgang-grisold-and-riccardo-soffietti-eds/

ebookultra.com

The Cultural Theorist s Book Of Quotations 1st Edition

Arthur Asa Berger

https://ebookultra.com/download/the-cultural-theorist-s-book-of-
quotations-1st-edition-arthur-asa-berger/

ebookultra.com
Manual of Neuro Ophthalmogy First Edition Agarwal

https://ebookultra.com/download/manual-of-neuro-ophthalmogy-first-
edition-agarwal/

ebookultra.com

Kuerer s Breast Surgical Oncology 1st Edition Ph.D. Kuerer

https://ebookultra.com/download/kuerer-s-breast-surgical-oncology-1st-
edition-ph-d-kuerer/

ebookultra.com

Neuro Ophthalmology 1st Edition Desmond P. Kidd

https://ebookultra.com/download/neuro-ophthalmology-1st-edition-
desmond-p-kidd/

ebookultra.com

Rook s Textbook of Dermatology Eighth Edition Tony Burns

https://ebookultra.com/download/rook-s-textbook-of-dermatology-eighth-
edition-tony-burns/

ebookultra.com

Textbook of Vitreoretinal Diseases and Surgery 1st Edition

Edition S Natarajan

https://ebookultra.com/download/textbook-of-vitreoretinal-diseases-
and-surgery-1st-edition-edition-s-natarajan/

ebookultra.com
Textbook of Neuro Oncology 1st Edition Mitchel S.
Berger Digital Instant Download
Author(s): Mitchel S. Berger, Michael Prados
ISBN(s): 0721681484
Edition: 1
File Details: PDF, 93.24 MB
Year: 2004
Language: english
The Curtis Center
170 S Independence Mall W 300E
Philadelphia, Pennsylvania 19106-3399

TEXTBOOK OF NEURO-ONCOLOGY ISBN: 0-7216-8148-4

Copyright © 2005 by Elsevier Inc. All rights reserved.

No part of this publication may be reproduced or transmitted in any form or by any means, electronic or
mechanical, including photocopy, recording, or any information storage and retrieval system, without
permission in writing from the publisher. Permissions may be sought directly from Elsevier Inc. Rights
Department in Philadelphia, Pennsylvania, USA: telephone: (+1)215-238-7869, fax: (+1)215-238-2239,
e-mail: [email protected]. You may also complete your request on-line via the Elsevier
homepage http://www.elsevier.com, by selecting “Customer Support” and then “Obtaining Permissions.”

NOTICE

Neurosurgery is an ever-changing field. Standard safety precautions must be followed, but as new research
and clinical experience broaden our knowledge, changes in treatment and drug therapy may become
necessary or appropriate. Readers are advised to check the most current product information provided by
the manufacturer of each drug to be administered to verify the recommended dose, the method and
duration of administration, and contraindications. It is the responsibility of the treating physician, relying
on experience and knowledge of the patient, to determine dosages and the best treatment for each
individual patient. Neither the Publisher nor the authors assume any liability for any injury and/or
damage to persons or property arising from this publication.
The Publisher

Library of Congress Cataloging-in-Publication Data

Textbook of neuro-oncology / [edited by] Mitchel S. Berger, Michael D. Prados. – 1st ed.
p. ; cm.
Includes index.
ISBN 0-7216-8148-4
1. Nervous system–Tumors. I. Berger, Mitchel S. II. Prados, Michael.
[DNLM: 1. Brain Neoplasms. 2. Neoplasms, Neuroepithelial. 3. Peripheral Nervous System
Neoplasms. 4. Skull Base Neoplasms. 5. Spinal Cord Neoplasms. WL 358 T355 2005]
RC280.N4T49 2005
616.99¢481–dc22 2004042996

Acquisitions Editor: Rebecca Schmidt Gaertner

Developmental Editor: Denise LeMelledo
Publishing Services Manager: Joan Sinclair
Project Manager: Cecelia Bayruns

Printed in China.

Last digit is the print number: 9 8 7 6 5 4 3 2 1

P R E FA C E

It takes an integrated multidisciplinary approach to provide the To this end, we enlisted the aid of some of the most
best care for patients with brain tumors, and in consulting with respected people working in the field. Ossama Al-Mefty, at the
our colleagues on the patient-care team we’ve often been frus- University of Arkansas, signed up to deal with tumors of the
trated by the lack of an all-inclusive resource that provides a cranial base. Ray Sawaya, at the University of Texas M.D.
comprehensive picture of the multifaceted aspects of the chal- Anderson Cancer Center, said he would take on the section on
lenges involved. In working with our colleagues doing basic brain metastasis. Paul McCormick of the Columbia Presbyter-
science research into the causes and mechanisms of brain ian Medical Center and the Neurological Institute of New York
tumors, we have been impressed by their interest in the clini- agreed to help out with a section on tumors of the spinal axis
cal details of translational therapies developed through their together with Andy Parsa, who has since joined us in San
creative work. With the enthusiasm and support of the pub- Francisco. Michel Kliot, at the University of Washington
lisher, we were encouraged to outline a book intended for School of Medicine in Seattle, agreed to edit the section on
everyone curious to understand more about the varied clinical peripheral nerve tumors together with Keith Kwok. Jim Rutka,
roles in the treatment of brain tumors and the basic biology and at the University of Toronto’s Hospital for Sick Children,
epidemiology of this most perplexing disease. agreed to edit the section on pediatric neuro-oncology, and our
This book is intended for everyone on the medical team colleague Evren Keles joined forces in editing the section on
caring for patients with brain tumors, from the physicians who the treatment of intracranial tumors in adults. With this world-
first suspect the diagnosis, to the neuroradiologists involved class team working with us, we felt we could produce a com-
in diagnostic work-ups, to the medical specialists in neuro- prehensive text that would be useful for everyone immersed in
oncology who monitor care, to radiation oncologists and or interested by the field of neuro-oncology. These section
neurosurgeons. It is a “Tumor Board in Print” where each spe- editors have done their part, enlisting a distinguished panel of
cific type of brain tumor occupies a chapter in which all of the specialists who have provided chapters that more than fulfill
multidisciplinary specialists involved in treating that specific our optimistic hopes for this textbook. Together with our fellow
tumor are heard. We have also tried to make this a book for editors, we hope you will find this volume both an interesting
other members of the patient-care team, for residents and and useful resource.
students interested in brain tumors, and in particular for basic
Mitchel S. Berger, MD
scientists wanting to understand more about the clinical com-
Michael D. Prados, MD
plexities of managing neuro-oncology patients.
Editors
CONTRIBUTORS
Joann Aaron, MA Edgardo J. Angtuaco, MD Michael E. Berens, PhD
Scientific Editor Professor, Department of Radiology Senior Investigator
Department of Neuro-Oncology University of Arkansas for Medical Sciences; TGen, The Translational Genomics Research
University of Texas M.D. Anderson Cancer Center Chief, Division of Neuroradiology Institute
Houston, Texas University of Arkansas for Medical Sciences Phoenix, Arizona
24: Mixed Gliomas 45: Meningiomas 3: Animal Models
48: Chordomas and Chondrosarcomas of the
Emad T. Aboud, MD Cranial Base Mitchel S. Berger, MD
Fellow, Department of Neurosurgery Professor and Chair of Neurological Surgery
University of Arkansas for Medical Sciences Henry E. Aryan, MD Kathleen M. Plant Distinguished Professor of
Little Rock, Arkansas Clinical Instructor of Neurosurgery Neurological Surgery
52: Epidermoid Tumors University of California, San Diego, Director, Brain Tumor Research Center
School of Medicine; University of California, San Francisco,
A. Leland Albright, MD Neurosurgeon School of Medicine
Professor of Neurosurgery University of California, San Diego, San Francisco, California
University of Pittsburgh School of Medicine; Medical Center; 1: Epidemiology of Brain Tumors
Chief of Pediatric Neurosurgery Children’s Hospital Medical Center 9: Surgical Strategies in the Management of Brain
Children’s Hospital of Pittsburgh San Diego, California Tumors
Pittsburgh, Pennsylvania 101: Pediatric Skull Base Tumors 17: Diffuse Astrocytoma
87: Hypothalamic Hamartomas 28: Gliomatosis Cerebri
Kurtis Ian Auguste, MD 40: Hemangioblastomas of the Central Nervous
Resident Physician System
Kenneth D. Aldape, MD
Associate Professor of Pathology Department of Neurological Surgery
University of Texas M.D. Anderson Cancer University of California San Francisco School of Mark Bernstein, MD, FRCSC
Center Medicine Professor of Surgery
Houston, Texas San Francisco, California The University of Toronto;
24: Mixed Gliomas 40: Hemangioblastomas Neurosurgeon
27: Astroblastoma Toronto Western Hospital
Samer Ayoubi, MD, FRCS University Health Network
Attending Physician Toronto, Ontario, Canada
Mubarak Al-Gahtany, MD, FRCSC 22: Subependymal Giant Cell Astrocytoma and
Clinical Fellow in Neurosurgery Damascus University School of Medicine;
Consultant Neurosurgeon Tuberous Sclerosis Complex
The University of Toronto
Toronto Western Hospital Al-Muwasat University Hospital
Toronto, Ontario, Canada Damascus, Syria Ratan D. Bhardwaj, MD
74: Malignant Peripheral Nerve Tumors 44: Nonacoustic Schwannomas of the Cranial Neurosurgical Resident
107: Peripheral Nerve Tumors in Children Nerves The University of Toronto
46: Meningeal Hemangiopericytomas and Toronto, Ontario, Canada
Sarcomas 94: Pituitary Tumors in Children
Ossama Al-Mefty, MD
Professor and Chair of Neurosurgery
University of Arkansas for Medical Sciences Matthew T. Ballo, MD Asis Kumar Bhattacharyya, MBBS, MS, MCh
Little Rock, Arkansas Associate Professor of Radiation Oncology Fellow, Division of Neurosurgery
44: Nonacoustic Schwannomas of the Cranial University of Texas M.D. Anderson Cancer Center University of Toronto
Nerves Houston, Texas Toronto Western Hospital
45: Meningiomas 61: Skull Base Metastasis Toronto, Ontario, Canada
46: Meningeal Hemangiopericytomas and 22: Subependymal Giant Cell Astrocytoma and
Sarcomas Geoffrey R. Barger, MD Tuberous Sclerosis Complex
48: Chordomas and Chondrosarcomas of the Associate Professor
Cranial Base Department of Neurology Juan M. Bilbao, MD
49: Paragangliomas of the Skull Base Wayne State University School of Medicine; Associate Professor of Pathology
Co-Chief, Neuro-Oncology The University of Toronto;
Nabeel Al-Shafai, MD Karmanos Cancer Institute Neuropathologist
Neurosurgery Resident Detroit, Michigan Sunnybrook Hospital
The University of Toronto 35: Medulloblastoma Toronto, Ontario, Canada
Toronto Western Hospital 80: Lhermitte-Duclos Disease
Toronto, Ontario, Canada Fred G. Barker II, MD
97: Choroid Plexus Tumors Assistant Professor of Neurosurgery Mark H. Bilsky, MD
Harvard Medical School; Assistant Attending, Department of Surgery
Christopher Ames, MD Assistant Visiting Neurosurgeon Memorial Sloan-Kettering Cancer Center
Assistant Professor of Neurological Surgery Massachusetts General Hospital Cornell University Medical Center
University of California, San Francisco, Boston, Massachusetts New York, New York
School of Medicine 20: Pilocytic Astrocytoma 71: Therapeutic Options for Treating Metastatic
San Francisco, California 34: Pineoblastoma Spine Tumors
69: Benign Tumors of the Spine
Gene H. Barnett, MD, FACS Devin K. Binder, MD, PhD
Lilyana Angelov, MD Professor of Surgery Chief Resident
Associate Staff Cleveland Clinic Lerner College of Medicine; Department of Neurological Surgery
Department of Neurological Surgery Chairman University of California, San Franciso,
The Cleveland Clinic The Brain Institute School of Medicine
Cleveland, Ohio Cleveland, Ohio San Francisco, California
30: Mixed Neuronal and Glial Tumors 30: Mixed Neuronal and Glial Tumors 14: Gene Therapy for Malignant Gliomas
 Contributors xiii

Deborah T. Blumenthal, MD Jeffrey N. Bruce, MD Marc C. Chamberlain, MD

Assistant Professor of Neurology and Oncology Associate Professor of Neurosurgery Professor of Neurology and Neurosurgery
University of Utah The Neurological Institute of New York University of Southern California Keck School of
Huntsman Cancer Institute Columbia-Presbyterian Medical Center Medicine;
Salt Lake City, Utah Columbia University College of Physicians and Co-Director of Neuro-Oncology
54: Small Cell Lung Carcinoma Surgeons Norris Cancer Center
New York, New York Los Angeles, California
Robert J. Bohinski, MD, PhD 33: Pineocytoma 36: Adult Supratentorial Primitive
Assistant Professor, Department of Neurosurgery Neuroectodermal Tumors
University of Cincinnati;
Mayfield Clinic and Spine Institute Derek A. Bruce, MD, ChB
Professor of Pediatric Neurosurgery Susan M. Chang, MD
Cincinnati, Ohio
George Washington University Associate Professor of Neurological Surgery
60: Multidisciplinary Management of Patients with
Children’s National Medical Center University of California, San Francisco
Brain Metastases from an Unknown Primary
Washington, D.C. San Francisco, California
Tumor
101: Pediatric Skull Base Tumors 16: Clinical Trials
Melissa L. Bondy, PhD 28: Gliomatosis Cerebri
Professor of Epidemiology 34: Pineoblastoma
University of Texas M.D. Anderson Cancer John M. Buatti, MD
Center Professor and Head
Michael Y. Chen, MD
Houston, Texas Department of Radiation Oncology
Resident
1: Epidemiology of Brain Tumors University of Iowa Hospitals and Clinics
Virginia Commonwealth University;
Iowa City, Iowa
Medical College of Virginia
Margaret Booth-Jones, PhD 58: Intracerebral Metastatic Colon Carcinoma
Richmond, Virginia
Assistant Professor of Oncology and Psychiatry 38: Fibrous Tumors
H. Lee Moffitt Cancer and Research Institute Ketan R. Bulsara, MD
University of South Florida Instructor of Neurosurgery
Tampa, Florida Thomas Chen, MD, PhD
University of Arkansas for Medical Sciences; Associate Professor, Neurosurgery and Pathology;
18: Anaplastic Astrocytoma Children’s Hospital Director of Surgical Neuro-Oncology
Little Rock, Arkansas University of Southern California;
Donald E. Born, MD, PhD 45: Meningiomas
Clinical Assistant Professor, Department of Staff Neurosurgeon
Pathology University of Southern California University
University of Washington School of Medicine Hospital;
Eric C. Burton, MD
Seattle, Washington Norris Cancer Hospital
Assistant Professor of Neurological Surgery
73: Evaluation and Management of Benign Los Angeles, California
University of California, San Francisco;
Peripheral Nerve Tumors and Masses 36: Adult Supratentorial Primitive
Neuro-Oncologist
Neuroectodermal Tumors
University of California, San Francisco, Medical
Kristin Bradley, MD Center
Assistant Professor of Human Oncology San Francisco, California Terri Chew, MPH
University of Wisconsin School of Medicine 17: Diffuse Astrocytoma Academic Coordinator, Neurosurgery Research
Madison, Wisconsin University of California, San Francisco
10: Basic Concepts Underlying Radiation San Francisco, California
Therapy Thomas Carlisle, MD, PhD 1: Epidemiology of Brain Tumors
Associate Professor, Clinical Internal Medicine
Steven Brem, MD University of Iowa
Professor of Oncology, Neurosurgery, and Roy J. and Lucille Carver College of Medicine John H. Chi, MD, MPH
Neurology; Iowa City, Iowa Resident, Department of Neurological Surgery
Director of Neurosurgery; 58: Intracerebral Metastatic Colon Carcinoma University of California, San Francisco
Program Leader, Neuro-Oncology Program San Francisco, California
H. Lee Moffitt Cancer Center and Research 72: Intrinsic Metastatic Spinal Cord Tumors
Institute James H. Carter, Jr., MHS, PAC 81: Cerebellar Astrocytomas
Tampa, Florida Assistant Clinical Professor of Surgery
18: Anaplastic Astrocytoma Division of Neurosurgery
Dean Chou, MD
Duke University Medical Center
Assistant Clinical Professor of Neurological
William C. Broaddus, MD, PhD Durham, North Carolina
Surgery
Associate Professor of Neurosurgery, Anatomy and 56: Brain Metastases from Malignant Melanoma
University of California, San Francisco
Neurobiology, Physiology, and Radiation San Francisco, California
Oncology Steve Casha, MD, PhD 70: Malignant Primary Tumors of the Vertebral
Virginia Commonwealth University; Neurosurgery Resident Column
Director of Neuro-oncology The University of Toronto
Medical College of Virginia Hospitals; Toronto Western Hospital
Chief of Neurosurgery H. Brent Clark, MD, PhD
Toronto, Ontario, Canada Professor of Neurology and Laboratory Medicine
Hunter Holmes McGuire VA Medical Center 92: Pediatric Meningiomas
Richmond, Virginia and Pathology
38: Fibrous Tumors University of Minnesota Medical School;
Soonmee Cha, MD Director of Neuropathology
Douglas L. Brockmeyer, MD Assistant Professor of Radiology and Neurological Fairview University Medical Center
Associate Professor of Neurosurgery and Surgery Minneapolis, Minnesota
Pediatrics University of California, San Francisco, 32: Dysembryoplastic Neuroepithelial Tumor
University of Utah School of Medicine; School of Medicine;
Attending Neurosurgeon Attending Neuroradiologist Aaron A. Cohen-Gadol, MD
Primary Children’s Medical Center University of California, San Francisco, Medical Neurosurgery Resident
Salt Lake City, Utah Center Mayo Clinic
104: Spinal Column Tumors in Pediatric San Francisco, California Rochester, Minnesota
Patients 4: Diagnostic Imaging 64: Spinal Meningiomas
xiv Contributors

Benedicto O. Colli, MD Egon M.R. Doppenberg, MD Allan H. Friedman, MD

Division of Neurosurgery Chief Resident, Department of Neurosurgery Guy L. Odom Professor of Neurological Surgery;
Department of Surgery Medical College of Virginia Neurosurgeon-in-Chief
Ribeirão Preto Medical School Virginia Commonwealth University Health Duke University Medical Center
University of São Paulo Center Durham, North Carolina
Ribeirão Preto, São Paulo, Brazil Richmond, Virginia 25: Ependymoma
48: Chordomas and Chondrosarcomas of the 38: Fibrous Tumors 56: Brain Metastases from Malignant Melanoma
Cranial Base
James G. Douglas, MD, MS Henry S. Friedman, MD
Ceasare Colosimo, MD Associate Professor James B. Powell Jr. Professor of Neuro-Oncology;
Professor of Radiology Departments of Radiation Oncology, Pediatrics, The Brain Tumor Center at Duke
Institute of Radiology and Neurological Surgery Duke University Medical Center
University of Chieti University of Washington Cancer Center; Durham, North Carolina
Chieti, Italy Chief, Division of Radiation Oncology 25: Ependymoma
85: Desmoplastic Infantile Ganglioglioma Children’s Hospital and Regional Medical
Center; Adam S. Garden, MD
Joseph F. Costello, PhD Co-director, University of Washington Gamma Associate Medical Director
Assistant Professor of Neurological Surgery Knife Center Head and Neck Center;
University of California, San Francisco Harborview Medical Center Associate Professor of Radiation Oncology;
San Francisco, California Seattle, Washington Associate Radiation Oncologist
2: Molecular and Cell Biology 35: Medulloblastoma Department of Radiation Oncology
University of Texas M.D. Anderson Cancer Center
Deborah L. Cummins, MD, PhD Houston, Texas
Christopher F. Dowd, MD 50: Carcinoma of the Paranasal Sinuses and
Associate Professor of Clinical Pathology Clinical Professor of Radiology, Neurological
University of Southern California Olfactory Neuroblastoma
Surgery, Neurology, and Anesthesia and
Keck School of Medicine Perioperative Care
Los Angeles, California University of California, San Francisco, Gary L. Gallia, MD, PhD
36: Adult Supratentorial Primitive School of Medicine Neurosurgery Resident
Neuroectodermal Tumors San Francisco, California Johns Hopkins University School of Medicine
12: Neurointerventional Techniques Baltimore, Maryland
Lisa M. DeAngelis, MD 29: Chordoid Glioma of the Third Ventricle
Professor of Neurology
Weill Medical College of Cornell University; James M. Drake, BSE, MBBCh, MSc, FRCSC Felice Giangaspero, MD
Chairman, Department of Neurology Professor Professor of Histopathology
Memorial Sloan-Kettering Cancer Center The University of Toronto; Department of Experimental Medicine and
New York, New York Chief of Neurosurgery Pathology
41: Lymphomas and Hemopoietic Neoplasms The Hospital for Sick Children University “La Sapienza”
Toronto, Ontario, Canada Rome, Italy
78: Midbrain Gliomas 85: Desmoplastic Infantile Ganglioglioma
Franco DeMonte, MD, FRCSC, FACS 92: Pediatric Meningiomas
Professor of Neurosurgery and Head and Neck
Surgery Wayne M. Gluf, MD
University of Texas M.D. Anderson Cancer Rose Du, MD, PhD Chief Resident, Department of Neurosurgery
Center Resident, Department of Neurological Surgery University of Utah
Houston, Texas University of California, San Francisco, Primary Children’s Medical Center
50: Carcinoma of the Paranasal Sinuses and School of Medicine Salt Lake City, Utah
Olfactory Neuroblastoma San Francisco, California 104: Spinal Column Tumors in Pediatric
61: Skull Base Metastasis 95: Craniopharyngioma Patients

Concezio Di Rocco, MD David H. Ebb, MD Karen Goddard, MBChB, FRCP

Professor of Pediatric Neurosurgery; Instructor and Assistant Professor, Department of Associate Professor of Radiation Oncology
Chief of Pediatric Neurosurgery Pediatrics University of British Columbia;
Catholic University of Rome Harvard Medical School; Radiation Oncologist
Rome, Italy Assistant Pediatrician British Columbia Cancer Agency
85: Desmoplastic Infantile Ganglioglioma Pediatric Hematology-Oncology Vancouver, British Columbia
Massachusetts General Hospital 103: Epidural Spinal Tumors
Eduardo M. Diaz, Jr., MD, FACS Boston, Massachusetts
Associate Professor; 20: Pilocytic Astrocytoma Ziya Gökaslan, MD, FACS
Center Medical Director Professor of Neurosurgery, Oncology, and
Department of Head and Neck Surgery Orthopedic Surgery;
University of Texas M.D. Anderson Cancer Said El-Shihabi, MD Director of Neurosurgical Spine Program
Center Resident in Neurological Surgery The Johns Hopkins Medical Institutions
Houston, Texas University of Arkansas for Medical Sciences Baltimore, Maryland
50: Carcinoma of the Paranasal Sinuses and Little Rock, Arkansas 21: Pleomorphic Xanthoastrocytoma
Olfactory Neuroblastoma 55: Management of Central Nervous System 70: Malignant Primary Tumors of the Vertebral
Metastases from Breast Carcinoma Column
Peter B. Dirks, MD, PhD, FRCSC
Assistant Professor of Neurosurgery Daryl R. Fourney, MD, FRCSC Ira M. Goldstein, MD
The University of Toronto; Assistant Professor of Neurosurgery Assistant Professor
Staff Neurosurgeon University of Saskatchewan; Department of Neurological Surgery
The Hospital for Sick Children Director of Residency Training Program University of Medicine and Dentistry of New
Toronto, Ontario, Canada Royal University Hospital Jersey
75: Supratentorial Low-Grade Gliomas Saskatoon, Saskatchewan Newark, New Jersey
90: Supratentorial Primitive Neuroectodermal 50: Carcinoma of the Paranasal Sinuses and 67: Spinal Cord Astrocytomas: Presentation,
Tumors Olfactory Neuroblastoma Management, and Outcome
 Contributors xv

Robert Goodkin, MD Cynthia Hawkins, MD, PhD, FRCPC Muhammad M. Husain, MD

Professor of Neurological Surgery Assistant Professor of Pathology Associate Professor, Department of Pathology
University of Washington School of Medicine The University of Toronto; University of Arkansas for Medical Sciences
Seattle, Washington Neuropathologist Little Rock, Arkansas
73: Evaluation and Management of Benign The Hospital for Sick Children 49: Paragangliomas of the Skull Base
Peripheral Nerve Tumors and Masses Toronto, Ontario, Canada
80: Lhermitte-Duclos Disease
Fredric Grannis, MD 91: Dysembryoplastic Neuroepithelial Tumor Laura F. Hutchins, MD
Assistant Clinical Professor 100: Langerhans Cell Histiocytosis Professor of Internal Medicine
Department of Surgery University of Arkansas for Medical Sciences;
University of California, San Diego; Director of Hematology-Oncology
Staff Physician and Head Glenda Hendson, MBBCh, FRCPC Arkansas Cancer Research Center
Section of Thoracic and Vascular Surgery Clinical Assistant Professor of Pathology and Little Rock, Arkansas
City of Hope National Medical Center Laboratory Medicine 55: Management of Central Nervous System
Duarte, California University of British Columbia; Metastases from Breast Carcinoma
53: Brain Metastasis from Non–Small Cell Lung Pediatric Neuropathologist
Cancer Children’s and Women’s Health Centre of British
Columbia W. Bradley Jacobs, MD
Vancouver, British Columbia, Canada Resident, Division of Neurosurgery
Abhijit Guha, MD, FRCSC, FACS
103: Epidural Spinal Tumors University of Toronto
Professor of Neurosurgery Toronto, Ontario, Canada
The University of Toronto; 74: Malignant Peripheral Nerve Tumors
Attending Neurosurgeon Stephen John Hentschel, MD 107: Peripheral Nerve Tumors in Children
University Health Network; Assistant Professor of Neurosurgery
Co-Director and Senior Scientist University of Saskatchewan;
The Arthur and Sonia Labatt Brain Tumour Neurosurgeon John A. Jane, Jr., MD
Research Centre Royal University Hospital Assistant Professor, Department of Neurosurgery
The Hospital for Sick Children; Saskatoon, Saskatchewan University of Virginia Health System
Alan and Susan Hudson Chair of Neuro-Oncology 103: Epidural Spinal Tumors University of Virginia
The Hospital for Sick Children Charlottesville, Virginia
Toronto, Ontario, Canada 47: Pituitary Adenomas
74: Malignant Peripheral Nerve Tumors Randall T. Higashida, MD
Clinical Professor of Radiology, Neurological
Nalin Gupta, MD, PhD Surgery, Neurology, and Anesthesiology; Abdul-Rahman Jazieh, MD, MPH
Assistant Professor of Neurosurgery and Pediatrics Chief, Division of Interventional Neurovascular Associate Professor of Medicine and
University of California, San Francisco, Radiology Environmental Health;
School of Medicine; University of California, San Francisco, Associate Professor and Acting Division Director
Chief, Division of Pediatric Neurosurgery School of Medicine Department of Internal Medicine, Division of
University of California, San Francisco, Medical San Francisco, California Hematology and Oncology
Center 12: Neurointerventional Techniques University of Cincinnati College of Medicine
San Francisco, California The Vontz Center for Molecular Studies
81: Cerebellar Astrocytomas Dominique B. Hoelzinger, MS Cincinnati, Ohio
95: Craniopharyngioma Research Assistant 60: Multidisciplinary Management of Patients with
Translational Genomics Research Institute Brain Metastases from an Unknown Primary
Van V. Halbach, MD Phoenix, Arizona Tumor
Clinical Professor 3: Animal Models
Department of Radiology, Neurological Surgery,
Neurology, and Anesthesia and Perioperative Jaque Jumper, MD
Care John K. Houten, MD Clinical Fellow in Neuroradiology
University of California, San Francisco, Assistant Professor of Neurosurgery University of California San Francisco
School of Medicine Albert Einstein College of Medicine of Yeshiva San Francisco, California
San Francisco, California University; 26: Choroid Plexus Tumors
12: Neurointerventional Techniques Director, Neurosurgery Spine Service
Montefiore Medical Center
Bronx, New York Peter Jun, MD
Walter A. Hall, MD
67: Spinal Cord Astrocytomas: Presentation, Howard Hughes Fellow
Professor of Neurosurgery, Radiation Oncology,
Management, and Outcome Department of Neurological Surgery
and Radiology
University of California, San Francisco
University of Minnesota Medical School;
San Francisco, California
Active Staff in the Clinical Service of Creig S. Hoyt, MD 26: Choroid Plexus Tumors
Neurosurgery Theresa and Wane Caygill Professor and
Fairview University Medical Center Chairman
Minneapolis, Minnesota Department of Ophthalmology Tetsuo Kanno, MD
32: Dysembryoplastic Neuroepithelial Tumor University of California, San Francisco Professor, Department of Neurosurgery;
59: Gynecologic Malignancies San Francisco, California Director
8: Neuro-Ophthalmology Fujita Health University
Fadi Hanbali, MD Toyoake, Aichi, Japan
Assistant Professor of Neurosurgery and 51: Craniopharyngioma
Orthopaedics Surgery Eugen B. Hug, MD
University of Texas Medical Branch Professor of Radiation Oncology and Pediatrics
Galveston, Texas Dartmouth Medical School Bruce Kaufman, MD
61: Skull Base Metastasis Hanover, New Hampshire; Professor of Neurosurgery
Chief, Radiation Oncology Director, Division of Pediatric Neurosurgery
Marlan R. Hansen, MD Dartmouth Hitchcock Medical Center and Norris Medical College of Wisconsin;
Assistant Professor of Otolaryngology Cotton Cancer Center Chief, Pediatric Neurosurgery
University of Iowa Lebanon, New Hampshire Children’s Hospital of Wisconsin
Iowa City, Iowa 48: Chordomas and Chondrosarcomas of the Milwaukee, Wisconsin
7: Neurotology Cranial Base 82: Brainstem Gliomas
xvi Contributors

Brian D. Kavanagh, MD, MPH Johan M. Kros, MD, PhD Edward R. Laws, Jr., MD
Associate Professor Professor of Neuropathology Professor of Neurosurgery and Medicine
Department of Radiation Oncology Department of Pathology University of Virginia Medical School
University of Colorado Health Sciences Center Erasmus Medical Center Charlottesville, Virginia
Denver, Colorado Rotterdam, The Netherlands 47: Pituitary Adenomas
57: Brain Metastases: Renal Cell Carcinoma 23: Oligodendroglial Tumors
Michael L. Levy, MD, PhD
G. Evren Keles, MD Abhaya V. Kulkarni, MD, PhD, FRCSC Head, Division of Pediatric Neurosurgery
Assistant Professor Assistant Professor of Surgery Children’s Hospital of San Diego;
Department of Neurological Surgery The University of Toronto; Co-Director, Residency Training Program
University of California, San Francisco, Staff Neurosurgeon University of California, San Diego
School of Medicine The Hospital for Sick Children San Diego, California
San Francisco, California Toronto, Ontario, Canada 101: Pediatric Skull Base Tumors
9: Surgical Strategies in the Management of Brain 100: Langerhans Cell Histiocytosis
Tumors
Kevin O. Lillehei, MD
17: Diffuse Astrocytoma Sandeep Kunwar, MD Professor, Department of Neurosurgery
39: Melanocytic Tumors Assistant Professor University of Colorado Health Sciences Center
Department of Neurological Surgery Denver, Colorado
John R.W. Kestle, MD, FRCS, FACS University of California, San Francisco, 37: Lipomatous Tumors
Associate Professor of Neurology School of Medicine; 57: Brain Metastases: Renal Cell Carcinoma
University of Utah Principal Investigator, Brain Tumor Research
Salt Lake City, Utah Center;
Co-Director, Pituitary Treatment Center Mark E. Linskey, MD
86: Pleomorphic Xanthoastrocytoma
University of California, San Francisco Associate Professor and Chair
San Francisco, California Department of Neurological Surgery
Sara H. Kim, MD 6: Neuroendocrinology University of California, Irvine
Attending Medical Staff Orange, California
Glendale Adventist Medical Center 55: Management of Central Nervous System
Glendale, California William J. Kupsky, MD Metastases from Breast Carcinoma
36: Adult Supratentorial Primitive Professor of Pathology
Neuroectodermal Tumors Wayne State University School of Medicine;
Chief of Neuropathology Russell R. Lonser, MD
Detroit Medical Center Staff Neurosurgeon
Bette K. Kleinschmidt-DeMasters, MD Detroit, Michigan Surgical Neurology Branch
Professor of Pathology, Neurology, and 35: Medulloblastoma National Institute of Neurological Disorders and
Neurosurgery Stroke
University of Colorado Health Sciences National Institutes of Health
Keith Kwok, MD
Center Bethesda, Maryland
Visiting Research Scholar
Denver, Colorado 68: Spinal Cord Hemangioblastomas
University of Washington School of Medicine
37: Lipomatous Tumors
Seattle, Washington
73: Evaluation and Management of Benign R. Loch Macdonald, MD, PhD
Michel Kliot, MD Peripheral Nerve Tumors and Masses Associate Professor of Surgery and Radiation and
Associate Professor Cellular Oncology
Department of Neurological Surgery Robert D. Labrom, MBBS, MSc, FRACS University of Chicago Hospitals
University of Washington School of Medicine Senior Lecturer Chicago, Illinois
Seattle, Washington University of Queensland; 98: Colloid Cysts in Children
73: Evaluation and Management of Director of Pediatric Orthopaedic Surgery and
Benign Peripheral Nerve Tumors Pediatric Spinal Surgeon Cormac O. Maher, MD
and Masses Royal Children’s Hospital Clinical Fellow, Department of Neurosurgery
Brisbane, Queensland, Australia Mayo Graduate School of Medicine;
Paul Kongkham, MD 103: Epidural Spinal Tumors Chief Resident Associate, Department of
Resident, Division of Neurosurgery Neurosurgery
The Hospital for Sick Children Enrico C. Lallana, MD Mayo Clinic
Toronto, Ontario, Canada Director of Medical Neuro-Oncology Rochester, Minnesota
108: Neurocutaneous Syndromes Seacoast Cancer Center 89: Medulloblastoma
Wentworth-Douglass Hospital
William E. Krauss, MD Dover, New Hampshire Todd Mainprize, MD
Neurosurgery Consultant 41: Lymphomas and Hemopoietic Neoplasms Resident, Division of Neurosurgery
Mayo Clinic The University of Toronto
Rochester, Minnesota Kathleen R. Lamborn, PhD The Hospital for Sick Children
64: Spinal Meningiomas Adjunct Professor Toronto, Ontario, Canada
Departments of Neurological Surgery and 80: Lhermitte-Duclos Disease
Epidemiology and Biostatistics 91: Dysembryoplastic Neuroepithelial Tumor
Ali F. Krisht, MD, FACS
Assistant Professor, Department of University of California, San Francisco,
Neurosurgery School of Medicine; Adam N. Mamelak, MD
University of Arkansas for Medical San Francisco, California Associate Professor
Sciences 16: Clinical Trials City of Hope Cancer Center;
Director, Neuroendocrine Clinic and 28: Gliomatosis Cerebri Visiting Associate in Biology
Cerebrovascular Clinic; California Institute of Technology;
Chief of Service, Department of Jeffrey L. Laurent, MD Director of Neurosurgery
Neurosurgery Resident, Department of Neurosurgery City of Hope Cancer Center
John L. McClellan Memorial Veterans Clinic University of Texas M.D. Anderson Cancer Center Duarte, California
Little Rock, Arkansas Houston, Texas 53: Brain Metastasis from Non–Small Cell Lung
52: Epidermoid Tumors 47: Pituitary Adenomas Cancer
 Contributors xvii

Timothy B. Mapstone, MD Minesh Mehta, MBChB W. Jerry Oakes, MD

Vice Chairman and Professor Chair and Professor, Department of Human Professor of Neurosurgery
Department of Neurological Surgery Oncology University of Alabama
Emory University; University of Wisconsin School of Medicine Birmingham, Alabama
Chief Pediatric Neurosurgeon Madison, Wisconsin 105: Pediatric Intradural and Extramedullary
Children’s Healthcare of Atlanta/Egleston 10: Basic Concepts Underlying Radiation Therapy Spinal Cord Tumors
Atlanta, Georgia
83: Dorsally Exophytic Brainstem Gliomas Hal S. Meltzer, MD Cian J. O’Kelly, MD
Associate Professor of Neurosurgery Resident Physician
Charles Matouk, MD University of California, San Diego, The Hospital for Sick Children;
Neurosurgical Resident School of Medicine; Resident Physician
The University of Toronto Department of Pediatric Neurosurgery Division of Neurosurgery
Toronto, Ontario, Canada Children’s Hospital of San Diego The University of Toronto
78: Midbrain Gliomas San Diego, California Toronto, Ontario, Canada
101: Pediatric Skull Base Tumors 76: Optic Pathway Gliomas
Masao Matsutani, MD
Department of Neurosurgery Thomas E. Merchant, DO, PhD Nezih Oktar, MD
Saitama Medical School Chief, Division of Radiation Oncology Professor
Iruma-gun, Saitama, Japan St. Jude Children’s Research Hospital Department of Neurosurgery
42: Germ Cell Tumors Memphis, Tennessee Ege University School of Medicine
88: Ependymoma Izmir, Turkey
Cordula Matthies, MD, PhD 39: Melanocytic Tumors
Professor of Neurosurgery Christina A. Meyers, PhD
Hannover Medical School; Edward H. Oldfield, MD
Professor of Neuropsychology Chief, Surgical Neurology Branch
Neurosurgical Attending University of Texas M.D. Anderson Cancer Center
Klinikum Hannover Nordstadt National Institute of Neurological Disorders and
Houston, Texas Stroke
Hannover, Germany 15: Functional Outcomes
43: Acoustic Neuromas (Vestibular Schwannomas) Bethesda, Maryland
68: Spinal Cord Hemangioblastomas
Rajiv Midha, MD, FRCS
J. Gordon McComb, MD Head and Chairman, Division of Neurosurgery
Professor, Department of Neurological Surgery Alessandro Olivi, MD
Department of Clinical Neurosciences Professor of Neurosurgery and Oncology
University of Southern California University of Calgary
Keck School of Medicine; Johns Hopkins University School of Medicine;
Foothills Medical Centre Director of Neurosurgical Oncology
Head, Division of Neurosurgery Calgary, Alberta, Canada
Los Angeles Childrens Hospital Johns Hopkins Hospital
74: Malignant Peripheral Nerve Tumors Baltimore, Maryland
Los Angeles, California 107: Peripheral Nerve Tumors in Children
102: Tumors of the Skull 29: Chordoid Glioma of the Third Ventricle

Yuriko Minn, MS Ian F. Parney, MD, PhD

Paul C. McCormick, MD, MPH
Medical Student Assistant Professor
Professor of Neurosurgery
Stanford University Departments of Clinical Neurosciences and
The Neurological Institute of New York
Palo Alto, California Oncology
Columbia-Presbyterian Medical Center
1: Epidemiology of Brain Tumors University of Calgary
Columbia University College of Physicians and
Calgary, Alberta
Surgeons
Robert J. Morgan, Jr., MD 11: Chemotherapy Principles
New York, New York
Assistant Clinical Professor of Medicine 19: Glioblastoma Multiforme
62: Spinal Axis Tumors: Incidence, Classification,
University of California; 65: Myxopapillary Ependymomas
and Diagnostic Imaging
66: Intramedullary Ependymomas Professor of Oncology;
Staff Physician; Andrew T. Parsa, MD, PhD
Director of Continuing Medical Education Assistant Professor
Michael W. McDermott, MD Department of Neurological Surgery
City of Hope Comprehensive Cancer Center
Associate Professor Principal Investigator, Brain Tumor Research
Duarte, California
Department of Neurological Surgery Center
53: Brain Metastasis from Non–Small Cell Lung
University of California, San Francisco, University of California, San Francisco,
Cancer
School of Medicine School of Medicine;
San Francisco, California San Francisco, California
26: Choroid Plexus Tumors: Choroid Plexus Reed Murtagh, MD 13: Immunotherapy
Papilloma and Choroid Plexus Carcinoma Neuro-Oncology Program 33: Pineocytoma
95: Craniopharyngioma H. Lee Moffitt Cancer Center; 62: Spinal Axis Tumors: Incidence, Classification,
Department of Radiology and Diagnostic Imaging
Patrick McDonald, MD, FRCSC University of South Florida 65: Myxopapillary Ependymomas
Assistant Professor of Neurosurgery Tampa, Florida 66: Intramedullary Ependymomas
University of Manitoba; 18: Anaplastic Astrocytoma 72: Intrinsic Metastatic Spinal Cord Tumors
Pediatric Neurosurgeon
Winnipeg Children’s Hospital Prithvi Narayan, MD James Pearlman, MD
Winnipeg, Manitoba, Canada Assistant Professor of Pediatric Neurosurgery Assistant Professor
78: Midbrain Gliomas Mount Sinai School of Medicine University of South Florida;
New York, New York Physician
Roger E. McLendon, MD 83: Dorsally Exophytic Brainstem Gliomas H. Lee Moffitt Cancer Center;
Professor of Pathology; Physician
Chief of Neuropathology; Tien T. Nguyen, MD VA-Hospital;
Director of Anatomic Pathology Services Division of Neurosurgery Physician
Duke University School of Medicine Los Angeles Children’s Hospital Tampa General Hospital
Durham, North Carolina Los Angeles, California Tampa, Florida
25: Ependymoma 102: Tumors of the Skull 18: Anaplastic Astrocytoma
xviii Contributors

Richard D. Pezner, MD Alfredo Quinones-Hinojosa, MD Timothy Ryken, MD

Clinical Professor, Department of Radiation Neurosurgery Chief Resident Associate Professor
Oncology University of California, San Francisco, Departments of Neurosurgery and Radiation
University of California Irvine School of Medicine Oncology
Orange, California; San Francisco, California University of Iowa College of Medicine;
Associate Chairman, Division of Radiation 26: Choroid Plexus Tumors University of Iowa Hospitals and Clinics
Oncology Iowa City, Iowa
City of Hope National Medical Center Ganesh Rao, MD 58: Intracerebral Metastatic Colon Carcinoma
Duarte, California Resident, Neurosurgery
53: Brain Metastasis from Non–Small Cell Lung University of Utah School of Medicine Madjid Samii, MD, PhD
Cancer Salt Lake City, Utah Professor Emeritus
86: Pleomorphic Xanthoastrocytoma Hannover Medical School;
Joseph H. Piatt, Jr., MD, FAAP Director, International Neuroscience Institute
Professor of Neurological Surgery and Pediatrics Hannover, Germany
Drexel University College of Medicine; Corey Raffel, MD, PhD 43: Acoustic Neuromas (Vestibular
Chief, Section of Neurosurgery Professor of Neurosurgery Schwannomas)
St. Christopher’s Hospital for Children Mayo Clinic College of Medicine
Philadelphia, Pennsylvania Mayo Clinic John H. Sampson, MD, PhD
106: Intramedullary Spinal Tumors Rochester, Minnesota Associate Professor of Neurosurgery and
89: Medulloblastoma Pathology
Russell O. Pieper, PhD
Associate Professor of Neurological Surgery Duke University Medical Center
University of California, San Francisco Vaneerat Ratanatharathorn, MD, MBA Durham, North Carolina
San Francisco, California Professor and Chairman of Radiation Oncology 56: Brain Metastases from Malignant Melanoma
2: Molecular and Cell Biology University of Arkansas for Medical Sciences;
Head of Radiation Oncology Robert A. Sanford, MD
Ian F. Pollack, MD Arkansas Cancer Research Center; Professor of Neurosurgery
Walter Dandy Professor of Neurological Surgery Co-Director Pediatric Neurosurgery, Semmes-Murphey Clinic
University of Pittsburgh School of Medicine; University of Arkansas Gamma Knife Center University of Tennessee College of Medicine
Co-Director Little Rock, Arkansas Semmes-Murphey Neurologic & Spine Institute
University of Pittsburgh Cancer Institute Brain 55: Management of Central Nervous System Memphis, Tennessee
Tumor Center; Metastases from Breast Carcinoma 88: Ependymoma
Professor of Neurosurgery
Children’s Hospital of Pittsburgh; Kevin P. Redmond, MD Michael Sargent, MD, MRCP, FRCR, FRCPC
Professor of Neurosurgery Associate Clinical Professor, Department of Clinical Associate Professor
University of Pittsburgh School of Medicine Radiation Oncology Department of Radiology
Pittsburgh, Pennsylvania University of Cincinnati College of Medicine; University of British Columbia;
79: Supratentorial High-Grade Gliomas Associate Clinical Professor Pediatric Neuroradiologist and Assistant
Division of Radiation Oncology Director
Martin G. Pomper, MD, PhD Department of Radiology Department of Radiology
Associate Professor Barrett Cancer Center British Columbia Children’s Hospital
Departments of Radiology, Pharmacology and Cincinnati, Ohio Vancouver, British Columbia
Molecular Sciences and Oncology 60: Multidisciplinary Management of Patients with 103: Epidural Spinal Tumors
Johns Hopkins University Brain Metastases from an Unknown Primary
Baltimore, Maryland Tumor Raymond Sawaya, MD
29: Chordoid Glioma of the Third Ventricle Professor and Chair of Neurosurgery
Mery Beth Pawelek Chair in Neurosurgery
Michael D. Prados, MD Amyn Rojiani, MD, PhD
Director, Brain Tumor Center
Professor of Neurological Surgery Professor of Oncology and Pathology;
University of Texas M.D. Anderson Cancer Center
Charles B. Wilson Endowed Chair in Neurological Director of Neuropathology
Houston, Texas
Surgery H. Lee Moffitt Cancer Center
Director, Clinical Neuro-Oncology Program University of South Florida
S. Clifford Schold, Jr., MD
Brain Tumor Research Center Tampa, Florida
Professor and Chair
University of California, San Francisco, 18: Anaplastic Astrocytoma
Department of Interdisciplinary Oncology
School of Medicine
University of Southern Florida
San Francisco, California Lucy B. Rorke, MD Associate Center Director for Clinical Affairs
11: Chemotherapy Principles Attending Neuropathologist H. Lee Moffitt Cancer Center and Research
17: Diffuse Astrocytoma Department of Neuropathology Institute
19: Glioblastoma Multiforme Childrens Hospital of Philadelphia Tampa, Florida
28: Gliomatosis Cerebri Philadelphia, Pennsylvania 18: Anaplastic Astrocytoma
99: Atypical Teratoid/Rhabdoid Tumors
Richard A. Prayson, MD
Professor of Pathology Ali Schwaiki, MD
Cleveland Clinic Lerner College of Medicine at James T. Rutka, MD, PhD, FRCSC Fellow, Hematology-Oncology
Case Western Reserve University; Professor and Chair of Neurosurgery Department of Internal Medicine
Section Head of Neuropathology Dan Family Chair in Neurosurgery University of Arkansas for Medical Sciences
Department of Anatomic Pathology Director, The Arthur and Sonia Labatt Brain Little Rock, Arkansas
The Cleveland Clinic Tumour Research Centre 55: Management of Central Nervous System
Cleveland, Ohio The University of Toronto Metastases from Breast Carcinoma
30: Mixed Neuronal and Glial Tumors Toronto, Ontario, Canada
76: Optic Pathway Gliomas Theodore Schwartz, MD
James N. Provenzale, MD 80: Lhermitte-Duclos Disease Assistant Professor of Neurological Surgery
Professor of Radiology; 84: Cervicomedullary Astrocytomas New York-Presbyterian Hospital
Chief, Neuroradiology 92: Pediatric Meningiomas New York Weill Cornell Medical Center
Duke University Medical Center 94: Pituitary Tumors in Children Cornell Medical College
Durham, North Carolina 97: Choroid Plexus Tumors New York, New York
25: Ependymoma 108: Neurocutaneous Syndromes 66: Intramedullary Ependymomas
 Contributors xix

Hilliard F. Seigler, MD Justin S. Smith, MD, PhD Aramis Teixeira, MD

Professor of Surgery and Immunology Resident Department of Neurosurgery
Duke University Medical Center School of Department of Neurological Surgery University of Arkansas for Medical Sciences
Medicine University of California, San Francisco, Little Rock, Arkansas
Durham, North Carolina School of Medicine 49: Paragangliomas of the Skull Base
56: Brain Metastases from Malignant Melanoma San Francisco, California
28: Gliomatosis Cerebri Tarik Tihan, MD, PhD
Timothy D. Shafman, MD Associate Professor of Pathology
Department of Radiation Oncology Susan Snodgrass, MD, RPh University of California San Francisco;
Duke University Medical Center Assistant Professor of Neuro-Oncology Attending Pathologist
Durham, North Carolina H. Lee Moffitt Cancer Center and Research University of California San Francisco School of
56: Brain Metastases from Malignant Melanoma Institute Medicine
University of South Florida San Francisco, California
Tampa, Florida 17: Diffuse Astrocytoma
Patrick Shannon, MSc, FRCPC 21: Pleomorphic Xanthoastrocytoma
Assistant Professor 18: Anaplastic Astrocytoma
26: Choroid Plexus Tumors
Department of Pathobiology and Laboratory 34: Pineoblastoma
Medicine Mark M. Souweidane, MD, FACS
Associate Professor of Neurological Surgery 39: Melanocytic Tumors
The University of Toronto; 62: Spinal Axis Tumors: Incidence, Classification,
Staff Pathologist, Neuropathology The Weill Medical College of Cornell University;
Vice Chairman and Director of Pediatric and Diagnostic Imaging
The Toronto Western Hospital;
Consulting Staff Pathologist, Fetal Neurosurgery
Tadanori Tomita, MD
Neuropathology The Weill Medical College of Cornell University
Professor of Neurosurgery
The Mount Sinai Hospital Memorial Sloan-Kettering Cancer Center
Northwestern University Feinberg School of
Toronto, Ontario, Canada New York, New York
Medicine;
22: Subependymal Giant Cell Astrocytoma and 77: Thalamic Gliomas
Chairman, Division of Pediatric Neurosurgery;
Tuberous Sclerosis Complex Director, Falk Brain Tumor Center
Paul Steinbok, MBBS, BSc, FRCSc Children’s Memorial Hospital
Dennis C. Shrieve, MD, PhD Professor Chicago, Illinois
Professor and Chair Department of Surgery 93: Pineal Region Tumors in Children
Department of Radiation Oncology University of British Columbia;
University of Utah School of Medicine Head, Division of Neurosurgery Ivo W. Tremont-Lukats, MD
Salt Lake City, Utah British Columbia’s Children’s Hospital Director of Neuro-Oncology and Brain Tumor
45: Meningiomas Vancouver, British Columbia, Canada Center
103: Epidural Spinal Tumors University of Texas M.D. Anderson Cancer Center
Sheila K. Singh, MD Houston, Texas
Resident Caron Stralendorf, MB, BCh, FCP 27: Astroblastoma
Division of Neurosurgery Assistant Clinical Professor
University of British Columbia; Eve C. Tsai, MD
The University of Toronto
Pediatric Hematology, Oncology, Bone Marrow Neurosurgery Resident
Toronto, Ontario, Canada
Transplantion Toronto Western Hospital Research Institute
75: Supratentorial Low-Grade Gliomas
Children’s and Women’s Health Centre Toronto, Ontario
Vancouver, British Columbia, Canada 84: Cervicomedullary Astrocytomas
William H. Slattery III, MD 103: Epidural Spinal Tumors
Clinical Professor Martin J. van den Bent, MD, PhD
Department of Otolaryngology Neurologist, Neuro-Oncologist
Mandeep S. Tamber, MD
University of Southern California; Daniel den Hoed Cancer Center
Resident
Director of Clinical Research Erasmus University Hospital
Division of Neurosurgery
House Ear Institute Rotterdam, The Netherlands
The University of Toronto
Los Angeles, California 23: Oligodendroglial Tumors
Toronto, Ontario, Canada
7: Neurotology
90: Supratentorial Primitive Neuroectodermal
Scott R. VandenBerg, MD, PhD
Tumors
Andrew E. Sloan, MD Professor of Pathology and Neurological Surgery;
Assistant Professor University of California, San Francisco,
Gianpiero Tamburrini, MD School of Medicine
Wayne State University School of Medicine; Assistant Professor of Neurosurgery
Chief of Neuro-Oncology San Francisco, California
Pediatric Neurosurgical Unit 5: Pathologic Classification
Karmanos Cancer Institute Catholic University of Rome
Detroit, Michigan 31: Neuronal Tumors
Rome, Italy
35: Medulloblastoma 85: Desmoplastic Infantile Ganglioglioma G. Edward Vates, MD, PhD
Assistant Professor
Jefferson C. Slimp, PhD Nancy J. Tarbell, MD University of Rochester School of Medicine;
Associate Professor Professor of Radiation Oncology Attending Neurosurgeon
Department of Rehabilitation Medicine; Harvard Medical School; Strong Memorial Hospital;
Director, Neurophysiology Monitoring Program Head, Pediatric Radiation Oncology Director of Cerebrovascular and Skull Base
University of Washington School of Medicine Massachusetts General Hospital Surgery
Seattle, Washington Boston, Massachusetts University of Rochester School of Medicine
73: Evaluation and Management of Benign 20: Pilocytic Astrocytoma Rochester, New York
Peripheral Nerve Tumors and Masses 28: Gliomatosis Cerebri
Michael D. Taylor, MD, PhD 40: Hemangioblastomas
Edward R. Smith, MD Neurosurgery Resident
Instructor in Neurosurgery Sonia and Arthur Labatt Brain Tumor Research Corneila S. von Koch, MD, PhD
Harvard Medical School; Centre Clinical Instructor
Shillito Staff Associate in Pediatric Neurosurgery The Hospital for Sick Children Department of Neurosurgery
Children’s Hospital Toronto, Ontario, Canada University of California San Francisco
Boston, Massachusetts 91: Dysembryoplastic Neuroepithelial Tumor San Francisco, California
20: Pilocytic Astrocytoma 99: Atypical Teratoid/Rhabdoid Tumors 81: Cerebellar Astrocytomas
xx Contributors

Frank Vrionis, MD, PhD Gordon A. Watson, MD, PhD Margaret Wrensch, PhD, MPH
Associate Professor of Neurosurgery, Oncology, Department of Radiation Oncology Professor of Neurological Surgery
and Neuro-Oncology Program LDS Hospital Co-Director, Division of Neuroepidemiology
University of South Florida College of Salt Lake City, Utah University of California, San Francisco,
Medicine; 54: Small Cell Lung Carcinoma School of Medicine
Associate Professor of Neurosurgery, San Francisco, California
Otolaryngology, and Oncology; K. Michael Webb, MD 1: Epidemiology of Brain Tumors
Director of Spinal and Skull Base Oncology Resident, Department of Neurosurgery
H. Lee Moffitt Cancer Center & Research University of Virginia Health System Raafat Yahya, MD
Institute Charlottesville, Virginia Resident, Division of Neurosurgery
Tampa, Florida 47: Pituitary Adenomas The University of Toronoto
18: Anaplastic Astrocytoma The Hospital for Sick Children
Jason S. Weinstein, MD Toronto, Ontario, Canada
Toshihiko Wakabayashi, MD, PhD Neurosurgery Resident 80: Lhermitte-Duclos Disease
Associate Professor University of Oregon 97: Choroid Plexus Tumors
Center for Genetic and Regenerative Medicine Portland, Oregon
Nagoya University Hospital 37: Lipomatous Tumors Jun Yoshida, MD, PhD
Nagoya, Aichi, Japan Professor and Chairman
96: Germ Cell Tumors in Children Philip R. Weinstein, MD Department of Neurosurgery
Chief, Neurological Service Nagoya University Graduate School of Medicine
James Waldron, MD Veterans Administration Hospital; Nagoya, Aichi, Japan
Resident Professor, Department of Neurological Surgery 96: Germ Cell Tumors in Children
Department of Neurosurgical Surgery Co-Director, Neurospinal Disorders Program;
University of California, San Francisco, Principal Investigator, Brain and Spinal Injury W.K. Alfred Yung, MD
School of Medicine Center Professor and Chairman
San Francisco, California University of California, San Francisco Department of Neuro-Oncology
13: Immunotherapy San Francisco, California University of Texas M.D. Anderson Cancer Center
63: Nerve Sheath Tumors of the Spine 63: Nerve Sheath Tumors of the Spine Houston, Texas
69: Benign Tumors of the Spine 24: Mixed Gliomas
John C. Wellons III, MD
Sabrina M. Walski-Easton, MD Assistant Professor of Surgery and Pediatrics Ping-pin Zheng, MD
Neurosurgery Resident Children’s Hospital of Alabama Department of Pathology
University of Minnesota Medical School Birmingham, Alabama Josephine Nefkens Institute
Minneapolis, Minnesota 105: Pediatric Intradural and Extramedullary Erasmus Medical Center
59: Gynecologic Malignancies Spinal Cord Tumors Rotterdam, The Netherlands
23: Oligodendroglial Tumors
Jeremy Wang, MD Richard Wheeler, MD
Chief Resident, Neurosurgery Service Professor, Department of Internal Medicine; J. Chris Zacko, MD
Department of Surgery Medical Director Resident in Neurosurgery
The Weill Medical College of Cornell University Clinical Trials Office Virginia Commonwealth University
Memorial Sloan-Kettering Cancer Center Huntsman Cancer Institute Medical College of Virginia
New York, New York Salt Lake City, Utah Richmond, Virginia
71: Therapeutic Options for Treating Metastatic 54: Small Cell Lung Carcinoma 38: Fibrous Tumors
Spine Tumors
John G. Wolbers, MD, PhD Imad T. Zak, MD
Ronald E. Warnick, MD Department of Neurosurgery Assistant Professor of Radiology
Medical Director Erasmus University Medical Center Director
The Jewish Hospital; Rotterdam, The Netherlands Neuroradiology Fellowship Program
Director, Neuro-Oncology Division; 23: Oligodendroglial Tumors Harper University Hospital
Professor, Department of Neurosurgery; Wayne State University
Director, Division of Surgical Neuro-Oncology Marguerite Wotoczek-Obadia, AHT Detroit, Michigan
University of Cincinnati College of Medicine Coordinator of Neuro-Oncology Research 35: Medulloblastoma
Cincinnati, Ohio Laboratory
60: Multidisciplinary Management of Patients with H. Lee Moffitt Cancer Center and Research
Brain Metastases from an Unknown Primary Institute
Tumor Tampa, Florida
18: Anaplastic Astrocytoma
S e c t i o n A Basic Science
Section Editors
Mitchel S. Berger and Michael D. Prados

Chapter 1
EPIDEMIOLOGY OF Margaret Wrensch, Terri Chew, Mitchel S. Berger, Yuriko Minn,
Melissa L. Bondy
BRAIN TUMORS

Epidemiology is concerned with measurements of health in brain tumor, such as diet, smoking, alcohol consumption, occu-
human populations—quantifying the incidence and prevalence pational and industry hazards, exposure to radiation, and inher-
of a disorder, disability, or injury; its distribution; and the risk ited genetic factors through either cohort studies (the risk of
factors that may determine its occurrence and course. At times, brain tumor in people with or without a specific characteristic)
epidemiologic evidence even suggests the root cause. Long or case-control studies (comparative histories of people with
restricted to evaluating mainly clinical, developmental, and and without brain tumors).
environmental exposures, epidemiologists now have open to Recent dramatic progress in the molecular classification of
them molecular and genetic techniques that provide new areas brain malignancies and the identification of genes suspected to
for exploration and may lead to refinements in exposure mea- play a part in their progression have resulted in new approaches
surements. In the study of brain tumors, molecular and genetic to delivering therapy to brain tumors and breaking down their
epidemiology may help define the disease. resistance to drugs and radiation. These advancements lend
Primary brain tumors are among the top 10 causes of renewed incentive in seeking means for prevention and cure of
cancer-related deaths in the United States, accounting for these devastating diseases.
approximately 1.4% of all cancers and 2.4% of all cancer-
related deaths.1 About 14 per 100,000 people in the United
States are diagnosed with a primary brain tumor each year, and
6 to 8 per 100,000 are diagnosed with a primary malignant DESCRIPTIVE EPIDEMIOLOGY:
brain tumor.8 Approximately 35,000 people in the United States PATIENTS’ CHARACTERISTICS AND THE
are newly diagnosed with a primary brain tumor each year. The
prognosis for patients who develop these tumors is bleak, with ASSOCIATED RISK OF BRAIN TUMOR
average survival after diagnosis ranging from 6 to 8 years for
those with low-grade astrocytoma or oligodendroglioma to 12 Issues in Describing the Incidence of
to 18 months for those with glioblastoma.1 The only proven
causes of brain tumors—inherited genetic syndromes,6 thera- Brain Tumors
peutic ionizing radiation,30 and immunosuppression causing Among children younger than 14 and adults 70 years old and
brain lymphoma16,35—account for a small number of the cases older, incidence rates for primary brain malignancies were sig-
diagnosed each year. nificantly higher from 1991 to 1995 than from 1975 to 1979.24
Epidemiologic studies have an important role in research In the 15- to 44-year-old age group, there were no meaningful
into the causes, progression, and outcome of primary brain differences in overall rates between the two periods. For people
tumors. Descriptive epidemiologic studies link histologic in the 45- to 64-year-old age group, rates were somewhat lower
tumor type and demographic characteristics (such as age, sex, for the more recent period. Increases in the incidence of
and race) of patients affected by primary brain tumors to asso- primary malignant brain tumors, particularly among the elderly,
ciated incidence, prevalence, mortality, and survival rates. Ana- have been attributed to several factors: improved diagnostic
lytic epidemiologic studies analyze suspected risk factors for procedures, such as computed tomography and magnetic reso-
3
4 S E C T I O N A Basic Science

nance imaging (MRI); a greater availability of neurosurgeons; with more accessible and highly developed medical care.20,30
changing patterns of access to medical care; and evolving However, an exception is in Japan, where access to medical
medical approaches toward elderly patients.11,19,24 Although care is similar to that in Northern Europe, but where the inci-
environmental factors have been implicated in some analytic dence rate for malignant brain tumors is less than half that in
epidemiologic studies, no risk factors accounting for a large Northern Europe. Among other influences are cultural, ethnic,
percentage of brain tumors have been identified. For this or geographic differences in risk factors. In the United States,
reason, researchers cannot yet hope to quantitatively explain glioma affects more whites than blacks, but the incidence of
temporal trends on the basis of changes in environmental meningioma is nearly equal among blacks and whites. Thus the
factors. At present, any comparisons across time periods or differences are not likely to be attributable only to differences
across studies are complicated by a variety of factors. Incidence between blacks and whites in their access to health care or in
rates can differ among studies simply because of differences in diagnostic practices.34 The absolute variation in brain tumor
definitions and methodologies in ascertaining and categorizing incidence rates from high-risk to low-risk areas, in both the
tumors.29 Registry data can suffer from ascertainment biases as United States and the world, is about fourfold to fivefold.20
a result of reporting differences and variability in the availabil-
ity of health care. The complexity of the anatomic, pathologic,
and clinical classifications of brain tumors is problematic, and Survival and Prognostic Factors
there is still dispute about how some tumor histologies, espe- The 5-year survival rate for all ages and all brain tumor types
cially mixed tumor types, should be classified. This complex- in the United States is 20% (95% confidence interval [CI], 18%
ity is apparent when one considers that the Central Brain Tumor to 22%).12 Another measurement of patients’ survival that is of
Registry of the United States categorizes 30 different histologic interest is the conditional probability of survival to 5 years
types of primary brain and central nervous system tumors based on survival for the first 2 years after diagnosis.14 In the
resulting from 215 individual histology codes.8 In the future, United States between 1979 and 1993, the conditional proba-
the use of genetic or other markers in conjunction with neu- bility of surviving another 3 years after survival to 2 years for
ropathologic diagnosis may help to achieve a more precise all patients with primary malignant brain and other tumors of
system of tumor classification. Above all, a uniform, accurate, the central nervous system was 76.2% (95% CI, 74.8% to
and unbiased method for the registration of both benign and 77.6%); for glioblastoma, the conditional probability of sur-
malignant brain tumors in adults and children would help to vival for the same period was 36.4%; and the conditional prob-
clarify variations in the incidence of brain tumors.11,18 abilities of survival for patients with other tumor types were
greater than 60%. Survival is known to be strongly related to
the patient’s age and tumor type.8 Patients with glioblastoma
Age multiforme consistently have the poorest survival rates in all
Patients’ median age at onset for all primary brain tumors is age groups. Within any tumor type, older patients have poorer
about 57 years.8 Age distributions differ by site and by the his- survival rates than younger patients. An exception is medul-
tologic type of tumor, suggesting that different histologic types loblastoma or embryonal primitive tumors, which rarely occur
may have different etiologies. Some of this variation may be in people older than 44 years. Children diagnosed before the
accounted for by differing diagnostic practices and access to age of 3 have poorer survival rates than children diagnosed at
diagnosis in different age groups. It seems probable that dura- ages 3 to 14.17 For all primary malignant brain tumors com-
tion of exposure required for malignant transformation, the bined, the 5-year survival rate in children younger than 14 years
number of genetic alterations required to produce clinical is 72%.17
disease, or poorer immune surveillance with advancing age Overall survival for patients with primary malignant brain
accounts for those tumor types that increase in incidence with tumors has not improved much since the early 1970s,24 but this
age. A peak in incidence of brain tumors among young chil- factor, too, varies by age and histologic tumor type. Modest
dren—some, but not all, of which can be attributed to medul- gains in survival time were achieved between 1975 and 1995
loblastoma and other tumors of primitive neuroectodermal for people younger than 65, but there was little or no change
origin—remains an intriguing and incompletely explained for patients 65 years and older. Little progress has been made
feature of brain tumor epidemiology. in survival from glioblastoma in the past 20 years, but 5-year
survival rates for patients with medulloblastoma increased 20%
from the 1970s to the 1980s. More recently, survival rates have
Sex leveled off.12
Overall, primary brain tumors are more common in males than Age and tumor histology are still the strongest prognostic
females.34 However, meningiomas are about twice as common indicators for patients who have a brain tumor, although some
in females as in males, and tumors of the cranial and spinal other factors have been shown to influence survival to some
nerves and of the sellar region affect males and females almost extent. In all but 2 of 17 European countries, 5-year survival
equally.8 rates were somewhat better for women with primary malignant
brain tumors than for men (20% versus 17%).32 The location of
a tumor and the extent of tumor resection are also factors pre-
Geographic and Ethnic Variations dictive of overall or progression-free survival.14
Both ascertainment bias and inconsistent reporting complicate
the interpretation of geographic and ethnic variations in the
incidence of brain tumors. Access to health care is an impor- Prevalence of Brain Tumors in the United States
tant geographic and ethnic factor because reported rates for The prevalence of brain tumors, which reflects their incidence
primary malignant brain tumors tend to be higher in countries and the duration of patients’ survival, gives an idea of the extent
 C H A P T E R 1 Epidemiology of Brain Tumors 5

Ta b l e 1 - 1 Potential Risk Factors for Primary Brain Tumors and Selected Epidemiologic Studies

Authors Potential Risk Factor Outcome of Selected Epidemiologic Studies

Wrensch et al.40 Alcohol Slight to no association.40
Schlehofer et al.33 Allergies An inverse association of allergic diseases with glioma.7,33,39 Statistically
Wiemels et al.39 significant inverse dose response of glioma with increasing numbers of
Brenner et al.7 allergens.39
Inskip et al.21 Cellular telephones No association found in current studies.21,28
Muscat et al.28
Caggana et al.9 Constitutive polymorphisms Silent polymorphisms in DNA repair genes linked to oligoastrocytoma,10
Chen et al.10 (in carcinogen metabolizing, glioblastoma,9 and astrocytoma.9 Conflicting results for associations of
Elexpuru-Camiruaga et al.15 DNA repair, and immune cytochrome p4502D6 and glutathione transferase theta polymorphisms
Kelsey et al.22 function genes) with increased risk.15,22,38
Trizna 38
Blot et al.2 Diet and vitamins Mixed support for association of N-nitroso compounds, antioxidants, or
Lubin et al.27 specific nutrients with adult or childhood brain tumors.2,28,40
Wrensch et al.40
Preston-Martin & Mack30 Drugs and medications Few statistically significant or consistent findings between brain tumor
development and studied drugs.30
Lote et al.26 Epilepsy, seizures, or Some association; causality is difficult to determine, because seizures are
Schlehofer et al.33 convulsions often a symptom of brain tumor.26,33
Wrensch et al.41 Family history of brain Reported relative risks range from almost 1 to 10.41 Genetics or possible
tumors common exposure to environmental factors may contribute.
Bondy et al.6 Hereditary syndromes* Convincing evidence that certain inherited genes may strongly influence
the risk of primary brain tumor development.6
Inskip et al.20 Infectious agents or Some viruses cause brain tumors in experimental animals,20 but except
Gavin & Yogev16 immunosuppression* for studies of HIV-related brain lymphomas,16,35 few epidemiologic studies
Taiwo35 have evaluated a potential role of viruses in human brain tumorigenesis.20
Loomis & Wolf25 Ionizing radiation Diagnostic radiation and nuclear material exposure may create elevated
Preston-Martin & Mack30 (therapeutic,* diagnostic, risk for some tumor types.25,30 Therapeutic ionizing radiation is a strong risk
Wrensch et al.40 other) factor for brain tumorigenesis.30,40
Bondy et al.4 Lymphocyte mutagen Significantly associated with a risk of glioma.4,5
Bondy et al.5 sensitivity (to gamma
radiation)
Thomas & Waxweiler37 Occupation and industry Difficult to establish an association between exposure to specific
occupational or industrial chemicals and human brain tumorigenesis.37
Exposure to multiple chemicals and chemical interactions are two of the
complicating factors.
Zahm43 Personal and residential Some studies have shown a link between pesticide exposure and risk of
chemical exposures childhood brain cancers.43
Wrensch et al.42 Power frequency No causal connection between EMF and risk of brain tumor has been
electromagnetic field established.42 Measurement and unknown latency are complicating factors.
Teppo et al.36 Prior cancers Increased risk among patients with small cell lung carcinoma and
adenocarcinoma, but interpretation complicated by possibility of confusing
metastatic and primary brain tumors.36
Boffetta et al.3 Tobacco smoke exposure No major contribution of maternal tobacco smoking to risk of childhood
Lee et al.25 brain tumors,3 small associations with paternal smoking. No important
contribution of smoking filtered cigarettes to adult brain tumors.23
Raaschou-Nielsen et al.31 Traffic-related air pollution No association between childhood brain tumor and exposure to
traffic-related air pollution.31

*The only factors so far proved to cause primary brain tumors.

Cited references often have more detailed reviews of studies of these factors.
6 S E C T I O N A Basic Science

of the disease burden in an area, especially for brain tumor gens. Unraveling the genetic, molecular, and cytogenetic errors
types associated with relatively long survival times, like in primary brain tumors is important to determining their patho-
meningioma. Davis and colleagues13 recently published the first genesis. As work on elucidating patterns of molecular change
estimates of prevalence of primary brain tumors in the United continues, a more precise classification of brain tumors may
States. They found that primary benign brain tumors had an evolve, making it possible to identify groups of tumors that are
estimated prevalence of 97.5 per 100,000 population for the more homogeneous than current histologic groupings with
year 2000 and emphasized the need for further studies of respect to causal factors. Table 1-1, which summarizes only
etiology and consideration of quality-of-life issues for this some of the vast number of epidemiologic studies of potential
group of patients. risk factors for brain tumor that have been performed, provides a
sample of some of the major factors currently being analyzed.

ANALYTIC EPIDEMIOLOGY OF BRAIN CONCLUSION

TUMORS: STUDIES OF RISK FACTORS
There are very few proven causes for brain tumors, and they
Within the brain tumor research community, there is little una- account for only a small proportion of cases. It is probable that
nimity about the nature and magnitude of the risk factors for primary brain tumors stem from multiple exogenous and
primary brain tumors. These tumors are highly heterogeneous endogenous events. Despite a persistently disheartening prog-
histologically, and definitions and classifications of tumors nosis for patients who have a malignant brain tumor, the
often differ from one study to another. These factors, combined promise of molecular and genetic approaches has intensified
with retrospective assessments of exposure to risk factors and efforts to define their biologic course and causes. Through these
undefined latency periods, make for imprecise estimates of efforts, new concepts about neuro-oncogenesis are emerging to
associations. Often, such limitations also make it difficult to explicate the epidemiology of brain tumors and provide hope
compare studies. Differences in the eligibility criteria estab- for a cure.
lished for patients and control groups and the use of proxies
further complicate the synthesis of results among studies. In
addition, certain biologic and physiologic characteristics of the ACKNOWLEDGMENTS
brain itself, such as the blood-brain barrier, add challenges to
determining the risk factors for brain tumors. This chapter was adapted with permission from Wrensch M,
Primary brain tumors are thought to develop through an Minn Y, Chew T, Bondy M, Berger MS. Epidemiology of
accumulation of genetic alterations that permit cells to evade primary brain tumors: current concepts and review of the liter-
normal regulatory mechanisms and escape destruction by the ature. Neuro-Oncol 4:278-99, 2002. The authors thank Sharon
immune system. In addition to inherited alterations in crucial Reynolds for editorial collaboration and Susan Eastwood for
genes that control the cell cycle, chemical, physical, and biologic helpful suggestions. Funding for this chapter was from National
agents that damage DNA are suspected potential neurocarcino- Institutes of Health grants R01CA52689 and 1P50CA97257.

References

1. ACS (American Cancer Society): Cancer Facts and Figures 2002. 9. Caggana M, Kilgallen J, Conroy JM, et al: Associations
Atlanta, American Cancer Society, 2002. between ERCC2 polymorphisms and gliomas. Cancer Epidemiol
2. Blot WJ, Henderson BE, Boice JD, Jr: Childhood cancer in rela- Biomarkers Prev 10:355–360, 2001.
tion to cured meat intake: review of the epidemiological evidence. 10. Chen P, Wiencke J, Aldape K, et al: Association of an ERCC1
Nutr Cancer 34:111–118, 1999. polymorphism with adult-onset glioma. Cancer Epidemiol
3. Boffetta P, Tredaniel J, Greco A: Risk of childhood cancer and Biomarkers Prev 9:843–847, 2000.
adult lung cancer after childhood exposure to passive smoke: a 11. Davis FG, Bruner JM, Surawicz TS: The rationale for standard-
meta-analysis. Environ Health Perspect 108, 73–82, 2000. ized registration and reporting of brain and central nervous system
4. Bondy ML, Kyritsis AP, Gu J, et al: Mutagen sensitivity and risk tumors in population-based cancer registries. Neuroepidemiology
of gliomas: a case-control analysis. Cancer Res 56:1484–1486, 16:308–316, 1997.
1996. 12. Davis FG, Freels S, Grutsch J, et al: Survival rates in patients with
5. Bondy ML, Wang LE, El-Zein R, et al: Gamma radiation sensi- primary malignant brain tumors stratified by patient age and
tivity and risk of glioma. J Natl Cancer Inst 93:1553–1557, 2001. tumor histological type: an analysis based on Surveillance,
6. Bondy M, Wiencke J, Wrensch M, et al: Genetics of primary brain Epidemiology, and End Results (SEER) data, 1973–1991.
tumors: a review. J Neurooncol 18:69–81, 1994. J Neurosurg 88:1–10, 1998.
7. Brenner AV, Linet MS, Fine HA, et al: History of allergies and 13. Davis FG, Kupelian V, Freels S, et al: Prevalence estimates for
autoimmune diseases and risk of brain tumors in adults. Int J primary brain tumors in the United States by behavior and major
Cancer 99:252–259, 2002. histology groups. Neuro-oncol 3:152–158, 2001.
8. CBTRUS (Central Brain Tumor Registry of the United States): 14. Davis FG, McCarthy BJ, Freels S, et al: The conditional proba-
Primary Brain Tumors in the United States Statistical Report bility of survival of patients with primary malignant brain tumors:
1995–1999. Chicago, Central Brain Tumor Registry of the United Surveillance, Epidemiology, and End Results (SEER) data.
States. 591–600, 2002–2003. Cancer 85:485–491, 1999.
 C H A P T E R 1 Epidemiology of Brain Tumors 7

15. Elexpuru-Camiruaga J, Buxton N, Kandula V, et al: Susceptibil- 30. Preston-Martin S, Mack W: Neoplasms of the nervous system. In
ity to astrocytoma and meningioma: influence of allelism at glu- Schottenfeld D, Fraumeni JF, Jr (eds): Cancer Epidemiology
tathione S-transferase (GSTT1 and GSTM1) and cytochrome and Prevention, ed 2. New York, Oxford University Press, 1996,
P-450 (CYP2D6) loci. Cancer Res 55:4237–4239, 1995. pp 1231–1281.
16. Gavin P, Yogev R: Central nervous system abnormalities in pedi- 31. Raaschou-Nielsen O, Hertel O, Thomsen BL, et al: Air pollution
atric human immunodeficiency virus infection. Pediatr Neurosurg from traffic at the residence of children with cancer. Am J
31:115–123, 1999. Epidemiol 153:433–443, 2001.
17. Grovas A, Fremgen A, Rauck A, et al: The National Cancer Data 32. Sant M, van der Sanden G, Capocaccia R: Survival rates for
Base report on patterns of childhood cancers in the United States. primary malignant brain tumours in Europe. EUROCARE
Cancer 80:2321–2332, 1997. Working Group. Eur J Cancer 34:2241–2247, 1998.
18. Gurney JG, Wall DA, Jukich PJ, et al: The contribution of non- 33. Schlehofer B, Blettner M, Preston-Martin S, et al: Role of medical
malignant tumors to CNS tumor incidence rates among children history in brain tumour development: results from the Interna-
in the United States. Cancer Causes Control 10:101–105, 1999. tional Adult Brain Tumour Study. Int J Cancer 82:155–160,
19. Helseth A: The incidence of primary central nervous system 1999.
neoplasms before and after computerized tomography availabil- 34. Surawicz TS, McCarthy BJ, Kupelian V, et al: Descriptive epi-
ity. J Neurosurg 83:999–1003, 1995. demiology of primary brain and CNS tumors: results from
20. Inskip PD, Linet MS, Heineman EF: Etiology of brain tumors in the Central Brain Tumor Registry of the United States, 1990–
adults. Epidemiol Rev 17:382–414, 1995. 1994. Neuro-oncol. [serial online], Doc. 98-13, January 19,
21. Inskip PD, Tarone RE, Hatch EE, et al: Cellular telephone use and 1999. URL neuro-oncology.mc.duke.edu. Neuro-oncol 1:14–25,
brain tumors. N Engl J Med 344:79–86, 2001. 1999.
22. Kelsey KT, Wrensch M, Zuo ZF, et al: A population-based case- 35. Taiwo BO: AIDS-related primary CNS lymphoma: a brief review.
control study of the CYP2D6 and GSTT1 polymorphisms and AIDS Read 10:486–491, 2000.
malignant brain tumors. Pharmacogenetics 7:463–468, 1997. 36. Teppo L, Salminen E, Pukkala E: Risk of a new primary cancer
23. Lee M, Wrensch M, Miike R: Dietary and tobacco risk factors for among patients with lung cancer of different histological types.
adult onset glioma in the San Francisco Bay Area (California, Eur J Cancer 37:613–619, 2001.
USA). Cancer Causes Control 8:13–24, 1997. 37. Thomas TL, Waxweiler RJ: Brain tumors and occupational risk
24. Legler JM, Ries LA, Smith MA, et al: Cancer surveillance series factors. Scand J Work Environ Health 12:1–15, 1986.
[corrected]: Brain and other central nervous system cancers: 38. Trizna Z, de Andrade M, Kyritsis AP, et al: Genetic
recent trends in incidence and mortality. J Natl Cancer Inst polymorphisms in glutathione S-transferase mu and theta, N-
91:1382–1390, 1999. acetyltransferase, and CYP1A1 and risk of gliomas. Cancer
25. Loomis DP, Wolf SH: Mortality of workers at a nuclear materials Epidemiol Biomarkers Prev 7:553–555, 1998.
production plant at Oak Ridge, Tennessee, 1947–1990. Am J Ind 39. Wiemels JL, Wiencke JK, Sison JD, et al: History of allergies
Med 29:131–141, 1996. among adults with glioma and controls. Int J Cancer 98:609–615,
26. Lote K, Stenwig AE, Skullerud K, et al: Prevalence and prognos- 2002.
tic significance of epilepsy in patients with gliomas. Eur J Cancer 40. Wrensch M, Bondy ML, Wiencke J, et al: Environmental risk
34:98–102, 1998. factors for primary malignant brain tumors: a review. J
27. Lubin F, Farbstein H, Chetrit A, et al: The role of nutritional habits Neurooncol 17:47–64, 1993.
during gestation and child life in pediatric brain tumor etiology. 41. Wrensch M, Lee M, Miike R, et al: Familial and personal medical
Int J Cancer 86:139–143, 2000. history of cancer and nervous system conditions among adults
28. Muscat JE, Malkin MG, Thompson S, et al: Handheld cellular with glioma and controls. Am J Epidemiol 145:581–593, 1997.
telephone use and risk of brain cancer. JAMA 284:3001–3007, 42. Wrensch MR, Minn Y, Bondy ML: Epidemiology. In Bernstein
2000. M, Berger MS (eds): Neuro-Oncology: The Essentials. New York,
29. Pobereskin LH, Chadduck JB: Incidence of brain tumours in two Thieme Medical Publishers, 2000.
English counties: a population based study. J Neurol Neurosurg 43. Zahm SH: Childhood leukemia and pesticides. Epidemiology
Psychiatry 69:464–471, 2000. 10:473–475, 1999.
 A

2
S e c t i o n

Chapter

MOLECULAR AND Russell O. Pieper and Joseph F. Costello

CELL BIOLOGY

CELL BIOLOGY all components necessary for duplication of DNA are created.
The DNA is duplicated in the S phase. The movement from G1
Our understanding of human cell biology has been facilitated to S phase is carefully regulated by the G1 checkpoint, which
by our understanding of the structure of the cell. The human in turn is carefully controlled by the expression of cyclin and
cell can be generally divided into two parts, the nucleus and the cyclin-associated proteins that monitor the process and allow
cytoplasm. The nucleus is the site of DNA and RNA synthesis or prohibit further cell-cycle progression. In S phase, the cell
and contains the genetic material of the cell arranged in 23 pairs replicates its DNA and moves into the G2 phase, in which com-
of chromosomes, each of which varies by size and which, upon ponents necessary for physical division of the cell are synthe-
staining, can be distinguished one from another under a simple sized. Movement of the cell out of G2 and into mitosis is also
light microscope. The approximate 3 m of human chromo- rigorously controlled by the G2 checkpoint, activation of which
somal DNA is packaged to fit inside a human cell of approxi- stalls entry into mitosis. Mitosis itself is divided into four
mately 20 mm in diameter by its association with histones, parts—prophase, metaphase, anaphase, and telophase—during
charged proteins around which the DNA is coiled and con- which the replicated chromosomes condense, align, are segre-
densed. In addition to associating with histones, the chromo- gated into two daughter cells, and ultimately decondense to
somal DNA is also attached in a transient manner to the nuclear take their normal histone-associated conformation. Following
matrix, a collection of scaffold proteins that further influences completion of replication, the cells can divide again or move
DNA packaging and accessibility of DNA to factors that influ- into the G0 phase of the cell cycle depending on the extracel-
ence gene expression. The nucleus is separated from the cyto- lular cues received.
plasm by the nuclear envelope, a double membrane breached The life of the cell following division consists of three
by nuclear pores that allow import and export of factors criti- primary functions: energy production, interaction with its envi-
cal in gene and protein expression. The cytoplasm itself is a ronment, and interaction with neighboring cells. The process of
viscous material filled with an interlaced pattern of protein fil- energy production is accomplished in the cytoplasm by mito-
aments called the cytoskeleton. The cytoskeleton organizes the chondria and takes one of two paths depending on the avail-
cytoplasm into compartments, which in turn contain the various ability of oxygen. In the presence of oxygen, cells rely on
intracellular organelles that carry out the intermediate metabo- aerobic glycolysis to convert glucose to pyruvate, after which
lism of the cell. These organelles include the ribosome-studded the pyruvate is converted via the citric acid cycle into CO2
endoplasmic reticulum, which carries out protein synthesis; the and H2O with the concomitant production of adenosine 5¢-
Golgi apparatus, which directs macromolecules to their appro- triphosphate (ATP). In tumors, however, the growth rate of cells
priate intracellular locations; the mitochondria, which produce often exceeds the availability of oxygen, a problem com-
energy; lysozomes, which degrade molecules; and peroxi- pounded by the disordered growth and poor vascularization
somes, which are involved in oxidative reactions. Surrounding characteristic of tumors. In tumors (and in normal tissues in
the cytoplasm is the cell membrane, a lipid bilayer embedded which oxygen is limited, such as exercising muscle), cells have
with proteins critical for interactions between the cell interior the ability to produce energy in the absence of oxygen by con-
and the extracellular environment. The proteins include trans- verting pyruvate to lactate, a reaction that yields considerably
porters, growth-factor receptors, neurotransmitter receptors, less ATP than complete oxidative phosphorylation. Although
and import and efflux proteins. Together, these components anaerobic glycolysis remains a secondary, fallback means of
allow the cell to carry out all functions necessary for its con- energy production in most cells, tumors, even under well-
tribution to the organism as a whole. oxygenated conditions, produce high levels of pyruvate and
New cells are created by the process of cell division. Cell often rely on anaerobic metabolism more than aerobic energy
division is in turn intimately linked to the cell cycle. The human generation.7 This so-called “Warburg effect” in part explains why
cell cycle is divided into four parts: G1, S, G2, and mitosis (M) many tumor cells appear to have very high rates of glucose
(with a fifth phase, G0, reserved for cells not in the cell cycle).14 metabolism and further illustrates the idea that, although all
Cells begin the process of cell division by preparing to repli- cells need to produce energy, the multiple paths by which this
cate their DNA in the G1 phase. In this phase, nucleotides and can occur allow for differences between normal and tumor cells.
8
 C H A P T E R 2 Molecular and Cell Biology 9

In addition to energy production, a second fundamen- not limited to viruses, bacterial toxins, and complement lead to
tal function of cells is interaction with the extracellular swelling of the cell, plasma membrane collapse, cell rupture,
environment. As noted, the cell membrane is studded with a and release of cellular contents into the environment. This type
variety of proteins whose primary role is to sense signals in of cell death, which is often considered to be nonprogrammed
the extracellular environment and to relay these signals to the and not dependent on expression of new genes, initiates an
interior and to the nucleus. The sensing of the extracellular immune response and can lead to the death of neighboring cells
environment is mediated by four major types of molecules: as well as the primary cell itself. In contrast, death by apopto-
transmembrane G–protein-linked receptors, receptor protein sis results not in cell lysis but rather in cell shrinkage, disinte-
tyrosine kinases, ligand-gated ion channels, and intracellular gration of the cell into small, well-enclosed apoptotic bodies,
receptors. Although the molecular details of how activation of and the display of unique lipids on the cell membrane that
these receptors is linked to cellular action vary, in each case allow engulfment of the apoptotic bodies by professional
ligand binding results in changes in the receptor or in receptor- macrophages or in many cases by neighboring cells. The apop-
associated proteins that lead either to alterations in ion move- totic process, like necrosis, can be triggered by external toxic
ment in and out of the cells or in changes in gene expression. stimuli, but can also be activated by receptor-ligand binding or
As an example, activation of the epidermal growth-factor by other physiologic processes. A specialized form of apopto-
receptor leads to activation of pathways that ultimately change sis, referred to as aniokis, can be triggered by detachment of
gene expression in ways that stimulate cell-cycle progression cells from components of the extracellular matrix. Because
and support cell survival. In addition, each signaling cascade apoptosis is considered to be a form of programmed cell death,
has the potential to amplify the signal such that a small stimu- the process is rigorously controlled by both proapoptotic and
lus can lead to prolonged cellular effects. Most significantly, antiapoptotic molecules that can swing the balance between
however, the overexpression of such receptors or the dysregu- cell survival and cell elimination. Furthermore, the process
lation of the intracellular pathways that link the receptors to appears to be critically linked to energy metabolism, because
cellular actions can lead to constitutive activation of these path- apoptosis appears to be initiated in the mitochondria and can
ways. In the case of growth-factor pathways controlled prima- take place in the absence of a cell nucleus. Although the cell
rily by receptor tyrosine kinases (such as the epidermal death process is critical for the remodeling of normal tissue, it
growth-factor receptor), this dysregulation can lead to unre- has special importance in cancer states, in which cells outside
stricted growth signaling and aberrant proliferation. Therefore, their normal environment must suppress activation of pro-
in cancers, and particularly in brain tumors, the pathways that grammed cell death pathways while at the same time main-
cells use to sense their environment can serve as a driving force taining activated growth pathways.
for the disease and potentially also as targets for therapeutic The normal life of the cell is a highly ordered process in
intervention. which cell replication gives rise to cells that produce energy,
Although energy production and interaction with the extra- interact with their environment and neighbors, and die in a
cellular environment are key cellular functions, from a larger highly ordered fashion. The initiation of the tumorigenic
organism-based view the ability of cells to interact appropri- process alters many, if not all, of these pathways, such that the
ately with neighboring cells is critical in keeping the organism surviving tumor cell must not only activate growth pathways
functioning properly. The ability of cells to interact with their and divide but also produce energy and suppress cell death in
neighbors is controlled by membrane-bound receptors known sometimes hostile and foreign cellular environments. Certainly
as integrins. Integrin complexes exist as dimers of one of 16 not all tumor cells can do so, and solid tumors such as gliomas
alpha subunits and one of 8 beta subunits, each dimer interact- often contain significant numbers of dead cells. Selection pres-
ing with a different extracellular component found in the extra- sures, however, can allow the emergence of cells with the
cellular matrix or in neighboring cell membranes.4 The appropriate combination of alterations that allow survival and
signaling pathways that emanate from ligated integrin com- proliferation. Our understanding of normal cell biology has
plexes are complicated and involve both positive and negative therefore provided significant clues about defects present in
signals regulating the cell cytoskeleton, cell adhesion, cell tumors, including brain tumors, and how to potentially use
movement, protease expression and secretion, protein degrada- these defects for therapeutic advantage.
tion, and cell division. Although appropriate balance of these
signals leads to cells that function in a normal ordered envi-
ronment, dysregulation of these signals can lead to degradation MOLECULAR BIOLOGY
of extracellular proteins, movement of cells, growth of cells in
the absence of cell contact, recruitment of blood vessels, and a All cellular functions depend on the coordinated activity of
variety of effects that are critical to the tumor phenotype. The molecules. Molecules can be very small and detectable only
disruption of normal cell-to-cell contact is therefore critical to through experimental molecular biology techniques, whereas
the formation of tumors that must not only divide but also move others are macromolecules that are visible in the light micro-
through or displace existing cells in the environment. Cell-to- scope. It is the nature and activity of these molecules that
cell communication therefore represents an important control together constitute the primary focus of molecular biology.
mechanism that is lost in tumorigenesis. The flow of information between molecules is perhaps the
At the end of the life span of the cell, a variety of processes most well-studied aspect of molecular biology. The central
allow for cell removal and eventual replacement. The cell death dogma of molecular biology is that information is encoded in
process can be subdivided into necrotic cell death and apopto- DNA, transcribed to RNA, and then translated into protein. The
sis.10 Necrotic cell death represents the most destructive form proteins carry out the vast majority of molecular activities.
of cell elimination. In this process, cellular insults including but Although this pattern is generally true, RNA itself can also
10 S E C T I O N A Basic Science

provide important structural or catalytic activities. Proteins and RNA copy through the exons and introns to the 3¢ end of the
RNA in turn regulate and maintain the information encoded in gene to produce a relatively large heteronuclear RNA
DNA, completing a continual and circular flow of information. (hnRNA).13 The elongation process can also be negatively reg-
Protein–protein interactions are essential to intrinsic cellular ulated by molecules such as the Von Hippel Lindau (VHL)
activities such as regulating the cell cycle. Proteins, lipids, and tumor suppressor gene. The introns are spliced out of the mole-
inorganic ions also mediate the cell’s response to extracellular cule to generate the mature messenger RNA (mRNA). Identical
signals such as growth factors or cell–cell interactions through hnRNA molecules from a gene may be spliced differently to
a multistep process called signal transduction. The cell’s produce different mRNAs (i.e., differential splicing) and ulti-
response to exogenous agents such as chemotherapy also mately a greater diversity of proteins. Similarly to DNA, RNA
involves transduction of information through proteins and the also contains four bases, adenine, guanine, cytosine, and uracil
cell signaling machinery, as well as through the induced DNA (U) (rather than the thymine in DNA). RNA does not form a
damage itself. Regulation and fine tuning of the information double helix like DNA but is composed of a single strand that
flow also comes in the form of normal postsynthesis modifica- adopts secondary structures, including regions of double-strand
tions of DNA, RNA, and proteins. character. More than 95% of the RNA in a cell is ribosomal
The starting point in the flow of information is the DNA RNA (rRNA) and transfer RNA (tRNA), whereas only a few
molecule. DNA is the heritable material of genes and chromo- percent are mRNA. In addition to the protein-coding portions,
somes that stores, or encodes, the information for creating pro- mRNAs contain 5¢ and 3¢ regulatory elements. The protein-
teins. The building blocks of DNA are nucleotides, which coding portion starts with the ATG codon, which encodes for a
contain one of four bases, either adenine (A), guanine (G), cyto- methionine in the corresponding protein. The protein-coding
sine (C), or thymine (T). A phosphodiester bond connects adja- portion, also called the open reading frame, is preceded by a 5¢
cent nucleotides asymmetrically, giving the DNA strand a 5¢ to untranslated region (5’ UTR) and flanked on the 3¢ end with a 3¢
3¢ directionality. The two strands are then held together by spe- UTR and a poly(adenylic acid) (polyA) tail. The very abundant
cific base pairing along the strands, A with T and C with G. The rRNA, in contrast, does not code for proteins but instead serves
combination of directionality and base pairing is the basis of as a component of ribosomes, the protein and RNA complex
the complementary, antiparallel nature of the two strands in a where proteins are synthesized. tRNA also participates in
DNA double helix. This structure, seen in each chromosome, protein synthesis by acting as an adapter to decode the mRNA
is essential for faithful DNA replication during mitosis, DNA through an anticodon and paired amino acid. Other noncoding
repair, and heredity. RNAs serve unique functions in the cell. For example, the RNA
The human genome consists of 23 different chromosomes produced from the X-inactivation-specific transcript (XIST)
and is approximately 3 ¥ 109 base pairs (bp) long.6,15 In addi- gene binds directly to one of the two X chromosomes in female
tion to this nuclear DNA, there are also 16,571 bp of DNA in cells to help initiate X inactivation.8 RNA produced from an
each mitochondrion. More than 50% of the human genome is RNA-component-of-telomerase gene helps in the synthesis of
composed of repetitive elements, short sequences derived from telomeres, which are located at the ends of chromosomes.12
formerly mobile genetic elements (transposons) that have been There is also evidence that very short (21 to 23 bp) double-
incorporated over millions of years, though their current func- strand RNAs, termed small inhibitory RNAs (siRNAs), are
tion is unknown. In contrast to the repeat sequences, genes produced in mammalian cells and appear to have a role in reg-
serve as the blueprint for RNA and proteins. The number of ulating gene expression. siRNAs are also a very useful experi-
genes in the human genome is estimated at 30,000 to 40,000, mental tool in molecular biology to selectively inhibit the
only two times more than some worms and three times more expression of particular genes.9 Thus DNA is transcribed into
than the fruit fly. The parts of genes that code for proteins, RNAs that either can be the guide for protein production or may
called exons, constitute less than 5% of the genome. Each gene feed back directly to maintain and regulate the DNA.
has at least one promoter, which is a regulatory DNA sequence Following transcription, information flows from the
located at the start (5¢ end) of the gene, and an average of nine mRNA to proteins through the process of translation. Transla-
exons separated by contiguous noncoding DNA called introns. tion is the production of proteins from amino acids, using the
Fifty to sixty percent of all promoters are contained within mRNA as a template and the tRNA as an amino acid donor.
DNA regions termed CpG islands.3 CpG islands are typically Translation follows the genetic code, a set of rules by which
1000 bp and are unusual in their base content: They have a each specific set of three nucleotides (i.e., a codon) in the
higher G and C content and approximately five times more CpG mRNA is translated into a particular amino acid in the protein
dinucleotides than the rest of the genome. A proportion of the being synthesized. The carboxyl terminus of one amino acid is
promoters within CpG islands appears to function bidirection- then linked to the amino terminal of the next amino acid
ally, regulating two genes that are in a head-to-head configura- through a peptide bond. After translation, a protein may also be
tion.1 Other gene regulatory elements in DNA include proteolytically processed to remove short pieces on either end,
enhancers and repressors, which positively and negatively reg- allowing the protein to adopt an active conformation. It is the
ulate transcription, respectively. These regulatory elements can precise sequence of amino acids in a protein that determines its
be nearby or distant from the gene they influence, and their final conformation. Many smaller, newly synthesized proteins
activity can be further influenced by intervening DNA undergo spontaneous folding into a three-dimensional structure
sequences termed insulators. (i.e., the tertiary structure), whereas larger proteins may require
Information flows from the DNA to the RNA by the process the assistance of additional proteins called chaperones (e.g.,
of transcription, the creation of an RNA copy of a gene. RNA Hsp70) to fold correctly. The tertiary structure imparts the
polymerases and transcription factors initiate transcription unique function to the protein. Proteins have one or more
within the promoter, and additional proteins such as positive domains that perform specific functions, such as a helix-loop-
transcription elongation factor b (P-TEFb) help elongate the helix domain that binds DNA, or a protein kinase domain that
 C H A P T E R 2 Molecular and Cell Biology 11

catalyzes the phosphorylation of other proteins. Some proteins of the information flow and is essential for controlled molecu-
function as a monomer, whereas others are assembled into mul- lar signaling and a finite molecular response to changing
timers such as dimers (e.g., platelet-derived growth factor, or cellular conditions.
PDGF), trimers (e.g., collagen), or tetramers (e.g., p53 tumor The coordinated modification of proteins and protein–
suppressor). Proteins can be subdivided by their cellular loca- protein interactions described previously is the primary driving
tion, being extracellular (e.g., soluble vascular endothelial force for the process of signal transduction. Unlike steroid hor-
growth factor, or VEGF), intracellular (e.g., p53 or cyclin- mones, many signals from outside the cell are not able to enter
dependent kinases), or transmembrane (e.g., epidermal growth- the cell and therefore require receptors on the cell surface to
factor receptor, or EGFR). Nuclear proteins such as begin the signal transduction into the cytoplasm and beyond.
transcription factors regulate transcription of specific genes by These receptors are proteins that can be classified according to
binding to the DNA, thereby completing the feedback loop of the manner in which they transduce the signal. G–protein-
information from DNA to RNA to protein and back to DNA. coupled receptors activate intracellular G-proteins, which then
In addition to this central flow of information, the activity bind guanosine triphosphate (GTP) and control downstream
of DNA, RNA, and protein can be regulated and fine tuned biochemical reactions using the energy release as GTP is con-
through endogenous modifications. Nearly all nucleated cells verted to guanosine diphosphate (GDP). Tryosine kinase recep-
in the human body contain exactly the same sequence of DNA, tors, such as EGFR, have intrinsic tyrosine kinase activity that
the same genes, and the same chromosomes. Clearly, additional can both autophosphorylate and add phospho groups to tyrosines
mechanisms are required to determine which subset of genes is on other intracellular proteins. Tryosine-kinase-associated
active in a given cell, although these are poorly understood. In receptors do not have intrinsic kinase activity but are associated
humans, the only known covalent modification of DNA is with kinases (e.g., JAK, or Janus kinase) that then phosphory-
methylation. DNA methylation involves the addition of a late downstream proteins (e.g., JAK phosphorylates signal
methyl group to cytosine within a CpG dinucleotide, catalyzed transducer and activator of transcription [STAT], which then
by proteins called DNA methyltransferases.2,5 The pattern of moves into the nucleus to regulate genes). Other transmembrane
methylation is critical for maintaining a properly functioning receptors include serine-threonine-kinase receptors and ion
genome. Methylation stabilizes chromosomes particularly in channel receptors that control the movement of ions and small
the centromeres, constrains the activity of parasitic DNA molecules through the cell membrane. Following the initiation
elements such as transposons, and maintains patterns of gene of signal transduction, the signal can reach the nucleus and
expression. On a larger scale, proper methylation is necessary genes directly, as in the case of phosphorylated STAT, or may
for normal brain development and function. In cancers, the involve a signaling cascade such as occurs with the mitogen-
pattern and level of DNA methylation is compromised, and this activated protein kinases (MAPKs). The activation of Ras, an
contributes to the overall instability of chromosomes and to important intermediate molecule in signal transduction pathways,
aberrant loss of gene expression. Similar to DNA methylation, also leads to a cascade of phosphorylation and second messenger
RNA can also be modified by methylation, typically on adenine activation and ties into other pathways such as that of MAPK.
residues. This modification may alter the properties of the Inappropriate activation of the Ras pathway underlies the exces-
RNA, although this is less well understood. sive growth of many human cancers, including brain tumors.
In contrast to DNA and RNA, the variety of posttransla- In addition to proteins, signal transduction can be more
tional protein modifications is much more extensive. Protein indirect by using second messengers such as adenosine 3´,5´-
activity can be modified by the addition of chemical groups in cyclic monophosphate (cAMP), ions such as calcium, or lipid
the processes of acetylation and methylation (e.g., on histones), components of the intracellular membrane such as phos-
phosphorylation or dephosphorylation (e.g., on retinoblastoma phatidylinositol-4,5-bisphosphate (PIP2). Upon activation of a
[RB], the cell-cycle regulator and tumor suppressor), hydroxy- phospholipase, PIP2 can be cleaved to inositol-1,4,5-trisphos-
lation (e.g., on collagen), and N-formylation (e.g., on melittin). phate (IP3), a calcium channel opener, and another second
In addition, many membrane and secreted proteins can be mod- messenger 1,2-diacylglycerol (DAG) that initiates additional
ified by the addition of sugar side chains (glycosylation). Lipids signaling cascades. Signal transduction often involves both the
can also be added to proteins (acylation, on many membrane protein and these nonprotein molecules.
proteins, and myristolylation, on some protein kinases). Each Protein–protein signaling and phosphoregulation of protein
modification occurs specifically on one or a few specific types activity also lie at the heart of the cell cycle.14 As stated earlier,
of amino acids. One type of modification can have completely the cell cycle has distinct phases, including M, G1, S, and G2.
opposite effects on the activity of a protein depending on where Transition from one phase to the next is tightly controlled at
in the protein the modification occurs. The function of histone cell-cycle checkpoints (G1 to S and G2 to M) that are regulated
proteins, the main protein component of chromosomes, is reg- by the activity state of cell-cycle proteins. Center stage in regu-
ulated by a variety of these modifications, including acetyla- lation of the cell cycle is the RB protein. In the hypophospho-
tion, methylation, and phosphorylation, the sum total of which rylated state, RB binds to the E2F family of transcription factors
is called the histone code. The histone code may participate and prevents them from activating genes required for the next
with various transcription factors in the differential regulation phase, S, thereby maintaining the cell in G1. Cell-cycle pro-
of genes. The activity of proteins is also regulated by their gression requires a cyclin (e.g., cyclin D) and a cyclin-depend-
cellular localization.11 For example, in response to a steroid ent kinase (CDK) (e.g., CDK4 and CDK6) to form a complex
hormone entering the cell, the hormone binds its receptor in the and phosphorylate RB. Further regulation is provided by
cytoplasm, and the complex is actively transported into the inhibitors of CDKs, such as CDK inhibitor 2A (CDKN2A, also
nucleus to allow the receptor to bind to DNA and regulate gene called p16), which blocks the phosphorylation of RB by CDK-
expression. Finally, proteins can be targeted for degradation by cyclin complexes. Similarly, the transition between the G2 and
the addition of ubiquitin. Ubiquitination represents an end stage M phases is regulated by the cyclin B–CDC2 complex, which is
12 S E C T I O N A Basic Science

in turn regulated by removal of a phosphate on CDC2 by a phos- example, the most common abnormalities are in the RB and p53
phatase, CDC25C. Exit from mitosis involves a ubiquitin-medi- pathways, including overproduction of CDK4 or CDK6 by gene
ated degradation of cyclin B and inactivation of CDC2. The amplification, loss of the inhibitor p16 by homozygous deletion,
molecular guardians of the checkpoints restrain the cell cycle in or mutation and deletion of the p53 gene.
response to stress to the cell and to the DNA. For example, in Understanding the exact molecular pathways of signal
response to DNA damage, the ataxia telangiectasia mutated transduction and the elaborate molecular control of cellular
(ATM) and Rad3-related (ATR) proteins induce cell-cycle arrest processes such as the cell cycle are important goals of molec-
in G2 by inhibiting CDC2. This inhibition is mediated in part by ular biology. Because it is the imbalance of these mechanisms
the tumor suppressor p53. The parallels between normal cell- that underlie tumorigenesis, elucidation of the relevant mol-
cycle regulation and the dysregulation seen in most human ecules and mechanisms is essential to the development of tar-
cancers are remarkable. In brain tumors of glial origin, for geted anticancer therapies.

References

1. Adachi N, Lieber MR: Bidirectional gene organization: a common 8. Ogawa Y, Lee JT: Xite, X-inactivation intergenic transcription ele-
architectural feature of the human genome. Cell 109:807–809, ments that regulate the probability of choice. Mol Cell 11:731–
2002. 743, 2003.
2. Baylin S, Bestor TH: Altered methylation patterns in cancer 9. Paddison PJ, Hannon GJ: RNA interference: the new somatic cell
cell genomes: cause or consequence? Cancer Cell 1:299–305, genetics? Cancer Cell 2:17–23, 2002.
2002. 10. Proskuryakov SY, Gabai VL, Konoplyannikov AG: Necrosis is an
3. Bird A, Taggart M, Frommer M, et al: A fraction of the mouse active and controlled form of programmed cell death. Biochem-
genome that is derived from islands of nonmethylated, CpG-rich istry (Mosc) 67:387–408, 2002.
DNA. Cell 40:91–99, 1985. 11. Schwoebel ED, Moore MS: The control of gene expression by
4. Hood JD, Cheresh DA: Role of integrins in cell invasion and regulated nuclear transport. Essays Biochem 36:105–113, 2000.
migration. Nat Rev Cancer 2:91–100, 2002. 12. Shay JW, Wright WE: Telomerase: a target for cancer therapeu-
5. Jones PA, Baylin SB: The fundamental role of epigenetic events tics. Cancer Cell 2:257–265, 2002.
in cancer. Nature Reviews Genetics 3:415–428, 2002. 13. Shilatifard A, Conaway RC, Conaway JW: The RNA polymerase
6. Lander ES, Linton LM, Birren B, et al: Initial sequencing and II elongation complex. Annu Rev Biochem 72:693–715, 2003.
analysis of the human genome. Nature 409:860–921, 2001. 14. Stewart ZA, Westfall MD, Pietenpol JA: Cell-cycle dysregulation
7. Lu H, Forbes RA, Verma A: Hypoxia-inducible factor 1 activa- and anticancer therapy. Trends Pharmacol Sci 24:139–145, 2003.
tion by aerobic glycolysis implicates the Warburg effect in car- 15. Venter JC, Adams MD, Myers EW, et al: The sequence of the
cinogenesis. J Biol Chem 277:23111–23115, 2002. human genome. Science 291:1289, 1304–1351, 2001.
 A
3

S e c t i o n
Chapter

ANIMAL MODELS Dominique B. Hoelzinger and Michael E. Berens

Animal models are tools used to study various aspects of TRANSPLANTABLE TUMORS IN
cancer. The complexity of the disease leads to the adoption of
a variety of models that, to varying degrees, embody oncogenic ANIMAL MODELS
transformation events, tumor response to therapy, strategies
for preventing cancer, testing mechanisms for novel agent The most demanding application for animal models of tumors
delivery, and reporter systems of drug metabolism by tumors. is the testing and development of new therapies. Because these
Sporadic cancers in humans and in animals show substantial therapies will be applied in human treatments, the growth of
variation in histopathology, pace of progression, and associated human tumors in animal systems should approximate a testing
molecular genetic events. Disparate genetic backgrounds, as environment close to the eventual clinical response of tumor
well as likely multiplicity in etiologic paths for cancer induc- cells. However, growing understanding of the role of the host
tion, ensure great heterogeneity in cancer presentation. In the environment on cancer cell behavior leads to the caveat
midst of this variability, development of reproducible animal that human tumors growing in nonhuman hosts may introduce
models affords schemes for testing hypotheses that would be artificial and uncontrollable changes in cell behavior and
untenable in spontaneous tumors. response.8,21 Techniques for inducing such tumors might also
Animal models have utility in developing an understand- engage molecular changes unrepresented by spontaneous tumors.
ing of the underlying genetic events of oncogenesis. Once
tumors arise in controlled animal populations, they can be
studied to learn how cancer progresses from local to dissemi- Brain Tumor Allografts in Mice
nated disease, from low-grade to high-grade patterns of behav- A variety of induced brain tumors in mice has been developed
ior. Such observations assist in the development of markers of that can be serially transplanted into cohorts of test animals of
tumor progression. Ideally, changes in tumor growth in animals the same strain of mice. Such tumors grow reproducibly and
consequent to therapy are indications of similar responses in are nonimmunogenic. These models have been used as tools
human tumors. In this regard, animal models provide systems for exploring tumor-host relationships that include immune
to test treatment response in the development of new interven- cell modulation for therapeutic application. These models are
tions. The complexity of tumor physiology, which includes not readily identified by searching the literature for the tumor
only the self-renewing tumor stem cells but also neovascular- names: GL261, CT-2A, VM/Dk, SR-B10.A, and EPEN. Some
ization, stromal tissue encounters, drug metabolism and distri- of these models have existed since the 1980s but have not
bution, and immune cell interactions with the neoplastic mass, enjoyed as widespread an application as have rat brain tumors.
is also open to investigation using animal models.
Whereas there is no one ideal animal model of cancer that
embodies all desired features, certain available methods work Advantages
within critical research constraints. These restraints include a Mice are inexpensive. The breeding paradigms of mice have
model that is highly reproducible, is affordable, shows fast developed an exquisite immunologic perspective that is unri-
tumor growth, shows features homologous to the human valed in any other species of research animal. Tools for evalu-
disease, and is amenable to scale-up so that nuances of research ating tumor-host relationships are likely to have the fastest
queries can be addressed. Different model systems manifest maturation using the inventory of information on mice.
alignment with these constraints to varying degrees. Selection
of the animal model to test a hypothesis should be determined
by the degree to which features of the model best adhere to the Disadvantages
research paradigm. In that sense, the abundance of different The mouse models of brain tumors have gained only a limited
model systems represents a toolbox with which to advance the amount of attention, largely because of the difficulties in ortho-
discovery, development, and delivery of new approaches to and topic placement of tumors. In addition, these tumors have not yet
treatments for cancer. received the degree of characterization that rat tumor models have.
13
14 S E C T I O N A Basic Science

Brain Tumor Allografts in Rats that of humans. Optimizing tumor response to novel delivery
approaches, such as convection-based delivery systems, ideally
Chemical treatment of pregnant rats with carcinogenic agents would be conducted in models of anatomic scale much larger
such as ethylnitrosourea leads to a heightened incidence of than a rodent.
brain tumors in litters.7 Consequently, induction of brain tumors
in inbred rats has received considerable attention and effort. A
panel of serially transplantable, syngeneic rat brain tumors Nude and Severe Combined
exists. Their use is traceable through searching the literature for Immunodeficient Mice
the names of the tumor models: C6, 9L, RG2, T9, BT4C, and
CNS-1. Their popularity in the field has led to the assembly of Selection of the host species in which to propagate a human
a substantial body of literature characterizing these tumors for tumor is a pragmatic decision driven by cost considerations.
a wide variety of features such as growth properties, histology, Their low cost is the reason that nude and severe combined
angiogenesis, molecular genetics, and utility as markers of immunodeficient (SCID) mice are the most used. Many com-
chemo- and radiation-responsiveness. The thousands of pub- mercial vendors provide well-characterized strains of im-
lished articles on these tumors attest to their wide acceptance munodeficient and immunocompromised animals suitable for
and utility. xenotransplantation of human tumors.

Advantages Nude and Severe Combined

Immunodeficient Rats
Rat brain tumors are widely accepted in the biomedi-
cal research community. Their wide use affords a common In instances where a larger host is desirable, nude and SCID
vocabulary. rats have been developed. These are models readily adopted by
vivaria in most academic medical centers and in biotechnology
and pharmaceutical companies. For reasons of scale, xenografts
Disadvantages in rats are useful for stereotactic implants into the natural site
To varying degrees and with limited exceptions, syngeneic rat of the tumor’s origin (orthotopic), which in the case of brain
brain tumors show only modest invasiveness in the brain and cancers is of considerable value.
do not always recapitulate the histopathology of human
primary gliomas. Subcutaneous Implants
Expanding masses of tumor cells grown in the subcutaneous
Xenografts space of mice have provided interpretations of response to
treatments. Such tumors are typically well vascularized, highly
The central challenge facing researchers who propagate human
proliferative, and amenable to caliper measurements of tumor
cells in nonhuman hosts is that of immune rejection of the
volume. In this regard, subcutaneous tumors in mice operate as
xenograft. Chemical or radiologic immunosuppression can
a fairly high throughput screening system in the preclinical
establish a host that is unable to reject a transplanted tumor;
assessment of tumor response to therapy. Implantable chip
these approaches have been employed in larger animals. Selec-
transponders for animal identification become tools that con-
tive breeding of rodents has generated strains of mice and rats
veniently allow controlled, blinded studies that ensure optimal
deficient in either T cells or additional major effectors of the
execution of the investigation.
immune system. These hosts necessitate husbandry in sterile
A recent appreciation for differences in vascular endothe-
environments, with extensive attention to preservation of
lial cells by tissue or organ system has highlighted the need
barrier protection. By definition, studies of human xenografts
to approach transplantable xenografts using orthotopic sites
in immune-deficient animals are unable to advance an under-
of injection. Similarly, stromal cell responses to a neoplastic
standing of the complex interactions between tumors and the
growth show variation by organ system, and a more keen
host’s immune system.
appreciation for differences in the parenchymal cells residing
Propagation of a human tumor in immunocompromised
in different tissues indicate a need for careful consideration of
animals has as its starting point the collection of a biopsy from
sites of implantation for xenografts into locations where the
a spontaneous growth. Not all resected masses demonstrate an
tumor would have naturally arisen.
ability to grow in an immune-deficient animal. Because tumor
transplantation techniques are more successful when highly
malignant cells are used as the innoculae, most xenograft Intracranial Implants, Stereotaxy in
studies of human tumors are uninformative for investigations
of tumor progression; the transplanted cells are already highly
Small Dimensions
malignant. Successful development of serially transplanted Intentionality for anatomic and tissue representation of the
human xenografts may also bias the model in uncontrolled tumor’s environment has fostered use of stereotactic deposition
ways because of the in vivo selection of cells most capable of of tumors in the brains of experimental model hosts. A common
surviving in the xenogenic host.17 deficiency of orthotopic xenograft models of brain tumors is
The range of species into which human tumors have been the absence of glial tumor invasion, which is a central element
transplanted includes mice, rats, and cats. Rodent models are to the behavior of such cancers in humans. It may be that
readily amenable to use in most research institutes’ vivaria; cats species differences in the cell–cell interactions or the cell-
necessitate a more sizable commitment to infrastructure. Cats matrix relationships that exist between rodents and humans
adapt the xenograft model to a system closer in dimension to are sufficiently large that the invasive phenotype of human
 C H A P T E R 3 Animal Models 15

xenografts is poorly supported in these hosts. Recent modifi- integration of the DNA construct could generate secondary
cations in the preimplant handling of human tumors before genetic modifications, making interpretation of results more
intracranial implants, coupled with a protracted interval for the difficult. Position effect variegation can cause varying degrees
emergence of intracranial tumors in athymic rats, yields a of penetrance, requiring the creation of various genetically
model more complementary to the human disease that also engineered mice (GEM) to verify the levels of reporter gene or
affords glioma cell invasion.2,13 transgene expression. Finally, germline manipulation does not
address the interplay between multiple genetic changes that
results in true tumor biology.
GENETICALLY ENGINEERED MOUSE
MODELS OF ORTHOTOPIC Somatic Cell Manipulation In Situ
TUMORIGENESIS Intracranial delivery of a virus carrying platelet-derived growth
factor B (PDGF-B), driven by a viral promoter,18 induces for-
Conditional control of gene expression that leads to cellular mation of histologically varied brain tumors. The advantages
transformation may more closely recapitulate the evolution of of this technique are the localized effect of PDGF-B overex-
genetic changes found in cancers as they progress through the pression and the temporal control of the gene activation. Retro-
stages of malignancy. Genetically engineered mouse models of viral gene delivery through the replication-competent avian
orthotopic tumorigenesis provide a platform for learning more leukosis virus (ALV) splice (RCAS) system ensures specific
about the early changes in a cell or the sequence of changes gene activation in a timed and tissue-specific manner. It
that lead to more malignant phenotypes.10 A sizable variety of requires GEM expressing the RCAS receptor tv-a from a tissue-
mouse models of human cancers has been generated, and the specific promoter (TSP) and the RCAS acceptor with the gene
descriptive characterization of these tumors is maturing. The of interest. The viral vector is a retrovirus that infects only spe-
National Institutes of Health National Cancer Institute hosts a cific cell types at one time: those expressing the ALV receptor
Web site for tracking the status of various iterations of these tv-a, which is not encoded in the mammalian genome. In
engineered models.14 gliomas two such mouse lines have been developed, one
expressing tv-a from the GFAP promoter4 and one expressing
Germline Manipulation it from the nestin promoter.3 The viral particles are delivered
by local injection of avian producer cells. Only a few hundred
Models employing germline manipulations are those wherein cells are infected, which reduces the probability of the occur-
the chromosomal DNA of every cell of the mouse carries rence of secondary genetic alterations (through insertional
an engineered construct. Transgenes are introduced into the mutagenesis).
genome by microinjection of fertilized eggs or embryonic stem
cells. At its most basic it entails tissue-specific expression of
an oncogene or mutated oncogene believed to be involved Advantages
in tumor initiation. The expression of an oncogene (v-src) is Somatic cell manipulation combines several conditionally
driven through the astrocyte-specific glial fibrillary acidic expressed genes or specific mutations at the same time in the
protein (GFAP) promoter,23 which results in the appearance of same tissue to more accurately reflect the tumor’s genetic or
astrocytomas but also of tumors of different histologic lineage genomic make-up or sequential genetic changes during tumor
such as schwannomas. Verification of transgene expression is progression. The approach allows targeted-tissue delivery at
obtained by immunohistochemical staining for v-src and GFAP. anatomically determined sites at designated times. Titration of
Easy visualization of transgene expression and concomitant virus to low titers results in a limited number of tumors being
confirmation of appropriate histologic expression is achievable formed, which allows monoclonality of the mass.
by using a bicistronic vector in which the reporter gene lacZ is
expressed in tandem with V12Ha-ras but translated separately.1
The addition of a reporter gene also allows the evaluation of Disadvantages
the number of transgene integrations into the genome by meas- The viruses have a small carrying capacity of 2.5 kb or less. To
uring the relative level of reporter gene expression, illustrating the extent that there may be leakiness of the ALV expression,
the importance of gene dosage in tumor formation and this approach may lack cell-type specificity in the infection, and
progression. thus induction of gene expression may leak across different cell
lineages.
Advantages
This system allows for a targeted re-creation of the activation Conditional Knock-Outs with the
of specific signaling pathways involved in the genesis of
specific tumors. Cre-lox System
The Cre-lox approach enables timed, precise somatic inactiva-
tion of tumor suppressor genes. Two engineered mouse strains
Disadvantages are required for conditional gene deletion: first, a conventional
Introduction of genes into the germline affords no temporal reg- transgenic mouse line with Cre driven by a specific tissue or
ulation of expression; the engineered gene is turned on through- cell-type promoter, such as the whey acidic protein (WAP)
out the range of biologic activity of the promoter. This could employed in a breast cancer model,24 and second, a mouse
potentially result in embryonal lethality. Multisite, multicopy strain that embodies a target gene (BRCA1) flanked by two
16 S E C T I O N A Basic Science

loxP sites in a direct orientation (floxed gene). Recombination of Dox from each tissue (1 to 7 days); this could result in slow
(excision and consequent inactivation of the target gene) occurs activation of the rtTA system and slow inactivation of the tTA
only in those cells expressing Cre recombinase. Hence, the system.
target gene remains active in all cells and tissues that do not
express Cre. The same system can be used to create conditional
activation (knock-in) of an oncogene. In a lung cancer model,5 Combination Approaches
a mouse strain with an activated K-ras G12D mutation pre- Temporal fine tuning of Cre expression or activity results in
ceded by a stop codon in the 5¢ untranslated region (UTR) the formation of one generation of tumors, which might more
flanked by two loxP sites has been created. The oncogene closely resemble individual cancerous lesions. Two approaches
remains untranscribed until the tissue-specific removal of the have been developed to control the transcription or the activity
lox-stop-lox sequences is activated. The Cre recombinase of Cre.12 Transcriptional control of Cre is achieved by using the
is delivered by adenoviral (AdenoCre) infection of the lung tet inducible system, such that Cre is produced only in the
epithelium. This mimics the early events in the genesis of lung desired tissue during a timed administration of Dox. Functional
tumors. Other knock-in models require the crossing of the control of Cre activity depends on a Cre variant that functions
oncogene harboring GEM with a GEM carrying the Cre gene only in the presence of an exogenous inducer such as RU486
driven from a TSP. or tamoxifen. The Cre variant is continuously present in the
desired tissue but can localize to the nucleus only in the pres-
Advantages ence of the inducer.
The utility of this approach has led to a growing palette of
tissue-specific Cre GEM lines. This affords an inventory for Advantages
manipulation of multiple genes in the same cell, including The combined approach provides very sophisticated tissue and
turning some on and others off. The timing of the viral activa- temporal control of the expression of genes of interest.
tion circumvents embryonic lethality.
Disadvantages
Disadvantages
The level of complexity within each approach is compounded.
Recombination is an irreversible genetic change; the targeted
cell and all its potential progeny will manifest the same expres-
sion profile for the manipulated gene and the pathways it VALIDATION OF THE TUMOR MODELS
affects. There remains an inability to obtain exact temporal
control over promoter activation. With increasing characterization of human cancers using
genomic and proteomic tools, the investigator has better plat-
Controlled Manipulation of Gene Expression forms with which to defend the alignment of the model systems
with the spontaneous disease. It should be anticipated that
Fine-tuned temporal control over gene expression is the hall- in the near future there will be extensive verification of the
mark of inducible systems such as the tetracycline-dependent model tumor’s histology, gene expression profile, and therapy
regulatory system. This binary transgenic approach requires a response.
GEM with the TSP driving transcription of the tetracycline-
controlled transactivator (tTA) of transcription, as well as car-
rying the tetracycline operon (tetO) fused to the open reading ASSESSMENT OF BRAIN TUMOR
frame (ORF) of choice. tTA is produced in the specified cell
type; systemic treatment with doxocyline (Dox) prevents tTA BEHAVIOR IN ANIMAL MODELS
from binding to tetO sequences to induce transcription. The
ORF is transcribed when the cell is depleted of Dox. Reverse The ultimate outcome for successful modeling of tumors in
tTA (rtTA) is another, related system in which the administra- animals is an end point for illumination of tumor biology or
tion of Dox is needed for transcriptional activation.6 response to therapy. These all require measurements, some of
which necessitate the termination of the test subject. Where pos-
sible and appropriate, nonlethal end points allow serial studies
Advantages of tumor growth, which allows a reduction in the number of
Because of the chemical control afforded in this system, gene animals needed for conclusive experiments. This is always
expression is temporally controlled. Thus it is possible to stop a high priority. Statistical programs by which to evaluate the
the transgene’s expression and study factors needed not only number of animals needed to generate results of sufficient power
for tumor induction but also for the maintenance of the tumor to draw statistical inference of the outcome of the study should
phenotype. Multiple genes can be turned on simultaneously, be accessed. A free-ware program that has proved instructive can
including reporter genes like luciferace.22 be accessed at http://calculators.stat.ucla.edu/powercalc/.

Disadvantages Invasive/Terminal Techniques

This system requires both constructs to be stably integrated and The growth and invasion behavior of brain tumors in animals
expressed to be tissue specific. The model may manifest leak- can be assessed using histologic techniques. These require that
iness. Once administered, there is a time lag for the clearance the animal be euthanized; such end points necessitate pilot
 C H A P T E R 3 Animal Models 17

studies of the growth properties of the tumor to time the col- be beyond the reach of many facilities. Some recent develop-
lection of the brain tissue. For xenograft studies, discrimination ments in photon imaging have introduced lower-cost options
of the human cells from host cells can be readily accomplished for animal studies.
by using immunostaining of HLA markers. Other reporter
genes such as green fluorescent protein or one of its derivatives
establish a method by which to readily detect the cells of inter-
Magnetic Resonance Imaging
est under fluorescence microscopy. Linking reporter genes to Imaging technology for real-time assessment of tumor volume
the tumor cells allows technologies such as laser capture and invasion patterns in experimental animals has kept abreast
microdissection to be applied to tumors in animals for probing of model development. These techniques are crucial allies in
tumor-host interactions, tumor invasion mechanisms, or angio- the support to fully utilize the applications of such model
genic responses to tumors. systems. Magnetic resonance imaging (MRI) has been suc-
cessfully applied to brain tumor studies in mice9,20 and in rats.16
Physiology of Transplantable Tumors
One novel recent application to tumor transplantation is the Photon Imaging: Bioluminescence
installation of physiologic windows over the tumor for intra- Genetic engineering techniques allow the installation of
vital microscopy of the emerging tumor.15,19 These studies afford reporter genes into tumor cells that afford novel detection
exploration of tumor-host interactions, especially as related to strategies of biochemical events. One such application is the
the neoangiogenesis coincident to tumor expansion. The inter- expression of a modified form of the luciferase gene that is only
species variations notwithstanding, such applications afford a conditionally activated when cellular apoptosis is activated.11
research tool by which to assess unique but critical features of Use of luciferin as a substrate results in the elaboration of
the tumor-host relationship as the neoplastic mass expands. photons coincident to apoptosis (caspase-3-specific activation
of luciferase) in the cells. This model has been used as an in
situ reporter of the efficacy of cancer agents against brain
Noninvasive/Vital Techniques tumors in rats. Matching of conditionally activated enzymes
When possible, maintenance of tumor-bearing animals during that become functional under specific biochemical signaling
the end point analysis reduces the numbers of animals needed events will raise the utility of in vivo models for more accurate
for the experiments. These end points, however, are only col- assessment of cancer interventions or tumor biology
lected with technologically sophisticated instruments and may experiments.

References

1. Ding H, Roncari L, Shannon P, et al: Astrocyte-specific expres- 10. Lampson LA: New animal models to probe brain tumor biology,
sion of activated p21-ras results in malignant astrocytoma forma- therapy, and immunotherapy: advantages and remaining concerns.
tion in a transgenic mouse model of human gliomas. Cancer Res J Neurooncol 53:275–287, 2001.
61:3826–3836, 2001. 11. Laxman B, Hall DE, Bhojani MS, et al: Noninvasive real-time
2. Guillamo JS, Lisovoski F, Christov C, et al: Migration pathways imaging of apoptosis. Proc Natl Acad Sci USA 99:16551–16555,
of human glioblastoma cells xenografted into the immunosup- 2002.
pressed rat brain. J Neurooncol 52:205–215, 2001. 12. Lewandoski M: Conditional control of gene expression in the
3. Holland EC, Celestino J, Dai C, et al: Combined activation of Ras mouse. Nat Rev Genet 2:743–755, 2001.
and Akt in neural progenitors induces glioblastoma formation in 13. Mahesparan R, Read TA, Lund-Johansen M, et al: Expression of
mice. Nat Genet 25:55–57, 2000. extracellular matrix components in a highly infiltrative in vivo
4. Holland EC, Varmus HE: Basic fibroblast growth factor induces glioma model. Acta Neuropathol (Berl) 105:49–57, 2003.
cell migration and proliferation after glia-specific gene transfer in 14. National Cancer Institute. MMHCC. The Mouse Models of Human
mice. Proc Natl Acad Sci USA 95:1218–1223, 1998. Cancers Consortium. http://emice.nci.nih.gov/mouse_models
5. Jackson EL, Willis N, Mercer K, et al: Analysis of lung tumor ini- (last accessed 9 June 2003).
tiation and progression using conditional expression of oncogenic 15. Read TA, Farhadi M, Bjerkvig R, et al: Intravital microscopy
K-ras. Genes Dev 15:3243–3248, 2001. reveals novel antivascular and antitumor effects of endostatin
6. Jonkers J, Berns A: Conditional mouse models of sporadic cancer. delivered locally by alginate-encapsulated cells. Cancer Res
Nat Rev Cancer 2:251–265, 2002. 61:6830–6837, 2001.
7. Kindler-Rohrborn A, Koelsch BU, Buslei R, et al: Allele-specific 16. Ross BD, Chenevert TL, Rehemtulla A: Magnetic resonance
losses of heterozygosity on chromosomes 1 and 17 revealed by imaging in cancer research. Eur J Cancer 38:2147–2156, 2002.
whole genome scan of ethylnitrosourea-induced BDIX x BDIV 17. Schmidt EE, Ichimura K, Goike HM, et al: Mutational profile of
hybrid rat gliomas. Mol Carcinog 26:163–171, 1999. the PTEN gene in primary human astrocytic tumors and cultivated
8. Kjaergaard J, Tanaka J, Kim JA, et al: Therapeutic efficacy of xenografts. J Neuropathol Exp Neurol 58:1170–1183, 1999.
OX-40 receptor antibody depends on tumor immunogenicity 18. Uhrbom L, Hesselager G, Nister M, Westermark B: Induction of
and anatomic site of tumor growth. Cancer Res 60:5514–5521, brain tumors in mice using a recombinant platelet-derived growth
2000. factor B-chain retrovirus. Cancer Res 58:5275–5279, 1998.
9. Koutcher JA, Hu X, Xu S, et al: MRI of mouse models for gliomas 19. Vajkoczy P, Ullrich A, Menger MD: Intravital fluorescence
shows similarities to humans and can be used to identify mice for videomicroscopy to study tumor angiogenesis and microcircula-
preclinical trials. Neoplasia 4:480–485, 2002. tion. Neoplasia 2:53–61, 2000.
18 S E C T I O N A Basic Science

20. Verhoye M, van der Sanden BP, Rijken PF, et al: Assessment of 22. Vooijs M, Jonkers J, Lyons S, Berns A: Noninvasive imaging of
the neovascular permeability in glioma xenografts by dynamic spontaneous retinoblastoma pathway-dependent tumors in mice.
T(1)MRI with Gadomer-17. Magn Reson Med 47:305–313, Cancer Res 62:1862–1867, 2002.
2002. 23. Weissenberger J, Steinbach JP, Malin G, et al: Development and
21. Visted T, Thorsen J, Thorsen F, et al: lacZ-neoR transfected malignant progression of astrocytomas in GFAP-v-src transgenic
glioma cells in syngeneic rats: growth pattern and charac- mice. Oncogene 14:2005–2013, 1997.
terization of the host immune response against cells trans- 24. Xu X, Wagner KU, Larson D, et al: Conditional mutation of Brca1
planted inside and outside the CNS. Int J Cancer 85:228–235, in mammary epithelial cells results in blunted ductal morphogen-
2000. esis and tumour formation. Nat Genet 22:37–43, 1999.
S e c t i o n B Diagnosis
Section Editors
Mitchel S. Berger and Michael D. Prados

Chapter 4
DIAGNOSTIC IMAGING Soonmee Cha

Brain tumors rank second as the cause of cancer-related deaths in BRAIN TUMOR IMAGING
children and young adults younger than the age of 34, and they
affect adults of all ages. After cancer of the lung and pancreas, Imaging plays a critical role in the diagnosis of a brain tumor
primary malignant brain tumors have the third-highest cancer- and in patient management. Imaging provides a preliminary
related mortality rate in the United States, and they take a dis- assessment of tumor type and grade, which is important in
proportionate toll in disability and morbidity.8 The Surveillance, determining the patient’s prognosis. Imaging also provides
Epidemiology, and End Results (SEER) program estimated that anatomic detail about the tumor, such as its location and size,
there were 17,400 new cases of brain tumors in the United States and information about the surrounding brain, which directly
in 1998 and 13,300 deaths from brain and other central nervous influence surgical planning. Once the diagnosis of brain tumor
system tumors.11 There is now convincing evidence that the inci- is confirmed on histopathologic examination, imaging contin-
dence of primary brain tumors is increasing.5 Gliomas, the most ues to play an important role in assessing residual or recurrent
common type of primary brain tumor, are extremely heteroge- tumor and in monitoring the patient’s response to treatment.
neous neoplasms that have defied many decades of therapeutic The timing of brain tumor imaging falls into five broad
efforts, and they are one of the most difficult tumors to treat and categories: (1) at initial presentation, (2) before surgery, (3)
cure. Glioblastoma multiforme (GBM), the most common and immediately after surgery, (4) before irradiation or chemother-
malignant type of glioma, has a poor prognosis, with fewer than apy, and (5) after adjuvant therapy.
10% of patients remaining alive after 2 years, even with the most
aggressive combination of therapies. Even the less common
low-grade gliomas tend to dedifferentiate over time into more Initial Presentation: Differential Diagnosis of
malignant varieties and eventually result in death.2 Despite many
years of research, technologic progress, and clinical trials, the Brain Tumors
diagnosis and therapy of brain tumors continue to challenge the Patients who have a brain tumor are usually initially seen in an
medical and scientific community. acute setting, often an emergency room, because of an acute
The current standard for diagnosis and classification of episode of seizure or the subacute development of focal neu-
gliomas is histopathologic analysis of tissue specimens after rologic deficit, personality changes, or headaches. In this acute
imaging characterization and surgery. A histopathologic diag- environment, the first line of imaging is often a noncontrast
nosis of glioma relies on recognition of the tumor’s presumed computed tomography (CT) scan of the brain to exclude
cell of origin, and grading is based on a qualitative histopatho- hemorrhage or a large area of infarction. It is important to note
logic assessment of cellular proliferation, vascular hyperplasia, that less than 3% to 5% of all patients with acute stroke
necrosis, and nuclear atypia or pleomorphism.1,4 Preoperative have seizure as their primary symptom. Therefore, when an
imaging, however, plays several important roles in (1) charac- adult patient has a seizure, brain tumor and not stroke should
terizing the brain tumor and providing preliminary information be considered the most likely diagnosis.
on tumor type and grade; (2) establishing a differential diag- Once a mass lesion is suspected on noncontrast CT, con-
nosis; (3) providing sufficient anatomic detail for surgical plan- trast-enhanced CT or magnetic resonance imaging (MRI) should
ning; and (4) identifying tumor-related complications such as be performed to better characterize the mass. Its multiplanar
hemorrhage, hydrocephalus, herniation, and mass effect. capability and superior soft-tissue contrast make MRI, as
19
20 S E C T I O N B Diagnosis

opposed to CT, the preferred modality in evaluating patients age is an especially important clinical factor in narrowing the
with a brain tumor. The contrast agent gadolinium should be differential diagnosis of brain tumor because certain tumor
used routinely for MRI because it further enhances soft tissue types tend to occur more frequently in a particular age group.
contrast, improves delineation of the tumor, and provides infor- For instance, juvenile pilocytic astrocytoma (JPA) and GBM
mation about the integrity of the blood-brain barrier, detecting are common glial tumors distinctly associated with age. JPA is
areas of blood brain–barrier breakdown that allow leakage a grade I tumor with a relatively favorable prognosis that most
of contrast agent from the intravascular to the interstitial often occurs in the pediatric age group.7,13 In contrast, GBM is
compartment. a grade IV tumor with aggressive biologic activity that tends to
When a brain tumor is suspected, the differential diagno- affect mostly adults, especially those older than 65 years.
sis can be narrowed by incorporating clinical information with Imaging features further narrow the differential diagnosis.
imaging characteristics such as the tumor’s location and con- Tumor location is helpful in determining tumor type. Brain
trast-enhancement pattern (Tables 4-1 and 4-2). The patient’s tumors can be classified by location into (1) intra-axial (within

Ta b l e 4 - 1 Differential Diagnosis of Brain Tumors Based on

Tumor Location

Intra-axial Extra-axial Intraventricular Supratentorial

Glial, Astrocytic Dural
Low-grade fibrillary astrocytoma Meningioma Choroid plexus tumor
Anaplastic astrocytoma Hemangiopericytoma Neurocytoma
Glioblastoma multiforme Metastases Meningioma
Metastases

Glial, Nonastrocytic Pituitary

Oligodendroglioma Adenoma
Ganglioglioma
Dysembryoblastic
neuroepithelial tumor

Nonglial Pineal
Primary cerebral lymphoma Pineocytoma
Metastases Germ cell tumor
Pineoblastoma

Suprasellar
Craniopharyngioma
Germ cell tumor
Lymphoma
Metastases
Juvenile pilocytic
astrocytoma

Skull Base
Chordoma
Plasmacytoma
Metastases
Chondroid tumor

Infratentorial
Glial, Astrocytic Dural
Juvenile pilocytic Meningioma Ependymoma/subependymoma
astrocytoma Hemangiopericytoma Choroid plexus tumor
Astrocytoma (low-grade, Metastases
anaplastic, glioblastoma)

Nonglial Cerebellopontine Angle

Medulloblastoma Meningioma
Hemangioblastoma Schwannoma
Metastases Epidermoid
 C H A P T E R 4 Diagnostic Imaging 21

Ta b l e 4 - 2 Differential Diagnosis of Brain Tumors Based on

Imaging Characteristics

Imaging
Characteristic Appearance Tumor Type
Contrast Homogeneous Juvenile pilocytic astrocytoma
Enhancement Glioblastoma
Pattern Metastases

Heterogeneous, irregular Glioblastoma

Anaplastic astrocytoma
Oligodendroglioma
Metastases

Focal nodular Oligodendroglioma

Ganglioglioma

Absent Low-grade fibrillary astrocytoma

Gliomatosis cerebri
T2 Signal Low to isointense to gray Meningioma
matter Primary cerebral lymphoma
Ependymoma
Medulloblastoma
Pineoblastoma
Metastases (mucinous variant)

Hyperintense to gray matter Glial tumor

Cyst Juvenile pilocytic astrocytoma

Ganglioglioma
Oligodendroglioma
Hemangioblastoma
Pleomorphic xanthoastrocytoma

Hemorrhage Glioblastoma
Anaplastic astrocytoma
Ependymoma
Oligodendroglioma

Calcification Oligodendroglioma
Ependymoma

Necrosis Glioblastoma
Metastases

brain parenchyma) and extra-axial (outside brain parenchyma, planned, all patients should undergo preoperative MRI with
including the intraventricular, pituitary, pineal, and meningeal intravenous gadolinium administration. The important goals of
locations); and (2) supratentorial and infratentorial tumors. preoperative imaging are to (1) characterize tumor location,
Tumors within the cerebral white matter tend to be glial in size, and imaging patterns to suggest a preliminary assessment
origin, GBM being the most common. Tumors in the suprasel- of tumor type and grade; (2) provide sufficient anatomic detail
lar location include pituitary adenoma and craniopharyngioma, for surgical planning; and (3) evaluate any tumor-related com-
whereas tumors in the pineal region include those originating plications, such as hemorrhage, hydrocephalus, herniation, and
from primitive neuroectodermal cells, such as pineoblastoma mass effect.
or pineocytoma. In the posterior fossa, medulloblastoma and
JPA are the most common tumor types in children, and metasta-
tic cancer and glial tumors are the most common in adults. Imaging after Surgery
The goals of immediate postoperative MRI are to assess the
amount of any residual tumor and to evaluate postoperative
Imaging before Surgery complications, such as hemorrhage, infarction, contusion, or
When a brain tumor is suspected based on initial imaging subdural collection of blood. The timing of immediate post-
studies, most patients are then evaluated by a neurosurgeon for operative MRI is crucial to differentiate residual tumor and the
surgical resection and tissue diagnosis. Once surgery is reaction of brain tissue to surgery. Because postoperative gran-
22 S E C T I O N B Diagnosis

ulation tissue, which can appear identical to residual tumor, is approximately three quarters of all primary glial neoplasms.
most conspicuous several days after surgery, immediate post- Astrocytomas are an extremely heterogeneous group of tumors,
operative imaging should be done within 24 to 48 hours. both histopathologically and prognostically. The grading of
Most patients with high-grade brain tumors undergo radi- astrocytomas, which is important for prognosis and therapy,
ation therapy, with or without adjuvant chemotherapy, within 6 is based on histologic evaluation of a tissue specimen. There
weeks after their initial cytoreductive surgery. The primary are several systems for the grading and classification of
aims of imaging before irradiation are to (1) assess residual or brain tumors, but a three-tiered grading system based on the
recurrent tumor; (2) determine the volume of tumor to be irra- World Health Organization (WHO) criteria is the most widely
diated; and (3) plan the radiation isodose field. accepted and used. Regardless of the different grading systems
Once irradiation is completed, postirradiation imaging and the degree of histopathologic differentiation, all infiltrating
should be performed to assess tumor response to radiation and astrocytomas are considered malignant because, despite treat-
any immediate radiation-related complications, such as edema, ment, even the low-grade tumors tend to dedifferentiate into
mass effect, or hemorrhage. Imaging at this time can also serve higher grade tumors.
as a baseline scan before the patient starts adjuvant chemother-
apy. For those patients enrolled in therapeutic clinical trials, a
regular follow-up clinical and imaging evaluation is performed
Low-Grade Astrocytoma
as part of the protocol—usually at intervals of 2 to 3 months, Compared with their more malignant counterpart, low-grade
depending on the patient’s clinical status. The goals of serial astrocytomas are far less common, tend to affect a younger
imaging studies during chemotherapy are to (1) assess response patient population, and have a more favorable prognosis. On
to treatment and (2) detect tumor recurrence and therapy- imaging, low-grade astrocytomas tend to have a diffuse infiltra-
related complications. tive tumor margin, but there are also those with a focal, well-
After the completion of irradiation and chemotherapy, circumscribed border. On angiography, they usually appear as an
patients should undergo surveillance imaging to monitor them avascular mass with an absence of either arteriovenous shunting
for tumor recurrence or emergence of a new tumor focus. or early venous draining. On CT, low-grade astrocytomas tend
to be isodense to hypodense to the adjacent brain, and intratu-
moral calcification may be seen, although it is not common. On
PRIMARY BRAIN TUMORS contrast-enhanced CT, low-grade astrocytomas tend not to be
enhanced. On MRI, they appear as fairly well-circumscribed
lesions despite their infiltrative growth pattern, and they are
Astrocytomas homogeneously hyperintense on T2-weighted images. Contrast
Astrocytomas are the most common primary intra-axial brain enhancement is usually minimal or absent and so is peritumoral
tumor and the most common glial neoplasm, constituting edema (Figure 4-1). When there is robust and irregular contrast

A B

Figure 4-1 Low-grade glioma. A, Axial fluid-attenuated inversion recovery image demonstrates a hyperintense cortically based mass
in the right frontal lobe. B, Axial postcontrast spoiled gradient recalled image shows no abnormal enhancement within the right frontal tumor.
 C H A P T E R 4 Diagnostic Imaging 23

enhancement of the lesion, the diagnosis of low-grade astrocy- Oligodendroglioma

toma should not be considered the most likely choice.
Oligodendrogliomas, which constitute less than 10% of all
primary glial neoplasms, are thought to have originated from
Anaplastic Astrocytoma and oligodendrocytes of the brain. Histopathologically, oligoden-
Glioblastoma Multiforme drogliomas can be pure or mixed with astrocytic tumor cells.
Variable contributions of the oligodendroglial and the astro-
High-grade astrocytomas—anaplastic astrocytoma and GBM cytic components make oligodendrogliomas appear different
—are much more common than low-grade astrocytomas and from one another on imaging. One of the commonly seen
tend to affect an older group of patients. GBM is not only the imaging hallmarks of oligodendroglioma is intratumoral calci-
most common but also the most malignant of all primary glial fication. The presence of a tumor-associated cyst and involve-
neoplasms. The prognosis for patients with high-grade astro- ment of the cortical surface are also not unusual (Figure 4-3).
cytomas, especially a GBM, is usually poor, even with the most The tumor can be quite large at initial presentation. There
aggressive multimodality therapy. On imaging, high-grade are no definitive imaging criteria to differentiate low-grade
astrocytomas tend to be vividly enhanced, albeit in a heteroge- astrocytomas from anaplastic oligodendroglioma, although
neous and irregular pattern. The tumor margins are often ill extensive edema and enhancement tend to be associated with
defined, and peritumoral edema tends to be moderate to severe. the anaplastic variant.
It is often difficult to differentiate anaplastic astrocytomas from
glioblastomas, but the presence of tumor necrosis—as evi-
denced by a nonenhanced central region with a hyperintense
T2 signal surrounded by a thick irregular enhanced rim (Figure Gliomatosis Cerebri
4-2)—is highly suggestive of glioblastoma. Some glioblas- Gliomatosis cerebri is a rare neoplasm of the brain character-
tomas show no contrast enhancement or peritumoral edema, ized by moderately pleomorphic glial cells that infiltrate along
although such a case is very unusual. the preexisting structures without causing demonstrable
As the name multiforme implies, GBMs are extremely het- destruction. One of the hallmarks of gliomatosis cerebri is
erogeneous and variegated tumors. Histologically, GBMs are increased cellularity without frank destruction of the infiltrated
characterized by increased cellularity, high proliferation parenchyma or presence of neovascularity. Invasion of both the
indices, endothelial proliferation, and focal necrosis. Similarly, gray and white matter is seen, but it is distinctly unusual to find
on imaging GBMs can appear markedly different from the mitosis, necrosis, or angiogenesis.
more classic rim-enhanced necrotic mass, appearing to be a On MRI, gliomatosis cerebri is seen as a focal or diffuse
nonenhanced infiltrative lesion that at times can even be con- infiltrative signal abnormality involving white matter and
fused with an acute infarct. expansion of cortex, usually without contrast enhancement.

A B

Figure 4-2 Glioblastoma multiforme. A, Axial postcontrast spoiled gradient recalled image demonstrates an irregularly rim enhanc-
ing, centrally necrotic mass in the right frontal lobe. B, Axial T2-weighted image shows edema surrounding and compression and displacement of
the adjacent right lateral ventricle.
24 S E C T I O N B Diagnosis

A B

Figure 4-3 Low-grade oligodendroglioma. A, Axial fluid-attenuated inversion recovery image demonstrates a large hyperintense right
frontal mass with complex architecture compressing the front of the right lateral ventricle. B, Axial postcontrast T1-weighted image shows rimlike
linear enhancement along the peripheral margins of the tumor.

There is minimal, if any, mass effect or extensive vasogenic intracranial neoplasms, PCLs are treated with combined high-
edema.6,15,17,19,20 When there is new emergence of contrast dose chemotherapy and radiation therapy without surgery. Sur-
enhancement, malignant degeneration into a higher grade gical intervention is usually limited to biopsy for obtaining
tumor should be highly suspected. tissue for a pathologic diagnosis.9,12,16 It has been shown that
surgical resection of PCL does not necessarily alter prognosis,
and it can lead to profound functional deficit and increased
Ganglioglioma postoperative morbidity. More important is that, when patients
Gangliogliomas are most often seen in the pediatric patient with PCL receive steroid therapy to reduce intracranial pres-
population and are characterized by a biologically indolent sure before biopsy, the pathologic findings can be difficult to
growth pattern. The most common location is the temporal interpret, and a definitive diagnosis cannot be made. In addi-
lobe, but they may be seen anywhere, including the posterior tion, because PCLs tend to occur near eloquent areas of the
fossa.10,14, On imaging, they tend to have a well-circumscribed brain and in a subependymal location, gross total resection may
border, often composed of a solid enhanced nodule and a cystic in fact be contraindicated, and performing a biopsy may be
component. Peritumoral edema is usually minimal, even in challenging. A minimally invasive and accurate diagnosis of
large lesions. PCL is therefore an important goal that could clearly alter
patient management and reduce risk to the patient.
On imaging, PCLs are usually homogeneously enhanced.
PRIMARY BRAIN TUMORS OF They can be multiple and tend to favor deep gray or subependy-
NONGLIAL ORIGIN mal locations, often involving the corpus callosum. At times
they are indistinguishable from GBM. However, unlike GBM,
which has a hyperintense signal on T2-weighted MR imaging,
Primary Cerebral Lymphoma PCLs tend to show a hypointense T2 signal, presumably
Over the past 2 decades, the incidence of primary cerebral lym- because of their dense cellularity and a high nuclear-to-
phoma (PCL) has substantially increased in both immunocom- cytoplasmic ratio that results in decreased extracellular fluid.
promised and immune-competent individuals—a phenomenon Therefore the T2-signal intensity of the tumor can be helpful
that cannot be entirely explained by changes in tumor classifi- in narrowing the differential diagnosis between GBM and PCL.
cation, the increased prevalence of AIDS, or the growing There are other brain tumors that are often hypointense in
number of organ transplantations. The incidence ratio of both signal intensity on T2-weighted images, however, including
PCL and GBM rose from 1 in 250 in 1974 to 1 in 6 in the period primitive neuroectodermal tumors (e.g., medulloblastoma and
1981 to 1990, with PCLs now accounting for 6.6% to 15.4% pineoblastoma) and metastatic tumors, usually those metastatic
of all primary brain tumors.5,18 Unlike other high-grade from a mucinous adenocarcinoma primary.
 C H A P T E R 4 Diagnostic Imaging 25

Primary Nonglial Tumors of restriction of water within the area of infarcted tissue. In cases
of subacute infarction or venous infarction, the imaging appear-
the Brain Coverings ance can be quite confusing and misleading and hence may lead
Although a broad spectrum of neoplasms can originate from to an erroneous diagnosis of tumor. In this situation, the clini-
the coverings of the brain—the leptomeninges and dura— cal history can be helpful in narrowing the differential diagno-
meningioma is by far the most common. Meningiomas are sis. It is important to recall that the most common presenting
usually well-circumscribed lesions with a clearly demarcated symptom of a brain tumor is seizure, whereas less than 3% of
tumor-brain interface. They tend to be biologically benign, all patients with acute infarction have seizure as the primary
although malignant variants occur. initial symptom.
On imaging, meningiomas show characteristics of an On CT, the early phase of acute infarction may not be dis-
extra-axial tumor comprising a buckling or displacement of the cernible at all. There may be a subtle effacement of gray-white
gray-white matter junction and a cleaving of the subarachnoid matter differentiation or a hyperdense artery suggestive of an
space and vessels that surround the mass, separating it from the occluded vessel. On MRI, the most important clue to the correct
adjacent brain (Figure 4-4). Meningiomas are isointense to diagnosis of infarction is the abnormality seen on DWI. The
brain on unenhanced T1-weighted and T2-weighted images and degree of mass effect and gyral swelling is most prominent
tend to be vividly and homogeneously enhanced after the during the early phase of infarction, whereas contrast enhance-
administration of contrast agent. The degree of peritumoral ment, usually of a gyral pattern, is most conspicuous several
edema can be minimal to quite extensive (see Figure 4-4C). days after the onset of acute symptoms. In venous infarction,
there may be a prominent hemorrhagic component as evidenced
by marked hypointensity of signal on T2-weighted images or
SECONDARY BRAIN TUMORS on gradient-echo images.
(METASTASES)
Brain Abscess
Intracranial metastasis makes up almost 50% of all supratento- Brain abscess can simulate necrotic brain tumors. The use of
rial brain tumors. Metastatic tumors spread into the central antibiotics has substantially reduced the incidence of intracra-
nervous system through a hematogenous route and induce neo- nial bacterial abscesses. With a growing number of AIDS and
vascularization as they grow and expand. Intracranial metasta- organ transplantation patients, however, the prevalence of
tic tumors, especially those that cause neurologic symptoms, opportunistic infections is on the rise, especially those caused
are associated with a variable degree of vasogenic edema. The by viral or protozoan organisms. Brain abscesses are usually
capillaries within the area of vasogenic edema are consistently the result of either a direct spread of infection from extracra-
normal in appearance, and no tumor cells are found beyond the nial sites, such as the paranasal sinuses, or hematogenous dis-
macroscopic boundaries of the tumor.21 The radiologic diagno- semination from systemic bacteremia. The radiologic diagnosis
sis of intracranial metastases is usually straightforward and of intracranial infections can be challenging because of the
uncomplicated. Metastatic tumors tend to be multiple and well variable appearance of lesions as a result of different offending
circumscribed and favor the gray-white matter junction. microbes and different stages of presentation.
Patients generally have a known history of systemic malig- On MRI, intracranial infection can present as a focal mass
nancy. Solitary metastases, however—which can occur in 30% or a diffuse process, depending on the offending organism and
to 50% of cases—may pose a diagnostic challenge in differen- the stage of infection. Bacterial infections can present as a non-
tiating them from a primary glioma, particularly when the specific, diffuse signal abnormality with or without enhance-
patient has no known history of systemic cancer or when the ment in the cerebritis stage, which precedes the formation of a
medical evaluation does not identify a primary systemic malig- well-defined abscess. Cerebral abscess, which is usually caused
nancy. In general, metastatic tumors tend to have a more cir- by pyogenic bacterial organisms, often presents as a ring-
cumscribed border than do primary gliomas. enhanced mass with marked surrounding edema. In imaging a
well-formed abscess, there is a fairly uniform peripheral
enhancement, a well-defined capsule that appears hypointense
TUMOR-MIMICKING LESIONS on T2-weighted images, and moderate to severe vasogenic
edema.
There are several non-neoplastic brain lesions that masquerade Similarly to acute infarcts, brain abscesses have a charac-
as intracranial neoplasms. These so-called tumor-mimicking teristic abnormality on DWI, presumably because of a high
lesions (TMLs) may present a considerable diagnostic chal- viscosity of the pus that results in a restriction of water and
lenge to both the clinician and the radiologist. Clinically, TMLs hence a hyperintense signal on DWI. There are, however,
can present with symptoms suggestive of a mass lesion. The brain tumors that have a diffusion abnormality similar to that
three most common TMLs in the brain are infarction, abscess, seen with an abscess, such as epidermoid tumors, mucinous
and large demyelinating plaques. metastases, and primary cerebral lymphoma.

Infarction Tumefactive Demyelinating Lesions

A radiologic diagnosis of acute infarct is fairly straightforward Tumefactive demyelinating lesions (TDLs) are large demyeli-
since the introduction and wide clinical application of diffu- nating lesions that can mimic high-grade glial tumors. The lit-
sion-weighted MRI. On diffusion-weighted imaging (DWI), erature contains many reports of large demyelinating lesions
acute infarcts are markedly hyperintense because of localized that masquerade as intracranial neoplasms.3,22 These TDLs may
26 S E C T I O N B Diagnosis

A B

Figure 4-4 Meningioma. A, Axial unenhanced T1-

weighted image shows a well-circumscribed left temporal extra-axial
mass that has similar signal characteristics as the adjacent normal
brain. B, Axial postcontrast T1-weighted image demonstrates avid and
homogeneous enhancement of the mass. C, Axial fluid-attenuated
inversion recovery image demonstrates extensive edema surrounding
C the mass.

pose a considerable diagnostic challenge to both the clinician atypical appearance of a large lesion should cause concern
and the radiologist. Clinically, TDLs may present with symp- about a concurrent neoplasm.
toms suggestive of a mass lesion. On MRI, TDLs are often
indistinguishable from high-grade glial neoplasms, demon-
strating ill-defined borders, mass effect, perilesional edema, CONCLUSION
central necrosis, contrast enhancement, and variable involve-
ment of gray-matter structures. Even in the case of an estab- The differential diagnoses and imaging characteristics of brain
lished demyelinating disease such as multiple sclerosis, the tumors are quite variable, depending on tumor type and grade.
 C H A P T E R 4 Diagnostic Imaging 27

By using both clinical and imaging information, the differen- requested. The multiplanar capability and superior soft-tissue
tial diagnosis can be narrowed and an accurate diagnosis can resolution of MRI with the intravenous administration of a con-
be made. Imaging goals for patients who have a brain tumor trast agent make MRI the preferred diagnostic modality for use
depend on the clinical situation in which the imaging study is with patients who have a brain tumor.

References

1. Burger P: Classification, grading, and patterns of spread of experience at the University of California San Francisco. Cancer
malignant gliomas. In Apuzzo ML (ed): Neurosurgical Topics: 73:2570–2582, 1994.
Malignant Cerebral Glioma. Park Ridge, Ill, American Associa- 13. Mamelak AN, Prados MD, Obana WG, et al: Treatment options
tion of Neurological Surgeons, 1990. and prognosis for multicentric juvenile pilocytic astrocytoma. J
2. Burger PC, Vogel FS: The brain: tumors. In Burger PC, Vogel FS Neurosurg 81:24–30, 1994.
(eds): Surgical Pathology of the Central Nervous System and Its 14. Miller DC, Lang FF, Epstein FJ: Central nervous system gangli-
Coverings. New York, Wiley, 1982. ogliomas. Part 1: Pathology. J Neurosurg 79:859–866, 1993.
3. Dagher AP, Smirniotopoulos J: Tumefactive demyelinating 15. Porta-Etessam J, Berbel A, Martinez-Salio A, et al: Gliomatosis
lesions. Neuroradiology 38:560–565, 1996. cerebri. MRI, SPECT and the study of pathology. Rev Neurol
4. Daumas-Duport C, Scheithauer B, O’Fallon J, Kelly P: Grading 29:287–288, 1999.
of astrocytomas: a simple and reproducible method. Cancer 16. Reni M, Ferreri AJ, Garancini MP, Villa E: Therapeutic manage-
62:2152–2165, 1988. ment of primary central nervous system lymphoma in immuno-
5. Eby NL, Grufferman S, Flannelly CM, et al: Increasing incidence competent patients: results of a critical review of the literature.
of primary brain lymphoma in the US. Cancer 62:2461–2465, 1988. Ann Oncol 8:227–234, 1997.
6. Essig M, Schlemmer HP, Tronnier V, et al: Fluid-attenuated 17. Rippe DJ, Boyko OB, Fuller GN, et al: Gadopentetate-dimeglu-
inversion-recovery MR imaging of gliomatosis cerebri. Eur mine-enhanced MR imaging of gliomatosis cerebri: appearance
Radiol 11:303–308, 2001. mimicking leptomeningeal tumor dissemination. AJNR Am J
7. Garcia DM, Fulling KH: Juvenile pilocytic astrocytoma of the Neuroradiol 11:800–801, 1990.
cerebrum in adults. A distinctive neoplasm with favorable prog- 18. Schabet M: Epidemiology of primary CNS lymphoma. J
nosis. J Neurosurg 63:382–386, 1985. Neurooncol 43:199–201, 1999.
8. Greenlee RT, Murray T, Bolden S, Wingo PA: Cancer statistics. 19. Schoenen J, De Leval L, Reznik M: Gliomatosis cerebri:
CA Cancer J Clin 50:7–33, 2000. clinical, radiological and pathological report of a case with a
9. Herrlinger U, Schabet M, Clemens M, et al: Clinical presentation stroke-like onset. Acta Neurol Belg 96:294–300, 1996.
and therapeutic outcome in 26 patients with primary CNS 20. Spagnoli MV, Grossman RI, Packer RJ, et al: Magnetic resonance
lymphoma. Acta Neurol Scand 97:257–264, 1998. imaging determination of gliomatosis cerebri. Neuroradiology
10. Lang FF, Epstein FJ, Ransohoff J, et al: Central nervous system 29:15–18, 1987.
gangliogliomas. Part 2: Clinical outcome. J Neurosurg 21. Strugar J, Rothbart D, Harrington W, Criscuolo GR: Vascular
79:867–873, 1993. permeability factor in brain metastases: correlation with vaso-
11. Landis SH, Murray T, Bolden S, Wingo PA: Cancer statistics. CA genic brain edema and tumor angiogenesis. J Neurosurg
Cancer J Clin 48:6–29, 1998. 81:560–566, 1994.
12. Ling SM, Roach M, Larson DA, Wara WM: Radiotherapy of 22. Zagzag D, Miller DC, Kleinman GM, et al: Demyelinating disease
primary central nervous system lymphoma in patients with and versus tumor in surgical neuropathology. Clues to a correct patho-
without human immunodeficiency virus. Ten years of treatment logical diagnosis. Am J Surg Pathol 17:537–545, 1993.
 B

5
S e c t i o n

Chapter

PATHOLOGIC CLASSIFICATION Scott R. VandenBerg

The classification of tumors of the central nervous system CENTRAL NERVOUS SYSTEM
(CNS) attempts to precisely define highly complex and
dynamic biologic processes by a static, empirical framework. HISTOGENESIS AND
This is primarily based on histopathologic features that are TUMOR CLASSIFICATION
defined by a combination of routine and immunohisto-
chemical stains. Although these techniques have permitted The complex histogenesis of the CNS from a primitive neu-
useful characterization of a variety of neoplastic cellular phe- roepithelium must be considered when classifying the types
notypes comprising the spectrum of CNS tumors,57,58 the and distribution of the neural neoplasms arising in both the
current classification scheme has limited utility for precisely developing and mature CNS. The development of the nervous
predicting the invasive growth capacity, extent of recurrence, system is an intricate and elaborate multistep process involv-
or the efficacy of specific adjuvant therapies for many tumor ing induction of the neuroepithelium, regional subdivision and
types. morphogenesis, and cell-type specification with diverse neu-
The development of experimental molecular morphologic ronal and glial differentiation that culminates in an intricate
methods, including molecular cytogenetic techniques, promises arrangement of approximately 1011 neurons and 1012 glial
to complement the current classification of CNS tumors. Laser cells.50 These region-specific processes are associated with
capture microdissection of tissue sections now permits identi- different constellations of regulated gene expression that are
fying specific gene amplification or expression, and the activa- mediated by multiple homeobox genes, cell lineage–specific
tion of signaling pathways regulating cell proliferation and transcriptional activity, and gene methylation. Such differences
invasion in selected tumor cell populations. Such genomic and in regulated gene expression may create cell targets for neo-
proteomic approaches, when integrated into the current his- plastic transformation that are associated with specific devel-
topathologic framework, should provide the neuro-oncologist opmental stages and regions.43 The resultant transformed cells
with a more accurate prediction of biologic activity and defined may likewise have different groups of activated regulatory
therapeutic targets. Molecular genetic assessment of brain pathways affecting cell migration, proliferation, senescence,
tumors, especially with respect to astrocytic and oligoden- and apoptosis. Such differences would affect the prognostic
droglial tumors, has already enhanced our understanding of capability of a classification system.
oncogenesis and tumor progression in subtypes of tumors with The obvious differences in children’s and adults’ brains
similar or indistinguishable histopathology. Recent studies with respect to lineage and phenotypic differentiation are con-
have suggested that gene-expression-based classification of sistent with the existence of different cellular targets for neo-
malignant gliomas may have a stronger correlation with sur- plastic transformation. On the simplest level, this distinction is
vival than only the histologic classification.87 reflected by the different histogenetic potential and molecular
Advances in neuroimaging and neurosurgical techniques biology of the most common malignant tumors, the medul-
for precise tumor localization, combined with the expanded use loblastoma25 and glioblastoma,109 in children’s and adults’
of stereotactic procedures, necessitate that pathologic diagno- brains, respectively. There are also differences in the genetic
sis be rendered on minute tissue specimens from heterogeneous lesions and expression profiles between both the adult and pedi-
tumors. Thus for the most precise application of the current atric low-grade astrocytomas122,145 and the pediatric and adult
brain tumor classification, a clear and consistent definition of glioblastomas.16,105,129,130
key pathologic features must be integrated with clinical and Although the adult CNS lacks the diverse constellation of
neuroimaging data for each tumor type. These histopathologic stem cells and the differentiating neural cells that are present
features will be enumerated in detail in Part II of this text, in the developing brain, recent studies unequivocally document
whereas the following discussion will be a selected overview the presence of neural progenitor cells that have specific
of CNS tumor classification with specific emphasis on the regional distributions and variable potentials to generate glial
advances in experimental neuro-oncology in which the molec- and neuronal cellular lineages.131 Experimental murine models
ular biology complements the World Health Organization suggest that primitive glial and neural progenitors and more dif-
(WHO) conceptual framework. ferentiated glial cell lineages have different susceptibilities to
28
 C H A P T E R 5 Pathologic Classification 29

neoplastic transformation and develop different constellations of warrant communication to the neuro-oncologists, gemistocytic
gliomas after the introduction of the same genetic lesions, includ- astrocytomas are currently considered grade II in the WHO
ing activation of signaling pathways associated with platelet- classification, and a better definition of small glial cell lineages
derived growth factor B (PDGF-B), Ras, and Akt.20,45,46,136 awaits further study.
While the intriguing oncogenic roles of human adult neural pro- Because the presence of mitotic cells is a key feature to
genitors, as compared with their counterparts in the developing distinguish WHO grade II from WHO grade III astrocytoma,
brain, remain to be more completely defined, the concept immunohistochemical techniques have been used to more accu-
that primitive progenitors and differentiated cell populations rately enumerate proliferative cell populations. Immunostain-
are targets with differing susceptibility for specific genetic and ing for the Ki-67 protein, as defined by monoclonal antibody
epigenetic events in the process of neoplastic transformation MIB 1, is currently the most well-documented marker for
will have an impact on the future conceptual framework of tumor “proliferative” cells. Although Ki-67 immunoreactivity alone
classification. Likewise, molecular biologic characterization of is insufficient to definitively distinguish between anaplastic
specific cell populations within tumors will also be defined astrocytomas and glioblastomas, there is a good correlation
within a developmental and brain region–specific context. between MIB-1 labeling indices and WHO grade in diffuse
astrocytomas.34
The Ki-67 protein is a nuclear, nonhistone protein that is
required for maintaining the cell cycle. It is expressed during
TUMORS ARISING FROM G1, S, G2 and M phases but is absent in the G0 phase.21,31–33
NEUROEPITHELIAL CELLS Ki-67 appears to be minimally expressed in mid to late G1 and
significantly accumulates throughout S phase, with maximum
levels attained in M phase. Levels rapidly drop in postmitotic
Astrocytic Tumors cells with a half-life of less than 16% of G1 phase.11 Therefore
caution must be exercised when interpreting Ki-67 labeling
Diffuse Astrocytomas indices in tumors that are treated with small molecular
The current WHO classification of CNS tumors57,58 recognizes inhibitors with diverse effects on cell signaling, including dis-
two different categories of astrocytic tumors. The first category, ruption of the cell cycle in G1 phase or the G1-S transition,
the diffuse astrocytomas, is so designated because these tumors because these may result in a paradoxical increase in Ki-67
invade the brain or spinal cord tissue by diffuse cellular infil- antigen indices despite a lower proliferative fraction.19
tration at regional margins and cellular dispersion to distant While the features of nuclear pleomorphism and mitotic
sites. Aside from the capacity to diffusely invade the adjacent activity are generally useful for broadly discriminating the
neuropil or white matter and to spread remotely, a diffuse-type WHO II tumors from either non-neoplastic, reactive lesions, or
astrocytoma has a significant potential to develop progressively from WHO III neoplasms, the current histopathologic clas-
more malignant biologic behavior with time. Therefore the sification oversimplifies the biologic complexity of this neo-
grading of both anaplasia and the capacity for aggressive inva- plastic spectrum. Overlapping morphologic phenotypes thus
sive growth is implicit in the histopathologic classification of preclude precise predictions as to the biologic behavior of indi-
the diffuse astrocytic tumors; and likewise, an increase in the vidual tumors in the spectrum of the WHO II to III lesions.
WHO grade is accompanied by increased intratumoral hetero- Molecular approaches, especially with respect to unique pat-
geneity. The classification defines three groups of diffuse astro- terns and timing of genetic aberrations, may complement the
cytic neoplasms: astrocytoma (WHO grade II); anaplastic current system to better define subtypes of low-grade astrocy-
astrocytoma (WHO grade III); and glioblastoma multiforme tomas and predict the biologic potential of individual tumors.71,86
(GBM) and GBM variants, giant cell glioblastoma and gliosar- A comparative meta-analysis of genomic hybridization data
coma (astrocytoma, WHO grade IV). demonstrated the correlation between the average number of
According to the WHO scheme, tumors with nuclear atypia copy alterations, the average number of affected GTG-bands,
alone are considered grade II; those with mitotic activity in identified by Giemsa staining, for gains and losses, and the
addition to nuclear atypia are grade III; and neoplasms showing WHO grade in diffuse-type astrocytomas.61 With respect to spe-
atypia, mitoses, microvascular hyperplasia (“endothelial pro- cific genetic loci, methylation of either p14ARF or p16 INK4a,
liferation”), or necrosis are considered grade IV. The WHO without deletions in either gene, occurs exclusively in low-
classification recognizes three morphologic variants of cells grade astrocytomas. Such an epigenetic attenuation of gene
comprising diffuse astrocytomas: fibrillary, protoplasmic, and function may be associated with blocking replicative senes-
gemistocytic. However, no cytoarchitectural variant appears to cence7,48,72,90 to sustain mitogenic growth-factor stimulation.
have any prognostic importance. The exception appears to be Although both TP53 gene mutations and the expression of
astrocytomas containing more than 20% gemistocytes.66 Within PDGF and insulin-like growth factor 1 (IGF-1) appear to occur
these tumors, there is a highly variable population of small glial early in the tumorigenesis of the diffuse-type astrocy-
cells that appears to be the proliferative component, in contrast tomas,41,57,70 these molecular characteristics are not differential
to the gemistocytic tumor cells that are not proliferative but are features for low-grade neoplasms.
resistant to apoptosis. Although both the gemistocytic and small- GBM subtypes, with the same constellation of classic
cell components typically demonstrate nuclear p53 immunore- histopathologic features, have been identified on the basis of
activity, only the gemistocytes are immunoreactive for bcl-2,139 different sets of molecular genetic alterations in combination
suggesting a block in apoptosis. Gemistocytic astrocytomas with the clinical features of age of onset and clinical duration
tend to behave aggressively,66 with approximately 80% of these of tumor growth.56 The primary or de novo subtype of glioblas-
tumors progressing to glioblastoma. Although these features toma develops in older patients (mean age at onset is 56 years)
30 S E C T I O N B Diagnosis

with a short clinical interval (mean 1.7 months). The most adults, these tumors tend to appear 1 decade earlier (mean of
common distinctive molecular biologic features of primary 22 years) than the diffuse astrocytomas.30 Pilocytic astrocy-
glioblastomas are a loss of PTEN (phosphatase and tensin tomas are characteristically located in midline structures (e.g.,
homolog deleted on chromosome 10) (32%), epidermal growth- cerebellum, third-ventricular region, optic chiasm or nerves,
factor receptor (EGFR) amplification (40%), and MDM2 and brainstem) and may be cystic with a “mural nodule” of
(human homolog of mouse double minute 2) amplification or tumor. These tumors are not commonly located in the cerebral
overexpression (7% and 50%, respectively) in addition to loss hemispheres but, when present, show a slight predilection for
of a number of growth-arrest pathways, more often by deletion the temporal or temporoparietal regions and thalamus.30 These
(p14ARF and p16INK4a loss at 50% and 35%, respectively). tumors tend to occur in the spinal cord in older patients, more
The secondary subtype of glioblastoma develops in younger than at other sites, and may constitute a significant proportion
patients (mean age at onset is 40 years) over a period of years (58%) of spinal astrocytic tumors in some series.79
from lower grade tumors. In this subtype, the most differen- In a large series of cerebellar pilocytic astrocytomas,133 de
tial genetic lesion occurs in the TP53 gene. Retinoblastoma 1 novo anaplastic progression occurred in only 0.9%, and those,
(RB1) promoter hypermethylation (43%), loss of 14ARF and by flow cytometry, had higher S-phase fractions (5% to 11%)
p16INK4a by methylation and deletion (75% and 31%, respec- compared with 3.19 (±0.237 standard error of the mean).
tively), and cyclin-dependent kinase 4 (CDK4) amplification Although mitotic activity (£4 per 10 high-power field [HPF])
(13%) are also relatively common in secondary glioblastomas. may be an unusual feature of otherwise typical tumors, brisk
The rare giant cell glioblastoma subtype appears to have mitotic activity, in combination with increased cellularity,
a distinctive “hybrid” molecular genetic profile between the microvascular hyperplasia, or necrosis, may indicate anaplas-
primary and secondary glioblastomas with three notable tic progression.
features: (1) no genetic deletions of the CDK4/6 inhibitors One histopathologic variant, the monomorphous pilomyx-
(CDKN2a gene); (2) no amplification of the EGFR or CDK4 oid pilocytic astrocytoma, warrants particular notice.132 This
genes;77,92 and (3) a 30% frequency of PTEN mutations while variant occurs in infants and young children in the region of
yet having a higher frequency of TP53 mutations.93 These the hypothalamus and optic chiasm. In comparison with typical
differences from the other GBM subsets may account for a rel- pilocytic astrocytomas, these tumors have a more monomor-
atively more favorable clinical prognosis, including an appar- phic histologic pattern, a prominent myxoid stroma, and incon-
ently decreased capacity for brain invasion. spicuous or absent Rosenthal fibers and eosinophilic granular
In contrast, the gliosarcoma variant of GBM2 appears to bodies. These tumors have a significantly higher recurrence rate
be more closely related to the primary GBM subtype with a and the progression-free survivals at 1 year appear to be only
high incidence of PTEN mutations. However, there are impor- approximately 56% of those for typical pilocytic astrocytomas.
tant distinctions: (1) the incidence of TP53 mutations is approx- The molecular biologic profiles of pilocytic astrocytomas,
imately twice that in the primary GBM subtypes and lower than like the histopathologic features, are clearly distinctive from the
in the giant cell GBM or secondary GBM subset of tumors, diffuse-type astrocytomas. While TP53 mutation and increased
whereas (2) MDM2 and EGFR gene amplifications are signi- expression of PDGF-A and PDGF-Ra are common and prob-
ficantly lower than in the primary GBM subtype. ably early events in the formation of diffuse-type astrocytomas,
there appears to be no role for either TP53 mutations or aber-
rant PDGF signaling in the development of pilocytic astrocy-
Localized Astrocytomas tomas.67,88 Comparative analyses of gene expression in sporadic
The second category of astrocytic tumors in the WHO classifi- astrocytomas demonstrates that these tumors are uniquely
cation are the astrocytic tumors that share the feature of a rela- delineated from non-neoplastic white matter and other low-
tively circumscribed pattern of growth and a limited capacity for grade gliomas and have more similarity to fetal astrocytes.37
infiltrative spread into the surrounding brain or spinal cord. Four However, pilocytic astrocytomas do also express a spectrum of
distinct groups of astrocytic tumors compose this category, and genes associated with oligodendroglial lineages (natural killer
all are more common in infants, children, and young adults. cell restricted epitope PEN5, proteolipid protein PLP, peri-
pheral myelin protein PMP-22, myelin basic protein MBP, and
Pilocytic astrocytoma (WHO I). Pilocytic astrocytomas are oligodendroglial myelin glycoprotein).37,67
typically composed of distinctive biphasic patterns of neo- Approximately 30% of pilocytic astrocytomas occur in
plastic astrocytes accompanied by Rosenthal fibers and patients with NF1 and, notably, these tumors typically develop
eosinophilic granular bodies. Whereas pilocytic astrocytomas in younger children (mean age of 4.5 years at diagnosis).67 In
do not have a capacity for aggressive invasive growth and this subset of tumors, the reduction or loss of NF1 gene expres-
malignant progression within the brain or the spinal cord, com- sion appears to be a primary event in their development,35 while
pared with the diffuse astrocytomas the pilocytic tumor cells do in the sporadic tumors, the NF1 gene may even be overexpres-
appear to have selectively increased motility. Heterozygosity sed.59,100,146 In addition to differences between NF1 and sporadic
for the neurofibromatosis type 1 (NF1) tumor suppressor results pilocytic astrocytomas with respect to NF1 gene expression, the
in abnormalities in cell attachment spreading and motility in elongation factor EF-1a2 gene appears to have increased
astrocytes,36 with the resultant capacity to infiltrate the adjacent expression only in the sporadic tumors. Thus the sporadic and
leptomeninges or white tracts, especially of the brain stem, NF1-associated pilocytic astrocytomas, despite the same histo-
optic nerves, and optic chiasm. Leptomeningeal infiltration logic features, may develop and grow via different genetic
may play a role in the rare disseminated forms that do not alterations that culminate in similar histologic phenotypes.
manifest anaplastic features. Pilocytic astrocytomas are slow-
growing tumors that commonly arise in children and adoles- Pleomorphic xanthoastrocytoma (WHO II, III). Pleomor-
cents, with a peak incidence around 10 to 12 years of age. In phic xanthoastrocytoma (PXA) occurs as a superficial cerebral
 C H A P T E R 5 Pathologic Classification 31

(95% supratentorial) tumor in children and young adults lateral ventricles at the level of the basal ganglia or, less com-
(average age younger than 20 years) who usually have an monly, adjacent to the third ventricle. Symptoms generally are
antecedent history of seizures. There is a predilection for related to the obstruction of the foramen of Monro.
involvement of the temporal lobe and the overlying lep- The tumor cell populations in SEGAs exhibit a wide range
tomeninges, but growth into the dura is not common. These of astroglial phenotypes but have the hallmark of giant pyram-
neoplasms are commonly cystic (approximately 50%) and may idal cells, with a “ganglioid” appearance. Most tumor cells
present as a mural nodule. Although there is usually a well- demonstrate variable immunoreactivity for glial fibrillary acidic
defined macroscopic border with the subjacent brain, there is protein (GFAP) and S-100 protein, thus confirming the essen-
always focal microscopic infiltration. The abundance of a tially astroglial nature of this tumor. A number of tumors,
reticulin-positive stroma is usually a conspicuous feature of however, demonstrate both glial and neuronal-associated epi-
this group of astrocytic tumors where it delineates fascicles topes.42,69 These tumors may also exhibit ultrastructural features
of cells and is variably distributed between single tumor cells. suggestive of neuronal differentiation, including microtubules,
However, a spectrum of PXA cases lacking this conspicuous occasional dense core granules, and rare synapse formation.
stroma, but with otherwise typical features, have been These features recall those in tubers, the hamartomatous cortical
described.51 The prominence of this stroma may also vary lesions of tuberous sclerosis. Divergent glio-neuronal differ-
between the initial and recurrent tumors, suggesting that this entiation, a hallmark of tubers, may also be present in SEGAs.
phenotypic feature of PXA results from interactions between In contrast to the diffuse-type astrocytomas, there is no
tumor cells and specific types of brain extracellular matrices. correlation in the SEGAs between the anaplastic features of
A small number of cases have demonstrated atypical gan- endothelial proliferation, necrosis, increased mitoses and
glionic cells admixed with otherwise typical PXAs.29,60,65,68,75 marked cellular pleomorphism and any change in biologic
The proportion and pattern of the neuronal and glial com- behavior from typically indolent, noninvasive growth of these
ponents appear to vary from clusters of possibly entrapped tumors. The rare examples of recurrent tumors have never
neuronal elements to tumors in which the PXA appears to be undergone malignant transformation.39
the glial component of a ganglioglioma.29,68 The histiogenic Molecular genetics of SEGAs have demonstrated loss of
relationship between these rare cases and other desmoplastic heterozygosity in the TSC2 gene (16p13) and loss of the TSC2
glioneuronal neoplasms is currently unclear. gene product, tuberin, a Rap 1 homologue and putative tumor
Although the initial clinical data of the PXA group sug- suppressor.40,80 In a mouse model of TSC2 mutations, astroglial
gested that these tumors should be considered a low-grade neo- cells heterozygous for TSC2± had decreased expression of
plasm,53 the biologic behavior of PXA and its potential for p27Kip1, implicating a dysregulation of astrocyte proliferation
anaplastic progression remain uncertain. A number of cases (manifested by decreased contact inhibition in monolayer cell
have recurred and exhibited anaplastic progression,55,143 but culture) in the development of these neoplastic lesions.135 In
this behavior appears to be significantly less common than for addition, a functional loss of tuberin may stimulate vascular
diffuse infiltrative astrocytomas. Nonetheless, when compared growth,85 and SEGAs commonly have an abundance of abnor-
with the other more circumscribed, prognostically favorable mal blood vessels.
variants of astrocytomas (i.e., pilocytic and subependymal
giant-cell astrocytomas), PXA should be generally considered
a tumor with a significant potential for aggressive biologic Oligodendroglial Tumors
behavior. Accordingly, these tumors are designated as grade II The complex histogenesis of oligodendroglial tumors was rec-
or III (≥5 mitoses per HPF) by the WHO classification.57,58 The ognized more than half a century ago and remains a current
histopathologic features at the infiltrating margin may be espe- challenge to neuro-oncologists. Bailey and Bucy5 highlighted
cially important for prospectively evaluating the potential for glial heterogeneity as an intrinsic property of “oligoden-
recurrence and anaplastic progression55,143; however, infiltration drogliomas” with the variable presence of pleomorphic astro-
of Virchow-Robin spaces may occur without conferring a worse cytes and “all stages of transition” between “true” astrocytes,
prognosis. PXA, in contrast to the diffuse-type astrocytomas, more primitive glia, and the predominant oligodendroglial
may recur after long intervals without anaplastic progression.138 tumor cells. Molecular genetic analyses combined with thera-
The limited molecular analyses of PXA confirm its dis- peutic response and survival data from clinical trials have sig-
tinction from the diffuse-type astrocytomas, although there are nificantly added more precision to the WHO histopathologic
no definitive data as to specific mechanisms of tumorigenesis classification of oligodendrogliomas. There are distinct clinico-
and malignant progression from WHO II to WHO III. The genotypic groups of tumors within the histologic continuum of
genetic lesions that commonly present in the diffuse-type astro- gliomas that fall within the WHO oligodendroglioma classifi-
cytomas, the TP53 mutation and EGFR gene amplification, cation. The first, which comprises the majority of oligoden-
appear to be far less common in the pathogenesis of PXA, while drogliomas, has genetic losses on 1p (1p34-35; 1p36.2,
genetic losses not common in diffuse-type astrocytomas (chro- 1p36.3-pter) (40% to 92%) and 19q (19q13.3) (50% to 80%).
mosome 8p) may play a more important role.57,149 There is a tight association between 1p and 19q deletions and
a mutual exclusion between this first group and the smaller
Subependymal giant cell astrocytoma (WHO I). Subependy- numbers of histologic oligodendrogliomas that have TP53
mal giant-cell astrocytoma (SEGA) is a tumor that typically mutation, EGFR amplification, or PTEN mutations.26,44,107
occurs in association with tuberous sclerosis in the first or While multiple studies have shown that oligoden-
second decades of life, but tumor development may be the drogliomas, like low-grade astrocytomas, have hypermethyla-
primary manifestation of the disease. The tumors are conspic- tion of p14ARFand CDKN2A (p16INK4a) genetic loci, these loci
uously circumscribed, often nodular and multicystic with cal- are more commonly hypermethylated in oligodendrogliomas.
cifications. They are most commonly located in the wall of the Similarly, 5¢-CpG island hypermethylation of multiple genetic
32 S E C T I O N B Diagnosis

loci may be an important epigenetic mechanism regulating time.18,110 Correlation with neuroimaging data for clinical
gene expression at a number of sites in the tumors with 1p and grading of oligodendrogliomas is strongly warranted. Magnetic
19q deletions.147 Thus the 1p and 19q deletion appears to define resonance imaging of oligodendrogliomas (WHO II) com-
a unique subset of oligodendrogliomas with a distinctive profile monly shows minimal peritumoral edema, sharply demarcated
of gene expression.82 This subset also has a significant respon- margins (60% to 70%), hypodensity on T1-weighted seque-
siveness to specific chemotherapeutic agents and a best overall nces, and hyperdensity on T2-weighted sequences. However,
survival. In contrast, those oligodendrogliomas with genetic variable cystic degeneration and hemorrhage can result in more
lesions other than 1p or 19q loss have poorer clinical responses heterogeneous signal intensities of mixed hypodensities and
and shorter overall survivals. Activation of both EGFR and isodensities on T1-weighted images.
PDGF receptor (PDGFR) pathways most likely play important
roles in the biology of oligodendrogliomas,142 but amplification
of either EGFR or PDGFR genes is rare, even in anaplastic Anaplastic Oligodendroglioma (WHO III)
oligodendrogliomas with 1p or 19q loss. Wild-type EGFR may Anaplastic oligodendrogliomas share the histopathologic fea-
be overexpressed without the common mutation (EGFR vIII) tures of oligodendrogliomas WHO II in addition to the features
detected in de novo glioblastomas.76,108 Likewise, all com- of necrosis, microvascular hyperplasia, and high cellularity with
ponents of the functional growth-regulated oncogene-1 (GRO1)- a high mitotic activity and increased nuclear pleomorphism. A
PDGF pathway (PDGFA, PDGFaR, the GRO1 chemokine and definitive quantitative “grading” of these features as prognostic
CXC chemokine receptor [CXCR2]) are expressed in the factors has been problematic. Although necrosis, increased
majority of oligodendrogliomas (WHO II and III), in contrast mitotic rate, marked microvascular hyperplasia, and nuclear
to diffuse-type astrocytomas (WHO II and III).114 atypia may, in combination, be significant, extensive geographic
Progression of oligodendrogliomas to anaplastic tumors necrosis may be a key prognostic feature. However, the molec-
may be accompanied by increasing alterations in cell-cycle ular genetic analysis for 1p or 19q loss and other genetic aberra-
regulatory genes, notably deletional mutations in addition tions, as described earlier, is an important prognostic complement
to methylation at the 9p21 loci affecting CDKN2A/B/p14ARF to the tumor histopathology. Anaplastic tumors with brisk
and RB1 methylation.44,107,140 Overexpression of the angiogenic growth are marked by significant neovascularity and increased
growth factor vascular endothelial growth factor (VEGF) is vascular enhancement with magnetic resonance imaging,
also common with malignant progression. In addition, there are although approximately two thirds of oligodendrogliomas
smaller subsets of anaplastic oligodendrogliomas that have a (WHO II) may have a limited, patchy vascular enhancement.
TP53 mutation,139 10q chromosomal loss of heterozygosity
(LOH) (including PTEN) and EGFR amplification. Four
subsets of histopathologic anaplastic oligodendrogliomas are Oligoastrocytoma (WHO II)
defined on the basis of a hierarchy of molecular genetic alter- The majority of oligodendrogliomas contain variable numbers
ations: (1) tumors with 1p and 19q loss without other detectable of reactive astrocytes that may be a conspicuous component of
genetic lesions; (2) tumors with 1p loss without 19q loss or the non-neoplastic stroma. In contrast, the mixed oligoastrocy-
other detectable genetic lesions; (3) tumors with an intact 1p tomas contain two populations of neoplastic cells that have
chromosome with TP53 mutations; and (4) tumors with an variable astrocytic and oligodendroglial phenotypes. These
intact 1p chromosome without TP53 mutations. These neo- different phenotypes may be focally or diffusely distributed.
plastic subsets have differences in the age of onset, common Smear preparations of the latter, in combination with GFAP and
sites of involvement, clinical responses, and survival times.107,141 vimentin immunohistochemistry, highlight the distinct cyto-
morphologic features of the two cell populations.
Similarly to the oligodendrogliomas, molecular genetic
Oligodendroglioma (WHO II) studies have significantly complemented the histopathologic
While approximately 5% to 15% of intracranial gliomas are classification and have demonstrated that mixed oligo-
classified as oligodendrogliomas, this tumor group, particularly astrocytomas are a heterogeneous group of tumors with respect
among the low-grade gliomas, is most probably underdiag- to molecular or genetic alterations.8,9 One subset of these mixed
nosed as a result of its histogenetic complexity, as discussed tumors has the genetic lesions more typically associated with
above, and may account for a higher percentage of tumors, oligodendrogliomas, while another shares those molecular
ranging from 19% to 25% of neoplasms. These tumors most genetic profiles associated with the diffuse-type astrocytomas.
often arise in adults, with an overall peak incidence in those in Between 30% and 70% of oligoastrocytomas have allelic losses
their fourth and fifth decade. In children, the peak incidence is on chromosomes 1p and 19q in both oligodendroglial and astro-
between 6 and 12 years of age.26 Any region of the neuroaxis cytic components,63,74,106 while approximately 30% of tumors
may be involved, but the white matter of the frontotemporal have genetic alterations often seen in diffuse astrocytomas,
region and the basal ganglia are common sites. including TP53 gene mutations or 17p loss.74,106 It appears that
For oligodendrogliomas defined by using only the the subsets of oligoastrocytomas with 1p and 19q losses com-
histopathologic features of the WHO classification, the criteria monly have a predominance of oligodendroglial histologic fea-
for grading oligodendroglial tumors as WHO II include (1) no tures and more commonly occur outside the temporal lobes,
or infrequent mitoses with overall low MIB-1 labeling indices while the mixed tumors with TP53 mutations or 17p loss are
(less than 5%); (2) no microvascular hyperplasia; and (3) no more often astrocytic predominant and more commonly arise
geographic necrosis. A study of Ki-67 in oligodendrogliomas in the temporal lobes.74,81 Four subtypes of mixed oligoastro-
showed that the mean MIB-1 labeling index (LI) in oligoden- cytomas have been defined on the basis of chromosomal imbal-
drogliomas (low-grade) is less than 2, and that tumors with ances detected by comparative genomic hybridization: (1)
MIB-1 LI greater than 5 were associated with shorter survival tumors with losses on 1p or 19q, (2) tumors with gain on chro-
 C H A P T E R 5 Pathologic Classification 33

mosome 7 (7q more than 7p) and losses on chromosome 10, nostic factor of particular significance in supratentorial lesions
(3) tumors with losses on 1p/19q, plus other chromosomal appears to be the mitotic index.120 The importance of the
imbalances, and (4) tumors with imbalances other than in proliferative component in predicting the biologic behavior
groups 1 to 3.49 Unlike oligodendrogliomas, the mixed oligoas- has been confirmed by tumor growth-fraction measurements,
trocytomas do not have a well-characterized clinical behavior either by bromodeoxyuridine (BUdR) uptake4 or immunohis-
based on the profile of genetic aberrations. The apparent het- tochemistry for MIB 1.102,111,116,117,123 These studies showed that
erogeneity of mixed oligoastrocytomas suggests that glial pro- high labeling indices have a positive correlation with the his-
genitor cells with differing lineage plasticity45 may develop tologic grade and early tumor recurrence. Alternatively, the
tumors with similar histopathologic phenotypes but with dis- presence of endothelial proliferation, as an independent feature,
similar genotypes and gene expression profiles. appears to be less relevant to prognosis.115,121

Anaplastic Oligoastrocytoma (WHO III) Myxopapillary Ependymoma (WHO I)

The precise frequency of anaplastic progression from oligoas- Myxopapillary ependymomas are a distinct variant of ependy-
trocytomas is unknown. The histologic features of anaplasia are moma, common in adults during their third or fourth decades,
thus similar to those previously described for anaplastic astro- that are essentially restricted to the cauda equina. The neo-
cytomas.83 Anaplastic progression in mixed oligoastrocytomas plasms are usually discrete, elongate masses arising from the
appears to be accompanied by more genetic aberrations than in filum terminale or a spinal nerve root of the cauda equina.
anaplastic oligodendrogliomas, including EGFR and PDGF-R Rarely, extradural lesions in the sacral region occur, presum-
amplification, deletional loss of CDKN2A/B/p14ARF, and PTEN ably from ependymal rests. Myxopapillary ependymomas
mutation. contain an admixture of fibrillated and epithelioid cells with an
exuberant connective tissue stroma, which may be mucin pos-
itive. Papillary arrangements of elongated fibrillary cells that
Ependymal Tumors extend delicate processes to hyalinized vessels are a histologic
hallmark of these tumors. The unique histopathologic and bio-
Ependymoma (WHO II) logic features of this ependymal neoplasm suggest that it may
Ependymomas represent approximately 10% of brain tumors arise from cells derived from the distinctive caudal cell mass
and 6% of intracranial gliomas. Children and adolescents are associated with the secondary caudal neural tube formation
affected most often, although the tumors may occur in adults. during development.
Ependymomas are most common in the second decade of life
and preferentially arise in the fourth ventricle. In adults these
tumors represent more than 60% of the spinal gliomas, with a Subependymoma (WHO I)
peak incidence in the fourth decade. Multiple spinal cord Subependymomas are well-circumscribed, asymptomatic
ependymomas are typically associated with von Reckling- nodules in the walls of the fourth ventricle (66% to 70% of
hausen’s disease. cases), lateral ventricles, septum pellucidum, foramen of
The WHO classification recognizes four histopathologic Monro, or spinal cord. They may present with increased
variants of ependymomas: cellular, papillary, clear cell, and intracranial pressure caused by obstruction of cerebrospinal
tanycytic. These tumors have essentially the same clinical fluid (CSF) flow or with hemorrhage within the tumor, or they
behavior, but their recognition as ependymomas is important to may be detected incidentally during autopsy. The cytoarchitec-
eliminate confusion with anaplastic gliomas, meningiomas and tures of the tumor cells exhibit composite features of ependy-
choroid plexus tumors, oligodendrogliomas, and pilocytic mal and astrocytic differentiation. Despite the occasional
astrocytomas, respectively. finding of increased atypia and mitotic activity, tumor location
Molecular genetic analyses demonstrate that the ependy- and successful resection are the most important prognostic
momas are distinct from astrocytic and oligodendroglial tumors factors. The histogenesis of these lesions remains to be deter-
in that the common genetic alterations found in either of these mined, but discriminating them from ependymomas is impor-
gliomas are not common in the ependymomas. Loss of het- tant to avoid needless adjuvant therapy.
erozygosity of chromosome 22 occurs in approximately 30%
of cases in adult tumors but is uncommon in pediatric tumors,61
and loss of the neurofibromatosis type 2 (NF2) gene on chro- Choroid Plexus Tumors
mosome 22 may be a subset localized to the spinal cord.24 Choroid plexus tumors arise from the specialized epithelium
constituting the choroid plexus and therefore can arise in the
lateral and fourth ventricles. However, they commonly arise in
Anaplastic Ependymoma (WHO III) the lateral ventricles in children as choroid plexus papillomas
Anaplastic progression of ependymomas can occur at most (WHO I), although cases in adults and tumors arising in the
sites, but rarely in the spinal cord. For ependymal tumors, in fourth ventricle may also occur. Clinical symptoms are com-
comparison with astrocytomas, there are no features that appear monly caused by increased intracranial pressure, secondary to
to accurately grade anaplasia. Nevertheless, anaplastic ependy- obstruction of the CSF pathways or an increased production of
momas are recognized by the combination of increased cellu- CSF. These clearly demarcated, often calcified tumors are dis-
larity and increased mitotic rate, microvascular proliferation, tinctive for a conspicuously papillated surface that resembles
and variable degrees of cellular pleomorphism with nuclear the normal choroid plexus. The choroid plexus carcinoma
atypia. Microscopic or geographic necrosis may be present, but (WHO III) shows anaplastic features and can be both locally
necrosis alone is not a robust discriminating feature. A prog- invasive and diffusely disseminating in the CSF pathways.
34 S E C T I O N B Diagnosis

These malignant tumors are rare (less than 20% of choroid of multiple hamartomas and neoplasias.114 The genetic lesion is
plexus tumors) and typically occur in infants. a germline mutation of the PTEN gene. Cowden’s disease has
the hallmark mucocutaneous manifestations of trichilemmo-
mas, related follicular malformations, and a distinctive type of
Neuroepithelial Tumors of Uncertain Origin hyalinizing, mucinous fibroma, in addition to acral keratoses
There are four neuroepithelial neoplastic lesions that have dis- and oral papillomas.128 A significant number of patients also
tinctive histopathologic features but for which the histogenesis have thyroid adenoma or multinodular goiter, fibrocystic
is unclear. These are the chordoid glioma of the third ventricle, disease or adenocarcinoma of the breast, gastrointestinal polyps
the astroblastoma, and gliomatosis cerebri. Although these (colon, gastric, and esophageal), and ovarian cysts and polyps.
tumors are considered by some authors to be individual enti- The development of the cerebellar lesions in adolescents may
ties, they may, in fact, represent only distinctive histopathologic herald the presence of Cowden’s disease that usually does not
variations of more well-defined ependymal, oligoastrocytoma, more fully develop until the second to third decades.144
and astrocytic tumors. With the perspective of the germline PTEN mutation asso-
ciated with this hypertrophic lesion, it is interesting to note that
the tumor is composed of abnormal ganglion cells with abnor-
Neuronal and Mixed Neuronal-Glial Tumors mally myelinated parallel fibers, often with a reduction of the
Neuronal and mixed neuronal-glial tumors are a heterogeneous adjacent granular layer and demyelination of the cortical white
group of neoplasms that includes (1) tumors with only neuronal matter. Thus the morphogenesis of this lesion may be an aber-
tumor cell phenotypes, comprising the gangliocytoma (WHO I) rant expansion of a specific neuroblastic population (possibly
and the dysplastic gangliocytoma of the cerebellum (Lhermitte- a block in apoptosis) in combination with altered, but limited,
Duclos disease) (WHO I); (2) tumors with a predominant neu- cellular migration. There is never any manifestation of inva-
ronal cell phenotype and a minor, variable glial component, siveness or evidence for a significant population of proliferat-
comprising central neurocytoma (WHO II) and cerebellar ing cells with dysfunctional cell-cycle regulation, as in the de
liponeurocytoma (WHO I to II); (3) tumors with mixed popula- novo GBMs that have PTEN mutations in combination with
tions of neoplastic neuronal–glial populations, comprising gan- dysfunctional cell-cycle regulation and increased proliferative
gliogliomas (WHO I to II), desmoplastic infantile astrocytomas activity. Immunohistochemical studies suggest a histogenetic
and gangliogliomas (WHO I), and dysembryoplastic neuroep- relationship to Purkinje cell lineages27,124; however, other data
ithelial tumor (DNT) (WHO I); and (4) tumors with neuroen- suggest an origin from granular cell progenitors.112,148 Most
docrine or neurosecretory phenotypes–neural crest derivation, likely, the neuronal proliferation arises from a more cytogenet-
comprising paraganglioma (WHO I). The biologic activity of ically heterogeneous group of cells. Although most patients
DNTs especially reflects close interface between hamartoma- have a very favorable postoperative course, recurrences fol-
tous and neoplastic processes in tumor development. lowing subtotal resection do occur.
With the exception of the dysplastic gangliocytoma of the
cerebellum (see following discussion), the histogenesis and
molecular genetic alterations of these tumors are not defined. Pineal Parenchymal Tumors
Recent studies identified neural stem cells located within spe- Pineal parenchymal neoplasms account for less than 0.1% of
cific zones in the adult brain that are bipotential and unipoten- all intracranial tumors and approximately 11% to 28% of the
tial progenitor cells, depending on both intrinsic and extrinsic pineal-region tumors in children. Although these neoplasms
cues.43,131 Such cells, as targets for neoplastic transformation, have a wide spectrum of features, the current WHO classifica-
could give rise to the neuronal and mixed neuronal-glial tion designates three distinct groups. These range from the
tumors. An important corollary to this hypothesis would be that malignant pineoblastomas (WHO IV) to the well-differentiated
the growth potential and cellular phenotypic composition of the pineocytomas (WHO I). The third group, the pineal parenchy-
tumors may vary according to both the temporal and spatial mal tumors of intermediate differentiation, are tumors com-
origin of the transformed progenitors and to the local extrinsic posed of a more heterogeneous cell population encompassing
signals in the specific brain region. It is interesting to note that the cytoarchitectural features, cellular patterns, and biologic
the glial component of gangliogliomas may undergo clonal behavior that are generally intermediate but without a pre-
evolution during tumor development.6 dictable clinical behavior.
The dysplastic gangliocytoma of the cerebellum is a rare
lesion of the CNS. The neoplasmic mass usually becomes clin-
ically apparent between the third and fourth decades in patients Embryonal Tumors
with a reported mean age of 34 years; however, the lesions can Embryonal tumors are primitive and clinically aggressive neo-
clinically present from the neonatal period to 74 years. The plasms that commonly arise during the first decade of life. They
multitude of diverse designations previously applied to this are histogenetically distinct from anaplastic variants of gliomas
lesion, including granular cell hypertrophy, granomolecular and glioneuronal tumors that more commonly arise in adults.
hypertrophy of the cerebellum, diffuse hypertrophy of the cere- All embryonal tumors, regardless of histogenesis, share the
bellar cortex, gangliomatosis of the cerebellum, and gan- common features of high cellularity, numerous mitoses, and
glioneuroma, pointedly reflect the lack of understanding about focal or more extensive necrosis. These features manifest a clin-
its histogenesis and disagreements as to its hamartomatous or ically aggressive behavior that corresponds to WHO IV tumors.
neoplastic nature. In addition, there is a common propensity for leptomeningeal
The majority of dysplastic gangliocytomas arise in associ- invasion and subsequent metastasis in the CSF pathways.
ation with the clinical manifestations of an autosomal dominant, The classification of embryonal tumors over the past
systemic syndrome, Cowden’s disease, leading to the development decade has been controversial with respect to the category of
 C H A P T E R 5 Pathologic Classification 35

central primitive neuroectodermal tumors (PNETs). This cate- and GFAP, respectively) is present in the epithelial structures,
gory originally embraced the concept that such tumors arise while these proteins can be readily demonstrated in the more
from primitive neuroepithelial progenitor cells with equivalent differentiated cell populations surrounding the primitive neo-
developmental plasticity throughout the neuraxis. Although plastic neuroepithelium.
there is experimental evidence for a degree of autonomous
developmental neural plasticity of stem cells,134 this does not
preclude the concept that embryonal tumors may arise from pro- Medulloblastoma (WHO IV)
genitor cell populations with a more limited, but different, his- Medulloblastoma is the most common and well characterized
togenetic potential (multipotential or bipotential). Accordingly, of embryonal tumors. In children it constitutes approximately
this group of tumors would not necessarily represent a distinct one quarter of all intracranial tumors. The peak incidence for
clinicopathologic entity, but rather a category of tumors that medulloblastomas is near the end of the first decade, and there
could have a wide spectrum of histopathologic features and bio- is a slight male predilection. Most are situated in the cerebel-
logic behavior. Gene-expression profiling of supratentorial lar midline, although pediatric tumors may also appear in the
PNETs has demonstrated a lack of clustering and a molecular cerebellar hemispheres. The latter tends to be a more common
heterogeneity, possibly reflecting the diverse spectrum of site in the rare adult cases. Rare variants of medulloblastoma
tumors in this category.101 Experimental immortalization of include tumors showing striated muscle cell differentiation
immature cells from the rodent cerebellum has demonstrated the with populations of more primitive cells resembling rhab-
existence of bipotential clonal stem cells that can differentiate domyosarcoma (medullomyoblastoma WHO IV) and pig-
into both neuronal or glial cell lineages, depending on extrinsic mented papillary forms (melanotic medulloblastoma WHO
cues.28 If analogous progenitor cells exist in selective regions of IV) that are dissimilar to melanotic neuroectodermal tumors
the immature human brain, their transformation would most of infancy. The histogenetic relationships of these tumors
likely result in the PNET histologic phenotypes with possibly to the more common forms of medulloblastoma are not
distinctive epigenetic and genotypic alterations. It is most likely understood.
that the embryonal or primitive tumors of the CNS result from Medulloblastomas exhibit various phenotypic manifesta-
a combination of defects in signaling pathways that regulate the tions of neuroblastic or neuronal differentiation as demon-
clonal expansion and differentiation of neural progenitor cell strated by standard immunohistochemistry for neuronal
populations. The molecular biology of these tumors contrasts cytoskeletal, membrane or vesicular, and neurosensory epi-
with the various combinations of attenuation or loss of replica- topes. Several histologic variants are recognized: (1) the classic
tive senescence, dysfunctional regulation of the cell cycle, and subtype constituted by sheets of undifferentiated cells with
activation cell survival pathways that most often occur in variable numbers of neuroblastic rosettes and palisades, (2) the
tumors arising within the mature nervous system. desmoplastic or nodular subtype with rows and nodules of
There are diverse cell populations in the immature nervous neoplastic cells separated by an abundant collagen stroma, (3)
system that may be vulnerable targets for neoplastic transfor- the “extensively nodular” subtype, and (4) the large-cell or
mation; however, special caution should be exercised in this anaplastic subtype.
context to avoid complete reliance on immunohistochemical In the classic form, sporadic morphologic differentiation
techniques to establish the cell lineages of otherwise poorly dif- to ganglion cells within the primitive populations may occur
ferentiated tumors. Transient or heteroplastic expression of cell without any apparent correlation with clinical behavior. In a
type–associated proteins, differential sensitivity of specific epi- large retrospective study,23 moderate to severe grades of anapla-
topes to routine tissue handling, and incomparable antibody sia occurred in approximately one quarter of the cases and were
reagents should always be carefully considered, and all results associated with aggressive clinical behavior. Anaplasia is
should be interpreted with these reservations. Despite this defined by increased nuclear size, abundant mitoses, and the
caveat, a small number of embryonal tumors may be charac- presence of a large-cell component resembling the large-cell or
terized on the basis of distinctive features. anaplastic subtype (see the following discussion).
The current classification scheme recognizes a number of The desmoplastic or nodular subtype has highly cellular
embryonal tumors with restricted and relatively limited neural sheets and trabeculae of typical tumor cells embedded in a con-
histogenetic potentials along neuronal and glial or ependymal spicuous collagen stroma. Within the reticular-free zones, there
cell lineages: the supratentorial cerebral PNETs (neuroblas- is a variable decrease in cellularity that is accompanied by vari-
toma and ganglioneuroblastoma) and the ependymoblastoma, ably fibrillated processes and cells with morphologic evidence
respectively. A rare embryonal neoplasm, the medulloepithe- of increasing neuronal and ganglionic differentiation. In con-
lioma, merits separate mention. In contrast to the other tumors trast, the desmoplastic areas are composed of highly cellular,
mentioned, this neoplasm has the hallmark of a mitotically proliferating cell populations. The designation of “extensive
active, pseudostratified columnar epithelium, often arranged in nodular subtype” is applied when this biphasic pattern becomes
ribbons of tubules or papillary rosettes with variable interposi- predominant with the formation of extensive, less cellular
tion of delicate stromal elements, recapitulating the primitive nodules of reticulin-free populations of prominently fibrillary,
epithelium of the neural tube. Approximately 50% of cases differentiated cells, as demonstrated by neuronal cytoskeleton
exhibit neuroblastic or neuronal, astroglial, or ependymal cell and synaptophysin immunoreactivity. The characteristic
populations that are either intimately admixed with the tubules nodular architectural pattern can be present in an initial speci-
or present in more well-demarcated fields. Immunohistochem- men and may not be present in recurrent tumors.
ical studies demonstrate abundant IGF-1 and fibroblast growth Recent studies suggest that a combination of histopatho-
factor (FGF)-2125 in this primitive epithelium. In contrast, only logic taxonomy23 and molecular analysis of medulloblas-
rare sporadic expression of neuroblastic and glial cytoskeletal tomas25,101 may be the future basis for a more precise prognostic
proteins (the neuron-specific class III b-tubulin isotype TUJI grading of these complex neoplasms. Gene-expression profil-
Exploring the Variety of Random
Documents with Different Content
graceful lines of a mountain, or even the crystal curves in a fountain,
without dwelling on that form which, of all created, is undeniably the
most beautiful without any of its associations, and dwelling on it, too,
with somewhat other feelings than those expressed by the Italian
priest when he remarked, in a tone of reproof, to a friend who wished
to call his attention to a fair lady at an assembly—
“Una bella creatura di Dio!”

Thus I do think that if this celibacy is to be continued, it would be a

great improvement to enjoin the study of pure mathematics on
college Fellows, with examinations at intervals to prove that their
time is only taken up in contemplating the affinities of triangles, and
the love of the angles (not of the angels). The whole series of classical
literature ought to be forbidden them for the time; ditto all galleries,
pictures, and statues, all music and poetry; and they ought, as a final
measure, to be relegated to that monastery mentioned by Mr Curzon,
somewhere in the Acroceraunian mountains, where there were some
Greek monks who had never seen a female face, and had even
forgotten their mothers. One of them asked him whether women
were like the Madonna. The poor fellow had better not have seen that
Madonna. Even now, some men in their undergraduate life grow
tired of the exclusively masculine aspect of the University, and some
very good lines on that subject, of which I only recollect the end,
were written by a now eminent poet, when he was an undergraduate
—
“As I am one who feels the full divinity
Of a fair face in woman, I protest
I’m sick of this unvaried regularity
Of whiskered cheeks and chins of black barbarity.”

And one painful consequence of the present system is, the violation
of the good old adage, “Happy’s the wooing that’s not long a-doing:”
the notorious evil of long engagements becomes, in this case,
exaggerated to a painful degree. There being no absolute, but only a
conditional prohibition, and the prospect of a living, certain though
distant, appearing to justify the formation of such ties, engagements
are formed in early life, the ratification of which seems ever near, but
never actually comes, till both parties have passed their meridian,
and the fulfilment takes place, if it is thought worth while that it
should take place at all, rather as a matter of course, than because
the parties really now desire it. The hope deferred which “maketh the
heart sick,” embitters the masculine temper, and withers the
feminine frame, even before their natural bloom would have
disappeared. The courage which, in earlier life, would have taken a
bold step, and dared the world to do its worst, becomes irresolution
and timidity; and as it often happens that those who have been kept
without food too long, only know the sensation of hunger through a
general faintness of the system, so the vacuum of the affections too
long kept up by circumstances, becomes at last a chronic disease,
which, to the end of life, remains irremediable. At the same time, the
life of the common-room, and the extreme ease with which material
wants are provided for, acts on the mind as opium acts on the
system, till at last it ceases to care for anything but the drug which
has become a habit. It may be with some of those who have felt the
enduring influence of this soporific regime, as with the lotos-eaters
of Tennyson; they even come to dread a change, and cling to the
indolence from which at first they would have fled:
“Our island home
Is far beyond the sea, we will no longer roam.”

But, on the other hand, it may be urged that the immediate

happiness of those concerned is not so much contemplated in the
foundations as their usefulness, and that they must be content to cull
the flowers which grow beside the path of duty. This may be
answered by urging that, in certain cases, a man’s usefulness is
diminished instead of being increased by his being denied certain
sources of happiness. The best workman is ever the man who is best
fed and clothed, and made most generally comfortable; even so in
the great work of human life is that individual most efficient whose
legitimate wants, both of body and soul, are satisfied. The motives
which actuated the founders of the Roman Catholic colleges were no
doubt, as most human motives are, of a mixed nature. On the one
hand, they wished their money to fructify and do as much good as
possible; on the other hand, they wished it to fructify in such a way
as to redeem their own souls from purgatory, by providing a
succession of those who should sing masses for them for all time; at
the same time, it was the prevailing notion in these times, and is
now, among Romanists, that celibacy, if not the happiest, is the
holiest state of man.[4] If there be any truth in this, even to the most
limited extent, there is something to be said for the system; but if the
poor founders have been cheated out of their masses, and may
remain, for all the present generation care, boiling and broiling in
purgatory to the end of time, it seems purely hypocritical to lean on a
notion which has no better foundation than the ruling opinions of
founders. All the great and imposing faith is gone which would
support a heavy burden with the supernatural sinews of religion, and
the burden remains still to be borne as it best may by human muscle
alone. But it may be also said, the fellowships of colleges are in
themselves eleemosynary institutions, and poverty was in most cases
made a condition of the enjoyment of them; and just as, under the
new poor-law system, we imagine that a man, though he has a right
to existence, has no right to encumbrances which others must
support, so some would argue that the charity of the founders ought
to be thankfully accepted under all its conditions. But in the first
place, the question may be asked, whether apparent necessity, rather
than humanity, did not suggest the new poor-law system? In the next
place, whether that can strictly be called eleemosynary of which
merit is made a condition? We give to a beggar sometimes, although
we know him to be utterly worthless, merely because he is destitute;
and even the utterly worthless have a certain claim, in right of their
Maker’s image; but we give to a good man as a tribute to his virtue,
and the application of these foundations to proficiency in knowledge
is to those who accept them usually accounted peculiarly honourable,
just as a national pension is to the wounded soldier. Besides, it might
be said that all bequests are in a manner eleemosynary, because the
legacy is not a payment for labour in most cases, but a free gift from
the testator to the legatee; nor is its character materially altered by
the fact of its having been given under conditions. It appears to some
that the college property is as much real property to those who have
the use of it, as any property bequeathed subject to conditions; such
as, for instance, the law of entail in England. Indeed, a case has been
mentioned, in which, for some peculiar reason, a very rich man
inherited his estates subject to this very condition of celibacy. And
eleemosynary institutions, strictly so called, are commonly
administered by trustees, not by those who reap the benefits of them,
as is the case with college fellowships. I think I have now, as well as I
can, stated the arguments, both pro and con, though perhaps it is
easy for you to see to which side I lean. I confess that I should regard
the repeal of celibacy as a conservative change, because it would give
individuals a more enduring interest in their University. I dread
innovation, and especially from profane hands; at the same time, I
feel the necessity of such wholesome repairs in the constitution of
Alma Mater as shall secure for her, as far as possible, a perpetuity of
youth, or at least a green old age. How other changes, such as the
admission of Dissenters, can be brought about without ignoring the
entire history, associations, and character of the University, I do not
well see. If Dissenters are admitted at all, Roman Catholics must be
admitted with the rest; and they may perhaps lay claim to a
participation in the good things of the University, seeing that the
ancient foundations were undoubtedly made in their favour; and if
this participation be allowed, the rights of the foundation will be
again disturbed; and they may push their claim to the entire
exclusion of all other communities, for, unless there be a reason for
disfranchising them, they will ask why others should share
advantages originally intended for them alone. They are not like the
Jews, a sect who keep to themselves, and seek not to domineer over
others; but universal dominion is as much the policy of pontifical as
of imperial Rome. Thus they will be sure to take every advantage.
Thus there is a primâ facie danger in mooting any integral question
concerning the constitution of the University, lest an opening should
be unwarily made which would destroy everything on which its
existence depends; and this is, in my opinion, the most plausible
argument in favour of continuing the celibacy of Fellows. But averse
as all well-wishers to Oxford would be to any change in the way of
subtraction or diminution of her privileges, no such one could look
with coldness on any proposed additions to her area of efficiency,
and especially on extensions which seem suggested by her natural
aptitudes. As Cambridge seems to possess the soil in which
everything connected, however remotely, with science, is destined
especially to thrive, such as natural history in its various branches, so
does Oxford appear to be that University which should assume a
prominently artistic character. The foundations of a new museum
have been laid, which is to be built on a grand and imposing scale. Is
its chief attraction, when completed, to consist in a collection of
dried beetles and stuffed humming-birds, or even a complete
skeleton of the megatherium, if such a thing is to be had; or is an
attempt to be made to bring together, by every possible means, a
collection of works of art which would really do credit to the
University? It must be remembered that we have in England no
national gallery worthy of the name; not that the pictures composing
the collection in Trafalgar Square are to be despised—far from it; but
the building which contains them shows them to so little advantage,
and is altogether so inadequate, that it presents few temptations to
large additions, either by purchase, gift, or bequest. The very
atmosphere of London is an argument against building a new
national gallery in the neighbourhood of any of the centres of
metropolitan life. Trees may be blackened, but flourish under the
soot; but the purity of the marble, and the freshness of the canvass,
are liable to be permanently discoloured by the constant action of an
air impregnated with smoke, in a manner far other than that in
which they receive the mere mellowness of age. This would be
conclusive against a central situation, and if such a building is to be
placed in the suburb, to arrive at it would cost a sacrifice of time and
effort little short of that necessary to arrive at a site at a moderate
railroad distance from the metropolis. As it is, Oxford is a great point
of attraction to all strangers, and no Englishman who had not seen it,
could pretend to an average knowledge of his own country. It is even
placed within reach of the working-classes of London by excursion-
trains, who are thus led in the pursuit of pure air to a place full of
associations, which are in every way likely to do them good. It seems
to me that it is worth considering whether the national gallery of
England might not with advantage be placed at Oxford, and
combined in some way with the scheme of the new museum. A
school of art would probably spring up around it, to which the
University would naturally present many advantages, and to which it
might well extend peculiar privileges. The present is not the worst
time to consider this matter, when the existence of a great war
postpones the execution of all plans of subordinate importance. It is
quite certain that everything cannot be concentrated in London; and
this being the case, it is well to consider what other places are
calculated, in their own way, to become capital cities. Oxford has
already received some of the Muses as its inmates, and it is
abundantly spacious to receive them all. With respect to the natural
scenery of its environs, very much might be said in favour of its being
suited as a residence for an artist. The banks of its rivers are
especially fertile in subjects for the brush, and though its upland
scenery is generally stamped with that mediocrity which seems
peculiar to the central counties of England, there are spots here and
there which, from their wildness or woodiness, are well adapted for
the sketcher. I am sorry to see many of the wild places round Oxford
either already enclosed, or in course of enclosure; but what I saw
with most regret was, that Bagley Wood had been surrounded with a
fence, and placed under a most rigorous taboo to the public in
general. Now, there is some excuse for bringing land into cultivation
which may be made available for the wants of the community, and
can only become so if enclosed; but when the better preservation of
game is the only object, to exclude the public from a place where they
have been accustomed for years to expatiate and “recreate
themselves,” and an intelligent public, such as that of the University;
—to exclude them from one of the spots which Arnold mentioned as
giving him especial delight on his return to Oxford, and as being one
of its chief glories,—this, though perfectly justifiable according to
law, is scarcely consistent with that Aristotelian equity which ought
to be above law, especially in the neighbourhood of those brought up
in his precepts, and whose philanthropy might naturally be expected
to be more expansive than that of other men. It appears, however,
that this mischief has been done for some time; and the only
compensation the public gain is that a fine wide road has been made,
which certainly makes the walk round the wood complete—a poor
consolation, indeed, to those who, like myself, look upon walking
along a road as one of the dreariest duties imaginable, and have an
irreclaimable vein of the savage in their composition. Why, to me the
sight of the stiff hedges and mathematical drains of Bagley Wood
would spoil half the pleasure of shooting there; but, of course, those
who have that privilege may say that the grapes are sour. I may
mention that on the walk which crosses the railway, and cuts across
into the Abingdon road, which leads through Bagley Wood, a large
reservoir has lately been made, which in one place is crossed by a
bridge, that seems as if it had been put there on purpose to give the
best near view of the city. The best distant views I consider to be
those about the Hinksey fields, near the spot where Turner, with
singular ignorance of the customs of the University, painted
gownsmen in their academicals among the haycocks; and at a place
near Elstree, called Stow Wood, well known as a fox-cover. But
perhaps the most characteristic view of all is that of the towers of
Oxford, seen reflected in the flooded surface of Christchurch meadow
under a red sky. This view is suggestive of Venice, especially if the
boats are magnified by a slight effort of the imagination into sea-
going ships, or softened into gondolas. I have mentioned the
advantages which an artist might derive from residence in Oxford,
alike from the models that might be placed there, the architectural
beauties of the place, and the natural scenery. To the second of these
advantages would belong the excellent studies of interiors that some
of the rooms present. The rooms of one of my friends, which were
those at first intended for the Head of the College, are quite a gem in
the profuseness of decoration, especially as applied to the ceiling.
The halls of many of the colleges are also remarkably fine, as
presenting studies of interiors of peculiar magnificence. Occasionally
the internal decoration of the rooms themselves, in which individual
taste has perhaps taken a wider range than in any other place I know,
would assist a painter in his composition. Pictures and engravings,
profuse in quantity, if not always good in quality, decorate the rooms
of most of the junior members, and a marked improvement has of
late years taken place in this matter, engravings from good masters,
and really good original pictures by modern artists, having taken the
place of trumpery hunting-prints and portraits of the nymphs of the
ballet. Other rooms are hung round “with pikes, and guns, and
bows,” now obsolete, and seemingly made, at the time of their
construction, for this ulterior object of ornamenting a room, which
they fulfil so much better than any modern invention. But perhaps
the most extraordinary rooms of all are those of a friend of mine, in
one of the most picturesque colleges. The whole centre of his room is
taken up by a kind of immense Christmas tree, formed by his own
labour and ingenuity, on which is hung every imaginable article that
would be chosen in an old curiosity-shop from mere oddness in form
or nature. It is a rare collection of what the French call specimens of
“bêtises,” ironically, as I suppose, considering the extreme cleverness
which imagined them all. There are, if I rightly remember, gods from
the Sandwich Islands and fetishes from Africa, clubs from New
Zealand and bows from Tartary, stuffed birds, pipes of all kinds and
sizes, skins of snakes and crocodiles, skulls of men and animals, and
everything, in fact, that ever entered into a skull to devise. The walls
are papered with engravings, and engravings are hung from the
ceiling because there is no room for them on the walls. There is a
collection of divers plants, native or exotic, flourishing in stands or
trailing over the windows, in each of which is a kind of caravanserai
for wild birds (not aviary), for the amiable proprietor does not detain
them there longer than they wish to stay, but invites them in by
abundant proffers of their peculiar kinds of food; and as he sits or
reclines by his fire (for he has abundant facilities for assuming either
position) by the motionless silence which he purposely observes—has
constant opportunities of watching their flittings and hearing their
twitterings, and studying their little habits with the gusto of a
naturalist. That such an inventory, which entirely passes my memory
to describe, should have been amassed in a single room by any
amount of time and trouble, is a marvel to me, only to be explained
by the perfect and lotos-eating repose of a college life. Long may our
friend enjoy his quaint and instructive rooms! Travellers see strange
things, but few can say that they have seen stranger than those that
are enshrined in the colleges of Oxford.
You see that I have carefully abstained in what I have said from
making invidious comparisons between Oxford and the sister
university; nor have I spoken of the universities of the north, with
which I am but little acquainted, but which I should imagine to hold
an intermediate place between the English and the German system.
On the whole, it appears to me that the function of education,
comprising theology, philosophy, science, and belles lettres, is to
impress upon the mind images of Beauty and Truth, and to enable
the mind which has received these impressions to act in like manner
through life. If education cannot make a man’s actions truthful and
beautiful, he remains to the end a savage, or rather, I should say, the
scion of a vulgar civilisation, even if he knows all the poets by heart,
or can discourse with the acumen of an Erasmus or a Crichton. That
Beauty and Truth are one and the same in that perfect sunlight which
our eyes cannot see, and from which all lesser lights proceed, few will
deny. But here on earth they may be considered as in a measure
apart, and as exciting, each for good in its way, separate influences
on the moral life of man. Men incline to one or the other light
according to their natural bent or the bias of their education. It
seems to me that if a distinction is to be made between our
universities, the tendency of Oxford studies is to look at Truth
through Beauty, while that of Cambridge studies is to look at Beauty
through Truth. It is therefore that I have laid so much stress on the
capabilities of Oxford as a school of Art. I confess that I am anxious
to gain a closer insight into the nature and life of your German
universities. Probably they are with us but imperfectly and unfairly
understood. If it be true that the Bursch preserves, under his
outwardly rough exterior, any remains of that antique chivalry of
thought which is so fast dying out in this country, he preserves a
treasure which is of inestimable value, and which ought to be secured
to him at any price. At the same time, I think you will allow that our
system has certain superiorities of its own, which deserve at least
careful study, if not active imitation. We, at least, are successful in
affixing an ineffaceable stamp to the character of the great majority,
while you seem only to succeed in permanently impressing the
nature of a few, and impressing only a limited part of that nature.
May you live and lecture many years, Herr Professor; and may your
brimming Rhine flow on for ever, free and German as of yore; and
may the vine-blight spare the clusters that yield that molten gold
which, unlike the morbid production of Australia and California,
brings nothing but innocent joy to the soul of your Fatherland. Vale!
and believe me,
Your loving friend,
Tlepolemus.
 THE ANCIENT COINS OF GREECE.[5]

Father Hardouin, a learned French Jesuit of the seventeenth

century, lived to the venerable age of eighty-three years, and died, as
he had lived, in the full persuasion that the only authentic
monuments which we possess of classical antiquity are comprised in
coins, a few Greek and Latin inscriptions, with the Georgics of Virgil,
the Satires and Epistles of Horace, and the writings of Pliny and
Cicero. Out of these materials he held that certain ingenious
“falsarii,” in the thirteenth century, whom he styles the “architects of
annals,” compiled those multifarious productions of poetry and
prose which we have been accustomed to regard as a most precious
legacy bequeathed to us by ancient Greece and Rome. This fact we
mention to our readers, not with any view to shake them in their old
and orthodox convictions upon the subject, but simply to show them
what a vast amount of matériel this learned Father had discovered in
the study of ancient numismatics. A coin indubitably presents,
within the smallest compass, the fullest view of ancient times that we
possess. Though silent, it is always waiting to communicate
knowledge; though small, it is always ready to teach great things.
“Inest sua gratia parvis,” is the motto of the Cabinet. It would be
difficult, indeed, to say what department of ancient lore—whether in
mythology, or economics, or politics, or chronology, or geography—
may not be elucidated and explained by the study of coins. A series of
coins are, in fact, a series of illustrative engravings, of
contemporaneous date with the literary works of Greece and Rome,
and of the noblest school of art. We may realise much of what we
read by turning to designs executed by artists who lived in those very
countries, and at that very period. The lordly oak is uprooted by the
tempest, the lowly willow is spared. While the temples of the gods
and their concomitant myriads of statues have been reduced to
unintelligible fragments, those coins which formed the medium of
ordinary traffic—the tetrobolus, the soldier’s daily pay—the drachma,
that of the mariner—and the tetradrachmon, which, by virtue of the
archaic visage of Pallas, with her rigid smile, passed current among
merchants of every state and province,—these have remained safe in
their hiding-places under the soil, and may be found in nearly the
same condition in which the Greeks handled them more than two
thousand years ago.
Cities have been built with the express intent of perpetuating the
glory of a founder, and after all the founder’s intent is achieved, not
by the enduring testimony of edifices and streets of marble, but by
that of its coins. Thus the Emperor Augustus thought to immortalise
the fame of his victory over Antony and Cleopatra at Actium, by
erecting a city on the shores of the Ambracian Gulf, which city he
called by the appropriate name of Nicopolis. It was supplied with the
usual complement of public edifices; a gymnasium and a stadium
were built in a sacred grove in the suburb; another sanctuary stood
on the sacred hill of Apollo, which surmounted the city. It was
admitted by the Emperor’s desire into the Amphictyonic council, and
was made a Roman colony. Sacred games were instituted,
accompanied by a sacrifice and a festival, equal in dignity to the four
great games of Greece. Coins of the city were struck: and in
commemoration of a favourable omen which had presented itself on
the morning of the day of battle, a group of bronze statues,
representing an ass and his driver,[6] were placed, among other
dedications, in the temple of Apollo Actius.
Such were the forward-looking expedients of the conqueror to
perpetuate his fame;—and what has been the result?
“Look, where the second Cæsar’s trophies rose,
Now,—like the hands that made them, withering.”[7]

A long succession of ruined edifices, in one part converted into a

sheep-pen. In fact, before four centuries had elapsed, a
contemporaneous author tells us that the town of Nicopolis had
fallen into lamentable decay. The palaces of the nobles were rent; the
aqueducts crushed; everything was smothered with dust and
rubbish.—The bronze statues of Eutyches and Nicon, after being
removed first to Rome, and then to adorn the Hippodrome at
Constantinople, were at last melted down by the barbarous Latins on
their capture of the city in A.D. 1204. All is gone of Nicopolis except
the coins. The coins may be seen in the cabinet of the numismatist,
by time as yet uninjured; and we find upon one of them the head of
Augustus himself with the description of Κτίστης or founder, and the
appropriate figure of Victory holding a garland in her extended right
hand.
In connection with this city of Nicopolis, we may mention the fact
that one of the most important transactions in Colonel Leake s
diplomatic career—namely, a conference with the celebrated
Albanian Vezír, Ali Pasha, which led to the ratification of a peace
with the Porte in 1808—took place on the sea-beach, near the ruins
of the ancient aqueduct of the city, on a stormy night in the winter of
1807. The crafty Vezír, in order to throw dust into the eyes of the
French consul, who was watching the proceedings with much
jealousy, had previously got up a sort of scene in his presence,—
receiving an English messenger, whom he had himself instructed to
ask for permission to purchase provisions, with affected sternness,—
haughtily refusing to grant his request,—and declaring that the two
nations were still at war;—although he had already made with
Colonel Leake a private arrangement to give him the meeting that
same evening on the beach. As the day declined, the weather became
so threatening that the captain of Colonel Leake’s ship was afraid to
anchor off the coast; and so dark was the night, that had not Ali
himself caused muskets to be discharged, the appointed place of
rendezvous on the beach could not have been discovered. At length
the boat neared the land, and the Vezír was found seated under a
little cliff attended by one or two of his suite, and a few guards. Dr
Johnson might seem to have anticipated this scene, in his tragedy of
Irene, where he describes an interview between the Greek Demetrius
and the Vezír Cali in these words:—
“He led me to the shore where Cali sate,
Pensive, and listening to the beating surge.
There, in soft hints and in ambiguous phrase,
With all the diffidence of long experience,
That oft had practised fraud, and oft detected,
The veteran courtier half revealed his project.”[8]

During the two hours the conference between Colonel Leake and
the Vezír lasted, the surf rose considerably; and it was not without a
good drenching from the rain and the sea, and some difficulty also in
finding the ship, which they could hardly have done without the aid
of the lightning, that the boat returned on board. The ship then stood
away from the coast.[9]
But to return to our subject. Every one who feels a thirst for
knowledge, must value coins as the medium of acquiring knowledge:
every one who has an eye for grace and beauty, must value them as
presenting unrivalled specimens of grace and beauty: every one who
is susceptible of the charms of fancy, must love to study the hidden
meaning of those imaginative devices, which sometimes, as Addison
says, contain as much poetry as a canto of Spenser. Let not the study
be condemned as dry and crabbed, for Petrarch was a numismatist.
Let it not be condemned as connected with only a bygone and
obsolete school of art, for Raffaelle and Rubens, Canova, Flaxman,
Thorwaldsen, and Chantrey, delighted to refresh their powers by it.
Condemn it not as beneath the notice of the philosopher, for Newton
and Clarendon were among its votaries. Say not that men of active
pursuits can find no time for it, when you hear of the collections of
Wren, Mead, and Hunter.
There were numismatists among the ancient Romans. Admirers
and collectors, as they were, of the other productions of Greek art, we
should conclude that they were admirers and collectors of Greek
coins also, even if we had no direct evidence upon the subject.
Suetonius, however, expressly informs us that the Emperor Augustus
was accustomed—probably at the Saturnalia—to distribute among
his guests a variety of valuable and interesting gifts, and, among the
rest, pieces of money—not modern money, but of ancient date—not
Roman, but foreign; and some of it the coin of ancient kings. May we
not recognise in this description the beautiful coins of Greece and
her colonies—the coins of Syracuse and of Tarentum—of the
Seleucidæ and other Asiatic kings—of the kings of Macedonia,
Epirus, and Thrace? A facetious friend of ours professes to enrol
Horace also in the list of numismatists; and we have often smiled at
the mock solemnity with which he argues his point. He holds, for
instance, that the passage,
“Nullus argento color est avaris
Abdito terris”—
refers, not as we have been taught to interpret it, to the unwrought
silver lying hidden as yet in the mine, but to those choice productions
of ancient art—Syracusan medallions, for instance, or the rarer
tetradrachms of the Seleucidæ—which blush unseen in their
subterranean lurking-places, and are kept out of our cabinets by that
churlish miser the earth. And he holds that the poet very
consistently, in the same ode, assigns the regal diadem, and the
laurel crown of virtue, not to the man who is simply master enough
of himself not to covet his neighbour’s money-bags,
“Quisquis ingentes oculo irretorto
Spectat acervos,”

but rather to the noble self-denial of that numismatist, who can pass
from the contemplation of the well-stored cabinet of his rival without
one sidelong glance of envy.
And in that well-known passage where Horace says, in a rather
boastful strain, that the fame of his lyric poetry will be more durable
than bronze, our friend observes that if the poet alluded to the
statues of bronze which met his eye at every turn in the city of Rome,
it did not follow that his lyric fame would be of any long duration; for
of all articles of bronze the statue was doomed to the earliest
destruction, and but few, in comparison with the number of marble
statues, have come down to our time. Many a graceful figure which
Horace had seen and admired in the palace of Mecænas, for instance,
ere many centuries had elapsed was melted down by greedy
plunderers, and played its part a second time in the brazen caldron
of the housewife. But the medal of bronze survives the wear and tear
of centuries full a score. The medal it is,
“Quod non imber edax, non Aquilo impotens
Possit diruere, aut innumerabilis
Annorum series, et fuga temporum.”

Our observation has been drawn by some modern writers to the

supposed existence of a sacred character or quality in the coin of the
ancients. It is the opinion of the most experienced numismatists that
the Greek coin was invested with a character of sanctity, arising from
the head, or figure, or symbol of some deity which it usually bore;
that the ἐικών or image upon it was really and truly an idol. We
believe that such a notion prevailed, to a certain extent, both among
the Greeks and the Romans. Not that we regard the worship of Juno
Moneta as a case in point. We think that the worship of Juno Moneta
was the worship of a deity who was supposed to have admonished
the Romans that there are other things in the world much better
worth attending to than money, and that money would not be
wanting to them, so long as the weapons they fought with were the
arms of justice. At the same time, there was indubitably a reverence
paid to the coin, even down to the Roman times, for the sake of its
religious symbol or device. The people of Aspendus, in Pamphylia,
professed to hold in such reverence the effigy of the Emperor
Tiberius upon his coin, that they found a certain fellow-citizen guilty
of impiety, simply on the ground of his having administered a little
wholesome chastisement to a refractory slave who happened to have
at the time one such coin in his pocket.
It has been thought that the practice which prevailed among the
Greeks, of placing a piece of coin in the mouth of the corpse,
originated in this notion of its sanctity, inasmuch as it was supposed
to insure the protection of the deity, whoever it might be, to whom
the coin was attached by the symbol it bore. But we must confess
that, for our own part, we still cling to the old story of the fee
required by the Stygian ferryman. Hercules informs Bacchus, in the
Ranæ of Aristophanes, when he is meditating a visit to the shades
below, that he will arrive at a wide unfathomable lake, and that an
old man who attends for the purpose will ferry him and his
companion across it, on receiving the fee of two oboli. Lucian, too,
has a joke about Charon’s complaining that, in consequence of the
slackness of his trade, he cannot raise money enough to supply the
necessary repairs for his boat. The mouth was so commonly used as a
purse by the Greek in his lifetime, that we can scarcely wonder at this
method being adopted for his carrying money into the other world
with him when dead. Colonel Leake mentions the discovery of a coin
of Motya in the mouth of a skeleton in the island of Ithaca, in a tomb
of the first century before Christ.
At the same time, although we believe that the myth of Charon was
more closely connected with this practice in the minds of the
common people than any other consideration, we doubt not that the
sanctity of the coin was also taken into account. We find that notion
of sanctity prevailing, not only among the Greeks and Romans, but
among other nations, to a considerable extent. The Mohammedan
coin bears invariably a passage from the Koran, or some other
religious text, quite sufficient to insure its reverential treatment by
the faithful Mussulman; and we read in Marsden’s Numismata
Orientalia of a certain class of very rare gold coins of ancient date, to
which the Hindoos avowedly paid religious worship. Of this coin the
Rajah of Tanjore was so fortunate as to possess two specimens.
Whether the sect of gold-worshippers is yet extinct is a question
which we must leave moralists to settle among themselves. It has
been remarked by an accomplished scholar and excellent
numismatist,[10] that “gold has been worshipped in all ages without
hypocrisy.” That there were many in ancient times who held the coin
in reverence for the sake of an indwelling sanctity connected with the
symbolic representations which it bore, we fully believe; and that
there may be some in modern times who hold it in reverence,—
ἀισχρου κέρδους χάριν,—we are by no means disposed to deny.
There is no doubt that pieces of antique coin have been frequently
carried in the purse or in the pocket as a sort of charm or amulet; but
we question whether this notion of their supernatural power has any
connection with the supposed sanctity of the legends or symbols with
which they are impressed. We should ascribe it rather to the same
feeling which induces some old women, and young ones too, to carry
a crooked sixpence in their purse—the charm being supposed to
reside, not in any device or legend of the coin, but simply in its
curvilinear shape. So in the cases we have just alluded to, the charm
lies in the mystery of the coin’s unknown and ancient origin—“omne
ignotum pro magnifico est.” Stukeley tells us that, in the
neighbourhood of one of the ancient Roman sites which he visited in
his “Iter Curiosum,” Roman coins were known among the peasantry
by the appellation of “swine pennies,” from the fact of their being
often turned up by that indefatigable excavator in his search after
something more succulent. To the mighty Cæsars this was truly a
degradation. But at Dorchester he found the same coins known by
the name, assigned with more semblance of respect, of “Dorn
pennies,” after some mythical king Dor, whom tradition states to
have once resided there. The rustic antiquary is wont to labour under
a sad confusion of ideas. The Roman he confounds perpetually with
the Roman Catholic. We remember ourselves—after visiting a sort of
bi-linguar monument near Hadleigh in Suffolk, which marks the spot
of the martyrdom of Dr Rowland Taylor, under Queen Mary—to have
asked a passer-by whether a certain antiquated mansion by the road-
side had ever been inhabited within his recollection; to which we
received the oracular reply that, to the best of our rustic friend’s
belief, it had never been inhabited since the Romans occupied it, in
the days of Dr Taylor!
This, however, is rather a digression. We learn from Trebellius
Pollio that, in the fourth century, the coins of Alexander the Great
were supposed to insure prosperity to any person who was prudent
enough to carry one of them constantly about his person; and we find
this, and all other such notions, strongly condemned by Chrysostom.
An Italian traveller tells us that, in 1599, the silver coins found in the
fields in a certain district in the island of Crete were called by the
people after the name of St Helen; and that the story went that this
saint, being in want of money, had made a number of coins of brass,
endowing them, at the same time, with such miraculous properties,
that the brass, in passing into the hands of another person, was at
once changed into silver; and, moreover, that any such silver coin
being held fast in the hand, will cure the falling-sickness. Mr Pashley,
who visited Crete in 1830, found that the possession of an ancient
coin is looked upon as a sovereign charm against maladies of the
eyes. In the year 1366, the discovery made by some children at play
of a number of ancient coins, at Tourves, near Marseilles, threw the
whole community of the district into a state of alarm and
consternation. The coins were some that had been struck at
Marseilles at that early period when, under the name of Massalia, it
ranked among the most thriving colonies of ancient Greece. They
bore on the one side a head of Apollo, and on the other a circle
divided into quadrants. In the chronicles of Provence, where this
discovery is recorded, they are described as bearing on the one side a
Saracen’s head, and on the other side a cross. This was interpreted
as bearing some portentous allusion to the Crusades. And the devout
writer intimates that, while one part of the community look upon it
as an omen of good, and the other part as an omen of evil, Heaven
only knows how it will turn out.
 We believe that some persons, sedulously devoted to other
branches of the study of classical antiquity, are deterred from
availing themselves of the aid of coins, by a fear of being imposed
upon by forgeries. This is an easy, but an idle mode of putting aside
that which we have not courage to investigate. We shall add a few
remarks upon the subject.
In the first place, we shall venture to ask these anti-numismatic
sceptics, whether they think we ought to cease to read and to admire
the dramas of Shakespeare, because it is questionable whether one or
two of those which pass under his name were really of his
composition?—or, whether we shall shut our eyes before all pictures
which pass under the names of the Old Masters, because spurious
ones have been palmed off upon the self-dubbed connoisseur?—or
whether all autographs of illustrious men are to be condemned as
trash, because Ireland attempted to impose upon the public with
some that were not genuine?—or whether all currency is to come to
an end, because clever knaves have succeeded in counterfeiting it?
Everything, in short, which is valuable, offers, in proportion to its
value, a temptation to ingenious and unscrupulous men to show
their cleverness by imposing upon the world with an imitation of it.
The Holy Scripture itself has not escaped.
And after all, in regard to coins as well as in regard to the other
subjects which we have mentioned, although forgers may be clever,
detectors are clever also. The numismatic phalanx of investigators
are more than a match for the “falsarii.” The skill of Cavino,
Gambello, and Cellini, has been met with equal skill on the part of
the numismatist. The eye that has been accustomed to wander over a
well-selected cabinet acquires a power of ready discrimination,—a
power difficult to teach by theory, but not so difficult to gain by
practice. Solitary instances may occur of a solitary numismatist
fondly persuading himself that some clever forgery which he
possesses is a genuine coin, but we would not give much for his
chance of beguiling others into the same belief. Unwilling he may be
to have the “gratissimus error” extracted from his own mind, but he
never will succeed in engrafting it upon others. Never does the eye of
man exert so much jealous vigilance as when it is employed upon the
coin of a rival numismatist claiming to be genuine upon insufficient
grounds, The House of Lords sitting upon a claim of some peerage in
abeyance is nothing to it. We apprehend that scarcely an instance is
on record of a forged coin having enjoyed for any length of time,
unquestioned, the honours of a genuine one. Nor do we think that
there are many instances of a forger’s attempting to falsify history.
He generally aims at making his invention tally with historical fact as
closely as he can. And if his inventive powers are not at all brought
into exercise, but he simply produces a coin which is a fac-simile or
reproduction of a genuine one, for purposes of study that fac-simile
will be equally available with the genuine coin, and no further harm
is done than the abstraction of a few shillings more than its value
from the pocket of the unwitting purchaser.
At the same time we would not let the forger go unpunished.
Though the evil actually done be small, the intention is bad. We
would have him tried by a jury of numismatists. Or if the offence
should have been committed in a country where the power of
punishing the offence resides in one magistrate, we should say that
that one magistrate ought to be a numismatist. It is said that a
distinguished archæologist who possessed this power in virtue of his
office as Her Majesty’s consul at Bagdad, very recently exercised it by
directing that a Jew “falsarius” should be bastinadoed. We applaud
his Excellency’s most righteous judgment. The man who had
counterfeited the famous sequins of Venice, and had aggravated his
crime by doing it badly,—
“Che male aggiusto ’l conio di Vinegia,”

is represented by Dante as worthy of an especial notice among those

sinners against laws divine and moral with whom he has peopled the
shades of his Inferno.
Seriously, however, we think that any clever work of art is worthy
of being preserved, and none the less for its having taken in some
who set themselves up as judges. Even in Pliny’s time a counterfeit
denarius of superior workmanship was sometimes thought cheap at
the price of sundry genuine denarii. The tasteful device of Cellini, or
of some cunning artist of Padua, must not be thrown to the dogs,
merely because it was produced with the intention of rivalling the
work of ancient artists, and of testing the acumen of the cognoscenti.
Those figures of Cellini, for instance, which some one brought and
exhibited to the artist himself as antiques, and respecting which the
nobleman who was their proprietor declared, when he saw a smile
playing upon the conscious visage of Cellini, that there had not lived
a man for these thousand years who could have wrought such;—
would not those figures have been worth preserving? And in like
manner a coin which, by the excellence of its workmanship, has
raised a doubt whether it may not have been really of ancient origin,
ought by no means to be treated with contempt, even though it
proves to be modern.
The learned work of Colonel Leake, now before us, has supplied a
desideratum in the archæological literature of our country. It is the
first work of the kind upon Greek coins which has been published by
an Englishman, and those of our readers who are acquainted with his
character will agree that no Englishman could have been found to do
it so well as Colonel Leake. The vast amount of knowledge which he
has been laying up for more than half a century, in regard to the
literature, the mythology, the political and social history, and the
geography of ancient Greece, supplies an infinity of streams which
flow over the pages of his work in the form of notes. No longer shall
we blush under the well-grounded reproach that all the standard
works upon Greek coinage are written by foreigners. Already, indeed,
we observe that Professor L. Müller, in his Numismatique
d’Alexandre, just published at Copenhagen, has made ample use of
Colonel Leake’s volume, which must necessarily become a text-book
in this branch of Greek archæology. For the convenience of those
who may consult it, not only is every ordinary variety of index
supplied to the coins themselves, but we observe that, in an
appendix, an index is added to the valuable information contained in
the notes. We observe, also, in the appendix, a very interesting and
learned dissertation upon the weights of Greek coins, in which
Colonel Leake traces the Attic didrachmon—which seems to have
been a sort of standard or unit in the monetary scales of Persia and
Lydia, as well as of the cities and colonies of Greece—to Phœnicia,
and from Phœnicia to Egypt. It would scarcely be in accordance with
our usual practice to enter into the more erudite part of this
important subject, and we shall therefore conclude our remarks by
making one reference to the work, in order to show how successful
its author has been in availing himself of the light which a coin may
throw upon the more obscure portions of ancient geography.
 In Colonel Leake’s collection there is a coin, recently brought to
light, of a people called the Orthians, bearing the Thessalian type of a
horse issuing from a rocky cavern, in allusion to the story that
Neptune produced the horse originally by a stroke of his trident upon
a Thessalian rock. Now a city, called “Orthe,” is mentioned by Homer
in the second book of the Iliad.[11] With regard to the site of this city,
there was a difference of opinion among geographers even in
Strabo’s time; the majority seem to have identified it with the
acropolis of a more modern city, which at that time was known by
the name of Phalanna. But inasmuch as there are coins now extant of
Phalanna, and of a date contemporaneous with that of Colonel
Leake’s coin of Orthe, it is evident that Phalanna and Orthe were two
separate and distinct places. The appearance, therefore, of this
previously unknown coin of Orthe corrects an error which prevailed
among geographers as far back as the time of Strabo. It shows that
Phalanna and Orthe were not the same place. Out of the five cities
mentioned by Homer in this passage, Strabo had well ascertained the
position of three; and Colonel Leake is now enabled to fix the
probable position of the fourth. In reference to such facts as this,
Colonel Leake observes in his preface that they have an important
bearing upon the great question as to the origin of the Homeric
poems.

“It seems impossible,” he says, “for any impartial reader of the Iliad, who is not
seeking for arguments in favour of a preconceived theory; who visits the scene of
the poem; and who, when making himself acquainted with the Dramatis Personæ
in the second book, identifies the sites of their cities, and thus finds the accuracy of
Homer confirmed by existing evidence,—to believe that no such city as Troy ever
existed, and that the Trojan war is a mere poetic invention; this, too, in defiance of
the traditions of all antiquity, and the belief of intelligent historians, who lived
more than two thousand years nearer the event than ourselves. The Iliad differs
not from any other poetical history or historical romance, unless it be in the great
length of time which appears to have elapsed between the events and the poem;
but which time was employed by an intelligent people in improving and perfecting
their language and poetry—in committing, by the latter, past occurrences to
memory; and the principal subjects of which, therefore, could not have been any
other than religious and historical.”
 The study of coins has been very much facilitated by recent
improvements in the art of electrotype, which now enables the
collector to obtain perfect copies of the rarer and more costly
specimens, and to render them as useful to art and literature as the
originals themselves. For purposes of reference we have a noble
collection in the National Museum, as well as another which,
although of much more limited extent, is nearer to ourselves, and
therefore more accessible to students on this side of the Tweed, at
Glasgow. In the concluding paragraph of his preface, Colonel Leake
mentions these two collections in connection with each other; and
with that paragraph we shall also conclude our remarks upon his
valuable work.

“Augmented as our National Collection has been by the bequest of Mr Payne

Knight, by the purchase of the Bargon Collection, and by similar acquisitions on
the dispersion of the Devonshire, Thomas, and Pembroke cabinets, it now rivals
most of those on the Continent. With the addition of the Hunterian at Glasgow,
which the Trustees of the British Museum have now, at the end of eighty or ninety
years, once more the opportunity of acquiring, with the assistance of Government,
it would be the richest in Europe.”
 TICKLER AMONG THE THIEVES!
EXTRACT FROM AN AUTOBIOGRAPHY, WITH A PREFATORY
NOTICE.

Poor Tickler! The thing happened in this wise.—But, by the way,

before coming to that, it may not be amiss to give the reader some
idea who Tickler is:—to wit, a very Skye of Skyes, with a mouth the
roof whereof is dark as midnight: his glittering eyes are black as jet;
his ears short, his legs none of the longest, but his body is: his tail is a
triumph, when fairly spread out; and as for the strength with which it
is attached to his body, you may hold him up by the aforesaid tail as
long as you can—with one hand. Then his hair is pepper-and-salt in
hue, long and curly, and—if I may so speak (though no one but
myself and the family will know exactly what I mean by it)—with a
kind of silken wiriness. And as for cleanliness, why, he is washed
thoroughly every Friday morning, and carefully combed afterwards;
and the recurring day of that jobation (to use a word of his own) he is
as perfectly acquainted with as the gentleman who performs the
operation, and has come, in process of time, even to like the thing:
witness how he jumps into the tub of warm water of his own accord,
alike in winter and summer, with a kind of alacrity. He makes no fuss
about it, except that sometimes, when the soapy water gets into his
eyes, they wink at you in silent suffering, which he unconsciously
aggravates, instead of alleviating, by putting up his wet paw to rub
them! Through this operation he has gone for now nearly twelve
years, and a sweeter dog there is not than Tickler. I may indeed
almost say as much in respect of his temper, which is excellent
whenever he has everything his own way. I have reflected a good deal
on the dog’s idiosyncrasy, and think I now know it well. ’Tis tinctured
by a warm regard for himself, with respect to the good things of this
life; he says, reasonably enough, that if there are good things to be
had, he cannot think why he should not try to get them, and like
them, since he is formed for the purpose, if he can get them; and as
for huge or little hungry dogs in the street, of the plebeian order, he
does not dislike to see them enjoying themselves, by way of giving a
zest, as it were, to starvation,—if he have no fancy himself for what
they have routed out of the gutter. He says he thinks they must often
be sore driven; for he has sometimes seen a gaunt dog crunching a
dirty bone till he has actually almost eaten it! I am sure Tickler is not
without feeling; for one day he was sitting on a chair, with his paws
resting on the top of it, near the window, in a warm dining-room, on
a blighting day in February—the dust-laden wind without seeming to
cut both man and beast to the very bone: and at the foot of our steps
there had presumed to sit a dirty half-starved cur, shivering
miserably in every muscle, but uttering no sound—neither whine nor
bark.
“He starved, and made no sign!”

Was it necessary for that lout of a fellow that passed, to kick the
unoffending brute (which did not belong to him) from our steps, it
showing, however, no resentment, but simply sitting and shivering a
foot or two farther on? Then Tickler (who is of patrician descent),
whose eyes had been for some time fixed wistfully upon his plebeian
brother, could hold his peace no longer, but gave a loud, fierce, little
bark, jumped down from his chair, and fawned whiningly on me; and
when I took two nice chicken-bones from his plate under the sofa,
and called the forlorn victim of man’s chance brutality into the hall,
and gave him the bones, which he was for a while too cold, and also
timid, to eat for fear of another kick,—Tickler stood by, not only
without growl or bark, though he knew the victuals were his, but very
complacently wagging his tail. He had pity for his poor brother, who
seemed such a wretched little outcast! And as for the poor voracious
creature before him, crouching guiltily as if he had done wrong in
enjoying himself, we could hardly find it in our hearts to put him out
again into the street. If he could have carried away sixpence to a
tripe-shop, he should have had it to get a complete feast for once in
his life. I think the incident made a deep impression on Tickler; for
when he returned into the dining-room, he went again to the
window, and sate for some time looking through it wistfully, and
whining; and then jumped down, went under the sofa, and lay there
for upwards of two hours, sighing several times, and without
touching his victuals.
But, on proper occasions, Tickler could show a proper spirit. We
have a cat; and if there be any force in the new saying, the right cat in
the right place, Tickler was the dog to insist on its being observed; for
if ever poor Tom presumed to steal up-stairs out of the kitchen
(which, it must be owned, was his proper place), there was no end of
uproar on the part of Tickler; though Tom would sometimes turn
round, on his way down stairs, and, curving up his back, and
showing his teeth, glare at his little tyrant with an expression that
was perfectly fiendish; and tended, moreover, effectually to keep the
right dog in the right place, viz. the dining-room, to which he would
on these occasions retreat in good order, perhaps, not without
needless delay. Thus Tickler had a notion of fitness.
He was also of a very contemplative character, shown by his long
sittings on the chair nearest one of the windows—in fact, always the
lefthand side window. He would sit on the chair, with his fore-paws
resting on the top of it, and his mouth between them, calmly
surveying so much of human nature as passed before our windows. It
would have been strange, indeed, if he could have lived so long with
us,—growing up with our children, and growing old alas! with
ourselves,—without having endeared himself to us all in a hundred
different ways, and becoming thoroughly familiar with our ways and
habits. Can any one persuade me that the little fellow did not know
6.30 P.M. o’clock, at which hour I pretty regularly returned to dinner,
when he used always to take his seat on his chair a quarter of an hour
before that time, with his jet-black nose and watchful eyes pointed in
the direction in which I always came; and when I approached the
steps, he would leap down and bark like mad, till the dining-room
door was opened,—and then the front door? And how he jumped up
against my legs, when I entered, and scampered wildly to and fro! I
know he liked me, and “no mistake,” as the Great Duke said. But
besides this, I am morally certain that he always knew the Sunday
morning. Even as early as breakfast-time, he was grave and
restrained, looking as though he knew that there was something or
other in the wind; and when we severally went out, he made no
indecent and clamorous attempts to accompany any of us, but lay
looking solemnly at us, as we respectively took our departure—and as
soon as we had all gone, he invariably went up to his bed, which was
under our own, never stirring till we returned; and who shall tell
what he was thinking of on such occasions? Did he sleep, dream?
That he does dream, no one knows better than I; for he talks—I beg
pardon, barks—in his sleep almost every night, often waking me from
my own dreams. But what has particularly pleased me in Tickler is,
that when I sit up after everybody else is gone to bed, he has, for
years, voluntarily remained with me, however long I may remain. I
wheel an easy-chair (my wife’s) towards the fire as soon as we are left
alone, he waiting for it quite as a matter of course, and jumping into
it, immediately turning round, slowly and thoughtfully, three or four
times, and then settling down into what he at length, I presume,
conceives to be a comfortable position—his mouth resting on his
paws, and his eyes fixed on me, till he falls asleep, with one eye open.
Bless his little soul (for something of that sort he assuredly has)—
how well I recollect one night, soon after Madame and the young
ones had retired, taking out of my pocket a hard-hearted and
insulting letter received during the day—laying it down after reading
it, with a sigh, and then gazing affectionately at my faithful Tickler,
whose watchful eyes were fixed all the while on me! Ay, my little
friend! this would try your temper; but dogs are mercifully spared
such anxieties, although you have your own sensibilities! In a long
series of years, I have sate up many hours engaged on my great work,
in seventeen folio volumes, entitled, The Essence of Everything from
the Beginning; and if it please Heaven to spare my life to finish it, I
undertake that it shall finish the reader. Well, it has been such a
comfort to me, night after night, every now and then to watch Tickler
watching me, as I cannot describe; and I do believe he has
contributed, whether consciously or unconsciously, to divers fine
ideas of mine—at least I think them fine, and tranquilly await the
judgment of the critics, or such of them as shall survive to see my
great work, and, above all, survive the reading of it. How snug he has
made me feel, with my huge easy-chair exactly opposite his smaller
one (which is my wife’s till she goes to bed), my table and one or two
chairs covered with books, the crimson curtain drawn close, and the
fire crackling briskly; many and many a time have I been inwardly
tickled by seeing and hear him dreaming, his breathing quickened,
and his bark short and eager, but suppressed. I am certain that he
sometimes has nightmare! How pleasantly we used thus to keep one
another company in the winter nights! When my work was over,
often not till two and even three o’clock in the morning, Tickler had
notice thereof by the act of shutting up my desk, till which moment
he never stirred; but that done, and before I had extinguished my
candles, he descended from his chair in a leisurely way, and yawned
and stretched himself; I often holding him up by his tail, just to let
him feel that all was right, and that he was really awake. Then we
both crept up-stairs to bed, as quietly as possible, lest we should
disturb the sleeping folk. And if I should happen to have to go down
stairs again to look at a book, or bring up my watch left on the table,
Tickler seemed to feel it his duty to get out of his snug bed, and come
pattering softly down stairs at my heels.
He was almost as vivacious as ever, though twelve summers had
passed over him at the period of that serious adventure which is
presently to be laid before the admiring reader. But no amount of
vitality has sufficed to prevent Mr Tickler’s face getting white; so
that, when he is in his lively humours, he suggests to my mind the
funny face of a frolicsome little elderly man, or a dog who had
plunged his nose into a flour-bag. I took him with me last autumn to
a place which I described, but without specifying, as may be seen in
the October and November numbers of Maga,[12] and the trip did him
a world of good. Do you recollect something that befell me there?
viz., that I lost him for a while, to my grievous discomfiture and
painful exertion—finding at last that the sweet little rogue was not
lost at all, but squatting comfortably on our drawing-room sofa?
How little I dreamed, however, that this might be deemed the
shadow cast before, of a coming event—a loss of Tickler!! in right
earnest? Only the very midnight before this startling occurrence he
was sitting in his old place, about twelve o’clock, opposite to me and
the table, whereon lay a portion of the stupendous accumulation of
MSS., through which I was patiently distilling off The Essence of
Everything. I got up from my seat and yawned with a sense of
weariness, when he did the very same thing, and thereby attracted
my attention to him. So I sate down beside him, and, tickling his
ears, said, “Ah, you little runaway! A pretty wild-goose chase you led
me at——!” on which he wagged his tail, and smiled: but no one can
tell a dog’s smile that has not studied his countenance as I have
Tickler’s. The next morning I lost him in right earnest—in dreary
earnest! He left our house at 10 A.M. on Monday the 4th December, in
company with a steady middle-aged servant, almost as much
attached to him as we were ourselves, and who had come down on an
errand to me—but having left with Tickler, he arrived at the place
where I pass most of my day-time, without his better half. “I
thought,” said I, on my arrival, and finding him sitting in the ante-
room, “that you were to bring Tickler with you, for a walk?”
“So I did, sir, but I’ve lost un, sir, I’m afraid,” he replied stolidly.
“Lost Tickler!” I echoed in consternation.
“Yes, sir. Missed un in a moment, like, and couldn’t vind un
anywhere!”
“Why, when did you leave our house, sir?”
“Just as the clock struck ten.”
“And now it’s not quite half-past!! What upon earth were you
about not to stop and look for him?”—Suffice it to say, that he
described himself as having suddenly missed Tickler, who had been
following as usual close at his heels, when at only two streets’
distance from our house,—had consumed five minutes in looking for
un—and then came quietly down without him, to me! He said he
thought the dog might have returned home “of his-self! as he had
done at ——!” I was disposed for a while to entertain a very particular
view of this strange transaction, but in the mean time sternly
despatched the delinquent back, at top speed, to acquaint our family
with the loss of Tickler; and also sent a trusty messenger after him, in
the forlorn hope that Tickler might have returned home “of his-self.”
Nothing of the kind; he was gone, poor little fellow, in earnest: and
as he wore his collar, with my name and address in full engraved
thereon, it was plain that unless he quickly made his appearance, he
must have experienced the professional attentions of a very vigilant
class of London practitioners. Every member of my family spent the
rest of the day in scouring the neighbourhood, especially the more
dubious (i. e., discreditable or suspicious) portions—but in vain. Our
baker, whom Tickler used to visit on business every day, saw him
walking past the shopwindow, alone, and at a leisurely pace, within
about ten minutes of the time of my servant’s missing him—but
supposed, as a matter of course, that he was in attendance upon
some member of the family! Inquiries were made of all our
tradespeople—only to be answered by exclamations—“What! Tickler
gone? poor little thing, we loved him like a child!” “He can’t be far
away—you’ll be sure to see him by nighttime, in particular as he had
his collar with his master’s name;” “and, ma’am,” added one more
sagacious than the rest, in a mysterious whisper—“if you don’t—why,
in course! he’s been stole!” “He was the hamiablest of dogs—so
petecler well bred!” “Oh, you see, Miss! he’ll be sure to come back!”
Then we betook ourselves to the Police Station; where the courteous
inspector, having listened to us, said, with a quiet oracular air, “He’s
not far away; he’s taken of course for the reward, and as he had his
collar on, they know where to find you when they choose. Is he an old
or young dog?” “He’s in his thirteenth year!” “Oh, then, you’ll have
him back very soon; the dog-stealers are knowing fellows, and he
won’t do. But take my advice—advertise him in to-morrow’s Times,
and offer only one pound reward, and be sure to add, no further
reward will be offered.” This we did; and the next morning appeared
the following public indication of our calamity, drawn up by my own
masterly pen, and all out of my own head: “Dog Lost. One Sovereign
Reward. On Monday the 4th inst., between —— and ——, a pepper-
and-salt Skye terrier, answering to the name of Tickler. Collar round
his neck with,” &c. “inscribed on it. To be brought to that address. No
further reward will be offered.” Having dropped this our little line
into the huge water of the Times advertisement sheet, we awaited a
nibble with such patience as we could command. But we got no
nibble at all, and very dull our house seemed, without our merry and
sagacious little Skye friend. Why, there was not a room in the house,
or a chair or sofa in it, that did not remind us of him; and as for my
wife’s little easy-chair opposite mine, when she had gone to bed, and
was no longer succeeded by Tickler, I wheeled it into the corner of
the room, and did not write at my Essence with anything like my
former satisfaction or spirit. The advertisement in the Times had
explained our disaster to all our friends; and no one called on us that
did not ask, “Well, any news of Tickler?” or say, “Poor little fellow,
how you must miss him!” At length an exceedingly knowing person
came, and said, “Have you been to ——’s? You can’t do anything
without him; he knows all the respectable dog-stealers in London,
and enjoys their confidence.” So my wife and daughter went to him
the next day; and following his advice (given after a minutely
accurate description of Tickler), I inserted in the particular
newspaper which he said was likely to be read by the parties
concerned, the following advertisement, which no false modesty shall
prevent my owning to be, in my opinion, a choice morsel of
expressive pithiness: “Tickler.—One sovereign reward, and no
more, will be paid for the recovery of a pepper-and-salt Skye Terrier,
answers to the above name, and lost near ——, on Monday the 4th
instant. Had on a collar, with the words,” &c. &c. “In its 13th year,
and many teeth gone. To be brought to the above address.” It grieved
me thus to publish to the world poor Tickler’s age and infirmities;
but needs must, when a certain Jehu drives:—and the way in which I
vindicated my advertisement against the reclamations of all Tickler’s
friends was the following: If I show the thieves that I am quite wide
awake to the poor little dog’s age and infirmities, it may certainly be
no news to those gentlemen, so experienced in those matters, but
will, peradventure, add force to the three pregnant words in italics in
the above advertisement, “and no more.” The more candid of my
opponents said that there was something in this; but they held that I
had, nevertheless, greatly hurt Tickler’s feelings, if ever he came to
hear of it. The more long-headed of my friends went so far as to say,
besides, that it was, after all, a toss-up whether I ever got him again!
Now comes a remarkable occurrence, and the reader may depend
upon its being told him exactly as it occurred, viz., that on my
returning to dinner, one day, a strange Skye terrier presented himself
to me, on entering our dining-room. He had followed home two
young ladies in the neighbourhood, who took him to be our dog, of
the loss of whom they had heard. So they brought him to us; and on
our saying that it was not Tickler, they left, followed by the stranger,
but refused to allow him to enter their house. Now it was a blighty
December afternoon, and this poor Waif and Stray sate outside their
door shivering in the cold: so our servants got leave to bring the poor
thing into our house, to be taken care of as a sort of locum-tenens of
poor Tickler. The Stranger behaved so well, and had so many nice
little tricks, that we all were satisfied he was a gentleman’s or lady’s
dog, and we began, in spite of ourselves, to like him very fast: for his
face reminded us of Tickler a good deal; but on a more narrow
investigation of Stranger’s pretensions to our affections, it was
discovered that he was not thorough-bred, as testified by the mottled
roof of his mouth; and also in respect of his configuration, he seemed
not like a canine homogeneity, but as it were two dogs joined