Diagnosis of Endometrial Biopsies and Curettings_ a Practical Approach ( PDFDrive )

Diagnosis of
Endometrial Biopsies
and Curettings
A Practical Approach
Tricia A. Murdock
Emanuela F. T. Veras
Robert J. Kurman
Michael T. Mazur
Third Edition
123
Diagnosis of Endometrial Biopsies
and Curettings
Tricia A. Murdock • Emanuela F. T. Veras
Robert J. Kurman • Michael T. Mazur
Diagnosis of
Endometrial Biopsies
and Curettings
A Practical Approach
Third Edition
Tricia A. Murdock Emanuela F. T. Veras
Department of Pathology Department of Pathology
The Johns Hopkins Hospital The Johns Hopkins Hospital
Baltimore, MD Baltimore, MD
USA USA
Robert J. Kurman Michael T. Mazur

Department of Pathology Department of Pathology and
The Johns Hopkins Hospital Laboratory Medicine, State University
Baltimore, MD of New York Upstate Medical University
USA Syracuse, NY
USA
ISBN 978-3-319-98607-4 ISBN 978-3-319-98608-1 (eBook)

https://doi.org/10.1007/978-3-319-98608-1
Library of Congress Control Number: 2018960871
© Springer Nature Switzerland AG 2019

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or
part of the material is concerned, specifically the rights of translation, reprinting, reuse of
illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way,
and transmission or information storage and retrieval, electronic adaptation, computer software,
or by similar or dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are
exempt from the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors, and the editors are safe to assume that the advice and information in
this book are believed to be true and accurate at the date of publication. Neither the publisher nor
the authors or the editors give a warranty, express or implied, with respect to the material
contained herein or for any errors or omissions that may have been made. The publisher remains
neutral with regard to jurisdictional claims in published maps and institutional affiliations.
This Springer imprint is published by the registered company Springer Nature Switzerland AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
Preface
The third edition of Diagnosis of Endometrial Biopsies and Curettings: A

Practical Approach developed from positive comments received from the
second edition and recognition that several topics have advanced since the
last publication. The emphasis of this book is to provide a foundation for a
daily, pragmatic approach for common entities found in endometrial samples.
Our understanding of endometrial pathology has evolved, and clinical termi-
nology accompanying these specimens also has changed. Therefore, the
entire text has been updated; all illustrations are now in color.
As with the first two editions, the focus of this book is to assist the patholo-
gist in the routine evaluation of endometrial tissue specimens. The book is not
intended to be an all-encompassing text of endometrial pathology but to
guide the pathologist through the more common diagnoses such as artifacts,
benign entities, precursor lesions, and neoplasia. We recognize that for cer-
tain areas, such as precursors of endometrial cancer, there is ongoing debate
and study regarding the terminology for these lesions. Different investigators
and practitioners prefer different terminologies (hyperplasia/atypical hyper-
plasia versus hyperplasia/endometrial intraepithelial neoplasia [EIN]), and
we therefore attempt to provide a balanced discussion of both so readers can
decide which they prefer. For the majority of the text, the terminology and
classification schemes mirror the most recent edition (2014) of the WHO
Classification of Tumors of Female Reproductive Organs in an effort to pro-
mote uniform and comprehensive communication with our clinical col-
leagues as ultimately this is the most important role of the diagnostic surgical
pathologist.
In the second edition, immunohistochemistry (IHC) was largely discussed
in the final chapter on methodology, but as its use in diagnosis has greatly
expanded, the current edition incorporates IHC into each of the individual
chapters. Furthermore, IHC now has an important role as a surrogate marker
of genetic aberrations, knowledge of which is becoming increasingly impor-
tant for diagnostic and therapeutic purposes. For example, the use of fumarate
hydratase staining to detect fumarate hydratase-deficient leiomyomas can
lead to the recognition of the hereditary leiomyomatosis and renal cell carci-
noma syndrome.
Although molecular diagnostics currently are infrequently used in routine
evaluation of endometrial specimens, pathologists need to be familiar with
the advances in molecular genetics as this is transforming our understanding
of many of the pathologic conditions that can be found in endometrial
v
vi Preface
s amples. For example, the classification of endometrial stromal tumors has

been modified based on fluorescence in situ hybridization and targeted RNA
sequencing that have demonstrated different types of high-grade endometrial
stromal sarcomas depending on their molecular genetic features. Another
example is that through molecular genotyping, we are now able to defini-
tively diagnose partial hydatidiform moles, whereas in the past morphologi-
cal, immunohistochemical, and cytogenetic studies were insufficiently
reliable to make a definitive diagnosis.
In summary, we hope that this new edition continues to be a foundation for
the diagnosis of endometrial specimens with a strong emphasis on the impor-
tant morphologic features and that this will be useful to pathologists and
gynecologists.
Baltimore, MD, USA Tricia A. Murdock

Baltimore, MD, USA Emanuela F. T. Veras
Baltimore, MD, USA Robert J. Kurman
Syracuse, NY, USA Michael T. Mazur
Acknowledgments
We are grateful to Norman Barker, a friend and colleague, who provided his
expertise in the field of medical illustration and tirelessly assisted us with the
images depicted in this edition. We are also grateful to Dr. Lora Ellenson,
another good friend and esteemed colleague, for generously supplying
updated information that helped enrich the contents of this book.
vii
Contents
1 Introduction�� 1
Indications for Biopsy�� 1
Clinical History and Biopsy Interpretation�� 2
Abnormal Uterine Bleeding�� 2
Infertility Biopsy�� 4
Products of Conception�� 4
Hormone Therapy�� 4
Other Considerations �� 5
Clinical Queries and Reporting�� 5
References�� 6
2 The Normal Endometrium�� 9
General Considerations in Histologic Evaluation�� 10
Histologic Features of Normal Cycling Endometrium�� 13
Proliferative Phase �� 13
Secretory Phase�� 16
Menstrual Endometrium�� 20
Pitfalls in The Histologic Assessment
of The Normal Endometrium�� 21
Sample Adequacy and Standardized Reporting �� 24
Artifacts and Contaminants�� 27
Irregular Secretory Endometrium�� 31
References�� 35
3 Pregnancy, Abortion, and Ectopic Pregnancy�� 39
Endometrial Glands and Stroma in Pregnancy�� 40
Early Gestational Endometrium (1–3 Weeks
Postfertilization)�� 40
Endometrium in Later Pregnancy (4 or More Weeks
Postfertilization)�� 41
Arias-Stella Reaction �� 45
Other Glandular Changes in Pregnancy�� 47
Trophoblast and Villi�� 49
Trophoblastic Cells�� 49
Immunohistochemistry of Trophoblastic Cells�� 51
Placental Implantation Site�� 53
Chorionic Villi and Villous Trophoblast in the First Trimester�� 57
ix
x Contents
Hydropic Change and Other Pathologic Changes

in Abortions�� 59
Chorionic Villi and Villous Trophoblast After
the First Trimester�� 62
Placental Polyps �� 63
Placenta Accreta�� 63
Endometrium Associated with Ectopic Pregnancy�� 64
References�� 71
4 Gestational Trophoblastic Disease�� 75
Hydatidiform Mole�� 76
General Features�� 76
Complete Hydatidiform Mole�� 76
Partial Hydatidiform Mole�� 81
Cytogenetics�� 83
Differential Diagnosis�� 85
Abnormal (Non-molar) Villous Lesions�� 90
Other Considerations �� 91
Persistent Postmolar Gestational Trophoblastic Disease
and Invasive Hydatidiform Mole �� 91
Clinical Queries and Reporting of Hydatidiform Mole �� 93
Gestational Trophoblastic Neoplasms�� 94
Choriocarcinoma�� 94
Placental Site Trophoblastic Tumor �� 100
Epithelioid Trophoblastic Tumor �� 107
Nonneoplastic Lesions �� 110
Exaggerated Placental Site�� 110
Placental Site Nodule and Plaque�� 111
Clinical Queries and Reporting of Trophoblastic Neoplasms�� 112
References�� 113
5 Abnormal Uterine Bleeding�� 121
Morphologic Features of Glandular and Stromal Breakdown
in Menstrual and Abnormal Bleeding�� 122
Abnormal Uterine Bleeding: Nonstructural Causes�� 126
Estrogen-Related Bleeding�� 129
Proliferative Endometrium with Glandular
and Stromal Breakdown�� 129
Disordered Proliferative Endometrium
and Persistent Proliferative Phase�� 132
Atrophy�� 133
Progesterone-Related Bleeding�� 135
Irregular Secretory Endometrium�� 136
Irregular Shedding �� 137
Abnormal Secretory Endometrium with Breakdown
of Unknown Etiology�� 138
Contents xi
6 Effects of Hormones�� 145

Estrogenic Hormones�� 146
Progestins, Oral Contraceptives, and Selective
Progesterone Receptor Modulators�� 146
Patterns of Response�� 148
Decidual Pattern�� 148
Secretory Pattern�� 149
Inactive Pattern�� 150
Other Stromal Changes�� 151
Combined Estrogen and Progestin as Replacement Therapy
for Menopausal Women �� 153
Progestin-Like Effects with No Hormone Use�� 154
Effects of Other Hormones�� 155
Selective Progesterone Receptor Modulators�� 155
Tamoxifen�� 156
Raloxifene�� 160
Clomiphene Citrate�� 160
Danazol�� 161
Human Menopausal Gonadotropins/Human
Chorionic Gonadotropin�� 162
Gonadotropin-Releasing Hormone Agonists �� 162
Antiprogestin RU486 �� 162
Postmenopausal Hormone Replacement�� 163
Abnormal Uterine Bleeding�� 163
Treatment of Hyperplasia and Endometrioid Carcinoma�� 164
Infertility Therapy�� 165
7 Endometritis �� 173
Nonspecific Endometritis�� 174
Inflammatory Cells�� 174
Stromal Changes�� 177
Abnormal Glandular Development�� 178
Epithelial Changes �� 178
Glandular and Stromal Breakdown�� 179
Specific Infections�� 179
Granulomatous Inflammation�� 180
Actinomycosis�� 182
Cytomegalovirus�� 183
Herpesvirus�� 184
Mycoplasma�� 185
8 Polyps�� 199
Classification and Histologic Features�� 200
Common Polyps�� 205
xii Contents
Proliferative/Hyperplastic Pattern�� 205

Atrophic Pattern �� 206
Functional Pattern�� 206
Mixed Endometrial–Endocervical Pattern �� 206
Adenomyomatous Pattern�� 207
Atypical Polypoid Adenomyoma (APA)�� 207
Adhesions�� 215
9 Precursors of Endometrial Carcinoma�� 225
Endometrial Hyperplasia/Atypical Hyperplasia�� 225
Behavior �� 240
Epithelial Cytoplasmic Change (Metaplasia)�� 241
Serous Endometrial Intraepithelial Carcinoma�� 251
Behavior �� 254
10 Endometrial Carcinoma�� 261
Classification of Endometrial Carcinoma�� 262
Important Issues in Interpretation of Biopsies �� 264
Criteria for the Diagnosis of Well-Differentiated
Endometrioid Carcinoma �� 264
Confluent Gland Pattern�� 264
Altered Fibrous or Desmoplastic Stroma�� 265
Extensive Papillary Pattern�� 265
Malignant Neoplasms: Classification, Grading, and Staging
of the Tumor�� 267
Classification�� 267
Grading�� 269
Clinically Important Histologic Subtypes�� 274
Typical (Endometrioid) Carcinoma �� 274
Carcinoma with Squamous Differentiation�� 278
Mucinous Carcinoma �� 281
Microglandular (Endocervical-Like)
Endometrial Carcinoma �� 283
Hereditary Syndromes �� 285
Hereditary Nonpolyposis Colorectal Cancer
(Lynch Syndrome) �� 285
Cowden Syndrome�� 288
Histologic Effects After Progestin Therapy �� 289
Serous Carcinoma�� 290
Clear Cell Carcinoma�� 295
Rare Histologic Subtypes�� 297
Contents xiii
Carcinosarcoma�� 301
Differential Diagnosis of Carcinosarcoma�� 305
Staging �� 309
Endometrial Versus Endocervical Carcinoma�� 310
Metastatic Carcinoma�� 317
Carcinoma Mimics�� 318
11 Mesenchymal Tumors and Other Rare Neoplasms�� 333
Smooth Muscle Tumors �� 333
Leiomyomas�� 333
Variants of Leiomyoma�� 334
Smooth Muscle Tumor of Uncertain Malignant Potential�� 335
Leiomyosarcoma�� 336
Tumorlets �� 337
Miscellaneous Mesenchymal Tumors�� 339
Perivascular Epithelioid Cell Tumor�� 339
Stromal Tumors�� 340
Endometrial Stromal Nodule and Low-Grade Endometrial
Stromal Sarcoma�� 341
High-Grade Endometrial Stromal Sarcoma �� 344
Undifferentiated Uterine Sarcoma �� 345
Uterine Tumors Resembling Ovarian Sex Cord Tumors �� 347
Mixed Epithelial and Mesenchymal Tumors �� 347
Adenofibroma and Adenosarcoma�� 347
Pathologic Features�� 347
Immunohistochemical Analysis �� 349
Molecular Analysis�� 350
Rare Neoplasms �� 351
Inflammatory Myofibroblastic Tumor�� 351
Lymphoma and Leukemia�� 351
Miscellaneous Tumors �� 352
Other Lesions and Tumor-Like Conditions �� 353
12 Methods of Endometrial Evaluation�� 363
Endometrial Sampling Techniques�� 363
Endometrial Biopsy �� 363
Dilation and Curettage �� 364
Hysteroscopy and Curettage�� 364
Other Aspiration Devices �� 365
Endometrial Imaging Studies�� 365
xiv Contents
Ultrasound�� 365
Magnetic Resonance Imaging�� 366
Histologic Techniques�� 366
Frozen Section �� 367
Index�� 373
Introduction
1
Contents
Indications for Biopsy 1
Clinical History and Biopsy Interpretation 2
Clinical Queries and Reporting 5
References 6
Endometrial biopsies and curettings are among the tion of any pathologic specimen, proper
most common tissue specimens received in the interpretation requires appropriate fixation, pro-
pathology laboratory. In several respects, these cessing, and sectioning of the tissue.
specimens present a unique challenge for the sur-
gical pathologist. The normal endometrium under-
goes a variety of morphologic changes, especially Indications for Biopsy
during the reproductive years, when cyclical hor-
monal influences and pregnancy affect uterine There are four main indications for endometrial
growth. Biopsy-induced artifacts confound this biopsy or curettage [5–9]:
heterogeneous group of morphologic changes.
Endometrial sampling techniques can vary from 1. Determination of the cause of abnormal uter-
hysteroscopy with curettage, which is considered ine bleeding
the “gold standard” [1–4], to a “blind” biopsy with 2. Evaluation of the status of the endometrium in
no visualization of the tissue sampled. The final infertile patients
specimen contains multiple, irregularly oriented 3. Evacuation of products of conception, either
tissue fragments mixed with blood and contami- spontaneous abortions, termination of preg-
nating cervical tissue and mucus. nancy, or retained tissue
Interpreting the biopsy material demands a 4. Assessment of the response of the endome-
logical approach that takes into account many trium to hormonal therapy, especially estrogen
factors, including patient history, the specific replacement in perimenopausal and postmeno-
requests of the clinician performing the biopsy, pausal women, progestin therapy in reproduc-
and an appreciation of the limitations, potential tive age women for treatment of endometrial
pitfalls, and complex array of patterns encoun- hyperplasia or endometrioid carcinoma, and
tered in the microscopic sections. As in evalua- tamoxifen therapy for breast cancer
© Springer Nature Switzerland AG 2019 1

T. A. Murdock et al., Diagnosis of Endometrial Biopsies and Curettings,
https://doi.org/10.1007/978-3-319-98608-1_1
2 1 Introduction
Other indications for biopsy may arise. An linical History and Biopsy
C
occasional patient will have atypical or abnor- Interpretation
mal glandular cells of undetermined signifi-
cance (AGUS) in a cervical–vaginal cytologic Abnormal Uterine Bleeding
specimen that requires endometrial sampling to
exclude hyperplasia or carcinoma. Uterine The most common reason for performing an
screening with transvaginal ultrasound can endometrial biopsy is abnormal uterine bleeding
show a thickened endometrial stripe in post- (AUB). Because of the inconsistent nomenclature
menopausal patients, and a biopsy can be per- used to describe variations of abnormal bleed-
formed to exclude significant pathology ing, the International Federation of Gynecology
[10–12]. Some clinicians sample the endome- and Obstetrics (FIGO) developed a classification
trium prior to hysterectomy to exclude signifi- system. Categorization is based on the acronym
cant pathology, although this procedure reveals PALM-COEIN (polyps, adenomyosis, leiomy-
little pathology in the absence of a history of oma, malignancy and hyperplasia – coagulopathy,
abnormal bleeding [13, 14]. Likewise, endome- ovulatory dysfunction, endometrial, iatrogenic,
trial biopsy for screening of endometrial cancer and not yet specified) and is used for AUB in non-
or precursor lesions in asymptomatic perimeno- gravid women of reproductive age. The first four
pausal and postmenopausal patients has a very terms are structural lesions, i.e., specific lesions.
low yield of significant abnormalities and is not The latter five (COEIN) are used to describe causes
cost-effective [15–17]. that are not defined by imaging or histopathology
At times, these indications for endometrial and were previously under the term dysfunctional
sampling overlap. For example, some compli- uterine bleeding (DUB) [18]. Abnormal uterine
cations of pregnancy, such as a missed abor- bleeding can be a sign of one or multiple uterine
tion or trophoblastic disease, are accompanied disorders ranging from nonstructural abnormalities
by abnormal uterine bleeding. Nonetheless, to structural lesions such as polyps, hyperplasia, or
these broad categories provide a clinicopatho- carcinoma [8, 11, 18–22]. For the FIGO classifica-
logic framework for approaching the micro- tion system, if the AUB is attributed to polyps, the
scopic analysis of endometrial biopsy patient chart would then read “AUB-P”; if a patient
specimens. The text has therefore been divided had multiple causes, such as a submucosal leiomy-
into chapters that correspond to these clinical oma and a coagulopathy, AUB-L(SM), C would be
indications. an acceptable term (Table 1.1).
Table 1.1 Clinical terms for abnormal uterine bleeding (AUB)

Abnormal uterine bleeding Abnormal uterine bleeding caused by structural lesions
(AUB) (PALM) Polyps (AUB-P)
Adenomyosis (AUB-A)
Leiomyoma (AUB-L)
Malignancy and hyperplasia (AUB-M)
Abnormal uterine bleeding Abnormal uterine bleeding with no structural cause
(AUB) (COEIN) Coagulopathies (AUB-C)
Abnormalities in ovulation (AUB-O)
Primary disorders of the endometrium (AUB-E)
Iatrogenic (AUB-I)
Other causes not yet specified (AUB-N)
Acute AUB Nongravid, reproductive-aged women with bleeding of sufficient quantity to
require immediate intervention to prevent further blood loss [18, 23]
Chronic AUB Bleeding that is abnormal in duration, volume, and/or frequency and has been
present for the majority of the last 6 months [18, 23]
Clinical History and Biopsy Interpretation 3
Age and menstrual/menopausal status are Table 1.3 Causes of abnormal uterine bleeding in peri-
menopausal years
especially important data to include in the pathol-
ogy requisition, as causes of abnormal uterine Common Uncommon
bleeding vary significantly according to parame- Ovulatory dysfunction Coagulopathies
Structural lesions Endometritis
ters, as discussed later. The prevalence of the
Hyperplasia Sarcoma
various abnormalities that lead to abnormal
Polyps (endometrial, Complications of
bleeding is difficult to determine precisely, vary- endocervical) pregnancya
ing with the patient population and the previous Iatrogenic
terms used by investigators [5–7]. The nomencla- Birth control
ture to describe menstrual bleeding related to Estrogen replacement
regularity or frequency of onset, duration, and Progestin therapy
heaviness (volume) of menstrual flow has been See Chap. 3 (Complications of pregnancy)
a
reclassified as well, and terms such as menorrha-

gia and menometrorrhagia have been largely patients [25, 32–34]. One consistent observation
abandoned. A practical approach to the possible in studies of postmenopausal patients is that atro-
diagnoses associated with abnormal bleeding phy is a common cause of abnormal bleeding,
takes age into account (Tables 1.2 and 1.3). In being found in 25% or more of cases [25, 26, 28,
adolescence, AUB may be secondary to ovula- 31, 33, 35, 36].
tory dysfunction, pregnancy, exogenous hormone There are a few exceptions where younger, pre-
administration, or coagulopathies. Pregnancy- menopausal women are at higher risk for endome-
related and nonstructural entities are more com- trial hyperplasia and carcinoma. Premenopausal
mon in younger patients, whereas atrophy and women (<45 years of age) with AUB and a body
structural lesions become more frequent in older mass index >30 kg/m2 are four times more likely
individuals [24]. Polyps in perimenopausal and to develop endometrial hyperplasia or carcinoma
postmenopausal patients have been found in than premenopausal women with a normal body
2–26% of patients [25–33] with a mean age of mass index [37]. Hereditary cancer syndromes
45.8 [34]. Hyperplasia is found in up to 16% of including Lynch, Cowden, Peutz-Jeghers, and
postmenopausal patients undergoing biopsy and Li-Fraumeni all have an elevated risk for endome-
endometrial carcinoma in fewer than 10% of trial cancer [38–41]. For Lynch syndrome, one
surveillance strategy is for annual transvaginal
ultrasound and/or endometrial sampling, followed
Table 1.2 Causes of abnormal uterine bleeding in the by risk-reducing hysterectomy upon completion
reproductive years
of childbearing [42]. In women aged 30–35 years
Common Uncommon
with Cowden syndrome, consideration for annual
Endometritis Neoplasia
transvaginal ultrasound and endometrial sampling
Complications of Endometrial
pregnancya carcinoma with a discussion of hysterectomy following
Ovulatory dysfunction Cervical carcinoma childbearing is a potential surveillance and risk
Anovulatory cycles Hyperplasia reduction strategy, respectively [42]. In addition
Inadequate luteal phase Coagulopathies to endometrial cancer, women with Peutz-Jeghers
Irregular shedding syndrome have a risk of developing sex cord

Structural lesions tumor with annular tubules (SCTAT) of the ovary
Leiomyomas and minimal deviation adenocarcinoma of the
Polyps (endometrial,
cervix [41]. Because of the additional ovarian and
endocervical)
Adenomyosis cervical cancer risk, an annual pelvic exam with
Iatrogenic Pap smear starting at age 18–20 years and consid-
Birth control eration of an annual transvaginal ultrasound is a
Progestin therapy reasonable surveillance approach. There are no
a
See Chap. 3 (Complications of pregnancy) clear surveillance recommendations for
4 1 Introduction
Table 1.4 Causes of abnormal uterine bleeding in post- Infertility Biopsy

menopausal years
Common Uncommon When a patient undergoes biopsy for evaluation of
Atrophy Endometritis infertility, the clinical information often is limited,
Structural lesions Sarcoma
but here, too, the history should include the date of
Hyperplasia Coagulopathies
the last menstrual period (LMP) to place an approx-
Polyps (endometrial)
Neoplasia imate time in the menstrual cycle. This information
Endometrial carcinoma is useful but not precise for determining the actual
Exogenous hormones day of the cycle, as ovulatory frequency and length
Estrogen replacement of the follicular phase are highly variable among
Progestin therapy patients. Usually the main objective of biopsies for
infertility is to determine whether there is morpho-
logic evidence of ovulation, i.e., secretory change
Li-Fraumeni syndrome and no clear risk reduc- (see Chap. 2). The gynecologist may seek other
tion strategies for Peutz-Jeghers or Li-Fraumeni specific information, such as response to hormone
syndromes [42]. therapy, so it is important that the pathologist be
Even among perimenopausal and postmeno- given any additional history that may be necessary
pausal patients, the proportion of cases attributable for the interpretation.
to any of the aforementioned conditions is age
dependent (Table 1.4). Atrophy and carcinoma
occur more frequently in patients older than Products of Conception
60 years of age, while polyps and hyperplasia are
more common in patients who are perimenopausal When endometrial sampling is performed to remove
or more recently postmenopausal. In addition to products of conception, clinical information often is
these uterine causes of bleeding, other abnormali- sparse, as the main goal of the p rocedure is simply
ties, such as genitourinary syndrome of menopause, to remove the placental and fetal tissue. Significant
can cause vaginal bleeding, and this may be difficult pathologic changes are rare. Nonetheless, it is help-
to distinguish from uterine bleeding until the patient ful to know if pregnancy is suspected, and, if so, the
undergoes thorough clinical evaluation. approximate gestational age of the pregnancy. If
A history of anovulation, obesity, hyperten- there is a suspicion of trophoblastic disease, this
sion, diabetes, and exogenous estrogen use should be stated. In such instances, the serum
should alert the pathologist that the patient is at human chorionic gonadotropin (hCG) titer is very
increased risk for hyperplasia and endometrioid important. If an ectopic pregnancy is suspected,
carcinoma, but this information is rarely included alerting the pathologist can ensure rapid processing
on the requisition. Typically, there is little and interpretation of the specimen.
accompanying clinical data except the patient’s
age and a short history of abnormal bleeding.
Consequently, hyperplasia and adenocarcinoma Hormone Therapy
must be diagnostic considerations for most
endometrial specimens received in the labora- Because the endometrium is responsive to hor-
tory. On rare occasions, hyperplasia or even ade- mones, the history of hormone use is important
nocarcinoma is found in biopsies performed information. Clinical uses of steroid hormones
during an infertility workup [28]. It should be (estrogens, progestins, or both) include oral, subcu-
kept in mind that women with unexplained infer- taneous or vaginal contraception methods, proges-
tility or diagnosed with polycystic ovarian syn- tin-releasing intrauterine devices, postmenopausal
drome (PCOS) at a young age are at risk of replacement therapy, and therapy for endometrio-
endometrial cancer [43, 44]. sis, infertility, hyperplasia, and endometrial endo-
metrioid or breast carcinoma. As with other facets
of the clinical data, this information may be absent Clinical Queries and Reporting
or, if present, unreadable on the requisition (in
which case the gynecologist should be contacted). Diagnostic terms such as “no pathologic diagno-
Consequently, the pathologist must be prepared to sis” or “no significant pathologic findings” are
recognize hormonal effects in the absence of his- unacceptable as there is a wide range of normal
tory indicating the use of hormones (see Chap. 6 – histology. When the tissue lacks abnormalities,
Effects of Hormones). stating the normal phase of the endometrium, for
example, menstrual, proliferative, or secretory,
provides useful information for the clinician.
Other Considerations In biopsies for abnormal uterine bleeding, the
pathologic information sought varies with the
Pregnancy history is useful, especially in pre- patient’s age and clinical history. The gynecolo-
menopausal patients, regardless of the indication gist wishes to know the following:
for biopsy, as recent and remote effects of preg-
nancy, such as a placental site nodule or gesta- 1. Is there an organic or structural lesion, such as
tional trophoblastic disease, may be encountered a complication of pregnancy, inflammation, or
in biopsy material. The history of recent or past a polyp?
pregnancies is expressed as gravidity and parity. 2. Is there evidence of active or old breakdown
The letter G (gravidity) followed by a number and bleeding?
(G1, G2, etc.) indicates the number of pregnan- 3. Is there evidence to suggest abnormalities in
cies, and the letter P (parity) followed by a num- ovulation?
ber indicates the number of deliveries. For 4. Is there evidence of hyperplasia or carcinoma?
example, G4, P2 indicates that a woman has had
four pregnancies and two deliveries. Further For example, in young premenopausal patients
information on parity often is designated by four with a normal BMI, the possibility of pregnancy
numbers indicating full-term pregnancies, prema- and related bleeding is a frequent question. In a
ture pregnancies (>20 but <37 weeks’ gestation), perimenopausal patient, the concern shifts to
abortions (<20 weeks’ gestation), and living chil- hyperplasia and carcinoma, and in postmeno-
dren. Thus a patient who is G5, P3013 is currently pausal patients, the importance of ruling out car-
pregnant and has had three previous full-term cinoma becomes paramount. In any of these
pregnancies and one abortion, and the three chil- conditions, glandular and stromal breakdown
dren from the term pregnancies are alive. may be present either focally or diffusely. It is the
The type of procedure, that is, biopsy versus underlying disorder that is most important to
curettage, is important for deciding whether focal report. The changes of breakdown and bleeding
changes represent significant abnormalities or are secondary and do not indicate a primary dis-
whether small specimens are adequate (see Chap. order by themselves. Nonetheless, when there is
12 Methods of Endometrial Evaluation). Although a history of abnormal bleeding, it can be helpful
office-based biopsies generally provide a repre- to note whether there is histologic evidence of
sentative sample, they may not contain sufficient glandular and stromal breakdown (see Chap. 5),
tissue to ensure that the endometrium has been especially if the tissue lacks evidence of an
adequately sampled. For example, the irregular organic process such as hyperplasia or carci-
glands of hyperplasia may resemble patterns seen noma. This information serves to document to
in some polyps, low-grade adenocarcinomas, and the gynecologist that bleeding is, in fact, endo-
even artifactually distorted normal endometrium. metrial in origin. Even when there is no evidence
Furthermore, atypia can be focal in hyperplasia; of active bleeding, foci of stromal foam cells or
therefore, biopsy specimens may preclude a defin- hemosiderin, sometimes with fibrosis, indicate
itive diagnosis. In these cases, a more thorough that abnormal bleeding has taken place and
biopsy or curettage is necessary. deserve comment.
6 1 Introduction
Besides reporting the morphologic changes tematic review and meta-analysis. Obstet Gynecol.
2017;130:803–13.
present, noting significant negative findings can 3. Gkrozou F, Dimakopoulos G, Vrekoussis T,
be helpful to the clinician. As an example, the Lavasidis L, Koutlas A, Navrozoglou I, Stefos T,
diagnosis of chronic endometritis is more helpful Paschopoulos M. Hysteroscopy in women with
if it includes a comment regarding the presence abnormal uterine bleeding: a meta-analysis on four
major endometrial pathologies. Arch Gynecol Obstet.
or absence of specific etiologic factors such as 2015;291:1347–54.
evidence of a recent pregnancy. Likewise, if an 4. Svirsky R, Smorgick N, Rozowski U, Sagiv R,
structural lesion such as a polyp is present, it is Feingold M, Halperin R, Pansky M. Can we rely
helpful to indicate whether noninvolved tissue is on blind endometrial biopsy for detection of focal
intrauterine pathology? Am J Obstet Gynecol.
present and, if so, its appearance. In perimeno- 2008;199:115.e1–3.
pausal and postmenopausal patients, if the gyne- 5. Baitlon D, Hadley JO. Endometrial biopsy. Pathologic
cologist indicates a specific concern regarding findings in 3,600 biopsies from selected patients. Am
the presence of hyperplasia, atypia, or carcinoma, J Clin Pathol. 1975;63:9–15.
6. Nickelsen C. Diagnostic and curative value of
then a statement noting the absence of these uterine curettage. Acta Obstet Gynecol Scand.
lesions is reassuring. 1986;65:693–7.
For all cases, specimen adequacy is a consid- 7. Van Bogaert LJ, Maldague P, Staquet JP. Endometrial
eration, but this needs to be specifically addressed biopsy interpretation. Shortcomings and problems
in current gynecologic practice. Obstet Gynecol.
only in limited samples in which the diagnosis is 1978;51:25–8.
not clear-cut. Scant tissue obtained by an office- 8. Galle PC, McRae MA. Abnormal uterine bleed-
based biopsy may be insufficient to allow thor- ing. Finding and treating the cause. Postgrad Med.
ough assessment of the status of the endometrium. 1993;93:73–6. 80–81
9. Merrill JA. The interpretation of endometrial biop-
In these cases, the pathologist should indicate in sies. Clin Obstet Gynecol. 1991;34:211–21.
the report that the specimen is scant. For instance, 10. Goldstein RB, Bree RL, Benson CB, Benacerraf BR,
small samples may reveal hyperplastic glands, Bloss JD, Carlos R, Fleischer AC, Goldstein SR, Hunt
but it may be difficult to determine whether the RB, Kurman RJ, Kurtz AB, Laing FC, Parsons AK,
Smith-Bindman R, Walker J. Evaluation of the woman
abnormality represents a localized polyp with a with postmenopausal bleeding: society of Radiologists
proliferative/hyperplastic pattern (see Chap. 8) or in ultrasound-sponsored consensus conference state-
a diffuse hyperplasia. Some assessment of the ment. J Ultrasound Med. 2001;20:1025–36.
endometrium can be done even on very limited 11. Patel V, Wilkinson EJ, Chamala S, Lu X, Castagno
J, Rush D. Endometrial thickness as measured
specimens, for example, noting whether the by transvaginal ultrasound and the correspond-
endometrium is proliferative or secretory phase. ing histopathologic diagnosis in women with
Atrophic endometrium typically yields a very postmenopausal bleeding. Int J Gynecol Pathol.
small amount of tissue, yet these specimens 2017;36:348–55.
12. Fleischer A, Abramowicz J, Goncalves L, Manning F,
should not be regarded as inadequate (see Chap. Monteagudo A, Timor I, Toy E, editors. Sonographic
5). The subsequent chapters consider in greater techniques for early detection of ovarian and endo-
detail the queries likely to arise in various cir- metrial cancers. Fleischer’s sonography in obstetrics
cumstances and the information that the patholo- and gynecology. 8th ed. New York: McGraw Hill
Education; 2018.
gist should incorporate in the final report. 13. Stovall TG, Solomon SK, Ling FW. Endometrial
sampling prior to hysterectomy. Obstet Gynecol.
1989;73:405–9.
References 14. Ramm O, Gleason JL, Segal S, Antosh DD, Kenton
KS. Utility of preoperative endometrial assess-
ment in asymptomatic women undergoing hysterec-
1. Clark TJ, Voit D, Gupta JK, Hyde C, Song F, Khan
tomy for pelvic floor dysfunction. Int Urogynecol J.
KS. Accuracy of hysteroscopy in the diagnosis of
2012;23:913–7.
endometrial cancer and hyperplasia: a systematic
15. Archer DF, McIntyre-Seltman K, Wilborn WW,
quantitative review. JAMA. 2002;288:1610–21.
Dowling EA, Cone F, Creasy GW, Kafrissen
2. Visser NCM, Reijnen C, Massuger LFAG, Nagtegaal
ME. Endometrial morphology in asymptomatic
ID, Bulten J, Pijnenborg JMA. Accuracy of endo-
postmenopausal women. Am J Obstet Gynecol.
metrial sampling in endometrial carcinoma: a sys-
1991;165:317–20. discussion 320–322.
References 7
16. Louie M, Canavan TP, Mansuria S. Threshold for 33. Van den Bosch T, Ameye L, Van Schoubroeck D,
endometrial sampling among postmenopausal Bourne T, Timmerman D. Intra-cavitary uterine
patients without vaginal bleeding. Int J Gynaecol pathology in women with abnormal uterine bleeding:
Obstet. 2016;132:314–7. a prospective study of 1220 women. Facts Views Vis
17. McPencow AM, Erekson EA, Guess MK, Martin DK, Obgyn. 2015;7:17–24.
Patel DA, Xu X. Cost-effectiveness of endometrial 34. Ricciardi E, Vecchione A, Marci R, Schimberni
evaluation prior to morcellation in surgical procedures M, Frega A, Maniglio P, Caserta D, Moscarini
for prolapse. Am J Obstet Gynecol. 2013;209:22.e1–9. M. Clinical factors and malignancy in endometrial
18. Munro MG, Critchley HOD, Broder MS, Fraser polyps. Analysis of 1027 cases. Eur J Obstet Gynecol
IS. FIGO working group on menstrual disorders: FIGO Reprod Biol. 2014;183:121–4.
classification system (PALM-COEIN) for causes of 35. Meyer WC, Malkasian GD, Dockerty MB, Decker
abnormal uterine bleeding in nongravid women of repro- DG. Postmenopausal bleeding from atrophic endo-
ductive age. Int J Gynaecol Obstet. 2011;113:3–13. metrium. Obstet Gynecol. 1971;38:731–8.
19. Goldfarb JM, Little AB. Current concepts: abnormal 36. Gambrell RD. Postmenopausal bleeding. J Am Geriatr
vaginal bleeding. N Engl J Med. 1980;302:666–9. Soc. 1974;22:337–43.
20. Povey WG. Abnormal uterine bleeding at puberty and 37. Wise MR, Gill P, Lensen S, Thompson JMD, Farquhar
climacteric. Clin Obstet Gynecol. 1970;13:474–88. CM. Body mass index trumps age in decision for
21. Kilbourn CL, Richards CS. Abnormal uterine bleed- endometrial biopsy: cohort study of symptom-
ing. Diagnostic considerations, management options. atic premenopausal women. Am J Obstet Gynecol.
Postgrad Med 2001;109:137–138, 141–144, 147–150. 2016;215:598.e1–8.
22. Wren BG. Dysfunctional uterine bleeding. Aust Fam 38. Stoffel E, Mukherjee B, Raymond VM, Tayob N,
Physician. 1998;27:371–7. Kastrinos F, Sparr J, Wang F, Bandipalliam P, Syngal
23. Fraser IS, Critchley HOD, Broder M, Munro MG. The S, Gruber SB. Calculation of risk of colorectal and
FIGO recommendations on terminologies and defi- endometrial cancer among patients with lynch syn-
nitions for normal and abnormal uterine bleeding. drome. Gastroenterology. 2009;137:1621–7.
Semin Reprod Med. 2011;29:383–90. 39. Pennington KP, Walsh T, Lee M, Pennil C, Novetsky
24. Fritz MA, Speroff L. Clinical gynecologic endocrinol- AP, Agnew KJ, Thornton A, Garcia R, Mutch D, King
ogy and infertility. 8th ed. Philadelphia: LWW; 2010. M-C, Goodfellow P, Swisher EM. BRCA1, TP53, and
25. Rubin SC. Postmenopausal bleeding: etiology, evaluation, CHEK2 germline mutations in uterine serous carci-
and management. Med Clin North Am. 1987;71:59–69. noma. Cancer. 2013;119:332–8.
26. Schindler AE, Schmidt G. Post-menopausal bleed- 40. Tan M-H, Mester JL, Ngeow J, Rybicki LA, Orloff
ing: a study of more than 1000 cases. Maturitas. MS, Eng C. Lifetime cancer risks in individuals
1980;2:269–74. with germline PTEN mutations. Clin Cancer Res.
27. Van Bogaert LJ. Clinicopathologic findings in endo- 2012;18:400–7.
metrial polyps. Obstet Gynecol. 1988;71:771–3. 41. van Lier MGF, Wagner A, Mathus-Vliegen EMH,
28. Choo YC, Mak KC, Hsu C, Wong TS, Ma Kuipers EJ, Steyerberg EW, van Leerdam ME. High
HK. Postmenopausal uterine bleeding of nonorganic cancer risk in Peutz-Jeghers syndrome: a system-
cause. Obstet Gynecol. 1985;66:225–8. atic review and surveillance recommendations. Am
29. Mencaglia L, Perino A, Hamou J. Hysteroscopy J Gastroenterol. 2010;105:1258–64. author reply
in perimenopausal and postmenopausal women 1265.
with abnormal uterine bleeding. J Reprod Med. 42. Ring KL, Garcia C, Thomas MH, Modesitt
1987;32:577–82. SC. Current and future role of genetic screening in
30. Pacheco JC, Kempers RD. Etiology of postmeno- gynecologic malignancies. Am J Obstet Gynecol.
pausal bleeding. Obstet Gynecol. 1968;32:40–6. 2017;217:512–21.
31. Lidor A, Ismajovich B, Confino E, David 43. Venn A, Watson L, Lumley J, Giles G, King C,
MP. Histopathological findings in 226 women with Healy D. Breast and ovarian cancer incidence
post-menopausal uterine bleeding. Acta Obstet after infertility and in vitro fertilisation. Lancet.
Gynecol Scand. 1986;65:41–3. 1995;346:995–1000.
32. Moghal N. Diagnostic value of endometrial curettage 44. Harris HR, Terry KL. Polycystic ovary syndrome and
in abnormal uterine bleeding – a histopathological risk of endometrial, ovarian, and breast cancer: a sys-
study. J Pak Med Assoc. 1997;47:295–9. tematic review. Fertil Res Pract. 2016;2:14.
The Normal Endometrium
2
Contents
General Considerations in Histologic Evaluation 10
Histologic Features of Normal Cycling Endometrium 13
Pitfalls in The Histologic Assessment of The Normal Endometrium 21
Sample Adequacy and Standardized Reporting 24
Artifacts and Contaminants 27
Irregular Secretory Endometrium 31
References 35
The histologic features of what constitutes “nor- under sampling. These examples emphasize the
mal” endometrium change with a woman’s age, importance of clinical information including
through the premenarchal, reproductive, peri- patient’s age and hormonal status. In biopsy
menopausal, and postmenopausal years [1–3]. specimens, the combination of these cyclical
Throughout the reproductive years, the cyclical changes along with potential processing arti-
hormonal changes of the menstrual cycle pro- facts, limited sampling, or the presence of mim-
vide a continuously changing morphologic spec- ics can make normal patterns difficult to
trum from proliferative to secretory to menstrual interpret. Deviations from normal, either in his-
phenotypes which is considered “normal.” Once tologic pattern or in temporal relationship to
menopause is reached, the presence of only rare ovulation, often indicate underlying abnormali-
strips of atrophic epithelium in biopsy/curettage ties that may contribute to female infertility and
sampling becomes the new “normal.” The same abnormal uterine bleeding.
“normal” in a postmenopausal woman would be Over the past 6 decades, pathologists have used
deemed “abnormal” in a premenopausal woman the histologic criteria originally described by Noyes
or, alternatively, would raise the possibility of et al. [1, 4] to date secretory phase endometrial biop-
exogenous hormonal effect. Likewise, finding sies (Table 2.1), as part of an infertility workup. The
only cervical or lower uterine segment tissue in a original study by Noyes et al. described discrete
woman known to have an ultrasonographic changes that varied daily following ovulation, cul-
lesion such as thickened endometrium supports minating with menstruation. Over the last decades,

10 2 The Normal Endometrium
multiple studies critically evaluating the Noyes dat- use of terms such as “early,” “mid,” and “late” secre-
ing criteria have shown that they are not reproduc- tory endometrium suffices [13]. Nonetheless, an
ible and that the criteria themselves are seriously appreciation of the various morphologic changes
flawed [5–12]. More importantly, gynecologists that occur in the secretory phase of the cycle is
appear to be less inclined to utilize this information important for pathologists so that normal phases of
in the evaluation of infertility. For these reasons, the secretory endometrium are not misinterpreted as
traditional dating schema is not discussed in detail in abnormal (Table 2.1).
this chapter but is briefly outlined in Table 2.1. Of
interest, the Gynecologic Pathology Interest Group
of the Canadian Association of Pathologists (GPIG- General Considerations
CAP) published a guideline for pathologists stan- in Histologic Evaluation
dardizing terminology to assist in communicating
with our clinical colleagues. One recommendation Histologic evaluation begins with identification
was that formal “dating” of secretory endometrium of surface epithelium, a prerequisite for orient-
is optional but that it should be provided if specifi- ing the underlying glands and stroma. The sur-
cally requested by the clinician. Consequently, the face epithelium is less responsive to sex steroid
Table 2.1 Secretory phase endometrial changes

Interval phase, 14–15d.a No datable histologic changes for 36–48 h after ovulation
Early secretory phase, 16–20d. Glandular changes predominate
Early Histologic features
16 Subnuclear, irregular vacuoles
17 Regular vacuolation—nuclei lined up with subnuclear vacuoles
18 Vacuoles decreased in size
18 Early secretions in lumen
18 Nucleus approaches base of cell with supranuclear vacuoles
19 Few vacuoles remain
19 Intraluminal secretion
19 No pseudostratification, no mitoses
20 Peak of intraluminal secretions
Mid- to late secretory phase, 21–27d. Stromal changes predominate, variable secretory exhaustion
Mid
21 Marked stromal edema
22 Peak of stromal edema—cells have “naked nuclei”
23 Periarteriolar predecidual change
23 Prominent spiral arteries
24 More prominent predecidual change
24 Stromal mitoses recur
25 Predecidual differentiation begins under surface epithelium
Late
25 Increased numbers of granular lymphocytes
26 Predecidua starts to become confluent
27 More numerous granular lymphocytes
27 Confluent sheets of predecidua
27 Focal necrosis
24–27 Secretory exhaustion of glands—tortuous with intraluminal tufts (saw-toothed), ragged luminal borders,
variable cytoplasmic vacuolization, and luminal secretions
d. = day of ideal 28-day menstrual cycle
a
General Considerations in Histologic Evaluation 11
hormones than the underlying glands, but it epithelial cells are important features in the histo-
often shows alteration in pathologic conditions, logic evaluation. Under normal conditions, the
especially when the abnormalities are subtle or glands should be regularly spaced and have a per-
focal. For example, during the proliferative pendicular arrangement from the basalis to the
phase, estrogenic stimulation results in ciliated surface epithelium. In the secretory phase, the
cells along the surface [10, 14]. Ciliated surface endometrium also shows a stratum compactum, a
epithelial cells are, however, far more frequent thin region beneath the surface epithelium. In the
in pathologic conditions, particularly those stratum compactum, the stroma is dense, and the
associated with unopposed estrogen stimula- glands are straight and narrow, even when the
tion, such as hyperplasia and metaplasia (see glands in the functionalis are tortuous. The basa-
Chap. 9) [2, 3, 15–17]. lis adjoins the myometrium, serving to regenerate
The subsurface endometrium is divided into the functionalis and surface epithelium following
two regions, the functionalis (stratum spongio- shedding during menses. The endometrium of the
sum) and the basalis (stratum basalis) (Fig. 2.1). basalis is less responsive to steroid hormones and
The functionalis is situated between the surface typically shows irregularly shaped, inactive-
epithelium and is important to evaluate because it appearing glands, dense stroma, and aggregates
shows the greatest degree of hormonal respon- of spiral arteries. The spiral arteries of the basalis
siveness. The size and distribution of glands as (basal arteries) have thicker muscular walls than
well as the cytologic features of the glandular those in the functionalis. In biopsies, tissue
Fig. 2.1 Normal secretory phase endometrium. Surface epi- blood vessels demonstrate the typical patterns of maturation
thelium orients the tissue (far right). The midportion of the through the menstrual cycle. The stratum compactum is com-
tissue consists of functionalis where glands, stroma, and posed of the surface-type epithelium and a subjacent thin
layer of dense stroma
fragments that contain basalis often do not have mal cells with minimal cytoplasm seen in the
surface epithelium. The glands and stroma of the corpus.
basalis are unresponsive to steroid hormones. Tangential orientation of the functionalis in
Lower uterine segment/isthmus is another biopsies and the tortuosity of the glands, partic-
region of the endometrium that is less responsive ularly in the late proliferative and secretory
to steroid hormones. In the lower uterine seg- phases, often lead to irregular cross sections of
ment, the endometrium has shorter, poorly devel- the tissue. In this instance, gland development
oped, inactive glands dispersed in a distinctive can be difficult to assess. Furthermore, not all
stroma (Figs. 2.2 and 2.3). The columnar cells fragments of tissue in a biopsy or curettage
lining the glands resemble those of the corpus. include surface epithelium, which helps to
Some glands near the junction with the endocer- orient the glands. Nonetheless, at least focally,
vix show a transition to mucinous endocervical- portions of better-oriented glands usually can be
type epithelium. The stromal cells in the lower traced through the functionalis to the surface
uterine segment are elongate and resemble fibro- epithelium, and these foci are critical for
blasts with more abundant eosinophilic assessing appropriate glandular and stromal

cytoplasm, in contrast to the oval to rounded stro- development.
Fig. 2.2 Lower uterine segment in curettage specimen. which are juxtaposed to hormonally responsive endome-
At the right side of the image, lower uterine segment trium (left)
shows inactive glands embedded in a fibrotic stroma
Fig. 2.3 Lower uterine segment. Small, poorly developed glands are seen in nonreactive stroma. Tissue from the lower
uterine segment cannot be dated
istologic Features of Normal

H [1]. Five of these features affect the glands,
Cycling Endometrium namely, tortuosity, gland mitoses, orientation of
nuclei (pseudostratified versus basal), basal sub-
The endometrium displays two distinct phases in nuclear cytoplasmic vacuoles, and glandular ser-
ovulatory cycles. The first is the proliferative rations with increased luminal secretions and
(follicular or preovulatory) phase which is secretory exhaustion. Four features relate to the
characterized by growth of glands, stroma, and stroma: edema, mitoses, predecidual change, and
vessels that is influenced by estradiol produced infiltration of granular lymphocytes. Practically,
mainly by granulosa cells in the ovarian follicles. the most important glandular features are orienta-
Following ovulation, the secretory (luteal or tion of nuclei, subnuclear cytoplasmic vacuoles,
postovulatory) phase reflects the effect of the and luminal secretions with secretory exhaustion,
combined production of progesterone and estra- and the most important stromal features are
diol by luteinized granulosa and theca cells of the edema, predecidual change, and granular lym-
corpus luteum [18]. phocytic infiltration (Table 2.1).
The day 1 of the menstrual cycle was arbi-
trarily defined as the first day of bleeding in a
“normal” cycle of 28 days [1]. Proliferative phase Proliferative Phase
changes are not as discrete as those in the secre-
tory phase; the latter can be roughly divided into During the proliferative phase, the endometrium
early, mid-, and late secretory phases. grows from about 0.5 mm up to 4.0–5.0 mm in
There are nine histologic features of the glands thickness, so by the late proliferative phase, a
and stroma that determine the phase of the cycle biopsy obtains a moderate amount of tissue.
Proliferative endometrium has three phases: small nucleoli and dense basophilic cytoplasm.
early, mid, and late [2]. There is considerable The pseudostratified nuclei remain oriented to
overlap between these phases so the diagnosis of the basement membrane, but some nuclei are
proliferative phase alone is sufficient, indicating raised above the basement membrane, giving a
that the endometrium is growing and shows a two-dimensional layering of the nuclei. The
normal glandular distribution and evidence of pseudostratification of the nuclei and the pres-
ovulation is not present. ence of mitotic activity in the glands and stroma
Growth of endometrium is the main charac- are two constant features of the proliferative
teristic of the proliferative phase (Figs. 2.4, 2.5, phase.
2.6, and 2.7). Glands and stroma show brisk In the proliferative phase, the stromal cells are
mitotic activity. In early proliferative endome- widely separated in the functionalis. They are
trium, the functionalis contains small, tubular small and oval, with dense nuclei, scant wisps of
glands. The glands progressively elongate and cytoplasm, and ill-defined cell borders. Some
become tortuous from the mid- to the late pro- stromal edema is normal at mid-proliferative
liferative phase because the gland growth is dis- phase. A few lymphocytes also are scattered
proportionate to the stromal growth. Despite throughout the stroma, being most prominent
the tortuosity, the glands maintain a relatively around the vessels. Small spiral arteries and thin-
regular spacing between each other. Throughout walled venules are present.
the proliferative phase, the epithelium lining The orientation and outline of proliferative
the glands has pseudostratified, oval nuclei with phase glands and their relationship to intact stroma
Fig. 2.4 Proliferative endometrium. Focal hemorrhage beneath the surface epithelium is a result of the biopsy and does
not represent a pathologic change
Fig. 2.5 Proliferative endometrium. Glands are tubular and dispersed in abundant stroma
Fig. 2.6 Proliferative endometrium. In this tangential section, the glands are regularly spaced. The gland to stroma ratio is 1:1
Fig. 2.7 Proliferative endometrium. The proliferative phase gland shows pseudostratified nuclei with mitotic activity.
The stromal cells have oval nuclei and indistinct cytoplasm. Scattered lymphocytes are normally present
are important features for recognizing this normal gests another diagnosis. Also, proliferative
pattern, as hyperplastic glands or glands in a polyp phase glands frequently show the telescoping
can have cytologic features identical to those of artifact (see below).
glands in the proliferative phase. The regular spac-
ing and uniform shape of the glands are character-
istics of normal proliferative endometrium. Secretory Phase
Assessing gland orientation can be complicated,
however, by biopsy-induced fragmentation, an In the secretory phase, the glands and stroma
especially common artifact in early to mid-prolif- develop in a somewhat orderly sequence, dis-
erative phase biopsies when the mucosa is still playing histologic features of (post-ovulation)
thin. Detached and disrupted glands may appear secretory activity. The endometrium attains a
abnormally crowded or irregular. thickness of up to 7.0–8.0 mm. Unlike the prolif-
To distinguish fragmentation artifact from erative phase, the changes in the glands and
true abnormalities, it is important to assess the stroma are relatively discrete, changing more
integrity of the stroma as well as the glands and abruptly from one day to the next. The first half
to use surface epithelium to help orient the tis- of the secretory phase is characterized primarily
sue fragments. Detached and poorly oriented by glandular changes, whereas in the second half,
glands that show pseudostratified nuclei and stromal alterations become more prominent.
mitotic activity usually represent proliferative The morphologic changes of the secretory
endometrium unless better-oriented tissue sug- phase begin within 48 hours after ovulation;
Fig. 2.8 Interval endometrium. The glands maintain proliferative phase characteristics and show scattered subnuclear
vacuoles. The extent of cytoplasmic vacuolization is not sufficient to be certain ovulation has occurred
however, that interval varies among women and c ulminating with a peak in luminal secretions
from cycle to cycle in the same woman. During (Figs. 2.12 and 2.13).
this interval, the glands become more tortuous During the mid-luteal phase, edema, the first
and begin to show subnuclear vacuoles (Fig. 2.8). noticeable stromal change, is most prominent.
The first diagnostic evidence of ovulation, how- Because of the edema, the stromal cells take on
ever, is the presence of abundant subnuclear gly- the so-called “naked nucleus” appearance. With
cogen vacuoles in the undulating, tortuous glands this change, the stromal cells are widely dis-
(Fig. 2.9a, b). At this time, the stroma is indistin- persed and have small nuclei with scant, imper-
guishable from that of the late proliferative phase. ceptible cytoplasm. This phase of pure stromal
If confirmation of ovulation is a requirement, edema is brief, and the subsequent predecidual
substantial amounts of glands with conspicuous transformation of the stroma becomes the main
vacuoles should be present. feature in the late secretory phase. Edema may
During the early secretory endometrium, occur in the earlier portion of the secretory phase
subnuclear vacuoles are abundant, progressively which does not connote an irregularity of matura-
moving from the basal to a supranuclear position. The glands show increasing tortuosity, and
tion (Figs. 2.10 and 2.11). Concurrently, the variable amounts of luminal secretions persist
nuclei become basally oriented and line up in a until just before menses.
single layer perpendicular to the basement Finally, during the late secretory phase, secre-
membrane. The cytoplasmic contents then form tory exhaustion is achieved (Figs. 2.14, 2.15).
mucin that is expelled into the gland lumen, Secretory exhaustion is characterized by the
a b
Fig. 2.9 Early secretory endometrium. (a) Postovulatory phase. (b) Every glandular cell contains a subnuclear vac-
changes are clearly present with a regular distribution of uole, resulting in a uniform alignment of nuclei away
subnuclear vacuoles in the serpiginous glands. The stroma from the basement membrane
shows no changes compared to the late proliferative
Fig. 2.10 Early secretory endometrium. Glandular cell vacuoles remain prominent but begin to migrate to the supra-
nuclear cytoplasm
Fig. 2.11 Early secretory endometrium. A portion of the endometrial stroma shows mild edema
presence of a single layer of cells that lie in dis- secretory phase, cellular necrosis (apoptosis)
array with loss of orientation. The cytoplasmic becomes evident with accumulation of nuclear
border along the luminal surface becomes ragged debris in the basal cytoplasm of the glandular
with gland serration or a “saw-toothed” luminal epithelial cells. Throughout the secretory phase,
border (Figs. 2.16a), and luminal secretions are the glands in the stratum compactum immedi-
usually, although not invariably, present. Spiral ately beneath the surface epithelium remain
arterioles, which started to become prominent small and tubular despite their increasing tortu-
during the mid-secretory phase, now are sur- osity in the functionalis.
rounded by predecidualized stromal cells with In the predecidua, there is a resurgence of
abundant cytoplasm (Fig. 2.16b). Predecidual stromal mitotic activity, while the glandular epi-
change, the main characteristic of the late secre- thelium lacks mitoses. Predecidual change
tory phase, makes the vessel walls appear expands, extending to the subsurface and form-
thicker, leading to their prominence. The glan- ing larger stromal sheets just before menstrual
dular cells may continue to show a variable endometrial breakdown. Predecidua is easy to
degree of vacuolization throughout the remain- recognize when advanced, but this change can be
der of the secretory phase. Toward the end of the subtle when it is early and not confluent. With
Fig. 2.12 Mid-secretory endometrium. Maximum intraglandular secretion is present, and the glands are dilated. The
stroma shows edema and there is no predecidual change
predecidual transformation, the stroma contains Menstrual Endometrium

an increasing number of leukocytes (Figs. 2.17,
2.18) [2, 3, 19, 20]. Immunohistochemical studies Menstrual endometrium shows glandular and stro-
revealed that most of these cells are T lympho- mal breakdown that rapidly affects the entire func-
cytes [21–23]. They are normally present in small tionalis. In this phase, there are fibrin thrombi in
numbers earlier in the cycle but become promi- small vessels, condensed and collapsed stroma,
nent by the late secretory phase, peaking during and necrotic debris (Figs. 2.19a, b). With this
menstruation. At the very end of the secretory necrosis, a true neutrophilic infiltrate becomes a
phase, the endometrium is premenstrual. part of the physiologic process [24]. When the
Predecidua is present in sheets with many inter- bleeding is extensive, it may not be possible to
spersed leukocytes. The glands are highly convo- assess the development of the glands or stroma or
luted and saw-toothed. The glands begin to show the “normality” of the tissue. Once breakdown
apoptosis with nuclear dust at their base starts, the stromal cells coalesce into aggregates
(Fig. 2.18). Fibrin thrombi begin to form in small and clusters that often show little cytoplasm. With
vessels, and hemorrhage follows with extravasa- extensive stromal collapse during menstruation, the
tion of erythrocytes into the stroma. predecidual change in the stromal cells becomes
Fig. 2.13 Mid-secretory endometrium. The glands are stroma is edematous and there is no predecidual change.
tortuous with intraluminal secretions that may appear The nuclei of stromal cells acquire a “naked nucleus”
washed out depending on the H&E preparation. The appearance with scant cytoplasm
indistinct. The extensive breakdown also can result pregnancy, but this tissue is not optimal unless
in striking morphologic alterations with artifactual obtained very early in the menstrual phase before
glandular crowding (Fig. 2.20). As a result, men- breakdown becomes extensive [18].
strual endometrium can be confused with hyper-
plasia or even carcinoma if the background
bleeding pattern is not recognized. Conversely, Pitfalls in The Histologic
hyperplasia and carcinoma are proliferative pro- Assessment of The Normal
cesses that rarely show extensive breakdown of the Endometrium
type displayed by menstrual endometrium. Because
of the artifacts induced by the breakdown and The preceding description summarizes the basic
bleeding of the menstrual phase, this tissue is not histologic changes of endometrial development.
suitable for evaluation of glandular and stromal In addition to understanding the normal morphol-
development. For infertility evaluation, some advo- ogy in ideal situations, it is important to consider
cate biopsy at the onset of bleeding to be certain several practical points when interpreting the
that the procedure does not interrupt an early endometrial biopsy. There are several caveats and
Fig. 2.14 Late secretory endometrium. The glands show a typical tortuous architecture and secretory exhaustion with
secretions pushed to the lumen. Predecidual stromal change is evident around spiral arteries
potential pitfalls, knowledge of which assists in certain that ovulation has occurred, a sub-
accurate diagnosis of normal endometrium. The stantial number of glands must show subnu-
following are especially important aspects to clear vacuoles.
consider: 5. The presence of secretions in the glandular
lumen does not indicate secretory endome-
1. Endometrium with surface epithelium is best trium. Proliferative, hyperplastic, and
for interpretation. Absence of surface epithe- neoplastic glands can contain luminal secre-
lium compromises the interpretation. tions. It is the glandular cytoplasm and
2. Tissue from the lower uterine segment or nuclear changes that are most important for
basalis is not satisfactory for endometrial determining the presence or absence of
assessment as these regions do not respond secretory changes.
fully to hormones (Fig. 2.21). 6. Focal glandular crowding caused by tangen-
3. Straight, tubular glands beneath the surface tial sectioning can occur in proliferative or
are normal and not a sign of irregularity in secretory endometrium (Fig. 2.22). This arti-
maturation in the late secretory phase. fact can result in back-to-back glands that do
4. Scattered subnuclear vacuoles in glands are not represent hyperplasia. A Ki-67 prolifera-
not sufficient evidence of ovulation. To be tion index will assist as the glands in
Fig. 2.15 Late secretory endometrium. The glands show Predecidual stromal change is extensive around spiral
a typical serrated or “saw-toothed” luminal border while arteries and beneath surface epithelium. Stromal lympho-
keeping a tortuous architecture and secretory exhaustion. cytes are numerous
a b
Fig. 2.16 Late secretory endometrium. (a) With secre- arteries show predecidual change with increased cyto-
tory exhaustion, the cytoplasmic border along the luminal plasm. The glands show secretory exhaustion with only
surface becomes ragged, showing gland serration or a patchy cytoplasmic vacuolization
“saw-toothed” appearance. (b) Stromal cells around spiral
Fig. 2.17 Late secretory endometrium. The glands in late secretory endometrium show tortuosity and variable amounts
of intraluminal secretions. Stromal granular lymphocytes are scattered throughout the stroma
s ecretory endometrium typically show mini- 11. Infiltrating leukocytes normally become
mal to no proliferative activity when com- prominent in the stroma of the late secretory
pared to true hyperplasia (see Chap. 9). phase. These do not represent inflammation.
7. Focal cystic glands or nonreactive glands 12. If the tissue is difficult to interpret because of
can occur in normal endometrium and have apparent discordance in features, the possi-
no significance. bility of chronic endometritis or a polyp
8. Stromal edema is normal in all stages of the should be considered.
secretory phase; however, it is maximal dur- 13. The endometrial phase cannot be accurately
ing the mid-secretory stage. assigned when polyps, inflammation, or
9. Identifying very early pregnancy based on other abnormalities are present.
endometrial changes alone is very difficult.
Apparent “hypersecretory” late secretory
phase glands with vacuolated cytoplasm Sample Adequacy
usually are a variation of normal develop- and Standardized Reporting
ment and do not, by themselves, indicate
early pregnancy (see Chap. 3). Sample adequacy assessment is the first step and
10. Compact predecidua with spindle-shaped major component of the endometrial biopsy/
stromal cells may not be appreciated as a curetting interpretation with few studies objec-
true predecidual reaction. Directing attention tively evaluating this parameter [25]. The volume
to stromal changes around spiral arteries of the sample received in the pathology laboratory
assists in the identification of predecidua. will depend on the procedure performed to obtain
Sample Adequacy and Standardized Reporting 25
Fig. 2.18 Premenstrual endometrium. Late secretory glands show apoptosis with nuclear debris at their base
a b
Fig. 2.19 Menstrual endometrium. (a) With stromal metrial stroma (breakdown) eventually becomes entirely
hemorrhage, the predecidual cells collapse and lose their fragmented with artifactual glandular crowding and
abundant cytoplasm. (b) Condensed and collapsed endo- telescoping
the specimen, as well as on the hormonal status of published series, the rates of inadequate biopsies
the patient (premenopausal, postmenopausal, may reach up to 33% [26]; however, the inade-
postmenopausal taking hormones). Currently, quacy criteria used by different practitioners may
most biopsies are obtained using a Pipelle instru- vary widely. In our experience, Pipelle samples
ment, as opposed to a curette (see Chap. 12). In have a much lower inadequate rate when the
Fig. 2.20 Menstrual endometrium. As the small vessels eventually becoming entirely fragmented. At this stage,
accumulate fibrin thrombi, the endometrial stroma starts some glands appear inactive and may show artifactual
to condense and collapse forming “stromal blue balls” and crowding, while others may still show secretory changes
s pecimen is carefully processed and sectioned. An be deemed inadequate, and repeat biopsy should
inadequate biopsy diagnosis will typically be left to clinical discretion.
prompt re-sampling; therefore, it should only be Terminology used to describe benign condi-
used when necessary. Most investigators have tions has been an issue in many subspecialties of
suggested that there is little chance of missing pathology, with most practitioners using the lexi-
significant pathology when minimal atrophic con adopted by their institutions or demanded by
endometrial tissue is obtained by Pipelle biopsy their clinicians. In gynecologic pathology, terms
of a postmenopausal patient lacking an ultraso- like “inactive” have been used interchangeably
nographic lesion [27]. Admittedly, determining with “atrophic.” While atrophic endometrium
what is inadequate in contrast to what is “nor- typically describes a single layered, mitotically
mal” atrophic endometrium is subjective. We inactive glandular epithelium with scant
consider a sample adequate, even if it is com- cytoplasm in a postmenopausal woman, inactive
posed of scant strips and fragments of atrophic endometrium may have somewhat similar histo-
endometrium (Figs. 2.23 and 2.24), if the logic features but with somewhat increased cyto-
gynecologist is confident that their sampling
plasm and retention of formed glands in biopsies.
instrument was in the endometrial cavity. If hys- Inactive changes can be a feature of exogenous
teroscopic examination the endometrial cavity hormonal effect in a premenopausal woman (see
is worrisome or if abnormal bleeding persists, Chap. 6). Weakly proliferative is another com-
hysteroscopy and curettage should be consid- monly used term, which we reserve for cases that
ered. Endometrial biopsies that only sample the most closely resemble proliferative endometrium
lower uterine segment or cervical tissue should with pseudostratified columnar epithelium and
Fig. 2.21 Basalis layer. Tissue from the endometrial basalis shows typical thick-walled blood vessels and should
not be misconstrued as elements of a polyp
inconspicuous mitotic activity. Most of these carcinoma. Fragmentation and close apposition
cases are seen in perimenopausal or postmeno- of disparate tissues such as cervical epithelium
pausal women with abnormal uterine bleeding. and functionalis also lead to confusing patterns.
Artifactually crowded glands lack a continuous
investment of tissue and are not connected by
Artifacts and Contaminants intervening stroma. These latter features help in
recognition of the artifact. Fragmentation also
Besides variations in the normal anatomy, such is a common feature of atrophy (Fig. 2.23) (see
as the basalis and lower uterine segment, several Chap. 5).
artifacts of the biopsy often complicate the his- Another frequent change is the so-called tele-
tologic patterns. One frequent artifact is tissue scoping of glands [28]. Telescoping may occur in
fragmentation caused by mechanical disruption either proliferative or secretory phase endome-
of the tissue by the sampling procedure. As a trium, but it also complicates many non-physiologic
result, glands are detached from the surround- conditions. Telescoping results in a pattern of an
ing stroma, and fragmented glands become ran- apparent gland within the lumen of another gland
domly oriented, often appearing closely spaced and can mimic hyperplasia or neoplasia (Figs. 2.26
(Fig. 2.25). This artifact should not be mistaken and 2.27). This artifact seems to be a result of
for real crowding that occurs in hyperplasia or mechanical disruption and “snap back” of the
Fig. 2.22 Artifactual crowding of proliferative endometrium. Glands are artifactually crowded on the right side. This
finding does not represent hyperplasia or other lesion
gland during curettage, resulting in intussusception absence of nuclear staining or “null” pattern, or
which rarely occurs in hysterectomy specimens. rarely, a cytoplasmic stain only with nuclear wild-
Tangential sectioning of tortuous glands also con- type pattern) is supportive of the diagnosis of SEIC,
tributes to this phenomenon. Fortunately, telescop- while papillary syncytial metaplasia related to
ing rarely presents difficulty in interpretation once breakdown is negative for these markers. A diffuse
the observer understands the phenomenon. In and strong p16 staining can be seen in both situa-
questionable cases, the bland cytology of the glan- tions and should not be taken into consideration
dular cells and comparison with surrounding tissue alone [29].
establishes this change as an artifact. Endometrial biopsies also often contain con-
Detached, superficial fragments of atypical- taminants from the cervix. Most of these contami-
appearing papillary syncytial metaplasia occasion- nants are obvious. Strips of bland squamous or
ally occur in a background of atrophy. The presence mucinous epithelium and irregular pools of extra-
of atypia may raise the concern for malignancy cellular mucin are common. The extracellular
such as serous endometrial intraepithelial carci- mucin may contain neutrophils, cell debris, macro-
noma (SEIC). The presence of mitotic figures cou- phages, or giant cells that are normal components
pled with a high Ki-67 proliferation index and an with no pathologic significance in the absence
aberrant pattern of p53 staining (strong/diffuse of inflammation in the endometrial stroma.
nuclear staining in >75% of tumor cells, complete Occasionally, benign cervical contaminants
Fig. 2.23 Atrophic endometrium. Scant strips of atro- (low columnar to cuboidal), no mitotic activity, and mini-
phic endometrium are characterized by a single layer of mal cytoplasm. The stroma is minimal or absent
inactive-appearing epithelium with darkly staining nuclei
become more complex and troublesome in biop- if it is of endometrial or endocervical type.

sies. Endocervical glands with squamous metapla- Rarely, an endometrial biopsy may also reveal
sia or microglandular hyperplasia yield complex an admixture of fragments of tissue from cervi-
patterns, but these elements are cytologically bland cal dysplasia, squamous carcinoma, or adeno-
and usually blend into more typical cervical epithe- carcinoma. Chapter 10 addresses the differential
lium (Figs. 2.28a, b and 2.29). Occasionally, sig- diagnosis of endocervical versus endometrial
net ringlike cells are present in the cervical stroma carcinoma.
following loop electrosurgical excision procedures Occasionally, curettage yields sheets of histio-
(LEEP). If an endometrial biopsy and LEEP proce- cytes with no associated mucin or other tissue.
dures are performed simultaneously, the endocer- These histiocytes apparently reside in the endome-
vical tissue with these signet ringlike cells is an trial cavity and show the typical histiocyte cytol-
important potential contaminant of the endometrial ogy with a lobulated nucleus and amphophilic
sampling [30]. cytoplasm. We have seen them in association with
In questionable cases, assessment of continu- hydrometra and with benign bleeding patterns.
ity with endometrial surface epithelium may They apparently represent a response to intracavi-
help to establish origin in the corpus. It is also tary debris and, when excessive sheets are formed,
helpful to look at the surrounding stroma and see have been referred to as “nodular histiocytic
Fig. 2.24 Atrophic endometrium. A few detached strips of atrophic epithelium are considered an adequate sample if
no hysteroscopic lesion is suspected
hyperplasia” because they can represent a nodular adipose tissue with clearly identifiable fat cells in
accumulation of histiocytes in sections [31]. an endometrial biopsy almost always represent
Nodular histiocytic hyperplasia is an uncommon omentum or extrauterine pelvic soft tissue and
phenomenon, however, which may raise the con- may indicate perforation of the uterus (Fig. 2.30a).
cern for neoplasia when histiocytes do not demon- In these circumstances, the clinician should be
strate their characteristic cytoplasmic foaminess. notified immediately. We have also occasionally
Instead, they may present with cytoplasmic vacu- seen colonic mucosa in endometrial biopsies,
oles that can mimic signet ring cells [32]. which may represent sample contamination dur-
When these histiocytic aggregates are promi- ing processing (“floater” tissue) or could represent
nent, the differential diagnosis can include other a fragment of bowel resulting from the biopsy or
more specific entities such as xanthogranuloma- curettage underscoring the importance of immedi-
tous endometritis, malakoplakia, Langerhans cell ate communication with the gynecologist.
histiocytosis, or even metastatic signet ring cell A relatively common artifact that merits rec-
carcinoma. More frequently though, we encoun- ognition is pseudolipomatosis (Fig. 2.30b).
ter small collections of histiocytes, both stromal Morphologically, pseudolipomatosis resembles
and intraglandular (see Chap. 7) [33, 34]. the presence of adipose tissue admixed with the
Immunohistochemical stains for histiocytes, endometrial sample. Nevertheless, it is secondary
such as CD68, can facilitate their recognition. to the introduction of air into the uterine cavity
Stromal foam cells, in contrast, represent stromal during insufflation for hysteroscopic procedures.
cells and macrophages that are filled with lipid Alternatively, it may be artifactually created by
from erythrocytes in areas of chronic non- the processing of specimens with scant cellular-
physiologic bleeding [33]. Separate fragments of ity and abundant hemorrhage [35].
Irregular Secretory Endometrium 31
Fig. 2.25 Artifactual fragmentation. Normal prolifera- arrangement but are detached from stroma and should not
tive phase endometrium is fragmented as a result of the be confused with a significant abnormality
sampling procedure. The glands have a haphazard
Finally, fallopian tube tissue can rarely pro- findings are non-specific but frequently include
lapse into the uterine cavity or cervix and be irregular glandular and stromal maturation with a
sampled during a biopsy or curettage. large variation in the pattern of endometrial devel-
Acquaintance with tubal morphology and its opment from field to field. For example, some
occurrence in this type of material prevent misdi- areas may show early secretory changes, while
agnosis of an otherwise benign condition. other areas show foci of predecidua consistent with
mid-late secretory phase. Clinically, this discordant
endometrial development is thought to contribute
Irregular Secretory Endometrium to a non-receptive endometrium which might cause
implantation failure [37, 38].
Irregular secretory endometrium is composed of The clinical significance of irregular secretory
secretory changes in a benign endometrial speci- endometrium is unclear as these patterns have
men that do not correlate with the normal expected been found in both fertile and infertile women
progression of the luteal phase. Examples of such [39]. Moreover, approximately one third of women
spectrum of changes include the changes previously with normal cycles have an isolated “out-of-
described as luteal phase defect and disordered phase” biopsy. Similarly, the rate of “out-of-
proliferative endometrium with glandular secre- phase” biopsies in infertile women does not appear
tory (postovulatory) changes [36]. Histologically, to be greater than what occurs by chance [40–43].
Fig. 2.26 Telescoping artifact. Late secretory endometrium with stromal predecidualization shows telescoping artifact
with a “gland-in-gland” appearance
Pathologic processes can cause irregular secre- infertility such as chronic endometritis or polyps.
tory endometrium (Table 2.2). When irregular Secretory phase changes indicate that ovulation
secretory endometrium is the result of specific has occurred. Assigning an early, mid-, or late
abnormalities such as inflammation or polyps, the secretory phase gives a general assessment of
primary abnormality may be evident. In other progesterone production by the corpus luteum
cases, there may be no identifiable etiology for and the ability of the endometrium to respond to
the abnormally developed secretory pattern. For progesterone.
this latter group of cases, the abnormality may be Dating the secretory phase, as previously noted,
dysfunctional, related to abnormal development has been shown by countless studies to be subjec-
of the corpus luteum, secondary to underlying tive and lack reproducibility. Therefore, it is also
pathology that is not adequately sampled, or important that everyone involved in the interpreta-
merely a result of a sporadic normal variation of tion and clinical application of histologic dating
the secretory phase. understands the limitations of this morphologic
assessment [8, 44]. Hormonal effects, various
structural lesions, and sampling problems all can
Clinical Queries and Reporting make dating difficult to impossible. Other patho-
logical factors such as inflammation, adhesions, or
Endometrial biopsies submitted for an infertility polyps may interfere with pregnancy (see Chap. 3,
workup should be evaluated for histologic evi- 7, and 8). These abnormalities affect fertility by
dence of ovulation and the exclusion of endome- altering the development of the glands and stroma,
trial abnormalities that may be responsible for thereby preventing normal implantation or
Fig. 2.27 Telescoping artifact. Proliferative endometrium can also show “gland in gland” artifactual change secondary
to the biopsy procedure. Telescoping has no clinical significance
a b
Fig. 2.28 Cervical contaminants. (a) Endocervical epi- monly harbors collections of lymphocytes and plasma
thelium with squamous metaplasia is very frequently cells. This should not be mistaken for chronic endometri-
encountered as a contaminant in endometrial curettings. tis and does not need a comment
(b) Endocervical tissue in an endometrial biopsy com-
mechanically disrupting the early implanting blas- to decide whether the endometrium is proliferative
tocyst. Therefore, when present, these abnormali- or secretory, because secretory phase development
ties should be reported. An attempt should be made generally indicates that ovulation has occurred.
Fig. 2.29 Cervical contaminants. Detached fragments with cervical epithelium and mucus frequently appear in endo-
metrial biopsy and curettage material
a b
Fig. 2.30 Uterine perforation. (a) Fragments of adipose fact created by air insufflation into the uterine cavity dur-
tissue in an endometrial biopsy almost always represent ing hysteroscopic procedures or the processing of
omentum or extrauterine pelvic soft tissue and indicate specimens with scant cellularity and abundant hemor-
perforation of the uterus. (b) Pseudolipomatosis, an arti- rhage, should not be confused with adipose tissue
References 35
Table 2.2 Causes of undatable secretory endometrium endometrium may evolve that have clinical util-
Hormonal effects ity. At present, however, routine histologic evalu-
Anovulation ation remains a cost-effective method of
Changes previously known as luteal phase defect determining the relative degree of endometrial
Persistent corpus luteum development through the menstrual cycle.
Exogenous hormones
Recently, a transcriptomics and bioinformatic
Pregnancy
tool based on machine-learning prediction (endo-
Structural/Pathological lesions
Polyps metrial receptivity array or ERA) has been used
Leiomyomas clinically in reproductive medicine to assess the
Chronic inflammation endometrium. The ERA test is a customized array
Hyperplasia containing 238 differentially expressed genes that
Carcinoma are coupled to a computational predictor, which is
Atrophy able to identify endometrial samples that are in
Sampling problems the optimal receptive period for implantation of a
Fragmentation free-lying blastocyst (window of implantation)
Lack of surface epithelium
regardless of their histologic appearance [51].
ERA and other molecular diagnostic tools may
On occasion, the biopsy shows an abnormal prove in the future more accurate and robust than
secretory pattern that cannot be histologically the standard H&E-based endometrial biopsy.
dated. The specimens can show either delayed Currently, however, morphologic evaluation of
development of glands or stroma or irregular the development of the endometrium remains a
maturation with a wide spectrum of secretory relatively rapid and cost-effective method of eval-
changes and does not necessarily indicate any uating normal and abnormal endometrial devel-
specific abnormality. In cases with altered secre- opment through the menstrual cycle.
tory phase development, it is best to use a care-
fully worded descriptive diagnosis. The term
“dyssynchronous endometrium” is a vague term References
that should not be used to describe apparent alter-
ations in secretory phase development. 1. Noyes RW, Hertig AT, Rock J. Dating the endometrial
biopsy. Fertil Steril. 1950;1:3–25.
As histologic endometrial dating has been 2. Dallenbach-Hellweg G. Histopathology of the endo-
shown to be both inaccurate and imprecise as a metrium. 4th ed. New York: Springer; 1987.
method for the diagnosis of luteal phase insuffi- 3. Hendrickson MR, Kempson RL. Surgical pathology
ciency or to guide the clinical management of of the uterine corpus (Major problems in pathol-
ogy series), vol. 12. Philadelphia: W.B. Saunders;
infertile women [10], morphometric analysis has 1980.
been attempted to increase the accuracy of the 4. Rock J, Bartlett MK. Biopsy studies of human endo-
former [45]. It was found that five morphometric metrium: criteria of dating and information about
measurements, including mitotic index in gland amenorrhea, menorrhagia, and time of ovulation. J
Am Med Assoc. 1937;108(24):2022–8.
cells, amount of luminal secretion, amount of 5. Murray MJ, Meyer WR, Zaino RJ, Lessey BA,
pseudostratification of gland cells, and amount of Novotny DB, Ireland K, Zeng D, Fritz MA. A criti-
predecidual reaction, added precision to histo- cal analysis of the accuracy, reproducibility, and clini-
logic dating [45]. Furthermore, immunohisto- cal utility of histologic endometrial dating in fertile
women. Fertil Steril. 2004;81(5):1333–43.
chemical analysis for specific secretory products 6. Koninckx P, Brosens JJ, Brosens I. Endometrial
of the endometrium and integrins may help to dating – still room for controversy. Fertil Steril.
identify specific causes of irregular secretory pat- 2005;83(6):1889–90. author reply 1890–1.
terns [46–49]. Other cell products such as cyclin 7. Coutifaris C, Myers ER, Guzick DS, Diamond MP,
Carson SA, Legro RS, McGovern PG, Schlaff WD,
E and p27 also may be useful in assessing secre- Carr BR, Steinkampf MP, Silva S, Vogel DL, Leppert
tory phase development [50]. These data indicate PC, NICHD National Cooperative Reproductive
that further refinements in evaluation of normal Medicine Network. Histological dating of timed
endometrial biopsy tissue is not related to fertility sta- metrium throughout the menstrual cycle. Hum
tus. Fertil Steril. 2004;82(5):1264–72. Immunol. 1989;24:195–205.
8. Duggan MA, Brashert P, Ostor A, Scurry J, Billson 23. Bulmer JN, Hollings D, Ritson
V, Kneafsey P, et al. The accuracy and interobserver A. Immunocytochemical evidence that endometrial
reproducibility of endometrial dating. Pathology. stromal granulocytes are granulated lymphocytes. J
2001;33:292–7. Pathol. 1987;153:281–8.
9. Myers ER, Silva S, Barnhart K, Groben PA, 24. Poropatich C, Rojas M, Silverberg
Richardson MS, Robboy SJ, Leppert P, Coutifaris C, SG. Polymorphonuclear leukocytes in the endo-
NICHD National Cooperative Reproductive Medicine metrium during the normal menstrual cycle. Int J
Network. Interobserver and intraobserver variability in Gynecol Pathol. 1987;6:230–4.
the histological dating of the endometrium in fertile 25. Phillips VP, McCluggage WG. Results of a question-
and infertile women. Fertil Steril. 2004;82(5):1278–82. naire regarding criteria for adequacy of endometrial
10. Fadare O, Zheng W. Histologic dating of the endome- biopsies. J Clin Pathol. 2005;58:417–9.
trium: accuracy, reproducibility, and practical value. 26. Gordon SJ, Westgate J. The incidence and man-
Adv Anat Pathol. 2005;12(2):39–46. agement of failed Pipelle sampling in a general
11. Haney AF. Endometrial biopsy: a test whose time has outpatient clinic. Aust N Z J Obstet Gynaecol.
come and gone. Fertil Steril. 2004;82(5):1295–6. dis- 1999;39:115–8.
cussion 1301–2. 27. Bakour SH, Khan KS, Gupta JK. Controlled analy-
12. Kazer RR. Endometrial biopsy should be abandoned sis of factors associated with insufficient sample on
as a routine component of the infertility evaluation. outpatient endometrial biopsy. Br J Obstet Gynaecol.
Fertil Steril. 2004;82(5):1297–8. discussion 1300–2. 2000;107:1312–4.
13. Parra-Herran C, Cesari M, Djordjevic B, Grondin 28. Numers CV. On the traumatic effect of curettage on the
K, Kinloch M, Köbel M, Pirzada A, Plotkin A, endometrial biopsy, with special reference to so-called
Gilks CB. Canadian consensus-based and evidence- invagination pictures and the “crumbling endome-
based guidelines for benign endometrial pathology trium”. Acta Obstet Gynecol Scand. 1942;28:305–13.
reporting in biopsy material. Int J Gynecol Pathol. 29. McCluggage WG, McBride HA. Papillary syncy-
2018. tial metaplasia associated with endometrial break-
14. Masterson R, Armstrong EM, More IAR. The cycli- down exhibits an immunophenotype that overlaps
cal variation in the percentage of ciliated cells in with uterine serous carcinoma. Int J Gynecol Pathol.
the normal human endometrium. J Reprod Fertil. 2012;31(3):206–10.
1975;42:537–40. 30. McKenna M, McCluggage WG. Signet ring cells of
15. Hendrickson MR, Kempson RL. Ciliated carcinoma-a stromal derivation in the uterine cervix secondary to
variant of endometrial adenocarcinoma. A report of cauterization: report of a previously undescribed phe-
10 cases. Int J Gynecol Pathol. 1983;2:1–12. nomenon. J Clin Pathol. 2008;61(5):648–51.
16. Hedrick Ellenson L, Ronnett BM, Kurman RJ. Precursor 31. Kim KR, Lee YH, Ro JY. Nodular histiocytic hyper-
lesions of endometrial carcinoma. In: Kurman RJ, edi- plasia of the endometrium. Int J Gynecol Pathol.
tor. Blaustein’s pathology of the female genital tract. 6th 2002;21:141–6.
ed. New York: Springer; 2011. p. 359–91. 32. Iezzoni JC, Mills SE. Nonneoplastic endometrial
17. Silverberg SG, Kurman RJ. Tumors of the uterine signet-ring cells. Vacuolated decidual cells and stro-
corpus and gestational trophoblastic disease, Atlas of mal histiocytes mimicking adenocarcinoma. Am J
tumor pathology, 3rd series, fascicle 3. Washington, Clin Pathol. 2001;15:249–55.
DC: Armed Forces Institute of Pathology; 1992. 33. Silver SA, Sherman ME. Morphologic and immuno-
18. Speroff L, Glass RH, Kase NG. Clinical gyneco- phenotypic characterization of foam cells in endome-
logic endocrinology and infertility. 6th ed. Baltimore: trial lesions. Int J Gynecol Pathol. 1998;17:140–5.
Lippincott Williams & Wilkins; 1999. 34. Parkash V, Domfeh AB, Fadare O. Nodular histiocytic
19. Buckley CH, Fox H. Biopsy pathology of the endo- aggregates in the endometrium: a report of 7 cases. Int
metrium. 2nd ed. London: Arnold; 2002. J Gynecol Pathol. 2014;33(1):52–7.
20. Mutter GL, Ferenczy A. Anatomy and histology of 35. Unger ZM, Gonzalez JL, Hannisian PD, Schned
the uterine corpus. In: Kurman RJ, editor. Blaustein’s AR. Pseudolipomatosis in hysteroscopically
pathology of the female genital tract. 5th ed. resected tissues from the gynecologic tract: patho-
New York: Springer; 2002. logic description and frequency. Am J Surg Pathol.
21. Bulmer JN, Lunny DP, Hagin 2009;33:1187–90.
SV. Immunohistochemical characterization of stromal 36. Mutter GL, Prat J. Pathology of the female reproduc-
leucocytes in nonpregnant human endometrium. Am J tive tract. 3rd ed. Cambridge, MA: Elsevier; 2014.
Reprod Immmunol Microbiol. 1988;17:83–90. 37. Jones GS. Luteal phase insufficiency. Clin Obstet
22. King A, Wellings V, Gardner L, Loke Gynecol. 1972;16:255–273.
YW. Immunocytochemical characterization of the 38. Wentz AC. Endometrial biopsy in the evaluation of
unusual large granular lymphocytes in human endo- infertility. Fertil Steril. 1980;33:121–4.
References 37
39. Bukulmez O, Arici A. Luteal phase defect: myth or 46. Graham RA, Seif MW, Aplin JD, Li TC, Cooke ID,
reality. Obstet Gynecol Clin N Am. 2004;31:727–44. Rogers AW, et al. An endometrial factor in unex-
40. Balasch J, Fabregues F, Creus M, Vanrell JA. The plained infertility. Br Med J. 1990;300:1428–31.
usefulness of endometrial biopsy for luteal 47. Seif MW, Aplin JD, Buckley CH. Luteal phase defect:
phase evaluation in infertility. Hum Reprod. the possibility of an immunohistochemical diagnosis.
1992;7:973–7. Fertil Steril. 1989;51:273–9.
41. Davis OK, Berkeley AS, Naus GJ, Cholst IN, 48. Lessey BA, Castelbaum AJ, Wolf L, Greene W,
Freedman KS. The incidence of luteal phase defect in Paulson M, Meyer WR, et al. Use of integrins to date
normal fertile women, determined by serial endome- the endometrium. Fertil Steril. 2000;73:779–87.
trial biopsies. Fertil Steril. 1989;51:582–6. 49. Ordi J, Creus M, Ferrer B, Fabregues F, Carmona F,
42. Wentz AC, Kossoy LR, Parker RA. The impact of Casamitjana R, et al. Midluteal endometrial biopsy
luteal phase inadequacy in an infertile population. Am and alphavbeta3 integrin expression in the evalua-
J Obstet Gynecol. 1990;162:937–45. tion of the endometrium in infertility: implications for
43. Shivapathasundram G, Kwik M, Chapman fecundity. Int J Gynecol Pathol. 2002;21:231–8.
M. Luteal phase defect: part of the infertility zeit- 50. Dubowy RL, Feinberg RF, Keefe DL, Doncel
geist or relic from the past? Hum Fertil (Camb). GF, Williams SC, McSweet JC, et al. Improved
2011;14(1):60–3. endometrial assessment using cyclin E and p27. Fertil
44. Zaino R, Novotney D, Ireland D, Meyer W, Fritz M, Steril. 2003;80:146–56.
Lessey B, et al. Dating the endometrial biopsy: a re- 51. Díaz-Gimeno P, Ruiz-Alonso M, Blesa D, Bosch N,
evaluation. Mod Pathol. 2003;16:216A. Martínez-Conejero JA, Alamá P, Garrido N, Pellicer
45. Li TC, Rogers AW, Dockery P, Lenton EA, Cooke A, Simón C. The accuracy and reproducibility of the
ID. A new method of histologic dating of human endometrial receptivity array is superior to histology
endometrium in the luteal phase. Fertil Steril. as a diagnostic method for endometrial receptivity.
1988;50:52–60. Fertil Steril. 2013;99(2):508–17.
Pregnancy, Abortion, and Ectopic
Pregnancy 3
Contents
Endometrial Glands and Stroma in Pregnancy 40
Other Glandular Changes in Pregnancy 46
Trophoblast and Villi 48
Endometrium Associated with Ectopic Pregnancy 62
References 71
Recognition of the features of gestational endo- woman and are occult [5, 6]. Most spontaneous
metrium, trophoblast, and villi, as well as the abortions occur before 12 weeks of pregnancy,
pathologic changes in chorionic tissues, is an and at least half of these are attributable to a
important part of endometrial biopsy interpreta- genetic (karyotypic) anomaly. An “incomplete
tion. The presence of intrauterine products of abortion” is a spontaneous abortion in which the
conception generally excludes the diagnosis of conceptus and decidua are incompletely passed,
ectopic pregnancy and can help explain other thus requiring curettage. A “missed abortion”
pathologic states such as abnormal bleeding or refers to an abortion with retained products of
chronic endometritis. conception but no abnormal bleeding for
An abortion before the 16th week of preg- 5–8 weeks after death of the embryo or fetus. The
nancy is the usual source of endometrial tissue criteria for diagnosis of a missed abortion vary
specimens that show gestational changes. The among practitioners and institutions. “Therapeutic
different types of abortions are defined as fol- abortions” are those in which the pregnancy is
lows. “Spontaneous abortions” are unexpected electively terminated.
and unplanned interruptions of pregnancy that Besides abortions, several other complications
present with bleeding and passage of tissue. of pregnancy, such as retained placenta or placen-
Approximately 15–20% of early pregnancies end tal implantation site, ectopic pregnancy, or gesta-
in a spontaneous abortion [1–4]. In addition, tional trophoblastic disease, lead to the need for
many other early pregnancies spontaneously endometrial curettage (Table 3.1). Specimens
abort before pregnancy is recognized by the from patients with these conditions show either

40 3 Pregnancy, Abortion, and Ectopic Pregnancy
Table 3.1 Complications of pregnancy recrudescence or accentuation of glandular

Spontaneous abortion secretions, distension of the glands, edema, and
Missed abortion an extensive predecidual reaction [7–10]. The
Retained placental tissue/implantation site coiled glands show secretory activity and a ser-
Ectopic pregnancy rated lumen, but they appear distended or wider
Placenta accreta, increta, and percreta
than those in the late secretory phase of a men-
Gestational trophoblastic neoplasms and nonneoplastic
lesions strual cycle (Figs. 3.1 and 3.2). Vascular promi-
nence with engorgement and dilation of
superficial veins and capillaries also occurs [8],
trophoblastic tissue, the effects of trophoblastic and the spiral arteries develop thicker walls.
tissue on the endometrium, or a combination of Other reported changes in the cycle of con-
trophoblastic tissue and its effects. ception include persistent basal cytoplasmic
This chapter first reviews the physiologic vacuoles in late secretory phase glands [9] and a
changes of the endometrium in pregnancy, espe- disparity between development of the glands
cially early pregnancy. This is followed by a dis- and stroma [11]. One study reported that in the
cussion of normal placental implantation and cycle of conception pronounced stromal edema
growth. The benign trophoblastic lesions of pla- and vascular engorgement of capillaries and
cental site nodule and exaggerated placental site, small veins correlated better with pregnancy
as well as gestational trophoblastic neoplasms, than did the glandular changes [7]. Because
are discussed in Chap. 4. menstruation does not occur, there is no sub-
stantial change in the number of true granular
lymphocytes at this time. Practically, however,
ndometrial Glands and Stroma
E the morphologic changes during the first
in Pregnancy 1–2 weeks after conception are subtle. It is dif-
ficult to decide whether the vacuolated cyto-
arly Gestational Endometrium
E plasm within glandular epithelium reflects
(1–3 Weeks Postfertilization) normal persistence of vacuoles in the late secre-
tory phase or the changes of pregnancy. It usu-
Fertilization occurs in the fallopian tube soon ally is not possible to be certain of
after ovulation, and implantation (nidation) of pregnancy-related changes until 2 weeks or
the developing blastocyst takes place on day 20 more after conception, when decidua, as
or 21 (postovulatory day 6 or 7). Implantation opposed to predecidua, is fully developed (see
occurs on the surface of the endometrium, usu- later).
ally on the midportion of the posterior wall. The Within 10–15 days of fertilization, the endo-
ovulatory cycle, during which fertilization and metrium gradually begins to show more charac-
implantation take place, is called the cycle of teristic changes of pregnancy as differentiation
conception. Immediately after implantation, of stromal cells into decidua progresses
subtle changes begin to appear in the glands and (Table 3.2). As compared to predecidual cells,
stroma, although the tissue retains the overall decidual cells are larger and contain more abun-
characteristics of the mid- to late secretory dant eosinophilic to amphophilic cytoplasm
phase for several days. An endometrial biopsy that may contain faint vacuoles (Figs. 3.3 and
or curettage performed inadvertently at this 3.4a, b). These cells become more clearly poly-
time, usually during infertility evaluation, may hedral with well- defined cell membranes.
not include trophoblast or disrupt the early ges- Nuclei of the decidualized stromal cells are
tation, yet will show very early pregnancy- round to oval and uniform, with smooth out-
related changes. These changes include lines, finely dispersed chromatin, and indistinct
Fig. 3.1 Early gestational endometrium, cycle of con- ened spiral arteries. In the absence of trophoblast or cho-
ception. Inadvertent endometrial biopsy in the cycle of rionic villi, these features are too subtle to be diagnostic of
conception shows distended, coiled glands and engorged early pregnancy until the stroma shows more advanced
vessels. Early decidual reaction is present around thick- decidual change
nucleoli. Occasional decidualized stromal cells hypersecretory activity, the glands become
are binucleate. Stromal granular lymphocytes highly coiled with prominent serrations and
persist in early pregnancy and are clearly evi- papillary folds of epithelium projecting into the
dent among decidual cells. The presence of lumen. The epithelial cells become stratified.
granular lymphocytes may suggest a chronic Concurrently, spiral arteries are more prominent
inflammatory infiltrate, but the granular lym- (Fig. 3.3).
phocytes, in contrast to inflammatory cells,
have characteristic lobated nuclei, and plasma
cells are not present. ndometrium in Later Pregnancy (4
E
Along with progressive decidual transforma- or More Weeks Postfertilization)
tion of the stromal cells, the glands and the ves-
sels undergo pronounced alterations. Secretory With advancing gestational age, pregnancy-
changes in the glands become more prominent related patterns become more pronounced and
with increased cytoplasmic vacuolization and distinctive (Table 3.2). The decidualized stro-
augmented luminal secretions that distend the mal cells are widespread and prominent, espe-
glands (Figs. 3.5, 3.6a, b). In addition to this cially as the cell borders become better defined,
Fig. 3.2 Early gestational endometrium. There is an coiled glands show secretory activity and a serrated
accentuation of glandular secretions, distension of the lumen, but they appear distended or wider than those in
glands, edema, and an extensive predecidual reaction. The the late secretory phase of a menstrual cycle
Table 3.2 Histologic changes of the endometrium in pregnancya

Duration of
pregnancyb Stroma Glands Vessels
1–3 weeks Edema, then Hypersecretory with cytoplasmic vacuoles and Spiral arteries begin to
progressive luminal secretions; saw-toothed, tortuous, thicken; superficial venules
decidual change distended; rare Arias-Stella reaction congested, dilated
4 or more Marked decidual Irregular and inactive; variable Arias-Stella Spiral arteries thickened;
weeks change reaction, clear cytoplasm, optically clear nuclei superficial venules dilated
Changes of endometrial glands, stroma, and vessels only
a
Duration from time of fertilization (day 15–16 of menstrual cycle). Add 2 weeks for time from last menstrual period
b
and they develop an epithelioid appearance. and placenta and as the decidua capsularis
Decidual cell nuclei become somewhat larger fuses with the decidua parietalis. These small
and may appear vesicular, but they maintain foci of necrosis, with a localized neutrophilic
their uniform contours. The decidua shows response, are physiologic. They do not reflect
small foci of physiologic necrosis during preg- an infectious or septic process and do not indi-
nancy, as it remodels during growth of the fetus cate a significant abnormality. The decidua
Fig. 3.3 Early gestational endometrium with decidual- cell borders, and uniform round to oval nuclei. Granular
ized stroma. Endometrial stroma from a first-trimester lymphocytes are numerous. Spiral artery walls are thicker
abortion shows prominent decidual transformation. The than in nongestational endometrium
decidualized cells have abundant, pale cytoplasm, distinct
a b
Fig. 3.4 Decidualized stroma. (a) Decidualized cells borders and uniform, round to oval nuclei. Their cyto-
become very prominent around spiral arterioles as preg- plasm displays scattered, small vacuoles
nancy advances. (b) Decidual cells have prominent cell
Fig. 3.5 Gestational endometrium. Hypersecretory pattern of endometrial glands in early pregnancy with extensive
cytoplasmic vacuolization
a b
Fig. 3.6 Gestational endometrium. (a) Hypersecretory display extensive cytoplasmic vacuolization. Decidualized
pattern of endometrial glands in early pregnancy with stroma was present in other areas of the sections
focal Arias-Stella change. (b) The hypersecretory glands
continues to contain a sprinkling of granular ing decidualization, the glands become inac-
lymphocytes that remain throughout gestation. tive (Fig. 3.7). Conversely, in areas where the
The hypersecretory pattern of the glands begins glands appear hypersecretory, the stroma often
to regress early in pregnancy, and with increas- is not decidualized (Fig. 3.6a). Usually a mix-
Fig. 3.7 Gestational endometrium. Endometrium from spontaneous abortion shows diffuse decidual reaction of
stroma. The glands are dilated and lined by inactive, flattened epithelial cells
ture of hypersecretory and inactive glands is p olyps when the endometrium grows but does
present. By the end of the first trimester, the not undergo cyclical shedding.
glands for the most part are inactive and have
lost their luminal secretions. In fact, as they
form irregular, dilated spaces with indistinct Arias-Stella Reaction
epithelium, they may be difficult to distinguish
from vascular channels. At 4–8 weeks after blastocyst implantation, the
As pregnancy advances, the spiral arteries endometrium often shows at least a focal Arias-
maintain thick walls, a feature that persists to Stella reaction in the glands [13–17]. This glan-
term and is helpful in recognizing gestational dular change is a physiologic response to the
changes. Some authors suggest that in the first presence of chorionic tissue either in the uterus or
trimester, the arteries develop a characteristic at an ectopic site. The morphologic features of
atherosclerosis-like change when an intrauterine the Arias-Stella reaction include nuclear enlarge-
pregnancy is present, characterized by subintimal ment up to three times normal size and nuclear
proliferation of myofibroblasts with foam cells hyperchromasia, often accompanied by abundant
[12]. In addition, the venules beneath the surface vacuolated cytoplasm (Figs. 3.8a, b, 3.9a, b). The
epithelium dilate. Dilated superficial venules are cells typically are stratified and the nuclei hob-
not a specific change of pregnancy, however, as nail-shaped, bulging into the gland lumen. These
they may also be observed as a result of progestin large nuclei may contain prominent cytoplasmic
stimulation, hyperplasia, and occasionally in invaginations [18]. Mitotic figures are rarely
a b
Fig. 3.8 Arias-Stella reaction. (a) Glands from an abor- (b) The glands also show marked cytoplasmic vacuoliza-
tion specimen show prominent Arias-Stella reaction with tion. Identical changes can be seen in an ectopic
hyperchromatic nuclei that bulge into glandular lumen. pregnancy
a b
Fig. 3.9 Arias-Stella reaction. (a) Arias-Stella reaction plasmic invaginations. (b) Arias-Stella reaction with a
with “regenerative” pattern shows hobnail cells with “hypersecretory” pattern shows stratified nuclei and vacu-
dense cytoplasm. Several nuclei show prominent cyto- olated cytoplasm
present, and a Ki-67 labeling index demonstrates change remains unsubstantiated. This pattern is
a very low proliferative rate. This process may be characterized by glands lined by enlarged hob-
extensive, involving many glands, or the reaction nail cells with little cytoplasmic secretory activ-
can be focal, involving only a few glands ity. In fact, the two patterns are not very distinct,
(Fig. 3.10). The change can even be limited to and there is frequent overlap between them.
part of a gland, leaving the remaining nuclei The degree and extent of the Arias-Stella
unaffected. reaction are highly variable in normal and
The Arias-Stella reaction has two histologic abnormal intrauterine gestation, in ectopic
patterns (Fig. 3.9a, b) [15]. One is a “hyperse- pregnancy, and in gestational trophoblastic
cretory” change characterized by highly convo- neoplasms [15, 19–22]. This change occurs as
luted glands lined by cells with stratified nuclei early as 4 days after implantation [14], although
and abundant clear to foamy cytoplasm. The it generally is seen after about 14 days [15].
other pattern has been termed “regenerative,” The Arias-Stella reaction persists up to at least
although this hypothesized etiology for the 8 weeks following delivery. There is no appar-
Other Glandular Changes in Pregnancy 47
Fig. 3.10 Gestational endometrium with Arias-Stella enlarged, and hyperchromatic nuclei (left of image),
reaction. A cluster of hypersecretory glands in decidual- while a dilated, inactive gland (right of image) does not
ized stroma shows Arias-Stella reaction with stratified, show Arias-Stella change
ent relationship between the presence and gland cells accumulate abundant amounts of
extent of the Arias-Stella reaction and the status clear, glycogen-rich cytoplasm. The nuclei in
of the fetus. The Arias-Stella reaction is almost areas of clear cell change can become stratified,
unique to pregnancy or gestational trophoblas- which, combined with the abundant clear cyto-
tic neoplasms. Similar phenomena are rarely plasm, can result in apparent obliteration of the
produced by administration of exogenous pro- gland lumens.
gestins [23]. Another pregnancy-related change is optically
clear nuclei of gland cells (Fig. 3.11) [24, 25]. This
alteration is also often associated with the Arias-
ther Glandular Changes
O Stella reaction but can occur independently.
in Pregnancy Optically clear nuclei usually are focal. They have
a clear to glassy appearance that is caused by accu-
Besides the Arias-Stella reaction, the endome- mulation of a filamentous material in the nuclei.
trial gland cells may undergo other specific There is an indication that the change is related to
changes in the presence of trophoblastic tissue. the intranuclear accumulation of biotin [24]. Clear
One such change is abundant clear cytoplasm nuclei can mimic the changes seen in herpesvirus
[13]. This phenomenon overlaps with the Arias- infection, although the optically clear nuclei asso-
Stella reaction yet does not show the nuclear ciated with pregnancy lack the Cowdry type A
enlargement of the latter. With this change the eosinophilic nuclear inclusions, nuclear molding,
Fig. 3.11 Gestational endometrium with optically clear nuclei. Crowded glands show prominent optically clear nuclei.
This gland cell change is infrequent and usually is focal. It may be seen throughout pregnancy, however
and associated necrosis seen in the viral infection. another unusual gland response to a concurrent
This alteration is infrequent, occurring in fewer pregnancy [26]. Rarely, frank endometrioid
than 10% of first-trimester abortion specimens. carcinoma also may be associated with intrauter-
These changes may persist until term, however. As ine gestation [27, 28].
in the Arias-Stella reaction, optically clear nuclei In early pregnancy, endometrial glands
simply reflect the presence of chorionic tissue. become strongly immunoreactive for S-100 pro-
Localized atypical-appearing endometrial tein [29, 30]. This immunoreactivity rapidly dis-
glandular proliferations rarely may be found dur- appears after the 12th week of gestation. Normal
ing gestation [26]. These focal abnormalities proliferative and secretory endometrium, as well
show glandular expansion with nuclear stratifica- as glands in patients with hyperplasia and neopla-
tion and cribriform change. Mitotic activity is sia, do not stain for S-100 protein. There are no
present, but nuclear cytology is bland. other markers of the glands that are generally
Intraglandular calcifications frequently are pres- practical for identifying pregnancy-related
ent. The few lesions studied have been benign, changes except for the low Ki-67 labeling index
and some patients have had subsequent pregnan- of the Arias-Stella reaction, which helps to indi-
cies, suggesting these proliferative foci represent cate a benign glandular change.
Trophoblast and Villi gestation, implanting trophoblast is the predomi-

nant component of placental tissue. The tropho-
In early pregnancy, trophoblastic proliferation blast continues to grow along this interface of
begins with the development of the blastocyst, maternal and placental tissue throughout preg-
the outer layer of which is termed the trophoblas- nancy. The decidua basalis where trophoblast
tic shell. Villous formation does not begin until interfaces with the endometrium and myome-
about 7 days after implantation of the blastocyst trium becomes the placental implantation site.
(13 days following conception) [4]. For morpho- The trophoblastic cells are the epithelial compo-
logic identification, the products of conception nent of the placenta and are divided into three
are divided into three components: (1) the villi cytologically and functionally distinct popula-
and their trophoblast (“villous” trophoblast) tions: cytotrophoblast (CT), syncytiotrophoblast
(Fig. 3.12), (2) the implantation site (“extravil- (ST), and intermediate trophoblast (IT)
lous” trophoblast), and (3) fetal tissues. Usually (Table 3.3) [31–35]. Trophoblastic cells can also
these tissues are easy to recognize. be classified according to their anatomic location
Identifying trophoblast or villi is essential for as “villous” and “extravillous” trophoblast [36].
confirming the diagnosis of an intrauterine ges- CT cells are the germinative cells from which
tation. Also, the presence of placental and fetal other trophoblastic cells differentiate.
tissue in curettage samples, for all practical pur- Accordingly, they are mitotically active. They
poses, rules out an ectopic pregnancy. The mor- are uniform cells about the size of a decidualized
phologic features of these abortion specimens stroma cell, with a single nucleus, one or two
can be highly varied. Occasionally the diagnostic nucleoli, pale to faintly granular cytoplasm, and
features of the products of conception are dif- prominent cell borders. ST cells, in contrast, are
ficult to identify, especially in early pregnancy, larger and multinucleate with dense amphophilic
when the placental component is very small and to basophilic cytoplasm. The nuclei of ST cells
often missed in small biopsy specimens, or if are dark and often appear pyknotic; they do not
most of the products of conception were expelled contain mitoses. Their cytoplasm also typically
prior to the curettage. When villi and fetal tissue contains small vacuoles (Fig. 3.13) and larger
are not present in curettage samples, trophoblas- lacunae in which maternal erythrocytes can be
tic cells should be searched for to confirm the identified. A microvillous brush border some-
diagnosis of intrauterine pregnancy. Recognizing times lines the lacunae of the ST cells. CT and
the full morphologic spectrum of normal tropho- ST cells typically display a dimorphic growth
blastic cells is important, not only for establish- pattern, with the two cell types growing in close
ing the presence of an intrauterine pregnancy but proximity. In early abortions, the CT and ST
also for distinguishing exaggerated but physi- cells are quite prominent compared with the
ologic changes from gestational trophoblastic amount of villi present (Fig. 3.14). In very early,
neoplasms. unanticipated abortions, the entire products of
conception consist of previllous trophoblast that
can be easily confused with choriocarcinoma
Trophoblastic Cells (Figs. 3.15 and 3.16) (see Chap. 4). In curettage
for suspected abortions, sometimes the only evi-
The trophoblast is extraembryonic but fetal in dence of an intrauterine pregnancy is the pres-
origin, growing in intimate association with host ence of a few isolated trophoblastic cells mixed
maternal tissues. Very early in pregnancy tropho- with blood, and these may be necrotic (Fig. 3.13).
blastic cells differentiate and invade decidua, Careful scrutiny may be necessary to identify
even before villi form [4]. At this stage of early these diagnostic cells.
Table 3.3 Morphologic and immunohistochemical features of intermediate trophoblastic cells in the first trimester
Villous IT Implantation site IT Chorionic-type IT
Morphology Polyhedral; abundant, Pleomorphic and large; Round to polyhedral,
eosinophilic to clear cytoplasm; abundant, eosinophilic regular abundant
prominent cell borders cytoplasm; occasional eosinophilic and clear
multinucleated cells cytoplasm
Immunohistochemical profile
Cytokeratina ++++b ++++ ++++
EMA − − +++
hCG − –; + in multinucleated IT −
hPL −/+ ++++ ++
Mel-CAM −/++++c ++++ ++
PLAP − − +++
p63 −/+ − +++
HSD3B1 +++ +++ +++
Inhibin-α − +/− ++
Ki-67 index >90%d 0 3–10%
β-catenin +++e +++e +++e
HLA-G +++f +++ +++
hCG human chorionic gonadotropin, hPL human placental lactogen, Mel-CAM melanoma cell adhesion molecule (CD
146), EMA epithelial membrane antigen, PLAP placental alkaline phosphatase
a
Positive for cytokeratins 7, 8, 18, 19, and AE1/AE3; variable for cytokeratin 20; and negative for high molecular weight
keratin
b
Semiquantitative scoring of a proportion of positive cells
c
Mel-CAM staining increases from the base to the tip of the trophoblastic column. While high levels of expression are
seen in implantation-type IT, only medium levels are seen in chorionic-type IT
d
Ki-67 labeling index decreases from the base to the tip of the trophoblastic column
e
β-catenin shows a high level of expression in all intermediate trophoblastic cells (membranous) and cytotrophoblast
(nuclear), while syncytiotrophoblast does not express this marker
f
HLA-G is highly expressed in all intermediate trophoblastic cells, while negative in cytotrophoblast and
syncytiotrophoblast
The intermediate trophoblast develops from features of these IT subtypes depend on their dif-
cytotrophoblast on the villous surface and in ferentiation status and their anatomic location
early pregnancy is manifested as sprouts and col- (Table 3.3). The immunohistochemical staining
umns that extend to and extensively infiltrate the is discussed in greater detail below.
underlying decidua at the implantation site (see The IT that extends from the trophoblastic
later). In fact, the predominant location of the IT column of the anchoring villi is designated “vil-
is at the implantation site, which explains why it lous” IT [37]. These cells are mononucleate and
is often called “extravillous cytotrophoblast” larger than CT cells. They have pale cytoplasm
[36]. This latter term is less precise, however, as and large, round nuclei. The implantation site IT
these cells also occur in association with villi, cells infiltrate the decidua and the myometrium
and they are immunohistochemically and physi- and have a heterogeneous appearance, as dis-
ologically different from cytotrophoblast. cussed in the following paragraphs. The
Another older term for IT is “X cells” [4]. The IT chorionic-type IT constitute the cells of the cho-
actually represents a heterogeneous population rion laeve where they form a cohesive layer of
of trophoblastic cells: the villous IT, the implan- epithelium. These latter IT cells are composed of
tation site IT, and the chorionic-type IT [31, 37]. relatively uniform cells with eosinophilic to
The morphologic and immunohistochemical clear (glycogen-rich) cytoplasm. These cells are
Fig. 3.12 Villous trophoblast. The small, dark, and uni- cytiotrophoblast (ST), while the mononucleate cells with
form cells overlying the villi are the cytotrophoblast (CT). intermediate size between CT and ST and pale cytoplasm
The multinucleated cells with pyknotic nuclei are the syn- are the intermediate trophoblast (IT)
smaller than implantation site IT although an

occasional cell is multinucleated. Immunohistochemistry
The villous and the implantation site IT consti- of Trophoblastic Cells
tute the two forms of IT usually seen in biopsy spec-
imens from early pregnancy. The villous IT cells are Trophoblastic cells express a number of proteins
readily recognized because they are associated with that can be detected by commercially available
anchoring villi, but the implantation site IT may antibodies. Cytokeratin is the most ubiquitous
pose a greater challenge in recognition because the cell product since trophoblasts are epithelial
implantation site may be seen with no associated cells [31]. Broad-spectrum antibodies such as
villi, or the implantation site can appear prominent AE1/AE3 diffusely stain all trophoblastic cells.
or “exaggerated” (see Chap. 4). Implantation site IT Trophoblast is also reactive for simple
cells are the primary cell type of the exaggerated epithelium-type cytokeratins, and cytokeratins 7
placental site (EPS) and placental site trophoblastic and 18 are especially notable for their reactivity
tumor (PSTT), while the chorionic-type IT consti- in trophoblast. Cytokeratin 20, in contrast, is
tute the cell population seen in the placental site variably reactive in trophoblast, and high molec-
nodule (PSN) and in the epithelioid trophoblastic ular weight cytokeratin (34βE12) is negative in
tumor (ETT) both discussed in Chap. 4. trophoblastic cells.
Other markers are more specific for tropho- trophoblastic cells although it too is nonspecific,
blast with some antigens showing specific differ- staining other cells and tumors including breast
ences between the various types of trophoblast, and bladder carcinoma [39, 43]. Cyclin E, which
ST, IT, and CT. These antigens include the participates in the cell cycle control, is expressed
β-subunit of human chorionic gonadotropin in the trophoblastic columns of the anchoring
(hCG), placental alkaline phosphatase (PLAP), villi and implantation site IT cells. Cyclin E is
human placental lactogen (hPL), melanoma cell differentially expressed in benign and malignant
adhesion molecule (Mel-CAM or CD 148), and chorionic-type IT cells [44].
inhibin-α. New diagnostic markers have emerged The proliferation marker Ki-67 also has util-
that assist in the detection of subtypes of tropho- ity in evaluating trophoblast. The Ki-67 label-
blastic populations. These include human leuko- ing index is differentially expressed in
cyte antigen-G (HLA-G) and hydroxyl-δ-5-steroid trophoblastic cells and their tumors and tumor-
(HSD3B1). Other useful but less specific markers like lesions. For instance, ST cells and implan-
include p63, GATA-3, and Cyclin E. tation site IT are typically negative for Ki-67,
The β-subunit of human chorionic gonadotro- reflecting their non-proliferative nature, while
pin (hCG) is synthesized by ST cells throughout up to 50% of CT or germinative trophoblastic
gestation and is the ideal marker for these cells in cells express such marker. Chorionic-type IT
normal and pathologic conditions. Placental alka- expresses up to 3–5% of proliferative activity
line phosphatase (PLAP) and cytokeratin all react by Ki-67 [45].
with trophoblastic cells, but the degree of reactiv- The ST cells stain for the common trophoblas-
ity varies among the cell types. Human placental tic markers including hCG, hPL, PLAP, inhibin-α,
lactogen (hPL) is a pregnancy-related hormone as well as cytokeratin [46]. The ST does not react
synthesized by ST and IT cells in implantation with antibodies to Mel-CAM [47]. In CT cells,
sites (see Chap. 4). Mel-CAM (CD 146) is seen in except for cytokeratins, immunoreactivity to
all types of IT and serves as a useful marker for commonly available antibodies is limited. The
these cells. Inhibin-α also is another useful marker CT cells do not stain for hCG, hPL, HLA-G,
for subtypes of IT (see below). PLAP, HSD3B1, and other antigens associated
HLA-G is expressed in extravillous tropho- with specialized trophoblast such as Mel-CAM
blasts, i.e., IT, and is very useful in demonstrating and inhibin-α [38].
these cells. IT cells stain with HLA-G in 78–100% Table 3.3 summarizes the staining pattern of
of cases [38–40]. different subtypes of IT cells [37]. All IT cells
HSD3B1 is expressed by implantation site and react strongly for cytokeratin. Mel-CAM and
chorionic-type IT cells, ST in early placentas and HSD3B1 also are seen in all types of IT,
complete hydatidiform moles, as well as tropho- although with Mel-CAM, the villous IT staining
blastic neoplasms (placental site trophoblastic increases from the base to the tip of the tropho-
tumor and epithelioid trophoblastic tumor, see blastic columns. The implantation site IT and
Chap. 4). The importance of HSD3B1 it that IT chorionic-
type IT are reactive for hPL and
appears to be specific to trophoblastic tissue and PLAP, but the villous IT cells generally are not.
can be useful in situations a non-trophoblastic Inhibin-α is typically seen only in chorionic-
lesion must be excluded [41, 42]. type IT. Chorionic-type IT cells also stain well
p63 is another useful immunomarker that for epithelial membrane antigen (EMA); it is the
shows specificity for chorionic-type IT while only type of trophoblastic cell that is EMA reac-
lacking expression in implantation-type IT [33]. tive. Ki-67 immunostaining also shows a vari-
GATA3 also has been shown to be a marker of able proliferation index. CT shows a Ki-67
Fig. 3.13 Isolated ST cell with vacuoles in an abortion An isolated ST cell such as this is often found in areas of
specimen. Occasionally, a few ST cells mixed with fibrin hemorrhage
and blood are the only evidence of intrauterine pregnancy.
index of 25–50%, which is consistent with its Placental Implantation Site

role as the germinative trophoblastic cell as pre-
viously mentioned. The Ki-67 proliferation The placental implantation site is composed largely
index in villous trophoblast is high at the junc- of IT cells that infiltrate the decidua basalis, mixing
tion with the CT at the base of the villus but with decidualized stromal cells, glands, and vessels
decreases progressively toward the tip (distal [4, 31, 32, 36, 48–50]. This site begins as a micro-
end) of the trophoblastic column. The Ki-67 scopic focus where the blastocyst implants into
proliferation index is zero at the junction of the secretory endometrium, and it expands with the
column with the decidua of the basal plate, as growing placenta to cover the entire area of decidua
well as in the implantation-type IT cells as they and superficial myometrium to which the placenta
“drop off” the column and infiltrate the endo- is attached. It can be diffuse or focal in endometrial
myometrium. The Ki-67 proliferation index is biopsy specimens. Previous terminology for the
low (<5%) in chorionic-type IT. The Ki-67 trophoblastic infiltrate in decidual and myometrial
index in ST cells is zero, which is consistent tissue included “syncytial endometritis” and “syn-
with that of terminally differentiated cells. cytial endomyometritis” or “placental giant cell
Fig. 3.14 Immature chorionic villi with trophoblast. inner layer of mononucleated CT cells capped by multinucle-
Syncytiotrophoblast and cytotrophoblast grow from the sur- ated ST cells. To the bottom of the villus, villous IT cells pro-
face of immature first-trimester villi. The villi are lined by an liferate toward the implantation site in the placental tissue
reaction.” None of these terms is correct, however, plasm [49]. They are larger than decidualized
because the process is physiologic, not inflamma- stromal cells, with which they are intimately
tory; the majority of the cells are not giant cells; admixed. They may have sharply outlined
and the change is not confined to the endometrium cytoplasmic vacuoles. The nuclear morphol-
but also involves the myometrium. ogy of implantation site IT, however, is the
Placental site intermediate trophoblast char- most important feature that distinguishes these
acteristically diffusely permeates the decidua cells from decidua. They are enlarged, lobated,
and implantation site (Figs. 3.17, 3.18, 3.19 and and hyperchromatic with irregular nuclear
3.20). Because they closely resemble decidual membranes. Sometimes they have deep clefts,
cells, they are often difficult to recognize. In and some nuclei appear smudged [51]. Most
fact, these IT cells have been misinterpreted of these IT cells contain a single nucleus, but
as degenerating decidual cells because of their bi- or multinucleate cells with similar nuclear
intimate association with the latter (Figs. 3.16 and cytoplasmic features occur as well. The
and 3.17). The implantation site IT cells have dark and irregular implantation site IT nuclei,
variable size and shape ranging from polygo- which often contain a prominent nucleolus,
nal to round to spindle-shaped, with a moderate contrast with the nuclei of decidualized stromal
amount of eosinophilic to amphophilic cyto- cells, which are uniform and round to oval
Fig. 3.15 Trophoblast of early pregnancy. A mixed pop- nuclei in the lower left part of the field are IT cells that are
ulation of ST and CT is encountered in early abortion infiltrating the decidua
biopsies. The cells with pleomorphic, hyperchromatic
with an even, delicate chromatin distribution. trated by intermediate trophoblast in curettage

Immunohistochemical stains for keratin and samples from abortions.
human placental lactogen (hPL) help identify In addition to infiltrating the decidua and myo-
intermediate trophoblast cells and distinguish metrium, the implantation site IT cells invade spiral
them from decidual cells which are negative for arteries, extensively replacing the wall and endothe-
such immunostains [32, 52]. lium while maintaining the integrity of the lumen
At the implantation site, intermediate tropho- (Figs. 3.19 and 3.22). When vascular infiltration is
blastic cells also infiltrate into myometrium, extensive, the IT cells may partially fill the lumen of
where they often have a spindled appearance. the vessel. This infiltration of vessels contributes to
The implantation site IT cells are often more con- the enlargement of the spiral arteries.
spicuous here, especially when some of them Besides their characteristic growth pattern and
become multinucleated (Fig. 3.21). In the myo- cytologic features, implantation site IT cells can
metrium they infiltrate between muscle bundles be recognized because they typically are associ-
and fibers, often with no evidence of a tissue ated with patches of extracellular eosinophilic
reaction. Many are spindle shaped and can fibrinoid material (Fig. 3.23). This fibrinoid
closely resemble smooth muscle cells. It is very matrix in the placental bed eventually becomes
common to see fragments of myometrium infil- the so-called Nitabuch’s layer. When this fibri-
Fig. 3.16 Trophoblast of early pregnancy. A prominent of villi, this pattern can resemble choriocarcinoma, and
trimorphic network of ST, IT and CT encountered in an clinical history may be required to determine the signifi-
early abortion biopsy. No villi are present. In the absence cance of the finding
noid material becomes disrupted in abortions, it absence of interspersed IT cells and the lack of
forms hyaline, eosinophilic strands termed the linear deposits of eosinophilic implantation
Rohr’s stria. The origin of this fibrinoid, hyaline site fibrinoid.
material is not fully understood, although it Changes in the spiral arteries of the placental
appears to be partly composed of fibronectin, site are significant. The presence of enlarged, hya-
laminin, type IV collagen, and a small amount of linized spiral arteries in the decidua of curettage
fibrin. In any event, fibrinoid is a distinctive part specimens is a valuable adjunct in diagnosing
of the implantation site. intrauterine gestation [52]. These vessels have
Placental site fibrinoid can resemble fibrin thickened hyalinized walls that are partially infil-
thrombi that occur as a result of chronic bleeding trated by IT and show an increased luminal diam-
from a variety of causes. Also, endometrial eter (Figs. 3.19 and 3.22). Often several
stroma can become hyalinized and fibrotic in implantation site vessels are seen in cross section
areas of continued breakdown such as the surface forming a prominent cluster. These vascular
of polyps, and this change, too, can mimic the changes, like the presence of IT in the decidua, are
fibrinoid deposition of the implantation site. characteristic of the implantation site and are not
Fibrin thrombi and hyalinized endometrial found in endometrium associated with ectopic
stroma are distinguished from fibrinoid by the pregnancy.
Fig. 3.17 Placental implantation site. Fibrin and implan- blood vessel. Chorionic villi are not seen, but the presence
tation site IT are interspersed among decidual cells. The of intermediate trophoblast in decidua establishes the
IT cells are partially replacing the endothelial cells of a diagnosis of intrauterine pregnancy
Histologic recognition of the placental implan- HSD3B1, Mel-CAM (CD 146), and, to a lesser
tation site usually is straightforward. Sometimes degree, hCG [31]. Immunostains, especially hPL,
IT cells are indistinct or are difficult to distin- are useful for detecting IT cells including the
guish from degenerating decidual cells that multinucleated IT in the myometrium, as smooth
develop dark nuclei. In these instances, ancillary muscle cells do not express hPL [32, 52–54]. The
immunohistochemical techniques are useful for implantation site IT show no proliferative activity
the detection of IT. Broad-spectrum keratin anti- and are not reactive with Ki-67 [55].
bodies are very useful for demonstrating IT [32,
52–54]. Endometrial glands also stain with kera-
tin, however, and therefore all keratin-positive Chorionic Villi and Villous
cells do not represent IT cells. It is the growth Trophoblast in the First Trimester
pattern along with immunoreactivity for keratin
that identifies IT cells. With the keratin immu- Villus formation in very early pregnancy
nostain, the IT cells appear as intensely staining depends on the existence of the embryonic disc.
single cells or irregular clusters of cells with Villi begin to develop on the 12th to 13th day
intervening decidua or smooth muscle that is postfertilization, and by days 12–15, the pla-
nonreactive for keratin. The IT cells, unlike centa can develop for a while independently,
decidual cells, also express hPL, HLA-G, without the presence of an embryo [4, 56]. In
Fig. 3.18 IT cells, decidua, and vessels. Placental implanta- the spiral arterioles where fibrin is abundant (left upper cor-
tion site in an abortion specimen contains numerous implan- ner). IT can be recognized at low magnification by their char-
tation site IT cells that infiltrate the decidua and cluster around acteristic large, irregular, and hyperchromatic nuclei
fact, in spontaneous abortions, the placenta can infrequent clinical question. These developmen-
persist for several weeks after the death of the tal intervals are stated in relation to the time of
embryo. In the early stages of pregnancy, villi fertilization, also known as the postcoital or post-
have a loose, edematous stroma (Fig. 3.24a) conception date. Taking into account the time
with few well-developed capillaries. Once the from the last known menstrual period adds
yolk sac and embryo develop, vascular circula- 2 weeks to these figures. The conceptus is tradi-
tion is established in the villous stroma. These tionally called an embryo in the first 2 months of
vessels contain nucleated red blood cells development; thereafter it is called a fetus.
(Fig. 3.24b). During this early period of placen- As placental development proceeds, the pres-
tal development, the trophoblastic covering of ence of nucleated erythrocytes produced in the
the villi consists of an inner layer of CT cells yolk sac helps determine the approximate gesta-
and an outer layer of ST cells. The CT and IT tional age [56]. The nucleated erythrocytes from
cells also proliferate at the implanting end of the the yolk sac appear in the villous circulation at
anchoring villi that grow along the basal plate of 4.5 weeks (Fig. 3.24b). By 5–6 weeks, nonnucle-
the developing placenta. ated erythrocytes from the embryonic liver also
A few histologic changes in the placenta help begin to appear in the villi, and from this point
to determine the age of the developing conceptus on, there is a shift in the proportion of nucleated
(Table 3.4), although the length of gestation is an to nonnucleated erythrocytes. By 9 weeks post-
Fig. 3.19 IT cells and spiral arteries. Spiral arteries in the center show the wall replaced by IT and fibrinoid material
fertilization, the percentage of nucleated erythro- to some extent throughout gestation, but a visible
cytes in the villi decreases from 100% to only inner layer of CT cells starts to disappear at about
10%. Consequently, if the embryo dies before 14 weeks of gestation. Between weeks 14 and 18,
4.5 weeks postfertilization, the villi contain no the percentage of villi showing an inner layer of
red blood cells. Death of the embryo between 4.5 CT cells decreases from 80% to 60% [4]. From
and 10 weeks leaves a mixture of nucleated and that point on, there is a continuing gradual
nonnucleated erythrocytes in villous capillaries, decrease in the CT layer. The large decrease in
and later death of the embryo usually leaves non- identifiable CT cells by week 18 is especially
nucleated red cells in the stroma. useful for determining the relative duration of
Other features also help to determine the rela- early pregnancy.
tive length of gestation. For example, normally
developing immature villi are relatively larger
than those of later pregnancy. They have a loose, ydropic Change and Other
H
myxoid stroma with widely spaced capillaries. Pathologic Changes in Abortions
As pregnancy progresses, the villi become
smaller but more vascular, and their stroma loses The microscopic features of the decidua and the
its edematous appearance. The morphologic products of conception in curettage samples
features of the trophoblast covering the villi
vary depending on the type of abortion [4, 57,
change along with growth of the placenta. The 58]. Villi are usually normal in therapeutic abor-
bilayered CT and ST covering of the villi persists tions, whereas they tend to reflect early death of
Fig. 3.20 IT cells and decidua. Implantation site IT with by the IT cells. The implantation site IT have irregular,
enlarged, hyperchromatic nuclei infiltrate decidua in an hyperchromatic nuclei that contrast with the round to
abortion specimen. Several inactive glands are surrounded oval, uniform nuclei of the decidualized stromal cells
the embryo in spontaneous or missed abortions. development at a very early stage of gestation.
Therapeutic abortion specimens may show Microscopically, villous edema in a hydropic
pathologic changes in the villi, however [59]. In abortion can appear especially prominent at first
spontaneous or missed abortions, placental mor- glance. Hydropic change affects most villi but is
phology is influenced by gestational age, karyo- minimal and microscopic; cistern formation in
type, and regressive changes [4, 57, 60–63]. the villi is rare but does occur (Fig. 3.26). There
With the death of the embryo, the villi often is no associated trophoblastic hyperplasia
show hydropic change because of loss of the vil- except the normal growth at one pole of the
lous vascular supply, especially if embryonic anchoring villi. Usually these microscopic
death occurs very early, often before 4.5 weeks abnormalities are less impressive when the
postfertilization age [56]. The avascular villi are gross, quantitative aspects are considered also.
mildly distended with fluid, and the curettage Hydropic abortions usually consist of one or
samples do not contain fetal tissue giving the two cassettes of tissue with villi, whereas moles
changes of the so-called blighted ovum typically yield multiple cassettes. It is important
(Fig. 3.25). This pattern of mild villous edema to recognize the changes of the blighted ovum in
and no evidence of fetal development indicates order to separate an abortion with hydropic
that the embryo either never developed or ceased changes from a hydatidiform mole (see Chap. 4).
Fig. 3.21 IT cells in myometrium. Curetted placental IT. These cells often become multinucleated when they
site from an abortion includes a fragment of myometrial invade the myometrium
smooth muscle infiltrated by large implantation site
Hydropic change also may be focally present in mine whether a chromosomal abnormality is
therapeutic abortions [59, 64], especially if the present [60, 61, 65].
tissue blocks sample villi of the chorion laeve, In incomplete abortions, the amount of vil-
where the villi normally degenerate, and mild lous tissue may be greatly reduced or even
hydropic change often is difficult to distinguish absent if portions of the placenta spontaneously
from the loose, myxoid stroma of the normal pass before curettage. The implantation site,
early placenta. however, with its characteristic features, usu-
Other morphologic changes in chorionic villi ally is present. In missed abortions, the villi
from first-trimester abortion specimens may be often are necrotic or hyalinized, and the decidua
found [4]. Irregular outlines of villi yielding a is necrotic (Fig. 3.27). Another change in villi
scalloped appearance and trophoblastic invagina- associated with death of the embryo is loss of
tion into the villous stroma forming pseudoinclu- villous vascularity and fibrosis of the villous
sions often are associated with abnormal stroma (Fig. 3.28) [62, 66, 67]. This change
karyotypes of the conceptus, particularly trip- occurs more frequently in missed abortions.
loidy, but the findings are not sufficiently specific Some or many villi may be necrotic. These vil-
by themselves to be diagnostic of a chromosomal lous changes have little if any clinical
abnormality. Karyotyping is necessary to deter- significance.
Fig. 3.22 IT cells. Prominent infiltration of blood vessel in the decidua by implantation site IT in an abortion specimen.
The cells infiltrate and replace the wall but preserve the lumen
fibrotic and hypovascular, with either no residual

Chorionic Villi and Villous erythrocytes or a few residual degenerating red
Trophoblast After the First Trimester blood cells present.
Following delivery of the term or near-term
After the first trimester or early second trimester, placenta, abnormal bleeding may require curet-
the placenta is sufficiently large that it is delivered tage. The histologic findings in the postpartum
spontaneously or by induction, and curettage endometrium vary. Failure of the implantation
specimens are rare. The villi are more numerous site to resolve quickly is called subinvolution.
and become more complex. By this time there is a With subinvolution the uterus contains remnants
mixture of larger stem villi with prominent ves- of necrotic decidua and the placental site with
sels and central stromal fibrosis and many smaller eosinophilic fibrinoid, trophoblastic cells, and
tertiary villi containing numerous capillaries [4]. enlarged vessels that often are infiltrated by the
By late in the second trimester, the inner CT cells intermediate trophoblast. Retained placental site
are indistinct over most villi, and the trophoblastic also serves as a nidus for inflammation, yielding
proliferation along the anchoring villi has largely inflamed and necrotic endometrium and implan-
ceased. Some abortion specimens present early in tation site (Fig. 3.29). Immunohistochemical
the second trimester following intrauterine fetal stains for keratin help to demonstrate the residual
death. In these cases the villous stroma often is intermediate trophoblastic cells.
Fig. 3.23 Fibrinoid (Rohr’s stria). A strand of densely eosinophilic fibrinoid material of the placental implantation site
overlies the decidua in curettings from an abortion specimen
Placenta Accreta
Placental Polyps
Placenta accreta is a form of abnormal implanta-
Placental polyps are a form of retained product
tion in which the placenta implants directly onto
of conception that represent polypoid portions of
the myometrium with no intervening decidua; it
chorionic villi from an incomplete abortion or a
may grow into or through the myometrium (pla-
term gestation retained in the uterine cavity [4].
centa increta and percreta, respectively) [4, 62].
The villi may be necrotic, hyalinized, or partially
Usually placenta accreta presents immediately
calcified (Fig. 3.30). These polyps are not
following delivery of a term pregnancy when the
tumors, but they do form a nidus for inflamma-
placenta or portions of the placenta cannot be
tion and bleeding. Often these pedunculated
delivered. The placental tissue fails to detach
masses of villi are found within days to weeks
from the implantation site in the myometrium
following abortion or delivery of a term pla-
and cannot be manually removed. The usual
centa. Rarely, they persist for months or years
management of extensive placenta accreta is
after pregnancy.
planned cesarean hysterectomy, but on occasion
a b
Fig. 3.24 Immature chorionic villi. (a) Trophoblast ema- stroma containing a few capillaries with nucleated eryth-
nates from one pole of two villi, which represent the rocytes. The CT cell layer with a single nucleus and pale
implanting portion of the anchoring villi. (b) Immature cytoplasm with overlaying multinucleated ST cells
villus from an abortion specimen show loose, edematous
Table 3.4 Events in first-trimester placental developmenta are contiguous with myometrium. Scattered IT
Time after cells without villi in the myometrium are a phys-
Development of placenta fertilization iologic phenomenon and not placenta accreta
Blastocyst implantation 6–7 days implants.
Villus formation beginsb 12 days
Nucleated RBCs from yolk sac 4.5 weeks
appear in villi
Nonnucleated RBCs from the liver 5–6 weeks
appear in villi
Endometrium Associated
Proportion of nucleated RBCs 4.5–9 weeks with Ectopic Pregnancy
decreases from 100% to 10% in
villi In most circumstances, the endometrium asso-
Decrease in prominence of inner 16–18 weeks ciated with an ectopic pregnancy shows the
cytotrophoblast layer
typical features of early gestation, yet tropho-
RBCs red blood cells
blastic
tissue is not present. A decidualized
a
From time of ovulation. For menstrual age add 2 weeks
b
After 12–15 days, placental development proceeds even stroma, hypersecretory to inactive glands, and
if embryo dies thick-walled spiral arterioles usually are pres-
ent [15, 16, 21, 68, 69]. Often the Arias-Stella
reaction is present, at least focally. The endo-
focal placenta accreta is encountered in curet- metrium associated with ectopic gestation can
tage for postpartum hemorrhage. The diagnostic be highly variable, however, depending on the
feature is villi in direct apposition to myome- status of the trophoblastic tissue. If ectopic tro-
trium without intervening decidua (Fig. 3.31). phoblast is actively proliferating, the endome-
Hyaline fibrinoid material with scattered IT cells trium continues to show the changes of
is interposed between the villi and myometrium, pregnancy with prominent stromal decidual
but decidual cells are absent. Some placental change. If the trophoblast begins to regress, the
polyps may represent focal placenta accreta, endometrium can display a variety of patterns
although the latter are diagnosed only when villi ranging from proliferative to secretory changes.
Endometrium Associated with Ectopic Pregnancy 65
Fig. 3.25 Immature chorionic villi with hydropic change. microscopic and not associated with hyperplasia of tro-
Abortion specimen shows avascular, edematous chorionic phoblast. Furthermore, there is a polar distribution of the
villi. This change is associated with the so-called blighted trophoblast toward the placental implantation site at the
ovum and usually indicates very early demise of the top of the field. These features distinguish this hydropic
embryo. Although the villi are edematous, the change is abortus from a partial hydatidiform mole
The endometrium can show features seen in to identify IT cells scattered in partially necrotic
dysfunctional bleeding, including anovulatory decidua, often around spiral arterioles. In the
bleeding patterns, irregular secretory patterns, absence of chorionic villi, the infiltrate of inter-
or progestin effects. Subtle clues, such as a mediate trophoblast at the placental implantation
focal Arias-Stella reaction or a small aggregate site can be very difficult to distinguish from
of gland cells with clear cytoplasm, can suggest decidua. Identification of IT cells of the placental
an ectopic pregnancy. site is crucial, and in such cases, immunostaining
Establishing the presence of an intrauterine for keratin can help show scattered trophoblast,
pregnancy effectively rules out an ectopic preg- usually IT cells in the decidua. A broad-spectrum
nancy. Evidence of an intrauterine pregnancy keratin stain will often assist in demonstrating IT
includes chorionic villi, trophoblastic cells, or the in difficult cases. A panel using several immunos-
placental implantation site, while the absence of tains can be helpful for identifying trophoblastic
these elements raises the possibility of an ectopic tissue if the keratin stain is inconclusive.
in the appropriate clinical setting. Occasionally Immunostains for hPL, hCG, inhibin-α, HLA-G,
individual ST giant cells may be detected HSD3B1, and Mel-CAM specifically identify tro-
enmeshed in blood or fibrin. In cases where there phoblast, but these tend to be less sensitive than
are no villi or ST cells, an attempt should be made staining with keratin antibodies. Nonetheless,
Fig. 3.26 Immature chorionic villi with hydropic change. This hydropic abortus shows only rare cisterns, no tropho-
blastic hyperplasia, and extensive villous edema, the combination of which is not diagnostic of hydatidiform mole
these immunostains can be helpful in ruling out Clinical Queries and Reporting
an ectopic pregnancy in questionable cases by
demonstrating the subtle presence of intrauterine For most endometrial biopsies or curettings
trophoblast [32, 34, 52, 54, 70–74]. Clearly, ade- related to pregnancy, three clinical questions
quate sampling of endometrial tissue is important need to be answered by pathologic examination:
to ensure recognition of chorionic tissue. We have (1) Does the endometrium show features of preg-
empirically found that three cassettes from abor- nancy? (2) If the changes indicate pregnancy, are
tion specimens are usually sufficient to establish chorionic villi and/or trophoblast present? (3) If
the presence of chorionic tissue. If no trophoblast villi and/or trophoblast are present, do they
or villi are present in the first three tissue blocks, appear normal? For example, endometrial
all of the residual tissue should be processed [34, changes of pregnancy without the presence of
52, 53, 71, 72]. chorionic villi or trophoblast suggest the possi-
Fig. 3.27 Missed abortion. The villi are extensively necrotic with interspersed hyalinization. Rare fibrotic villi are seen
at the top of the image
bility of an ectopic pregnancy. In spontaneous endometrial patterns but there is no evidence of

abortions, the villi may reflect pathologic devel- chorionic villi, trophoblast, or fetal tissue, then an
opment of the embryo, a feature that helps ectopic pregnancy must be considered.
explain the occurrence of the abortion. In other Occasionally the entire placenta with implanta-
cases, edematous villi or proliferative trophoblast tion site is expelled before the curettage, so lack
can raise the question of a hydatidiform mole or of identifiable products of conception does not
gestational trophoblastic neoplasms such as cho- unequivocally indicate ectopic pregnancy.
riocarcinoma or placental site trophoblastic Nonetheless, an ectopic pregnancy can result in
tumor (see Chap. 4). sudden, life-threatening intra-abdominal hemor-
The pathology report should consider the clini- rhage, so immediate notification of the clinician
cal questions asked regarding pregnancy as well managing the patient is imperative. Also, all the
as the pathologic findings. Most cases require residual tissue should be processed. The clinician
only documentation of the presence of placental should be informed if immunohistochemical
or fetal tissue. The most urgent question is that of stains to identify trophoblast are pending or if
an ectopic pregnancy. If pregnancy is suspected, residual tissue is being processed. Another call
or if the morphology shows pregnancy-induced should follow as soon as the results are available.
Fig. 3.28 Immature chorionic villi with villous fibrosis. Villous trophoblast remains confined to one pole of the
Abortion specimen showing immature villi with the villi. Scattered necrotic villi are also noted (right of the
absence of vessels and stromal fibrosis (left of image). image)
Occasional spontaneous abortion specimens cally relevant. In spontaneous abortions, for

show only a small amount of early placental site example, mild villous edema (hydropic change)
with intermediate trophoblast and no chorionic and absence of fetal tissues including erythrocytes
villi. This finding is sufficient to establish the diag- in villous vessels indicates that the gestation was
nosis of an intrauterine pregnancy. The specimen abnormal and may help the clinician in the coun-
represents products of conception, and a descrip- seling of a patient. Although the microscopic
tive diagnosis such as “implantation site and findings can help indicate early death of the
decidua” or “intermediate trophoblast” serves to embryo, cytogenetic analysis of tissue has more
verify the presence of an intrauterine gestation. value for assessing significant abnormalities that
When villi are present in first-trimester speci- may lead to recurrent abortion. Sometimes the term
mens, some morphologic findings may be clini- “hydropic villi” raises the specter of hydatidiform
Fig. 3.29 Retained placental tissue. Inactive endometrial tissue with associated inflammatory infiltrate is seen juxta-
posed to retained placental products, including a fibrotic villus (left side of image)
mole, so this term should be used cautiously unless men containing more than a small amount of
the clinician fully understands the significance. If trophoblast without villi, or an exaggerated pla-
hydropic change is diagnosed, it may be useful to cental site without villi or unusually hydropic
add a comment to indicate that this does not repre- villi that are not clearly molar (i.e., abnormal
sent a mole. villous morphology), should be reported (see
Because pregnancy can be complicated by Chap. 4). In such cases, a comment regarding
gestational trophoblastic neoplasms, the status the uncertainty of the finding and recommenda-
of the trophoblast, especially any abnormal pro- tion for follow-up with serum β-hCG titers helps
liferative activity, deserves comment. A speci- in the management of the patient.
Fig. 3.30 Placental polyp. Polypoid fragment of retained placental tissue removed by curettage several weeks after
term delivery. The mature villi are degenerate and hyalinized
References 71
Fig. 3.31 Placenta accreta. Curetting for bleeding fol- fibrinoid material but no intervening decidua are all that
lowing delivery of a term gestation shows mature chori- separate the villi from the myometrial smooth muscle at
onic villi at the right of the field. Implantation site IT and the left
7. Mazur MT, Duncan DA, Younger JB. Endometrial

References biopsy in the cycle of conception. Histologic
and lectin histochemical evaluation. Fertil Steril.
1. Mishell DRJ. Spontaneous and recurrent abortion. 1989;51:764–7.
In: Stenchever MA, Droegemueller W, Herbst AL, 8. Foss BA, Horne HW, Hertig AT. The endometrium
Mishell Jr DR, editors. Comprehensive gynecology. and sterility. Fertil Steril. 1958;9:193–206.
2nd ed. St. Louis: Mosby; 2001. p. 413–41. 9. Karow WG, Gentry WC, Skeels RF, Payne
2. Kalousek DK. Pathology of abortion: chromosomal SA. Endometrial biopsy in the luteal phase of the
and genetic correlations. In: Kraus FT, Damjanov I, cycle of conception. Fertil Steril. 1971;22:482–95.
Kaufman N, editors. Pathology of reproductive fail- 10. Hertig AT. Gestational hyperplasia of endometrium.
ure. Baltimore: Williams & Wilkins; 1991. p. 228–56. A morphologic correlation of ova, endometrium, and
3. Miller JF, Williamson E, Glue J, Gordon YB, corpora lutea during early pregnancy. Lab Investig.
Grudzinskas JG, Sykes A. Fetal loss after implanta- 1964;13:1153–91.
tion. A prospective study. Lancet. 1980;2:554–6. 11. Wentz AC, Herbert CMI, Maxson WS, Hill GA,
4. Benirschke K, Kaufmann P. Pathology of the human Pittaway DE. Cycle of conception endometrial biopsy.
placenta. 4th ed. New York: Springer; 2000. Fertil Steril. 1986;46:196–9.
5. Wilcox AJ, Weinberg CR, O’Conner JF, Baird DD, 12. Lichtig C, Korat A, Deutch M, Brandes JM. Decidual
Schlatterer JP, Camfield RE, et al. Incidence of early vascular changes in early pregnancy as a marker
loss of pregnancy. N Engl J Med. 1988;319:189–94. for intrauterine pregnancy. Am J Clin Pathol.
6. Simpson JL. Incidence and timing of pregnancy 1988;90:284–8.
losses. Am J Med Genet. 1990;35:165–73. 13. Clement PB, Young RH, Scully RE. Nontrophoblastic
pathology of the female genital tract and peritoneum
associated with pregnancy. Semin Diagn Pathol. 31. Shih I-M, Mazur MT, Kurman RJ. Gestational tro-
1989;6:372–406. phoblastic tumors and related tumor-like lesions.
14. Holmes EJ, Lyle WH. How early in pregnancy In: Kurman RJ, editor. Blaustein’s pathology of the
does the Arias–Stella reaction occur? Arch Pathol. female genital tract. 6th ed. New York: Springer;
1973;95:302–3. 2011. p. 1075–135.
15. Arias-Stella J. Atypical endometrial changes pro- 32. Kurman RJ, Main CS, Chen HC. Intermediate tropho-
duced by chorionic tissue. Hum Pathol. 1972;3:450–3. blast: a distinctive form of trophoblast with specific
16. Arias-Stella J. Atypical endometrial changes associ- morphological, biochemical and functional features.
ated with the presence of chorionic tissue. AMA Arch Placenta. 1984;5:349–70.
Pathol. 1954;58:112–28. 33. Kurman RJ, Young RH, Norris HJ, Main CS, Lawrence
17. Arias-Stella J. The Arias-Stella reaction: facts WD, Scully RE. Immunocytochemical localization of
and fancies four decades after. Adv Anat Pathol. placental lactogen and chorionic gonadotropin in the
2002;9:12–23. normal placenta and trophoblastic tumors, with empha-
18. Dardi LE, Ariano L, Ariano MC, Gould VE. Arias- sis on intermediate trophoblast and the placental site tro-
Stella reaction with prominent nuclear pseudoin- phoblastic tumor. Int J Gynecol Pathol. 1984;3:101–21.
clusions simulating herpetic endometritis. Diagn 34. Yeh IT, O’Connor DM, Kurman RJ. Intermediate tro-
Gynecol Obstet. 1982;4:127–32. phoblast: further immunocytochemical characteriza-
19. Beswick IP, Gregory MM. The –Arias-Stella phe- tion. Mod Pathol. 1990;3:282–7.
nomenon and the diagnosis of pregnancy. J Obstet 35. Yeh I-T, Kurman RJ. Functional and morphologic
Gynaecol. 1971;78:143–8. expressions of trophoblast. Lab Investig. 1989;61:1–4.
20. Bernhardt RN, Bruns PD, Drose VE. Atypical 36. Wells M, Bulmer JN. The human placental bed:
endometrium associated with ectopic pregnancy. histology, immunohistochemistry and pathology.
The Arias-Stella phenomenon. Obstet Gynecol. Histopathology. 1988;13:483–98.
1966;23:849–53. 37. Shih IM, Seidman JD, Kurman RJ. Placental site nod-
21. LaIuppa MA, Cavanagh D. The endometrium in ecto- ule and characterization of distinctive types of inter-
pic pregnancy. A study based on 35 patients treated by mediate trophoblast. Hum Pathol. 1999;30:687–94.
hysterectomy. Obstet Gynecol. 1963;21:155–64. 38. Mao T-L, Kurman RJ, Huang C-C, Lin M-C, Shih
22. Roach WR, Guderian AM, Brewer JI. Endometrial I-M. Immunohistochemistry of choriocarcinoma: an
gland cell atypism in the presence of trophoblast. Am aid in differential diagnosis and in elucidating patho-
J Obstet Gynecol. 1960;79:680–91. genesis. Am J Surg Pathol. 2007;31:1726–32.
23. Huettner PC, Gersell DJ. Arias-Stella reaction in non- 39. Kalhor N, Ramirez PT, Deavers MT, Malpica A, Silva
pregnant women: a clinicopathologic study of nine EG. Immunohistochemical studies of trophoblastic
cases. Int J Gynecol Pathol. 1994;13:241–7. tumors. Am J Surg Pathol. 2009;33:633–8.
24. Yokoyama S, Kashima K, Inoue S, Daa T, Nakayama 40. Singer G, Kurman RJ, McMaster MT, Shih
I, Moriuchi A. Biotin-containing intranuclear inclu- I-M. HLA-G immunoreactivity is specific for inter-
sions in endometrial glands during gestation and puer- mediate trophoblast in gestational trophoblastic
perium. Am J Clin Pathol. 1993;99:13–7. disease and can serve as a useful marker in differen-
25. Mazur MT, Hendrickson MR, Kempson RL. Optically tial diagnosis. Am J Surg Pathol. 2002;26:914–20.
clear nuclei. An alteration of endometrial epithelium 41. Mao TL, Kurman RJ, Jeng YM, Huang W, Shih
in the presence of trophoblast. Am J Surg Pathol. IM. HSD3B1 as a novel trophoblast-associated
1983;7:415–23. marker that assists in the differential diagnosis of tro-
26. Genest DR, Brodsky G, Lage JA. Localized endome- phoblastic tumors and tumorlike lesions. Am J Surg
trial proliferations associated with pregnancy: clinical Pathol. 2008;32(2):236–42.
and histopathologic features of 11 cases. Hum Pathol. 42. Chou YY, Jeng YM, Mao TL. HSD3B1 is a spe-
1995;26(11):1233–40. cific trophoblast-associated marker not expressed in
27. Schammel DP, Mittal KR, Kaplan K, Deligdisch L, a wide spectrum of tumors. Int J Gynecol Cancer.
Tavassoli FA. Endometrial adenocarcinoma associ- 2013;23(2):343–7.
ated with intrauterine pregnancy. A report of five 43. Mirkovic J, Elias K, Drapkin R, Barletta JA, Quade
cases and a review of the literature. Int J Gynecol B, Hirsch MS. GATA3 expression in gestational
Pathol. 1998;17:327–35. trophoblastic tissues and tumours. Histopathology.
28. Kovacs AG, Cserni G. Endometrial adenocarci- 2015;67(5):636–44.
noma in early pregnancy. Gynecol Obstet Investig. 44. Bamberger AM, Aupers S, Milde-Langosch K, Löning
1996;41:70–2. T. Expression pattern of the cell cycle promoter cyclin
29. Nakamura Y, Moritsuka Y, Ohta Y, Itoh S, Haratake A, E in benign extravillous trophoblast and gestational
Kage M, et al. S-100 protein in glands within decidua trophoblastic lesions: correlation with expression of
and cervical glands during early pregnancy. Hum Ki-67. Int J Gynecol Pathol. 2003;22(2):156–61.
Pathol. 1989;20:1204–9. 45. Shih IM. Trophogram, an immunohistochemistry-
30. Agarwal S, Singh UR. Immunoreactivity with S100 based algorithmic approach, in the differential diag-
protein as an indicator of pregnancy. Indian J Med nosis of trophoblastic tumors and tumorlike lesions.
Res. 1992;96:24–6. Ann Diagn Pathol. 2007;11(3):228–34.
References 73
46. Shih IM, Kurman RJ. Immunohistochemical local- mality in early abortion specimens. Histopathology.
ization of inhibin-alpha in the placenta and gesta- 1993;22:25–9.
tional trophoblastic lesions. Int J Gynecol Pathol. 61. Novak R, Agamanolis D, Dasu S, Igel H, Platt M,
1999;18:144–50. Robinson H, et al. Histologic analysis of placen-
47. Shih IM, Kurman RJ. Expression of melanoma cell tal tissue in first trimester abortions. Pediatr Pathol.
adhesion molecule in intermediate trophoblast. Lab 1988;8:477–82.
Investig. 1996;75:377–88. 62. Gersell DJ, Kraus FT. Diseases of the placenta. In:
48. Silverberg SG, Kurman RJ. Tumors of the uterine Ellenson LH, Ronnett BM, Kurman RJ, editors.
corpus and gestational trophoblastic disease, Atlas of Blaustein’s pathology of the female genital tract. 6th
tumor pathology, 3rd series, fascicle 3. Washington, ed. New York: Springer; 2011. p. 999–1073.
DC: Armed Forces Institute of Pathology; 1992. 63. Genest DR. Estimating the time of death in still-
49. Kurman RJ. The morphology, biology, and pathology born fetuses: II. Histologic evaluation of the pla-
of intermediate trophoblast—a look back to the pres- centa; a study of 71 stillborns. Obstet Gynecol.
ent. Hum Pathol. 1991;22:847–55. 1992;80:585–92.
50. Young RE, Kurman RJ, Scully RE. Proliferations and 64. Ladefoged C. Hydrop degeneration: a histopathologi-
tumors of intermediate trophoblast of the placental cal investigation of 260 early abortions. Acta Obstet
site. Semin Diagn Pathol. 1988;5:223–37. Gynecol Scand. 1980;59:509–12.
51. Wan SK, Lam PWY, Pau MY, Chan JKC. Multiclefted 65. Rehder H, Coerdt W, Eggers R, Klink F, Schwinger
nuclei. A helpful feature for identification of interme- E. Is there a correlation between morphological and
diate trophoblastic cells in uterine curetting speci- cytogenetic findings in placental tissue from early
mens. Am J Surg Pathol. 1992;16:1226–32. missed abortions? Hum Genet. 1989;82:377–85.
52. O’Connor DM, Kurman RJ. Intermediate trophoblast 66. Fox H. Morphological changes in the human pla-
in uterine curettings in the diagnosis of ectopic preg- centa following fetal death. J Obstet Gynaecol Br
nancy. Obstet Gynecol. 1988;72:665–70. Commonw. 1968;75:839–43.
53. Daya D, Sabet L. The use of cytokeratin as a sensitive 67. Hustin J, Gaspard U. Comparison of histologi-
and reliable marker for trophoblastic tissue. Am J Clin cal changes seen in placental tissue cultures and in
Pathol. 1991;95:137–41. placentae obtained after fetal death. Br J Obstet
54. Angel E, Davis JR, Nagle RB. Immunohistochemical Gynaecol. 1977;84:210–5.
demonstration of placental hormones in the diagnosis 68. Buckley CH, Fox H. Biopsy pathology of the endo-
of uterine versus ectopic pregnancy. Am J Clin Pathol. metrium. 2nd ed. London: Arnold; 2002.
1985;84:705–9. 69. Romney SL, Hertig AT, Reid DE. The endometria
55. Shih IM, Kurman RJ. Ki-67 labeling index in the dif- associated with ectopic pregnancy. A study of 115
ferential diagnosis of exaggerated placental site, pla- cases. Surg Gynecol Obstet. 1950;91:605–11.
cental site trophoblastic tumor, and choriocarcinoma: 70. Shih I-M, Mazur M, Kurman R. Gestational tro-
a double immunohistochemical staining technique phoblastic tumors and related tumor-like lesions.
using Ki-67 and Mel-CAM antibodies. Hum Pathol. Blaustein’s pathology of the female genital tract. 6th
1998;29:27–33. ed. New York: Springer; 2011. p. 1075–135.
56. Szulman AE. Examination of the early conceptus. 71. Kaspar HG, To T, Dinh TV. Clinical use of immuno-
Arch Pathol Lab Med. 1991;115:696–700. peroxidase markers in excluding ectopic gestation.
57. Rushton DI. Examination of products of conception Obstet Gynecol. 1991;78:433–7.
from previable human pregnancies. J Clin Pathol. 72. Sørensen FB, Marcussen N, Daugaard HO, Kristiansen
1981;34:819–35. JD, Møller J, Ingerslev HJ. Immunohistological dem-
58. Redline RW, Zaragoza M, Hassold T. Prevalence of onstration of intermediate trophoblast in the diagnosis
developmental and inflammatory lesions in nonmolar of uterine versus ectopic pregnancy: a retrospective
first-trimester spontaneous abortions. Hum Pathol. survey and results of a prospective trial. Br J Obstet
1999;30(1):93–100. Gynaecol. 1991;98:463–9.
59. Jurkovic I, Muzelak R. Frequency of pathologic 73. Shih I, Wang T, Wu T, Kurman RJ, Gearhart
changes in the young human chorion in therapeu- JD. Expression of Mel-CAM in implantation site
tic abortions of normal pregnancies: a report of 500 intermediate trophoblastic cell line, IST-1, limits its
cases studied histologically. Am J Obstet Gynecol. migration on uterine smooth muscle cells. J Cell Sci.
1970;108:382–6. 1998;111(Pt 17):2655–64.
60. Vanlijnschoten G, Arends JW, Delafuente AA, 74. Hiby SE, King A, Sharkey A, Loke YW. Molecular
Schouten HJA, Geraedts JPM. Intra-observer and studies of trophoblast HLA-G: polymorphism, iso-
inter-observer variation in the interpretation of his- forms, imprinting and expression in preimplantation
tological features suggesting chromosomal abnor- embryo. Tissue Antigens. 1999;53:1–13.
Gestational Trophoblastic Disease
4
Contents
Hydatidiform Mole�� 76
Gestational Trophoblastic Neoplasms�� 94
Nonneoplastic Lesions �� 110
Clinical Queries and Reporting of Trophoblastic Neoplasms�� 112
References �� 113
The World Health Organization (WHO) classifi- important, as they have different clinical presen-
cation of gestational trophoblastic disease (GTD) tations and behavior.
includes disorders of placental development With improved early detection and diagnosis
(hydatidiform mole) and neoplasms of the tro- of molar pregnancies, most are evacuated at
phoblast (choriocarcinoma, placental site tropho- earlier gestational stages than in the past.
blastic tumor [PSTT], and epithelioid Therefore, the morphologic features are less
trophoblastic tumor [ETT]) (Table 4.1) [1]. well developed, and this makes it challenging
Nonneoplastic lesions include the exaggerated for pathologists to diagnose. Hydatidiform
placental site and placental site nodules and mole, either partial or complete, is the most
plaques. A common feature of all trophoblastic common form of GTD, and this also is the tro-
lesions is that they produce human chorionic phoblastic lesion most commonly encountered
gonadotropin (hCG), which serves as a clinical in endometrial curettings [2]. Gestational tro-
marker for the presence of persistent or progres- phoblastic neoplasms (GTN) including chorio-
sive trophoblastic disease. Because these lesions, carcinoma, PSTT, and ETT are infrequent.
especially postmolar trophoblastic disease, are Recognition of any of these lesions can be dif-
often treated in the absence of a histologic diag- ficult, however, because the morphologic fea-
nosis, they may be clinically classified as GTD tures of all forms of GTD overlap with the
without designation of the morphologic subtype. features of trophoblastic growth encountered in
Nonetheless, identification and separation of the early pregnancy, abortion, and persistent pla-
different pathologic forms of the disease are cental implantation sites.

76 4 Gestational Trophoblastic Disease
Table 4.1 Modified World Health Organization classifi- Native American women in New Mexico were
cation of gestational trophoblastic disease
found to have an increased risk for complete
Molar lesions and partial hydatidiform mole, invasive mole,
Complete hydatidiform mole and choriocarcinoma over other ethnic groups
Partial Hydatidiform mole
[16, 17].
Invasive mole
The separation of hydatidiform mole into
Gestational trophoblastic neoplasms
Choriocarcinoma two subtypes, complete and partial, represents a
Placental site trophoblastic tumor significant advance in our understanding of
Epithelioid trophoblastic tumor molar pregnancy as these two forms of hyda-
Nonneoplastic lesions tidiform mole have different cytogenetic pat-
Exaggerated placental site terns that are accompanied by different
Placental site nodule and plaque clinicopathologic profiles and different degrees
Abnormal (non-molar) villous lesions of risk for the development of GTN (Table 4.2)
Modified from Kurman et al. [1], with permission [2, 9, 20–24]. Maternal age does not have an
effect on the incidence of partial hydatidiform
mole (PHM) but does for complete hydatidi-
Hydatidiform Mole form mole (CHM). Specifically, adolescents
(<20 years) are 7.0 times more likely, and
General Features advanced maternal age women (>40 years) are
twice as likely to develop CHM when com-
Hydatidiform mole, either partial or complete, pared to average age women (20–39 years)
is infrequent in the United States and Europe, [25]. Both forms of hydatidiform mole typi-
occurring in about one in 600 to one in 2000 cally present in the first trimester, often as an
pregnancies [3–8], although some studies have abortion.
suggested that partial hydatidiform mole may Familial recurrent hydatidiform mole (FRHM)
be even more frequent, occurring in up to 1 in represents <2% of all hydatidiform moles [26,
700 pregnancies [9]. The incidence of GTD is 27]. Gene mutations of NLRP7 or KHDC3L lead
usually reported in relation to the total num- to a silencing of maternal imprinted genes and
ber of pregnancies in a study cohort rather expression of paternal counterparts [28, 29].
than the total population [10]. In other parts of Although the conception is digynic triploid [30],
the world, including Asia and Latin America, the phenotype is that of a CHM.
these disorders are more common, although
problems in methodology often complicate
studies of their frequency when deliveries Complete Hydatidiform Mole
take place at home [2, 10, 11]. In the United
States racial data is inconsistent [12]; glob- Complete hydatidiform mole (CHM) has been
ally, however, Asians, Hispanics, and Native recognized and studied for many years [5]. In the
Americans continue to have an increased risk past, typical CHM presented at approximately
of developing gestational trophoblastic dis- the 16th week of pregnancy, but many hydatidi-
ease [13–17]. In one study, Asian women were form moles are now detected earlier in gestation
more likely than white women to develop com- at approximately 9 weeks [31–34]. The earlier
plete hydatidiform mole (CHM) but less likely detection of CHM has been attributed to wide-
to develop partial hydatidiform mole (PHM) spread use and increased sensitivity of pelvic
[18]. Hispanic women had an increased risk ultrasound during pregnancy [35] in combination
of PHM when compared to white and African with serum hCG measurement. As a result, the
American women [15]. Alaska Natives had a classic presenting symptoms of vaginal bleeding,
3–4 times higher incidence than white women uterine enlargement greater than that expected
for developing hydatidiform mole [19], and for the gestational age, anemia, hyperemesis,
Hydatidiform Mole 77
Table 4.2 Clinical and cytogenetic comparison of hydatidiform moles and hydropic abortus
Complete hydatidiform
mole Partial hydatidiform mole Hydropic abortus
Presentation Spontaneous abortion Missed abortion Spontaneous or missed
abortion
Gestational 9 weeks 11–12 weeks 6–14 weeks
age
β-hCG titer Typically elevated Low to normal Low to normal
Uterine size Often enlarged for date Often small for date Often small for date
Amount of Variable, may be Variable, usually decreased Usually decreased
tissue increased
Karyotype 46XX (all paternal) 69XXY or XXX (2: 1, paternal: Variable, often abnormal
maternal)
Persistent 15–20% 0.5–5% None
GTD
β-hCG human chorionic gonadotropin
hyperthyroidism, theca lutein ovarian cysts, and mole. In fact, the classic features of CHM are now
preeclampsia are less frequently encountered relatively uncommon, as noted above. Whether a
[34, 36–38]. Abortion with abnormal bleeding CHM is detected very early in gestation or at a
and passage of molar tissue remains a frequent more advanced stage, both diffuse villous edema
presentation. Although historically many CHMs and irregular hyperplasia of trophoblast from the
were diagnosed before curettage due to advanced surface of some villi should be seen. The classic
gestational age at clinical presentation, in current features of CHM are considered first, followed by
practice CHM may be undetected clinically and a description of the early CHM, defined as a CHM
diagnosed only when the pathology is reviewed. detected prior to 12 weeks of gestation. Clearly
The serum β-human chorionic gonadotropin there are transitions between these patterns.
(β-hCG) titers may be markedly elevated, but lev- Classic features of a complete hydatidiform
els are often unreliable for establishing the diag- mole obtained at 16–18 weeks of gestation include
nosis. However, a β-hCG titer above 82,350 mIU/ a voluminous gross specimen showing many
ml, coupled with absence of fetal heart move- grapelike translucent villi [2]. The grapelike villi
ment, is correlated with the presence of hydatidi- typically range from several millimeters to 2.0 cm
form mole [39]. In a cohort of women diagnosed or more in the largest dimension (Table 4.3).
with CHM that developed GTN, the median Microscopically, these advanced CHMs
serum β-hCG levels prior to evacuation were show marked villous edema with cistern forma-
179,944 mIU/mL and 654,280 mIU/mL for tion (Fig. 4.1) coupled with hyperplasia of tro-
patients diagnosed prior to and after 10 weeks phoblast that includes circumferential growth
gestation, respectively [34]. A pre-evacuation over the villous surfaces (Fig. 4.2). The edema
serum β-hCG less than 150,000 mIU/mL was affects all villi, although the degree of enlarge-
associated with a lower risk of developing GTN ment caused by the edema is variable. Sometimes
irrespective of gestational age at diagnosis, and one of the two features, circumferential hyper-
early diagnosis was not associated with a reduc- plasia or edema, predominates, but both features
tion in risk of developing a GTN [34]. should be present. A cistern is a completely
The morphologic features of CHM differ to acellular central cavity within a villus that is
some extent depending on the gestational age. filled with edema fluid and surrounded by a
Consequently, the diagnosis of CHM requires rec- sharply demarcated stromal border. In complete
ognition of both classic changes and the subtler hydatidiform mole, many, but not all, of the villi
changes of very early complete hydatidiform show cistern formation, although all the villi are
Table 4.3 Comparison of pathologic features of complete hydatidiform mole, partial hydatidiform mole, and hydropic
abortus
Complete hydatidiform mole Partial hydatidiform mole Hydropic abortus
Villous hydrops Generalized, often grossly Partial, often grossly visible Microscopic, limited
(swelling) visible
Villous shape Round to bulbous Irregular, scalloped Round, small
Cisterns Present Present Usually absent
Trophoblastic Rare Common Rare
inclusions
Fetal tissue Rare Common Rare
Trophoblast Circumferential, multifocal Circumferential, multifocal Polar at anchoring villi
distribution only
Proliferation Variable, may be marked Focal, minimal, with ST Limited to anchoring
sprouts villi
Atypia Often present Rare Absent
Implantation site Exaggerated Normal or exaggerated Normal
ST syncytiotrophoblast
Fig. 4.1 Complete hydatidiform mole. Low- tended villus to the right of center has a central cistern.
magnification view shows generalized villous edema and Irregular, haphazard hyperplasia of trophoblast is present
marked enlargement of many villi. The massively dis- along the surface of several villi
edematous (Fig. 4.3). Scattered small villi with- presence of this type of hydatidiform mole. This
out cisterns often are admixed (Fig. 4.1). Also, stroma has a blue-gray tint, or “myxoid hue,”
in CHM the villous stroma has a subtle but dis- that is generally distinctly different from the
tinctive appearance that helps to alert to the stroma of partial hydatidiform mole and
Fig. 4.2 Complete hydatidiform mole. Marked circumferential hyperplasia of trophoblast is seen along the surface of
most of the edematous villi in this field. Cisterns are ill defined, but all the villi are edematous
hydropic abortions that lack this feature polar orientation at the anchoring villi and is not pres-
(Fig. 4.3). ent on other villi. The degree of trophoblastic hyper-
In addition to these changes, some villi may plasia in a CHM is highly variable. Often the
be necrotic and occasional villi can show par- hyperplasia is moderate to marked with large masses
tial calcification. Because fetal development of trophoblast that may be confluent, extending from
ends very early in placentation, the villi usu- the surface of the villi. Occasional cases, however,
ally do not have visible blood vessels. Although have only minimal trophoblastic hyperplasia.
stromal vessels are indistinct, occasional ves- Early complete hydatidiform mole, detected at
sels in the villous stroma can be found. In fact, a younger gestational age (<12 weeks), demon-
CD34 immunostaining will show numerous strates less advanced degree of villous abnormal-
blood vessels; these may contain cellular ities at both the gross and microscopic levels. The
debris [40]. gross specimen may not show visible or obvi-
Circumferential trophoblastic hyperplasia seen in ously hydropic villi. This is attributable to the
complete hydatidiform mole is characterized by early gestational age [31, 43], villi being passed
masses of trophoblast, often confluent, that project spontaneously before curettage, or villi collaps-
randomly along the surface of edematous villi ing during the suction curettage.
(Figs. 4.1 and 4.2) [2, 5, 24, 41, 42]. The trophoblas- Microscopically, the villous features are sub-
tic hyperplasia can involve smaller villi as well as the tle, and well-formed cisterns may be absent [31,
large villi with cisterns. This haphazard trophoblastic 32, 44–46]. Early CHM shows bulbous,
proliferation contrasts with the normal trophoblastic cauliflower-like terminal villi with hypercellu-
proliferation in an immature placenta that maintains lar villous stroma and karyorrhexis (Figs. 4.4,
Fig. 4.3 Complete hydatidiform mole. Portions of edem- Note the smaller villus with edema on the right, one with
atous, avascular villi in a complete mole show hyperplasia a blue-gray tint to the stroma, or “myxoid hue.” Occasional
of the trophoblast covering in a “lacy pattern” (right). A smaller villi such as these are commonly found in com-
trophoblastic inclusion is present in the left upper corner. plete mole specimens
4.5 and 4.6) [31, 44, 47]. At this stage, villi may be brisk. These findings contrast with those found
also show relatively smooth contours. Even in the trophoblast of an abortus that does not dis-
these small villi show edema, however, with play marked nuclear atypia. In addition, “layer-
sparsely cellular stroma showing widely sepa- ing” of atypical implantation type trophoblast on
rated fibroblasts. In very early CHM, the tro- fibrin is a very characteristic feature of complete
phoblastic hyperplasia tends to be focal but hydatidiform moles. The amount of trophoblast
definite, with proliferation of both cytotropho- and the degree of atypia present in hydatidiform
blastic (CT) and syncytiotrophoblastic (ST) moles have no apparent bearing on the subsequent
cells from the villous surface. If present, the tro- clinical course, so grading the trophoblast is not
phoblast along the chorionic plate also is helpful [42].
hyperplastic. With hydatidiform mole, especially CHM, the
The trophoblast that accompanies CHMs also trophoblastic infiltration of the placental implan-
often shows atypia, with enlarged, pleomorphic, tation site typically is exaggerated, even in early
and hyperchromatic nuclei. Mitotic activity may CHM [31, 48, 49]. Curettage, especially sharp
Fig. 4.4 Early complete hydatidiform mole. The villi are shows circumferential hyperplasia, however, which dis-
bulbous with mildly hypercellular stroma that shows min- tinguishes this early complete mole from a hydropic
imal edema with no well-formed cisterns. The trophoblast abortus
curettage after suction extraction, can yield abun- influence the diagnosis or prognosis of these
dant, atypical trophoblast including many inter- lesions.
mediate trophoblastic cells from the implantation
site (Fig. 4.7). This trophoblastic proliferation is
a standard feature of hydatidiform moles and Partial Hydatidiform Mole
should not be misinterpreted as a coexisting pla-
cental site trophoblastic tumor or choriocarci- Partial hydatidiform moles (PHM) tend to pres-
noma. The Ki-67 proliferation index can be ent slightly later in gestation, occurring between
slightly elevated in an exaggerated implantation 11 and 12 weeks gestation [33]. These hydatidi-
site associated with a CHM but is often zero form moles also may be clinically unsuspected,
(Fig. 4.8). As with the trophoblastic proliferation presenting as a spontaneous or missed abortion
that covers the villi, the amount of trophoblast (Table 4.2) [20, 21]. The uterus often is small for
and the atypia in the implantation site do not gestational age, and serum β-hCG titers are in
Fig. 4.5 Early complete hydatidiform mole. Portion of a strates slight but definite circumferential hyperplasia with
villus from early complete mole shows minimal edema minimal atypia
and karyorrhexis of the stroma. The trophoblast demon-
the low or normal range for that time in preg- Frequently the nonedematous villi are fibrotic,
nancy; preeclampsia is less frequent than in the especially in PHM that are greater than 12 weeks
case of CHM. [43]. Typically the enlarged villi have irregular,
Grossly, PHM may demonstrate edematous scalloped borders with deep infoldings (Figs. 4.9
villi, but to a lesser extent than CHM, and may and 4.10) that contrast to the smooth or bulbous,
contain remnants of a fetus. Morphologically, par- rounded contours of villi in complete hydatidi-
tial hydatidiform mole, as the name implies, the form mole. Transverse sectioning of the invagi-
villous abnormalities affect only a portion of the nations yields trophoblastic “inclusions” in the
placenta, resulting in two populations of villi [2, villous stroma. Trophoblastic hyperplasia usually
20, 23, 24]. The first population is composed of is limited, with only small foci of syncytiotro-
edematous villi with trophoblastic hyperplasia phoblast projecting randomly from the surface of
(Table 4.3) [2, 20, 22, 24, 50], while the second the affected villi (Fig. 4.11). No single or com-
population is composed of small villi that do not bined morphologic features are sufficient for a
show edema (Fig. 4.9). definitive diagnosis of PHM. The presence of any
Fig. 4.6 Early complete hydatidiform mole. Prominent hyperplasia of the trophoblast from the surface of a villus of an
early complete mole
one of the following morphologic features should usually shows only focal and mild atypia com-
prompt further workup (molecular genotyping pared to the implantation site seen in complete
encouraged) to rule out PHM: cistern formation, hydatidiform mole [49].
round or oval pseudoinclusions, two populations
of villi, and villous size ≥2.5 mm [51] (see below
Hydatidiform Mole Versus Hydropic Abortus for Cytogenetics
more information on genotyping).
Another frequent finding in PHM is micro- Cytogenetic studies show that complete hydatidi-
scopic evidence of fetal development, such as form moles have a normal diploid DNA content,
fetal tissue, erythrocytes in villous capillaries, or usually 46XX [53–56]. The entire chromosomal
fetal membranes. Fetal tissue is not invariably complement, however, is paternal, resulting from
present in PHM, however, and in some studies duplication of a haploid paternal genome (23X); the
this feature has been found in fewer than one half CHM lacks maternal DNA [57]. More than 90% of
of cases [52]. The implantation site trophoblast CHMs contain this composition of paternal chromo-
Fig. 4.7 Exaggerated placental implantation site of com- villi, but other fields showed markedly edematous villi of
plete hydatidiform mole. Abundant, trophoblastic cells a complete mole. This exaggerated placental site has no
are present at the implantation site of a complete mole. clinical significance
The trophoblastic cells in this field are not associated with
somes (androgenetic diploidy [homozygous]) [22, triploid chromosomal composition is 69XXY;

58–60]. The remaining CHMs also are androgenetic less often it is XXX, and rarely it is XYY [55,
but are 46XY and formed by dispermy, that is, fertil- 62, 66, 67]. Cytogenetic distinctions between
ization by two spermatozoa of an ovum lacking complete and partial hydatidiform moles have
functional maternal chromosomes (androgenetic not been absolute, however. There have been
diploidy [heterozygous]) [54, 55, 59, 61, 62]. reports of partial hydatidiform moles that are
Cytogenetically, partial hydatidiform mole is diploid [68–70] and occasional complete hyda-
triploid (69 chromosomes); the majority (>90%) tidiform moles that are triploid [68, 71].
have two sets of chromosomes of paternal origin Furthermore, both complete and partial hyda-
(diandric triploidy [dispermic/heterozygous]) tidiform moles have been reported to show
and a haploid maternal set (monogynic) [22, marked heterogeneity in ploidy patterns.
62–64]. The remaining cases develop from one Haploid, aneuploid, and tetraploid moles, both
haploid sperm fertilizing an egg followed by partial and complete, have been reported [55,
reduplication of the paternal chromosome set or 71–75]. DNA ploidy analysis can be helpful in
one sperm that is diploid due to failure of meio- classifying cases that lack clear-cut morpho-
sis I or II (monospermic, homozygous) [55, logic features to identify the mole as either com-
65–67]. In more than two thirds of cases, the plete or partial [65, 69, 71, 72, 76, 77]. A review
Fig. 4.8 Exaggerated placental implantation site of com- lights the importance of a dual immunohistochemical
plete hydatidiform mole. A dual immunohistochemical stain, as the Ki-67 is staining infiltrating lymphocytes and
stain for HLA-G (red) and Ki-67 (brown) demonstrates no not the trophoblastic cells
proliferative activity in the trophoblastic cells. This high-
of apparent nontriploid partial moles found that of CHM and PHM as well as a hydropic abortus
errors in pathologic or ploidy classification may not be present in all sections, and therefore
were common and suggested that nontriploid adequate sampling is essential.
partial moles may not exist [66]. Some cases have obvious features of a hyda-
tidiform mole, but the distinction between a
CHM and a PHM is not clear cut. In these cases,
Differential Diagnosis the morphologic features of CHM and PHM are
ambiguous, even after extensive histologic sam-
Complete Versus Partial pling. Diagnosis of a hydatidiform mole early in
Hydatidiform Mole pregnancy can be particularly difficult, and more
The subtle clinical presentation of most hydatidi- hydatidiform moles are now detected as early as
form moles, either complete or partial, under- the 7th or the 8th week of pregnancy [31–34, 43,
scores the need for careful pathologic evaluation 46]. In such cases the differential diagnosis is
of abortion specimens [43]. Tissue from hyda- especially challenging, because the cistern for-
tidiform moles should be generously sampled (at mation and trophoblastic hyperplasia are less
least four cassettes) to ensure accurate diagnosis. pronounced than those found several weeks later
The histologic features that permit the distinction in gestation.
Fig. 4.9 Partial hydatidiform mole. A mixture of large, a central cistern. Irregular, patchy hyperplasia of tropho-
edematous villi and small, fibrotic villi characterizes par- blast is present along the surface of the larger villi
tial mole. A markedly enlarged villus to the right contains
The distinction between CHM and PHM is p57 has been used in the differential diagno-
straightforward when the characteristic features of sis of hydatidiform moles [78–80, 91]. The p57
either entity are pronounced. Multiple edematous gene is located on chromosome 11p15.5 and
villi with large cisterns and diffuse trophoblastic is expressed predominantly from the maternal
hyperplasia of CHM contrast with the more limited allele in most tissues, and, because complete
villous edema and focal trophoblastic hyperplasia hydatidiform moles contain only paternal DNA,
of PHM. Another feature that separates CHM from p57 has absent expression. Therefore, antibodies
PHM is the generalized edema of the villi in the against the p57 protein generally show little to no
CHM, which contrasts with the mixture of two expression in cytotrophoblastic cells and villous
populations of villi, with fibrosis of some of the stromal cells of CHM (Fig. 4.12). In contrast, the
villi in PHM. Partial hydatidiform moles often cytotrophoblastic cells and villous mesenchyme
show evidence of development of a fetus or embryo of partial hydatidiform moles as well as spontane-
that generally is not seen in CHM. ous abortions show strong immunoreactivity for
Ancillary testing including immunohisto- p57 [79, 80]. It is the presence or absence of p57 in
chemistry for p57KIP2 (p57) can be extremely CT and villous stromal cells that aids in separat-
useful for distinguishing the two forms of hyda- ing CHM from PHM (see Fig. 4.13, Algorithmic
tidiform mole [55, 78–90]. Immunohistochemical approach to diagnosis of hydatidiform moles).
analysis of the paternally imprinted gene product In addition, p57 immunostaining can be seen in
Fig. 4.10 Partial hydatidiform mole. Many of the enlarged villi show irregular outlines. Note multiple trophoblastic
inclusions formed by invagination of the surface of the villi into the stroma
villous intermediate trophoblastic cells [80] and and negative staining in stromal and cytotropho-
in intervillous trophoblast islands composed of blastic cells within individual villi and is based
intermediate trophoblast within complete hyda- on maternal genetic material in those stromal and
tidiform moles [79]. Syncytiotrophoblastic cells cytotrophoblastic cells. The androgenetic com-
are nonreactive for p57 and maternal decidual- ponent does not express p57, while the biparental
ized stromal cells are strongly reactive for p57 in component is positive for p57 [81, 82, 96, 97]. A
all types of gestations. case of a twin gestation with no p57 expression
Rarely a complete hydatidiform mole may in the CHM component and p57 expression in the
occur as a twin gestation in conjunction with a non-molar abortus component highlights this pat-
normal placenta [50, 71, 92–95]. In these cases the tern [95].
curettings contain a mixture of normal sized and Other rare exceptions to the p57 expression
molar villi, which mimics a partial hydatidiform patterns described above exist. Two genotype
mole. Thus, it may not always be possible to clas- confirmed complete hydatidiform moles express-
sify a specimen accurately as a complete or a PHM ing p57 secondary to retained maternal
by morphology alone. Immunohistochemistry for chromosome 11 (location of p57 gene) have been
p57 will display a mosaic (androgenic/biparental reported [98, 99]. Two reported cases of geno-
mosaic conceptions) pattern in this setting. This type confirmed partial hydatidiform moles with
is demonstrated by any combination of positive loss of p57 expression were attributable to loss of
Fig. 4.11 Partial hydatidiform mole. Edematous villi show small foci of haphazard trophoblastic hyperplasia along the
surface and infoldings that form inclusions in the stroma
chromosome 11 [55, 100]. In keeping with the molar hydropic abortion with villous edema
CHM phenotype, the majority of familial recur- (see Chap. 3) [9, 41, 104, 105]. Microscopically,
rent hydatidiform mole cases had loss of p57 the edema of the hydropic abortus can be strik-
expression. However, depending upon the sever- ing. Gross specimens from hydropic abortions
ity of the NLRP7 mutation in FBCHM, a subset generally are smaller, however, and villous
expressed p57 [101]. enlargement is not seen by either the clinician or
To date, most studies find no consistent asso- the pathologist on gross examination (Table 4.3).
ciation between DNA ploidy and the subsequent It is important to keep microscopic observations
clinical course of either partial or complete hyda- in context with the gross findings. Most hydropic
tidiform mole [71]. One report suggested that abortions yield only one or a few cassettes of
aneuploidy predicts persistence in CHM [102], tissue, whereas hydatidiform moles tend to be
but another study found that aneuploid CHM is voluminous. These generalizations usually hold,
associated with less risk for progressive disease but in some cases of hydatidiform mole, the villi
than diploid or tetraploid CHMs [103]. are also not grossly visible. This is especially
true if part of the molar tissue was spontane-
ydatidiform Mole Versus Hydropic
H ously aborted prior to curettage, if the hydatidi-
Abortus or Abnormal (Non-molar) form mole is evacuated very early in gestation,
Villous Lesions if there is collapse of villi secondary to suction
Another frequent consideration in the differen- curettage, or if the specimen is a PHM with lim-
tial diagnosis of hydatidiform mole is the non- ited tissue.
Fig. 4.12 Complete hydatidiform mole. Complete stromal cells (left) and intermediate trophoblast (bottom
absence of p57 immunostaining within cytotrophoblastic center) serve as a positive internal control
cells and villous stromal cells. Maternal decidualized
Several microscopic features distinguish a tial hydatidiform mole, thorough sampling may
hydropic abortus from a hydatidiform mole [2, be needed to establish the diagnosis.
24]. In hydropic abortion the villi are edematous Immunostaining with p57 has been shown to
and avascular. Some also may show trophoblas- improve the recognition of complete hydatidiform
tic inclusions [106]. Occasional small cisterns moles (see above), but had no impact on the dis-
occur in hydropic abortuses, but they are infre- tinction of partial hydatidiform moles and non-
quent and do not cause gross villous enlarge- molar abortus [107, 108] because immunostaining
ment. The most useful feature for separating a for p57 is positive in the villus cytotrophoblast and
hydatidiform mole from a hydropic abortus is stroma of both PHM and non-molar hydropic abor-
the distribution of the villous trophoblast. In a tus. Molecular genotyping via PCR amplification
hydatidiform mole at least occasional villi show of short tandem repeat (STR) loci has been demon-
circumferential hyperplasia of trophoblast along strated to differentiate between the two [55, 65, 95,
their surface, whereas in the hydropic abortus 109–111]. Genotyping will discern androgenetic
the proliferating trophoblast has a polar distribu- diploidy (CHM), diandric triploidy (PHM), and
tion, projecting only from one surface of the biparental diploidy (non-molar abortus) [55, 65, 95,
anchoring villi. Because trophoblastic hyperpla- 109–111], whereas other ancillary techniques, such
sia may be limited and focal, especially in a par- as flow cytometry, conventional cytogenetics, and
Fig. 4.13 Algorithmic approach to diagnosis of hydatidi- non-molar specimens. B. In another approach, potentially
form moles. A. Per one approach, potentially molar speci- molar specimens are universally subjected to genotyping.
mens are universally subjected to immunohistochemical C. In yet another approach, some triage to p57 immuno-
analysis of p57 expression, with triage to genotyping histochemistry versus genotyping, respectively, is per-
based on this result. p57-negative cases are diagnosed as formed based on morphologic assessment as favoring
complete hydatidiform mole, and p57-positive cases are complete hydatidiform mole versus partial hydatidiform
genotyped to distinguish partial hydatidiform moles from mole, respectively (Courtesy of Dr. Brigitte Ronnett)
fluorescence in situ hybridization (FISH), cannot be followed by serum β-hCG titers to determine if
discriminate paternal versus maternal chromosome the lesion represents a molar pregnancy that may
complements. Because diagnostic reproducibility require further therapy (Fig. 4.13, Algorithmic
among gynecologic pathologists is low for PHM approach to diagnosis of hydatidiform moles).
versus non-molar specimens based on morphol-
ogy and p57 immunostaining alone, ancillary
testing with genotyping is encouraged for defini- Abnormal (Non-molar) Villous
tive diagnosis [107, 108]. Genotyping provides Lesions
definitive diagnosis for the ~25–50% of cases that
were misclassified by morphology, especially the Abnormal (non-molar) villous lesions are a cat-
cases with unresolved p57 immunostaining [107]. egory recognized by the WHO and are defined
Genotyping is encouraged as a definitive diagnos- as various non-molar, villous lesions with his-
tic tool and is available at larger medical institu- tological features simulating a partial hydatidi-
tions, and the tissue blocks can be submitted for form mole [1]. The morphology has a similar
analysis. It is recognized, however, that this testing spectrum to a PHM, with villous irregularity
may not be available for all pathologists evaluat- in size and shape, varying degrees of tropho-
ing these specimens. In these cases, a descriptive blastic hyperplasia with occasional syncytiotro-
diagnosis as “abnormal villous morphology,” the phoblastic buds or “snouts,” scalloped villous
result of p57 immunostaining, and the differential outlines, and rarely trophoblastic inclusions
diagnosis in a note may suffice since patients can (Fig. 4.14). These morphologic characteristics
Fig. 4.14 Abnormal (non-molar) villous lesion. Villous abortus). This case demonstrates the overlapping morpho-
irregularity to size and shape, with hydropic forms (right). logic features with a PHM and the importance of geno-
Fetal tissue (left) and trophoblastic “inclusions” are seen. type testing
Genotyping confirmed biparental diploidy (non-molar
can be manifest of hydropic abortions, digynic show marked hydrops of placental stem villi
triploid conceptions, and chromosomal trisomy that can mimic a hydatidiform mole in later
syndromes [1]. DNA genotyping is encour- pregnancy. In contrast to a hydatidiform mole,
aged for distinction between PHM and these however, no abnormal trophoblast proliferation
diagnoses. or trophoblastic “inclusions” are seen in the
enlarged villi [112].
Other Considerations
ersistent Postmolar Gestational
P
Placental abnormalities such as placental angio- Trophoblastic Disease and Invasive
matous malformation or placental mesenchymal Hydatidiform Mole
dysplasia, which is associated with Beckwith-
Wiedemann syndrome, are typically seen in After a hydatidiform mole has been evacuated,
more advanced gestations, in the second half of a subsequent curettage may be done for persis-
pregnancy, and are not a significant issue in the tence or elevation of follow-up hCG titers or
differential diagnosis of early abortion speci- for significant uterine bleeding [114–116].
mens [50, 112, 113]. These abnormalities may The repeat curettage may show persistent
hydatidiform mole, choriocarcinoma, retained myometrium. Trophoblastic tissue without

implantation site, no evidence of trophoblastic villi is not choriocarcinoma but persistent tro-
tissue, or, rarely, a new pregnancy [114]. If the phoblast [117].
specimen contains persistent hydatidiform With invasive hydatidiform mole, hydropic
mole, it will show residual molar villi mixed molar villi and hyperplastic trophoblast either
with trophoblast. Usually the amount of tissue invade myometrium or are present at other sites,
and villi are greatly reduced compared to the usually the vulva, vagina, or lungs [2, 24, 118].
original curettage specimen, but if villi are To establish the diagnosis, it is necessary to
present, the diagnosis remains that of persis- clearly identify molar villi beyond the
tent intracavitary mole (Fig. 4.15). However, endometrium. In curettings this requires finding
the trophoblastic proliferations may be exu- the villi within myometrial smooth muscle, an
berant (Fig. 4.16). Choriocarcinoma is diag- extremely rare event. Consequently, invasive
nosed when there is abundant trophoblast that mole is almost never diagnosed by endometrial
shows a trimorphic arrangement of syncytio- biopsy or curettage. It is important to remember
trophoblast, cytotrophoblast, and intermediate that the presence of residual hydatidiform mole
trophoblast without villi. In addition, it is nec- in a curettage specimen does not represent inva-
essary to see tumor necrosis or destructive sive mole in the absence of demonstrable myo-
infiltrative growth by trophoblast into the metrial invasion.
Fig. 4.15 Persistent trophoblast following hydatidiform after evacuation of a hydatidiform mole. This should be
mole. Prominent trophoblastic proliferation extending diagnosed as “persistent hydatidiform mole”
from the surface of a villus found in curettage specimen
Fig. 4.16 Persistent complete hydatidiform mole. ing cytotrophoblast with small, uniform nuclei and multi-
Persistent hydatidiform mole obtained by curettage sev- nucleated syncytiotrophoblast. A villus was present in
eral weeks after initial evacuation of a complete hydatidi- another field indicating the diagnosis is a persistent hyda-
form mole. The trophoblastic proliferation is striking, tidiform mole and not choriocarcinoma
with pleomorphic implantation type IT cells (left), overly-
linical Queries and Reporting

C times hydatidiform mole is clinically suspected
of Hydatidiform Mole when visibly enlarged, edematous villi are
encountered at curettage. In such cases the logical
One of the most important clinical questions in the questions are whether a hydatidiform mole is
evaluation of an abortion specimen is whether ges- present and, if so, whether it is a CHM or
tational trophoblastic disease is present, as hydatidi- PHM. With the aid of ancillary diagnostic tools
form mole, choriocarcinoma, or PSTT can present such as p57 immunohistochemical staining and
as a spontaneous or missed abortion. Even a thera- molecular genotyping, the distinction between
peutic abortion for an apparently normal gestation CHM, PHM and non-molar abortus is possible.
may reveal an unsuspected hydatidiform mole. With either type of hydatidiform mole, the grad-
Often the clinician suspects a hydatidiform ing of the trophoblast does not have clinical sig-
mole from the clinical history and from findings nificance in terms of the overall risk for persistent
such as rapid uterine enlargement, abnormally GTD [42]. If there is any doubt about whether a
high β-hCG titers, or preeclampsia in the first tri- specimen should be classified as an abortus with
mester. Ultrasound findings may support the clini- abnormal villous morphology or as a hydatidi-
cal impression of a hydatidiform mole. At other form mole, more tissue should be submitted if
available, with subsequent ancillary testing (geno- About 2–3% of patients with complete hyda-
typing when available) as discussed above for a tidiform mole will develop choriocarcinoma.
definitive diagnosis. If a case remains equivocal, Among patients with partial hydatidiform mole,
the gynecologist should be cautioned to follow the risk of choriocarcinoma is much lower, as it
the patient to be certain that β-hCG titers return to occurs in 0.5% [2, 20, 21, 133–139]. Most cases of
normal before pregnancy is attempted again. persistent GTD represent persistent hydatidiform
Management of either type of hydatidiform mole in the uterine cavity or invasive mole in the
mole requires monitoring of serum β-hCG titers myometrium. Less often a patient has invasive
after evacuation in conjunction with contraception mole with villi and trophoblast that migrates to the
until undetectable levels are obtained. Patients with lungs, vulva, or vagina or lung metastases that are
known hydatidiform mole who are being followed not biopsied [140]. Development of choriocarci-
may undergo repeat curettage for continued bleed- noma following a PHM is very rare [136–138,
ing or for abnormally persistent or elevated β-hCG 141, 142]. Genotyping and chromosome in situ
titers. If the biopsy shows trophoblastic tissue, then hybridization analysis has shown that some cases
the presence or absence of chorionic villi is impor- of genetically confirmed PHM have metastatic tro-
tant and should be clearly reported. Molar villi gen- phoblastic disease in the lungs and liver [139]. In
erally indicate persistent intrauterine hydatidiform 2000, the International Federation of Obstetrics
mole rather than invasive mole or development of and Gynecology (FIGO) devised a staging and
choriocarcinoma [117]. Immature but normal villi risk factor scoring system for women with malig-
indicate a new pregnancy unrelated to GTD. nant gestational trophoblastic disease [143, 144].
The FIGO system uses eight parameters for clini-
cal risk assessment (Table 4.4).
Gestational Trophoblastic
Neoplasms
Choriocarcinoma
A gestational trophoblastic neoplasm (GTN) follow-
ing a complete or partial hydatidiform mole (or General Features
rarely a normal pregnancy) is detected by serum Choriocarcinoma can be gestational, originating
hCG titers that fail to return to normal or plateau from a pregnancy (normal or abnormal), or non-
[119]. The risk of GTN is greater with a complete as gestational. Gestational choriocarcinoma can occur
compared to a partial hydatidiform mole [120, 121]. in the uterine cavity following any type of preg-
Up to 20% of patients with CHM and up to 5% of nancy [2, 5, 24, 121]. As a rule, the risk of chorio-
patients with PHM require further therapy, usually carcinoma increases with the abnormality of the
chemotherapy, for a plateau or increase in the serum antecedent gestation. Complete hydatidiform mole
β-hCG titer after evacuation [2, 122–129]. Rarely is a major predisposing factor, and about half the
GTN can develop after serum β-hCG normalization. cases of choriocarcinoma follow a CHM [145,
In one study GTN developed in 0.4% of patients 146]. Choriocarcinoma also can arise from the tro-
with a mean time to diagnosis of 18 months; nearly phoblast of an abortion or a term pregnancy.
all developed following a CHM [120]. When moni- Consequently, this lesion may be present whenever
toring serum β-hCG titers, a false-positive serum abnormal vaginal bleeding occurs during the post-
β-hCG level, also known as “phantom β-hCG” is partum period in a woman who has had a preg-
seldom encountered. This is caused by non-specific nancy of any type. Patients with choriocarcinoma
heterophilic antibodies in the patient’s sera [130] and also can present with metastatic disease without
can lead to overtreatment. If suspected, a urine hCG uterine signs or symptoms. Typically, the patient
test should be performed as heterophilic antibodies with choriocarcinoma has markedly elevated serum
will not affect a urine hCG test [131, 132]. β-hCG titers.
Gestational Trophoblastic Neoplasms 95
Table 4.4 FIGO/WHO scoring system based on prognostic factors

FIGO risk score 0 1 2 4
Age <40 ≥40 _ _
Type of pregnancy Hydatidiform mole Abortion Term _
Months from index pregnancy <4 4–6 7–12 >12
Pretreatment β-hCG <103 103–104 104–105 >106
Largest tumor size (cm) <3 3–5 >5 _
Site of metastases Lung Spleen, kidney Gastrointestinal tract Liver, brain
Number of metastases _ 1–4 5–8 >8
Previous failed chemotherapy _ _ Single agent Two or more agents
Low-risk score ≤6; high-risk score ≥7
Data from Kohorn [143] and Ngan et al. [144]
Pathologic Features often contain erythrocytes (Figs. 4.18 and 4.19). In

Choriocarcinoma is hemorrhagic and necrotic, contrast to ST, CT cells are small (about the size of
composed of trophoblastic cells without villi that a decidualized stromal cell) and uniform. They
invade normal tissues (Fig. 4.17). The two main have a single nucleus with a prominent nucleolus,
diagnostic features are an absence of chorionic pale to clear cytoplasm, and distinct cell borders.
villi and a trimorphic population of trophoblast Large IT cells with polygonal shapes and one or
cells (Figs. 4.17, 4.18 and 4.19). The first criterion, two large, hyperchromatic nuclei also occur in
absence of villi, is important, as the proliferative choriocarcinoma (Figs. 4.17 and 4.18). The per-
trophoblast of hydatidiform moles or of early nor- centage of IT cells in choriocarcinoma is highly
mal pregnancy can closely simulate the tropho- variable, ranging from rare cells to the large major-
blast of choriocarcinoma. However, the presence ity of mononucleate trophoblastic cells. Typically,
of villi does not exclude the diagnosis, as demon- there is generalized enlargement of the trophoblas-
strated in exceedingly rare cases, choriocarcinoma tic cells with increased nuclear atypia in chorio-
can occur with a CHM [121]. The second criterion carcinoma compared to normal trophoblastic cell
of choriocarcinoma, a trimorphic pattern of syncy- populations in early pregnancy.
tiotrophoblastic (ST) cells alternating with nests or The amount of trophoblast in cases of chorio-
sheets of mononucleate trophoblast (cytotropho- carcinoma is highly variable. There may be
blastic [CT] or intermediate trophoblastic [IT]) abundant neoplastic tissue, but frequently only a
cells should be found, at least focally, to establish small amount of viable tumor associated with
a histologic diagnosis of choriocarcinoma. Often, extensive hemorrhage is present. Small amounts
the characteristic pattern of choriocarcinoma is of tumor may pose problems in diagnosis. While
clear, but at times the trimorphic population of tro- ST cells generally are prominent in uterine ges-
phoblast may be difficult to recognize (Fig. 4.20). tational choriocarcinoma, in occasional cases
The admixture of ST alternating with CT or IT these cells are indistinct, and CT and IT cells
cells yields a plexiform pattern. In these cases, predominate.
identification of ST cells is an important diagnos- Choriocarcinoma is largely an H&E diagnosis,
tic feature. These cells contain multiple nuclei, but the ST cells are strongly reactive for hCG,
ranging from 3 to more than 20 per cell, which are which highlights the plexiform pattern, whereas
variable in size. Often the nuclei are pyknotic, but CT and IT cells are generally nonreactive. The syn-
they can be vesicular with prominent nucleoli. ST cytiotrophoblastic cells of choriocarcinoma also
cells have dense eosinophilic to amphophilic cyto- show strong immunoreactivity for inhibin-α. The
plasm with small vacuoles or large lacunae that ST and IT cells will stain for HSD3B1 [147] while
Fig. 4.17 Choriocarcinoma. A mass of trophoblast with prominent syncytiotrophoblast in a curetting. Abundant tro-
phoblastic tissue is present with no associated villi. The trophoblast shows invasion of the endomyometrium (right)
the IT cells will stain with human leukocyte anti- villi is present in an abortion specimen. In such
gen (HLA-G) in 78–100% of cases [148–150]. cases the trophoblast is small in quantity and
lacks necrosis, significant hemorrhage, and atypi-
Differential Diagnosis cal nuclear features. The finding of a significant
The differential diagnosis of choriocarcinoma amount of secretory endometrium, decidua, or
includes physiologic trophoblastic proliferations placental implantation site favors the presence of
associated with normal pregnancies and the normal trophoblast and not choriocarcinoma.
trophoblast associated with hydatidiform moles, The presence of atypia, including nuclear pleo-
as well as PSTT, ETT, and nontrophoblastic morphism, macronucleoli, and abnormal mitotic
tumors including exaggerated placental site figures, strongly suggests choriocarcinoma.
(EPS) and placental site nodule (PSN)(see Rarely, choriocarcinoma may arise in a devel-
Nonneoplastic Lesions below). Trophoblast of oping placenta, and this usually is seen in a pla-
normal pregnancy can appear highly prolifera- centa from a second- or third-trimester gestation
tive, especially at the implantation site of anchor- [121, 151]. We have seen choriocarcinoma aris-
ing villi. In contrast to choriocarcinoma, the ing in association with retained intrauterine villi
trophoblast in normal pregnancy usually is asso- from either an abortion or a term pregnancy. In
ciated with small, immature chorionic villi and these very unusual cases, curettage specimens
decidua. Occasionally, only trophoblast without show a few villi that tend to be hyalinized or
Fig. 4.18 Choriocarcinoma. Trimorphic population of with hyperchromatic nuclei and dense, eosinophilic cyto-
ST, CT, and IT cells in choriocarcinoma. Large ST cells plasm predominate in the right field. Mononucleate CT
with abundant, vacuolated cytoplasm and numerous cells are scattered on the right field. The trophoblast
mitotic figures predominate in the left field, while IT cells shows invasion of the myometrium (bottom right)
fibrosed, mixed with fragments of tissue showing sistent hydatidiform mole can show large sheets
choriocarcinoma. of trophoblast with an alternating arrangement of
Because hydatidiform mole proliferations can cytotrophoblast and syncytiotrophoblast mixed
be significantly atypical and florid, there is con- with hemorrhage. Trophoblast may be prominent
troversy regarding the diagnosis of choriocarci- in the original curettage sample (Fig. 4.6) or in
noma in the setting of concurrent CHM. The subsequent curettings done for abnormal eleva-
marked trophoblastic hyperplasia and cytologic tion of β-hCG titers (Fig. 4.16).
atypia found in some cases of CHM and persis- Nonetheless, if any villi are present the diag-
tent hydatidiform mole closely resemble the pat- nosis is likely that of hydatidiform mole.
terns found in choriocarcinoma (Figs. 4.6 and However, cases of polymerase chain reaction
4.16). In general, if edematous chorionic villi are amplification (genotype confirmation testing) of
present, no matter how much trophoblastic pro- tumor tissue and the associated CHM supported
liferation is present, the lesion is a hydatidiform the origin of the choriocarcinoma from the con-
mole. This feature can be particularly problem- current molar pregnancy [121]. Genotyping also
atic, however, in persistent hydatidiform mole can be useful to differentiate gestational from
that is found in repeat curettings, since villi often non-gestational choriocarcinomas, as gestational
are sparse in these specimens. These cases of per- associated choriocarcinomas have a better prog-
Fig. 4.19 Choriocarcinoma. Endometrial curettings fol- nent, the juxtaposition of syncytiotrophoblast forming a
lowing a complete mole show ST cells containing multi- trimorphic population establishes the diagnosis of
ple nuclei and cytoplasmic vacuoles with CT and IT choriocarcinoma
interspersed. Although intermediate trophoblast is promi-
nosis [121]. Even without villi, a diagnosis of tory helps resolve this question, as high-grade
choriocarcinoma should be made only when carcinomas and sarcomas generally occur in older
atypical, trimorphic trophoblast without villi are postmenopausal patients, whereas trophoblastic
present along with hemorrhage and tumor necro- tumors occur in women of reproductive age.
sis or unequivocal invasion of the myometrium Furthermore, many patients with choriocarci-
by a trimorphic population of trophoblastic cells. noma have a history of a prior hydatidiform mole.
In the absence of these features, the diagnosis In equivocal cases, immunohistochemical stains
should be persistent trophoblast. for hCG are useful for demonstrating syncytiotro-
Nontrophoblastic tumors may mimic chorio- phoblast and HLA-G will demonstrate the inter-
carcinoma when they show a large component of mediate trophoblast (Table 4.5). In addition,
giant cells. Anaplastic carcinomas and sarcomas HSD3B1 will stain both ST and IT cells in chorio-
with tumor giant cells may simulate choriocarci- carcinomas [147]. Because both choriocarcinoma
noma, at least focally, and may rarely show cho- and anaplastic carcinoma are epithelial tumors,
riocarcinomatous differentiation although this immunostains for cytokeratin are not useful in
occurs in older women. Usually the clinical his- establishing the diagnosis.
Fig. 4.20 Choriocarcinoma. In this field ST cells are indistinct and CT cells predominate, yielding a pattern resem-
bling poorly differentiated nontrophoblastic carcinoma
Table 4.5 Immunohistochemistry of uterine trophoblastic tumors compared with nontrophoblastic tumors
Choriocarcinoma PSTT ETT Nontrophoblastic uterine tumors
HLA-G + to +++ +++ + −
HSD3B1 +++ +++ +++ +++
hCG +++ +/− +/− +/−
Inhibin-α ++ ++ ++ −
hPL + +++ +/− −
PLAP +/− − ++ +/−
Mel-CAM ++ +++ +/− −
Keratina +++ +++ +++ +++b
EMA + ++ ++ ++b
p63 + − +++ +
PSTT placental site trophoblastic tumor, HLA-G human leukocyte antigen, hCG human chorionic gonadotropin, hPL
human placental lactogen, PLAP placental alkaline phosphatase, EMA epithelial membrane antigen
a
Immunostaining for keratin AE1/AE3
b
Keratin and EMA immunostaining for carcinomas, only
Rarely choriocarcinoma may be found in a genetic analysis has established the trophoblastic
postmenopausal patient. In such cases the tumor origin of these tumors [163]. The β-hCG titer is
may represent gestational choriocarcinoma with generally low and may not be noticeably elevated
a long latent period, or it may represent somatic if a sensitive assay method is not used. Because
carcinoma with choriocarcinomatous transfor- PSTTs extensively infiltrate the myometrium, the
mation [152, 153]. PSTT and ETT are discussed uterus can be perforated during curettage. These
and compared with choriocarcinoma in the fol- neoplasms usually are benign, despite destructive
lowing sections. growth in the myometrium. About 15% of
reported tumors have shown aggressive malig-
nant behavior with disseminated metastases.
Placental Site Trophoblastic Tumor
Pathologic Features
General Features PSTT typically produces a mass lesion. These
The placental site trophoblastic tumor (PSTT) is tumors range from focal lesions 1 to 2 cm in
a rare form of trophoblastic neoplasia composed diameter to large masses that replace much of the
predominantly of implantation site IT [2, 24, 48, corpus. Therefore, curettings of PSTT typically
154–157]. Like other forms of GTD, it almost yield multiple fragments of neoplastic tissue.
always occurs during the reproductive years. Microscopically PSTT is composed predomi-
PSTT is typically diagnosed at the time a curet- nantly of implantation site IT cells that invade
tage is performed in a woman who is thought to normal tissues (Fig. 4.21) [164]. These cells gen-
be pregnant, usually with a preoperative diagno- erally are polyhedral and grow in cohesive
sis of a missed abortion, but in contrast to chorio- masses that often show areas of necrosis. The
carcinoma, this tumor rarely is directly associated curettings usually include fragments of myome-
with a recent pregnancy [158–162]. Molecular trium infiltrated by IT cells. The IT cell cyto-
Fig. 4.21 Placental site trophoblastic tumor. Low magnification shows PSTT invading endomyometrial junction.
Residual endometrium spans the top of the micrograph
plasm is generally amphophilic with occasional confluent masses of cells infiltrate and dissect between
clear vacuoles and distinct cell borders. Some smooth muscle fibers (Fig. 4.23). Furthermore, in
parts of the tumor, especially in areas of PSTT there is a characteristic pattern of vascular inva-
myometrial invasion, are composed of spindle- sion in which the intermediate trophoblast surrounds
shaped cells. Most cells have a single irregular and replaces the vessel wall while retaining its overall
and hyperchromatic nucleus, but binucleate and structural integrity (Figs. 4.24 and 4.25). One other
multinucleate IT cells are also present. Marked constant finding in this tumor is patchy deposition to
variation in nuclear size and shape is often a fea- complete replacement of the vessel wall by eosino-
ture of the tumor (Fig. 4.22). Some nuclei have philic fibrinoid material (Figs. 4.24–4.26). The hya-
deep folds or grooves, and others may have pseu- line, amorphous deposits of fibrinoid also occur
doinclusions because of large cytoplasmic invag- randomly throughout the tumor, often entrapping
inations. Scattered ST cells with several nuclei individual cells (Fig. 4.27).
and vacuolated cytoplasm are also present. In PSTT the IT cells are diffusely immunore-
Mitotic activity is usually low but can be brisk active for HSD3B1, human leukocyte antigen
including occasional abnormal forms. (HLA-G), human placental lactogen (hPL), and
Besides the characteristic cytologic features of the Mel-CAM (CD 146), as well as inhibin-α, epithe-
individual cells, the growth pattern of intermediate lial membrane antigen, and cytokeratins (AE1/
trophoblast in PSTT is an important diagnostic fea- AE3 and cytokeratin 18) [2, 147–150, 165]
ture. When myometrium is present in the biopsy, the (Table 4.5). These tumors are rarely or only
Fig. 4.22 Placental site trophoblastic tumor. The cells in nuclear size. In the center, the IT infiltrate and replace
are pleomorphic and generally contain a single hyper- the wall of a blood vessel while preserving the lumen.
chromatic nucleus. There is moderate to marked variation Fibrinoid material is present around the vessel
Fig. 4.23 Placental site trophoblastic tumor. Confluent masses of intermediate trophoblast dissecting between smooth
muscle fibers
focally positive for hCG, PLAP, and p63 [2, 149, malignant cases [2, 169]. In contrast, the benign
157, 164, 166, 167]. The Ki-67 proliferation tumors usually show a mitotic rate of about two
index of this tumor is approximately 20% [2]. mitoses per ten HPFs, with the highest reported
It is difficult to reliably predict the behavior rate being five mitoses per ten HPFs. In several
of PSTT based on the microscopic features, and clinically malignant PSTTs, the mitotic rate was
therefore this neoplasm is not divided into only two per ten HPFs [170, 171], so it appears
benign and malignant categories. The reported that some overlap exists in the mitotic rates of
malignant cases of PSTT generally show some malignant and benign PSTT. Abnormal mitotic
features that predict aggressive behavior. These figures occur in benign as well as malignant
clinically malignant tumors are composed of PSTT. Preliminary findings suggest that Ki-67
larger sheets and masses of cells with more labeling index may be a significant prognostic
extensive necrosis than benign tumors [2, 24, 48, indicator as it is usually greater than 50% in
156, 157]. In malignant PSTT the cells also tend malignant tumors but only about 14% in benign
to have clear instead of amphophilic cytoplasm PSTT [2]. One series found only stage and clear
[168]. Finally, the mitotic rate usually is higher cytoplasm to be independent predictors of over-
in the malignant tumors, with more than five all survival, while age and stage were predictors
mitoses per ten high-power fields (HPFs) in most of disease-free survival [168].
Fig. 4.24 Placental site trophoblastic tumor. Early-stage tion, in the bottom left and right corners of the micro-
vessel involvement. In this field, the IT cells are starting to graph, IT cells are embedded in amorphous fibrinoid
invade a blood vessel wall (vessel left of center). In addi- material around vessels (late-stage vessel involvement)
Differential Diagnosis suspicious for PSTT. The Ki-67 labeling index is

The differential diagnosis of PSTT includes very useful in the differential diagnosis because
exaggerated placental site (EPS), choriocarci- the index is near zero in the normal and exagger-
noma, and other, nontrophoblastic tumors. An ated implantation site but 14% ± 6.9% in PSTT
EPS is one of the most important considerations [172]. It may be slightly elevated (<5%) in EPS
in the differential diagnosis, as it can have fea- associated with a CHM [172]. We advocate for
tures that are very similar to those of PSTT. The the use of a dual expression marker, one for IT
distinction is largely one of degree (Table 4.6). cells such as HLA-G or HSD3B1 in addition to
The EPS usually is a focal finding, maintaining Ki-67, therefore, highlighting proliferating IT
the overall architecture, and in other portions of cells. The Ki-67 index helps to differentiate an
the tissue there is decidua and/or chorionic villi. EPS (IT cells with a proliferation index <1%)
PSTT, in contrast, is composed of sheets and from PSTT (IT cells with a proliferation index
masses of cells typically accompanied by necro- >10%) [172]. The use of Ki-67 alone is discour-
sis with little normal tissue in the sections. In aged as natural killer cells, and activated T lym-
addition, the EPS tends to have more ST giant phocytes in the EPS or PSTT will also stain with
cells, the nuclei tend to have smudged (degenera- Ki-67 and hinder the accurate assessment of pro-
tive) chromatin, and mitotic activity is absent. liferating IT cells. Implantation site IT cells of
Any evidence of unequivocal mitotic figures is EPS and PSTT will not stain with p63, whereas
Fig. 4.25 Placental site trophoblastic tumor. Mid-stage lumens. Eosinophilic fibrinoid material deposits are seen
vessel involvement. In this field the trophoblastic cells in the blood vessel walls
have invaded blood vessel walls and protrude into the
the chorionic leave IT cells of placental site nod- for HSD3B1, HLA-G [148–150], hPL, and hCG,
ules (PSN) and ETT are typically positive [149, the ratio of the number of immunoreactive cells for
173, 174] (Fig. 4.28 Trophogram). hPL and hCG differs in the two tumors. In PSTT
Separation of PSTT from choriocarcinoma the hPL/hCG ratio is typically 3:1, whereas in
is important, as these two tumors behave differ- choriocarcinoma it is 1:3. Occasionally, however,
ently and are treated differently. Choriocarcinoma PSTT shows a ratio of hPL and hCG staining that
may have a monomorphic appearance in some more closely resembles that of choriocarcinoma.
areas and can have large numbers of IT. In con- The Ki-67 proliferative index can also assist in the
trast to PSTT, however, a network of syncytio- differential diagnosis, as it is very high (>50%) in
trophoblast, cytotrophoblast and intermediate choriocarcinoma and significantly lower in PSTT
trophoblast in choriocarcinoma results in a trimor- (15–20%) [172].
phic population, at least focally (Table 4.6). The Rarely, a trophoblastic tumor may show fea-
syncytiotrophoblast in the PSTT is composed of tures of both choriocarcinoma and PSTT. This is
isolated giant cells that do not show the trimor- called a mixed PSTT-choriocarcinoma. There is
phic plexiform pattern found in choriocarcinoma. insufficient experience with these tumors to pre-
Immunohistochemical stains for hCG can be espe- dict their behavior accurately.
cially helpful in highlighting the network of syncy- Now that the histologic features of intermedi-
tiotrophoblast in choriocarcinoma. Although both ate trophoblast are better recognized, distinction
PSTT and c horiocarcinoma show immunostaining of PSTT from other forms of malignancy is less
Fig. 4.26 Placental site trophoblastic tumor. Late-stage vessel involvement. A blood vessel wall replaced by eosino-
philic fibrin resulting in an intact, flaccid lumen. Monotonous trophoblastic cells surround the vessel
Fig. 4.27 Placental site trophoblastic tumor. The tumor amount of eosinophilic cytoplasm. Extracellular eosino-
is composed of a monomorphic population of IT cells philic fibrinoid material surrounds the cells in a haphazard
with hyperchromatic, irregular nuclei and a moderate fashion
Table 4.6 Comparison of microscopic features. Choriocarcinoma, PSTT, and exaggerated placental site
Exaggerated placental
Choriocarcinoma PSTT site
Amount of lesional tissue Variable, often abundant Variable, often Usually limited
abundant
Villi Absent Absenta Usually present, focal
Trophoblast growth pattern and cell Trimorphic: ST with CT and Monomorphic: IT Monomorphic: IT
types IT cells cells cells
Mitoses Present, usually high rate Present, usually low Absent or rare
rate
Nuclear atypia Variable, may be marked Moderate to marked Moderate
Necrosis Usually present Usually present Absent
Immunohistochemistry
hCG ++++ + +
hPL + ++++ ++++
*
Ki-67 labeling index High Usually about 15% Very low to zero
PSTT placental site trophoblastic tumor, ST syncytiotrophoblast, CT cytotrophoblast, IT intermediate trophoblast,
hCG human chorionic gonadotropin, hPL human placental lactogen
a
Rarely present and, if so, very focal
*
We advocate for the use of a dual expression marker, one for IT cells such as HLA-G or HSD3B1 in addition to Ki-67.
The use of Ki-67 alone is discouraged as natural killer cells, and activated T lymphocytes will also stain with Ki-67 and
hinder the accurate assessment of proliferating IT cells
Trophoblastic lesion?
HSD 3B +++ (diffuse)

LMW cytokeratin +++ (diffuse)
Trophoblastic lesion
HSD3B1
p63 – p63 +++ beta-hCG positive
hPL+++ hPL–/+ syncytiotrophoblast
Lesions of implantation site Lesions of chorionic type

intermediate trophoblast intermediate trophoblast Choriocarcinoma
Ki-67 Ki-67 Ki-67 Ki-67

< 1% > 10% < 8% > 12%
hPL cyclin E – cyclin E ++
EPS PSTT PSN ETT
Ki-67
Fig. 4.28 Trophogram- an algorithmic approach to trophoblastic proliferations

problematic. PSTT is a tumor of the reproductive riocarcinoma and PSTT with features resembling
years, whereas many of the malignant tumors those of somatic carcinomas. In half of the cases,
that enter the differential diagnoses tend to occur it arises from the lower uterine segment or cervix
at a more advanced age. Cytologic features of and has overlapping morphologic features with
intermediate trophoblast combined with the typi- squamous cell carcinoma. This lesion was ini-
cal patterns of vascular invasion and fibrinoid tially observed in a few patients with persistent
deposition usually allow differentiation of PSTT lung metastases following intensive chemother-
from other neoplasms. In biopsies, PSTT may be apy for documented choriocarcinoma [179].
confused with keratinizing squamous cell carci- Similar lesions were reported as multiple nodules
noma when the keratin has an amorphous eosino- of intermediate trophoblast in the uteri of patients
philic appearance that superficially resembles the following evacuation of hydatidiform moles
fibrinoid of PSTT. Squamous cell carcinoma usu- [180]. These tumors have also been seen in
ally arises in the cervix, however, whereas PSTT patients without a history of antecedent GTD. We
occurs in the corpus; clinical features can help have observed similar tumors that merged imper-
distinguish these two neoplasms. Furthermore, ceptibly with typical choriocarcinoma and
with squamous cell carcinoma, a transition to a PSTT. This lesion also has been found in the
more obvious squamous epithelium or normal uterus adjacent to placental site nodules (PSN)
endocervix often is found. Conversely, in the case following hydatidiform mole.
of PSTT, the curettage samples often contain The epithelioid trophoblastic tumor is preceded
fragments of endometrium or myometrium that by a term gestation in two thirds of cases, with
have been invaded by the tumor. Cervical carci- spontaneous abortions and hydatidiform moles
noma is normally positive for p16 immunostain- being the antecedent gestation in the remaining
ing and in one study, about 50% of intermediate cases. Usually there is a long interval following the
trophoblast tumors also stained for p16 making gestation and the diagnosis of this tumor with a
this marker less helpful in distinguishing the two range of 1–18 years (average, 6.2 years). A case
lesions [149]. has been reported in a postmenopausal woman
On occasion PSTT can mimic leiomyosar- [181]. Serum β-hCG levels are usually elevated at
coma, especially in areas where intermediate tro- the time of diagnosis, although the levels are gen-
phoblast invades myometrium and becomes erally low (<2500 mIU/ml).
intimately admixed with smooth muscle cells.
The trophoblastic cells of the PSTT are strongly Pathologic Features
immunoreactive for cytokeratin, inhibin-α, hPL, Epithelioid trophoblastic tumor lacks the trimor-
and Mel-CAM, whereas leiomyosarcomas, except phic pattern of choriocarcinoma and is composed
for occasional staining with keratin, are negative of chorionic-type IT (Figs. 4.29 and 4.30) [2,
for these other markers (Table 4.5). The epitheli- 164, 175]. ST cells are indistinct. The tumor dis-
oid appearance of intermediate trophoblast can plays a nodular growth pattern and has a striking
also resemble high-grade nontrophoblastic carci- epithelioid appearance, both in its cytologic fea-
nomas. In questionable cases, immunohistochem- tures and in its pattern of invasion. The neoplasm
ical stains for HLA-G, HSD3B1, hPL, Mel-CAM, is composed of small nests and cords of cells.
and hCG are very helpful in distinguishing PSTT The nests often contain dense central hyaline
from a nontrophoblastic neoplasm. material and necrotic debris, and the cords are
encompassed by a hyaline matrix (Fig. 4.31). The
predominant cells are relatively uniform in size
Epithelioid Trophoblastic Tumor and are mononucleate with round, uniform nuclei
and eosinophilic or clear cytoplasm. They are
General Features larger than CT cells, smaller than implantation
Epithelioid trophoblastic tumor is another rare site IT cells, but closely resemble chorionic-type
form of trophoblastic tumor [2, 163, 175–178]. IT cells seen in the mature placenta. Rarely,
This trophoblastic neoplasm is distinct from cho- larger cells resembling implantation site IT cells
Fig. 4.29 Epithelioid trophoblastic tumor. Cords and nests of trophoblastic cells with indistinct ST. The epithelioid
cells have pale to vacuolated cytoplasm
are admixed or scattered within the extracellular CAM (CD 146) have variable or only focal
hyaline material. Apoptotic cells and islands of expression (Table 4.5) [2, 173, 175]. The epitheli-
necrotic debris are abundant in most tumors. oid trophoblastic tumor has a Ki-67 proliferative
Focal areas resembling PSN, PSTT, or choriocar- index of 10–25% with a mean of about 20% [2].
cinoma may be seen within these tumors and
mixed tumors may rarely occur. Differential Diagnosis
These tumors have immunohistochemical pro- The differential diagnosis includes PSN, PSTT,
files like that seen in normal chorionic-type choriocarcinoma, and keratinizing squamous cell
IT. They are diffusely reactive for cytokeratin carcinoma. The distinction from a PSN is usually
(AE1/AE3 and cytokeratin 18) as well as epithe- not difficult because the PSN is typically a micro-
lial membrane antigen. The IT cells of ETT scopic, well-circumscribed nodule with low cel-
express p63 and the trophoblastic markers lularity (see below, Nonneoplastic Lesions),
HSD3B1 and HLA-G [148–150], while other tro- while ETTs are larger, cellular neoplasms that
phoblastic markers including hCG, hPL, and Mel- show necrosis. The IT cells in PSNs are bland
Fig. 4.30 Epithelioid trophoblastic tumor. The tumor uniform, polygonal, mononucleate trophoblastic cells of
lacks the trimorphic pattern of CT, IT and ST of chorio- chorionic-type IT. Necrosis (bottom left) is a common
carcinoma, and is composed of a population of relatively feature of ETT
and are widely spaced in a hyalinized stroma In contrast to choriocarcinoma, ETT does not
(Figs. 4.32 and 4.33). The cells of a PSN show a have a trimorphic pattern with interspersed ST
low Ki-67 proliferation index of 3–10% [182] cells. Furthermore, ETT is not as hemorrhagic as
that contrasts with the higher proliferation index choriocarcinoma. Immunohistochemically, ETT
that is typically seen in ETT (>12%) [176]. shows only random β-hCG reactivity, in contrast
The nodular growth pattern with the distinctive to the abundant staining of ST cells in choriocar-
hyaline matrix of the ETT contrasts with the dif- cinoma [164].
fuse, infiltrative pattern of PSTT. In addition, the ETT also can resemble a nontrophoblastic
cells in ETT are smaller than the IT cells of the tumor, especially squamous cell carcinoma of the
PSTT and tend to grow in nests and cords. cervix, because of their epithelioid appearance,
Immunohistochemistry can be useful in distin- resemblance of the hyaline and necrotic debris to
guishing ETT from PSTT as the antibody against keratin, and propensity for the cells of ETT to
p63 reacts with chorionic-type IT that comprises grow along the surface of the cervix. In addition,
ETT but not with implantation type IT of PSTT about 50% of these tumors present in the cervix.
[173]. Typically, ETT has only limited reactivity Immunostains for inhibin-α and cytokeratin 18
for hPL and Mel-CAM, in contrast to the PSTT and CK5/6 can be useful, as these are normally
which is diffusely positive for hPL and Mel-CAM positive in ETT and usually negative in squa-
and does not express p63 [149, 173, 174] (Fig. 4.28 mous cell carcinoma (Table 4.5) [149, 175, 182].
Trophogram). Furthermore, the Ki-67 proliferative index is
Fig. 4.31 Epithelioid trophoblastic tumor. In this field, the tumor is composed of cords and nests of monotonous cells
in a hyaline matrix arranged in a perivascular fashion. Necrosis is a common feature of ETT
lower in ETT (10–25%) compared to squamous PSTT. In CHM, the placental site is typically
cell carcinomas, which have Ki-67 labeling indi- exaggerated, but exaggerated placental site can
ces of greater than 50%. Immunohistochemical occur in association with a normal gestation as
staining for p16 is positive in cervical carcinoma well. The exaggerated placental site is character-
and in one study none of the ETTs examined ized by an increase in the number and size of
stained for p16 [183]. However, in another study individual IT cells. In addition, widely dispersed
about 50% of intermediate trophoblastic tumors multinucleated IT cells are a component of the
also stained for p16 making this marker less help- trophoblastic infiltrate. Often several fragments
ful for distinguishing an intermediate trophoblas- of tissue in curettage samples contain portions of
tic tumor from a cervical carcinoma [149]. the lesion, and this process can extensively infil-
trate fragments of myometrium. A few chorionic
villi may be present. In the exaggerated placental
Nonneoplastic Lesions site, IT cells appear larger and more hyperchro-
matic than normal. Despite their apparent promi-
Exaggerated Placental Site nence, these IT cells show no mitotic activity, and
Ki-67 immunostaining is zero when the exagger-
An exaggerated placental site represents one end ated implantation site is associated with an abor-
of the morphologic spectrum of the normal tus [172]. As mentioned above, the Ki-67
implantation site [24, 48]. It is not a tumor. It is proliferation index can be slightly elevated in
an unusually prominent but physiologic placental exaggerated implantation site associated with a
site that may be difficult to distinguish from a CHM but is often zero (Figs. 4.7 and 4.8).
Nonneoplastic Lesions 111
Fig. 4.32 Placental site nodule. Low-power magnification of a PSN (center) found in endometrial curettings. The
chorionic-type IT cells are bland and widely spaced in a hyalinized stroma
Necrosis is not a feature of the exaggerated pla- abnormal uterine bleeding. The nodules may be
cental site, although the surrounding decidua present in biopsies taken several years after tubal
often shows degeneration and necrosis typical of ligation, suggesting that they are retained in the
spontaneous abortions. A PSTT is an important endometrium for extended periods of time [182,
consideration in the differential diagnosis of this 184, 186]. The known antecedent pregnancy dates
lesion (see PSTT above). back 2–108 months, demonstrating the long dura-
tion and benign course of the lesions. They show a
propensity for the lower uterine segment and cervix.
Placental Site Nodule and Plaque In the latter location, when large and atypical, the
differential diagnosis includes an ETT (see above) or
Placental site nodules (PSN) are focal abnormalities, a cervical squamous cell carcinoma.
usually associated with proliferative or secretory Generally, these lesions are microscopic,
endometrium elsewhere in the sections [164]. It is although hysterectomies may yield gross lesions
composed of small, circumscribed foci of hyalinized as large as 1 or 2 cm in diameter. Occasionally,
extracellular matrix with chorionic-type IT cells that multiple nodules are present, and they may form
occasionally present in an endometrial biopsy or elongated plaques rather than rounded nodules.
curettage (Figs. 4.32 and 4.33) [24, 48, 164, 182, In the past these nodules and plaques were con-
184, 185]. These benign lesions occur in women of sidered hyalinized decidua, but they are now rec-
reproductive age, although often the pregnancy his- ognized as a distinctive benign lesion of
tory is remote [184, 186]. Usually they are incidental chorionic-type IT. The lesion itself is circum-
findings, although they may be associated with scribed, nodular, or plaque-like with densely
Fig. 4.33 Placental site nodule. High-power magnification of a PSN demonstrates hyalinized extracellular matrix with
chorionic-type IT cells haphazardly distributed. A benign endometrial gland is present (top left)
eosinophilic, hyalinized stroma containing aggre- separate them from the PSTT and ETT (see
gates of IT cells. The cells of a placental site above respective sections).
nodule vary in size; many have small, uniform Occasionally, cytologic atypia and prolifera-
nuclei and some larger cells show irregular, tion exceed what is acceptable for a PSN but fall
hyperchromatic nuclei. Most cells are mononu- short of the diagnostic criteria for an ETT. These
cleated; however, occasional multinucleated cells lesions have been referred to as atypical PSNs.
are present. Often focal chronic inflammation They have not been well characterized and their
including plasma cells surrounds the nodule, clinical significance is unknown [3].
while the rest of the endometrium shows no
inflammation.
Similar to other IT lesions, the trophoblastic linical Queries and Reporting
C
cells of PSN are strongly reactive for keratin of Trophoblastic Neoplasms
and epithelial membrane antigen (EMA) as well
as PLAP and inhibin-α [184, 186, 187]. Other In curettings, choriocarcinoma is most com-
trophoblastic markers such as HSD3B1 and monly found in a repeat curettage during follow-
HLA-G are positive, while hPL and Mel-CAM up of a hydatidiform mole. Sometimes, however,
(CD 146) may be positive but only in a few this neoplasm is an unsuspected finding in a
cells. Reactivity for hCG is usually absent. reproductive-age patient with abnormal uterine
From a morphologic standpoint, the small size, bleeding. PSTT, because of its rarity, often is not
circumscription, and extensive hyalinization are clinically suspected, and the patient presents with
consistent features of these lesions that help to amenorrhea or an apparent missed abortion.
References 113
Obvious cases of either choriocarcinoma or miological and morphological study. Placenta.

1985;6:93–105.
PSTT do not require more than a concise diagno- 5. Hertig A, Mansell H. Hydatidiform mole and cho-
sis. Because of the rarity, a diagnosis of PSTT or riocarcinoma. Tumors of the female sex organs
ETT often is most useful if accompanied by an Part I Atlas of Tumor Pathology, section 9, Fascicle
explanatory comment that describes this as a 33. Washington, DC: Armed Forces Institute of
Pathology; 1956.
form of GTN derived from intermediate tropho- 6. Hayashi K, Bracken MB, Freeman DH, Hellenbrand
blast. Furthermore, for PSTT the cytoplasmic K. Hydatidiform mole in the United States (1970–
features and mitotic count should be stated, as 1977): a statistical and theoretical analysis. Am J
these features may help to predict aggressive Epidemiol. 1982;115:67–77.
7. Atrash HK, Hogue CJ, Grimes DA. Epidemiology
behavior. All PSTTs should be considered poten- of hydatidiform mole during early gestation. Am J
tially malignant, even those with a low mitotic Obstet Gynecol. 1986;154:906–9.
index. Because of the rarity and the potential for 8. Savage PM, Sita-Lumsden A, Dickson S, Iyer
aggressive growth of any trophoblastic tumor, but R, Everard J, Coleman R, Fisher RA, Short D,
Casalboni S, Catalano K, Seckl MJ. The relationship
especially choriocarcinoma, we also recommend of maternal age to molar pregnancy incidence, risks
verbal communication with the gynecologist for chemotherapy and subsequent pregnancy out-
whenever possible. come. J Obstet Gynaecol. 2013;33:406–11.
At times a specific diagnosis of a trophoblastic 9. Jeffers MD, O’Dwyer P, Curran B, Leader M, Gillan
JE. Partial hydatidiform mole: a common but under-
tumor may not be possible, yet the lesion is suspi- diagnosed condition. A 3-year retrospective clinico-
cious for neoplasia. The term “unclassified” tro- pathological and DNA flow cytometric analysis. Int
phoblastic lesion is reserved for those unusual J Gynecol Pathol. 1993;12:315–23.
cases that cannot clearly be placed in one of the 10. Bracken MB, Brinton LA, Hayashi K. Epidemiology
of hydatidiform mole and choriocarcinoma.
defined subgroups of the disease. An example of Epidemiol Rev. 1984;6:52–75.
such a case would be a small amount of prolifera- 11. Grimes DA. Epidemiology of gestational trophoblas-
tive trophoblast without villi, which can lead to a tic disease. Am J Obstet Gynecol. 1984;150:309–18.
difficult differential diagnosis of choriocarci- 12. McCorriston CC. Racial incidence of hydatidiform
mole. A study in a contained polyracial community.
noma versus trophoblast of a normal pregnancy. Am J Obstet Gynecol. 1968;101:377–82.
Also, prominent IT within decidua and myome- 13. Matsuura J, Chiu D, Jacobs PA, Szulman
trium may lead to a differential diagnosis of AE. Complete hydatidiform mole in Hawaii:
exaggerated placental site versus PSTT. When an epidemiological study. Genet Epidemiol.
1984;1:271–84.
the diagnosis is not straightforward, a descriptive 14. Chong CY, Koh CF. Hydatidiform mole in Kandang
diagnosis of “atypical trophoblast” is best. With Kerbau Hospital – a 5-year review. Singap Med J.
this type of diagnosis, the clinician is alerted to 1999;40:265–70.
the possibility of trophoblastic disease. Then the 15. Drake RD, Rao GG, McIntire DD, Miller DS,
Schorge JO. Gestational trophoblastic disease
patient can be followed with hCG titers or rebi- among Hispanic women: a 21-year hospital-based
opsied if symptoms persist. study. Gynecol Oncol. 2006;103:81–6.
16. Smith HO, Qualls CR, Hilgers RD, Verschraegen
CF, Rayburn WF, Cole LW, Padilla LA, Key
CR. Gestational trophoblastic neoplasia in American
References Indians. J Reprod Med. 2004;49:535–44.
17. Smith HO, Hilgers RD, Bedrick EJ, Qualls CR,
1. Kurman RJ, Carcangiu ML, Herrington CS, Young Wiggins CL, Rayburn WF, Waxman AG, Stephens
RH, editors. WHO classification of tumours of female ND, Cole LW, Swanson M, Key CR. Ethnic differ-
reproductive organs (World Health Organization clas- ences at risk for gestational trophoblastic disease in
sification of tumours). 4th ed. Geneva: World Health New Mexico: a 25-year population-based study. Am
Organization; 2014. J Obstet Gynecol. 2003;188:357–66.
2. Shih I-M, Mazur M, Kurman R. Gestational tro- 18. Melamed A, Gockley AA, Joseph NT, Sun SY,
phoblastic tumors and related tumor-like lesions. Clapp MA, Goldstein DP, Berkowitz RS, Horowitz
Blaustein’s pathology of the female genital tract. 6th NS. Effect of race/ethnicity on risk of complete and
ed. New York: Springer; 2011. p. 1075–135. partial molar pregnancy after adjustment for age.
3. Craighill MC, Cramer DW. Epidemiology of complete Gynecol Oncol. 2016;143:73–6.
molar pregnancy. J Reprod Med. 1984;29:784–7. 19. Yen S, MacMahon B. Epidemiologic features
4. Womack C, Elston CW. Hydatidiform mole of trophoblastic disease. Am J Obstet Gynecol.
in Nottingham: a 12-year retrospective epide- 1968;101:126–32.
20. Czernobilsky B, Barash A, Lancet M. Partial Disease Center over the past three decades: does
moles: a clinicopathologic study of 25 cases. Obstet early diagnosis alter risk for gestational trophoblas-
Gynecol. 1982;59:75–7. tic neoplasia? Gynecol Oncol. 2015;138:46–9.
21. Szulman AE, Surti U. The clinicopathologic profile 35. Jauniaux E, Memtsa M, Johns J, Ross JA, Jurkovic
of the partial hydatidiform mole. Obstet Gynecol. D. New insights in the pathophysiology of complete
1982;59:597–602. hydatidiform mole. Placenta. 2018;62:28–33.
22. Szulman AE, Surti U. The syndromes of hydatidi- 36. Joneborg U, Marions L. Current clinical features of
form mole. I. Cytogenetic and morphologic correla- complete and partial hydatidiform mole in Sweden.
tions. Am J Obstet Gynecol. 1978;131:665–71. J Reprod Med. 2014;59:51–5.
23. Szulman AE, Surti U. The syndromes of hydatidiform 37. Felemban AA, Bakri YN, Alkharif HA, Altuwaijri
mole. II. Morphologic evolution of the complete and SM, Shalhoub J, Berkowitz RS. Complete molar
partial mole. Am J Obstet Gynecol. 1978;132:20–7. pregnancy. Clinical trends at King Fahad Hospital,
24. Silverberg S, Kurman R. Tumors of the uterine Riyadh, Kingdom of Saudi Arabia. J Reprod Med.
corpus and gestational trophoblastic disease. Atlas 1998;43:11–3.
of tumor pathology. 3rd series. Washington, DC: 38. Ben-Arie A, Deutsch H, Volach V, Peer G, Husar M,
Armed Forces Institute of Pathology; 1992. Lavie O, Gemer O. Reduction of postmolar gesta-
25. Gockley AA, Melamed A, Joseph NT, Clapp M, Sun tional trophoblastic neoplasia by early diagnosis and
SY, Goldstein DP, Horowitz NS, Berkowitz RS. The treatment. J Reprod Med. 2009;54:151–4.
effect of adolescence and advanced maternal age on 39. Romero R, Horgan JG, Kohorn EI, Kadar N, Taylor
the incidence of complete and partial molar preg- KJ, Hobbins JC. New criteria for the diagnosis of
nancy. Gynecol Oncol. 2016;140:470–3. gestational trophoblastic disease. Obstet Gynecol.
26. Sand PK, Lurain JR, Brewer JI. Repeat gestational tro- 1985;66:553–8.
phoblastic disease. Obstet Gynecol. 1984;63:140–4. 40. Qiao S, Nagasaka T, Nakashima N. Numerous
27. Parazzini F, La Vecchia C, Franceschi S, Mangili vessels detected by CD34 in the villous stroma of
G. Familial trophoblastic disease: case report. Am J complete hydatidiform moles. Int J Gynecol Pathol.
Obstet Gynecol. 1984;149:382–3. 1997;16:233–8.
28. Kou YC, Shao L, Peng HH, Rosetta R, del Gaudio D, 41. Conran RM, Hitchcock CL, Popek EJ, Norris HJ,
Wagner AF, Al-Hussaini TK, Van den Veyver IB. A Griffin JL, Geissel A, McCarthy WF. Diagnostic
recurrent intragenic genomic duplication, other considerations in molar gestations. Hum Pathol.
novel mutations in NLRP7 and imprinting defects in 1993;24:41–8.
recurrent biparental hydatidiform moles. Mol Hum 42. Genest DR, Laborde O, Berkowitz RS, Goldstein
Reprod. 2008;14:33–40. DP, Bernstein MR, Lage J. A clinicopathologic
29. El-Maarri O, Seoud M, Coullin P, Herbiniaux U, study of 153 cases of complete hydatidiform mole
Oldenburg J, Rouleau G, Slim R. Maternal alleles (1980–1990): histologic grade lacks prognostic sig-
acquiring paternal methylation patterns in biparen- nificance. Obstet Gynecol. 1991;78:402–9.
tal complete hydatidiform moles. Hum Mol Genet. 43. Fukunaga M. Early partial hydatidiform mole: prev-
2003;12:1405–13. alence, histopathology, DNA ploidy, and persistence
30. Fallahian M, Sebire NJ, Savage PM, Seckl MJ, Fisher rate. Virchows Arch. 2000;437:180–4.
RA. Mutations in NLRP7 and KHDC3L confer a 44. Mosher R, Goldstein DP, Berkowitz R, Bernstein
complete hydatidiform mole phenotype on digynic M, Genest DR. Complete hydatidiform mole.
triploid conceptions. Hum Mutat. 2013;34:301–8. Comparison of clinicopathologic features, current
31. Keep D, Zaragoza MV, Hassold T, Redline RW. Very and past. J Reprod Med. 1998;43:21–7.
early complete hydatidiform mole. Hum Pathol. 45. Paradinas FJ. The diagnosis and prognosis of
1996;27:708–13. molar pregnancy: the experience of the National
32. Sebire NJ, Rees H, Paradinas F, Seckl M, Newlands Referral Centre in London. Int J Gynaecol Obstet.
E. The diagnostic implications of routine ultra- 1998;60(Suppl 1):S57–64.
sound examination in histologically confirmed early 46. Coukos G, Makrigiannakis A, Chung J, Randall
molar pregnancies. Ultrasound Obstet Gynecol. TC, Rubin SC, Benjamin I. Complete hydatidiform
2001;18:662–5. mole. A disease with a changing profile. J Reprod
33. Gockley AA, Joseph NT, Melamed A, Sun SY, Med. 1999;44:698–704.
Goodwin B, Bernstein M, Goldstein DP, Berkowitz 47. Paradinas FJ, Browne P, Fisher RA, Foskett M,
RS, Horowitz NS. Effect of race/ethnicity on clini- Bagshawe KD, Newlands E. A clinical, histopatho-
cal presentation and risk of gestational trophoblas- logical and flow cytometric study of 149 complete
tic neoplasia in patients with complete and partial moles, 146 partial moles and 107 non-molar hydropic
molar pregnancy at a tertiary care referral center. Am abortions. Histopathology. 1996;28:101–10.
J Obstet Gynecol. 2016;215:334.e1–6. 48. Kurman RJ. The morphology, biology, and pathol-
34. Sun SY, Melamed A, Goldstein DP, Bernstein ogy of intermediate trophoblast: a look back to the
MR, Horowitz NS, Moron AF, Maestá I, Braga A, present. Hum Pathol. 1991;22:847–55.
Berkowitz RS. Changing presentation of complete 49. Montes M, Roberts D, Berkowitz RS, Genest
hydatidiform mole at the New England Trophoblastic DR. Prevalence and significance of implantation
References 115
site trophoblastic atypia in hydatidiform moles cytometric reevaluation of nontriploid specimens. J

and spontaneous abortions. Am J Clin Pathol. Reprod Med. 2002;47:363–8.
1996;105:411–6. 67. Lawler SD, Fisher RA, Pickthall VJ, Povey S, Evans
50. Genest DR. Partial hydatidiform mole: clinicopatho- MW. Genetic studies on hydatidiform moles. I. The
logical features, differential diagnosis, ploidy and origin of partial moles. Cancer Genet Cytogenet.
molecular studies, and gold standards for diagnosis. 1982;5:309–20.
Int J Gynecol Pathol. 2001;20:315–22. 68. Hemming JD, Quirke P, Womack C, Wells M, Elston
51. Buza N, Hui P. Partial hydatidiform mole: histologic CW, Bird CC. Diagnosis of molar pregnancy and
parameters in correlation with DNA genotyping. Int persistent trophoblastic disease by flow cytometry. J
J Gynecol Pathol. 2013;32:307–15. Clin Pathol. 1987;40:615–20.
52. Iyengar V, Robers D, Lage J. Histopathologic study 69. Davis JR, Kerrigan DP, Way DL, Weiner
of triploid partial moles. Mod Pathol. 1993;6:74A. SA. Partial hydatidiform moles: deoxyribonucleic
53. Szulman AE. Syndromes of hydatidiform moles. acid content and course. Am J Obstet Gynecol.
Partial vs. complete. J Reprod Med. 1984;29:788–91. 1987;157:969–73.
54. Jacobs PA, Wilson CM, Sprenkle JA, Rosenshein 70. Lage JM, Berkowitz RS, Rice LW, Goldstein DP,
NB, Migeon BR. Mechanism of origin of complete Bernstein MR, Weinberg DS. Flow cytometric anal-
hydatidiform moles. Nature. 1980;286:714–6. ysis of DNA content in partial hydatidiform moles
55. Banet N, DeScipio C, Murphy KM, Beierl K, Adams with persistent gestational trophoblastic tumor.
E, Vang R, Ronnett BM. Characteristics of hydatidi- Obstet Gynecol. 1991;77:111–5.
form moles: analysis of a prospective series with p57 71. Lage JM, Mark SD, Roberts DJ, Goldstein DP,
immunohistochemistry and molecular genotyping. Bernstein MR, Berkowitz RS. A flow cytometric
Mod Pathol. 2014;27:238–54. study of 137 fresh hydropic placentas: correlation
56. Ohama K, Okamoto E, Nomura K, Fujiwara A, between types of hydatidiform moles and nuclear
Fukuda Y. Genetic studies of hydatidiform mole with DNA ploidy. Obstet Gynecol. 1992;79:403–10.
46,XY karyotype (author’s transl). Nippon Sanka 72. Lawler SD, Fisher RA, Dent J. A prospective genetic
Fujinka Gakkai Zasshi. 1981;33:1664–8. study of complete and partial hydatidiform moles.
57. Kajii T, Ohama K. Androgenetic origin of hydatidi- Am J Obstet Gynecol. 1991;164:1270–7.
form mole. Nature. 1977;268:633–4. 73. Fisher RA, Tommasi A, Short D, Kaur B, Seckl MJ,
58. Wake N, Takagi N, Sasaki M. Androgenesis as a Sebire NJ. Clinical utility of selective molecular
cause of hydatidiform mole. J Natl Cancer Inst. genotyping for diagnosis of partial hydatidiform
1978;60:51–7. mole; a retrospective study from a regional tropho-
59. Kajii T, Kurashige H, Ohama K, Uchino F. XY and blastic disease unit. J Clin Pathol. 2014;67:980–4.
XX complete moles: clinical and morphologic cor- 74. Murphy KM, Descipio C, Wagenfuehr J, Tandy S,
relations. Am J Obstet Gynecol. 1984;150:57–64. Mabray J, Beierl K, Micetich K, Libby AL, Ronnett
60. Vassilakos P, Riotton G, Kajii T. Hydatidiform BM. Tetraploid partial hydatidiform mole: a case
mole: two entities. A morphologic and cytogenetic report and review of the literature. Int J Gynecol
study with some clinical consideration. Am J Obstet Pathol. 2012;31:73–9.
Gynecol. 1977;127:167–70. 75. Surti U, Szulman AE, Wagner K, Leppert M,
61. Surti U, Szulman AE, O’Brien S. Dispermic origin O’Brien SJ. Tetraploid partial hydatidiform
and clinical outcome of three complete hydatidiform moles: two cases with a triple paternal contri-
moles with 46,XY karyotype. Am J Obstet Gynecol. bution and a 92,XXXY karyotype. Hum Genet.
1982;144:84–7. 1986;72:15–21.
62. Surti U, Szulman A, Buschsbaum H. Genetic con- 76. Lage JM, Weinberg DS, Yavner DL, Bieber FR. The
cepts and techniques. Gestational trophoblastic dis- biology of tetraploid hydatidiform moles: histopa-
ease. New York: Springer-Verlag; 1987. p. 111–21. thology, cytogenetics, and flow cytometry. Hum
63. Jacobs PA, Szulman AE, Funkhouser J, Matsuura JS, Pathol. 1989;20:419–25.
Wilson CC. Human triploidy: relationship between 77. Fukunaga M, Ushigome S, Fukunaga M, Sugishita
parental origin of the additional haploid complement M. Application of flow cytometry in diagnosis of
and development of partial hydatidiform mole. Ann hydatidiform moles. Mod Pathol. 1993;6:353–9.
Hum Genet. 1982;46:223–31. 78. Genest DR, Dorfman DM, Castrillon DH. Ploidy and
64. Redline RW, Hassold T, Zaragoza MV. Prevalence of imprinting in hydatidiform moles. Complementary
the partial molar phenotype in triploidy of maternal use of flow cytometry and immunohistochemistry of
and paternal origin. Hum Pathol. 1998;29:505–11. the imprinted gene product p57KIP2 to assist molar
65. Bifulco C, Johnson C, Hao L, Kermalli H, Bell S, classification. J Reprod Med. 2002;47:342–6.
Hui P. Genotypic analysis of hydatidiform mole: an 79. Castrillon DH, Sun D, Weremowicz S, Fisher RA,
accurate and practical method of diagnosis. Am J Crum CP, Genest DR. Discrimination of complete
Surg Pathol. 2008;32:445–51. hydatidiform mole from its mimics by immu-
66. Genest DR, Ruiz RE, Weremowicz S, Berkowitz RS, nohistochemistry of the paternally imprinted
Goldstein DP, Dorfman DM. Do nontriploid partial gene product p57KIP2. Am J Surg Pathol.
hydatidiform moles exist? A histologic and flow 2001;25:1225–30.
80. Fukunaga M. Immunohistochemical characteriza- maternally transcribed gene p57(KIP2) (CDNK1C)

tion of p57(KIP2) expression in early hydatidiform is abnormally expressed in both androgenetic and
moles. Hum Pathol. 2002;33:1188–92. biparental complete hydatidiform moles. Hum Mol
81. Hoffner L, Parks WT, Swerdlow SH, Carson JC, Genet. 2002;11:3267–72.
Surti U. Simultaneous detection of imprinted gene 92. Steller MA, Genest DR, Bernstein MR, Lage JM,
expression (p57(KIP2)) and molecular cytogenetics Goldstein DP, Berkowitz RS. Clinical features of
(FICTION) in the evaluation of molar pregnancies. J multiple conception with partial or complete molar
Reprod Med. 2010;55:219–28. pregnancy and coexisting fetuses. J Reprod Med.
82. Hoffner L, Dunn J, Esposito N, Macpherson T, 1994;39:147–54.
Surti U. P57KIP2 immunostaining and molecu- 93. Steller MA, Genest DR, Bernstein MR, Lage JM,
lar cytogenetics: combined approach aids in diag- Goldstein DP, Berkowitz RS. Natural history of twin
nosis of morphologically challenging cases with pregnancy with complete hydatidiform mole and
molar phenotype and in detecting androgenetic cell coexisting fetus. Obstet Gynecol. 1994;83:35–42.
lines in mosaic/chimeric conceptions. Hum Pathol. 94. Choi-Hong SR, Genest DR, Crum CP, Berkowitz R,
2008;39:63–72. Goldstein DP, Schofield DE. Twin pregnancies with
83. Chilosi M, Piazzola E, Lestani M, Benedetti A, complete hydatidiform mole and coexisting fetus:
Guasparri I, Granchelli G, Aldovini D, Leonardi use of fluorescent in situ hybridization to evalu-
E, Pizzolo G, Doglioni C, Menestrina F, Mariuzzi ate placental X- and Y-chromosomal content. Hum
GM. Differential expression of p57kip2, a mater- Pathol. 1995;26:1175–80.
nally imprinted cdk inhibitor, in normal human 95. Ronnett BM, DeScipio C, Murphy KM. Hydatidiform
placenta and gestational trophoblastic disease. Lab moles: ancillary techniques to refine diagnosis. Int J
Investig. 1998;78:269–76. Gynecol Pathol. 2011;30:101–16.
84. Fukunaga M. Immunohistochemical characteriza- 96. Lewis GH, DeScipio C, Murphy KM, Haley L,
tion of p57Kip2 expression in tetraploid hydropic Beierl K, Mosier S, Tandy S, Cohen DS, Lytwyn
placentas. Arch Pathol Lab Med. 2004;128:897–900. A, Elit L, Vang R, Ronnett BM. Characterization
85. Jun S-Y, Ro JY, Kim K-R. p57kip2 is useful in the of androgenetic/biparental mosaic/chimeric con-
classification and differential diagnosis of complete ceptions, including those with a molar compo-
and partial hydatidiform moles. Histopathology. nent: morphology, p57 immunohistochemistry,
2003;43:17–25. molecular genotyping, and risk of persistent gesta-
86. McConnell TG, Murphy KM, Hafez M, Vang R, tional trophoblastic disease. Int J Gynecol Pathol.
Ronnett BM. Diagnosis and subclassification of 2013;32:199–214.
hydatidiform moles using p57 immunohistochem- 97. Hui P, Buza N, Murphy KM, Ronnett
istry and molecular genotyping: validation and BM. Hydatidiform moles: genetic basis and preci-
prospective analysis in routine and consultation sion diagnosis. Annu Rev Pathol. 2017;12:449–85.
practice settings with development of an algorithmic 98. Fisher RA, Nucci MR, Thaker HM, Weremowicz S,
approach. Am J Surg Pathol. 2009;33:805–17. Genest DR, Castrillon DH. Complete hydatidiform
87. Merchant SH, Amin MB, Viswanatha DS, Malhotra mole retaining a chromosome 11 of maternal origin:
RK, Moehlenkamp C, Joste NE. p57KIP2 immu- molecular genetic analysis of a case. Mod Pathol.
nohistochemistry in early molar pregnancies: 2004;17:1155–60.
emphasis on its complementary role in the differ- 99. McConnell TG, Norris-Kirby A, Hagenkord JM,
ential diagnosis of hydropic abortuses. Hum Pathol. Ronnett BM, Murphy KM. Complete hydatidiform
2005;36:180–6. mole with retained maternal chromosomes 6 and 11.
88. Popiolek DA, Yee H, Mittal K, Chiriboga L, Prinz Am J Surg Pathol. 2009;33:1409–15.
MK, Caragine TA, Budimlija ZM. Multiplex short 100. DeScipio C, Haley L, Beierl K, Pandit AP, Murphy
tandem repeat DNA analysis confirms the accu- KM, Ronnett BM. Diandric triploid hydatidiform
racy of p57(KIP2) immunostaining in the diagno- mole with loss of maternal chromosome 11. Am J
sis of complete hydatidiform mole. Hum Pathol. Surg Pathol. 2011;35:1586–91.
2006;37:1426–34. 101. Nguyen NMP, Zhang L, Reddy R, Déry C, Arseneau
89. Romaguera RL, Rodriguez MM, Bruce JH, Zuluaga J, Cheung A, Surti U, Hoffner L, Seoud M, Zaatari
T, Viciana A, Penalver MA, Mehrdad N. Molar G, Bagga R, Srinivasan R, Coullin P, Ao A, Slim
gestations and hydropic abortions differentiated R. Comprehensive genotype-phenotype correlations
by p57 immunostaining. Fetal Pediatr Pathol. between NLRP7 mutations and the balance between
2004;23:181–90. embryonic tissue differentiation and trophoblastic
90. Sarmadi S, Izadi-Mood N, Abbasi A, Sanii S. proliferation. J Med Genet. 2014;51:623–34.
p57KIP2 immunohistochemical expression: a useful 102. Martin DA, Sutton GP, Ulbright TM, Sledge GW,
diagnostic tool in discrimination between complete Stehman FB, Ehrlich CE. DNA content as a prog-
hydatidiform mole and its mimics. Arch Gynecol nostic index in gestational trophoblastic neoplasia.
Obstet. 2011;283:743–8. Gynecol Oncol. 1989;34:383–8.
91. Fisher RA, Hodges MD, Rees HC, Sebire NJ, Seckl 103. Fukunaga M. Flow cytometric and clinicopathologic
MJ, Newlands ES, Genest DR, Castrillon DH. The study of complete hydatidiform moles with special
References 117
reference to the significance of cytometric aneu- 116. Berkowitz RS, Birnholz J, Goldstein DP, Bernstein
ploidy. Gynecol Oncol. 2001;81:67–70. MR. Pelvic ultrasonography and the management
104. Messerli ML, Parmley T, Woodruff JD, Lilienfeld of gestational trophoblastic disease. Gynecol Oncol.
AM, Bevilacqua L, Rosenshein NB. Inter- and 1983;15:403–12.
intra-pathologist variability in the diagnosis of ges- 117. Elston CW, Bagshawe KD. The diagnosis of tro-
tational trophoblastic neoplasia. Obstet Gynecol. phoblastic tumours from uterine curettings. J Clin
1987;69:622–6. Pathol. 1972;25:111–8.
105. Javey H, Borazjani G, Behmard S, Langley 118. Takeuchi S. Nature of invasive mole and its rational
FA. Discrepancies in the histological diagno- management. Semin Oncol. 1982;9:181–6.
sis of hydatidiform mole. Br J Obstet Gynaecol. 119. Berkowitz RS, Goldstein DP. Current advances in
1979;86:480–3. the management of gestational trophoblastic disease.
106. Koenig C, Demopoulos RI, Vamvakas EC, Mittal Gynecol Oncol. 2013;128:3–5.
KR, Feiner HD, Espiritu EC. Flow cytometric DNA 120. Braga A, Maestá I, Matos M, Elias KM, Rizzo J,
ploidy and quantitative histopathology in partial Viggiano MGC. Gestational trophoblastic neoplasia
moles. Int J Gynecol Pathol. 1993;12:235–40. after spontaneous human chorionic gonadotropin
107. Gupta M, Vang R, Yemelyanova AV, Kurman RJ, normalization following molar pregnancy evacua-
Li FR, Maambo EC, Murphy KM, DeScipio C, tion. Gynecol Oncol. 2015;139:283–7.
Thompson CB, Ronnett BM. Diagnostic repro- 121. Savage J, Adams E, Veras E, Murphy KM, Ronnett
ducibility of hydatidiform moles: ancillary tech- BM. Choriocarcinoma in women: analysis of a
niques (p57 immunohistochemistry and molecular case series with genotyping. Am J Surg Pathol.
genotyping) improve morphologic diagnosis for 2017;41:1593–606.
both recently trained and experienced gynecologic 122. Curry SL, Hammond CB, Tyrey L, Creasman WT,
pathologists. Am J Surg Pathol. 2012;36:1747–60. Parker RT. Hydatidiform mole: diagnosis, man-
108. Vang R, Gupta M, Wu L-S-F, Yemelyanova AV, agement, and long-term followup of 347 patients.
Kurman RJ, Murphy KM, Descipio C, Ronnett Obstet Gynecol. 1975;45:1–8.
BM. Diagnostic reproducibility of hydatidiform 123. Lurain JR, Brewer JI, Torok EE, Halpern B. Natural
moles: ancillary techniques (p57 immunohistochem- history of hydatidiform mole after primary evacua-
istry and molecular genotyping) improve morpho- tion. Am J Obstet Gynecol. 1983;145:591–5.
logic diagnosis. Am J Surg Pathol. 2012;36:443–53. 124. Bagshawe KD, Wilson H, Dublon P, Smith A,
109. Bell KA, Van Deerlin V, Addya K, Clevenger CV, Baldwin M, Kardana A. Follow-up after hydatidi-
Van Deerlin PG, Leonard DG. Molecular genetic form mole: studies using radioimmunoassay for
testing from paraffin-embedded tissue distinguishes urinary human chorionic gonadotrophin (HCG). J
nonmolar hydropic abortion from hydatidiform Obstet Gynaecol Br Commonw. 1973;80:461–8.
mole. Mol Diagn. 1999;4:11–9. 125. Sebire NJ, Fisher RA, Foskett M, Rees H, Seckl
110. Lai CYL, Chan KYK, Khoo U-S, Ngan HYS, Xue MJ, Newlands ES. Risk of recurrent hydatidiform
W-C, Chiu PM, Tsao S-W, Cheung ANY. Analysis mole and subsequent pregnancy outcome following
of gestational trophoblastic disease by genotyping complete or partial hydatidiform molar pregnancy.
and chromosome in situ hybridization. Mod Pathol. BJOG. 2003;110:22–6.
2004;17:40–8. 126. Wielsma S, Wiesma S, Kerkmeijer L, Bekkers R,
111. Lipata F, Parkash V, Talmor M, Bell S, Chen S, Maric Pyman J, Tan J, Quinn M. Persistent trophoblast dis-
V, Hui P. Precise DNA genotyping diagnosis of hyda- ease following partial molar pregnancy. Aust N Z J
tidiform mole. Obstet Gynecol. 2010;115:784–94. Obstet Gynaecol. 2006;46:119–23.
112. Paradinas FJ, Sebire NJ, Fisher RA, Rees HC, 127. Eagles N, Sebire NJ, Short D, Savage PM, Seckl
Foskett M, Seckl MJ, Newlands ES. Pseudo-partial MJ, Fisher RA. Risk of recurrent molar pregnancies
moles: placental stem vessel hydrops and the asso- following complete and partial hydatidiform moles.
ciation with Beckwith-Wiedemann syndrome and Hum Reprod. 2016;31:1379.
complete moles. Histopathology. 2001;39:447–54. 128. Hancock BW, Nazir K, Everard JE. Persistent gesta-
113. Sander CM. Angiomatous malformation of pla- tional trophoblastic neoplasia after partial hydatidi-
cental chorionic stem vessels and pseudo-partial form mole incidence and outcome. J Reprod Med.
molar placentas: report of five cases. Pediatr Pathol. 2006;51:764–6.
1993;13:621–33. 129. Feltmate CM, Growdon WB, Wolfberg AJ, Goldstein
114. Schlaerth JB, Morrow CP, Rodriguez M. Diagnostic DP, Genest DR, Chinchilla ME, Lieberman ES,
and therapeutic curettage in gestational trophoblastic Berkowitz RS. Clinical characteristics of persis-
disease. Am J Obstet Gynecol. 1990;162:1465–70. tent gestational trophoblastic neoplasia after par-
discussion 1470–1471 tial hydatidiform molar pregnancy. J Reprod Med.
115. Berkowitz RS, Desai U, Goldstein DP, Driscoll 2006;51:902–6.
SG, Marean AR, Bernstein MR. Pretreatment 130. Cole LA. Phantom hCG and phantom choriocarci-
curettage-A predictor of chemotherapy response in noma. Gynecol Oncol. 1998;71:325–9.
gestational trophoblastic neoplasia. Gynecol Oncol. 131. Rotmensch S, Cole LA. False diagnosis and needless
1980;10:39–43. therapy of presumed malignant disease in women
with false-positive human chorionic gonadotropin 147. Mao T-L, Kurman RJ, Jeng Y-M, Huang W, Shih
concentrations. Lancet. 2000;355:712–5. I-M. HSD3B1 as a novel trophoblast-associated
132. Berglund L, Holmberg NG. Heterophilic antibodies marker that assists in the differential diagnosis of
against rabbit serum causing falsely elevated gonadotro- trophoblastic tumors and tumorlike lesions. Am J
pin levels. Acta Obstet Gynecol Scand. 1989;68:377–8. Surg Pathol. 2008;32:236–42.
133. Goto S, Yamada A, Ishizuka T, Tomoda 148. Mao T-L, Kurman RJ, Huang C-C, Lin M-C, Shih
Y. Development of postmolar trophoblastic dis- I-M. Immunohistochemistry of choriocarcinoma: an
ease after partial molar pregnancy. Gynecol Oncol. aid in differential diagnosis and in elucidating patho-
1993;48:165–70. genesis. Am J Surg Pathol. 2007;31:1726–32.
134. Berkowitz RS, Goldstein DP, Bernstein MR. Natural 149. Kalhor N, Ramirez PT, Deavers MT, Malpica A,
history of partial molar pregnancy. Obstet Gynecol. Silva EG. Immunohistochemical studies of tropho-
1985;66:677–81. blastic tumors. Am J Surg Pathol. 2009;33:633–8.
135. Zalel Y, Dgani R. Gestational trophoblastic disease 150. Singer G, Kurman RJ, McMaster MT, Shih
following the evacuation of partial hydatidiform I-M. HLA-G immunoreactivity is specific for inter-
mole: a review of 66 cases. Eur J Obstet Gynecol mediate trophoblast in gestational trophoblastic dis-
Reprod Biol. 1997;71:67–71. ease and can serve as a useful marker in differential
136. Seckl MJ, Fisher RA, Salerno G, Rees H, Paradinas diagnosis. Am J Surg Pathol. 2002;26:914–20.
FJ, Foskett M, Newlands ES. Choriocarcinoma and 151. Zanetta G, Maggi R, Colombo M, Bratina G,
partial hydatidiform moles. Lancet. 2000;356:36–9. Mangioni C. Choriocarcinoma coexistent with intra-
137. Medeiros F, Callahan MJ, Elvin JA, Dorfman DM, uterine pregnancy: two additional cases and a review
Berkowitz RS, Quade BJ. Intraplacental choriocar- of the literature. Int J Gynecol Cancer. 1997;7:66–77.
cinoma arising in a second trimester placenta with 152. Savage J, Subby W, Okagaki T. Adenocarcinoma of
partial hydatidiform mole. Int J Gynecol Pathol. the endometrium with trophoblastic differentiation
2008;27:247–51. and metastases as choriocarcinoma: a case report.
138. Ma N, Litkouhi B, Mannion CM. FIGO stage III Gynecol Oncol. 1987;26:257–62.
metastatic gestational choriocarcinoma developed 153. Pesce C, Merino MJ, Chambers JT, Nogales
from an antecedent partial Hydatidiform molar preg- F. Endometrial carcinoma with trophoblastic dif-
nancy bearing a numerical chromosomal aberration ferentiation. An aggressive form of uterine cancer.
68, XX: a case report and literature review. Int J Cancer. 1991;68:1799–802.
Gynecol Pathol. 2016;35:162–6. 154. Eckstein RP, Paradinas FJ, Bagshawe KD. Placental
139. Cheung ANY, Khoo US, Lai CYL, Chan KYK, site trophoblastic tumour (trophoblastic pseudo-
Xue W-C, Cheng DKL, Chiu P-M, Tsao S-W, Ngan tumour): a study of four cases requiring hyster-
HYS. Metastatic trophoblastic disease after an ini- ectomy including one fatal case. Histopathology.
tial diagnosis of partial hydatidiform mole: genotyp- 1982;6:211–26.
ing and chromosome in situ hybridization analysis. 155. Kurman RJ, Scully RE, Norris HJ. Trophoblastic
Cancer. 2004;100:1411–7. pseudotumor of the uterus: an exaggerated form
140. Menczer J, Girtler O, Zajdel L, Glezerman of “syncytial endometritis” simulating a malignant
M. Metastatic trophoblastic disease following partial tumor. Cancer. 1976;38:1214–26.
hydatidiform mole: case report and literature review. 156. Young RH, Scully RE. Placental-site trophoblas-
Gynecol Oncol. 1999;74:304–7. tic tumor: current status. Clin Obstet Gynecol.
141. Gardner HA, Lage JM. Choriocarcinoma follow- 1984;27:248–58.
ing a partial hydatidiform mole: a case report. Hum 157. Young RH, Kurman RJ, Scully RE. Proliferations
Pathol. 1992;23:468–71. and tumors of intermediate trophoblast of the pla-
142. Matsui H, Iizuka Y, Sekiya S. Incidence of invasive cental site. Semin Diagn Pathol. 1988;5:223–37.
mole and choriocarcinoma following partial hyda- 158. Bower M, Paradinas FJ, Fisher RA, Nicholson SK,
tidiform mole. Int J Gynaecol Obstet. 1996;53:63–4. Rustin GJ, Begent RH, Bagshawe KD, Newlands
143. Kohorn EI. The new FIGO 2000 Staging and risk ES. Placental site trophoblastic tumor: molecular
factor scoring system for gestational trophoblastic analysis and clinical experience. Clin Cancer Res.
disease: description and critical assessment. Int J 1996;2:897–902.
Gynecol Cancer. 2001;11:73–7. 159. Chang YL, Chang TC, Hsueh S, Huang KG, Wang
144. Ngan HYS, Kohorn EI, Cole LA, Kurman RJ, PN, Liu HP, Soong YK. Prognostic factors and treat-
Kim SJ, Lurain JR, Seckl MJ, Sasaki S, Soper ment for placental site trophoblastic tumor-report
JT. Trophoblastic disease. Int J Gynaecol Obstet. of 3 cases and analysis of 88 cases. Gynecol Oncol.
2012;119(Suppl 2):S130–6. 1999;73:216–22.
145. Palmer JR. Advances in the epidemiology of gestational 160. Feltmate CM, Genest DR, Goldstein DP, Berkowitz
trophoblastic disease. J Reprod Med. 1994;39:155–62. RS. Advances in the understanding of placental site
146. Hoffner L, Surti U. The genetics of gestational tro- trophoblastic tumor. J Reprod Med. 2002;47:337–41.
phoblastic disease: a rare complication of pregnancy. 161. Papadopoulos AJ, Foskett M, Seckl MJ, McNeish
Cancer Genet. 2012;205:63–77. I, Paradinas FJ, Rees H, Newlands ES. Twenty-five
References 119
years’ clinical experience with placental site tropho- 174. Zhang H-J, Xue W-C, Siu MK-Y, Liao X-Y, Ngan
blastic tumors. J Reprod Med. 2002;47:460–4. HY-S, Cheung AN-Y. P63 expression in gestational
162. Wright JD, Powell MA, Horowitz NS, Huettner PC, trophoblastic disease: correlation with prolifera-
White F, Herzog TJ. Placental site trophoblastic tion and apoptotic dynamics. Int J Gynecol Pathol.
tumor presenting with a pneumothorax during preg- 2009;28:172–8.
nancy. Obstet Gynecol. 2002;100:1141–4. 175. Shih IM, Kurman RJ. Epithelioid trophoblastic
163. Oldt RJ, Kurman RJ, Shih I-M. Molecular genetic tumor: a neoplasm distinct from choriocarcinoma
analysis of placental site trophoblastic tumors and and placental site trophoblastic tumor simulating car-
epithelioid trophoblastic tumors confirms their tro- cinoma. Am J Surg Pathol. 1998;22:1393–403.
phoblastic origin. Am J Pathol. 2002;161:1033–7. 176. Hamazaki S, Nakamoto S, Okino T, Tsukayama C,
164. Shih IM, Kurman RJ. The pathology of intermedi- Mori M, Taguchi K, Okada S. Epithelioid tropho-
ate trophoblastic tumors and tumor-like lesions. Int J blastic tumor: morphological and immunohisto-
Gynecol Pathol. 2001;20:31–47. chemical study of three lung lesions. Hum Pathol.
165. Shih IM, Kurman RJ. Immunohistochemical local- 1999;30:1321–7.
ization of inhibin-alpha in the placenta and gesta- 177. Ohira S, Yamazaki T, Hatano H, Harada O, Toki
tional trophoblastic lesions. Int J Gynecol Pathol. T, Konishi I. Epithelioid trophoblastic tumor
1999;18:144–50. metastatic to the vagina: an immunohistochemi-
166. Kurman RJ, Main CS, Chen HC. Intermediate tro- cal and ultrastructural study. Int J Gynecol Pathol.
phoblast: a distinctive form of trophoblast with 2000;19:381–6.
specific morphological, biochemical and functional 178. Shen D-H, Khoo US, Ngan HYS, Ng T-Y, Chau
features. Placenta. 1984;5:349–69. M-T, Xue W-C, Cheung ANY. Coexisting epithe-
167. Kurman RJ, Young RH, Norris HJ, Main CS, lioid trophoblastic tumor and choriocarcinoma of
Lawrence WD, Scully RE. Immunocytochemical the uterus following a chemoresistant hydatidiform
localization of placental lactogen and chorionic mole. Arch Pathol Lab Med. 2003;127:e291–3.
gonadotropin in the normal placenta and trophoblas- 179. Mazur MT. Metastatic gestational choriocarcinoma.
tic tumors, with emphasis on intermediate tropho- Unusual pathologic variant following therapy.
blast and the placental site trophoblastic tumor. Int Cancer. 1989;63:1370–7.
J Gynecol Pathol. 1984;3:101–21. 180. Silva EG, Tornos C, Lage J, Ordonez NG, Morris
168. Baergen RN, Rutgers JL, Young RH, Osann K, M, Kavanagh J. Multiple nodules of intermedi-
Scully RE. Placental site trophoblastic tumor: a ate trophoblast following hydatidiform moles. Int J
study of 55 cases and review of the literature empha- Gynecol Pathol. 1993;12:324–32.
sizing factors of prognostic significance. Gynecol 181. Coulson LE, Kong CS, Zaloudek C. Epithelioid tro-
Oncol. 2006;100:511–20. phoblastic tumor of the uterus in a postmenopausal
169. Feltmate CM, Genest DR, Wise L, Bernstein MR, woman: a case report and review of the literature.
Goldstein DP, Berkowitz RS. Placental site tro- Am J Surg Pathol. 2000;24:1558–62.
phoblastic tumor: a 17-year experience at the New 182. Shih IM, Seidman JD, Kurman RJ. Placental
England Trophoblastic Disease Center. Gynecol site nodule and characterization of distinctive
Oncol. 2001;82:415–9. types of intermediate trophoblast. Hum Pathol.
170. Gloor E, Dialdas J, Hurlimann J, Ribolzi J, Barrelet 1999;30:687–94.
L. Placental site trophoblastic tumor (trophoblastic 183. Mao T-L, Seidman JD, Kurman RJ, Shih I-M. Cyclin
pseudotumor) of the uterus with metastases and fatal E and p16 immunoreactivity in epithelioid tropho-
outcome. Clinical and autopsy observations of a blastic tumor – an aid in differential diagnosis. Am J
case. Am J Surg Pathol. 1983;7:483–6. Surg Pathol. 2006;30:1105–10.
171. Fukunaga M, Ushigome S. Metastasizing placental 184. Young RH, Kurman RJ, Scully RE. Placental site
site trophoblastic tumor. An immunohistochemical nodules and plaques. A clinicopathologic analysis of
and flow cytometric study of two cases. Am J Surg 20 cases. Am J Surg Pathol. 1990;14:1001–9.
Pathol. 1993;17:1003–10. 185. Santos LD, Fernando SS, Yong JL, Killingsworth
172. Shih IM, Kurman RJ. Ki-67 labeling index in the MC, Wu XJ, Kennerson AR. Placental site nodules
differential diagnosis of exaggerated placental site, and plaques: a clinicopathological and immunohis-
placental site trophoblastic tumor, and choriocarci- tochemical study of 25 cases with ultrastructural
noma: a double immunohistochemical staining tech- findings. Pathology. 1999;31:328–36.
nique using Ki-67 and Mel-CAM antibodies. Hum 186. Huettner PC, Gersell DJ. Placental site nodule: a
Pathol. 1998;29:27–33. clinicopathologic study of 38 cases. Int J Gynecol
173. Shih IM, Kurman RJ. p63 expression is useful in the Pathol. 1994;13:191–8.
distinction of epithelioid trophoblastic and placental 187. Shitabata PK, Rutgers JL. The placental site nod-
site trophoblastic tumors by profiling trophoblastic ule: an immunohistochemical study. Hum Pathol.
subpopulations. Am J Surg Pathol. 2004;28:1177–83. 1994;25:1295–301.
Abnormal Uterine Bleeding
5
Contents
orphologic Features of Glandular and Stromal Breakdown in Menstrual
M
and Abnormal Bleeding 122
Abnormal Uterine Bleeding: Nonstructural Causes 126
Estrogen-Related Bleeding 129
Progesterone-Related Bleeding 135
References 141
This chapter specifically addresses uterine bleed- most common morphologic change associated
ing resulting from alterations in the normal cycli- with this type of bleeding is endometrial glandu-
cal changes of the endometrium that are not due lar and stromal breakdown. Glandular and stro-
to structural causes. It is termed “abnormal uter- mal breakdown can be found in a variety of
ine bleeding” (AUB); in the past AUB was organic disorders, as well. Conversely, not all
referred to as “dysfunctional uterine bleeding” biopsy specimens of patients with a history of
(DUB) [1–5]. Clinically, this type of bleeding is bleeding show evidence of breakdown.
addressed in the acronym COEIN (coagulopathy, Nonetheless, endometrial breakdown and bleed-
ovulatory dysfunction, endometrial, iatrogenic, ing is commonly encountered, and the morpho-
and not yet classified) (see Chap. 1). Abnormal logic features of bleeding need to be recognized
uterine bleeding, therefore, excludes postmeno- to allow clear separation of these nonspecific
pausal bleeding or bleeding caused by the pres- artifacts and degenerative/regenerative changes,
ence of specific pathologic processes such as from other, more specific histologic changes
inflammation, polyps, hyperplasia, carcinoma, associated with hyperplasia or carcinoma.
exogenous hormones, and complications of preg- Because glandular and stromal breakdown
nancy. It is important to recognize the endome- tends to be most extensive when associated with
trial changes associated with nonstructural AUB, AUB, the morphologic features of this form of
because they may be confused with more serious early tissue breakdown are first presented in
lesions such as hyperplasia or carcinoma. The detail.

122 5 Abnormal Uterine Bleeding
orphologic Features of Glandular

M Table 5.1 Features of glandular and stromal breakdown
in menstrual and abnormal bleeding
and Stromal Breakdown
in Menstrual and Abnormal Menstrual Abnormal bleeding
bleeding
Bleeding
Background Late Proliferative or
endometrium secretory secretory
Breakdown and bleeding patterns have been endometrium or
referred to by several different descriptive terms, structural lesion
such as “lytic,” “shedding,” “slough,” or even Heterogeneous − +
background
“menstrual” endometrium. The histologic changes
Stromal “collapse” + +
seen in abnormal endometrial bleeding are some-
Stromal cell +/− +
what different from the changes seen in menstrual clusters “stromal
endometrium, because unlike menstrual bleeding, blue balls”
glandular and stromal breakdown are usually asso- Fibrin thrombi +/− +
ciated with anovulatory cycles and therefore occur Nuclear debris at + +
base of gland cells
on a nonsecretory background.
Nuclear debris in − +
Glandular and stromal breakdown can also stroma
occur with structural lesions such as inflammation, Eosinophilic − +
polyps, hyperplasia, or carcinoma and usually syncytial change
does not uniformly involve the entire endome- Hemosiderin − +
trium. As a result, abnormal bleeding typically Foam cells − +
leads to a heterogeneous pattern with fragments of Stromal fibrosis − +
and hyalinization
intact, nonshedding endometrium admixed with
endometrium showing glandular and stromal
breakdown. Menstrual endometrium also shows Among the distinctive features of acute break-
breakdown, but the changes affect all the tissue down and bleeding is a particular pattern of hem-
and occur on a background of late secretory phase orrhage in the stroma. This process is characterized
glands (see Chap. 2). Furthermore, in menstrual by collapse of the stroma and coalescence of stro-
endometrium the breakdown is acute and lacks the mal cells into small aggregates and clusters sepa-
changes of chronic bleeding, such as hemosiderin rated by lakes of blood (Fig. 5.1). As stromal
deposition, eosinophilic syncytial change, or foam condensation becomes more advanced, small,
cell accumulation, seen in abnormal bleeding pat- rounded clusters of stromal cells, sometimes called
terns (Table 5.1). Accordingly, as AUB is not a “stromal blue balls,” become detached from the
normal physiologic change, it is important to surrounding tissue (Fig. 5.2). These cellular clus-
describe the morphologic changes as glandular ters are characterized by tightly packed aggregates
and stromal breakdown not menstrual endome- of cells with hyperchromatic small nuclei and
trium so that the gynecologist understands the scant cytoplasm mixed with karyorrhectic debris.
bleeding is not related to a normal ovulatory cycle. With loss of cytoplasm, the collapsed stromal cells
Two factors complicate recognition of break- in these aggregates can show nuclear molding.
down and bleeding in endometrial biopsies. First, They often are capped by attenuated surface epi-
the operative procedure itself causes tissue frag- thelium or by eosinophilic epithelial cells forming
mentation and hemorrhage. Second, when break- a syncytium (see later) (Fig. 5.3). Sometimes these
down does occur, the pattern of tissue necrosis is condensed aggregates of cells form small polypoid
unlike that seen in other organs. This unique mor- extrusions along the endometrial surface as they
phologic expression of necrosis is due to the detach from the intact stroma below. These are not
rapid expulsion of tissue into the uterine cavity. true polyps.
Consequently, extensive necrosis and autolysis Another characteristic feature of early break-
often do not occur, and the tissue shows only down, occurring before actual stromal collapse, is
early signs of degeneration. the accumulation of nuclear debris in the basal
Morphologic Features of Glandular and Stromal Breakdown in Menstrual and Abnormal Bleeding 123
Fig. 5.1 Glandular and stromal breakdown. Abnormal gated into rounded clusters and are surrounded by blood.
bleeding pattern associated with chronic inflammation The glands show poorly developed secretory changes in
secondary to retained placental site shows extensive glan- this specimen. Other fields showed residual placental site
dular and stromal breakdown. Stromal cells are aggre-
c ytoplasm of glandular cells. This feature is promi- Thrombi form and extend into the stroma. Fibrin
nent in premenstrual and menstrual endometrium by itself does not always signify true bleeding;
but also occurs, to a lesser degree, with abnormal some fibrin may be a result of the mechanical dis-
bleeding patterns (Fig. 5.3). The debris appears to ruption that occurs during the biopsy procedure.
represent nuclear karyorrhexis from apoptosis of Another feature of bleeding due to abnormal
individual cells within the glands [6]. glandular and stromal breakdown as opposed to
Along with stromal collapse, fibrin thrombi menstrual bleeding is eosinophilic syncytial
typically form in small vessels, representing change of the surface epithelium. This lesion,
another sentinel of abnormal bleeding [7, 8]. The previously termed “papillary syncytial metapla-
thrombi form either in superficial portions of the sia,” [9] “papillary syncytial change,” [10] or
spiral arteries or in ectatic venules in the stroma “surface syncytial change,” [11] is a result of
beneath the surface epithelium (Fig. 5.4). These focal, irregular breakdown [10]. Because of its
ectatic vessels develop in association with any consistent association with breakdown, eosino-
condition in which prolonged endometrial philic syncytial change appears to be a degenera-
growth is not followed by physiologic shedding. tive or regressive alteration and is not a
Examples include anovulatory cycles, metaplastic transformation [9, 10, 12]. This is
hyperplasia, polyps, and progestin effect.
demonstrated by a low Ki-67 proliferation index.
Fig. 5.2 Glandular and stromal breakdown. Proliferative endometrium with glandular and stromal breakdown. Foci of
breakdown with collapsed stroma and characteristic clusters or “stromal blue balls” (lower right)
In eosinophilic syncytial change, portions of the eosinophilic syncytial and hobnail change as a
surface epithelium coalesce into syncytial aggre- component of a bleeding pattern. Eosinophilic
gates of cells with eosinophilic cytoplasm syncytial change has been associated with men-
(Figs. 5.5, 5.6, and 5.7). The cells often ran- strual, proliferative, secretory and atrophic endo-
domly pile up into small papillae infiltrated by metrium, as well as in hyperplasia and
neutrophils, and these papillae may contain endometrial polyps. [13]
microcystic spaces (Fig. 5.5). In this change the Because of stromal collapse and tissue frag-
pink cells have sparsely vacuolated cytoplasm mentation associated with bleeding, the endome-
and indistinct cell borders, leading to the appear- trial glands and surface epithelium become
ance of a syncytium (Figs. 5.5 and 5.7). Nuclei disrupted and crowded, yielding a pattern that
vary from oval to rounded but have a uniform can mimic complex hyperplasia or even a well-
chromatin distribution (Fig. 5.7). At times the differentiated endometrioid carcinoma (Fig. 5.9).
nuclei may be hyperchromatic with irregular Glandular and stromal breakdown is recognized
borders. Occasional nucleoli and rare mitotic by the pattern of stromal collapse as well as the
figures are present. These changes should not be disrupted glands. With breakdown and bleeding,
interpreted as atypical. Hobnail surface change these fragmented glands lose the stroma that
may also occur in abnormal glandular and stroma should be present between intact glands, even in
breakdown (Fig. 5.8). The underlying stromal conditions such as hyperplasia, in which intact
collapse and condensation helps to identify glands show marked crowding.
Morphologic Features of Glandular and Stromal Breakdown in Menstrual and Abnormal Bleeding 125
Fig. 5.3 Glandular and stromal breakdown. Focus of results in distortion of the glands and condensation of
glandular and stromal breakdown in proliferative endo- stromal cells into tight clusters
metrium due to anovulatory cycles. Collapse of the tissue
Other epithelial changes associated with trial carcinoma (see Chap. 10), but they can
breakdown are highly variable. As bleeding occur in hyperplasia or other benign pathologic
becomes chronic, variable amounts of hemosid- conditions and therefore are a nonspecific find-
erin deposits [14] and foam cells [15] appear in ing. The architecture of the glands and the cyto-
the stroma. Neither hemosiderin nor foam cells logic features of the epithelial cells, not the
occur in normal cycling endometrium during foam cells, are features that determine the
the reproductive years, apparently because the pathologic diagnosis. Chronic bleeding occa-
tissue sloughs during menstruation. Hemosiderin sionally results in patches of stromal hyaliniza-
deposits can also be found in the endometrium tion and fibrosis when b leeding is more severe
of women using levonorgestrel intrauterine and prolonged, and this appears to be related to
devices [16]. Foam cells, originally thought to fibrin deposition.
represent macrophages, are endometrial stromal Necrotic debris or old blood is sometimes pres-
cells that become distended with lipid following ent in the lumens of otherwise normal endometrial
erythrocyte breakdown in areas of nonphysio- glands (Fig. 5.10). This debris seems to result from
logic hemorrhage [15]. These cells have abun- breakdown with entrapment of the debris within
dant pale, faintly granular cytoplasm and small glands. The association with endometrial break-
oval nuclei and can stain for macrophage immu- down and bleeding is poorly defined, however, and
nohistochemical markers, including CD68 [17]. often no definite abnormalities are present when
Foam cells usually are associated with endome- luminal debris is seen. Without other morphologic
Fig. 5.4 Fibrin thrombi with glandular and stromal breakdown. Proliferative endometrium shows ectatic venules
that contain fibrin thrombi
abnormalities, luminal debris is a nonspecific find-

mone administration, will be discussed further in
ing with no known clinical significance. Chap. 6, Effects of Hormones. Those changes
related to ovulatory dysfunction (AUB-O) will be
the focus of discussion further below.
Abnormal Uterine Bleeding: AUB may be a presenting symptom in women
Nonstructural Causes and adolescents with coagulopathies, (AUB-C)
or in patients receiving anticoagulation therapy
Nonstructural causes of abnormal uterine bleed- [21–24]. In one study of women initiating oral
ing are frequent in perimenopausal and perimen- anticoagulation, up to 70% of women reported a
archeal women, and they occur to a lesser extent change in their menstrual cycle with prolonged
in women of reproductive age [1, 2, 18–20]. This (50%) and heavy menstrual bleeding (66%) [25].
includes the latter half of the acronym PALM- Women on anticoagulation therapy may receive
COEIN (coagulopathy, ovulatory dysfunction, progestin therapy [26, 27] for heavy menstrual
endometrial, iatrogenic, and not yet classified). bleeding and therefore induce iatrogenic hor-
These etiologies do not have specific histopatho- monal effects on the endometrium (see Chap. 6,
logic findings, except for ovulatory dysfunction “Effects of Hormones”). Although no definitive
and iatrogenic causes, which may show hormonal histopathologic features have been defined in
changes. Endometrial changes related to iatro- patients with a coagulopathy, there have been
genic causes (AUB-I), including exogenous hor- associations with recurrent pregnancy loss [28]
Abnormal Uterine Bleeding: Nonstructural Causes 127
Fig. 5.5 Glandular and stromal breakdown with eosino- proliferative endometrium. The surface is heaped into a
philic syncytial change. A focus of prominent eosino- syncytium of cells with eosinophilic cytoplasm and indis-
philic syncytial change in an area of breakdown in tinct cell borders
and implantation failure in women undergoing AUB related to ovulatory dysfunction (AUB-
in vitro fertilization embryo transfer [29–31]. O) occurs either because of lack of ovulation fol-
The exact mechanisms and interactions between lowing follicular development (anovulatory
maternal endometrium and embryo development cycles) or because of luteal phase abnormalities
have yet to be elucidated. [4]. The latter includes abnormal persistence of
AUB-E is a primary disorder of the endome- the corpus luteum (irregular shedding). Often,
trium in the setting of cyclic ovulatory menstrual before a biopsy is performed, AUB-O is managed
bleeding in the absence of other identifiable by hormonal therapy. When bleeding persists,
causes. At the local endometrial level, there may curettage often becomes necessary to control
be difficulty in regulating the vascular changes as bleeding and exclude structural lesions. Clinical-
evidenced by increased production of prostaglan- pathologic correlations of the full spectrum of
din E2 and prostacyclin (I2) leading to vasodila- morphologic changes associated with AUB are
tion [32, 33], excess plasminogen activator not known. Nonetheless, certain endometrial
leading to lysis of endometrial clot formation alterations can be correlated with abnormalities
[34], and deficiencies of production of vasocon- in the pattern of sex steroid hormone production.
strictors prostaglandin F2α, which have all been Anovulatory cycles are, by far, the most com-
associated with heavy menstrual bleeding [34]. mon cause of nonstructural AUB [4]. The preva-
However, no specific histopathologic findings are lence of postovulatory luteal phase abnormalities
associated with AUB-E. as an etiology for AUB is not known. Ovarian
Fig. 5.6 Glandular and stromal breakdown with eosino- change that forms pseudopapillary structures. The stroma
philic syncytial change. At low magnification the endo- is collapsed into dense aggregates. Other areas showed
metrium shows extensive breakdown. The detached underdeveloped secretory changes. The cause of the
epithelium demonstrates marked eosinophilic syncytial breakdown was not determined, but the pattern is benign
dysfunction with anovulatory cycles or luteal tions of bleeding disorders, atrophy is not regarded
phase abnormalities also may present with infer- as a form of abnormal uterine bleeding, yet it is a
tility (see Chap. 3) rather than AUB. significant cause. The second, less frequent cate-
To place ovulatory abnormalities in the appro- gory of AUB-O is progesterone-related and reflects
priate pathophysiologic context, this disorder can abnormal endogenous progesterone levels.
be grouped into two broad categories: estrogen- All these disorders, classified here as AUB-O,
related and progesterone-related bleeding. The reflect variations in ovarian hormone production.
more common is estrogen-related AUB-O, which Exogenous hormones may produce endometrial
refers to episodes of bleeding that are related to patterns that are indistinguishable from the patterns
lack of ovulation with alterations in endogenous seen in AUB-O caused by endogenous hormone
estrogen levels. Although not really a manifesta- fluctuations. Although this chapter specifically
tion of AUB, atrophy is included as a form of addresses AUB and hormone effects of ovulatory
estrogen-related bleeding because it occurs when dysfunction, similar morphologic changes may be
the endometrium is deprived of estrogen for a rela- attributable to exogenous hormones; these effects
tively long period of time. In clinical classifica- are discussed further in Chap. 6.
Fig. 5.7 Glandular and stromal breakdown with eosinophilic syncytial change. A syncytium of surface epithelial cells
and stromal breakdown. The epithelial cells show oval to rounded nuclei and have a uniform chromatin distribution
Estrogen-Related Bleeding 35]. Exceptions to this include obese women,

women with chronic anovulation associated with
Proliferative Endometrium polycystic ovary syndrome (Stein-Leventhal syn-
with Glandular and Stromal drome), or those with hereditary cancer syn-
Breakdown dromes, most notably Lynch syndrome, because
these women have an increased risk of develop-
This term describes the endometrial changes ment of hyperplasia or carcinoma (see Chaps. 1
resulting from anovulatory cycles. It is probably and 10 for further details on Lynch syndrome)
the most common abnormality found in biopsies [36–40].
performed for abnormal bleeding in perimeno- Anovulatory cycles result when a cohort of
pausal women. Anovulatory cycles with bleeding ovarian follicles begins to develop but ovulation
also occur in perimenarcheal adolescents in does not occur. Chronic anovulation may be the
whom regular ovulatory cycles are not estab- result of a variety of disorders, including hypo-
lished. Anovulatory bleeding even occurs spo- thalamic dysfunction and obesity, because of
radically in women throughout the reproductive peripheral conversion of androgens to estrogen in
years. Usually, this bleeding does not lead to the adipose tissue, as well as increased androgen pro-
need for biopsy in younger patients, as the risk of duction by the adrenal glands or the ovaries [1,
other lesions, especially hyperplasia and carci- 2]. Causes of anovulation following recruitment
noma, is low in individuals of this age [18, 19, of follicles are complex. They include defects in
Fig. 5.8 Glandular and stromal breakdown with hobnail change. The endometrium with breakdown shows prominent
hobnail change in a papillary-like configuration. The epithelium overlies several clusters of condensed stromal cells
the hypothalamic-pituitary-ovarian axis such as branes and vasoconstriction with bleeding. In

hyperprolactinemia, abnormal feedback mecha- contrast to estrogen withdrawal bleeding, estro-
nisms of hormonal control, and local ovarian fac- gen breakthrough bleeding results from persist-
tors that interfere with appropriate follicular ing follicles that produce estradiol; the
development [1]. Whatever the pathogenesis, if proliferating endometrium becomes thicker and
ovulation does not occur, a corpus luteum does outgrows its structural support. Focal vasocon-
not develop, and progesterone is not produced. striction and thrombosis of dilated capillaries fol-
The follicles produce estradiol, which stimulates low. In either event, the result is irregular
endometrial growth. The developing follicles breakdown and bleeding of the endometrium.
may persist for variable periods of time before Although the terms “withdrawal” and “break-
undergoing atresia. As long as the follicles per- through” help to describe the mechanisms for
sist, estradiol is secreted, and the endometrium estrogen-related bleeding, various authors have
proliferates. defined these terms in different ways. The lack of
When these follicles undergo atresia, estradiol uniform interpretation of these terms limits their
production falls precipitously, and estrogen with- usefulness for reporting the pathologic changes
drawal bleeding occurs. This results in the loss of associated with endometrial bleeding.
estrogenic support of endometrial proliferation In most cases of anovulatory AUB, the endome-
resulting in destabilization of lysosome mem- trium shows a proliferative phase pattern with glan-
Fig. 5.9 Glandular and stromal breakdown with artifac- crowding of the glands. The discontinuous, collapsed
tual crowding. Nonmenstrual secretory phase endome- stroma around the glands helps to identify this pattern as
trium undergoing extensive breakdown and artifactual an artifact
dular and stromal breakdown (Figs. 5.11 and 5.12; subepithelial stroma and often thrombose. Because
see also Figs. 5.2, 5.3, 5.4, and 5.5). The amount of of continuous estrogenic stimulation, the tissue
tissue and the architectural pattern of the glands often shows estrogen-induced epithelial changes
depend on the duration of unopposed estrogenic (“metaplasia”), especially ciliated cell and eosino-
stimulation, not necessarily the level of estrogen. In philic cell change (see Chap. 9). The glands also
addition, the extent of breakdown can be highly may show focal subnuclear vacuolization as a
variable, ranging from minute areas to extensive response to estrogen stimulation, but the extent and
involvement of the specimen. Sporadic anovulation uniformity of the vacuolization are less than that
often results in rapid atresia of follicles with estro- seen in normal early secretory glands. The cyto-
gen withdrawal bleeding. This results in minimal plasmic changes and subnuclear vacuoles compli-
endometrial proliferation. A small amount of endo- cate the interpretation of the h istologic pattern but
metrium with poorly developed, weakly prolifera- do not change the diagnosis. Prolonged, unop-
tive glands and stroma develops (Fig. 5.13). posed estrogenic stimulation also can lead to the
Chronic anovulation results in persistence of folli- development of varying degrees of hyperplasia,
cles and sustained unopposed estrogen stimulation. atypical hyperplasia, and even well differentiated
A greater quantity of endometrial tissue develops endometrioid carcinoma, but these structural
with actively proliferating glands and augmented lesions are not functional disorders and, as such,
glandular tortuosity. Dilated venules appear in the are not considered causes of AUB.
Fig. 5.10 Debris in glandular lumens. A gland contains entrapped cellular debris within the lumen. By itself, this find-
ing has no known significance. Stromal condensation with breakdown is present at the surface
When proliferative endometrium shows break- nization characterized by focal glandular dilation
down and bleeding, the pattern strongly suggests may occur. Usually these are regarded as variants
anovulatory cycles. Exogenous estrogens can of normal proliferative endometrium. Sometimes
cause similar patterns, and therefore a complete more sustained estrogen stimulation may result
clinical history is needed to be certain that the in the focal branching and some dilation of
bleeding pattern is truly due to anovulation. The glands, yielding a proliferative phase pattern that
differential diagnosis of proliferative phase endo- is neither normal nor hyperplastic (Fig. 5.14).
metrium with glandular and stromal breakdown The terms “disordered proliferative endometrium”
also includes inflammation, polyps, and leiomyo- and “persistent proliferative phase” have been
mas. In such cases, the presence of other features, applied to describe this pattern of proliferative
such as plasma cells in chronic endometritis or the endometrium with tortuous and mildly disorga-
dense stroma and thick-walled vessels of polyps, nized glands. A designation of a “disordered” or
establishes the proper diagnosis. “persistent proliferative phase” has utility in cor-
relating the morphologic findings with the appar-
ent pathophysiology. When used, these terms
Disordered Proliferative should be clarified in a note so that the gynecolo-
Endometrium and Persistent gist understands the clinical significance of the
Proliferative Phase change.
In our experience, “disordered prolifera-
When chronic anovulatory cycles result in abun- tive” often is inappropriately applied to a vari-
dant proliferative tissue, mild degrees of disorga- ety of patterns, including normal proliferative
Fig. 5.11 Proliferative endometrium with glandular and stromal collapse along the surface. There is a stromal cluster
stromal breakdown. Anovulatory bleeding pattern in peri- (top center). This patient was not on hormones, but estrogen
menopausal women shows proliferative phase pattern with replacement therapy can present a similar appearance
endometrium, proliferative endometrium with glands because of fragmentation (see

breakdown, artifactual crowding, basalis, and Fig. 5.12), but again this change is not that of a
simple hyperplasia. The diagnosis of disor- true disordered proliferative phase pattern.
dered proliferative phase should be reserved
for cases in which assessment is based on
intact, well- oriented fragments of tissue. In Atrophy
these areas the abnormal glands should be
focal. These glands are qualitatively similar to As previously noted, atrophy is an important
those seen in simple hyperplasia, but they are cause of abnormal uterine bleeding in postmeno-
limited in extent and interspersed among pausal patients, found in 25% or more of cases
glands with a normal proliferative phase pat- coming to biopsy [35, 41, 42]. The percentage of
tern. This criterion helps to separate the focal patients with atrophy varies greatly among stud-
disordered proliferative phase pattern from ies, probably reflecting different patient popula-
simple hyperplasia, a more diffuse abnormal- tions as well as variations in indication for biopsy
ity. If the tissue is extensively fragmented or and criteria for diagnosis of atrophy among dif-
disrupted by the procedure and contains mainly ferent pathologists [43–48]. In many laborato-
detached proliferative glands, it is best to diag- ries, atrophy is found in up to 50% of biopsy
nose the change only as proliferative. Extensive specimens taken for postmenopausal bleeding,
breakdown in proliferative endometrium can and in one study, 82% of cases of postmeno-
also display a disorganized appearance to the pausal bleeding were attributable to atrophy [45].
Fig. 5.12 Proliferative with glandular and stromal break- continue to show proliferative features with tubular out-
down. The tissue is extensively fragmented secondary to lines. Other portions of the epithelium show breakdown
the breakdown and the procedure. A few detached glands with eosinophilic change and underlying stromal collapse
Besides being common in postmenopausal sies because tissue fragmentation from the proce-
patients, atrophic endometrium can occur in dure disrupts the glands. Breakdown and bleeding
reproductive-age patients with premature ovarian may be superimposed on the features of atrophy,
failure, either idiopathic or due to radiation or although often, even when there is a history of
chemotherapy for malignancies. abnormal uterine bleeding, the sections show no
With atrophy, tissue obtained at biopsy is typi- evidence of glandular and stromal breakdown.
cally scant, often consisting only of a small Although there is a paucity of tissue in
amount of mucoid material. Characteristically, biopsy specimens of atrophic endometrium,
atrophic endometrium is composed of tiny strips these specimens are not insufficient or inade-
and wisps of surface endometrium and detached, quate. The scant tissue may be all that is pres-
fragmented endometrial glands (Fig. 5.15). The ent and therefore is completely representative
epithelium is low columnar to cuboidal with of the lining of the uterine cavity. The minimal
small, dark nuclei and minimal cytoplasm. quantity of tissue should serve as a clue to the
Stroma is scant or absent, consisting of a few diagnosis; it does not represent an insufficient
clusters of small spindle cells. Mitotic activity is specimen (see Clinical Queries and Reporting).
absent. The cystic change seen in atrophic glands This cannot be overstated because a diagnosis
in hysterectomy specimens is rarely seen in biop- of “insufficient” will result in the gynecologist
Fig. 5.13 Weakly proliferative endometrium. are also present. The glandular epithelium is minimally
Fragmented endometrium with isolated glands, extensive stratified and shows only rare mitotic activity; it is there-
blood, and scant stroma from perimenopausal patient with fore designated weakly proliferative
anovulatory bleeding. Glandular and stromal breakdown
repeating the endometrial sampling, which yet lack marked tortuosity and secretory exhaus-
subjects the patient to an unnecessary invasive tion, while the stroma lacks extensive predecid-
procedure. ual change (Figs. 5.16 and 5.17). In other cases,
the glands appear to show a “hypersecretory”
pattern, with vacuolated cytoplasm, marked tor-
Progesterone-Related Bleeding tuosity, and luminal secretion, while the stroma
lacks predecidual change. In addition, the tissue
Biopsy specimens from reproductive-age and shows foci of breakdown with characteristic
perimenopausal women occasionally show changes such as nuclear dust, fibrin thrombi, and
abnormal secretory phase patterns with associ- dense cell clusters, like those occurring in the
ated nonmenstrual breakdown and bleeding. In proliferative endometrium with glandular and
such cases the pattern is secretory owing to ovar- stromal breakdown (Figs. 5.1, 5.2, 5.3, and 5.4).
ian progesterone production, but the glandular Often in abnormal secretory bleeding patterns,
and stromal changes usually are less advanced the glands show stellate shapes as they involute.
than those seen in normal late secretory endome- This latter pattern of collapsing, star-shaped
trium [4, 49–53]. The endometrial pattern does secretory glands is nonspecific, however, and
not correlate with any date of the normal luteal simply shows secretory gland regression that
phase. The glands may show secretory changes could be due to a variety of factors.
Fig. 5.14 Disordered proliferative phase pattern. Portion of endometrium from a patient with apparent anovulatory
bleeding shows disorganized proliferative phase glands with focal branching and glandular dilatation
These changes may reflect AUB-O due to Irregular Secretory Endometrium

luteal phase abnormalities and irregular shedding
(see Chap. 2). The etiology and frequency of Luteal phase abnormalities may cause abnor-
abnormal bleeding caused by luteal phase abnormal bleeding as discussed in Chap. 2; the corpus
malities are not known, however, as these disor- luteum is “insufficient,” either regressing prema-
ders appear to be sporadic and do not persist long turely or failing to produce an adequate amount of
enough to permit clinical-pathologic correla- progesterone to sustain normal secretory phase
tions. Alterations in the morphology of the endo- development. This is a sporadic disorder of the
metrium due to changes in the absolute or relative reproductive and perimenopausal years. With
levels of estrogen and progesterone have been luteal phase abnormalities, ovulation occurs, so
well established in experimental systems [54], so secretory changes develop, but do not fit within the
it is likely that abnormal secretory bleeding pat- physiological progression of the luteal phase [55].
terns are, at least in part, the result of ovulatory If abnormal bleeding is the result, the appearance
abnormalities that involve the luteal phase. is that of breakdown and bleeding in a nonmen-
Nonetheless, the secretory patterns with nonmenstrual secretory phase pattern. One pattern of irreg-
strual bleeding are not well characterized. When ular secretory endometrium is disordered
a pattern of nonmenstrual phase secretory endo- proliferative endometrium with superimposed
metrium with breakdown is present, the abnor- ovulation changes. This is a variation in gland size
mality may be due to defined or undefined luteal and shape, similar to disordered proliferative
phase abnormalities or other causes that are not endometrium, with glandular secretory change
evident from the sections. [55]. The second pattern is glands with secretory
Fig. 5.15 Atrophy with glandular and stromal break- tiny detached wisps and strips of epithelium with almost
down. Atrophic endometrium consists of scant tissue that no stroma. A characteristic stromal cluster or “stromal
shows fragmentation. The specimen consists largely of blue ball” is present (center)
changes. Basally oriented nuclei and vacuolated causes, including early pregnancy, inflammation,
cytoplasm lacking the tortuosity of late secretory and exogenous hormone effects, but we prefer to
phase glands characterize this pattern. Focal restrict this term to bleeding caused by true ovula-
breakdown is present with “stromal blue balls” tory dysfunction. One pattern of irregular shedding
and karyorrhectic debris. This pattern is nonspe- yields a mixed phase pattern composed of secre-
cific, however, and may be attributable to other tory and proliferative endometrium (Fig. 5.18). The
factors discussed in the following. diagnosis is reserved for those specimens in which
there is a mixed pattern of secretory and prolifera-
tive glands at least 5 days after the onset of bleed-
Irregular Shedding ing. Irregular shedding is also manifested by
irregular secretory phase development in which
Irregular shedding is attributed to a persistent cor- multiple phases of secretory endometrium (early,
pus luteum with prolonged progesterone produc- mid, or late) are present. Breakdown and bleeding
tion [19, 49, 51, 52]. This is the least studied and with glandular and stromal collapse is present, usu-
consequently the most poorly understood form of ally focally, but occasionally in a diffuse pattern.
AUB patterns related to ovulatory dysfunction. Although the frequency of irregular shedding as a
Some authors have used this term to refer to irregu- cause of AUB is not known, it is an unusual event
lar secretory phase bleeding from a variety of in our experience.
Fig. 5.16 Abnormal secretory phase with glandular and responding to early secretory endometrium. The stroma
stromal breakdown. In this field the gland on the left is shows evidence of breakdown with early collapse
starting to develop subnuclear vacuolated cytoplasm cor-
sions, all can show patterns of abnormal

bnormal Secretory Endometrium
A secretory development and bleeding. Other
with Breakdown of Unknown poorly understood ovarian disorders, such as a
Etiology luteinized unruptured follicle, presumably could
result in abnormal secretory endometrial
Some examples of secretory endometrium with changes. With this latter entity, developing fol-
abnormal, nonmenstrual bleeding patterns pre- licles are believed to undergo luteinization of
sumably reflect the specific luteal phase abnor- the granulosa and theca cells with progesterone
malities described in the preceding section. A production in the absence of ovulation. In addi-
variety of other factors may also be associated tion to these considerations, management of
with a pattern of aberrant secretory phase devel- AUB often involves progestational therapy. If
opment with superimposed bleeding (Table 5.2). bleeding is not controlled, curettage is per-
For example, endometrial changes associated formed. Accordingly, progestin effects superim-
with abortions or ectopic pregnancies, response posed on the underlying abnormality may
to exogenous progestins, tissue near a polyp, complicate the histology.
endometrium overlying leiomyomas, and endo- These patterns show glands with secretory
metrium involved with inflammation or adhe- changes such as basally oriented nuclei and dif-
Fig. 5.17 Abnormal secretory phase. Some glands show tuous. The stroma is dense. Other areas in the sections
secretory changes with basal nuclei and slightly vacuo- showed glandular and stromal breakdown
lated cytoplasm but are small and tubular rather than tor-
fuse cytoplasmic vacuolization and absence of gests the possibility of a luteal phase abnor-
mitotic activity. The glands may be tortuous. mality if no other pathologic condition is
Often the stroma is dense, lacking edema or pre- clinically manifest.
decidual. The endometrium in such cases cannot
be assigned to any phase (early, mid, or late) of
the normal secretory phase of the menstrual cycle. Clinical Queries and Reporting
The changes of glandular and stromal breakdown
are similar to those found in any bleeding phase When a biopsy is performed for AUB, the report
endometrium with glandular and stromal col- should address the presence or absence of mor-
lapse, “stromal blue balls,” and eosinophilic syn- phologic changes of breakdown and bleeding as
cytial change. With early breakdown, tortuous well as any specific lesions. If the pattern is that
secretory glands often show star-shaped outlines. of proliferative endometrium with breakdown
There are no clear-cut clinical correlations and if the clinical history is appropriate, the
for many abnormal secretory phase patterns changes can accurately be attributed to anovula-
with breakdown, so the alterations at times tory cycles. A descriptive diagnosis such as “pro-
defy precise pathologic diagnosis. Nonetheless, liferative endometrium with glandular and
recognition of the general category of abnor- stromal breakdown” offers a precise morphologic
mal secretory bleeding patterns helps to interpretation of the anovulatory bleeding pattern
exclude other specific organic lesions and sug- that often is sufficient for clinical management.
Fig. 5.18 Irregular shedding with mixed phase pattern. A be due to a persistent corpus luteum, but the etiology can-
mixed phase pattern with secretory changes (left) and pro- not be determined by morphology alone, and clinical cor-
liferative changes (right). This mixed phase pattern may relation is required
Table 5.2 Possible causes of nonmenstrual secretory as “irregular secretory endometrium with glandu-
phase bleeding lar and stromal breakdown” communicate the
Luteal phase abnormalities observation of an abnormal yet benign appear-
Persistent corpus luteum (irregular shedding) ance while not assigning definite morphologic
Structural lesions etiology. In general, a comment regarding the
Submucosal leiomyomas absence of other possible causes of bleeding,
Intrauterine adhesions
such as hyperplasia, inflammation, pregnancy, or
Inflammation
polyps, is most useful in addressing specific clin-
Complications of pregnancya
Progestin effects
ical concerns.
Because atrophy is one of the most frequent
See Chap. 3, Table 3.1
a
causes of abnormal bleeding in postmenopausal
patients, it is important to recognize the morpho-
An additional comment indicating that the change logic features of atrophy and correctly report the
is compatible with anovulatory cycles helps to findings. A scant amount of endometrium con-
clarify the diagnosis. If the changes show non- sisting of detached strips of atrophic endometrial
menstrual secretory endometrium with break- epithelium with little stroma should be regarded
down but these are not diagnostic of a defined as consistent with atrophy and not “insufficient
luteal phase abnormality, descriptive terms such for diagnosis.” A brief comment or description
References 141
of the findings helps the clinician understand the logic diagnoses because they lack clear defini-
basis of the diagnosis while providing reassur- tions in the clinical literature regarding
ance that the endometrium has, in fact, been endometrial bleeding.
sampled.
Occasional biopsies show extensive break-
down and bleeding that largely obscures the cyto- References
logic details of the glands and stroma. Although it
is usually possible to exclude neoplastic processes 1. Fritz MA, Speroff L. Abnormal uterine bleeding. In:
Clinical gynecologic endocrinology and infertility.
in such cases, detailed assessment of the endome- 8th ed. Philadelphia: Lippincott Williams & Wilkins;
trium to determine the underlying pathologic pro- 2010. p. 591–620.
cess becomes difficult. Unless the breakdown is 2. Bayer SR, DeCherney AH. Clinical manifestations
clearly menstrual, i.e., reflecting the shedding at and treatment of dysfunctional uterine bleeding.
JAMA. 1993;269:1823–8.
the end of a normal ovulatory cycle, breakdown 3. Galle PC, McRae MA. Abnormal uterine bleed-
patterns should not be diagnosed as “menstrual.” ing. Finding and treating the cause. Postgrad Med.
Instead, it is better to use descriptive diagnoses 1993;93:73–6, 80–1.
that reflect the morphologic changes. On rare 4. Vakiani M, Vavilis D, Agorastos T, Stamatopoulos
P, Assimaki A, Bontis J. Histopathological find-
occasions endometrial biopsy in a postmeno- ings of the endometrium in patients with dysfunc-
pausal patient may show only evidence of abnor- tional uterine bleeding. Clin Exp Obstet Gynecol.
mal bleeding such as foam cells. In such cases the 1996;23:236–9.
subtle evidence of abnormal bleeding should be 5. Ryntz T, Lobo RA. Abnormal uterine bleeding.
In: Lobo RA, Gershenson DM, Lentz GM, Valea
recognized and reported; further sampling may be FA, editors. Comprehensive gynecology. 7th ed.
needed to determine the cause, as endometrial Philadelphia: Elsevier; 2017. p. 621–33.
carcinoma becomes a more likely cause of abnor- 6. Stewart CJ, Campbell-Brown M, Critchley HO,
mal bleeding in this age group. Farquharson MA. Endometrial apoptosis in patients
with dysfunctional uterine bleeding. Histopathology.
It is imperative that the terms used in 1999;34:99–105.
descriptive pathologic diagnoses be clearly 7. Ferenczy A. Pathophysiology of endometrial bleed-
understood by the clinician. The terms ing. Maturitas. 2003;45:1–14.
employed can be self-explanatory, especially 8. Picoff RC, Luginbuhl WH. Fibrin in the endometrial
stroma: its relation to uterine bleeding. Am J Obstet
when a brief microscopic description accompa- Gynecol. 1964;88:642–6.
nies the report to clarify the histologic finding 9. Clement PB. Pathology of the uterine corpus. Hum
and to exclude more specific structural lesions. Pathol. 1991;22:776–91.
The diagnosis of “proliferative endometrium 10. Zaman SS, Mazur MT. Endometrial papillary
syncytial change. A nonspecific alteration asso-
with glandular and stromal breakdown” is an ciated with active breakdown. Am J Clin Pathol.
example of a clinically relevant term that can be 1993;99:741–5.
applied to apparent anovulatory bleeding. 11. Lack L, Silverberg S, Kurman R. Tumors of the uter-
Diagnoses such as “disordered proliferative ine corpus and gestational trophoblastic disease –
Third Series. 4th ed. Washington, DC: Armed Forces
endometrium” or “persistent proliferative Institute of Pathology; 1992.
phase” applied to the proliferative endome- 12. Shah SS, Mazur MT. Endometrial eosinophilic syn-
trium of anovulatory cycles are most useful cytial change related to breakdown: immunohisto-
when there is a clear understanding of the chemical evidence suggests a regressive process. Int
J Gynecol Pathol. 2008;27:534–8.
microscopic findings and the physiologic and 13. Nicolae A, Preda O, Aneiros-Fernández J, Palacios J,
clinical significance of the pattern. Therefore, Biscuola M, Nogales FF. p16 INK4A positivity iden-
an explanatory note added to the diagnosis is tifies endometrial surface papillary syncitial change
helpful to describe the changes and indicate as a regressive feature associated with desquamation.
Histopathology. 2011;58:483–6.
that the lesion lacks atypia or clear-cut criteria 14. Reeves G, Sommers SC. Endometrial hemosid-
for hyperplasia. The terms “withdrawal” and erin as evidence of metrorrhagia. Obstet Gynecol.
“breakthrough” should be avoided in patho- 1962;19:790–2.
15. Fechner RE, Bossart MI, Spjut HJ. Ultrastructure of 33. Smith SK, Abel MH, Kelly RW, Baird
endometrial stromal foam cells. Am J Clin Pathol. DT. Prostaglandin synthesis in the endometrium of
1979;72:628–33. women with ovular dysfunctional uterine bleeding.
16. Phillips V, Graham CT, Manek S, McCluggage Br J Obstet Gynaecol. 1981;88:434–42.
WG. The effects of the levonorgestrel intrauterine 34. Gleeson NC. Cyclic changes in endometrial tissue
system (Mirena coil) on endometrial morphology. J plasminogen activator and plasminogen activator
Clin Pathol. 2003;56:305–7. inhibitor type 1 in women with normal menstruation
17. Silver SA, Sherman ME. Morphologic and immuno- and essential menorrhagia. Am J Obstet Gynecol.
phenotypic characterization of foam cells in endome- 1994;171:178–83.
trial lesions. Int J Gynecol Pathol. 1998;17:140–5. 35. Van den Bosch T, Ameye L, Van Schoubroeck D,
18. Altchek A. Dysfunctional uterine bleeding in adoles- Bourne T, Timmerman D. Intra-cavitary uterine
cence. Clin Obstet Gynecol. 1977;20:633–50. pathology in women with abnormal uterine bleeding:
19. Scommegna A, Dmowski WP. Dysfunctional uterine a prospective study of 1220 women. Facts Views Vis
bleeding. Clin Obstet Gynecol. 1973;16:221–54. Obgyn. 2015;7:17–24.
20. Aksel S, Jones GS. Etiology and treatment of 36. Wise MR, Gill P, Lensen S, Thompson JMD, Farquhar
dysfunctional uterine bleeding. Obstet Gynecol. CM. Body mass index trumps age in decision for
1974;44:1–13. endometrial biopsy: cohort study of symptom-
21. Ewenstein BM. The pathophysiology of bleeding dis- atic premenopausal women. Am J Obstet Gynecol.
orders presenting as abnormal uterine bleeding. Am J 2016;215:598.e1–8.
Obstet Gynecol. 1996;175:770–7. 37. Stoffel E, Mukherjee B, Raymond VM, Tayob N,
22. Davies J, Kadir RA. Heavy menstrual bleeding: an Kastrinos F, Sparr J, Wang F, Bandipalliam P, Syngal
update on management. Thromb Res. 2017;151(Suppl S, Gruber SB. Calculation of risk of colorectal and
1):S70–7. endometrial cancer among patients with Lynch syn-
23. Davies J, Kadir RA. Endometrial haemosta- drome. Gastroenterology. 2009;137:1621–7.
sis and menstruation. Rev Endocr Metab Disord. 38. Pennington KP, Walsh T, Lee M, Pennil C, Novetsky
2012;13:289–99. AP, Agnew KJ, Thornton A, Garcia R, Mutch D, King
24. Díaz R, Dietrich JE, Mahoney D, Yee DL, Srivaths M-C, Goodfellow P, Swisher EM. BRCA1, TP53, and
LV. Hemostatic abnormalities in young females with CHEK2 germline mutations in uterine serous carci-
heavy menstrual bleeding. J Pediatr Adolesc Gynecol. noma. Cancer. 2013;119:332–8.
2014;27:324–9. 39. Tan M-H, Mester JL, Ngeow J, Rybicki LA, Orloff
25. Huq FY, Tvarkova K, Arafa A, Kadir RA. Menstrual MS, Eng C. Lifetime cancer risks in individuals
problems and contraception in women of reproductive with germline PTEN mutations. Clin Cancer Res.
age receiving oral anticoagulation. Contraception. 2012;18:400–7.
2011;84:128–32. 40. van Lier MGF, Wagner A, Mathus-Vliegen EMH,
26. Bradley LD, Gueye N-A. The medical management Kuipers EJ, Steyerberg EW, van Leerdam ME. High
of abnormal uterine bleeding in reproductive-aged cancer risk in Peutz-Jeghers syndrome: a systematic
women. Am J Obstet Gynecol. 2016;214:31–44. review and surveillance recommendations. Am J
27. Boonyawat K, O’Brien SH, Bates SM. How I treat Gastroenterol. 2010;105:1258–64; author reply 1265.
heavy menstrual bleeding associated with anticoagu- 41. Gredmark T, Kvint S, Havel G, Mattsson
lants. Blood. 2017;130:2603–9. LA. Histopathological findings in women with
28. Pritchard AM, Hendrix PW, Paidas MJ. Hereditary postmenopausal bleeding. Br J Obstet Gynaecol.
thrombophilia and recurrent pregnancy loss. Clin 1995;102:133–6.
Obstet Gynecol. 2016;59:487–97. 42. Patel V, Wilkinson EJ, Chamala S, Lu X, Castagno
29. Ivanov P, Tsvyatkovska T, Konova E, Komsa- J, Rush D. Endometrial thickness as measured by
Penkova R. Inherited thrombophilia and IVF failure: transvaginal ultrasound and the corresponding histo-
the impact of coagulation disorders on implantation pathologic diagnosis in women with postmenopausal
process. Am J Reprod Immunol. 2012;68:189–98. bleeding. Int J Gynecol Pathol. 2017;36:348–55.
30. Qublan HS, Eid SS, Ababneh HA, Amarin ZO, 43. Rubin SC. Postmenopausal bleeding: etiology,
Smadi AZ, Al-Khafaji FF, Khader YS. Acquired evaluation, and management. Med Clin North Am.
and inherited thrombophilia: implication in recur- 1987;71:59–69.
rent IVF and embryo transfer failure. Hum Reprod. 44. Schindler AE, Schmidt G. Post-menopausal bleed-
2006;21:2694–8. ing: a study of more than 1000 cases. Maturitas.
31. Coulam CB, Jeyendran RS, Fishel LA, Roussev 1980;2:269–74.
R. Multiple thrombophilic gene mutations are risk 45. Choo YC, Mak KC, Hsu C, Wong TS, Ma
factors for implantation failure. Reprod Biomed HK. Postmenopausal uterine bleeding of nonorganic
Online. 2006;12:322–7. cause. Obstet Gynecol. 1985;66:225–8.
32. Smith SK, Abel MH, Kelly RW, Baird DT. A role for 46. Meyer WC, Malkasian GD, Dockerty MB, Decker
prostacyclin (PGi2) in excessive menstrual bleeding. DG. Postmenopausal bleeding from atrophic endo-
Lancet. 1981;1:522–4. metrium. Obstet Gynecol. 1971;38:731–8.
References 143
47. Gambrell RD. Postmenopausal bleeding. J Am 52. Hendrickson MR, Kempson RL. Surgical pathol-
Geriatr Soc. 1974;22:337–43. ogy of the uterine corpus. Major Probl Pathol.
48. Lidor A, Ismajovich B, Confino E, David 1979;12:1–580.
MP. Histopathological findings in 226 women with 53. Buckley H, Fox H. Biopsy pathology of the endome-
post-menopausal uterine bleeding. Acta Obstet trium. 2nd ed. London: Arnold; 2002.
Gynecol Scand. 1986;65:41–3. 54. Good RG, Moyer DL. Estrogen-progesterone rela-
49. McCluggage WG. Benign diseases of the endome- tionships in the development of secretory endome-
trium. In: Kurman RJ, Hedrick Ellenson L, Ronnett trium. Fertil Steril. 1968;19:37–49.
BM, editors. Blaustein’s pathology of the female 55. Parra-Herran C, Cesari M, Djordjevic B, Grondin
genital tract. 6th ed. New York: Springer; 2011. K, Kinloch M, Köbel M, Pirzada A, Plotkin A, Gilks
p. 305–58. CB. Canadian consensus-based and evidence-based
50. Dallenbach-Hellweg G. The endometrium of infertil- guidelines for benign endometrial pathology report-
ity. A review. Pathol Res Pract. 1984;178:527–37. ing in biopsy material. Int J Gynecol Pathol. 2018 Apr
51. Dallenbach-Hellweg G. Histopathology of the endo- 19. https://doi.org/10.1097/PGP.0000000000000481.
metrium. 4th ed. New York: Springer; 1987. [Epub ahead of print].
Effects of Hormones
6
Contents
Estrogenic Hormones 146
rogestins, Oral Contraceptives, and Selective Progesterone Receptor
P
Modulators 146
Patterns of Response 148
ombined Estrogen and Progestin as Replacement Therapy for Menopausal
C
Women 153
Effects of Other Hormones 155
References 165
Women receive hormone preparations for a variety hyperplasia managed with progestin therapy or
of reasons, including birth control and treatment routine follow-up of patients on hormone replace-
for abnormal uterine bleeding, perimenopausal ment therapy. In other circumstances, the endo-
and postmenopausal symptoms, endometriosis, metrial sampling is coincidental with another
endometrial hyperplasia, endometrioid carcinoma, procedure, such as tubal ligation, where the
breast carcinoma, and certain types of infertility. patient has received hormone therapy to ensure no
Usually the exogenous hormone is some form of interval pregnancy. The hormone, the dosage, and
progestin, but estrogenic and even androgenic hor- the duration of therapy influence the appearance
mones are used for some disorders. Since the of the endometrium. Clinical information regard-
endometrium has estrogen and progesterone ing hormone use helps in the pathologic interpre-
receptors, it shows the effects of these hormones. tation, but this history sometimes is incomplete
An endometrial biopsy or curettage may be when the specimen is received. Consequently, the
performed when abnormal bleeding occurs or possibility of e xogenous hormonal effects should
when hormone therapy does not correct abnormal always be kept in mind.
bleeding that is thought to be related to anovula- This chapter reviews the effects of different
tory bleeding. Sometimes, however, the biopsy is types of hormones on the endometrium: (1) hor-
intended to evaluate the status of the endometrium mones used in women of reproductive age that
following hormonal therapy, as in the case of clearly have estrogenic or progestogenic effects,

146 6 Effects of Hormones
such as oral contraceptives; (2) estrogen– In some patients continued estrogen use
progestin hormone replacement therapy in leads to atypical hyperplasia and carcinoma
postmenopausal women; (3) tamoxifen therapy [13, 14]. The risk of malignancy increases with
for breast cancer; and (4) other hormones with the duration of therapy. After 1 year of treat-
less well-
established effects on the ment, low-dose unopposed estrogen was asso-
endometrium. ciated with a marginally nonsignificant increase
in endometrial hyperplasia [15]. Unopposed
estrogen use for 2–3 years significantly
Estrogenic Hormones increased the risk of developing endometrial
hyperplasia [15] and endometrioid carcinoma
Estrogen therapy is largely used in perimeno- [16, 17]. The highest risk is in patients who
pausal or postmenopausal women to treat symp- have taken estrogens for 10 years or longer
toms of menopause, such as vasomotor [16]. When carcinoma develops, it usually is
instability, genitourinary syndrome of meno- low grade and superficially invasive, but high-
pause (previously termed atrophic vaginitis), grade lesions may occur.
and osteoporosis [1–8]. Estrogenic substances All the estrogen-related changes are
include conjugated estrogens, such as Premarin reviewed elsewhere in the text, including nor-
(Wyeth Pharmaceuticals, Collegeville, PA), mal proliferative phase patterns (Chap. 2), pro-
micronized estradiol such as Estrace (Allergan liferative with glandular and stromal breakdown
Incorporated, Dublin, Ireland), or transdermal (Chap. 5), hyperplasia and cytoplasmic change
estradiols such as Vivelle-Dot (Novartis (Chap. 9), and carcinoma (Chap. 10). The
Pharmaceuticals Corp, East Hanover, NJ). Use reader should refer to those chapters for
of estrogenic hormones by themselves is associ- detailed morphologic descriptions of the spe-
ated with an increased risk of developing endo- cific entities.
metrial hyperplasia and endometrioid
carcinoma, so the use of these hormones alone
is now unusual in patients with a uterus. Progestins, Oral Contraceptives,
Consequently, the effects of unopposed exoge- and Selective Progesterone
nous estrogen are seen less frequently in biopsy Receptor Modulators
specimens than the effects of combined estro-
gen–progestin compounds, as progestins abro- Although progestin effects are common, the sub-
gate the effect of estrogen stimulation on the ject of progestin-related changes is complex.
uterus. Nonetheless, some patients do receive Various forms of these synthetic analogues of
estrogen replacement only. progesterone, also termed “progestogens” or
Unopposed estrogenic stimulation causes the “progestagens,” are widely used, either alone or
endometrium to proliferate [9–12]. The result is in combination with an estrogen. Progestin-only
variable, depending on the dose and duration of therapy is useful in the empirical medical man-
use. Often the pattern is that of proliferative agement of abnormal uterine bleeding that clini-
phase endometrium, showing tubular to tortuous cally appears to be anovulatory. These hormones,
glands and abundant stroma. The patterns can be such as medroxyprogesterone acetate or noreth-
identical to those seen with anovulatory cycles indrone acetate, suppress ovulation and endome-
and may include superimposed breakdown and trial growth. They also lead to secretory changes
bleeding (see Chap. 5). Continued, prolonged followed by gland involution and progesterone
estrogenic stimulation can lead to disordered pro- withdrawal bleeding, effecting a medical “curet-
liferative phase patterns and hyperplasia. tage.” Consequently, progestins are especially
Estrogen-related epithelial cytoplasmic changes, helpful in managing ovulatory disorders where
especially squamous differentiation and ciliated irregular, noncyclical endometrial growth results
cell change, also often occur. in abnormal bleeding. Often a trial of progestin is
Progestins, Oral Contraceptives, and Selective Progesterone Receptor Modulators 147
given to alleviate apparent abnormal bleeding, depends on the underlying status of the endome-
and if the bleeding does not resolve, biopsy or trium as well as the dose, potency, and duration of
curettage follows to exclude other structural progestin therapy [12, 14, 24–26]. In fact, brief use
pathology. of oral contraceptives for emergency contracep-
Progestins, usually given in combination with tion shows no significant histologic effects on the
estrogens, are the basis for the oral contraceptive endometrium [27–30]. When used for emergency
or “birth control pill.” Most oral contraceptives contraception, both levonorgestrel and the selec-
are used in a fixed-dose formulation, with small tive progesterone receptor modulator (SPRM), uli-
doses of both the estrogen and progestin taken pristal acetate, act by delaying or inhibiting
daily. Some oral contraceptives use a “phasic” ovulation [31]. Ulipristal acetate (UPA) acts by
combination with increasing amounts of proges- reversibly blocking progesterone receptors and is
tin over a 21-day medication period. In either also used as intermittent or preoperative treatment
case the combination estrogen and progestin is for women with uterine leiomyomas.
administered over 3 weeks, and no medication is With more prolonged use of progestin, how-
given in the 4th week to allow withdrawal bleed- ever, its effects can persist for several weeks to
ing to occur. Some oral contraceptives contain months following cessation of their use. To help
only progestin and are taken continuously with- simplify this complex subject, the effects of the
out a hormone-free fourth week. The dose of pro- progestins can be placed into three general mor-
gestin and estrogen used in modern oral phologic patterns that form the basis for under-
contraceptives is much lower than that used in the standing the entire spectrum of progestin-mediated
initial formulations of oral contraceptives in the changes. These patterns include (1) decidual
1960s. Consequently, the pharmacologic effects (pregnancy-like) changes, (2) secretory changes,
of these steroid contraceptives are somewhat dif- and (3) inactive changes (Table 6.1). The pattern
ferent than those originally described. Other encountered depends on the degree of estrogen
combined hormone contraceptive methods “priming” of the endometrium and the dose and
include the etonogestrel/ethinyl estradiol vaginal duration of administration of the progestin. There
ring, which is inserted vaginally for 3 weeks and often is an overlap between the various patterns
the norelgestromin/ethinyl estradiol transdermal of progestin effect in the endometrium.
patch, which is placed on the skin weekly for
3 weeks. Both are removed following the third
week, and similar to oral contraceptive pills, have
Table 6.1 Morphologic features of progestin effects
a “hormone-free” fourth week.
Long-acting progestin-only contraceptive Decidual (pregnancy-like) effects
Abundant tissue, often polypoid
methods include a medroxyprogesterone injec-
Glands are predominantly inactive, but rarely show
tion every 12 weeks, the subdermally placed marked secretory activity
etonogestrel implant and the levonorgestrel intra- Stroma appears decidualized with lymphoid
uterine device (IUD), which are effective infiltrate
3–5 years, respectively. The levonorgestrel- Vascular ectasia
releasing intrauterine device may also be used to Secretory effects
treat hyperplasia or endometrioid carcinoma of Moderate to sparse amount of tissue
the endometrium. Other noncontraceptive bene- Mildly tortuous secretory glands lined by columnar
cells
fits of hormonal contraception include cycle reg- Stromal cells are plump, oval (predecidual)
ularity, prevention of menstrual migraines; Vascular ectasia
treatment of acne or hirsutism, leiomyomas, or Inactive effects
endometriosis; and a decreased risk of endome- Sparse tissue
trial, ovarian, or colorectal c ancer [18–23]. Glands are small and inactive, not coiled
The morphologic appearance of the endome- Variable amount of stroma with plump to spindle-
trium following progestin therapy is variable and shaped cells
Patterns of Response cases the amount of tissue can be copious, and

the biopsy or curettage can yield large polypoid
Decidual Pattern tissue fragments. Although the tissue is polypoid,
this finding does not indicate the presence of true
The decidual or pregnancy-like pattern, as the polyps. With marked progestin effect, the stromal
term implies, features differentiation of endome- cells become enlarged and show abundant cyto-
trial glands and stroma to a point where they plasm and prominent cell borders, resembling the
resemble the endometrium in pregnancy with decidua of pregnancy (Fig. 6.1). The stroma can
decidual transformation of the stroma. Although show occasional mitotic figures. The glands are
the term “decidua” applies most strictly to the predominantly inactive, but with marked proges-
endometrium of pregnancy, this term also is use- tin effect, the glands develop a hypersecretory
ful for describing this progestin-induced pattern. pattern with vacuolated cytoplasm and abundant
This exaggerated effect typically occurs in endo- luminal secretions. An Arias-Stella-like reaction
metrium which is influenced by high estrogen with nuclear enlargement and hyperchromasia
levels and therefore is actively growing and pro- may occur in glands, but this is very rare. Some
liferating. This morphology is most common fol- glands are dilated (Fig. 6.2). The spiral arteries
lowing high-dose progestin therapy for also can show marked thickening with endothe-
anovulatory cycles or for hyperplasia. In these lial and smooth muscle hyperplasia [12]. The
Fig. 6.1 Progestin effect, decidual pattern. Marked decidual transformation in pregnancy. Inactive glands
decidual reaction following progestin therapy of anovula- with secretory exhaustion are surrounded by abundant
tory bleeding. The changes resemble the endometrial decidualized stroma
Fig. 6.2 Progestin effect, decidual pattern. Plump stro- Signet-ring-like cells (lower left) and ectatic vessels are
mal cells with cystic dilation of a gland. The gland is lined seen (courtesy of Dr. Chengbao Liu, The Johns Hopkins
by cuboidal epithelium with vacuolated cytoplasm. Hospital)
venules in the superficial portion of the endome- ing the patterns of development. In these areas
trium become ectatic (Fig. 6.3). Often, especially of breakdown, the decidua-like character of the
in treated hyperplasia, the glands show promi- stromal cells is lost as the cells degenerate and
nent eosinophilic or mucinous cytoplasmic lose cytoplasm. The glands fragment and
change [32]. Occasional cases of decidua-like become haphazardly oriented. Consequently, it
progestin effect show prominent squamous remains important to avoid areas of active
change within glands. This change is usually bleeding and find intact tissue in order to accu-
seen in cases in which the biopsy results that led rately assess the changes associated with the
to therapy demonstrate hyperplasia, often with no progestin effect.
squamous differentiation [26, 33].
Cases with advanced decidual changes often
show areas of breakdown and bleeding, espe- Secretory Pattern
cially as the dilated venules thrombose. As a
result, many of the features of breakdown and The secretory pattern of progestin effect mimics
bleeding described in Chap. 5 are superim- the glandular and stromal changes seen in the
posed on the progestin effect. With breakdown, luteal (secretory) phase of the menstrual cycle.
the collapse of the stroma and glands signifi- With the secretory pattern, the glands are tortu-
cantly alters their appearance, partially mask- ous, and the glandular cells have basally ori-
Fig. 6.3 Progestin effect, decidual pattern. Marked decidual pattern of progestin effect. Ectatic venules are
decidual change in the stroma and small inactive glands present in the upper left of the field
lined by a single layer of epithelium characterize the
ented nuclei (Figs. 6.4, 6.5 and 6.6). These low underdeveloped, lacking tortuosity. Stromal
columnar cells typically have a small amount of predecidual change tends to be confluent, lack-
pale-staining supranuclear cytoplasm and may ing the intermittent edema that characterizes
show small, randomly distributed vacuoles. The most of the normal secretory phase. Scattered
apical border often becomes smooth and well mitotic figures can be found in the stroma. As in
defined, unlike the ragged luminal border in other progestin-related patterns, the superficial
normal secretory endometrium [12]. The lumens stroma contains ectatic venules and may throm-
may have a small amount of dense, eosinophilic bose (Fig. 6.7).
secretions. The stromal cells show weak prede-
cidual change as they gain cytoplasm and
become mildly enlarged. These predecidualized Inactive Pattern
cells are ovoid with identifiable pale cytoplasm
but are not as large or polygonal as fully decidu- The inactive pattern represents the other end of
alized stromal cells. Although the glandular and the spectrum of progestin effect, in which the
stromal changes superficially resemble the endometrium is hypoplastic. This pattern evolves
secretory phase endometrium of a menstrual following prolonged progestin therapy or with
cycle, neither the glands nor the stroma are continued use of contraceptive hormones includ-
appropriately developed for any day of the ing levonorgestrel-releasing IUD and medroxy-
normal cycle. Usually the glands appear to be progesterone acetate injection (Depo-Provera,
Fig. 6.4 Progestin effect, secretory pattern. The glands show marked secretory changes, and the stroma is transformed
into decidua-like cells
Pfizer, New York, NY) (Fig. 6.8). The glands Other Stromal Changes
atrophy, although they continue to show weak
secretory changes. The glands lose their tortuos- Whereas the glands atrophy with prolonged
ity and are small and tubular with scant to absent progestin effect, with high doses of progestin,
luminal secretions. The epithelium is low colum- the stroma retains a decidua-like appearance.
nar with basal nuclei and a small amount of pale In such cases the stroma can show alterations
cytoplasm. that can be confusing or alarming. For instance,
When the progestin dose is low, the stromal the stroma can appear hyperplastic and pseu-
cells remain mildly enlarged with a small amount dosarcomatous with increased cellularity as
of discernible cytoplasm but lose their decidua- well as nuclear hyperchromasia, enlarged
like appearance (Fig. 6.9). Instead, they are nucleoli, and variation in cell and nuclear size
plump and ovoid with only a moderate amount [12, 34, 35]. A pseudosarcomatous change is
of cytoplasm. Cell borders become indistinct, rare with modern progestin therapy, however,
and stromal mitoses are not found. Vascular and is infrequent in our experience. The stroma
channels beneath the surface epithelium become can show other peculiar alterations. One
ectatic. In contrast to the physiologic atrophy change occasionally seen is clustering of
pattern of the postmenopausal endometrium, groups of enlarged stromal cells with interven-
progestin-induced atrophy often has more abun- ing areas of myxoid change or edema. This
dant stroma, while the glands become tiny and change can impart an epithelioid appearance to
indistinct. some of the decidualized cells, especially when
Fig. 6.5 Progestin effect, secretory pattern. High-magnification view of a pattern resembling secretory phase endome-
trium shows a tortuous gland with secretions. The stromal cells are plump, having a moderate amount of cytoplasm
the cells are enlarged with prominent cell bor- helpful features to separate this progestin effect
ders. The decidualized stromal cells can from true inflammation (see Chap. 7). Also,
develop other epithelioid features such as especially in pregnancy-like patterns of proges-
eccentric nuclei and vacuolated cytoplasm tin effect, multiple small foci of breakdown are
(Fig. 6.10) [36, 37]. In such cases the decidual- accompanied by a neutrophilic response. These
ized stroma can mimic signet-ring cells of met- neutrophils, however, are a localized response
astatic carcinoma [37]. to tissue necrosis and do not represent an infec-
In some cases of progestin effect, infiltrates tious process.
of lymphocytes or neutrophils yield patterns Often there is overlap between the various pat-
that can suggest endometritis (Fig. 6.11). For terns of progestin effect, which depend on the
instance, with prolonged progestin effect of the duration of progestin use, the dose of the proges-
levonorgestrel IUD, the stroma often contains a tin, and underlying endogenous estrogen levels.
moderate infiltrate of stromal granular lympho- In some cases, different fields from the same
cytes and mononuclear cells [38]. These are the specimen show different patterns of progestin
normal lymphoid cells of the endometrium that effect that can range from decidualized stroma to
appear exaggerated owing to the relative atro- secretory or inactive change. Consequently, mor-
phy of the other components. This striking infil- phologic identification of progestin effects
trate can mimic chronic inflammation. An requires recognition of the spectrum of changes
absence of plasma cells and no evidence of that may be found. Furthermore, some patients
gland infiltration by inflammatory cells are on progestins, especially those on oral contracep-
Combined Estrogen and Progestin as Replacement Therapy for Menopausal Women 153
Fig. 6.6 Progestin effect, secretory pattern. The endome- increased relative to the normal secretory phase. The
trium has some resemblance to secretory endometrium of glands, while slightly tortuous, are markedly underdevel-
the normal luteal phase with tortuous glands and abun- oped relative to glands in the normal luteal phase
dant, predecidualized stroma. The amount of stroma is
tives, can even show proliferative phase patterns not increase the risk of hyperplasia [40–42].
when the estrogen influence is present, but the Combined estrogen–progestin hormonal replace-
progestin influence is temporarily decreased or ment can be given either sequentially or in com-
absent. Long-term use of oral contraceptives bination [43, 44]. The most common forms of
rarely may result in permanent endometrial atro- estrogen prescribed are conjugated equine estro-
phy after the agent is discontinued [39]. gens, synthetic conjugated estrogens, micronized
17β-estradiol, and ethinyl estradiol. Sequential
medication uses daily estrogen for the first
ombined Estrogen and Progestin
C 21–25 days of the month and daily progestin
as Replacement Therapy added for the last 10–13 days. This regimen
for Menopausal Women results in withdrawal bleeding. The continuous
regimen uses both estrogen and progestin daily.
Because of the possible deleterious consequences In the continuous regimen, breakthrough bleed-
of unopposed estrogen therapy on the endome- ing may occur during the first 6 months, but then
trium, estrogen replacement is nearly always bleeding usually stops. Patients receiving either
given with a progestin in perimenopausal and the sequential or the combined regimen may
postmenopausal patients with a uterus. Another undergo biopsy as part of the routine surveillance
treatment option is the combination of estrogen to ensure that no neoplasm develops.
and bazedoxifene, a selective estrogen receptor With the sequential estrogen–progestin regimen,
modulator, which prevents bone loss and does the endometrium often shows a weakly proliferative
Fig. 6.7 Progestin effect, secretory pattern. Underdeveloped Thrombi such as this result in bleeding that frequently leads
secretory glands in abundant stroma with decidual change to biopsy
near the surface. A thrombus is present to the right of center.
pattern with small, tubular glands in scant stroma regimen may have a more significant lesion in the
[45]. The epithelium can have occasional mitotic biopsy specimen [53, 56–59]. Polyps, hyperplasia,
figures. The pattern is identical to the weakly pro- and carcinoma are lesions that have been found in
liferative phase pattern seen in association with a few cases. In general, however, estrogen–proges-
anovulatory bleeding caused by estrogen with- tin and estrogen–bazedoxifene replacement therapy
drawal (see Chap. 5). Sometimes the tissue shows control endometrial proliferation, and significant
a superimposed progestin effect with poorly devel- proliferative and neoplastic lesions are less com-
oped secretory changes in the glands (Fig. 6.12) mon than in women not receiving this therapy [9,
[10, 46–48]. This latter pattern of secretory changes 45, 58, 60–64].
is especially likely to be seen if the biopsy is taken
during the period of progestin administration. In
these cases, the glandular cells show mild tortuosity, rogestin-Like Effects with No
P
basal nuclei, some cytoplasmic vacuoles, and scant Hormone Use
luminal secretions. In biopsy material, focal glandu-
lar and stromal breakdown also may be seen. With On rare occasions endometrial tissue will show
the combination regimens, and for that matter also morphologic features of a progestin effect even
with sequential therapy, the endometrium usually though there is clearly no history of exogenous hor-
is inactive [26, 45, 46, 49–54]. Secretory changes mone use. These changes may be seen in both pre-
can be seen especially if higher doses of estrogen menopausal and postmenopausal women, and their
and progestin are used [55]. Occasionally a patient etiology is poorly understood. In premenopausal
receiving either the sequential or the combination women, these patterns can be decidua-like [65] or
Fig. 6.8 Depo-Provera effect, inactive pattern. Small glands in abundant stroma; endometrium following long-term
progestin use often has more abundant stroma while the glands become tiny and indistinct
can resemble “pill effect” changes, with hypoplas- biopsy, can cause increased decidual changes in
tic secretory glands and plump stromal cells. It is the progesterone-primed endometrium [25, 66].
possible that this alteration is due either to abnor- Also, an intrauterine device (IUD) may lead to an
mal persistence of a functioning corpus luteum or enhanced decidual reaction in the endometrium
to the so-called luteinized unruptured follicle. This [25, 67, 68].
latter entity, as the name implies, occurs when a
follicle develops, does not rupture (ovulate), and
persists with luteinization of the granulosa and Effects of Other Hormones
theca cells. If progesterone is produced by the
unruptured follicle, then the result could be a pro- elective Progesterone Receptor
S
gestin effect from the endogenous source. Modulators
There have been a few examples of idiopathic
endometrial decidual reaction in postmenopausal These drugs are used in management of endome-
women who are not taking hormones [36]. These triosis and uterine leiomyomas. Histologic
patients have tended to present with abundant pol- changes can range from inactive and normal-
ypoid tissue. The etiology of the change is not appearing cycling endometrium to a spectrum of
known, but it may be the result of local m
echanical unique features [69]. With short- or long-term
factors rather than a response to progesterone-like use of ulipristal acetate (UPA), a variety of non-
hormones. Mechanical stimulation, including physiologic changes can occur and given the
Fig. 6.9 Progestin effect, inactive pattern. Small inactive moderate amount of cytoplasm, distinguishes this pattern
glands in spindle cell stroma show scant secretory changes. from atrophy due to lack of estrogen. This pattern is com-
The abundant stroma, composed of plump cells with a monly seen in women on continuous oral contraceptives
collective term “progesterone receptor modula-

tor-associated endometrial change (PAEC).” Tamoxifen
These changes can include an increase in cysti-
cally dilated glands lined by flattened epithelium Tamoxifen is a nonsteroidal antiestrogen that is
with low mitotic activity and apoptosis, while the widely used in the hormonal therapy of breast car-
stroma can be edematous or have a fibroblastic cinoma. This drug is a selective estrogen receptor
appearance resembling the stroma of endometrial modulator (SERM) with its action mediated
polyps [70, 71]. The stromal vessels have a range through the estrogen receptor. The effect of
of findings from “chicken-wire capillaries,” to tamoxifen on the endometrium appears to depend
ectatic thin-walled or thickened walls with on the ambient estradiol concentration, meno-
smooth muscle [70]. Because of the antiproges- pausal status, and the dose and duration of tamox-
terone effects of UPA, and therefore unopposed ifen use [73–75]. Current data suggest it can act as
estrogenic effects, endometrial hyperplasia was both an estrogen antagonist and an agonist [76].
observed in up to 1.1% of patients [70]. However, Most normally cycling premenopausal patients
similar to the nonphysiologic findings, the hyper- taking tamoxifen continue to have regular men-
plasia regressed to benign 6 months following strual cycles, but some develop amenorrhea. With
discontinuation of the drug [70, 72]. continued use, serum estrogen and progesterone
Fig. 6.10 Progestin effect, decidual pattern. The decidual cells have vacuolated cytoplasm and eccentric nuclei result-
ing in a signet-ring cell appearance
levels often are increased to two or three times the monitored for symptoms of hyperplasia and car-
normal levels. In postmenopausal women, tamox- cinoma [79]. Premenopausal women on tamoxi-
ifen has estrogenic effects on vaginal epithelium. fen have no increased risk of endometrioid
A large volume of literature has accumulated carcinoma, and routine gynecologic care is rec-
on the effects of tamoxifen. Before beginning ommended [80]. Both transvaginal ultrasound
tamoxifen therapy, asymptomatic postmeno- and endometrial biopsy have been used to moni-
pausal women with ER-positive breast cancer tor patients on tamoxifen, and in the absence of
were found to have a high prevalence of subclini- symptoms, few pathologic lesions are found [81].
cal endometrial abnormalities, including 29.6% In asymptomatic postmenopausal patients on
with polyps (up to three polyps harbored simple tamoxifen, atrophy is the most common finding
hyperplasia) and 0.8% with endometrial hyper- [82, 83]. Endometrial abnormalities are more
plasia [77, 78]. Baseline endometrial pathology commonly found in symptomatic patients [82,
correlated with increasing patient age, BMI, and 84], although some women with no symptoms or
time since menopause; therefore, endometrial ultrasound abnormalities will be found to have a
screening in obese, older women before tamoxi- pathologic lesion [82, 85, 86]. Patients receiving
fen therapy may be useful [77, 78]. In general, a progestin after initial tamoxifen therapy can
postmenopausal women on tamoxifen should be show decidual reaction of the stroma [87, 88].
Fig. 6.11 Progestin effect. Abundant stroma containing a rich infiltrate of stromal granular lymphocytes mimicking an
endometritis
Both endometrial hyperplasia and carcinoma stage and well differentiated [104]. When com-
occasionally occur in patients on tamoxifen, and paring tamoxifen to aromatase inhibitors (AI),
some studies suggest an increased frequency of another endocrine treatment for early breast can-
both these disorders in patients receiving tamoxi- cer, vaginal bleeding, endometrial polyps, and
fen [74, 79, 89–98]. The relative risk of atypical hyperplasia was more common in the
developing endometrial carcinoma appears to be tamoxifen treated group [105]. Carcinosarcomas
within the same range as reported with unop- (malignant mixed müllerian tumors [MMMTs])
posed estrogen use [89]. The apparent increase and other sarcomas including endometrial stro-
may be the result of increased rate of detection of mal sarcoma and adenosarcoma also have been
otherwise asymptomatic, “silent” tumors, how- reported in patients on tamoxifen [106–109].
ever. There are conflicting results regarding the When comparing carcinosarcomas occurring in
histologic grade of endometrial carcinomas that women with and without tamoxifen use, those
develop in women receiving tamoxifen treat- with tamoxifen-related carcinosarcomas may
ment. Some patients develop high-grade have favorable tumor characteristics but compa-
carcinoma [91, 99–101], but most studies indi- rable stage-specific survival outcomes [110].
cate those carcinomas associated with tamoxifen Endometrial polyps appear to be one of the
use do not differ in grade from carcinomas that most common pathologic findings in patients on
occur in patients not receiving this hormone [91, tamoxifen [73, 82, 99, 100, 105, 111–117].
102, 103]. Most endometrial cancers are low These patients are postmenopausal and have
Fig. 6.12 Estrogen–progestin therapy. Endometrium with extensive subnuclear vacuoles. This progestin effect
from a postmenopausal woman receiving estrogen–pro- resembles early secretory phase endometrium
gestin replacement therapy shows small secretory glands
received long-term tamoxifen therapy for meta- least one reported polyp, the stroma also showed
static breast carcinoma. The polyps tend to be decidual changes that could not be attributed to
large and multiple [118], and they may be recur- any exogenous progestin use [112]. Some pol-
rent [119]. The stroma is variably edematous, yps, however, show markedly hyperplastic
myxoid, or fibrous. Often these polyps show glands, and others show cystic glands with focal
mildly hyperplastic changes (Fig. 6.13) (see atypia [111, 112]. A few examples of atypical
Chap. 8). Various types of cytoplasmic change or hyperplasia and carcinoma arising within these
metaplasia have been described in the glands, polyps also have been reported [100, 118, 121,
especially mucinous and clear cell change [114, 122]. The levonorgestrel intrauterine system was
118, 120]. Occasionally endometrial polyps in found to decrease the frequency of benign endo-
patients receiving tamoxifen show foci of secre- metrial polyps and endometrial hyperplasia in
tory changes in the glands with clear to vacuo- women taking tamoxifen, but it was not clear if
lated cytoplasm (Fig. 6.14). The mechanism of endometrial cancer was prevented [123]. With
the secretory effects is not known. Some authors the expanding role of tamoxifen in the treatment
have found that the polyps show glands polar- and possible prevention of breast carcinoma,
ized along the long axis of the polyp with stag- more clinical information will be accumulated
horn-shaped glands and a cambium layer of regarding the effect of this drug on the
stromal condensation around them [114]. In at endometrium.
Fig. 6.13 Tamoxifen-related polyp. Portion of a large ular, staghorn shapes and showed weak proliferative
polyp removed by curettage in a postmenopausal patient activity. The stroma varies from fibrous to edematous
on tamoxifen for breast carcinoma. The glands have irreg-
Raloxifene Clomiphene Citrate
Raloxifene is another selective estrogen receptor Clomiphene citrate is another SERM that is
modulator (SERM) and may be useful in prevent- used to induce ovulation in the treatment of
ing postmenopausal osteoporosis. This drug infertile patients who are anovulatory [1]. This
appears to be a more pure estrogen antagonist, hormone stimulates multiple follicles to
lacking the weak estrogen agonist effects of develop, and ovulation follows. It is thought to
tamoxifen [124, 125]. When compared to pla- act by competitively binding to estrogen recep-
cebo, raloxifene showed no difference in the inci- tors in the hypothalamus, causing increased lev-
dence of ovarian cancer, postmenopausal els of follicle-stimulating hormone (FSH) and
bleeding, endometrial hyperplasia, or carcinoma luteinizing hormone (LH) that induce ovulation.
[126]. In comparison to those receiving tamoxi- Like tamoxifen, clomiphene citrate has been
fen, there appears to be no increase in endome- found to have estrogenic as well as antiestro-
trial pathologic lesions with raloxifene [127–130]. genic activity.
In patients diagnosed with endometrial cancer in Morphologic effects of clomiphene on the
the general population, tamoxifen compared to endometrium are difficult to assess. Biopsies often
raloxifene users were three times more likely to are performed in the luteal phase of clomiphene-
develop endometrial cancer [131]. In that same induced cycles to assess the endometrial develop-
population, the risk of endometrial cancer was ment. Usually the pattern is that of normally
50% less among raloxifene users compared to developing secretory phase endometrium. One
nonusers [131]. Usually the endometrium is study of the morphologic effects of clomiphene
inactive. citrate on the endometrium suggested that the drug
Fig. 6.14 Tamoxifen effect with secretory change. Endometrial biopsy from a woman with breast cancer shows weak
secretory effect manifested by small cytoplasmic vacuoles
causes significant alterations in secretory phase Danazol

development [132]. Decreased gland tortuosity
with scant secretions in clomiphene-treated cases Danazol is structurally related to testosterone and
was described. The gland-to-stroma ratio was is a weak androgen [137]. Its main metabolite,
reportedly decreased relative to secretory endome- ethisterone, is a weak progestin, however [138].
trium in untreated women. In early secretory endo- This steroid is used for the treatment of endome-
metrium, the subnuclear vacuoles appeared to be triosis [1]. Because it suppresses endometrial
larger and more sharply defined, and later in the growth, it also may be used to treat heavy men-
secretory phase, the luminal secretions appeared to strual bleeding [139–142], symptomatic leiomy-
be hyalinized and inspissated. The stroma also omas [142], and endometrial hyperplasia [143,
showed decreased predecidual change compared 144]. The side effect profile of acne, hirsutism,
to untreated endometrium. This report suggests and weight gain makes it a less desirable treat-
that clomiphene may cause reproducible morpho- ment option [139].
logic changes in secretory phase endometrium. The few studies of the effects of danazol on
Another study reported advanced secretory activ- the endometrium show changes similar to those
ity in clomiphene citrate-induced cycles [133]. found in progestin therapy [138, 145, 146].
Other investigators have found no changes in Within a few months of use, the amount of tissue
endometrial morphology following clomiphene is reduced [147]. Glands show weak and irregular
citrate administration [134–136]. In our opinion, secretory changes with mild tortuosity, basal
any morphologic changes associated with clomi- nuclei, and some cytoplasmic vacuolization. The
phene citrate are subtle and difficult to appreciate stroma is hypercellular. With prolonged therapy,
by routine microscopy. the glands become inactive with scant to no
secretory activity [137, 146–148]. Vascular ecta- in vivo matured oocytes [157]. For the patholo-
sia also can occur [149]. Occasionally there is gist interpreting biopsies from women who
some proliferative activity with stromal and glan- receive these hormones, accurate histologic clas-
dular mitoses. sification into early, mid or late secretory phase is
the most important consideration.
uman Menopausal Gonadotropins/

H
Human Chorionic Gonadotropin Gonadotropin-Releasing Hormone
Agonists
Human menopausal gonadotropins (hMG or
menotropins) are extracted from the urine of Gonadotropin-releasing hormone (GnRH) ago-
postmenopausal women and consist of FSH and nists include leuprolide acetate, buserelin acetate,
LH. These drugs are used to induce ovulation in and goserelin acetate. They are also referred to as
the treatment of infertility due to anovulation, “luteinizing hormone-releasing hormone ago-
such as polycystic ovarian disease. Human chori- nists.” These preparations are used to suppress
onic gonadotropin (hCG) has structural and bio- the endometrium prior to resectoscopic ablation
logic similarities to LH and is used to simulate and to decrease the size of leiomyomas before
and improve on the midcycle LH surge associ- surgical removal [158, 159]. They are also used
ated with ovulation. This hormone is used in con- to prevent spontaneous LH surges before oocyte
junction with hMG and can be used along with retrieval and to improve follicular development
clomiphene citrate-induced ovulation. for ovulation induction during in vitro fertiliza-
The effects of hMG and hCG on endometrial tion (IVF) and gamete intrafallopian transfer
morphology are difficult to define. Some studies (GIFT) [1]. In conjunction with progestins, these
suggest that the main change caused by adminis- compounds may also have utility for contracep-
tration of hMG/hCG is endometrial “inadequacy” tion. When they are used to suppress endometrial
with delayed development of more than 2 days growth, GnRH agonists cause marked atrophy of
when the histologic date is compared to the the endometrium [141, 149]. When GnRH ago-
chronologic date [150–153]. In one study, 27% of nists are used in conjunction with a progestin, the
patients treated with hMG/hCG showed inade- endometrium shows apparent secretory changes
quacy in development with out-of-phase histo- consistent with progestin effect [160].
logic dates [150]. Other studies, however, have
reported “advanced” histology with more highly
developed secretory changes than expected for Antiprogestin RU486
the cycle day, gland-stromal dyssynchrony, or
normal glandular development [154, 155]. The synthetic progestogenic steroid RU486, or
Despite the apparent discrepancies in the findings mifepristone, has high affinity for progesterone
between studies, there are no specific morpho- receptors in the endometrium, which causes its
logic alterations that can be definitely correlated antiprogesterone action [161]. Its main use is for
with the effects of these hormones. There is a medical abortion, which involves the use of med-
positive relationship between endometrial thick- ications rather than a surgical procedure for ter-
ness and clinical pregnancy with in vitro matura- mination of early pregnancy. The effects on the
tion (IVM) treatment [156]. When comparing uterus include necrosis of the capillary endothe-
low-dose human menopausal gonadotropin and lium [162] and decidua [161] uterine contractility
micronized 17beta-estradiol supplementation to and sensitivity to prostaglandin [163]. The drug
improve endometrial thickness during IVM, both is used in high dose to induce early abortion and
showed significant improvement in endometrial is most effective up to 63 days of gestation (cal-
thickness, but hMG had the additional advantage culated from LMP). The more common regimen
of larger ovarian follicles and a greater number of is oral administration of mifepristone followed
by misoprostol, a prostaglandin E1 analogue, evaluation of the status of the endometrium

48 h later. following hormone induction of ovulation or

Mifepristone also has been used to treat leio- endometrial growth in infertility therapy. For
myomas and pain in women with endometriosis each of these indications, there are specific con-
[164]. Daily administration has contraceptive siderations in interpretation and reporting.
potential and has been shown to block ovulation
[165, 166]. Studies suggest that with longer-term
use, the secretory phase is not appropriately Postmenopausal Hormone
developed [161, 167–174]. Replacement
One case report of an adolescent female osteo-
porosis treated with high-dose mifepristone for its Surveillance biopsy of asymptomatic postmeno-
antiglucocorticoid effect describes recurrent dif- pausal patients receiving estrogen or estrogen–
fuse simple hyperplasia of the endometrium progestin replacement therapy usually is an
[175]. Controversy has been associated with the office-based procedure intended to provide a
histological effects on women receiving low-dose small representative sample. The primary con-
mifepristone for 6 months for the treatment of cern is whether the endometrium shows evidence
fibroids. One study reported simple non-atypical of hyperplasia or neoplasia that would require
hyperplasia in 28% [176]. A National Institutes of cessation or change of the hormone therapy. The
Health sponsored workshop evaluated the effects presence or absence of these changes should be
of mifepristone (including the specimens from the explicitly stated in the report. This information
aforementioned study), asoprisnil, ulipristal ace- can appear either in the diagnosis or as a com-
tate, and JNJ-17072341. Premalignant lesions ment. In these cases, accurate reporting of the
such as atypical hyperplasia/EIN were not identi- degree of proliferative activity, if any, also is
fied. There was a subset of cases with asymmetry important.
of stromal and epithelial growth resulting in
prominent cystically dilated glands with admixed
estrogen (mitotic) and progestin (secretory) Abnormal Uterine Bleeding
effects [69, 177]. These effects are consistent with
progesterone receptor modulators (PRM) and the There are several situations in which hormone
previously mentioned PRM-associated endome- therapy is related to abnormal bleeding.
trial changes (PAEC) in the ulipristal acetate sec- Sometimes patients on oral contraceptives or
tion above. In postmenopausal women receiving other progestins experience “breakthrough bleed-
estrogen alone, this drug has effects similar to ing,” which indicates bleeding of a noncyclical
those of progesterone, suggesting it acts as a pro- type. This may occur as irregular bleeding during
gesterone agonist [161, 167, 168, 178]. the first few months of oral contraceptive use or
later after many months of oral contraceptive use.
Often the bleeding is caused by glandular atrophy
Clinical Queries and Reporting and increased “fragility” of the tissue with focal
glandular and stromal breakdown. In other situa-
A wide variety of indications for hormone use tions, hormone therapy may be used as primary
influence the reasons for biopsy and the clinical therapy for abnormal bleeding, especially AUB
questions posed to the pathologist. Most biopsies due to ovulatory dysfunction (see Chap. 5).
that show hormone effects are performed for one Usually these patients have anovulatory cycles,
of the following reasons: (1) surveillance during and a progestin is given to suppress proliferation.
postmenopausal hormone replacement therapy, If the bleeding is not controlled by progestin,
(2) evaluation of abnormal uterine bleeding then curettage is performed.
related to hormone therapy, (3) assessment of Other hormone therapies may also lead to
treatment of hyperplasia or carcinoma, and (4) abnormal bleeding. For example, tamoxifen ther-
apy for breast cancer, or estrogen replacement are suppressed by progestin treatment, and the
therapy for menopausal symptoms, may be asso- gland cells show rounded nuclear contours, con-
ciated with bleeding. Biopsy in such cases is per- densed chromatin, and loss of nucleoli. It is
formed to rule out significant structural lesions. important to state clearly that progestin effect is
Also, megestrol acetate (Megace) therapy for present and that it may not be possible to assess
breast cancer may lead to decidual transforma- fully the degree of the glandular abnormality. An
tion of the endometrium with focal glandular and effort also should be made to review the previous
stromal breakdown. In any of these cases in material to assess the effect of the therapy on the
which hormone therapy is associated with abnor- hyperplasia.
mal bleeding, the biopsy is performed to assess Studies have evaluated the effect of progestin
the status of the endometrium and to rule out pos- therapy on complex atypical hyperplasia and
sible underlying inflammation or structural disor- well-differentiated endometrioid carcinoma to
ders such as polyps, hyperplasia, or neoplasia. In assess features that would predict the effective-
addition, the biopsy may be therapeutic, remov- ness of the therapy [32, 179]. These patients often
ing the abnormal tissue that is bleeding. Most of are poor surgical candidates or wish to preserve
these biopsies show progestin effects that range fertility if premenopausal. Most patients were
from decidual patterns to inactive patterns. When treated with an oral progestin or a progestin-
the patient has received unopposed estrogen, releasing IUD. In women with a pre-treatment
however, the pattern usually is proliferative. A diagnosis of complex atypical hyperplasia
specific diagnosis should be rendered. If this is (CAH), progestin therapy generally decreases the
not possible, a descriptive diagnosis should be glandular crowding and complexity of the indi-
made. When glandular and stromal breakdown vidual glands; however, small foci of glandular
also is present, this should be reported. In all confluence (cribriform pattern) may be present
these cases, it is important to exclude other [179]. Metaplastic changes including eosino-
lesions such as hyperplasia or carcinoma, so a philic, squamous, and mucinous changes are fre-
comment regarding the absence of these lesions quent [179]. In women with a pretreatment
is generally indicated. diagnosis of well-differentiated carcinoma,
glands displaying complex outlines and conflu-
ency (cribriform and/or papillary patterns) are
Treatment of Hyperplasia often present and may be more pronounced in
and Endometrioid Carcinoma post- rather than pre-treatment samples [179].
Cytoplasm is usually eosinophilic, and the meta-
Hyperplasia, either with or without atypia, may plastic changes are similar to those encountered
be managed by progestin therapy to suppress in progestin-treated CAH [179].
gland growth and then re-biopsied to determine To fully evaluate the response to treatment, it
the efficacy of the therapy. In these cases, it is is important to review the pre-treatment biopsy
important to note whether or not there is evidence and all subsequent posttreatment specimens
of continued persistent hyperplasia and whether available for review [179]. This is important
or not atypia is found. In progestin-treated hyper- because studies found that most cases showed
plasia, the underlying pattern of gland growth resolution of the process, but persistence of cyto-
and cellular differentiation often becomes dis- logic atypia and architectural abnormalities after
torted. Frequently the tissue resembles that seen 6 months of therapy were associated with treat-
in early pregnancy. When there is a well- ment failure. One study, however, did find that
developed decidua-like response, it is difficult to confluent or cribriform gland patterns did not
fully assess the degree of gland complexity, and imply disease persistence or progression and that
the secretory changes of the glands make evalua- only cytologic atypia was important in determin-
tion of cytologic atypia difficult or impossible. ing treatment failure [32]. That study also noted
Generally, cytologic atypia and mitotic activity that eosinophilic and mucinous cytoplasmic
References 165
changes were frequently seen following proges- the biopsy clearly shows a progestin effect, we
tin therapy and occasional cases also showed recommend a brief description of the gland and
squamous change. See Chap. 9 (Endometrial stroma changes followed by the term “consistent
Hyperplasia, Endometrial Intraepithelial with exogenous hormone therapy effect” as the
Carcinoma, and Epithelial Cytoplasmic Change) diagnosis. For example, “Inactive decidualized
and Chap. 10 (Endometrial Carcinoma) for fur- endometrium, consistent with exogenous hor-
ther details. mone therapy effect.”
In some cases the glands and stroma show
some changes that suggest progestin effect but
Infertility Therapy the features are not diagnostic. Usually these
cases show glands with poorly developed secre-
When a biopsy is done during hormone therapy tory changes and plump stromal cells that appear
in the infertile patient, the clinical questions to be partly decidualized. If the clinical history
depend on the hormone used and the indication does not establish progestin use and the changes
for biopsy. Often before beginning treatment, the by themselves are not diagnostic of progestin
clinician would like to rule out a structural cause effect, then it is best to give a descriptive diagno-
for infertility such as endometrial polyps, uterine sis. This diagnosis can reflect the fact that
septa, submucosal leiomyomas, chronic endome- irregular secretory changes are present and sug-
tritis, or endometrial hyperplasia or carcinoma. gest the possibility of progestin effect. A descrip-
During treatment with a drug, such as clomi- tive diagnosis would serve to indicate that the
phene citrate, the reproductive endocrinologist or changes are benign and caused by an imbalance
gynecologist wishes to know whether secretory in the amounts of sex steroid hormones, regard-
phase development has occurred (see Chap. 2). less of their source.
When ovulation-inducing medication is used,
one obvious question is whether the endome-
trium shows secretory changes that reflect ovula-
tion or if the pattern is proliferative. Historically, References
this determination required accurate histologic
dating of secretory endometrium (see Chap. 2). 1. Fritz MA, Speroff L. Clinical gynecologic endo-
crinology and infertility. 8th ed. Philadelphia:
However, when the clinician requests accurate Lippincott Williams & Wilkins; 2010.
dating of secretory phase endometrium, a diagno- 2. Maclennan AH, Broadbent JL, Lester S, Moore
sis of early, mid-, or late secretory phase endome- V. Oral oestrogen and combined oestrogen/progesto-
trium is recommended [180]. gen therapy versus placebo for hot flushes. Cochrane
Database Syst Rev. 2004;18:CD002978.
In addition to the specific situations for hor- 3. Greendale GA, Reboussin BA, Hogan P, Barnabei
mone use given in the preceding, hormone VM, Shumaker S, Johnson S, Barrett-Connor
effects, especially progestin-related changes, can E. Symptom relief and side effects of postmeno-
be found in other biopsy specimens when the pausal hormones: results from the Postmenopausal
Estrogen/Progestin Interventions Trial. Obstet
hormone use is incidental. These cases include Gynecol. 1998;92:982–8.
biopsies performed during evaluation of pelvic 4. Cintron D, Lahr BD, Bailey KR, Santoro N,
pain, pre-hysterectomy sampling, or curettage at Lloyd R, Manson JE, Neal-Perry G, Pal L, Taylor
the time of tubal ligation. Furthermore, progestin HS, Wharton W, Naftolin F, Harman SM, Miller
VM. Effects of oral versus transdermal menopausal
effects can persist for a variable period of one or hormone treatments on self-reported sleep domains
more months after the drug is discontinued. and their association with vasomotor symptoms
Occasionally, some of these cases may represent in recently menopausal women enrolled in the
an endogenous progestin effect with no exoge- Kronos Early Estrogen Prevention Study (KEEPS).
Menopause. 2018;25:145–53.
nous hormone use. Consequently, the finding of 5. Faubion SS, Sood R, Kapoor E. Genitourinary syn-
the histologic changes of apparent progestin drome of menopause: management strategies for the
effect is not negated by the clinical history. When clinician. Mayo Clin Proc. 2017;92:1842–9.
6. Gandhi J, Chen A, Dagur G, Suh Y, Smith N, Cali B, photoepilation therapy alone). Cochrane Database
Khan SA. Genitourinary syndrome of menopause: Syst Rev. 2015;4:CD010334.
an overview of clinical manifestations, pathophysi- 21. Arowojolu AO, Gallo MF, Lopez LM, Grimes
ology, etiology, evaluation, and management. Am J DA. Combined oral contraceptive pills for treat-
Obstet Gynecol. 2016;215:704–11. ment of acne. Cochrane Database Syst Rev.
7. Gartlehner G, Patel SV, Feltner C, Weber RP, Long 2012;6:CD004425.
R, Mullican K, Boland E, Lux L, Viswanathan 22. Iversen L, Sivasubramaniam S, Lee AJ, Fielding S,
M. Hormone therapy for the primary preven- Hannaford PC: Lifetime cancer risk and combined
tion of chronic conditions in postmenopausal oral contraceptives: the Royal College of General
women: evidence report and systematic review for Practitioners’ Oral Contraception Study. Am J
the US Preventive Services Task Force. JAMA. Obstet Gynecol. 2017;216:580.e1–9.
2017;318:2234–49. 23. Chiaffarino F, Parazzini F, La Vecchia C, Marsico S,
8. Schiff I. Menopause. In: Kistner RW, editor. Surace M, Ricci E. Use of oral contraceptives and
Gynecology: principles and practice. 7th ed. uterine fibroids: results from a case-control study. Br
Chicago: Year Book Medical Pub; 1999. J Obstet Gynaecol. 1999;106:857–60.
9. Effects of hormone replacement therapy on endo- 24. Ober WB. Effects of oral and intrauterine adminis-
metrial histology in postmenopausal women. The tration of contraceptives on the uterus. Hum Pathol.
Postmenopausal Estrogen/Progestin Interventions 1977;8:513–27.
(PEPI) Trial. The Writing Group for the PEPI Trial. 25. Dallenbach-Hellweg G. Histopathology of the endo-
JAMA. 1996;275:370–75. metrium. 4th ed. New York: Springer; 1987.
10. Ferenczy A, Gelfand MM. Endometrial histology 26. Deligdisch L. Effects of hormone therapy on the
and bleeding patterns in post-menopausal women endometrium. Mod Pathol. 1993;6:94–106.
taking sequential, combined estradiol and dydroges- 27. Raymond EG, Lovely LP, Chen-Mok M, Seppälä M,
terone. Maturitas. 1997;26:219–26. Kurman RJ, Lessey BA. Effect of the Yuzpe regimen
11. Gomes AM, Baracat EC, Simões MJ, Haidar of emergency contraception on markers of endome-
MA, Focchi GR, Evêncio-Neto J, de Lima trial receptivity. Hum Reprod. 2000;15:2351–5.
GR. Morphologic and morphometric aspects of the 28. Swahn ML, Westlund P, Johannisson E, Bygdeman
endometrium of postmenopausal women before and M. Effect of post-coital contraceptive methods on
after cyclic oestrogen replacement treatment. Eur J the endometrium and the menstrual cycle. Acta
Obstet Gynecol Reprod Biol. 1997;74:79–82. Obstet Gynecol Scand. 1996;75:738–44.
12. Deligdisch L. Hormonal pathology of the endome- 29. Marions L, Hultenby K, Lindell I, Sun X, Ståbi B,
trium. Mod Pathol. 2000;13:285–94. Gemzell Danielsson K. Emergency contraception
13. Grady D, Rubin SM, Petitti DB, Fox CS, Black D, with mifepristone and levonorgestrel: mechanism of
Ettinger B, Ernster VL, Cummings SR. Hormone ther- action. Obstet Gynecol. 2002;100:65–71.
apy to prevent disease and prolong life in postmeno- 30. Vargas MF, Tapia-Pizarro AA, Henríquez SP, Quezada
pausal women. Ann Intern Med. 1992;117:1016–37. M, Salvatierra AM, Noe G, Munroe DJ, Velasquez
14. Whitehead MI, Fraser D. The effects of estrogens LA, Croxatto HB. Effect of single postovulatory
and progestogens on the endometrium. Modern administration of levonorgestrel on gene expression
approach to treatment. Obstet Gynecol Clin N Am. profile during the receptive period of the human endo-
1987;14:299–320. metrium. J Mol Endocrinol. 2012;48:25–36.
15. Furness S, Roberts H, Marjoribanks J, Lethaby 31. Gemzell-Danielsson K, Berger C. P G L L: emer-
A. Hormone therapy in postmenopausal women and gency contraception – mechanisms of action.
risk of endometrial hyperplasia. Cochrane Database Contraception. 2013;87:300–8.
Syst Rev. 2012;8:CD000402. 32. Mentrikoski MJ, Shah AA, Hanley KZ, Atkins
16. Grady D, Gebretsadik T, Kerlikowske K, Ernster V, KA. Assessing endometrial hyperplasia and car-
Petitti D. Hormone replacement therapy and endo- cinoma treated with progestin therapy. Am J Clin
metrial cancer risk: a meta-analysis. Obstet Gynecol. Pathol. 2012;138:524–34.
1995;85:304–13. 33. Miranda MC, Mazur MT. Endometrial squa-
17. Sjögren LL, Mørch LS, Løkkegaard E. Hormone mous metaplasia. An unusual response to proges-
replacement therapy and the risk of endometrial can- tin therapy of hyperplasia. Arch Pathol Lab Med.
cer: a systematic review. Maturitas. 2016;91:25–35. 1995;119:458–60.
18. Steinauer J, Autry AM. Extended cycle combined 34. Dockerty MB, Smith RA, Symmonds
hormonal contraception. Obstet Gynecol Clin N Am. RE. Pseudomalignant endometrial changes induced
2007;34:43–55, viii. by administration of new synthetic progestins. Proc
19. Allais G, Chiarle G, Sinigaglia S, Benedetto Staff Meet Mayo Clin. 1959;34:321–8.
C. Menstrual migraine: a review of current and 35. Cruz-Aquino M, Shenker L, Blaustein
developing pharmacotherapies for women. Expert A. Pseudosarcoma of the endometrium. Obstet
Opin Pharmacother. 2018;19:123–36. Gynecol. 1967;29:93–6.
20. van Zuuren EJ, Fedorowicz Z, Carter B, Pandis 36. Clement PB, Scully RE. Idiopathic postmenopausal
N. Interventions for hirsutism (excluding laser and decidual reaction of the endometrium. A clinico-
References 167
pathologic analysis of four cases. Int J Gynecol medroxyprogesterone acetate and norethindrone ace-
Pathol. 1988;7:152–61. tate on bleeding patterns and endometrial histologic
37. Iezzoni JC, Mills SE. Nonneoplastic endometrial diagnosis. Am J Obstet Gynecol. 1992;167:1–7.
signet-ring cells. Vacuolated decidual cells and stro- 51. Piegsa K, Calder A, Davis JA, McKay-Hart D, Wells
mal histiocytes mimicking adenocarcinoma. Am J M, Bryden F. Endometrial status in post-menopausal
Clin Pathol. 2001;115:249–55. women on long-term continuous combined hor-
38. Critchley HO, Wang H, Jones RL, Kelly RW, mone replacement therapy (Kliofem). A comparative
Drudy TA, Gebbie AE, Buckley CH, McNeilly AS, study of endometrial biopsy, outpatient hysteroscopy
Glasier AF. Morphological and functional features and transvaginal ultrasound. Eur J Obstet Gynecol
of endometrial decidualization following long-term Reprod Biol. 1997;72:175–80.
intrauterine levonorgestrel delivery. Hum Reprod. 52. Sturdee DW. Endometrial morphology and bleed-
1998;13:1218–24. ing patterns as a function of progestogen sup-
39. Bernardini L, Araujo FE, Balmaceda JP. A case plementation. Int J Fertil Menopausal Stud.
of permanent endometrial hypotrophy after long- 1996;41:22–8.
term use of oral contraceptives. Hum Reprod. 53. Udoff L, Langenberg P, Adashi EY. Combined con-
1993;8:543–6. tinuous hormone replacement therapy: a critical
40. Taylor HS, Ohleth K. Using bazedoxifene plus review. Obstet Gynecol. 1995;86:306–16.
conjugated estrogens for treating postmenopausal 54. Fernández-Murga L, Hermenegildo C, Tarín JJ,
women: a comprehensive review. Menopause. García-Pérez M-Á, Cano A. Endometrial response to
2012;19:479–85. concurrent treatment with vaginal progesterone and
41. The NAMS. Hormone therapy position statement transdermal estradiol. Climacteric. 2012;15:455–9.
advisory panel: the 2017 hormone therapy posi- 55. Moyer DL, de Lignieres B, Driguez P, Pez
tion statement of The North American Menopause JP. Prevention of endometrial hyperplasia by pro-
Society. Menopause. 2017;2017(24):728–53. gesterone during long-term estradiol replacement:
42. Lello S, Capozzi A, Scambia G. The tissue-selective influence of bleeding pattern and secretory changes.
estrogen complex (bazedoxifene/conjugated Fertil Steril. 1993;59:992–7.
estrogens) for the treatment of menopause. Int J 56. Bjarnason K, Cerin A, Lindgren R, Weber
Endocrinol. 2017;2017:5064725. T. Adverse endometrial effects during long cycle
43. Gelfand MM, Ferenczy A. A prospective 1-year hormone replacement therapy. Scandinavian Long
study of estrogen and progestin in postmeno- Cycle Study Group. Maturitas. 1999;32:161–70.
pausal women: effects on the endometrium. Obstet 57. Sturdee DW, Ulrich LG, Barlow DH, Wells M,
Gynecol. 1989;74:398–402. Campbell MJ, Vessey MP, Nielsen B, Anderson MC,
44. Ryntz T, Lobo RA. Abnormal uterine bleeding. Bragg AJ. The endometrial response to sequential
In: Lobo RA, Gershenson DM, Lentz GM, Valea and continuous combined oestrogen-progestogen
FA, editors. Comprehensive gynecology. 7th ed. replacement therapy. BJOG. 2000;107:1392–400.
Philadelphia: Elsevier; 2017. p. 621–33. 58. Elliott J, Connor ME, Lashen H. The value of outpa-
45. Feeley KM, Wells M. Hormone replacement therapy tient hysteroscopy in diagnosing endometrial pathol-
and the endometrium. J Clin Pathol. 2001;54:435–40. ogy in postmenopausal women with and without
46. Byrjalsen I, Thormann L, Meinecke B, Riis BJ, hormone replacement therapy. Acta Obstet Gynecol
Christiansen C. Sequential estrogen and progesto- Scand. 2003;82:1112–9.
gen therapy: assessment of progestational effects on 59. Omodei U, Ferrazzi E, Ramazzotto F, Becorpi A,
the postmenopausal endometrium. Obstet Gynecol. Grimaldi E, Scarselli G, Spagnolo D, Spagnolo L,
1992;79:523–8. Torri W. Endometrial evaluation with transvaginal
47. Carranza-Lira S, Martínez-Chéquer JC, Santa ultrasound during hormone therapy: a prospective
Rita MT, Ortiz de la Peña A, Pérez Y, Fernández multicenter study. Fertil Steril. 2004;81:1632–7.
RL. Endometrial changes according to hor- 60. Heikkinen J, Vaheri R, Kainulainen P, Timonen
mone replacement therapy schedule. Menopause. U. Long-term continuous combined hormone
1998;5:86–9. replacement therapy in the prevention of postmeno-
48. Habiba MA, Bell SC, Al-Azzawi F. Endometrial pausal bone loss: a comparison of high- and low-
responses to hormone replacement therapy: histolog- dose estrogen-progestin regimens. Osteoporos Int.
ical features compared with those of late luteal phase 2000;11:929–37.
endometrium. Hum Reprod. 1998;13:1674–82. 61. Nand SL, Webster MA, Baber R, O’Connor
49. Magos AL, Brincat M, Studd JW, Wardle P, V. Bleeding pattern and endometrial changes during
Schlesinger P, O’Dowd T. Amenorrhea and endome- continuous combined hormone replacement therapy.
trial atrophy with continuous oral estrogen and pro- The Ogen/Provera Study Group. Obstet Gynecol.
gestogen therapy in postmenopausal women. Obstet 1998;91:678–84.
Gynecol. 1985;65:496–9. 62. Oosterbaan HP, van Buuren AH, Schram JH, van
50. Hillard TC, Siddle NC, Whitehead MI, Fraser DI, Kempen PJ, Ubachs JM, van Leusden HA, Beyer
Pryse-Davies J. Continuous combined conjugated GP. The effects of continuous combined transder-
equine estrogen-progestogen therapy: effects of mal oestrogen-progestogen treatment on bleeding
patterns and the endometrium in postmenopausal 76. Leslie KK, Walter SA, Torkko K, Stephens JK,
women. Maturitas. 1995;21:211–9. Thompson C, Singh M. Effect of tamoxifen on endo-
63. Varila E, Wahlström T, Rauramo I. A 5-year follow- metrial histology, hormone receptors, and cervical
up study on the use of a levonorgestrel intrauterine cytology: a prospective study with follow-up. Appl
system in women receiving hormone replacement Immunohistochem Mol Morphol. 2007;15:284–93.
therapy. Fertil Steril. 2001;76:969–73. 77. López DML, Fernández YG, Sánchez AV, Alvarez
64. Kurman RJ, Félix JC, Archer DF, Nanavati N, Arce MI, Reynaldo MIL, Delgado RC. Baseline hystero-
J, Moyer DL. Norethindrone acetate and estradiol- scopic assessment of endometrium in asymptomatic
induced endometrial hyperplasia. Obstet Gynecol. postmenopausal women with estrogen receptor-
2000;96:373–9. positive breast cancer. Menopause. 2013;20:64–71.
65. Te Linde RW, Henriksen E. Decidua like changes in 78. Lubián López DM, Orihuela López F, García-Berbel
the endometrium without pregnancy. Am J Obstet Molina L, Boza Novo P, Pozuelo Solís E, Menor
Gynecol. 1940;39:733–49. Almagro D, Comino Delgado R. Endometrial pol-
66. Dallenbach-Hellweg G, Hohagen F. On the problem yps in obese asymptomatic pre and postmenopausal
of premature decidualization induced by the embryo patients with breast cancer: is screening necessary?
transfer. Lab Investig. 1985;52:17A. Gynecol Oncol. 2014;133:56–62.
67. Tamada T, Okagaki T, Maruyama M, Matsumoto 79. Bissett D, Davis JA, George WD. Gynaecological
S. Endometrial histology associated with an intra- monitoring during tamoxifen therapy. Lancet.
uterine contraceptive device. Am J Obstet Gynecol. 1994;344:1244.
1967;98:811–7. 80. Committee Opinion No. 601. Tamoxifen and uterine
68. Bonney WA, Glasser SR, Clewe TH, Noyes RW, cancer. Obstet Gynecol. 2014;123:1394–7.
Cooper CL. Endometrial response to the intrauterine 81. Bertelli G, Valenzano M, Costantini S, Rissone R,
device. Am J Obstet Gynecol. 1966;96:101–13. Angiolini C, Signorini A, Gustavino C. Limited value
69. Mutter GL, Bergeron C, Deligdisch L, Ferenczy of sonohysterography for endometrial screening in
A, Glant M, Merino M, Williams ARW, Blithe asymptomatic, postmenopausal patients treated with
DL. The spectrum of endometrial pathology induced tamoxifen. Gynecol Oncol. 2000;78:275–7.
by progesterone receptor modulators. Mod Pathol. 82. Cohen I, Perel E, Flex D, Tepper R, Altaras MM,
2008;21:591–8. Cordoba M, Beyth Y. Endometrial pathology in
70. Williams ARW, Bergeron C, Barlow DH, Ferenczy postmenopausal tamoxifen treatment: compari-
A. Endometrial morphology after treatment of uter- son between gynaecologically symptomatic and
ine fibroids with the selective progesterone receptor asymptomatic breast cancer patients. J Clin Pathol.
modulator, ulipristal acetate. Int J Gynecol Pathol. 1999;52:278–82.
2012;31:556–69. 83. Marchesoni D, Driul L, Fabiani G, Di Loreto
71. Nogales FF, Crespo-Lora V, Cruz-Viruel N, Chamorro- C, Cataldi P, Mozzanega B. Endometrial histo-
Santos C, Bergeron C. Endometrial changes in surgi- logic changes in post-menopausal breast cancer
cal specimens of perimenopausal patients treated patients using tamoxifen. Int J Gynaecol Obstet.
with ulipristal acetate for uterine leiomyomas. 2001;75:257–62.
Int J Gynecol Pathol. 2017 Sep 13. doi: 10.1097/ 84. Seoud M, Shamseddine A, Khalil A, Salem Z, Saghir
PGP.0000000000000450. [Epub ahead of print]. N, Bikhazi K, Bitar N, Azar G, Kaspar H. Tamoxifen
72. Donnez J, Donnez O, Matule D, Ahrendt H-J, and endometrial pathologies: a prospective study.
Hudecek R, Zatik J, Kasilovskiene Z, Dumitrascu Gynecol Oncol. 1999;75:15–9.
MC, Fernandez H, Barlow DH, Bouchard P, Fauser 85. Ozşener S, Ozaran A, Itil I, Dikmen Y. Endometrial
BCJM, Bestel E, Loumaye E. Long-term medical pathology of 104 postmenopausal breast cancer
management of uterine fibroids with ulipristal ace- patients treated with tamoxifen. Eur J Gynaecol
tate. Fertil Steril. 2016;105:165–73.e4. Oncol. 1998;19:580–3.
73. Vosse M, Renard F, Coibion M, Neven P, Nogaret 86. LE Donne M, Alibrandi A, Ciancimino L, Azzerboni
JM, Hertens D. Endometrial disorders in 406 breast A, Chiofalo B, Triolo O. Endometrial pathology in
cancer patients on tamoxifen: the case for less inten- breast cancer patients: effect of different treatments
sive monitoring. Eur J Obstet Gynecol Reprod Biol. on ultrasonographic, hysteroscopic and histological
2002;101:58–63. findings. Oncol Lett. 2013;5:1305–10.
74. Goldstein SR. The effect of SERMs on the endome- 87. Cohen I, Figer A, Altaras MM, Tepper R, Shapira
trium. Ann N Y Acad Sci. 2001;949:237–42. J, Cordoba M, Yigael D, Arbel Y, Beyth Y. Common
75. Cohen I, Altaras MM, Shapira J, Tepper R, Rosen endometrial decidual reaction in postmenopausal
DJ, Cordoba M, Zalel Y, Figer A, Yigael D, Beyth breast cancer patients treated with tamoxifen and
Y. Time-dependent effect of tamoxifen therapy on progestogens. Int J Gynecol Pathol. 1996;15:17–22.
endometrial pathology in asymptomatic postmeno- 88. Cohen I, Shapira J, Altaras M, Cordoba M, Rosen
pausal breast cancer patients. Int J Gynecol Pathol. D, Beyth Y. Endometrial decidual changes in a
1996;15:152–7. postmenopausal woman treated with tamoxi-
References 169
fen and megestrol acetate. Br J Obstet Gynaecol. breast cancer patients with tamoxifen treatment. Int
1992;99:773–4. J Gynecol Pathol. 1999;18:304–9.
89. Andersson M, Storm HH, Mouridsen HT. Incidence 103. Barakat RR, Wong G, Curtin JP, Vlamis V, Hoskins
of new primary cancers after adjuvant tamoxifen WJ. Tamoxifen use in breast cancer patients who
therapy and radiotherapy for early breast cancer. J subsequently develop corpus cancer is not associ-
Natl Cancer Inst. 1991;83:1013–7. ated with a higher incidence of adverse histologic
90. Segna RA, Dottino PR, Deligdisch L, Cohen features. Gynecol Oncol. 1994;55:164–8.
CJ. Tamoxifen and endometrial cancer. Mt Sinai J 104. Fisher B, Costantino JP, Wickerham DL, Redmond
Med. 1992;59:416–8. CK, Kavanah M, Cronin WM, Vogel V, Robidoux
91. Magriples U, Naftolin F, Schwartz PE, Carcangiu A, Dimitrov N, Atkins J, Daly M, Wieand S, Tan-
ML. High-grade endometrial carcinoma in Chiu E, Ford L, Wolmark N. Tamoxifen for preven-
tamoxifen- treated breast cancer patients. J Clin tion of breast cancer: report of the National Surgical
Oncol. 1993;11:485–90. Adjuvant Breast and Bowel Project P-1 Study. J Natl
92. Gusberg SB. Tamoxifen for breast cancer: associ- Cancer Inst. 1998;90:1371–88.
ated endometrial cancer. Cancer. 1990;65:1463–4. 105. Wright JD, Desai VB, Chen L, Burke WM, Tergas
93. Fornander T, Rutqvist LE, Cedermark B, Glas U, AI, Hou JY, Accordino M, Ananth CV, Neugut AI,
Mattsson A, Silfverswärd C, Skoog L, Somell A, Hershman DL. Utilization of gynecologic services
Theve T, Wilking N. Adjuvant tamoxifen in early in women with breast cancer receiving hormonal
breast cancer: occurrence of new primary cancers. therapy. Am J Obstet Gynecol. 2017;217:59.
Lancet. 1989;1:117–20. e1–12.
94. Seoud MA, Johnson J, Weed JC. Gynecologic 106. Wickerham DL, Fisher B, Wolmark N, Bryant
tumors in tamoxifen-treated women with breast can- J, Costantino J, Bernstein L, Runowicz
cer. Obstet Gynecol. 1993;82:165–9. CD. Association of tamoxifen and uterine sarcoma.
95. Carcangiu ML. Uterine pathology in tamoxifen- J Clin Oncol. 2002;20:2758–60.
treated patients with breast cancer. Anat Pathol. 107. Clement PB, Oliva E. Young RH: üü adenosarcoma
1997;2:53–70. of the uterine corpus associated with tamoxifen ther-
96. Braithwaite RS, Chlebowski RT, Lau J, George apy: a report of six cases and a review of tamoxifen-
S, Hess R, Col NF. Meta-analysis of vascular and associated endometrial lesions. Int J Gynecol Pathol.
neoplastic events associated with tamoxifen. J Gen 1996;15:222–9.
Intern Med. 2003;18:937–47. 108. McCluggage WG, Abdulkader M, Price JH, Kelehan
97. Fisher B, Costantino JP, Redmond CK, Fisher ER, P, Hamilton S, Beattie J, Al-Nafussi A. Uterine
Wickerham DL, Cronin WM. Endometrial cancer carcinosarcomas in patients receiving tamoxi-

in tamoxifen-treated breast cancer patients: find- fen. A report of 19 cases. Int J Gynecol Cancer.
ings from the National Surgical Adjuvant Breast and 2000;10:280–4.
Bowel Project (NSABP) B-14. J Natl Cancer Inst. 109. Swerdlow AJ, Jones ME. British Tamoxifen Second
1994;86:527–37. Cancer Study Group: Tamoxifen treatment for breast
98. Davies C, Pan H, Godwin J, Gray R, Arriagada R, cancer and risk of endometrial cancer: a case-control
Raina V, et al. Adjuvant Tamoxifen: Longer Against study. J Natl Cancer Inst. 2005;97:375–84.
Shorter (ATLAS) Collaborative Group: long-term 110. Matsuo K, Ross MS, Bush SH, Yunokawa M, Blake
effects of continuing adjuvant tamoxifen to 10 years EA, Takano T, Ueda Y, Baba T, Satoh S, Shida M,
versus stopping at 5 years after diagnosis of oestro- Ikeda Y, Adachi S, Yokoyama T, Takekuma M,
gen receptor-positive breast cancer: ATLAS, a ran- Takeuchi S, Nishimura M, Iwasaki K, Yanai S,
domised trial. Lancet. 2013;381:805–16. Klobocista MM, Johnson MS, Machida H, Hasegawa
99. Dallenbach-Hellweg G, Schmidt D, Hellberg P, K, Miyake TM, Nagano T, Pejovic T, Shahzad MM,
Bourne T, Kreuzwieser E, Dören M, Rydh W, Im DD, Omatsu K, Ueland FR, Kelley JL, Roman
Rudenstam G, Granberg S. The endometrium in LD. Tumor characteristics and survival outcomes of
breast cancer patients on tamoxifen. Arch Gynecol women with tamoxifen-related uterine carcinosar-
Obstet. 2000;263:170–7. coma. Gynecol Oncol. 2017;144:329–35.
100. Deligdisch L, Kalir T, Cohen CJ, de Latour M, Le 111. Nuovo MA, Nuovo GJ, McCaffrey RM, Levine
Bouedec G, Penault-Llorca F. Endometrial histopa- RU, Barron B, Winkler B. Endometrial polyps in
thology in 700 patients treated with tamoxifen for postmenopausal patients receiving tamoxifen. Int J
breast cancer. Gynecol Oncol. 2000;78:181–6. Gynecol Pathol. 1989;8:125–31.
101. Bland AE, Calingaert B, Secord AA, Lee PS, Valea 112. Corley D, Rowe J, Curtis MT, Hogan WM, Noumoff
FA, Berchuck A, Soper JT, Havrilesky L. Relationship JS, Livolsi VA. Postmenopausal bleeding from
between tamoxifen use and high risk endometrial can- unusual endometrial polyps in women on chronic
cer histologic types. Gynecol Oncol. 2009;112:150–4. tamoxifen therapy. Obstet Gynecol. 1992;79:111–6.
102. Cohen I, Azaria R, Fishman A, Tepper R, Shapira 113. Lahti E, Blanco G, Kauppila A, Apaja-Sarkkinen M,
J, Beyth Y. Endometrial cancers in postmenopausal Taskinen PJ, Laatikainen T. Endometrial changes
in postmenopausal breast cancer patients receiving eight years in a clinical trial setting. Curr Med Res
tamoxifen. Obstet Gynecol. 1993;81:660–4. Opin. 2005;21:1441–52.
114. Kennedy MM, Baigrie CF, Manek S. Tamoxifen and 127. Boss SM, Huster WJ, Neild JA, Glant MD, Eisenhut
the endometrium: review of 102 cases and compari- CC, Draper MW. Effects of raloxifene hydrochloride
son with HRT-related and non-HRT-related endo- on the endometrium of postmenopausal women. Am
metrial pathology. Int J Gynecol Pathol. 1999;18: J Obstet Gynecol. 1997;177:1458–64.
130–7. 128. Davies GC, Huster WJ, Shen W, Mitlak B, Plouffe
115. Cohen I, Azaria R, Bernheim J, Shapira J, Beyth L, Shah A, Cohen FJ. Endometrial response to ral-
Y. Risk factors of endometrial polyps resected from oxifene compared with placebo, cyclical hormone
postmenopausal patients with breast carcinoma replacement therapy, and unopposed estrogen in post-
treated with tamoxifen. Cancer. 2001;92:1151–5. menopausal women. Menopause. 1999;6:188–95.
116. Liedman R, Lindahl B, Andolf E, Willen R, Ingvar 129. Delmas PD, Bjarnason NH, Mitlak BH, Ravoux AC,
C, Ranstam J. Disaccordance between estimation of Shah AS, Huster WJ, Draper M, Christiansen C. Effects
endometrial thickness as measured by transvaginal of raloxifene on bone mineral density, serum choles-
ultrasound compared with hysteroscopy and directed terol concentrations, and uterine endometrium in post-
biopsy in breast cancer patients treated with tamoxi- menopausal women. N Engl J Med. 1997;337:1641–7.
fen. Anticancer Res. 2000;20:4889–91. 130. Goldstein SR, Scheele WH, Rajagopalan SK, Wilkie
117. Sasaki LMP, Andrade KRC, Figueiredo ACMG, JL, Walsh BW, Parsons AK. A 12-month compara-
Wanderley MDS, Pereira MG. Factors associated tive study of raloxifene, estrogen, and placebo on
with malignancy in hysteroscopically resected the postmenopausal endometrium. Obstet Gynecol.
endometrial polyps: a systematic review and meta- 2000;95:95–103.
analysis. J Minim Invasive Gynecol. 2018;25:1–9. 131. DeMichele A, Troxel AB, Berlin JA, Weber AL,
118. Schlesinger C, Kamoi S, Ascher SM, Kendell M, Bunin GR, Turzo E, Schinnar R, Burgh D, Berlin M,
Lage JM, Silverberg SG. Endometrial polyps: a Rubin SC, Rebbeck TR, Strom BL. Impact of raloxi-
comparison study of patients receiving tamoxi- fene or tamoxifen use on endometrial cancer risk: a
fen with two control groups. Int J Gynecol Pathol. population-based case-control study. J Clin Oncol.
1998;17:302–11. 2008;26:4151–9.
119. Biron-Shental T, Tepper R, Fishman A, Shapira J, 132. Benda JA. Clomiphene’s effect on endometrium in
Cohen I. Recurrent endometrial polyps in postmeno- infertility. Int J Gynecol Pathol. 1992;11:273–82.
pausal breast cancer patients on tamoxifen. Gynecol 133. Birkenfeld A, Navot D, Levij IS, Laufer N,
Oncol. 2003;90:382–6. Beier-Hellwig K, Goecke C, Schenker JG, Beier
120. Di Benedetto L, Giovanale V, Caserta D. Endometrial HM. Advanced secretory changes in the prolifera-
tubal metaplasia in a young puerperal woman after tive human endometrial epithelium following clomi-
breast cancer. Int J Clin Exp Pathol. 2015;8:7610–3. phene citrate treatment. Fertil Steril. 1986;45:462–8.
121. Cohen I, Bernheim J, Azaria R, Tepper R, Sharony 134. Thatcher SS, Donachie KM, Glasier A, Hillier SG,
R, Beyth Y. Malignant endometrial polyps in Baird DT. The effects of clomiphene citrate on
postmenopausal breast cancer tamoxifen-treated the histology of human endometrium in regularly
patients. Gynecol Oncol. 1999;75:136–41. cycling women undergoing in vitro fertilization.
122. McCluggage WG, Sumathi VP, McManus Fertil Steril. 1988;49:296–301.
DT. Uterine serous carcinoma and endometrial 135. Li TC, Warren MA, Murphy C, Sargeant S, Cooke
intraepithelial carcinoma arising in endometrial pol- ID. A prospective, randomised, cross-over study
yps: report of 5 cases, including 2 associated with comparing the effects of clomiphene citrate and
tamoxifen therapy. Hum Pathol. 2003;34:939–43. cyclofenil on endometrial morphology in the
123. Dominick S, Hickey M, Chin J, Su HI. Levonorgestrel luteal phase of normal, fertile women. Br J Obstet
intrauterine system for endometrial protection in Gynaecol. 1992;99:1008–13.
women with breast cancer on adjuvant tamoxifen. 136. Triwitayakorn A, Suwajanakorn S, Triratanachat S,
Cochrane Database Syst Rev. 2015;12:CD007245. Sampatanukul P, Pruksananonda K, Sereepapong
124. Cummings SR, Eckert S, Krueger KA, Grady D, W. Effects of initiation day of clomiphene citrate on
Powles TJ, Cauley JA, Norton L, Nickelsen T, the endometrium of women with regular menstrual
Bjarnason NH, Morrow M, Lippman ME, Black cycles. Fertil Steril. 2002;78:102–7.
D, Glusman JE, Costa A, Jordan VC. The effect of 137. Floyd WS. Danazol: endocrine and endometrial
raloxifene on risk of breast cancer in postmeno- effects. Int J Fertil. 1980;25:75–80.
pausal women: results from the MORE randomized 138. Fedele L, Marchini M, Bianchi S, Baglioni A,
trial. Multiple Outcomes of Raloxifene Evaluation. Bocciolone L, Nava S. Endometrial patterns dur-
JAMA. 1999;281:2189–97. ing danazol and buserelin therapy for endometrio-
125. Goldstein SR. Update on raloxifene to pre- sis: comparative structural and ultrastructural study.
vent endometrial- breast cancer. Eur J Cancer. Obstet Gynecol. 1990;76:79–84.
2000;36(Suppl 4):S54–6. 139. Beaumont H, Augood C, Duckitt K, Lethaby
126. Martino S, Disch D, Dowsett SA, Keech CA, A. Danazol for heavy menstrual bleeding. Cochrane
Mershon JL. Safety assessment of raloxifene over Database Syst Rev. 2007;3:CD001017.
References 171
140. Robins JC. Therapies for the treatment of abnor- rionic gonadotropin for in vitro fertilization. Fertil
mal uterine bleeding. Curr Womens Health Rep. Steril. 1984;41:31–5.
2001;1:196–201. 155. Sharma V, Whitehead M, Mason B, Pryse-Davies
141. Tan YH, Lethaby A. Pre-operative endometrial thin- J, Ryder T, Dowsett M, Campbell S, Collins
ning agents before endometrial destruction for heavy W. Influence of superovulation on endometrial
menstrual bleeding. Cochrane Database Syst Rev. and embryonic development. Fertil Steril.
2013;11:CD001124. 1990;53:822–9.
142. Bradley LD, Gueye N-A. The medical management 156. Child TJ, Gulekli B, Sylvestre C, Tan
of abnormal uterine bleeding in reproductive-aged SL. Ultrasonographic assessment of endometrial
women. Am J Obstet Gynecol. 2016;214:31–44. receptivity at embryo transfer in an in vitro matura-
143. Soh E, Sato K. Clinical effects of danazol on endo- tion of oocyte program. Fertil Steril. 2003;79:656–8.
metrial hyperplasia in menopausal and postmeno- 157. Elizur SE, Son W-Y, Yap R, Gidoni Y, Levin D,
pausal women. Cancer. 1990;66:983–8. Demirtas E, Tan S-L. Comparison of low-dose
144. Tamaoka Y, Orikasa H, Sumi Y, Sakakura K, Kamei human menopausal gonadotropin and micronized
K, Nagatani M, Ezawa S. Direct effect of danazol 17beta-estradiol supplementation in in vitro matura-
on endometrial hyperplasia in adenomyotic women: tion cycles with thin endometrial lining. Fertil Steril.
treatment with danazol containing intrauterine 2009;92:907–12.
device. Hum Cell. 2000;13:127–33. 158. Casper RF. Clinical uses of gonadotropin-releasing
145. Wentz AC, Jones GS, Sapp KC, King hormone analogues. CMAJ. 1991;144:153–8.
TM. Progestational activity of danazol in the 159. Lethaby A, Puscasiu L, Vollenhoven B. Preoperative
human female subject. Am J Obstet Gynecol. medical therapy before surgery for uterine fibroids.
1976;126:378–84. Cochrane Database Syst Rev. 2017;11:CD000547.
146. Marchini M, Fedele L, Bianchi S, Di Nola G, Nava 160. Lemay A, Jean C, Faure N. Endometrial histology
S, Vercellini P. Endometrial patterns during ther- during intermittent intranasal luteinizing hormone-
apy with danazol or gestrinone for endometriosis: releasing hormone (LH-RH) agonist sequentially
structural and ultrastructural study. Hum Pathol. combined with an oral progestogen as an antiovulatory
1992;23:51–6. contraceptive approach. Fertil Steril. 1987;48:775–82.
147. Jeppsson S, Mellquist P, Rannevik G. Short-term 161. Spitz IM, Bardin CW. Mifepristone (RU 486)—a
effects of danazol on endometrial histology. Acta modulator of progestin and glucocorticoid action. N
Obstet Gynecol Scand Suppl. 1984;123:41–4. Engl J Med. 1993;329:404–12.
148. Fedele L, Bianchi S, Marchini M, Di Nola 162. Johannisson E, Oberholzer M, Swahn ML,
G. Histological impact of medical therapy – clinical Bygdeman M. Vascular changes in the human
implications. Br J Obstet Gynaecol. 1995;102(Suppl endometrium following the administration of the
12):8–11. progesterone antagonist RU 486. Contraception.
149. Brooks PG, Serden SP, Davos I. Hormonal inhibition 1989;39:103–17.
of the endometrium for resectoscopic endometrial 163. Swahn ML, Bygdeman M. The effect of the anti-
ablation. Am J Obstet Gynecol. 1991;164:1601–6; progestin RU 486 on uterine contractility and sen-
discussion 1606–1608. sitivity to prostaglandin and oxytocin. Br J Obstet
150. Reshef E, Segars JH, Hill GA, Pridham DD, Gynaecol. 1988;95:126–34.
Yussman MA, Wentz AC. Endometrial inadequacy 164. Spitz IM. Mifepristone: where do we come from and
after treatment with human menopausal gonado- where are we going? Clinical development over a
tropin/human chorionic gonadotropin. Fertil Steril. quarter of a century. Contraception. 2010;82:442–52.
1990;54:1012–6. 165. Croxatto HB, Kovács L, Massai R, Resch BA,
151. Bonhoff A, Naether O, Johannisson E, Bohnet Fuentealba B, Salvatierra AM, Croxatto HD, Zalányi
HG. Morphometric characteristics of endometrial S, Viski S, Krenács L. Effects of long-term low-dose
biopsies after different types of ovarian stimulation mifepristone on reproductive function in women.
for infertility treatment. Fertil Steril. 1993;59:560–6. Hum Reprod. 1998;13:793–8.
152. Campbell BF, Phipps WR, Nagel TC, Tagatz 166. Croxatto HB, Salvatierra AM, Croxatto HD,
GE. Endometrial biopsies during treatment with sub- Fuentealba B. Effects of continuous treatment with
cutaneous pulsatile gonadotropin-releasing hormone low dose mifepristone throughout one menstrual
and luteal-phase human chorionic gonadotropin. Int cycle. Hum Reprod. 1993;8:201–7.
J Fertil. 1988;33:329–33. 167. Swahn ML, Bygdeman M, Cekan S, Xing S,
153. Sterzik K, Dallenbach C, Schneider V, Sasse V, Masironi B, Johannisson E. The effect of RU 486
Dallenbach-Hellweg G. In vitro fertilization: administered during the early luteal phase on bleed-
the degree of endometrial insufficiency varies ing pattern, hormonal parameters and endometrium.
with the type of ovarian stimulation. Fertil Steril. Hum Reprod. 1990;5:402–8.
1988;50:457–62. 168. Li TC, Dockery P, Thomas P, Rogers AW, Lenton
154. Garcia JE, Acosta AA, Hsiu JG, Jones HW. Advanced EA, Cooke ID. The effects of progesterone recep-
endometrial maturation after ovulation induction tor blockade in the luteal phase of normal fertile
with human menopausal gonadotropin/human cho- women. Fertil Steril. 1988;50:732–42.
169. Batista MC, Cartledge TP, Zellmer AW, Merino low-dose mifepristone on the endometrium. Hum
MJ, Axiotis C, Loriaux DL, Nieman LK. Delayed Reprod. 2003;18:61–8.
endometrial maturation induced by daily admin- 175. Newfield RS, Spitz IM, Isacson C, New MI. Long-
istration of the antiprogestin RU 486: a potential term mifepristone (RU486) therapy resulting in
new contraceptive strategy. Am J Obstet Gynecol. massive benign endometrial hyperplasia. Clin
1992;167:60–5. Endocrinol. 2001;54:399–404.
170. Cameron ST, Critchley HO, Thong KJ, Buckley CH, 176. Eisinger SH, Meldrum S, Fiscella K, le Roux HD,
Williams AR, Baird DT. Effects of daily low dose Guzick DS. Low-dose mifepristone for uterine leio-
mifepristone on endometrial maturation and prolif- myomata. Obstet Gynecol. 2003;101:243–50.
eration. Hum Reprod. 1996;11:2518–26. 177. Horne FM, Blithe DL. Progesterone receptor modu-
171. Heikinheimo O, Hsiu JG, Gordon K, Kim S, Williams lators and the endometrium: changes and conse-
RF, Gibbons WE, Hodgen GD. Endometrial effects quences. Hum Reprod Update. 2007;13:567–80.
of RU486 in primates – antiproliferative action 178. Koering MJ, Healy DL, Hodgen GD. Morphologic
despite signs of estrogen action and increased response of endometrium to a progesterone recep-
cyclin-B expression. J Steroid Biochem Mol Biol. tor antagonist, RU486, in monkeys. Fertil Steril.
1996;59:179–90. 1986;45:280–7.
172. Gemzell-Danielsson K, Westlund P, Johannisson 179. Wheeler DT, Bristow RE, Kurman RJ. Histologic
E, Swahn ML, Bygdeman M, Seppälä M. Effect of alterations in endometrial hyperplasia and well-
low weekly doses of mifepristone on ovarian func- differentiated carcinoma treated with progestins. Am
tion and endometrial development. Hum Reprod. J Surg Pathol. 2007;31:988–98.
1996;11:256–64. 180. Parra-Herran C, Cesari M, Djordjevic B, Grondin K,
173. Jimenez-Moreno V, Billeter M, Liu CQ, Gordon Kinloch M, Köbel M, Pirzada A, Plotkin A, Gilks
K, Mahony M. Local effects of mifepristone on CB. Canadian consensus-based and evidence-based
the nonhuman primate endometrium. Fertil Steril. guidelines for benign endometrial pathology report-
2000;74:134–40. ing in biopsy material. Int J Gynecol Pathol. 2018
174. Baird DT, Brown A, Critchley HOD, Williams AR, Apr 19. doi:10.1097/PGP.0000000000000481.
Lin S, Cheng L. Effect of long-term treatment with [Epub ahead of print].
Endometritis
7
Contents
Nonspecific Endometritis 174
Specific Infections 179
Differential Diagnosis 185
References 196
Endometritis is a persistent inflammation of the between chronic endometritis and endometrio-

endometrial mucosa often caused by pathogens sis, as women affected by the latter were more
ascending into the uterine cavity. Accordingly, frequently found to have chronic endometrial
it may accompany pelvic inflammatory disease inflammation [7, 8]. Although it has been esti-
of the upper genital tract [1, 2]. Common infec- mated this disorder affects approximately
tious agents associated with chronic endometri- 10–11% of women based on biopsies of patients
tis are Chlamydia, Mycoplasma/Ureaplasma, undergoing hysterectomy for benign gyneco-
Enterococcus faecalis, Escherichia coli, and logic conditions [9]; the prevalence of endome-
Streptococcus agalactiae, among others [3]. tritis has been reported to be as high as 72%
Chronic endometritis may also be associated based on analysis of endometrial sampling of
with a recent pregnancy, either an abortion or a women suspected to have pelvic inflammatory
term pregnancy [4–6]. Instrumentation, such as disease [10] and 57% in women with unex-
a prior biopsy, an intrauterine contraceptive plained infertility [11].
device, cervical stenosis, and the presence of a Since endometritis can be entirely asymptom-
structural lesion such as a polyp, leiomyoma, atic or oligosymptomatic, assessment of preva-
hyperplasia, or carcinoma have all been docu- lence studies in the general population is
mented as possible causes or triggering factors problematic. When symptomatic, however, some
in the development of endometritis [6]. Any of the common presentations include abnormal
process that dilates the internal cervical os can uterine bleeding, dyspareunia, pelvic pain, or
lead to endometritis as an ascending infection leukorrhea. Chronic endometritis has also been
from the lower genital tract. Some recent stud- frequently identified during the investigation and
ies have suggested an etiopathogenetic link workup for infertility, recurrent miscarriages,

174 7 Endometritis
and in vitro fertilization embryo transfer with ing this diagnosis on H&E sections. Some investi-
implantation failure. It is also found in associa- gators have suggested the presence of eosinophils
tion with obstetrical and neonatal complications as a useful marker of chronic endometritis and his-
such as spontaneous abortions and retained prod- tologic indicator of the need to diligently search
ucts of conception [12–14]. Recent studies have for plasma cells [16], although eosinophils can be
suggested the persistent inflammation caused by present without inflammation, too.
chronic endometritis alters the normal hormonal In addition to inflammatory cells, there is a
response of endometrial stromal cells during the constellation of histologic findings that facili-
decidualization process in early pregnancy tate recognition of endometrial inflammation
stages, hence the increased rates of failed implan- (Table 7.1). The other morphologic changes
tation [15]. In our experience, however, endome- include reactive stroma, epithelial changes,
tritis is an infrequent diagnosis, and the prevalence abnormal glandular development, and evidence
of this disorder appears to vary greatly depending of glandular and stromal breakdown [17].
on the patient population and practice setting.
Endometrial inflammation often is nonspe-
cific and rarely has morphologic features that Inflammatory Cells
indicate a definite etiology. The nonspecific
forms of endometritis traditionally have been Plasma cells are the critical histologic feature for
separated into chronic and acute forms, depend- the diagnosis of endometritis [5, 6, 18, 19]. Their
ing on the type of inflammatory infiltrate; most presence is required to establish the diagnosis of
are referred to as chronic nonspecific endome- chronic endometritis, because, in contrast to lym-
tritis. The inflammatory infiltrate often is a phocytes, they are not present in normal endome-
mixed acute and chronic inflammatory process, trium. Within the inflammatory infiltrate, plasma
however, and neutrophils, as well as plasma cells may be diffuse (Fig. 7.1) and easily recog-
cells and lymphocytes can be present. Rigorous nizable but more commonly are focal and widely
separation of the type of inflammatory process dispersed. Plasma cells generally are most
is less important than recognition of the pres- numerous in the periglandular and subepithelial
ence of inflammation. Acute endometrial stroma and around lymphoid aggregates
inflammation is relatively infrequent except for (Fig. 7.2). Plasma cells should be readily identifi-
puerperal-related infections, and these latter able before a diagnosis of chronic endometritis is
cases rarely come to biopsy or curettage. established unless associated features of inflam-
mation are clearly present (see later). The diag-
nosis of endometritis should not rest on the
Nonspecific Endometritis finding of an apparent plasma cell, however, in
endometrium that otherwise appears normal. As
Endometritis may be diffuse or focal and can range previously mentioned, plasmacytoid or pre-
from a subtle inflammatory infiltrate to a pro- decidualized stromal cells closely mimic plasma
nounced inflammatory reaction. Although endo- cells (Fig. 7.3), the latter which should display an
metritis typically shows a pattern of a mixed eccentric nucleus with heterochromatin
inflammatory infiltrate containing plasma cells, distributed in a “clock face” arrangement. In

lymphocytes, occasional neutrophils (polymor- cases in which the plasma cell infiltrate appears
phonuclear leukocytes), and eosinophils, the pri- subtle or equivocal, the background pattern is as
mary histologic hallmark of this process, as important as the quantity of plasma cells for
discussed in the next section, is plasma cells establishing the diagnosis of endometritis
(Fig. 7.1). Morphologic overlap between plasma (Figs. 7.4 and 7.5). The number of plasma cells
cells and plasmacytoid-appearing pre-decidualized does not appear to correlate with the severity of
stromal cells adds to the complexity of establish- the lesion [4].
Fig. 7.1 Nonspecific endometritis. A dense plasma cell (“cartwheel”) chromatin pattern and perinuclear clear
infiltrate is seen in this image. Plasma cells can be recog- zone (hof). The cytoplasm is deep blue to purple
nized by their eccentrically located nuclei, clockface
Table 7.1 Morphologic features of nonspecific (immunoglobulin cytoplasmic accumulation)

endometritis facilitate plasma cell recognition.
Plasma cell infiltrate While plasma cells may be the predominant
Increased number of lymphocytes and lymphoid inflammatory component of endometritis, more
follicles severe inflammation commonly shows a mixed
Variable presence of neutrophils in surface epithelium
and glands
inflammatory infiltrate. Often the inflammatory
Reactive stromal response infiltrate includes numerous lymphocytes that
Altered gland development tend to concentrate in the subepithelial stroma
Breakdown and bleeding (Fig. 7.5). Lymphoid follicles become prominent
and may show germinal centers; larger trans-
formed lymphocytes and immunoblasts also may
Normal endometrial stromal cells, especially be interspersed with scattered surrounding
pre-decidualized cells in the late secretory phase, plasma cells [20–22].
can resemble plasma cells, having eccentric Neutrophils as a part of the inflammatory infil-
nuclei and a pale perinuclear zone. The plasma trate indicate an acute process. The neutrophilic
cell, however, is identified by its distinctive, inflammatory population typically infiltrates the
clumped chromatin arrangement yielding a cart- surface epithelium and extends into gland
wheel pattern. Occasionally, Russell bodies lumens, sometimes forming microabscesses in
176 7 Endometritis
Fig. 7.2 Nonspecific chronic endometritis. Plasma cells and lymphocytes surround a small endometrial gland. The
stromal cells are reactive, with elongated, spindle-shaped nuclei, and they swirl around the gland
the glands (Fig. 7.6) [17–19]. Neutrophils, how- patient (Fig. 7.7). Patients with pregnancy-related
ever, also can be present in menstrual endome- acute inflammation rarely come to biopsy or
trium, where foci of glandular and stromal curettage, however. When acute endometritis is
breakdown are present, too, without signifying present, there is a neutrophilic infiltrate in the
an infectious process. Additionally, intraluminal glands with microabscess formation and infiltra-
neutrophils are occasionally seen in endome- tion of neutrophils into the surface epithelium.
trium without a specific pathology. Therefore, Marked inflammation also will result in formation
like the presence of lymphocytes, the presence of of granulation tissue with a network of small ves-
neutrophils alone is not sufficient to indicate sels in a fibroblastic stroma. Acute e ndometritis
endometritis. The distribution pattern of these also can occur in postmenopausal women with
cells in the endometrium and the accompanying severe cervical stenosis. In this case the specimen
cellular infiltrate, usually including plasma cells, usually consists of a purulent exudate which may
must be considered before making the diagnosis show a few remnants of atrophic endometrial
of endometritis. In order to be pathologic, neutro- epithelium.
phils must invade the epithelial glandular On occasion, eosinophils may be present as a
compartment. part of the inflammatory infiltrate [9]. Although
Acute endometritis without a chronic (plasma they are not normally present in the endome-
cell) component is extremely unusual and occurs trium, their presence alone does not necessarily
most frequently in the postpartum or postabortal indicate endometritis [16]. Like lymphocytes or
Fig. 7.3 Plasmacytoid-appearing pre-decidualized stro- changes and mixed inflammatory infiltrate, assists in the
mal cells can mimic plasma cells. Identification of other diagnosis of endometritis
features of endometritis, particularly the reactive stromal
neutrophils, eosinophils should be present in a these cells in the stroma surrounded by plasma
background of inflammatory changes to be a cells and lymphocytes. When the histiocytes
component of endometritis. Eosinophilic infil- develop abundant, foamy cytoplasm, the pro-
trates also can occur following curettage, appar- cess becomes xanthogranulomatous [10, 19,
ently as a result of instrumentation [23]. 25–27].
Sometimes occasional eosinophils are found with
no other associated abnormality [16]. After curet-
tage, endometrial ablation, or biopsy, an eosino- Stromal Changes
philic xanthogranulomatous and/or foreign body
granulomatous endometritis can occur, none of With endometritis the stroma typically shows
which appear to be of clinical significance reactive changes [4, 19]. Stromal cells become
(Figs. 7.8, 7.9 and 7.10) [23, 24]. spindle-shaped, resembling fibroblasts, and are
Endometritis also can have a component of elongate and bipolar, in contrast to the rounded,
histiocytes. Usually these cells are widely ovoid shape of the nonreactive stromal cell
distributed in a mixed inflammatory infiltrate. (Fig. 7.11). The reactive process is also charac-
Hemosiderin-laden stromal cells and histiocytes terized by a swirling, interlacing pattern of the
often are interspersed [4, 6]. Sometimes histio- spindle cells, which may form radial, “pin-
cytes can be prominent, with large aggregates of wheel” arrangements (Figs. 7.2, 7.4). Plasma
178 7 Endometritis
Fig. 7.4 Nonspecific chronic endometritis. A dense inflammatory infiltrate composed of plasma cells and lymphocytes is
present in the stroma. The stroma has a reactive pattern, with spindle-shaped cells distributed in a cartwheel arrangement
cells usually are interspersed in the reactive developed, lacking tortuosity and luminal secre-
stroma. Superficial stroma may become tions. Plasma cells may rarely be seen in
edematous. histologically normal secretory phase endome-
trium, however, and represent a subtle inflamma-
tory change that can be important in the evaluation
Abnormal Glandular Development of infertility.
In cycling patients, the endometrial response to

hormones is often diminished. Usually the endo- Epithelial Changes
metrium has proliferative phase characteristics,
with tubular glands showing mitotic activity. In Reactive cellular changes also affect the endome-
the secretory phase, the glands may lose their trial surface and glandular epithelium. The epithe-
normal pattern of reactivity. Secretory changes lium may show squamous, ciliated, and
occur in ovulating women, but they often show eosinophilic cell change (see Chap. 9), especially
abnormal development with less gland tortuosity when the inflammation is long standing and
and distension than is seen in a normal, nonin- intense [17]. The reactive epithelial cells may
flamed secretory phase. The changes can include become stratified, with prominent nucleoli,
irregular or delayed maturation of secretory cleared chromatin, and increased mitotic activity
phase endometrium. Glands may appear under- (Figs. 7.12 and 7.13).
Fig. 7.5 Nonspecific chronic endometritis. Numerous often associated with Chlamydia infection. The stroma is
lymphocytes are part of a mixed inflammatory infiltrate. reactive
Scattered transformed lymphocytes and immunoblasts are
Glandular and Stromal Breakdown Specific Infections
Endometritis also results in focal glandular and Although the etiology of chronic endometritis
stromal breakdown. With severe and prolonged usually is not apparent in biopsy specimens,
chronic inflammation, the changes of irregular some cases show morphologic features that offer
bleeding with breakdown and regeneration clues to the etiology. For example, the endome-
become prominent. This pattern of haphazard trium may harbor other changes, such as a
bleeding leads to a corrugated surface with foci retained subinvoluted implantation site following
of regenerating and shedding endometrium inter- a pregnancy or abortion, a placental site nodule
spersed. The inflamed stroma becomes dense and from a remote pregnancy, or other lesions, such
less responsive to hormonal changes. Because of as a polyp (Figs. 7.14a, b and 7.15).
the irregular growth, the tissue may become pol- The inflammatory response associated with
ypoid, and the resemblance to a polyp is accentu- Chlamydia trachomatis infection is usually
ated by the dense stroma that accompanies the marked. The inflammatory infiltrate tends to be
inflammation. A mixture of acute and chronic diffuse, with plasma cells, lymphocytes, and
bleeding patterns also may be present, with areas lymphoid follicles with transformed lympho-
of stromal collapse, glandular breakdown, cytes [1, 17, 28–30]. The inflammatory response
stromal fibrosis, macrophages, and hemosiderin to chlamydia also may be mixed, with an infil-
deposition (see Chap. 5). trate of acute, as well as chronic inflammatory
180 7 Endometritis
Fig. 7.6 Nonspecific acute and chronic endometritis. Gland lumen with neutrophilic microabscesses and reactive
stroma with numerous lymphocytes. Plasma cells are rare
cells. Stromal necrosis and reactive atypia of the gonorrhoeae are most frequently associated with
epithelium also can be present [17]. These these findings of a marked acute and chronic
marked inflammatory changes are not specific inflammatory infiltrate, although infections with
for chlamydia, however, but appear to reflect the C. trachomatis produce a greater concentration
presence of upper genital tract infection and of plasma cells (Fig. 7.16) and more lymphoid
acute salpingitis [1]. One study showed that neu- follicles than N. gonorrhoeae [1]. In cases with
trophils in the endometrial surface epithelium plasma cell endometritis, the presence of C. tra-
and in gland lumens, along with a dense subepi- chomatis infection can be established by using
thelial stromal lymphocytic infiltrate, plasma plasmid-based polymerase chain reaction (PCR)
cells, and germinal centers containing trans- or immunohistochemistry of paraffin-embedded
formed lymphocytes, are features that are pre- sections [29, 30].
dictive of a diagnosis of upper genital tract
infection and acute salpingitis [1]. The finding of
one or more plasma cells per ×120 field in the Granulomatous Inflammation
stroma and five or more neutrophils per ×400
field in surface epithelium was strongly associ- Granulomatous inflammation of the endome-
ated with upper genital tract infection and sal- trium is infrequent. Often the process is caused
pingitis [1]. Chlamydia trachomatis or Neisseria by mycobacterium, especially Mycobacterium
Fig. 7.7 Postpartum deciduitis. This curettage sample disclosed residual decidua and villi (not shown in this image)
infiltrated by abundant neutrophils. Acute villitis was also present (not shown)
tuberculosis, and the infection usually indicates granulomas may be difficult to identify unless
systemic disease. Although rare in the United the endometrium is biopsied in the late secretory
States, in some countries tuberculous endome- phase when the granulomas have had sufficient
tritis is more commonly encountered in endo- time to develop [6, 18, 36]. The surrounding
metrial biopsies undertaken during assessment stroma can show a lymphocytic infiltrate. As
of primary or secondary female infertility, as with any form of inflammation, gland develop-
endometrial involvement is a reflection of more ment may be altered, lacking an appropriate
widespread disease that also affects the fallo- secretory response if the biopsy is taken in the
pian tubes in most cases [31, 32]. Tuberculous luteal phase [36]. Acid-fast stains rarely demon-
endometritis also can cause abnormal uter- strate the characteristic organism in endometrial
ine bleeding in postmenopausal patients [33]. infections, and culture of fresh tissue or PCR of
Hysteroscopically, it may present as micropol- paraffin-embedded tissue may be needed to
yps; however, the latter are more an indicator establish the diagnosis [32].
of chronic endometritis than a specific etiologic Fungal infections, including crypto-
agent [34, 35]. coccosis, coccidioidomycosis, and blasto-
In tuberculous infection, the granulomatous mycosis, rarely involve the endometrium,
response is variable. Often the granulomas are resulting in granulomatous inflammation [5].
non-necrotizing (Fig. 7.17) [32]. Well-formed Cytomegalovirus infection has been seen in
182 7 Endometritis
Fig. 7.8 Postprocedural foreign body giant cell reaction. A xanthogranulomatous or foreign body giant cell reaction is
present in reactive stroma. This change can be seen in biopsies that follow a previous diagnostic procedure
association with poorly formed endometrial Actinomycosis

granulomas [37]. Sarcoidosis, too, may rarely
lead to non- necrotizing granuloma formation Infection by Actinomyces israelii is another rare
in the endometrium [38], and we have seen this cause of endometritis. This organism typically is
more frequently than infectious granulomas in found in endometritis associated with use of the
the endometrium. Patients with sarcoid involv- intrauterine device (IUD). When actinomycosis-
ing the endometrium typically have a history associated inflammation is present, the inflamma-
of sarcoid diagnosed at other sites. Necrotizing tory response usually is intense, with many plasma
granulomatous inflammation or chronic acute cells, lymphocytes, and neutrophils present
endometritis have been seen following endo- throughout the tissue (Figs. 7.20, 7.21, and 7.22a).
metrial hysteroscopic ablation therapy, as well The organisms show the typical sulfur granules
as in association with levonorgestrel intrauter- characterized by basophilic peripheral radiating
ine device (Mirena) (Fig. 7.18), where amor- filaments and a central, dense eosinophilic core.
phous, hyaline material accumulates within the Actinomycotic granules are highlighted with
endometrium [39–42]. A foreign body giant cell Brown and Brenn stain (Gram positive) and methe-
reaction often is present in addition to the granu- namine silver stain (Fig. 7.22b) while negative
lomas (Fig. 7.19a, b) [41]. when stained with modified acid-fast stain [19].
Fig. 7.9 Foreign body granulomatous reaction following necrotic center contains amorphous debris and carbon
endometrial ablation therapy. Chronic inflammation with deposits in the lower portion of the image that accumu-
foreign body giant cells surrounds an area of necrosis. The lated following thermal ablation of benign endometrium
Pseudoactinomycotic radiate granules have no Immunohistochemistry may be a useful tool when

relation to any infectious process and are dis- there is a high degree of suspicion and the charac-
cussed later (see Differential Diagnosis). teristic inclusions are not readily apparent on
H&E sections. The stroma may show a sparse
plasma cell infiltrate, but other changes associated
Cytomegalovirus with inflammation, such as a spindle-cell reactive
stroma, may not be found. One reported case had
Rarely, endometrial biopsy will show evidence of small, ill-defined, non-necrotizing granulomas in
cytomegalovirus infection. This may occur in the endometrium but no visible inclusions,
postpartum specimens and immunosuppressed although the presence of the virus was demon-
patients, or it may be found in women with no strated by PCR for viral DNA [37]. Chemotherapy-
known underlying disorder [43–46]. Regardless related bizarre nuclear atypia is occasionally seen
of immunologic status, the tissue shows the in surface epithelial endometrial samples which
characteristic nuclear and cytoplasmic inclusions may suggest viral-induced changes. Alkylating
in epithelial cells and occasional endothelial cells. agents, pyrimidine antagonists, taxanes, and
184 7 Endometritis
Fig. 7.10 Endometrial changes following endometrial ablation therapy. Chronic inflammation with necrosis and super-
imposed cauterization artifact are some of the characteristic features of an ablated endometrium
topoisomerase inhibitors have all been associated partum endometritis, especially in cases unre-
with striking epithelial endometrial atypia that sponsive to broad-spectrum antibiotic therapy
occasionally mimics malignancy or viral cyto- [52, 53]. When present in the endometrium, HSV
pathic inclusions, such as cytomegalovirus or her- can cause patchy necrosis of the glands and
pesvirus [47]. Immunohistochemistry for viral stroma [51]. The diagnosis is established by iden-
antigens can be helpful in excluding a viral etiol- tifying cells that show typical herpesvirus cyto-
ogy in questionable cases. pathic effect. Eosinophilic Cowdry type A
nuclear inclusions and multinucleate cells with
molded ground-glass nuclei can be found in the
Herpesvirus glandular epithelium or the stroma in areas of
necrosis. Several non-viral alterations, including
Herpesvirus (HSV) rarely infects the endome- optically clear nuclei associated with the pres-
trium, but it may occur, usually as an ascending ence of trophoblast [54] and cytoplasmic nuclear
process associated with cervical infection [48– invaginations in the Arias-Stella reaction [55]
51]. Although not as common as Group B (see Chap. 3), may superficially resemble the
Streptococci, anaerobic Gram-positive cocci, or herpesvirus effect. Immunohistochemical stains
anaerobic Gram-negative bacilli, HSV should be for herpesvirus antigens can be helpful in docu-
considered as a potential causative agent of post- menting the presence of the virus.
Fig. 7.11 Normal, nonreactive endometrial stroma. Nonreactive endometrial stroma typically shows ovoid cells and
scattered lymphoid aggregates
Mycoplasma tinguished from normal lymphoid tissue of the

endometrium. Subacute focal inflammation also
The morphologic changes associated with myco- has been linked with the presence of pelvic adhe-
plasma, especially Ureaplasma urealyticum, sions [57].
have been described [56, 57]. The inflammatory A study analyzing the correlation among hys-
pattern, termed “subacute focal inflammation,” is teroscopic, histologic, and culture findings in a
subtle but distinctive. In this condition the inflam- large cohort of women with proven chronic endo-
matory infiltrate is patchy, focal, and difficult to metritis found that more than 70% of cases were
discern. It is best seen in the mid-secretory phase caused by non-gonococcal, non-chlamydial
when stromal edema accentuates the inflamma- infections with common bacteria and Mycoplasma
tory foci. The areas of inflammation consist being the most frequent agents. Ureaplasma ure-
mainly of lymphocytes with macrophages and alyticum was detected in 10% of positive endo-
only rare plasma cells or neutrophils [56–58]. metrial cultures [3].
These small inflammatory foci tend to be located
beneath surface epithelium, around spiral arteri-
oles, or adjacent to glands [56]. Chronic lesions Differential Diagnosis
may appear granulomatous [56]. Biopsies that
are not timed for the edematous portion of the One of the most difficult problems in the differ-
secretory phase may miss these abnormalities ential diagnosis of endometritis is to decide
because the inflammatory infiltrate cannot be dis- whether apparent inflammatory cells represent
186 7 Endometritis
Fig. 7.12 Nonspecific endometritis with reactive epithe- chronic inflammatory infiltrate with many plasma cells
lial changes. The reactive glandular cells have abundant, surrounds the glands
eosinophilic cytoplasm and enlarged nuclei. A dense
true inflammation or whether they are a part of especially if the tissue is poorly preserved.
the normal cellular infiltrate of the endometrium. Normal lymphoid aggregates, however, typically
Normal stromal cells can resemble plasma cells, are widely spaced, located near the basalis, and
having the same size and an eccentric nucleus. do not include plasma cells. Stromal granular
The cytologic features of the nuclei distinguish lymphocytes are uniformly distributed throughout
stromal cells from plasma cells, as the latter have the stroma and are most prominent in the late
a characteristic clockface chromatin pattern. An secretory phase. These cells have dark, irregular
immunostain for CD138 has been used as a tool nuclei that often appear bilobed; their cytoplasm
to help assess for the presence or absence of is faintly granular. Stromal granular lymphocytes
plasma cells in difficult cases. In our practices, normally occur in tissue that lacks other features
however, we do not routinely use plasma cell of inflammation, including plasma cells, reactive
markers when evaluating possible chronic endo- stroma, and glands that appear out of phase.
metritis cases. Granular lymphocytes become especially promi-
Normal lymphoid aggregates and stromal nent in decidualized gestational endometrium or
granular lymphocytes can be especially difficult in endometrium that shows a progestin effect,
to distinguish from an inflammatory infiltrate, especially when the stroma shows a decidua-like
Fig. 7.13 Nonspecific endometritis with squamous epithelial changes. Reactive glandular cells are surrounded by
dense chronic inflammatory infiltrate with many plasma cells, while the overlying epithelium shows squamous change
a b
Fig. 7.14 Subinvoluted postpartum endometrium. (a) mately associated with the glandular epithelium. (b)
Abundant hyalinized material in which chronic inflamma- Hemosiderin deposition is present beneath inflamed sur-
tory cells, including plasma cells, and neutrophils are face epithelium
seen. The inflammatory infiltrate is extensive and inti-
188 7 Endometritis
Fig. 7.15 Subinvoluted products of conception. The is intimately associated with reactive glandular epithe-
stroma has a dense inflammatory infiltrate including lium. Hemosiderin deposition is present
plasma cells and neutrophils. The inflammatory infiltrate
reaction (see Chap. 6). In these cases, the infil- as part of the physiologic tissue breakdown.
trate can be so marked that at casual inspection it These neutrophilic infiltrates secondary to necro-
resembles the lymphoid response seen with sis and breakdown do not represent infection.
endometritis. In such cases the absence of plasma They are recognized and separated from
cells is especially helpful in distinguishing this infection-related inflammation by their lack of
pattern from a true inflammatory response. The infiltration into the epithelium as well as their
presence of decidualized stromal cells, usually association with glandular and stromal break-
containing inactive glands showing faint secre- down. On occasion, neutrophils also may be
tory activity, indicates that the process is an effect found only in gland lumens but not infiltrating
of progesterone or a synthetic progestin. the epithelium (Fig. 7.23a, b). This phenomenon
Neutrophils, like lymphocytes, can be present apparently is caused by entrapment of cellular
without indicating an infectious process. They debris from a previous cycle and is a finding of
normally occur in areas of stromal necrosis asso- no known significance.
ciated with bleeding and in necrotic tissues such Inflamed endocervix may also be sampled
as decidua or degenerating polyps. Menstrual during endometrial biopsy or curettage. This tis-
endometrium also shows neutrophilic infiltrates sue is a contaminant that has no relevance, as
Fig. 7.16 Chlamydia endometritis. A marked plasma- Chlamydia trachomatis endometritis and should trigger
cytic stromal infiltrate resembling granulation tissue, investigation of tubal involvement (salpingitis)
although not specific, suggests the possibility of
long as the endometrium itself is free of an inflammatory infiltrate in the endometrial stroma
inflammatory infiltrate. Foci of nonspecific cervi- that includes plasma cells, these free-floating
cal inflammation usually are a minor component histiocytes do not indicate inflammation [58].

of the tissue in endometrial samples. Endocervical Immunostains for histiocytes, such as lysozyme
epithelium that shows squamous metaplasia or or CD-68, can be very useful for demonstrating
microglandular hyperplasia often is present and these cells.
helps to identify these foci (see Chap. 2). Pseudosulfur granules, also known as “pseu-
Another possible contaminant in biopsy speci- doactinomycotic radiate granules,” occasionally
mens consists of aggregates of free-floating his- occur in the lower female genital tract, especially
tiocytes that do not infiltrate endometrial glands in association with an IUD [59, 60]. These radi-
or stroma. These histiocytes are a response to ate structures mimic the appearance of true acti-
extracellular mucin in the endocervical canal or nomycotic organisms, however, represent an
cellular debris in the cavity (see Chap. 2). Large unusual response (Splendore–Hoeppli phenome-
sheets of histiocytes may be found in curettings non) to foreign bodies or bacteria (Fig. 7.25).
following cervical stenosis with obstruction of Pseudosulfur granules do not stain with tissue
the os (Figs. 7.24a, b). In the absence of an Gram stains or with Gomori methenamine silver
190 7 Endometritis
Fig. 7.17 Tuberculous endometritis. A non-necrotizing stains although stains are often negative. Alternatively,
granuloma is present in the stroma. Identification of the paraffin-embedded tissue can be analyzed by PCR
acid-fast microorganisms can be attempted with acid-fast
(GMS), whereas true actinomyces do. These lying abnormalities, such as hyperplasia and car-
peculiar structures are not associated with other cinoma, may become secondarily inflamed,
endometrial abnormalities, including endometri- especially if the lesion or associated bleeding
tis, and should not be mistaken for actinomyces. results in loss of the cervical mucus barrier or
Inflamed glands may show reactive changes, dilation of the internal os with secondary infec-
with nuclear enlargement and prominent nucleoli, tion. These lesions should retain the typical mor-
and the cytologic features may suggest hyperpla- phology of the glands and stroma, which would
sia or neoplasia. In addition, the spindle-shaped, allow their diagnosis in the absence of an inflam-
reactive stromal cells of endometritis may be dif- matory response. There are no data to suggest that
ficult to distinguish from the fibrous stroma of a chronic inflammation has a significant relation-
polyp or from the desmoplasia of carcinoma. The ship to the genesis of either hyperplasia or carci-
architecture of tubular and uniform glands is usu- noma of the endometrium. Polyps should be
ally preserved during inflammation, and it is localized abnormalities with at least a partial lin-
important to evaluate glands in areas that do not ing of surface epithelium, and these generally do
show fragmentation or breakdown. Other under- not contain plasma cells unless they are second-
Fig. 7.18 Post-intrauterine device endometritis shows a superficial fibrinous exudate, mixed acute and chronic inflam-
matory cells, including plasma cells, as well as dilated, inactive glands
arily inflamed as a result of extension into the most common hematologic malignancy to involve
endocervix. Submucosal leiomyomas may trigger the endometrium is non- Hodgkin lymphoma,
an endometrial inflammatory reaction, especially usually the diffuse large-cell type, but this is a rare
when large and obstructive (Fig. 7.26). finding in biopsy specimens (see Chap. 11).
Severe chronic endometritis can produce an
intense lymphoid infiltrate with large lymphoid
cell immunoblasts that can resemble signs of Clinical Queries and Reporting
malignant lymphoma or a leukemic infiltrate [20–
22]. Usually with severe inflammation, the cellu- Accurate diagnosis of endometritis is important,
lar infiltrate is mixed, with a combination of as the presence of inflammation can establish a
plasma cells, neutrophils, and lymphoid cells with cause for abnormal bleeding, unexplained infer-
follicle formation, whereas lymphoma or leuke- tility, or significant upper genital tract infection/
mic infiltrates usually are composed of a rela- pelvic inflammatory disease. Most cases of
tively monotonous cell population. Involvement endometritis have no specific etiology that can
of the endometrium by malignant lymphoma is be determined by histologic study of the biopsy
rare in the absence of disseminated disease. The specimens. When endometritis is present, how-
192 7 Endometritis
a b
Fig. 7.19 Foreign body giant cell reaction with polariz- had a Mirena IUD removed. (b) The foreign body giant
able material. (a) A foreign body giant cell in the stroma cell demonstrates a polarizable particle
is engulfing refractile material. Prior to the biopsy, patient
Fig. 7.20 Actinomycotic endometritis. A florid mixed inflammatory infiltrate surrounds “sulfur granules” composed of
Actinomyces
Fig. 7.21 Actinomycotic endometritis. Notice the florid predominantly neutrophilic inflammatory infiltrate in proxim-
ity to the sulfur granules (Courtesy of Dr. Andres Chiesa-Vottero, Cleveland Clinic)
a b
Fig. 7.22 Actinomycotic endometritis. (a) Basophilic radiating filamentous organisms are surrounded by extensive
chronic inflammation. (b) The Gomori methenamine silver (GMS) stain strongly decorates such filaments
194 7 Endometritis
a b
Fig. 7.23 Incidental intraluminal neutrophils. (a, b) not be diagnosed as acute endometritis, especially without
Proliferative (A) and secretory (B) endometria displaying stromal inflammatory changes
intraluminal neutrophilic infiltrates. This finding should
a b
Fig. 7.24 Histiocytes. (a) A collection of intraluminal secondary to chronic bleeding from a variety of causes
histiocytes, like intraluminal neutrophils, has no signifi- including carcinoma. Hemosiderin also is present
cance by itself. (b) Stromal foamy histiocytes accumulate
Fig. 7.25 Pseudoactinomycotic radiate granules. These for organisms with Gram or silver stains. The patient had
granular structures superficially resemble sulfur granules an IUD in place at the time of the biopsy. The endome-
of actinomycosis. The filamentous structures are thicker trium was not inflamed (Courtesy of Dr. Andres Chiesa-
than actinomycosis filaments, however, and do not stain Vottero, Cleveland Clinic)
ever, a statement to indicate whether the sec- neutrophils in the surface epithelium, a dense
tions also show a demonstrable cause of the subepithelial stromal lymphocytic infiltrate, and
inflammation, such as evidence of a recent preg- lymphoid follicles with transformed lympho-
nancy or an structural lesion, such as a polyp, cytes raise the possibility of upper genital tract
may be helpful for subsequent clinical manage- inflammation and salpingitis.
ment. The presence of such lesions helps to Specific infections, such as tuberculosis or
establish the clinical cause and significance of cytomegalovirus, should clearly be indicated
the inflammation. when they are present. If special stains or immu-
The intensity of the inflammation also should nostains for organisms or plasma cells are per-
be noted, especially as severe inflammation with formed, the results should be given in the report.
196 7 Endometritis
Fig. 7.26 Submucosal leiomyoma (right) with a florid chronic endometritis (left) due to obstruction of the cervical os
6. Rotterdam H. Chronic endometritis. A clinicopatho-

References logic study. Pathol Annu. 1978;13:209–31.
7. Takebayashi A, Kimura F, Kishi Y, Ishida M,
1. Kiviat NB, Wolner-Hanssen P, Eschenbach DA, Takahashi A, Yamanaka A, Takahashi K, Suginami
Wasserheirt JN, Paavonen JA, Bell TA, et al. H, Murakami T. The association between endo-
Endometrial histopathology in patients with culture- metriosis and chronic endometritis. PLoS One.
proved upper genital tract infection and laparoscopi- 2014;9(2):e88354.
cally diagnosed acute salpingitis. Am J Surg Pathol. 8. Cicinelli E, Trojano G, Mastromauro M, Vimercati
1990;14:167–75. A, Marinaccio M, Mitola PC, Resta L, de Ziegler
2. Paavonen J, Aine R, Teisala K, Heinonen PK, D. Higher prevalence of chronic endometritis in
Punnonen R. Comparison of endometrial biopsy and women with endometriosis: a possible etiopathoge-
peritoneal fluid cytologic testing with laparoscopy in netic link. Fertil Steril. 2017;108(2):289–95.
the diagnosis of acute pelvic inflammatory disease. 9. Polisseni F, Bambirra EA, Camargos AF. Detection
Am J Obstet Gynecol. 1985;151:645–50. of chronic endometritis by diagnostic hysteroscopy
3. Cicinelli E, Matteo M, Tinelli R, Pinto V, Marinaccio in asymptomatic infertile patients. Gynecol Obstet
M, Indraccolo U, De Ziegler D, Resta L. Chronic Investig. 2003;55(4):205–10.
endometritis due to common bacteria is prevalent in 10. Ladefoged C, Lorentzen M. Xanthogranulomatous
women with recurrent miscarriage as confirmed by inflammation of the female genital tract.
improved pregnancy outcome after antibiotic treat- Histopathology. 1988;13:541–51.
ment. Reprod Sci. 2014;21(5):640–7. 11. Cicinelli E, Matteo M, Trojano G, Mitola PC, Tinelli
4. Cadena D, Cavanzo FJ, Leone CL, Taylor HB. Chronic R, Vitagliano A, Crupano FM, Lepera A, Miragliotta
endometritis. A comparative clinicopathologic study. G, Resta L. Chronic endometritis in patients with
Obstet Gynecol. 1973;41:733–8. unexplained infertility: prevalence and effects of anti-
5. Dallenbach-Hellweg G. Histopathology of the endo- biotic treatment on spontaneous conception. Am J
metrium. 4th ed. New York: Springer-Verlag; 1987. Reprod Immunol. 2018;79:e12782.
References 197
12. Kitaya K, Matsubayashi H, Yamaguchi K, Nishiyama 29. Mount S, Mead P, Cooper K. Chlamydia trachomatis
R, Takaya Y, Ishikawa T, Yasuo T, Yamada H. Chronic in the endometrium: can surgical pathologists identify
endometritis: potential cause of infertility and plasma cells? Adv Anat Pathol. 2001;8:327–9.
obstetric and neonatal complications. Am J Reprod 30. Paukku M, Puolakkainen M, Paavonen T, Paavonen
Immunol. 2016;75(1):13–22. J. Plasma cell endometritis is associated with
13. Moreno I, Cicinelli E, Garcia-Grau I, Gonzalez Chlamydia trachomatis infection. Am J Clin Pathol.
M, Bau D, Vilella F, Ziegler D, Resta L, Valbuena 1999;112:211–5.
D, Simon C. The diagnosis of chronic endometri- 31. Bazaz-Malik G, Maheshwari B, Lal N. Tuberculous
tis in infertile asymptomatic women: a comparative endometritis. A clinicopathological study of 1000
study of histology, microbial cultures, hysteroscopy, cases. Br J Obstet Gynaecol. 1983;90:84–6.
and molecular microbiology. Am J Obstet Gynecol. 32. Nogales-Ortiz F, Tarancon I, Nogales FF. The pathol-
2018;218:602–e1. ogy of female genital tuberculosis. A 31-year study of
14. Cicinelli E, Matteo M, Tinelli R, Lepera A, Alfonso 1436 cases. Obstet Gynecol. 1979;53:422–8.
R, Indraccolo U, Marrocchella S, Greco P, Resta 33. Schaefer G, Marcus RS, Kramer EE. Postmenopausal
L. Prevalence of chronic endometritis in repeated endometrial tuberculosis. Am J Obstet Gynecol.
unexplained implantation failure and the IVF suc- 1972;112:681–7.
cess rate after antibiotic therapy. Hum Reprod. 34. Scrimin F, Limone A, Wiesenfeld U, Guaschino
2015;30(2):323–30. S. Tubercular endometritis visualized as endometrial
15. Wu D, Kimura F, Zheng L, Ishida M, Niwa Y, Hirata micropolyps during hysteroscopic procedure. Arch
K, Takebayashi A, Takashima A, Takahashi K, Gynecol Obstet. 2010;281(6):1079–80.
Kushima R, Zhang G, Murakami T. Chronic endome- 35. Cicinelli E, Resta L, Nicoletti R, Zappimbulso V,
tritis modifies decidualization in human endometrial Tartagni M, Saliani N. Endometrial micropolyps at
stromal cells. Reprod Biol Endocrinol. 2017;15(1):16. fluid hysteroscopy suggest the existence of chronic
16. Adegboyega PA, Pei Y, McLarty J. Relationship endometritis. Hum Reprod. 2005;20(5):1386–9.
between eosinophils and chronic endometritis. Hum 36. Govan ADT. Tuberculous endometritis. J Pathol
Pathol. 2010;41(1):33–7. Bacteriol. 1962;83:363–72.
17. Winkler B, Reumann W, Mitao M, Gallo L, Richart 37. Frank TS, Himebaugh KS, Wilson
RM, Crum CP. Chlamydial endometritis. A histologi- MD. Granulomatous endometritis associated with
cal and immunohistochemical analysis. Am J Surg histologically occult cytomegalovirus in a healthy
Pathol. 1984;8:771–8. patient. Am J Surg Pathol. 1992;16:716–20.
18. Buckley CH, Fox H. Biopsy pathology of the endo- 38. Ho K-L. Sarcoidosis of the uterus. Hum Pathol.
metrium. 2nd ed. London: Arnold; 2002. 1979;10:219–22.
19. McCluggage WG. Benign diseases of the endome- 39. Ferryman SR, Stephens M, Gough D. Necrotising
trium. In: Kurman RJ, editor. Blaustein’s pathology granulomatous endometritis following endo-
of female genital tract. 6th ed. New York: Springer- metrial ablation therapy. Br J Obstet Gynaecol.
Verlag; 2011. p. 305–58. 1992;99:928–30.
20. Young RH, Harris NL, Scully RE. Lymphoma-like 40. Ashworth MT, Moss CI, Kenyon WE. Granulomatous
lesions of the lower female genital tract. A report of endometritis following hysteroscopic resection of the
16 cases. Int J Gynecol Pathol. 1985;4:289–99. endometrium. Histopathology. 1991;18:185–7.
21. Ferry JA, Young RH. Malignant lymphoma, pseudo- 41. Colgan TJ, Shah R, Leyland N. Post-hysteroscopic
lymphoma, and hematopoietic disorders of the female ablation reaction: a histopathologic study of the
genital tract. Pathol Annu. 1991;26:227–63. effects of electrosurgical ablation. Int J Gynecol
22. Skensved H, Hansen A, Vetner M. Immunore-active Pathol. 1999;18:325–31.
endometritis. Br J Obstet Gynaecol. 1991;98:578–82. 42. Charlton F, Otton G, Hedges A, Scurry J. Endometrial
23. Miko TL, Lampe LG, Thomazy VA, Molnar P, Endes granulomas with Levonorgestrel intrauterine system
P. Eosinophilic endomyometritis associated with diag- (Mirena). Pathology. 2014;46(5):457–8.
nostic curettage. Int J Gynecol Pathol. 1988;7:162–72. 43. Sayage L, Gunby R, Gonwa T, Husberg B, Goldstein
24. Lax SF. Endometritis : rare disease with clinical R, Klintmalm G. Cytomegalovirus endometritis after
importance? Pathologe. 2016 Nov;37(6):521–5. liver transplantation. Transplantation. 1990;49:815–7.
25. Russack V, Lammers RJ. Xanthogranulomatous endo- 44. Brodman M, Deligdisch L. Cytomegalovirus endome-
metritis. Arch Pathol Lab Med. 1990;114:929–32. tritis in a patient with AIDS. Mt Sinai J Med (NY).
26. Lopez JI, Nevado M. Exuberant xanthogranulomatous- 1986;53:673–5.
like reaction following endometrial curettage. 45. Dehner LP, Askin FB. Cytomegalovirus endometritis:
Histopathology. 1989;15:315–22. report of a case associated with spontaneous abortion.
27. Shintaku M, Sasaki M, Baba Y. Ceroid-containing Obstet Gynecol. 1975;45:211–4.
histiocytic granuloma of the endometrium. 46. McCracken AW, A’Agostino AN, Brucks AB,
Histopathology. 1991;18:169–72. Kingsley WB. Acquired cytomegalovirus infection
28. Paavonen J, Aine R, Teisala K, Heinonen PK, Punonen presenting as viral endometritis. Am J Clin Pathol.
R, Lehtinen M, et al. Chlamydial endometritis. J Clin 1974;61:556–60.
Pathol. 1985;38:726–32.
198 7 Endometritis
47. Kim EK, Yoon G, Kim HS. Chemotherapy- 55. Dardi LE, Ariano L, Ariano MC, Gould VE. Arias-
induced endometrial pathology: mimicry of malig- Stella reaction with prominent nuclear pseudoin-
nancy and viral endometritis. Am J Transl Res. clusions simulating herpetic endometritis. Diagn
2016;8(5):2459–67. Gynecol Obstet. 1982;4:127–32.
48. Goldman RL. Herpetic inclusions in the endome- 56. Horne HW, Hertig AT, Kundsin RB, Kosasa TS. Sub-
trium. Obstet Gynecol. 1970;36:603–5. clinical endometrial inflammation and T-mycoplasma.
49. Abraham AA. Herpes virus hominis endometritis in a A possible cause of human reproductive failure. Int J
young woman wearing an intrauterine contraceptive Fertil. 1973;18:226–31.
device. Am J Obstet Gynecol. 1978;131:340–2. 57. Burke RK, Hertig AT, Miele CA. Prognostic value
50. Schneider V, Behm FG, Mumau VR. Ascending her- of subacute focal inflammation of the endometrium,
petic endometritis. Obstet Gynecol. 1982;58:259–62. with special reference to pelvic adhesions as observed
51. Duncan DA, Varner RE, Mazur MT. Uterine herpes on laparoscopic examination. An eight-year review. J
virus infection with mutifocal necrotizing endometri- Reprod Med. 1985;30:646–50.
tis. Hum Pathol. 1989;20:1021–4. 58. Kim KR, Lee YH, Ro JY. Nodular histiocytic hyper-
52. Giraldo-Isaza MA, Jaspan D, Cohen AW. Postpartum plasia of the endometrium. Int J Gynecol Pathol.
endometritis caused by herpes and cytomegalovi- 2002;21:141–6.
ruses. Obstet Gynecol. 2011;117(2 Pt 2):466–7. 59. O’Brien PK, Roth-Moyo LA, Davis BA. Pseudo-
53. McGill AL, Bavaro MF, You WB. Postpartum herpes sulfur granules associated with intrauterine contracep-
simplex virus endometritis and disseminated infec- tive devices. Am J Clin Pathol. 1981;75:822–5.
tion in both mother and neonate. Obstet Gynecol. 60. Bhagavan BS, Ruffier J, Shinn B. Pseudoactinomycotic
2012;120(2 Pt 2):471–3. radiate granules in the lower female genital tract.
54. Mazur MT, Hendrickson MR, Kempson RL. Optically Relationship to the Splendore–Hoeppli phenomenon.
clear nuclei. An alteration of endometrial epithelium Hum Pathol. 1982;13:898–904.
in the presence of trophoblast. Am J Surg Pathol.
1983;7:415–23.
Polyps
8
Contents
Classification and Histologic Features 200
Common Polyps 205
Atypical Polypoid Adenomyoma (APA) 207
Adhesions 215
References 220
Most endometrial polyps appear to originate HMGI-C gene and specific clonal rearrange-
from localized overgrowth of the basalis with ments [3–9]. Metabolic syndrome (obesity,
glands and stroma participating in the lesion. The dyslipidemia, hyperglycemia) in postmeno-
pathogenesis of polyps has not been firmly estab- pausal women has been recently shown to be a
lished although molecular mechanisms have predictive factor in the development of endo-
been invoked in several studies, and it is sug- metrial polyps potentially due to local, ele-
gested that polyps originate as a monoclonal stro- vated levels of IGF-1 [10].
mal overgrowth with secondary endometrial The prevalence of endometrial polyps in the
gland proliferation [1]. The abnormality results general population is estimated to range from
in a disorganized growth comprised of endome- 7.8% to 50% [11, 12], while in patients present-
trial glands, stroma, and vessels which may be ing with abnormal uterine bleeding or other
polypoid or sessile, possibly over a large portion symptoms, at least one endometrial polyp was
of the endometrium. The pathogenesis of endo- identified as the main pathological finding in up
metrial polyps appears to be linked to imbalances to 26% of patients [13, 14]. Epidemiologically,
between endometrial proliferation and apoptosis endometrial polyps occur over a large age range
[2], as well as molecular alterations of the but are most common in women in the fourth and

200 8 Polyps
fifth decades, becoming less frequent after age 60 intraepithelial carcinoma (SEIC) [50] (see Chap.
[14–16]. Although endometrial polyps may be 9), underscoring the necessity of carefully sec-
asymptomatic, they typically present with irregu- tioning polyps in this age group. Nonetheless,
lar uterine bleeding [17, 18], either spotting or polyps are not regarded as a major risk factor for
heavy periods, and are found in between 2% and the development of carcinoma. Most lesions are
23% of patients coming to biopsy for abnormal successfully treated with curettage or hystero-
uterine bleeding [14, 19–21]. They also have scopic excision.
been implicated as a possible cause of infertility, The large variation in the reported prevalence
either by physically interfering with blastocyst of polyps reflects the difficulties in establishing
implantation or by altering the development of the histologic diagnosis. Polyps often are frag-
secretory phase endometrium, making it less mented and removed piecemeal at curettage,
receptive to the implanting embryo [22–25]. and therefore they can be difficult to recognize.
Tamoxifen therapy is a risk factor for endome- Hysteroscopy can be useful to confirm the diag-
trial polyps [18, 26], and polyps are the most fre- nosis of polyp [19, 51], although in one study as
quent pathologic lesion found in patients many as 13% of polyp-like structures seen by
receiving tamoxifen therapy for breast carcinoma hysteroscopy were not confirmed histologically
[27–30]. Tamoxifen-related polyps may be mul- [14]. Given the difficulty in the histologic diag-
tiple and large (Chap. 6). Although polyp size nosis of polyps, it is conceivable that true pol-
and duration of tamoxifen therapy do not appear yps detected by hysteroscopy were not properly
to be directly correlated with increased risk of classified on microscopic examination. Focal
malignancy [31–34]; endometrial hyperplasia, lesions such as polyps are often difficult to diag-
endometrioid carcinoma and adenosarcoma have nose accurately even by dilation and curettage
been directly associated with the use of tamoxi- (D&C), and up to 58% of polyps were missed
fen (see Chap. 6). by D&C in one study [52]. In our experience,
Hormone replacement therapy is not a risk we have seen multiple cases in which limited
factor for the development of most polyps [26, biopsies reveal only small amounts of prolifera-
35]. Nonetheless, there appears to exist hormone tive or atrophic endometrium, but on follow-up,
receptor expression differences between benign a more thorough sampling, often with hysteros-
and premalignant/malignant polyps, as well as copy, reveals a polyp.
polyps of tamoxifen users and nonusers [36, 37].
In general, polyps are benign, pedunculated,
or sessile growths with no malignant potential Classification and Histologic
[17]. Studies have shown that among all patients Features
with endometrial polyps, the prevalence of pre-
malignant and malignant lesions is low (3.57%), Polyps vary greatly in size, ranging from
with postmenopausal women having a higher microscopic abnormalities that are only a few
prevalence of malignancy than premenopausal millimeters across to huge lesions that can fill
women [38]. the uterine cavity or prolapse through the
Atypical hyperplasia and occasional cases of endocervical canal. Usually polyps are small
carcinoma, including serous carcinoma and car- and solitary, but they can be large or multiple.
cinosarcoma (malignant mixed Müllerian tumor), They may be sessile or pedunculated. The
can be confined to a polyp [16, 39–45]. In addi- glands and stroma of endometrial polyps can
tion, polyps containing hyperplasia, particularly show diverse histologic patterns [15–17, 53–
atypical hyperplasia, are significantly associated 55]. For practical purposes, polyps are divided
with hyperplasia and carcinoma in the nonpolyp- into benign polyps and the atypical polypoid
oid endometrium [16, 46–49]. In older women, adenomyoma. There are five morphologic
particularly in their 60s, polyps appear to have an forms of common benign endometrial polyps:
increased frequency of serous endometrial proliferative/hyperplastic, atrophic, func-
tional, mixed endometrial–endocervical, and Table 8.1 Histologic features of polyps

adenomyomatous. These patterns often over- Larger tissue fragments
lap, and have little clinical significance, serv- Polypoid shape
ing only to demonstrate the spectrum of Surface epithelium on three sides
morphologic forms that may be seen. Common Dense stroma, may be fibrous
Thick-walled arteries
polyps also can contain focal lesions such as
Glands more irregular, tortuous, and dilated than
hyperplasia (with or without atypia) (see normal glands
Chap. 9), serous endometrial intraepithelial Glands appear “out of phase” or hyperplastic
carcinoma (see Chap. 9), or carcinoma (see Distinct fragments that “stand apart” from the
Chap. 10). The atypical polypoid adenomy- remaining endometrial tissue
oma represents a distinctive form of polyp that
should be segregated from the variants of com-
mon polyps [15]. sue fragments (Figs. 8.1–8.4). For pedunculated
Despite their diverse growth patterns, all pol- polyps, these large fragments tend to be lined on
yps show several common histologic features three sides by surface epithelium (Fig. 8.1). The
that facilitate their diagnosis (Table 8.1) [15, polypoid fragments appear distinctly different
17]. One important feature of polyps in biopsy from fragments of normal endometrium, which
specimens is the presence of large, polypoid tis- often are admixed. The stroma in polyps can be
Fig. 8.1 Benign endometrial polyp. Large and polypoid tissue fragment entirely lined by a thin surface epithelial layer
and composed of irregularly shaped glands. Thick-walled blood vessels are a constant feature
202 8 Polyps
Fig. 8.2 Benign endometrial polyp. Large and polypoid gland pattern, including cystic atrophic and small prolif-
tissue fragment composed of irregularly shaped glands or erative glands, can mimic hyperplasia
varying sizes embedded in a dense stroma. The irregular
highly variable. Most polyps have a dense, hya- been found that p16 immunoreactivity is
linized stroma (Figs. 8.4a and 8.5) with tightly increased in the stroma of endometrial polyps,
packed spindle cells that generally resemble the although p16 can also be seen focally in nonpol-
stroma of proliferative endometrium (Fig. 8.4b). ypoid functionalis and does not appear to be
This stroma can appear mildly hypercellular, sufficiently specific to be diagnostically useful
and mitotic activity can be increased [56]. Rare in most cases [59, 60]. Variable amounts of
polyps show enlarged, atypical, or bizarre stro- edema and myxoid change also can be found,
mal cells that resemble benign atypical stromal and these changes are especially notable in
cells seen at other sites in the female genital tamoxifen-related polyps (see Chap. 6) [61]. In
tract such as the vulva and vagina (Fig. 8.6) addition, thick-walled vessels may be promi-
[57]. Bizarre stromal cells are more commonly nent in the stroma, especially when the polyp is
seen in postmenopausal women and appear to large and pedunculated. Small veins in the
have myofibroblastic characteristics [58]. There superficial stroma can become ectatic and
is no reliable immunohistochemical stains for extravasate red blood cells into a fibrotic subepi-
the identification of endometrial polyps. It has thelial compartment (Fig. 8.7).
Fig. 8.3 Benign endometrial polyp. Polypoid tissue fragment lined by surface epithelium on at least three sides dis-
plays glands of varying sizes, including cystically atrophic glands
a b
Fig. 8.4 Benign endometrial polyp. (a) The stroma in iniscent of proliferative endometrium. Irregular glands
polyps is characteristically hyalinized and dense. (b) In show telescoping artifact and are seen in proximity to
this polyp, the glandular and stromal components are rem- thick-walled blood vessels
204 8 Polyps
Fig. 8.5 Benign endometrial polyp. This high magnification of a benign polyp discloses the characteristic hyalinized
stroma where proliferative-type glands appear haphazardly arranged
The glands in polyps are irregular in shape, The surface and glandular epithelium in pol-
typically inactive, but occasionally proliferative yps often show metaplastic changes that include
or secretory and have a highly variable ciliated cell, eosinophilic, mucinous, hobnail,
architecture. They may be focally crowded, a
clear cell, and squamous changes (Figs. 8.8 and
finding that should not be overdiagnosed as 8.9) which can be focal or extensive. Hyperplastic
hyperplasia. The glands lack the uniform orienta- papillary proliferations (see Chap. 9) have been
tion of normal endometrial glands and may lose described as a focal component of polyps, usu-
their perpendicular orientation to surface epithe- ally along the surface [63].
lium and parallel course to the surface. In secre- In addition to these characteristic features, pol-
tory endometria, the irregularly shaped glands yps may show evidence of focal glandular and
often lack normal development and appear to be stromal breakdown, as well as infarct-type
“out of phase” [14]. changes if they outgrow their vascular supply or
It is normal in endometrial polyps to find areas undergo thrombosis of dilated superficial veins
of gland crowding alternating with gland-free (Fig. 8.10) (see Chap. 5), creating a range of mor-
spaces. In the absence of definitive cytologic phologic alterations resembling hyperplastic or
atypia, these focally clustered glands should not neoplastic changes that may pose a diagnostic
be interpreted as hyperplasia. To diagnose hyper- challenge. These changes include cellular syncy-
plasia in a polyp, it is best to identify a gland-to- tia formation (papillary syncytial metaplasia) or,
stroma ratio that exceeds 2:1 within an extensive occasionally, glandular nuclear atypia (Fig. 8.11).
area of the lesion [1, 62]. Although the degree of cytologic atypia seen in
Common Polyps 205
Fig. 8.6 Benign polyp with atypical stromal cells. Increased stromal cellularity with enlarged stromal cell nuclei.
Mitotic figures are absent
these cases could raise the concern for malig- nuclei and mitotic activity. These polyps usu-
nancy, a low Ki-67 proliferation index helps dis- ally have a moderate amount of intervening
tinguish them from true premalignant and stroma between the glands, the latter which are
malignant lesions [64]. Chronic bleeding leads occasionally closely packed. These crowded
also to hemosiderin deposition or foam cells in proliferative glands of different sizes and
the stroma. shapes can therefore resemble a disordered
proliferative pattern or hyperplasia without
atypia (see Chap. 9). In fact, these irregular
Common Polyps glands of a proliferative pattern polyp fre-
quently mimic focal hyperplasia in a back-
Proliferative/Hyperplastic Pattern ground of proliferative or atrophic
endometrium. Identifying endometrium not
Proliferative/hyperplastic polyps are the most involved by the polyp is especially helpful for
common. They are highly variable in size and accurate diagnosis of a proliferative/hyperplas-
can measure up to several centimeters in great- tic polyp, as the polyp may otherwise resemble
est dimension (Figs. 8.1 and 8.2). They often diffuse hyperplasia. The endometrium that is
are diagnosed only on microscopic examina- not involved by the polyp is usually prolifera-
tion. Regardless of size, they show irregular, tive or atrophic, but sometimes it is secretory
proliferating glands with pseudostratified in premenopausal women.
206 8 Polyps
Fig. 8.7 Benign endometrial polyp with ectatic vessels. The superficial aspect of a polyp is typically fibrotic and popu-
lated by thin, ectatic blood vessels associated with extravasated blood
Atrophic Pattern functional polyp lose their orientation to surface

epithelium and have a haphazard distribution. In
Atrophic polyps are usually seen in postmenopausal well-oriented fragments, the glands may appear
women. These polyps contain atrophic glands lined to branch toward the surface like veins on a leaf.
by low columnar epithelium showing no mitotic The stroma in a polyp may show edema or prede-
activity. The glands often are dilated and cystic, and cidual change but often is dense and inactive.
the stroma appears dense and fibrotic (Fig. 8.12). When polyps have secretory changes, the glands
Many of these polyps apparently represent prolif- often are not as well developed as those in the
erative/hyperplastic polyps that no longer show pro- surrounding endometrium. The result is that they
liferative activity. Atrophic endometrium in appear to be “out of phase,” and the stroma shows
postmenopausal women frequently acquires a pol- decreased edema and predecidual formation.
ypoid shape; however, this change does not have all Some examples of secretory endometrium with
the features of a polyp (Fig. 8.13). irregular maturation probably represent func-
tional polyps that are too fragmented for accurate
diagnosis.
Functional Pattern
These polyps, like the endometrium around them, Mixed Endometrial–Endocervical

are hormonally responsive and show proliferative Pattern
or secretory changes (Fig. 8.14). They occur in
premenopausal patients and can be difficult to Some polyps originate in the upper endocervix
diagnose. Unlike normal glands, the glands in the and lower uterine segment and show both endo-
Atypical Polypoid Adenomyoma (APA) 207
Fig. 8.8 Benign endometrial polyp with squamous change. This traumatized polyp demonstrates reactive squamous
change replacing the lining of some of the haphazardly arranged glands. Extravasated blood is also seen in the stroma
cervical and endometrial-type gland develop- typical Polypoid Adenomyoma

A
ment. Such polyps tend to have a fibrous stroma (APA)
resembling the stroma of the lower uterine
segment. This is an unusual and distinctive polyp charac-
terized by glands that are lined by atypical epi-
thelium and surrounded by cellular smooth
Adenomyomatous Pattern muscle and variable amounts of fibrous tissue
(Figs. 8.15a and b, 8.16, and 8.17) [66–68]. The
These polyps have smooth muscle in their stroma, atypical polypoid adenomyoma typically occurs
usually as irregular bundles and strands in prox- in premenopausal or perimenopausal women,
imity to thick-walled vessels. Although smooth with a mean age of about 40 years. Premenopausal
muscle is present, the glands are invested by women often are nulliparous [68]. We have seen
stroma. These polyps usually display prolifera- these lesions in women as old as 56, however,
tive/hyperplastic or functional gland changes. and there is a report of a case in an 81-year-old
Uterine adenomyomas can occasionally present woman [69]. A few cases have been associated
as an endometrial polyp [65]. with Turner syndrome and appear to be a compli-
208 8 Polyps
Fig. 8.9 Benign polyp. Polyp with irregular, crowded glands and extensive squamous change. Benign endometrium
separate from the polyp is present in the left upper corner
cation of long-term estrogenic stimulation of the polypoid adenomyoma are h aphazardly arranged
endometrium [70]. Often these lesions arise in but generally are not markedly crowded or back-
the lower uterine segment [67]. to-back. The cells comprising the glands have
Several immunohistochemical and molecular enlarged, stratified, and rounded nuclei with a
studies of the APA have provided conflicting vesicular chromatin pattern and prominent
results. One study revealed MLH-1 promoter nucleoli similar to the glands of atypical hyper-
hypermethylation in two cases while nuclear plasia. A very characteristic, though not specific,
expression of β-catenin in five cases. No feature is the presence of squamous change
CTNNB1 mutations were identified [71]. A sub- (metaplasia), containing central nonkeratinizing
sequent study found β-catenin (CTNNB1) muta- nests of squamous cells (Figs. 8.16 and 8.17).
tions in all analyzed cases, specifically in the Central necrosis may occur in the squamous
epithelial but not stromal component, suggesting nests.
this could be a clue to their histogenesis [72]. A Smooth muscle typically surrounds the
larger analysis of 21 cases found similarities glands, and endometrial stroma is largely
between APAs, atypical hyperplasia, and endo- absent. The smooth muscle is arranged in short
metrial carcinomas with loss of PTEN expres- interlacing fascicles that contrast with the
sion and KRAS mutations and no TP53 elongate bundles of smooth muscle found in
abnormalities [73]. The glands of the atypical normal myometrium or in adenomyomatous
Fig. 8.10 Infarcted endometrial polyp. This scanning such material may pose a diagnostic challenge by mim-
magnification shows that the entire polyp surface has icking a mesenchymal lesion
undergone infarct-type changes. A superficial biopsy of
polyps. The smooth muscle component can The rare endometrioid carcinomas that develop
show increased mitotic activity, with up to two in APAs are typically low grade [68, 74–76].
mitoses per ten high-power fields (HPFs), but Hysterectomy is performed for older women, but
there is no evidence of cytologic atypia. The conservative treatment has been an option in
smooth muscle is diffusely reactive for des- selected cases. If the latter is chosen, careful fol-
min. Occasionally, the smooth muscle compo- low-up with hysteroscopy and repeat curettage is
nent is replaced by a fibromyomatous stroma. necessary. Recurrence rates have been reported as
Some have speculated that is why APAs show high as 30%. Recurrences typically resemble the
a predilection to involve the lower uterine seg- original tumor [77].
ment, where the stroma has a fibrous appear-
ance [73].
As with any type of polyp, the endometrium Differential Diagnosis
not involved by the atypical polypoid adeno-
myoma can be highly variable and can show Endometrial polyps may be difficult to identify in
proliferative, secretory, gestational, or hyper- curettage specimens and are often missed. When
plastic changes. The atypical polypoid adeno- the specimen is large and polypoid-shaped, has
myoma often presents in curettage specimens surface epithelium covering three sides, and con-
as large polypoid tissue fragments admixed tains dense stroma with thick-walled vessels, the
with small fragments of noninvolved diagnosis is straightforward. Often polyps are
endometrium. microscopic abnormalities or they are highly
210 8 Polyps
Fig. 8.11 Glandular nuclear atypia in a polyp. Significant a p53 immunostain is useful in this scenario since significant
atypia can also be seen in the epithelium of a polyp. The lesions are highly proliferative and typically display an aber-
latter may raise the concern for serous endometrial intraepi- rant p53 immunohistochemical staining pattern (see Chap.
thelial carcinoma. A Ki-67 proliferation index coupled with 10). In this case, the atypia was of degenerative nature
fragmented by the biopsy procedure. Because of ents with a polypoid shape (Fig. 8.18). The ves-
fragmentation and lack of clear-cut diagnostic sels in basalis consist of small aggregates of basal
features, the differential diagnosis often includes arteries, usually numbering six or more in cross
normal proliferative or secretory endometrium. section. Levels through the tissue block often
Subtle features, including irregular gland shape show transitions to normal functionalis layer if
and distribution as well as a somewhat denser the area in question represents basalis.
stroma, are helpful in the recognition of a small Conversely, if the tissue represents a true polyp,
polyp that is admixed with normal proliferative the step sections often reveal surface epithelium,
or secretory endometrium. at least focally.
Tangential sectioning of normal endometrium The fibroblastic spindle-cell stromal response
is a frequent consideration in the differential that occurs in endometritis may resemble the fibrous
diagnosis of small and fragmented polyps. stroma of a polyp. However, with inflammation,
Basalis, especially, can resemble a polyp because there is a plasma cell infiltrate in the stroma, and the
it has irregular glands, dense stroma, and promi- epithelium also may be infiltrated by neutrophils.
nent arteries. Basalis, however, lacks a lining of Infrequently, florid acute and chronic endometritis
surface epithelium and only occasionally pres- may appear polypoid (Fig. 8.19).
Fig. 8.12 Atrophic polyp. Cystic glands are lined by atrophic epithelium, while the stroma is fibrotic and dense. Thick-
walled vessels were present in other areas
Irregular polypoid tissue due to progestin effect trium, while polyps are focal and admixed with
also may superficially resemble a polyp in a biopsy normal, non-hyperplastic endometrium. Most
or curettage specimen. These specimens usually polyps have characteristically dense stroma and
contain atrophic glands and a decidua-like stroma appear clearly different from the surrounding
(Fig. 8.20). The finding of normal endometrium uninvolved endometrium. However, it should be
admixed with polypoid tissue fragments is helpful noted that bona fide hyperplasia may arise in an
for identifying polyps in questionable cases. endometrial polyp (Fig. 8.21b). In small biopsies,
Diffuse hyperplasia (see Chap. 9) is also however, the distinction may not be possible, and
included in the differential diagnosis of polyp, the question of whether a small polypoid frag-
especially polyps with the proliferative/hyper- ment of tissue represents a portion of a polyp or a
plastic pattern. Both polyps and hyperplasia have hyperplasia cannot be resolved. Dilation and
an irregular proliferation of glands with a vari- curettage or hysteroscopy may be necessary to
able amount of stroma (Fig. 8.21a and b). establish the correct diagnosis.
Hyperplasia, furthermore, often is polypoid. A large polyp with proliferative/hyperplastic
Hyperplasia is generally a diffuse process, how- features may also resemble adenosarcoma (see
ever, usually affecting all or most of the endome- Chap. 11), as both lesions are polypoid with
212 8 Polyps
Fig. 8.13 Atrophic endometrium mimicking a polyp. Polypoid cystic atrophy displays markedly dilated glands lacking
mitotic activity. The stroma, however, does not show all the features necessary to diagnose a polyp
hyperplastic-appearing glands. Adenosarcoma and the phyllodes-like architecture, such as that

typically has a leaflike pattern with broad-based seen in adenosarcoma, is absent or only focal. In
papillae lined by surface epithelium. In contrast, addition, stromal atypia is generally character-
a polyp has a smooth outline. In addition, adeno- ized by mild nuclear enlargement and rare bizarre
sarcoma, unlike benign polyps, has a more nuclei. Mitotic activity is usually less than 2/10
cellular stroma with increased mitotic activity HPFs (Fig. 8.22) [78]. As noted previously, occa-
that aggregates in cuffs around the glands. Polyps sional atypical stromal cells in polyps can be seen
generally have a more dense or fibrous stroma and have no further significance [57]. Rarely, car-
that is less cellular and has a low mitotic rate. cinosarcoma has been found confined to endome-
Large polyps typically have thick-walled vessels, trial polyps [43].
which are lacking in adenosarcoma. The differential diagnosis for functional pol-
Rare polyps will have mildly atypical stroma yps includes changes due to ovulatory dysfunc-
that can suggest adenosarcoma or even carcino- tion, such as irregular secretory patterns, or
sarcoma, and correct recognition of this benign exogenous hormone effects. These polyps show
change is important to avoid overtreatment. In secretory changes that are out of phase with the
general, these atypical polyps are less than 3 cm, surrounding endometrium. Functional polyps are
Fig. 8.14 Functional polyp. High magnification shows crowded secretory glands (right of the image) contrasting with
an inactive (nonresponsive) gland (left of the image) in a fibrotic stroma of a polyp
a b
Fig. 8.15 Atypical polypoid adenomyoma. (a) Irregular, occasionally may show focal necrosis, a finding without
atypical glands are haphazardly distributed in smooth prognostic significance. Florid squamous change may
muscle. The smooth muscle has a characteristic pattern of obscure some of the cytologically atypical glands. The
short, interlacing fascicles. (b) Superimposed squamous squamous cells are cytologically benign
metaplastic changes are characteristic of this lesion and
214 8 Polyps
Fig. 8.16 Atypical polypoid adenomyoma. Atypical lacing fascicles. Notice a small focus of necrosis (upper
glands are haphazardly distributed in smooth muscle. The left corner) within the squamous morules. This finding
smooth muscle has a characteristic pattern of short, inter- does not affect prognosis
composed of polypoid tissue fragments covered is extremely unusual to find well-differentiated or

with surface epithelium and containing dense moderately differentiated carcinoma invading
stroma and thick-walled vessels. Surrounding myometrial smooth muscle in endometrial curet-
fragments of endometrium that show normal tings. Accordingly, in curettings atypical glands
secretory changes are especially helpful for surrounded by smooth muscle represent an atypi-
establishing the diagnosis of a functional polyp. cal polypoid adenomyoma until proven other-
The APA may resemble well-differentiated wise. Squamous change is common in the
endometrioid carcinoma invading the myome- atypical polypoid adenomyoma but can also
trium because it shows atypical glands in smooth occur in carcinoma or hyperplasia. In atypical
muscle. The glands in the atypical polypoid ade- polypoid adenomyoma with squamous change,
nomyoma, however, do not fulfill criteria for however, the affected glands are invested by
well-differentiated carcinoma, lacking features smooth muscle, and this is an important feature
of stromal invasion such as cribriform or conflu- that helps to identify the lesion. The rare exam-
ent growth (see Chap. 10). Also, the smooth mus- ples of carcinoma occurring in association with
cle of the atypical polypoid adenomyoma grows atypical polypoid adenomyoma show typical
in short, interlacing fascicles in contrast to the endometrioid features without smooth muscle
elongated fibers seen in normal myometrium. It investing the malignant glands [74, 75].
Adhesions 215
Fig. 8.17 Atypical polypoid adenomyoma. Superimposed squamous changes may obscure the underlying glands
Finally, metastatic tumors may rarely second- pregnancy, or following a D&C and their pres-
arily colonize an endometrial polyp. The breast is ence is known as Asherman syndrome [15, 23,
the most common extragenital site to metastasize 85]. In these cases, the endometrium is largely
to an endometrial polyp; however, other primary replaced by a proliferation of fibroblasts, but
sites including the gallbladder, colon, stomach, this is an extraordinarily rare finding in biopsy
kidney, urinary bladder, pancreas, lung, thyroid, material.
and cutaneous melanoma have been reported [79– We have observed, on occasion, another type
83]. A synchronous metastatic breast carcinoma to of intrauterine adhesion formed by a band of
an endometrial polyp with endometrial serous car- endometrial tissue with dense stroma and small,
cinoma has been reported [84]. nonreactive glands. This form of adhesion typi-
cally is found in premenopausal women, and its
etiology is not known. Because such adhesions
Adhesions consist of dense stroma and poorly developed
glands, they may be indistinguishable from frag-
Endometrial adhesions are bands of tissue that ments of a polyp (Figs. 8.23 and 8.24). Correlation
bridge the endometrial cavity. These may con- of histologic features with other information,
sist of fibrous tissue that forms secondarily to such as hysteroscopic findings, may be necessary
inflammation associated with an abortion or to establish the diagnosis.
216 8 Polyps
Fig. 8.18 Endometrial basalis. Tangential sectioning of and prominent thick-walled arteries. Fragments of basalis
basalis in an endometrial biopsy or curettage can resem- usually lack surface epithelium
ble a polyp because of the irregular glands, dense stroma,
Clinical Queries and Reporting the background endometrium not involved by the
polyp should be provided to the gynecologist.
Diagnosis of a polyp is important because its For example, if curettings in a postmenopausal
presence often explains abnormal bleeding or patient show a polyp and separate fragments of
accounts for a thickened stripe on vaginal ultra- atrophic endometrium, a comment or diagnosis
sound. It is important to recognize polyps and not that reflects these observations is warranted,
mistake them for a more serious lesion. since it also indicates the status of the endome-
Subclassification of polyps can unnecessarily trium not involved by the primary lesion. This
complicate the pathologic report and should type of interpretive comment helps the gynecolo-
therefore be avoided. Diagnosis as a benign polyp gist understand that the endometrial cavity has
is sufficient. A comment regarding the status of been sampled beyond the focal polyp.
Fig. 8.19 Polypoid endometritis. Extensive acute and chronic endometritis may occasionally present as an “inflamma-
tory pseudopolyp”
When an atypical polypoid adenomyoma is limited samples, small fragments of polypoid

diagnosed, a comment should be included to tissue with irregular glands may represent por-
describe the lesion, because most gynecologists tions of a polyp, or they may represent part of a
are not familiar with it. It is necessary to indicate hyperplasia. In such cases, a descriptive diagno-
that the lesion is benign and that it does not rep- sis to indicate the presence of an abnormality,
resent atypical hyperplasia or carcinoma. coupled with a microscopic description of the
At times endometrial sampling may yield tis- findings, is most helpful to the clinician. If the
sue suggestive, but not diagnostic, of a polyp. diagnosis is not straightforward, further evalua-
Small samples and markedly fragmented tissue tion, including a dilation and curettage or hyster-
may show features suggestive of a polyp but can oscopy, may be necessary to establish the correct
be especially difficult to accurately diagnose. In diagnosis.
218 8 Polyps
Fig. 8.20 Polypoid exogenous hormonal effect. With glands, however, are inactive. The characteristic vessels of
extensive stromal decidua-like change, fragments of a polyp are lacking
endometrial tissue may acquire a polypoid shape. The
a b
Fig. 8.21 Crowded glands in a polyp. (a) Finding of a focus of distinctively crowded glands with scant inter-
focally crowded glands separated by dense fibrous stroma vening stroma represents hyperplasia within a polyp
should not be interpreted as hyperplasia. (b) Occasionally
Fig. 8.22 Atypical polyp. Rare polyps display focal stro- The leaflike architecture of an adenosarcoma is lacking. If
mal atypia with mild nuclear enlargement and rare bizarre the changes are only focal, these lesions are likely benign
nuclei. Mitotic activity is usually less than 2/10 HPFs.
Fig. 8.23 Polypoid adhesion. Intrauterine adhesion presenting as a polypoid tissue with dense stroma (scar tissue) and
abundant hemorrhage. The precise etiology of this change is unknown
220 8 Polyps
Fig. 8.24 Intrauterine adhesion. Scar tissue with fibrinoid material and associated superficial eosinophilic changes.
Rare histiocytes populate the fibrinoid tissue
5. Tallini G, Dal CP. HMGI(Y) and HMGI-C dysregu-

References lation: a common occurrence in human tumors. Adv
Anat Pathol. 1999;6(5):237–46.
1. Ip PP. Benign endometrial proliferations mimick- 6. Fletcher JA, Pinkus JL, Lage JM, Morton CC, Pinkus
ing malignancies: a review of problematic enti- GS. Clonal 6p21 rearrangement is restricted to the
ties in small biopsy specimens. Virchows Arch. mesenchymal component of an endometrial polyp.
2018;472(6):907–17. Genes Chromosomes Cancer. 1992;5(3):260–3.
2. Banas T, Pitynski K, Mikos M, Cielecka-Kuszyk J. 7. Hennig Y, Wanschura S, Deichert U, Bartnitzke S,
Endometrial polyps and benign endometrial hyperpla- Bullerdiek J. Rearrangements of the high mobility
sia present increased prevalence of DNA fragmenta- group protein family genes and the molecular genetic
tion factors 40 and 45 (DFF40 and DFF45) together origin of uterine leiomyomas and endometrial polyps.
with the antiapoptotic B-cell lymphoma (Bcl-2) pro- Mol Hum Reprod. 1996;2(4):277–83.
tein compared with normal human endometria. Int J 8. Bol S, Wanschura S, Thode B, Deichert U, Van de
Gynecol Pathol. 2017;37(5):431–40. Ven WJ, Bartnitzke S, Bullerdiek J. An endometrial
3. Dal Cin P, Vanni R, Marras S, Moerman P, Kools polyp with a rearrangement of HMGI-C underly-
P, Andria M, Valdes E, Deprest J, Van de Ven W, ing a complex cytogenetic rearrangement involving
Van den Berghe H. Four cytogenetic subgroups can chromosomes 2 and 12. Cancer Genet Cytogenet.
be identified in endometrial polyps. Cancer Res. 1996;90(1):88–90.
1995;55:1565–8. 9. Dal Cin P, Wanschura S, Kazmierczak B, Tallini G,
4. Indraccolo U, Di Iorio R, Matteo M, Corona G, Greco Dei Tos A, Bullerdiek J, Van den Berghe I, Moerman
P, Indraccolo SR. The pathogenesis of endometrial P, Van den Berghe H. Amplification and expression
polyps: a systematic semi-quantitative review. Eur J of the HMGIC gene in a benign endometrial polyp.
Gynaecol Oncol. 2013;34(1):5–22. Genes Chromosomes Cancer. 1998;22(2):95–9.
References 221
10. Bueloni-Dias FN, Spadoto-Dias D, Delmanto LR, thology in 700 patients treated with tamoxifen for
Nahas-Neto J, Nahas EA. Metabolic syndrome as a breast cancer. Gynecol Oncol. 2000;78:181–6.
predictor of endometrial polyps in postmenopausal 29. Vosse M, Renard F, Coibion M, Neven P, Nogaret
women. Menopause. 2016;23(7):759–64. JM, Hertens D. Endometrial disorders in 406 breast
11. Dreisler E, Stampe Sorensen S, Ibsen PH, Lose cancer patients on tamoxifen: the case for less inten-
G. Prevalence of endometrial polyps and abnormal sive monitoring. Eur J Obstet Gynecol Reprod Biol.
uterine bleeding in a Danish population aged 20–74 2002;101:58–63.
years. Ultrasound Obstet Gynecol. 2009;33(1):102–8. 30. Berlière M, Charles A, Galant C, Donnez J. Uterine
12. Tanos V, Berry KE, Seikkula J, Abi Raad E, Stavroulis side effects of tamoxifen: a need for system-
A, Sleiman Z, Campo R, Gordts S. The management atic pretreatment screening. Obstet Gynecol.
of polyps in female reproductive organs. Int J Surg. 1998;91(1):40–4.
2017;43:7–16. 31. Ramondetta LM, Sherwood JB, Dunton CJ, Palazzo
13. Van den Bosch T, Ameye L, Van Schoubroeck D, JP. Endometrial cancer in polyps associated with
Bourne T, Timmerman D. Intra-cavitary uterine tamoxifen use. Am J Obstet Gynecol. 1999;180(2 Pt
pathology in women with abnormal uterine bleeding: 1):340–1.
a prospective study of 1220 women. Facts Views Vis 32. Cohen I, Bernheim J, Azaria R, Tepper R, Sharony
Obgyn. 2015;7(1):17–24. R, Beyth Y. Malignant endometrial polyps in post-
14. Van Bogaert LJ. Clinicopathologic findings in endome- menopausal breast cancer tamoxifen-treated patients.
trial polyps. Obstet Gynecol. 1988 May;71(5):771–3. Gynecol Oncol. 1999;75(1):136–41.
15. McCluggage WG. Benign diseases of the endome- 33. Nuovo MA, Nuovo GJ, McCaffrey FM, Levine
trium. In: Kurman RJ, Ellenson LH, Ronnett BM, edi- RU, Barron B, Winkler B. Endometrial polyps in
tors. Blaustein’s pathology of the female genital tract. postmenopausal patients receiving tamoxifen. Int J
6th ed. New York: Springer-Verlag; 2011. p. 305–58. Gynecol Pathol. 1989;8:125–31.
16. Peterson WF, Novak ER. Endometrial polyps. Obstet 34. Corley D, Rowe J, Curtis MT, Hogan WM, Noumoff
Gynecol. 1956;8:40–9. JS, Livolsi VA. Postmenopausal bleeding from
17. Schlaen I, Bergeron C, Ferenczy A, Wong P, Naves A, unusual endometrial polyps in women on chronic
Bertoli R. Endometrial polyps: a study of 204 cases. tamoxifen therapy. Obstet Gynecol. 1992;79:111–6.
Surg Pathol. 1988;1:375–82. 35. Perrone G, DeAngelis C, Critelli C, Capri O, Galoppi
18. Reslova T, Tosner J, Resl M, Kugler R, Vavrova P, Santoro G, et al. Hysteroscopic findings in post-
I. Endometrial polyps. A clinical study of 245 cases. menopausal abnormal uterine bleeding: a compari-
Arch Gynecol Obstet. 1999;262:133–9. son between HRT users and non-users. Maturitas.
19. Loffer FD. Hysteroscopy with selective endometrial 2002;43:251–5.
sampling compared with D&C for abnormal uterine 36. Antunes A Jr, Vassallo J, Pinheiro A, Leão R, Neto
bleeding. The value of a negative hysteroscopic view. AM, Costa-Paiva L. Immunohistochemical expres-
Obstet Gynecol. 1989;73:16–20. sion of estrogen and progesterone receptors in endo-
20. Schindler AE, Schmidt G. Post-menopausal bleed- metrial polyps: a comparison between benign and
ing: a study of more than 1000 cases. Maturitas. malignant polyps in postmenopausal patients. Oncol
1980;2:269–74. Lett. 2014;7(6):1944–50.
21. Choo YC, Mak KC, Hsu C, Wong TS, Ma 37. Leão RB, Andrade L, Vassalo J, Antunes A Jr, Pinto-
HK. Postmenopausal uterine bleeding of nonorganic Neto A, Costa-Paiva L. Differences in estrogen and
cause. Obstet Gynecol. 1985;66:225–8. progesterone receptor expression in endometrial pol-
22. Dallenbach-Hellweg G. The endometrium of infertil- yps and atrophic endometrium of postmenopausal
ity. A review. Pathol Res Pract. 1984;178:527–37. women with and without exposure to tamoxifen. Mol
23. Wallach EE. The uterine factor in infertility. Fertil Clin Oncol. 2013;1(6):1055–60.
Steril. 1972;23:138–58. 38. Costa-Paiva L, Godoy CE Jr, Antunes A Jr, Caseiro
24. Foss BA, Horne HW, Hertig AT. The endometrium JD, Arthuso M, Pinto-Neto AM. Risk of malignancy
and sterility. Fertil Steril. 1958;9:193–206. in endometrial polyps in premenopausal and post-
25. Sillo-Seidl G. The analysis of the endometrium of menopausal women according to clinicopathologic
1,000 sterile women. Hormones. 1971;2:70–5. characteristics. Menopause. 2011;18(12):1278–82.
26. Bakour SH, Gupta JK, Khan KS. Risk factors asso- 39. Dinh TV, Slavin RE, Bhagavan BS, Hannigan EV,
ciated with endometrial polyps in abnormal uterine Tiamson EM, Yandell RB. Mixed mullerian tumors of
bleeding. Int J Gynaecol Obstet. 2002;76:165–8. the uterus. A clinicopathologic study. Obstet Gynecol.
27. Cohen I, Azaria R, Bernheim J, Shapira J, Beyth 1989;74:388–92.
Y. Risk factors of endometrial polyps resected from 40. Salm R. The incidence and significance of early carcino-
postmenopausal patients with breast carcinoma mas in endometrial polyps. J Pathol. 1972;108:47–53.
treated with tamoxifen. Cancer. 2001;92:1151–5. 41. Wolfe SA, Mackles A. Malignant lesions arising
28. Deligdisch L, Kalir T, Cohen CJ, de Latour M, Le from benign endometrial polyps. Obstet Gynecol.
Bouedec G, Penault-Llorca F. Endometrial histopa- 1962;20:542–50.
222 8 Polyps
42. Silva EG, Jenkins R. Serous carcinoma in endometrial 60. Moritani S, Ichihara S, Hasegawa M, Iwakoshi A,
polyps. Mod Pathol. 1990;3:120–8. Murakami S, Sato T, Okamoto T, Mori Y, Kuhara H,
43. Kahner S, Ferenczy A, Richart RM. Homologous Silverberg SG. Stromal p16 expression differenti-
mixed mullerian tumors (carcinosarcoma) con- ates endometrial polyp from endometrial hyperpla-
fined to endometrial polyps. Am J Obstet Gynecol. sia. Virchows Arch. 2012;461(2):141–8. https://doi.
1975;121:278–9. org/10.1007/s00428-012-1276-10.
44. Barwick KW, Livolsi VA. Heterologous mixed mul- 61. Kennedy MM, Baigrie CF, Manek S. Tamoxifen and
lerian tumor confined to an endometrial polyp. Obstet the endometrium: review of 102 cases and compari-
Gynecol. 1979;53:512–4. son with HRT-related and non-HRT-related endome-
45. Silverberg SG, Major FJ, Blessing JA, Fetter B, Askin trial pathology. Int J Gynecol Pathol. 1999;18:130–7.
FB, Liao S-Y, et al. Carcinosarcoma (malignant mixed 62. Seto MT, Ip PP, Ngu SF, Cheung AN, Pun TC. Positive
mesodermal tumor) of the uterus. A Gynecologic predictive value of endometrial polyps in Pipelle aspi-
Oncology Group pathologic study of 203 cases. Int J ration sampling: a histopathological study of 195 cases.
Gynecol Pathol. 1990;9:1–19. Eur J Obstet Gynecol Reprod Biol. 2016;203:12–5.
46. Kelly P, Dobbs SP, McCluggage WG. Endometrial 63. Lehman MB, Hart WR. Simple and complex hyper-
hyperplasia involving endometrial polyps: report of a plastic papillary proliferations of the endometrium:
series and discussion of the significance in an endome- a clinicopathologic study of nine cases of apparently
trial biopsy specimen. BJOG. 2007;114(8):944–50. localized papillary lesions with fibrovascular stromal
47. Armenia CS. Sequential relationship between endo- cores and epithelial metaplasia. Am J Surg Pathol.
metrial polyps and carcinoma of the endometrium. 2001;25:1347–54.
Obstet Gynecol. 1967;30:524–9. 64. Fadare O, Renshaw IL, Parkash V. The spectrum of
48. Gray LA, Robertson RWJ, Christopherson morphologic alterations associated with infarction in
WM. Atypical endometrial changes associated with endometrial polyps: a report of 41 cases. Int J Gynecol
carcinoma. Gynecol Oncol. 1974;2:93–100. Pathol. 2017.
49. Pettersson B, Adami H-O, Lindgren A, Hesselius 65. Tahlan A, Nanda A, Mohan H. Uterine adenomyoma:
I. Endometrial polyps and hyperplasia as risk fac- a clinicopathologic review of 26 cases and a review of
tors for endometrial carcinoma. A case-control study the literature. Int J Gynecol Pathol. 2006;25:361–5.
of curettage specimens. Acta Obstet Gynecol Scand. 66. Mazur MT. Atypical polypoid adenomyomas of the
1985;64:653–9. endometrium. Am J Surg Pathol. 1981;5:473–82.
50. Spiegel GW. Endometrial carcinoma in situ 67. Young RH, Treger T, Scully RE. Atypical polypoid
in postmenopausal women. Am J Surg Pathol. adenomyoma of the uterus. A report of 27 cases. Am J
1995;19:417–32. Clin Pathol. 1986;86:139–45.
51. Fraser IS. Hysteroscopy and laparoscopy in 68. Fukunaga M, Endo Y, Ushigome S, Ishikawa
women with menorrhagia. Am J Obstet Gynecol. E. Atypical polypoid adenomyomas of the uterus.
1990;162:1264–9. Histopathology. 1995;27:35–42.
52. Epstein E, Ramirez A, Skoog L, Valentin L. Dilatation 69. Kuwashima Y, Uehara T, Kurosumi M, Kishi K, Ohno
and curettage fails to detect most focal lesions in the H, Maruyama M, et al. Atypical polypoid adenomy-
uterine cavity in women with postmenopausal bleed- oma of the uterus in a very old woman. Report of a
ing. Acta Obstet Gynecol Scand. 2001;80:1131–6. case with immunohistochemical characterization of
53. Dallenbach-Hellweg G. Histopathology of the endo- its stromal components and proliferative status. Eur J
metrium. 4th ed. New York: Springer-Verlag; 1987. Gynaecol Oncol. 1995;16:115–9.
54. Silverberg SG, Kurman RJ. Tumors of the uterine 70. Clement PB, Young RH. Atypical polypoid adenomy-
corpus and gestational trophoblastic disease. Atlas of oma of the uterus associated with Turner’s syndrome.
tumor pathology, 3rd series, Fascicle 3. Washington, Int J Gynecol Pathol. 1987;6:104–13.
DC: Armed Forces Institute of Pathology; 1992. 71. Ota S, Catasus L, Matias-Guiu X, et al. Molecular
55. Buckley CH, Fox H. Biopsy pathology of the endo- pathology of atypical polypoid adenomyoma of the
metrium. 2nd ed. London: Arnold; 2002. uterus. Hum Pathol. 2003;34:784–8.
56. Hattab EM, Allam-Nandyala P, Rhatigan RM. The 72. Takahashi H, Yoshida T, Matsumoto T, Kameda Y,
stromal component of large endometrial polyps. Int J Takano Y, Tazo Y, Inoue H, Saegusa M. Frequent
Gynecol Pathol. 1999;18:332–7. β-catenin gene mutations in atypical polypoid adeno-
57. Tai LH, Tavassoli FA. Endometrial polyps with myoma of the uterus. Hum Pathol. 2014;45(1):33–40.
atypical (bizarre) stromal cells. Am J Surg Pathol. 73. Němejcová K, Kenny SL, Laco J, Škapa P, Staněk
2002;26:505–9. L, Zikán M, Kleiblová P, McCluggage WG, Dundr
58. Heller D, Barrett T. Bizarre stromal cells in an endo- P. Atypical polypoid adenomyoma of the uterus: an
metrial polyp. Int J Surg Pathol. 2016;24:320–1. immunohistochemical and molecular study of 21
59. Stewart CJ, Bharat C, Crook M. p16 immunoreactiv- cases. Am J Surg Pathol. 2015;39(8):1148–55.
ity in endometrial stromal cells: stromal p16 expres- 74. Lee KR. Atypical polypoid adenomyoma of the endo-
sion characterizes but is not specific for endometrial metrium associated with adenomyomatosis and ade-
polyps. Pathology. 2015;47(2):112–7. nocarcinoma. Gynecol Oncol. 1993;51:416–8.
References 223
75. Staros EB, Shilkitus WF. Atypical polypoid adeno- gallbladder to an endometrial polyp detected by endo-
myoma with carcinomatous transformation: a case metrial curettage: case report and review of the litera-
report. Surg Pathol. 1991;4:157–66. ture. Int J Gynecol Pathol. 2009;28(4):343–6.
76. Sugiyama T, Ohta S, Nishida T, Okura N, Tanabe K, 82. Houghton JP, Ioffe OB, Silverberg SG, McGrady B,
Yakushiji M. Two cases of endometrial adenocarci- McCluggage WG. Metastatic breast lobular carci-
noma arising from atypical polypoid adenomyoma. noma involving tamoxifen-associated endometrial
Gynecol Oncol. 1998;71:141–4. polyps: report of two cases and review of tamoxifen-
77. Heatley MK. Atypical polypoid adenomyoma: associated polypoid uterine lesions. Mod Pathol Off J
a systematic review of the English literature. U S Can Acad Pathol Inc. 2003;16(4):395–8.
Histopathology. 2006;48:609–10. 83. Martinez MR, Marazuela MA, Vallejo MR, Bernabeu
78. Howitt BE, Quade BJ, Nucci MR. Uterine polyps RÁ, Medina TP. Metastasis of lobular breast cancer to
with features overlapping with those of Müllerian endometrial polyps with and without the presence of
adenosarcoma: a clinicopathologic analysis of 29 vaginal bleeding. Int J Gynaecol Obstet Off Organ Int
cases emphasizing their likely benign nature. Am J Fed Gynaecol Obstet. 2016;134(1):101–2.
Surg Pathol. 2015;39(1):116–26. 84. Vicioso L, Ortega MV, Cívico V, López-Beltrán
79. Mazur MT, Hsueh S, Gersell DJ. Metastases to the A. Synchronous metastasis from lobular carcinoma
female genital tract. Analysis of 325 cases. Cancer. and primary carcinoma of the endometrium in a
1984;53:1978–84. patient after tamoxifen therapy. Int J Gynecol Pathol
80. Kumar NB, Hart WR. Metastases to the uterine cor- Off J Int Soc Gynecol Pathol. 2013;32(1):66–70.
pus from extragenital cancers. A clinicopathologic 85. Friedler S, Margalioth EJ, Kafka I, Yaffe H. Incidence
study of 63 cases. Cancer. 1982;50:2163–9. of post-abortion intra-uterine adhesions evaluated
81. Kefeli M, Gonullu G, Can B, Malatyalioglu E, by hysteroscopy. A prospective study. Hum Reprod.
Kandemir B. Metastasis of adenocarcinoma of the 1993;8:442–4.
Precursors of Endometrial
Carcinoma 9
Contents
Endometrial Hyperplasia/Atypical Hyperplasia 225
Epithelial Cytoplasmic Change (Metaplasia) 241
Serous Endometrial Intraepithelial Carcinoma 251
References 257
It has been well recognized since the 1970s that of the Female Reproductive Organs was revised
unopposed estrogenic stimulation leads to endo- in 2014, and several changes were made. This
metrial hyperplasia, which in turn has the poten- chapter reflects those changes.
tial to progress to the usual type of endometrial
carcinoma, termed “endometrioid carcinoma.”
More recently, another lesion designated “endo- Endometrial Hyperplasia/Atypical
metrial intraepithelial carcinoma (EIC)” has been Hyperplasia
shown to precede the development of invasive
uterine serous carcinoma (USC) [1]. Because of In 1994 and 2003, the World Health Organization
its association with USC and not endometrioid (WHO) and the International Society of
carcinoma, it has been proposed that endometrial Gynecologic Pathologists (ISGYP) promoted a
intraepithelial carcinoma be designated “serous system for endometrial hyperplasia classification
endometrial intraepithelial carcinoma” [2]. that quickly gained widespread acceptance
This chapter discusses the various types (Table 9.1) [3–6].
of endometrial carcinoma precursors and The classification subdivided hyperplasia into
focuses on their mimickers. The World Health four groups according to their nuclear alterations
Organization’s WHO Classification of Tumours (atypia vs. without atypia) and degree of

226 9 Precursors of Endometrial Carcinoma
a rchitectural crowding defined by the extent of atypical or not, ignoring the extent of glandular
back-to-back glandular crowding (complex vs. crowding.
simple). Endometrial hyperplasia without atypia is a
Although this classification system consider- proliferation that encompasses a spectrum of
ably improved the approach to endometrial abnormal glandular changes and altered architec-
hyperplasia diagnosis, subsequent studies showed ture without cytologic atypia [3, 4, 9–13]. Altered
poor to moderate interobserver agreement architecture is typically evident on low-power
(kappa = 0.337–0.60) [7, 8]. Accordingly, the magnification, but the important feature that
current 2014 WHO Classification of Tumours of determines prognosis, and the likely association
the Female Reproductive Organs (4th edition) [2] with endometrial carcinoma, is cytologic atypia
simplified the four-tier system into two groups (see below). All types of hyperplasia are charac-
based on whether the hyperplastic process is terized by an increase in the gland-to-stroma
ratio, irregularities in gland shape, and variation
in gland size (Fig. 9.1 and Table 9.2) [3, 4, 11, 14,
Table 9.1 Original WHO classification of endometrial 15]. In hyperplasia without atypia, the glandular
hyperplasia epithelium cytologically resembles proliferative
Simple hyperplasia Without atypia endometrium (Fig. 9.2). The cells are columnar
With atypia with amphophilic cytoplasm and have pseu-
Complex hyperplasia Without atypia dostratified nuclei that maintain their orientation
With atypia to the underlying basement membrane. Nuclei
Fig. 9.1 Hyperplasia without atypia (formerly known as simple). Irregular glands showing variation in size and shape
are separated by abundant stroma. Some glands are dilated and mildly cystic
Table 9.2 Morphologic features of hyperplasia without are oval with smooth contours, evenly dispersed
atypia
chromatin, and small, inconspicuous nucleoli.
Cytologic features Mitotic activity can be variable and may be less
Nuclei than in proliferative endometrium. The mitotic
Pseudostratified
rate has no influence on the diagnosis. Tissue
Cigar-shaped to oval with smooth contours
obtained at curettage may be considerable, some-
Uniform chromatin distribution
Small to indistinct nucleoli times yielding enough to fill three or more tissue
Variable mitotic activity cassettes. On the other hand, office-based biop-
Variable amount of cytoplasm, often amphophilic sies may yield only a limited volume of tissue. In
Glands the 1994 WHO classification, the amount of
Irregular, variable size, some dilated stroma separating the glands was used to distin-
Branching, infolding, and outpouching guish simple and complex forms of hyperplasia,
Haphazardly spaced in abundant stroma (formerly whereas in the current 2014 WHO classification,
known as simple)
the presence of cytologic atypia is of paramount
Closely spaced with decreased stroma (formerly
known as complex) importance. The degree of glandular crowding,
Highly irregular outlines i.e., complexity, can be optionally included in the
Frequently associated features diagnosis. Hyperplasia is generally diffuse but
Polypoid growth may be a focal abnormality, possibly because of
Ciliated cells regional differences in estrogen and progesterone
Ectatic venules receptor content in the endometrium or due to the
Breakdown and bleeding
Fig. 9.2 Hyperplasia without atypia. The glandular cell glandular and stromal cells are cytologically similar to
nuclei are oval and pseudostratified with uniform outlines, those of proliferative phase endometrium. A few mitotic
lacking cytologic atypia and maintaining orientation to figures are present
the basement membrane. Nucleoli are indistinct. Both
Fig. 9.3 Hyperplasia without atypia (formerly known as complex). The glands are closely packed, lacking the abun-
dant stroma seen in a simple hyperplasia. There is no atypia
development of clonal lesions in a background of around thrombosed ectatic venules. As hyperpla-

unaffected endometrium. sia progresses, the glands become more densely
Considerable stroma is present in simple crowded with increased structural complexity,
hyperplasia. The stroma resembles that seen in more outpouchings and infoldings (Fig. 9.3).
the normal proliferative phase, consisting of Usually the glands are closely apposed and often
small, oval cells with scant cytoplasm. Like the back-to-back, although a small amount of inter-
glands, the stroma shows mitotic activity. When vening stroma is always present (Figs. 9.4 and
the hyperplasia is polypoid, the stroma may con- 9.5). As previously mentioned, although the dis-
tain thick-walled arteries like those seen in pol- tinction between “simple” and “complex” is no
yps. In simple hyperplasia, ectatic vascular longer the priority, it is noteworthy that most
channels (venules) are often present in the super- truly cytologically distinct endometrial prolifera-
ficial stroma beneath the surface epithelium. The tions (compared to an uninvolved background
pathogenesis of this change is not well under- endometrium) are “complex,” displaying an
stood but appears to be associated with nonphysi- increased gland-to-stroma volume or glandular
ologic, noncyclical endometrial growth. crowding. Especially for the novice in the field,
Morphologic evidence of active breakdown and screening the slides at low power for areas of
bleeding (see Chap. 5) may also be present crowded architecture may facilitate the finding of
Fig. 9.4 Hyperplasia without atypia. The glands vary in size and are separated by only a small amount of stroma.
Nuclei are oval and pseudostratified. There is no atypia
foci harboring cytologic atypia. Although atypia Papillary infolding or tufts lacking a fibrovascu-
can be seen in areas classified as “simple,” much lar core may project into the lumen (Fig. 9.8).
more frequently atypia will be encountered in Even with apparent back-to-back glandular
“complex” areas. Finally, although no longer crowding, each gland has a basement membrane
required by the current WHO classification sys- with a thin rim of stroma separating it from adja-
tem of endometrial hyperplasia, we still include cent glands. In some cases, however, the glands
in our reports the qualifiers “simple” and “com- are widely dispersed. Glands displaying no cyto-
plex,” as part of the description of the degree of logic atypia may be admixed with those showing
architectural changes encountered in any given cytologic atypia.
sample. The specific nuclear features of atypical
Hyperplasia with atypia (atypical hyperpla- hyperplasia include stratification, nuclear
sia) is based on the presence of nuclear atypia. In enlargement with altered chromatin, and nucleoli
contrast to non-atypical hyperplasia, however, (Table 9.3). The nuclei characteristically show
most cases of atypical hyperplasia have a com- true stratification ranging from two to four layers,
plex architecture with closely packed glands with loss of polarity in relation to the basement
(Figs. 9.6a, b, 9.7, 9.8, and 9.9). The glands tend membrane, giving an appearance of disarray to
to be highly irregular in size and shape (Table 9.3). the nuclei that contrasts with the pseudostratifica-
Fig. 9.5 Hyperplasia without atypia. Although the glands are separated by scant stroma, the nuclei remain small and
pseudostratified with oval contours, resembling cells in the normal proliferative phase. Nucleoli are indistinct
a b
Fig. 9.6 Atypical hyperplasia/EIN. (a) The glands are closely spaced, with little intervening stroma. (b) The lining
epithelium shows atypia, characterized by rounded, stratified, and vesicular nuclei with conspicuous nucleoli
Fig. 9.7 Atypical hyperplasia/EIN. The glands are highly irregular, but a small amount of stroma encompasses each gland.
The cells show scattered nuclear atypia with some stratified, round nuclei. Scattered apoptotic bodies are evident
tion in proliferative endometrium and most useful techniques in determining whether

non-atypical hyperplasia. The nuclei are enlarged atypia is present is to compare the nuclear fea-
and rounded rather than oval and may have irreg- tures of the glands in question to adjacent nor-
ular nuclear membranes (Figs. 9.6a, b and 9.9). mal-appearing proliferative endometrium. If the
The chromatin is eccentrically dispersed and nuclei in the glands in question appear more
forms clumps along the nuclear membrane, atypical than those in the proliferative glands,
resulting in a distinctive vesicular appearance provided metaplastic changes or other mimics
that is highly characteristic of endometrial atypia. are excluded, the lesion in question qualifies as
Nucleoli may be prominent. Determination of atypical hyperplasia or EIN. If the nuclear fea-
cytologic atypia can, at times, be difficult. The tures of the lesion in question are similar to the
changes described above can be subtle, and as a adjacent proliferative glands, no matter how
result, the interpretation can be somewhat sub- crowded the proliferative process is, the lesion
jective. This problem is inherent in whichever qualifies as hyperplasia without atypia. Atypical
classification system is used, namely, hyperpla- hyperplasia can be focally present in tissue along
sia/atypical hyperplasia or endometrioid intraepi- with non-atypical hyperplasia [16]. The minimal
thelial neoplasia (EIN) (see below). One of the criteria for the diagnosis of focal atypia have not
Fig. 9.8 Atypical hyperplasia/EIN. The complex glands philic cells project into the lumen of many glands. The
are closely spaced, but each gland has a stromal invest- stroma displays a dense inflammatory infiltrate
ment with basement membrane. Papillary tufts of eosino-
been defined in the endometrial hyperplasia clas- endometrium is a very helpful tool to correctly
sification. Nonetheless, for focal atypia to be a identify cytologic atypia. Surface epithelium
significant finding, it should be readily discern- should be avoided in assessing the presence of
ible from a background of clearly benign glands. atypia, as this is a common site for “metaplastic
In equivocal cases where there is a question of changes,” the latter which closely mimic the for-
clustered focal atypia in a background of hyper- mer. Squamous metaplasia, for instance, can be
plasia, a diagnosis of atypical hyperplasia is best focal or extensive in atypical hyperplasia. When
limited to those cases in which clearly atypical present, squamous metaplasia sometimes fills
nuclei are readily identified without diligent and expands the glands, accentuating the crowded
searching. We recommend that atypia not be appearance and leaving only a partial rim of
diagnosed unless clearly atypical nuclei involve columnar gland cells (Figs. 9.10a, b and 9.11).
several well-visualized glands in cross section. Finally, any attempts to grade atypia are fruitless
Comparison with the background uninvolved and lack reproducibility.
Fig. 9.9 Atypical hyperplasia/EIN. High magnification have a haphazard distribution, losing their orientation to
shows glandular nuclei with features of atypia. The nuclei the underlying basement membrane
are rounded and vesicular, with prominent nucleoli. They
a b
Fig. 9.10 Hyperplasia with atypia (atypical hyperplasia) that bridges the lumen. (b) This change may partially
showing squamous change. (a) Many of the glands are obscure the glandular component, but cytologic atypia can
partially filled with nonkeratinizing squamous epithelium still be appreciated
Table 9.3 Morphologic features of atypical hyperplasia Differential Diagnosis

Cytologic featuresa
Nuclei Several artifacts, as well as a variety of benign
Stratification with loss of polarity and malignant lesions, can be confused with
Enlarged, rounded with irregular shapes
hyperplasia, especially in endometrial biopsies.
Coarsening of chromatin creating a vesicular appearance
Prominent nucleoli
Artifactual changes to glands include fragmen-
Variable mitotic activity tation during biopsy or curettage, active bleed-
Cytoplasm ing with stromal collapse, and poor orientation.
Eosinophilic, diffuse or focal With any of these artifacts, the glands can
Glands appear irregular and crowded on casual inspec-
Irregular, variable size, some dilated tion. Fragmented proliferative or normal late
Haphazardly spaced in abundant stroma (simple)
secretory glands may become closely posi-
Closely spaced with decreased stroma (complex)
Highly irregular outlines
tioned, giving the illusion of crowded, disorga-
Frequently associated features nized glands with irregular shapes and sizes.
Papillary infoldings into glands (no bridging) The artifact of “telescoping,” resulting in a
Decreased stroma “gland within a gland” appearance, frequently
Ciliated cells occurs in association with fragmentation (see
Squamous metaplasia Chap. 2). This change can be mistaken for
Atypical nuclei should be readily apparent, involving
a
hyperplasia, as it often occurs in proliferative
most of the cells lining affected glands
endometrium. Glandular and stromal break-
Fig. 9.11 Squamous cell change in hyperplasia. When this change is extensive, such as in this case, it may partially
obscure the glandular component and hamper evaluation of cytologic atypia
down and bleeding distort the tissue, causing Chap. 8). In addition, they may show ciliated cell
irregular crowding of glands around areas of or squamous change. Nonetheless, they are sepa-
collapse that can be mistaken for hyperplasia rated from hyperplasia because they are gener-
(see Chap. 5). Likewise, basalis has irregular ally not estrogen-related abnormalities. In
glands that focally resemble the glands in hyper- general, polyps are focal lesions, and the sur-
plasia. These potential pitfalls of interpretation rounding endometrium is normal. The polypoid
are avoided by ensuring that the tissue on which shape, dense stroma, and thick-walled vessels are
the diagnosis is based has intact glands and helpful features in recognizing the ordinary
stroma without areas of breakdown. Surface polyp. In summary, it is the focal nature of the
epithelium is a very important anatomic land- polyp that separates this lesion from the more dif-
mark to orient the tissue and can help steer one fuse hyperplasia. Sometimes, however, it is diffi-
away from artifactual changes (see Chap. 2). cult to make this distinction with certainty in a
Occasionally, early secretory endometrium with small biopsy. Complicated cases may have
abundant glandular crowding may mimic hyper- hyperplasia involving a polyp (Fig. 9.12) with or
plasia with superimposed secretory features. In without hyperplasia in endometrium beyond the
this scenario, a Ki-67 proliferation index is use- polyp. Repeat curettage and hysteroscopy may be
ful to distinguish the two as mitotic rates are necessary to establish whether there is hyperpla-
significantly lower in secretory endometria
sia beyond the polyp.
(2.6%) compared to non-atypical (17%) and In atypical hyperplasia the differential diagno-
atypical hyperplasia (36%) [17]. sis is broader than in non-atypical hyperplasia. At
Disordered proliferative endometrium fre- one end of the spectrum, the cytologic changes of
quently enters in the differential diagnosis of atypia must be distinguished from benign abnor-
hyperplasia because in this condition the glands malities, such as eosinophilic syncytial change.
display irregular sizes and shapes. So-called At the other end of the spectrum, the differential
disordered proliferative patterns have mild diagnosis includes well-differentiated endometri-
gland irregularities that do not fulfill the criteria oid carcinoma, as both lesions can be composed
for hyperplasia. Some pathologists use the term of closely packed glands with cytologic atypia.
“disordered proliferative” to avoid assigning the Criteria for the diagnosis of well-differentiated
term “hyperplasia” to a patient’s case, although endometrioid carcinoma are discussed later and
we do not recommend this usage. Often this in Chap. 10. Typically, the benign cellular
change results from estrogen stimulation that changes that mimic atypia are those that result in
leads to focal glandular irregularities (see Chap. cytoplasmic eosinophilia, as the cells of atypical
5). Sometimes it might also be applied to mildly hyperplasia also frequently have eosinophilic
irregular proliferative endometrium that is dif- cytoplasm [see later, Epithelial Cytoplasmic
ficult to classify in a small sample but may Change (Metaplasia)].
reflect the presence of a polyp or other focal Endometritis may at times result in glandular
benign abnormality that is not fully represented changes that mimic hyperplasia with atypia (see
in the sections. We use the term when only a few Chap. 7). In cases with marked inflammation,
glands are dilated or branched, being confined especially those with acute and chronic inflam-
to no more than scattered foci within the func- mation, the glands will show reactive changes
tionalis. Other areas show tubular to tortuous with an irregular distribution in a reactive, spin-
proliferative glands. With diffuse glandular dle stroma. The reactive process includes cyto-
irregularities, the process is better classified as logic changes with enlarged, stratified nuclei, but
hyperplasia. these are generally limited findings. With endo-
Polyps are another frequent source of confu- metritis the glands are not irregular and crowded
sion in the differential diagnosis of hyperplasia. unless there is fragmentation artifact. Usually
Many polyps represent focal hyperplasia of the these reactive changes associated with inflamma-
basalis, which contains irregular glands (see tion occur in premenopausal patients.
Fig. 9.12 Hyperplasia in a polyp. Although a certain hyperplasia with closely spaced glands involves a polyp
degree of glandular crowding is always expected in a (left of the image)
benign polyp (right of the image), occasionally bona fide
Chronic inflammation with plasma cells can atypical polypoid adenomyoma, the smooth mus-
be seen in hyperplasia, and sometimes the inflam- cle cells around the glands set this lesion apart
matory infiltrate is striking. This inflammation in from atypical hyperplasia. Instead of endometrial
hyperplasia probably occurs secondary to the stroma, as is seen in hyperplasia, the glands are
constant abnormal bleeding or the polypoid more widely separated by smooth muscle in
growth of the tissue, as either of these processes short, interlacing fascicles. Immunohistochemical
can dilate the internal os. When the internal os is stains for desmin or actin can assist in the diagno-
dilated, the endometrium is susceptible to infec- sis of this lesion by demonstrating the smooth
tion and subsequent inflammatory response. muscle.
Consequently, it is important to recognize the Rarely, the endometrium may show papillary
architectural and cytologic characteristics of proliferations composed of papillae with fibro-
hyperplasia and not dismiss the changes as only vascular stromal cores and variable degrees of
chronic endometritis. branching lined by cuboidal to low columnar epi-
The atypical polypoid adenomyoma (see thelium with no atypia and little, if any, mitotic
Chap. 8) also enters the differential diagnosis, activity (Fig. 9.13) [4]. Two architectural patterns
because in this lesion the glands are highly irreg- were originally described by Lehman and Hart: a
ular and the cytologic changes of the epithelium simple pattern with focal glandular involvement
are similar to those of atypical hyperplasia. In and limited epithelial proliferation and a complex
Fig. 9.13 Focal papillary proliferation in a benign endometrial polyp. A focus of glands near the surface of a polyp
shows small papillary tufts projecting into the lumens. This limited lesion is not atypical
pattern with more diffuse glandular involvement, number of women (27% and 44%) were taking
more papillary branching, and complex intracys- some sort of exogenous hormones [18, 19].
tic proliferations with intraluminal tufting. In addition to these specific alterations that
Superimposed metaplastic epithelial changes may be confused with atypical hyperplasia, there
were frequent. This distinct process, termed are occasional situations in which normal, non-
“papillary proliferation” [4] or “simple and com- hyperplastic endometrium can superficially
plex hyperplastic papillary proliferations of the mimic an atypical proliferation. Often in such
endometrium,” [18, 19] is extremely unusual. cases, the cells lining glands of proliferative
Lehman and Hart also described an association endometrium can appear atypical at high magni-
with endometrial polyps in two thirds of their fication; the nuclei seem to be stratified and
cases [18]. A subsequent study of a larger cohort rounded and show a coarse chromatin distribu-
(59 cases) found that the simple papillary prolif- tion. Artifactual distortion of the tissue also may
erations appeared to follow a benign course, yield changes that superficially resemble atypical
while the more architecturally complex lesions proliferations. For example, there is an infrequent
had a higher risk of concurrent or subsequent but peculiar artifact of biopsies in which the glan-
development of hyperplasia or carcinoma and dular cells appear to be stratified, with a hobnail-
hence should be regarded as a type of “atypical like pattern (Fig. 9.14). In addition, mitotically
hyperplasia” [19]. Interestingly, in both studies, a active cells protrude into the lumen. This hobnail-
Fig. 9.14 Hobnail-like artifact of proliferative endome- glance, this artifact typically occurs as a focal finding at
trium. Curettage specimen of proliferative endometrium the edge of tissue fragments in curettings. This change
shows a focus where glandular epithelium appears strati- does not represent atypia
fied and disorderly. Although appearing worrisome at first
like artifact usually involves only a few glands, When atypia is suspected, it is important to
generally occurring at the edge of tissue frag- identify areas with intact glands and surrounding
ments, while the remainder of the tissue is free of stroma. If atypical hyperplasia is present, the glands
the abnormality. It usually is found in fragmented should have the architectural as well as the cyto-
proliferative endometrium and can be mistaken logic features that establish the diagnosis. If the
for atypia unless the overall pattern of normal glands in these foci are tubular and lack the irregu-
gland architecture is recognized, and fragmented lar outlines and altered gland-to-stroma ratio of
areas are avoided. This alteration has also been hyperplastic glands, the apparent atypia probably
termed “exfoliative artifact” and may be due to has no clinical significance. An occasional case of
local infusion of fluids during hysteroscopy [20]. proliferative endometrium can show foci of appar-
Superimposed ciliated metaplasia can be found ent crowded glands in which the nuclei look
in cases of hyperplasia, raising the atypia ques- enlarged and vesicular, that is, “atypical.” In such
tion. In these cases, the nuclei may appear cases, areas of obvious proliferative endometrium
rounded and vesicular, but the lack of true strati- which show identical nuclei indicate that this does
fication and presence of cilia usually assist in the not represent atypical hyperplasia. This empha-
exclusion of atypia. sizes the importance in questionable cases of com-
paring the “atypical” cells to normal proliferative hyperplasia (clinically insignificant) and EIN
endometrium in the same specimen. Several levels based on glandular crowding and genetic abnor-
through the tissue block can help resolve equivocal malities that were interpreted as clonal and
cases. If serial sections do not resolve the question, regarded as neoplastic, i.e., a true carcinoma pre-
then additional sampling such as a dilation and cursor [24–27]. This classification applies histo-
curettage may be necessary, especially if the first logic criteria that emulate previously studied
specimen was from an office-based biopsy. morphometric analysis parameters, the latter
Pathologists should not feel reluctant about recom- which approached the diagnosis of carcinoma
mending a more generous sampling when a biopsy precursors based on computerized morphometry
alone proves suspicious but insufficient. [28–32]. A “D-Score” for “multivariate discrimi-
Once true atypia is identified and benign nant score” focused on three crucial features: (1)
lesions that mimic hyperplasia are excluded, the volume percentage stroma (VPS), (2) gland
differential diagnosis includes well-differentiated branching/convolution (outer surface density of
endometrioid carcinoma, the latter which is easily glands), and (3) standard deviation of the shortest
established when there is myometrial invasion, nuclear axis which they correlated with nuclear
but this is a very rare finding in curettings. Thus, a atypicality [28, 29] A low “D-Score” is associ-
diagnosis of carcinoma is based on identifying ated with clonality. VPS, which is a measurement
invasion of endometrial stroma that can be a sub- of glandular volume within a field of view, has
tle change in well-differentiated neoplasms. There been shown to be a useful parameter, with a VPS
are three criteria, any of which identifies endome- below 55%, i.e., having decreased stroma, likely
trial stromal invasion: (1) an irregular infiltration to be present in lesions that are precursors to car-
of glands associated with an altered fibroblastic cinoma [26]. Additional H&E criteria for EIN
stroma (desmoplastic response); (2) a confluent derived from the morphometric studies included
glandular pattern in which individual glands, glands cytologically different than non-crowded
uninterrupted by stroma, merge and create a crib- areas and a maximum linear dimension of the
riform pattern; and (3) an extensive papillary pat- lesion greater than 1 mm. Mimics and cancer
tern [3, 9, 21–23]. Increasing degrees of nuclear must be excluded (Table 9.4). Mimics include
atypia, mitotic activity, and stratification in curet- benign conditions such as basalis, secretory
tage specimens also are associated with a higher endometrium, and polyps. Carcinoma with a
frequency of carcinoma but are of limited value solid or cribriform pattern or maze-like glands
compared to the main criterion of stromal inva- must be excluded before diagnosing EIN [31,
sion. These three criteria for determining the pres-
ence of invasion allow the diagnosis of atypical
hyperplasia or carcinoma to be made more objec- Table 9.4 Criteria for the diagnosis of EIN on H&E
tively and are discussed further and illustrated in stained sections
the next chapter. Endometrial intraepithelial car- EIN criteria Comments
cinoma, discussed later in this chapter, also must Architecture Area of glands exceeds that of
be distinguished from atypical hyperplasia as it is stroma
Cytology Cytology differs between
a different type of precursor lesion.
architecturally crowded focus
Endometrioid intraepithelial neoplasia (EIN) and background
was introduced in 2000. Parenthetically, the sys- Diameter > 1 mm Maximum linear dimension of
tem was initially termed “endometrial intraepi- the lesion exceeds 1 mm
thelial neoplasia,” but when it was adopted by the Exclude mimics Benign conditions with overlapping
criteria: basalis, secretory
2014 WHO as an alternate to atypical hyperpla-
endometrium, polyps, repair
sia, “endometrial” was changed to “endometri- Exclude cancer Carcinoma if maze-like
oid” to reflect the association specifically with meandering glands, solid areas,
endometrioid carcinoma. The EIN classification or appreciable cribriforming
separates endometrial proliferations into benign Adapted from [25], with permission
32]. Accordingly, the WHO schema that incorpo- surprising that a two-tier system will have greater
rates EIN has three categories: endometrial reproducibility than a four-tier system. A head-
hyperplasia (EH), endometrioid intraepithelial to-head comparison of all three systems with the
neoplasia (EIN), and endometrioid carcinoma. two-tier WHO system has not been reported to
Despite this apparent simplification of the EIN our knowledge. Finally, the introduction of the
format, studies have shown that like the older term “endometrial neoplasia” by the European
hyperplasia system, the diagnosis of EIN suffers Working Group creates confusion between the
from lack of interobserver reproducibility and EIN and the European Working Group Systems.
does not predict outcome any better than the Endometrioid intraepithelial neoplasia in the
hyperplasia system [33]. EIN system does not include lesions that qualify
Although EIN often correlates with atypical as well-differentiated carcinoma, whereas
hyperplasia, in other instances it corresponds to “endometrial neoplasia” as employed in the
hyperplasia without atypia. We have also noticed European Working Group (EWG) classification
that the diagnosis of EIN is at times indiscrimi- does.
nately applied to several lesions that lack the
strict criteria proposed. Accordingly, the EIN
system, in our view, can lead to overdiagnosis Behavior
and unnecessary treatment, and therefore we con-
tinue to use and recommend the hyperplasia/ Hyperplasia without atypia, either simple or
atypical hyperplasia nomenclature. complex, usually is a self-limited lesion that will
The European Working Group (EWG) has regress. Atypical hyperplasia, however, is associ-
also developed a classification for endometrial ated with a high risk for the development of car-
proliferative lesions. According to the EWG cinoma [11, 16, 34–40]. In one study that
classification, simple and complex hyperplasia examined untreated hyperplasia in detail, 80% of
without atypia are referred to as “hyperplasia,” both simple and complex hyperplasia without
while atypical hyperplasia and well-differenti- atypia regressed (Table 9.5) [38]. Furthermore,
ated endometrioid carcinomas are combined into the risk of progression to carcinoma was slight,
a single category designated “endometrial neo- 1% in simple hyperplasia and 3% in complex
plasia” [8, 34]. A reproducibility study compar- hyperplasia. Approximately 60% of the cases of
ing the WHO (four-tier), EIN (two-tier), and atypical hyperplasia also regressed, but the risk
European Working Group (two-tier) classifica- of progression to carcinoma was significantly
tion systems demonstrated that the EWG system greater compared to hyperplasia without atypia
appeared to have superior performance (interob- (Table 9.5). In the same study, 8% of cases of
server agreement kappa = 0.530 compared to simple atypical hyperplasia and 29% of cases of
0.337 and 0.419 for the 4-tier WHO and 2-tier complex atypical hyperplasia progressed to car-
EIN systems, respectively). Of course, it is not cinoma. These differences in progression rates
Table 9.5 Follow-up comparing cytologic and architectural abnormalities in 170 patients
Regressed Persisted Progressed to carcinoma
Type of hyperplasia No. of patients No. (%) No. (%) No. (%)
Simple 93 74 (80) 18 (19) 1 (1)
Complex 29 23 (80) 5 (17) 1 (3)
Simple atypical 13 9 (69) 3 (23) 1 (8)
Complex atypical 35 20 (57) 5 (14) 10 (29)
Adapted from [37], with permission
between simple and complex atypical hyperpla- 42–46]. The term “metaplasia” refers to
sia did not achieve statistical significance, but transformation of cells to a type not normally
overall there was a 23% rate of progression to found in an organ. By this definition, most of the
carcinoma from atypical hyperplasia compared alterations commonly classified as endometrial
to 2% in hyperplasia without atypia, and this was metaplasia do not qualify as such. Consequently,
a statistically significant difference. Accordingly, some of the cytologic transformations of the epi-
the presence of atypia is the most important prog- thelium previously referred to as endometrial
nostic feature for endometrial hyperplasia. Other metaplasia are better classified as a “change,” as
studies also found that up to 25% of patients who most represent either a degenerative process or a
undergo a hysterectomy soon after the diagnosis form of cytoplasmic differentiation. Squamous
of atypical hyperplasia at biopsy have well-dif- differentiation and secretory-like vacuolization
ferentiated endometrioid carcinoma in the uterus are examples of cytoplasmic differentiation,
[41]. When carcinoma is present following a while eosinophilic syncytial change likely repre-
biopsy diagnosis of atypical hyperplasia, the neo- sents a degenerative/regenerative change related
plasm is almost always well differentiated, focal, to endometrial breakdown (discussed in greater
and either confined to the endometrium or mini- detail in Chap. 5).
mally invasive into the myometrium. There are six general types of epithelial cyto-
Subsequent studies comparing the accuracy plasmic transformation that occur in the endome-
of disease progression prediction of the EIN and trium. These are squamous, ciliated, eosinophilic,
WHO 1994 classification systems revealed that mucinous, secretory (clear cell), and hobnail cell
13% of atypical hyperplasia and 2.3% of non- changes [3, 15, 44]. Other rare types of endome-
atypical hyperplasia progressed to malignant trial change/metaplasia include cartilaginous and
disease in the WHO 1994 system, while a pro- osseous [47], glial, adipose tissue, and endome-
gression rate of 19% and 0.6% was encountered trial extramedullary hematopoiesis [48].
for EIN and non-EIN cases, respectively. Recently, a synovial-like metaplasia has been
Accordingly, the EIN system claims to show a added to the list of possible endometrial changes,
better sensitivity to detect precancerous lesions the latter which has been seen in cases associated
than the WHO 1994 (92% vs. 67%) [32]. A sub- with the use of Levonorgestrel-releasing intra-
sequent multi-institutional analysis engaging a uterine device, although this is very rare
panel of expert gynecological pathologists (Fig. 9.15) [49]. The terminology for these
(advocates of both systems) was undertaken and changes continues to evolve as greater experi-
showed that the progression risk associated with ence with them is gained. For example, ciliated
EIN is much lower than previously reported and cell change has also been termed tubal metapla-
quantitatively similar to the progression risk sia, and the terms eosinophilic and pink cell
seen with the atypical hyperplasia WHO-based change are synonymous. The relative frequencies
classification [33]. of these cytoplasmic changes are difficult to
determine. Ciliated and squamous changes are
the most widely recognized, but, in our experi-
pithelial Cytoplasmic Change
E ence, eosinophilic change is the most common of
(Metaplasia) the nonspecific cellular changes.
The association of cytoplasmic change with
Epithelial cytoplasmic alterations, commonly hyperplasia and carcinoma suggests that many
designated metaplasia, often occur in the endo- forms of cytoplasmic differentiation or transfor-
metrium. They are frequently encountered in mation are induced by chronic estrogen stimula-
association with hyperplasia and may also mimic tion. Cytoplasmic changes also may be associated
significant glandular abnormalities [3, 4, 9, 10, with trauma, polyps, or inflammation, however.
Fig. 9.15 Synovial-like metaplasia. This metaplastic lates the normal synovial membrane histology. (Courtesy
change has been rarely seen in association with from Dr. Julieta Barroeta, Cooper University Hospital)
Levonorgestrel-releasing intrauterine device. It recapitu-
Occasionally these cellular changes may be oval nuclei with small nucleoli and rare or
found in atrophy with no other known underlying absent mitoses. The nuclei are centrally placed
pathology. It is important to recognize the various in dense, eosinophilic cytoplasm. When the
types of cellular change and determine whether squamous change forms morules, the gland is
they accompany hyperplasia or not, as these largely filled with a round to oval mass of uni-
changes by themselves have no neoplastic form cells with indistinct cell borders. The
potential. intraglandular nests of squamous epithelium
Squamous differentiation (squamous meta- may show central necrosis, but this feature has
plasia) often is nonkeratinizing, forming so- no effect on the diagnosis or prognosis of the
called morules because of their three-dimensional lesion. Squamous change predominantly occurs
resemblance to mulberries [50]. The squamous within gland lumens, and in most cases surface
epithelium is rarely keratinized in hyperplasia, epithelium shows minimal involvement. Surface
keratinization occurring more frequently in car- squamous change is occasionally observed sec-
cinoma with squamous differentiation (see ondary to inflammation. In our experience focal
Chap. 10). The nonkeratinizing morules have a squamous morular change can be found in pol-
characteristic appearance, forming solid nests of yps, especially beneath the surface epithelium.
bland eosinophilic cells that fill gland lumens Detached nests of squamous morules can occa-
(Fig. 9.16). The cells have uniform, round to sionally be found in biopsies with no further
Fig. 9.16 Nonkeratinizing squamous change has a characteristic appearance, forming sheets of bland eosinophilic
cells that can fill gland lumens
changes suggestive of polyps. Their presence, normal, however. Ciliated cells usually are prom-
however, raises the possibility of the lesion inent in endometrium stimulated by unopposed
when no other abnormality is found. Squamous estrogen. Accordingly, ciliated cells are fre-
differentiation is generally found in endometria quently present in hyperplastic lesions, with and
that show signs of estrogenic stimulation, espe- without atypia. These cells often are interspersed
cially hyperplasia [43–45]. Although squamous in small groups among nonciliated columnar
change tends to occur more frequently in low- cells, but sometimes they are extensive and line
grade endometrioid carcinoma [51, 52], it can most of the gland. Ciliated cells have pale to
occur in association with all grades of endome- eosinophilic cytoplasm. The luminal border of
trioid carcinoma. these cells may show a cuticle of dense cyto-
Squamous change is immunoreactive for plasm formed by the ciliary basal bodies
CDX2 and CD10 [53], and immunostaining can (Figs. 9.17 and 9.18). Often the nuclei are mildly
be useful in identifying this process in difficult stratified, yet they remain cytologically bland
cases. with round to oval shapes, an even chromatin dis-
Ciliated cell change (tubal metaplasia) is tribution, and small nucleoli. The rounding and
arguably not a true metaplasia, as ciliated cells slight nuclear enlargement that characteristically
are normally present along the surface epithelium, occurs should not be considered as evidence of
being most numerous in proliferative endome- atypia. Mitoses generally do not occur in ciliated
trium [54]. Glands lined by ciliated cells are not cells.
Fig. 9.17 Hyperplasia with superimposed ciliated cell would not qualify as EIN as metaplastic changes must be
change. These endometrial glands resemble the lining of a excluded for that diagnosis to be rendered
fallopian tube, and there is no cytologic atypia. This
Eosinophilic (pink) cell change also is com- lacking luminal cilia. Eosinophilic cell change
mon. This change represents several types of also merges with squamous change in some
cytoplasmic transformation. Eosinophilic cells cases; in these instances, the cells are rounded to
may be a variant of ciliated cells, squamous cells, polygonal and pavement-like, resembling cells
or oncocytes as well as eosinophilic syncytial seen in squamous differentiation but lacking the
change [3, 4]. All of these cytoplasmic transfor- solid, morule-like growth pattern. In other cases,
mations are without clinical consequence, per se. eosinophilic cells contain abundant, granular
Eosinophilic cytoplasm also is a frequent feature cytoplasm resembling oncocytes or Hürthle cells
of glands in atypical hyperplasia and low-grade seen in other organs (Figs. 9.19 and 9.20).
endometrioid carcinoma, so it is important to Eosinophilic cell change may even show inter-
determine if there is a coexisting neoplastic spersed cells with a small amount of cytoplasmic
process. mucin, suggesting overlap with mucinous cell
Eosinophilic cell change that resembles cili- change. In all these forms of eosinophilic cell
ated cell change is common. In this situation, the change, the nuclei are often round rather than
cells are columnar or slightly rounded and have oval and somewhat stratified. Luminal cell bor-
a moderate amount of pale pink cytoplasm, ders are sharply demarcated. The nuclei are
resembling the cytoplasm of ciliated cells but smaller and more uniform and lack the irregular
Fig. 9.18 Ciliated cell change. A gland in simple hyper- The nuclei lack features of atypia, and they have smooth,
plasia is partially lined by ciliated cells. The cytoplasm is uniform contours, a delicate chromatin pattern, and tiny
eosinophilic, and some nuclei are enlarged and rounded. nucleoli
nuclear membrane, chromatin condensation superficially resembles the cells in squamous

along the membrane, and prominent nucleoli metaplasia. Eosinophilic syncytial change
that characterize cells with true cytologic atypia. should not be mistaken for squamous metapla-
As in other forms of cytoplasmic change, mito- sia or interpreted as an “early” form of squa-
ses are extremely rare. Occasionally, eosino- mous differentiation, however. The constant
philic cell change occurs in non-estrogenic association of eosinophilic syncytial change
patterns such as atrophy. with breakdown and bleeding indicates that this
As noted, eosinophilic syncytial change is change is degenerative rather than metaplastic
not a metaplastic transformation, yet it has been (Fig. 9.21) [55]. Syncytial change is recognized
commonly described as such. In several studies by its prominent localization along surface epi-
this cellular alteration has been termed “papil- thelium, although it may also occur in glands
lary syncytial metaplasia,” “surface syncytial (see Chap. 5). Eosinophilic syncytial change
change,” or an “early” form of squamous meta- usually is accompanied by karyorrhectic debris,
plasia [4, 15, 55, 56]. The classification of these neutrophils, and adjacent glandular and stromal
eosinophilic cells as a metaplastic phenomenon breakdown with stromal collapse. Furthermore,
is attributable to the fact that the syncytial in this change, nuclei have a haphazard distribu-
aggregation of eosinophilic cells in this change tion, whereas with the other cytoplasmic
Fig. 9.19 Eosinophilic cell change. Endometrium shows epithelial cells with abundant eosinophilic cytoplasm, resembling
oncocytes. The nuclei lack atypical features, and mitotic figures are absent. This alteration, by itself, has no significance
changes, nuclei generally have a uniform infrequent. Very rarely mucinous change can
distribution. include transformation into goblet cells, and
Mucinous change is characterized by the the change has been designated “intestinal
presence of abundant mucinous cytoplasm, metaplasia.” This goblet cell change should be
resembling normal endocervical glandular differentiated from true colonic epithelium that
cells (Fig. 9.22). Often with this change, the rarely is inadvertently obtained if the uterus is
epithelium is also thrown into small papillary perforated during biopsy or curettage. The
projections. This pattern is not as common as presence of true intestinal-type glands with
the other cytoplasmic changes and is seen most goblet cells in endometrial biopsy will occa-
often in association with atypical hyperplasia sionally be secondary to involvement by an
or carcinoma. These cells are columnar, with ascending endocervical adenocarcinoma (see
basal nuclei and abundant pale supranuclear Chap. 10) [57] or secondary involvement by
cytoplasm that contains mucin. Histochemical colorectal carcinoma.
stains, such as mucicarmine or periodic acid- Several studies of various endometrial muci-
Schiff with diastase digestion, demonstrate the nous proliferations found that mucinous change
abundant cytoplasmic mucin. As in ciliated occurs across a morphologic spectrum from
cell change and eosinophilic change, the nuclei bland cytoplasmic change in simple mucinous
remain small and uniform, although they may proliferations to mucinous carcinoma [58, 59].
contain small nucleoli. Mitotic figures are The absence of cytologic atypia and architectural
Fig. 9.20 Eosinophilic cell change in background of small, round to oval nuclei. The cells are similar to those
hyperplasia without atypia. The glands to the right of this seen in ciliated cell change, but they lack visible cilia.
image display abundant eosinophilic cytoplasm and Glands on the left side of the field lack this change
complexity in simple mucinous proliferations This is usually a focal alteration, limited to scat-
was associated with a low risk of neoplasia. tered glands. As the names imply, the cells con-
Alternatively, lesions showing architectural com- tain clear, glycogen-rich cytoplasm and resemble
plexity or cytologic atypia but not clearly neo- those found in secretory or gestational endome-
plastic were associated with the concurrent or trium (Fig. 9.23). Rarely the cells develop a hob-
subsequent presence of well-differentiated endo- nail pattern with nuclei that protrude into the
metrioid carcinoma. Consequently, it is impor- gland lumen, resembling the Arias-Stella reac-
tant to identify the background cytologic and tion (Fig. 9.24a and b). The secretory/clear cell
architectural features when confronted with change usually occurs in endometrium that shows
mucinous change in a biopsy. When a lesion is estrogenic effects that range from a proliferative
encountered that shows worrisome architectural pattern to carcinoma. Sometimes secretory endo-
or cytologic features, the abnormality could be metrium shows extensive cytoplasmic clear cell
termed “complex atypical endometrial mucinous change that exceeds the amount of vacuolization
proliferation” with a comment that there is a sub- seen during normal luteal phase of the menstrual
stantial risk of well-differentiated carcinoma in cycle, and this, too, can be considered a form of
the uterus. clear cell change. We prefer the term “secretory”
Secretory and clear cell change is very infre- change as opposed to “clear cell” change to make
quent once progestin-related effects are excluded. certain that the gynecologist does not misinter-
Fig. 9.21 Eosinophilic syncytial change. This change is Occasionally, slight nuclear atypia may be present; how-
usually associated with breakdown and bleeding, which ever, mitotic figures are lacking
indicates that it is degenerative rather than metaplastic.
pret this change as related to clear cell addition, superimposed secretory changes can
carcinoma. increase the tortuosity of the glands, complicat-
Diffuse secretory changes sometimes occur in ing the interpretation of gland crowding. In such
hyperplasia [60], and this has been called “secre- cases, re-biopsy may be necessary to assess the
tory hyperplasia” (Fig. 9.23). This process can be endometrium after the secretory change has
seen in the premenopausal or perimenopausal resolved.
patient with hyperplasia who has sporadic ovula-
tion or who has been treated with progestins [60].
However, some examples are found with no evi- Differential Diagnosis
dence of either ovulation or exogenous progestin
use. Regardless of the cause, in secretory hyper- Squamous, ciliated cell, and the various types of
plasia the glands maintain the disordered archi- eosinophilic cell change all may superficially
tecture of hyperplasia, but they also show resemble the epithelium in atypical hyperplasia,
secretory changes with variably vacuolated cyto- or even well-differentiated endometrioid carci-
plasm and luminal secretions. Atypia is difficult noma, because they have pale, often pink cyto-
to recognize in these cases, because the secretory plasm and nuclei that appear to be stratified [42,
changes result in differentiation of the glandular 51, 52]. To complicate matters further, these
cells that creates a rather bland appearance. In changes often occur in hyperplastic endome-
Fig. 9.22 Mucinous change. The glandular epithelium contains mucin vacuoles reminiscent of endocervical
epithelium
trium. In fact, metaplasia in the absence of an the hyperplastic process is atypical or not. In
underlying proliferative process is quite uncom- other words, the diagnosis of atypia in
mon except for surface ciliated cell change. hyperplastic lesions containing eosinophilic cells
When these cellular changes occur in hyperpla- must be based on cells that do not demonstrate
sia, their recognition and separation from true eosinophilic change.
glandular atypia require attention to the nuclear Cytoplasmic change may occur in benign,
features. With atypia, the nuclei are enlarged and non-hyperplastic endometrium including prolif-
rounded with central vesicular chromatin and erative or secretory endometrium or in other con-
irregular nuclear membranes. Usually the nuclei ditions such as endometritis or polyps.
are stratified. These cytologic findings contrast Eosinophilic cell change also can occur focally in
with the relatively bland nuclear features of the otherwise atrophic endometrium. In these situa-
various cellular cytoplasmic changes. Mitoses tions, the lack of hyperplastic glandular architec-
are very infrequent in the latter, another helpful ture is helpful in recognizing these alterations as
feature in the differential diagnosis. Cells dis- incidental processes with no biologic signifi-
playing eosinophilic change, in particular, typi- cance. Therefore, it is important to assess the
cally have enlarged, rounded nuclei mimicking overall configuration of the glands and to be cer-
cytologic atypia. Since these cells are frequently tain that intact endometrium without glandular
present lining the glands of hyperplasia, it is and stromal breakdown is studied to determine
important to disregard them when determining if whether or not variations in cytoplasmic features
Fig. 9.23 Secretory change in hyperplasia. Glands in ondary to cytoplasmic vacuoles. In this case, nuclear
atypical hyperplasia show extensively vacuolated cyto- atypia was better seen in other areas. This pattern often is
plasm indicating secretory change. The secretory change the result of progestin therapy prior to biopsy but may
masks some of the features of atypia as the cells lose their occur in the absence of this history
pseudostratified nuclei and eosinophilic cytoplasm sec-
a b
Fig. 9.24 Hobnail cell change. (a) The cells show a hob- cells bulge into the few remaining glandular lumens and
nail pattern in an area of breakdown and bleeding that outer aspect of the tissue. This is regarded as a degenera-
somewhat resembles the Arias-Stella reaction seen in tive change
early pregnancy. (b) In atrophic epithelium, the hobnail
represent nonspecific “change” or a more signifi- cent to serous carcinoma and is also observed
cant lesion. adjacent to carcinosarcomas (malignant mixed
Sometimes in biopsy specimens, small, Müllerian tumors [MMMTs]) [61]. Occasionally,
detached fragments of tissue may contain areas SEIC is seen without an invasive carcinoma pres-
of squamous, eosinophilic, or mucinous change ent. The changes often occur within a polyp
that are suggestive of a more significant abnor- (Fig. 9.25a) [62].
mality, even in the absence of glands with iden- This lesion also has been termed “carcinoma
tifiable atypia. Such foci are especially in situ” and “uterine surface carcinoma,” but
worrisome when they occur in postmenopausal the term “endometrial intraepithelial carci-
patients, as their endometria should be atrophic. noma” is preferred because these lesions may
Detached fragments of squamous or mucinous rarely metastasize, even in the absence of inva-
epithelium may reflect the presence of more sig- sive carcinoma in the endometrium [63]. At
nificant glandular abnormalities, including atyp- times it may be difficult to determine whether
ical hyperplasia or even carcinoma that has not crowded glands involved by SEIC are invasive.
been adequately sampled. Further sampling, In these cases, if the lesion measures less than
usually by dilation and curettage, may be neces- 1.0 cm, it is designated minimal uterine serous
sary to determine the significance of the focal carcinoma (MUSC) [64]. These small, equivo-
alteration. cally invasive lesions seem to behave the same
Secretory changes in hyperplasia must be dis- as SEIC.
tinguished from secretory phase patterns in In contrast to atypical hyperplasia, SEIC is
which there is marked glandular crowding due to not associated with hyperplasia and does not
fragmentation, haphazard orientation, or crowd- typically occur in the clinical setting of increased
ing that is greater than usual as these are much estrogen exposure but instead is seen in atrophic
more common than this rare form of hyperplasia. endometrium of older postmenopausal women
A Ki-67 proliferation index is helpful in separat- [1]. SEIC can show flat or stratified cells with
ing these conditions from secretory hyperplasia very-high-grade nuclei, sometimes with papil-
since there is no or only minimal Ki-67 labeling lary tufts (Fig. 9.25b). The lesion involves sur-
in the nuclei of glands in normal or crowded face epithelium and may extend into atrophic
secretory endometrium, whereas labeling is pres- glands, showing an abrupt transition to normal
ent in the glands of secretory hyperplasia [17]. epithelium. The cells are polygonal and hobnail
Proper orientation of the tissue with regard to with highly irregular nuclei having coarse to
surface epithelium, surrounding stroma, and smudged chromatin. Mitoses, including atypical
basalis avoids this pitfall (see Chap. 2). In sum- mitoses, and apoptotic bodies are readily seen
mary, it is important to reiterate that all the vari- (Figs. 9.26, 9.27 and 9.28). The cells of SEIC are
ous metaplastic lesions, in and of themselves, typically strongly reactive for p53 by immuno-
have no bearing on outcome. histochemistry (mutant pattern), reflecting over-
expression of p53 protein (Fig. 9.29a and b) [64,
65]. This finding correlates closely with p53
erous Endometrial Intraepithelial
S gene mutations, which are seen in close to 90%
Carcinoma of cases of SEIC and serous carcinoma [1, 66,
67]. In a subset of cases, due to a loss of function
Serous endometrial intraepithelial carcinoma TP53 mutation, endometrial serous carcinoma
(SEIC) is the precursor lesion of serous carci- shows a complete absence of p53 expression
noma [1]. The lesion is characterized by epithe- (null pattern) with uninvolved (internal control)
lial cells that show marked nuclear abnormalities endometrial glands showing moderate to strong
identical to that seen in serous carcinoma of the but variable staining [68]. A third and less com-
endometrium yet lacking an invasive component. mon staining pattern that has been associated
SEIC is commonly seen in endometrium adja- with TP53 mutations is seen as p53 cytoplasmic
a b
Fig. 9.25 Serous endometrial intraepithelial carcinoma ground. (b) The SEIC cells display markedly atypical
in a polyp. (a) At low magnification, a cluster of darker nuclei with high N/C ratio; multiple, prominent nucleoli;
and crowded glands stands out at the polyp surface (right and numerous mitotic figures. Tufting is common as the
of the image) when compared to the normal polyp back- cells become discohesive
Fig. 9.26 Serous endometrial intraepithelial carcinoma. Several glands are lined by malignant cells with high-grade
nuclei interspersed among atrophic glands. The nuclei are stratified with papillary tufts
Fig. 9.27 Serous endometrial intraepithelial carcinoma. The degree of nuclear atypia far exceeds that seen in atypical
hyperplasia and resembles that seen in serous carcinoma
staining only with a nuclear wild-type staining from the main malignancy that are lined partially
pattern (see Chap. 10). or wholly by cells that are identical (cytologically
The immunohistochemical proliferation marker, and in mitotic indices) to the associated invasive
MIB-1, which stains the Ki-67 nuclear protein, malignancy or rare clusters of glands that exhib-
shows a very high proliferative index in SEIC and ited no confluent growth. In a large study includ-
therefore is another useful diagnostic tool [3, 64, ing 50 rigorously classified endometrial clear cell
69]. p16Ink4a, encoded by the CDKN2A gene, is a carcinomas, an intraepithelial component was
tumor suppressor protein that plays an important identified in 41.7% of cases [72].
role in the cell cycle. The p16 immunostain also
shows a strong and diffuse positive immunolabel-
ing of SEIC cases, which reflects its hyperprolifer- Differential Diagnosis
ation and cell cycle abnormality (Fig. 9.29c) [70].
Clear cell endometrial intraepithelial carci- The differential diagnosis of SEIC in biopsy
noma has been proposed as a precursor lesion to specimens includes invasive serous carcinoma
the rare endometrial clear cell carcinoma, akin to and atypical hyperplasia. SEIC is separated from
serous endometrial intraepithelial carcinoma [71]. invasive serous carcinoma by its pattern of
In clear cell carcinoma, this lesion is character- growth, with preservation of normal surface and
ized as having endometrial surface growth over a glandular architecture as the lesion extends along
benign endometrium involving glands isolated the epithelial surfaces. This pattern contrasts
Fig. 9.28 Serous endometrial intraepithelial carcinoma, glandular pattern. The cells show high-grade nuclear atypia
and scant cytoplasm. Apoptosis is extensive
with the irregular papillary and confluent gland In addition, SEIC often occurs in a background of
patterns of frank serous carcinoma (see Chap. atrophic endometrium, while atrophy is unusual in
10). As noted earlier, in small lesions (<1.0 cm) association with atypical hyperplasia.
where it is difficult to determine the presence of Immunostains for p53 and Ki-67 proliferation
invasion, the term MUSC is applied [64]. index are helpful as SEIC, but not atypical hyper-
SEIC also should not be confused with atypical plasia, is typically strongly and diffusely positive
hyperplasia even though both lesions are composed for p53 with a markedly elevated Ki-67.
of nonconfluent glands lined by atypical epithe-
lium. SEIC, in contrast to atypical hyperplasia, fea-
tures a much greater degree of nuclear abnormality Behavior
with cells showing clearly malignant features and
numerous mitotic figures including abnormal SEIC is often found in the presence of frank i nvasive
mitotic figures. Conversely, atypical hyperplasia carcinoma, typically serous type. SEIC or serous
shows lesser degrees of nuclear abnormalities. carcinoma confined to the endometrium generally
a has a good prognosis [64]. However, a few cases of

pure SEIC without an invasive component have
been reported in which the patient had metastases
involving extrauterine organs [63, 64, 73]. Even
microscopic disease at extrauterine sites was associ-
ated with recurrence and death despite aggressive
chemotherapy. Consequently, the diagnosis of SEIC
in an endometrial biopsy is an indication for further
evaluation of the uterus and surgical staging at the
time of hysterectomy. If there is extrauterine dis-
ease, even microscopic, the prognosis is poor. In
contrast, if there is no evidence of extrauterine dis-
b ease, the prognosis is excellent [64].
Clinical Queries and Reporting
It is important for the pathologist to consider the

clinical history when attempting to differentiate
normal variations and artifacts from true atypia.
Atypical hyperplasia is unusual in premeno-
pausal women; however, when there are certain
predisposing factors such as obesity, atypical
hyperplasia becomes more likely. For instance,
c the prevalence of obesity among women in the
USA during 2011–2014 was 38.3% [74].
Accordingly, it is not surprising that more cases
of atypical hyperplasia and endometrial carci-
noma are occurring in younger women. The
SEER 18 (2011–2015) indicated that 23.4% of
new cases of uterine carcinoma affected women
≤54 years of age [75]. Nonetheless, atypical
glandular proliferations are more common in
perimenopausal or postmenopausal patients.
Patients with hyperplasia typically present
with abnormal uterine bleeding, and the diagno-
Fig. 9.29 Serous endometrial intraepithelial carcinoma.
sis establishes a cause for the bleeding. In 2014,
(a) The highly malignant cells entirely replace some of the the WHO published its 4th edition of the WHO
atrophic epithelium. Residual atrophic glands are seen at Classification of Tumours of the Female
the bottom. (b) The glands lined by intraepithelial carci- Reproductive Organs which endorses use of both
noma show aberrant expression of p53 protein (strong/
diffuse pattern) in the high-grade nuclei. (c) Diffuse
endometrial hyperplasia and the EIN classifica-
nuclear and cytoplasmic p16 immunolabeling is also tion systems. In summary, the diagnosis of hyper-
characteristic plasia should be qualified as to whether atypia is
present with the understanding that atypical atypia, can be managed conservatively, as these
hyperplasia and EIN do not encompass the exact lesions tend to be self-limited. Patients with poly-
same spectrum of lesions [2]. cystic ovaries or postmenopausal patients on hor-
The American College of Obstetricians and mone replacement therapy do warrant especially
Gynecologists (ACOG) along with the Society of close follow-up. Atypical hyperplasia does not
Gynecologic Oncology (SGO) released their necessarily require hysterectomy either. This
committee opinion favoring the EIN system, but lesion can be managed medically with suppres-
the committees only referred to the 1994 WHO sive progestin therapy in young women who wish
four-tier scheme with no reference to the current to retain their fertility and in older women in
2014 two-tier system. Their major criticism to whom surgery is contraindicated [77]. Close fol-
the 1994 WHO scheme was the lack of specific low-up with endometrial biopsies is necessary to
management algorithms and poor reproducibility monitor the response to therapy in patients man-
of the classification among pathologists [76]. aged medically, however.
Despite these criticisms, there are several alterna- Separating cases of hyperplasia according to
tive considerations that continue to justify the whether they are simple or complex is of lesser
traditional WHO classification of hyperplasia importance than determining the presence or
and atypical hyperplasia. First, management of absence of atypia. Often areas of simple hyper-
non-atypical hyperplasia and atypical hyperpla- plasia and complex hyperplasia are admixed, and
sia is straightforward for any gynecologist (see these lesions can be classified as mixed simple
below) and does not require an algorithm. and complex hyperplasia in an explanatory note,
Second, we believe that the two-tier WHO sys- if one chooses to do so (not a requirement). In
tem is based on fundamental and long-established contrast, almost all cases of atypical hyperplasia
methodology that is highly reliable if morpho- are complex. They may coexist with non-atypical
logic criteria are strictly applied. Further com- hyperplasia, but the diagnosis should be based on
parisons between the two-tier WHO and the EIN the worst lesion.
systems may change the recommendations, espe- Serous endometrial intraepithelial carcinoma
cially if the EIN system cannot be reliably used is a relatively unusual diagnosis, and clinicians
in a manner to prevent overdiagnosis of non- may not be familiar with this entity. Accordingly,
atypical lesions. it is important to indicate the significance of this
Pathologists should become acquainted with diagnosis. The gynecologist should understand
both systems and report according to the prefer- that this lesion is distinctly different from atypi-
ences of their clinicians. Likewise, gynecologists cal hyperplasia and, despite lack of invasion, may
who are not familiar with the EIN classification be associated with metastatic disease outside the
may need clarification of the terminology and uterus. Because of the association with serous
correlation with the terms for hyperplasia that and clear cell carcinomas and, possibly, serous
they have used previously. We recommend using carcinosarcomatous component, these patients
the current WHO terminology with explanatory should undergo a hysterectomy and careful stag-
notes when deemed appropriate. ing. Cellular cytoplasmic changes (metaplasia)
If atypical hyperplasia or EIN is present, the should be clearly separated from hyperplasia, as
gynecologist will be concerned about the possi- they have no effect on prognosis. Findings such
bility of carcinoma, as the former lesion is a rec- as ciliated cell and eosinophilic cell changes usu-
ognized risk factor for the latter. Even focal ally do not need to be reported as long as the
atypia carries a greater risk for the presence or other underlying conditions are evident.
subsequent development of carcinoma. Eosinophilic syncytial change also does not need
Conversely, hyperplasia without atypia has little to be reported, as it is simply a marker of break-
neoplastic potential, and often it is useful to down and bleeding. The importance of these
append the statement “no atypia seen” to these changes lies in their recognition as benign cyto-
diagnoses. Hyperplasia, especially without plasmic transformations. Generally, it is best that
References 257
these changes not be specified in the report, but if 3. Ellenson LH, Ronnett BM, Kurman RJ. Precursor
lesions of endometrial carcinoma. In: Ellenson LH,
they are, it is important to add a comment to Ronnett BM, Kurman RJ editors. Blaustein’s pathol-
explain that the alteration has no clinical ogy of the female genital tract. 6th ed. New York:
significance. Springer-Verlag; 2011 Chap 8 p. 359–91.
Finally, there are situations in which the 4. Silverberg SG, Kurman RJ. Tumors of the uterine
corpus and gestational trophoblastic disease. Atlas of
biopsy sample may not be sufficient to com- tumor pathology, 3rd series, Fascicle 3. Washington,
pletely determine the full extent of the underlying DC: Armed Forces Institute of Pathology; 1992.
abnormality. For example, architectural features 5. Scully RE, Poulson H, Sobin LH. International histo-
may be suggestive, but not conclusive, of hyper- logical classification and typing of female genital tract
tumours. Berlin: Springer-Verlag; 1994.
plasia, especially when the specimen is small or 6. Silverberg SG, Kurman RJ, Nogales F, Matter GL,
shows extensive fragmentation. Likewise, nuclear Kubik-Nutch RA, Tavassoli FA. Tumors of the uter-
changes may be present that suggest atypia but ine corpus. Epithelial tumours and related lesions. In:
are inconclusive. In such cases, it is best to World Health Organization classification of tumours.
Pathology and genetics of tumours of the breast and
describe the abnormality as thoroughly as possi- female genital organs. Travassoli FA, Devilee P, eds.
ble, indicating that artifact or limited sampling Lyon: IARC Press; 2003: 221–232.
precludes a definitive diagnosis. A descriptive 7. Kendall BS, Ronnett BM, Isacson C, Cho KR,
diagnosis that indicates the uncertainty of the Hedrick L, Diener-West M, et al. Reproducibility of
the diagnosis of endometrial hyperplasia, atypical
findings is appropriate in these cases. hyperplasia, and well – differentiated carcinoma. Am
At other times, the distinction between atypi- J Surg Pathol. 1998;22:1012–9.
cal hyperplasia and well-differentiated endome- 8. Bergeron C, Nogales FF, Masseroli M, Abeler V,
trioid carcinoma may be difficult because of Duvillard P, Muller-Holzner E, et al. A multicentric
European study testing the re-producibility of the
small amounts of tissue. If the specimen is an WHO classification of endometrial hyperplasia with
endometrial biopsy, a recommendation to per- a proposal of a simplified working classification for
form a curettage should be made. If the specimen biopsy and curettage specimens. Am J Surg Pathol.
is a curetting, a diagnosis of “atypical hyperpla- 1999;23:1102–8.
9. Norris HJ, Tavassoli FA, Kurman RJ. Endometrial
sia approaching the level of endometrioid carci- hyperplasia and carcinoma. Diagnostic consider-
noma” can be made. In this latter situation, it is ations. Am J Surg Pathol. 1983;7:839–47.
likely that the gynecologist will manage the 10. Silverberg SG. Hyperplasia and carcinoma of the
patient as if she has a low-grade endometrioid endometrium. Semin Diagn Pathol. 1988;5:135–53.
11. Kraus FT. High–risk and premalignant lesions of
carcinoma. In young women or older women the endometrium. Am J Surg Pathol. 1985;9(Suppl):
who are not candidates for surgery, suppression 31–40.
with progestins and close follow-up are an option. 12. Welch WR, Scully RE. Precancerous lesions of the
On the other hand, if preservation of fertility is endometrium. Hum Pathol. 1977;8:503–12.
13. Silverberg SG. Problems in the differential diagno-
not a consideration or the patient is older and is a sis of endometrial hyperplasia and carcinoma. Mod
surgical candidate, the standard of care is a hys- Pathol. 2000;13:309–27.
terectomy [78]. 14. Hendrickson MR, Kempson RL. Surgical pathology
of the uterine corpus (Major problems in pathology
series), vol. 12. Philadelphia: W.B. Saunders; 1980.
15. Buckley CH, Fox H. Biopsy pathology of the endo-
metrium. 2nd ed. London: Arnold; 2002.
References 16. Tavassoli F, Kraus FT. Endometrial lesion in uteri
resected for atypical endometrial hyperplasia. Am J
1. Sherman ME. Theories of endometrial carcino- Clin Pathol. 1978;70:772–7.
genesis: a multidisciplinary approach. Mod Pathol. 17. Gurda GT, Baras AS, Kurman RJ. Ki-67 index as an
2000;13:295–308. ancillary tool in the differential diagnosis of prolifera-
2. Zaino R, Carinelli S, Ellenson L. Uterine corpus: tive endometrial lesions with secretory change. Int J
epithelial tumours and precursors. In: Kurman R, Gynecol Pathol. 2014;33(2):114–9.
Carcangiu M, Herrington C, Young R, editors. World 18. Lehman MB, Hart WR. Simple and complex hyper-
Health Organization classification of tumours of the plastic papillary proliferations of the endometrium:
female reproductive organs. 4th ed. Lyon: IARC; A clinicopathologic study of nine cases of apparently
2014. p. 125–35. localized papillary lesions with fibrovascular stromal
cores and epithelial metaplasia. Am J Surg Pathol. 33. Lacey JV Jr, Mutter GL, Nucci MR, Ronnett BM, Ioffe
2001;25:1347–54. OB, Rush BB, Glass AG, Richesson DA, Chatterjee
19. Ip PP, Irving JA, McCluggage WG, Clement PB, N, Langholz B, Sherman ME. Risk of subsequent
Young RH. Papillary proliferation of the endome- endometrial carcinoma associated with endometrial
trium: a clinicopathologic study of 59 cases of simple intraepithelial neoplasia classification of endometrial
and complex papillae without cytologic atypia. Am J biopsies. Cancer. 2008 Oct 15;113(8):2073–81.
Surg Pathol. 2013;37(2):167–77. 34. Ordi, et al. Reproducibility of current classifications
20. Crum C, Lee K. Altered endometrial differen- of endometrial endometrioid glandular proliferations:
tiation. In: Diagnostic gynecologic and obstet- further evidence supporting a simplified classifica-
ric pathology. Pennsylvania: Elsevier-Saunders; tion. Histopathology. 2014;64:284–92.
2006. p. 538. 35. Gusberg SB, Kaplan AL. Precursors of corpus can-
21. Kurman RJ, Norris HJ. Evaluation of criteria for distin- cer. IV. Adenomatous hyperplasia as stage 0 car-
guishing atypical endometrial hyperplasia from well– cinoma of the endometrium. Am J Obstet Gynecol.
differentiated carcinoma. Cancer. 1982;49:2547–57. 1963;87:662–77.
22. Gray LA, Robertson RWJ, Christopherson 36. Hertig AT, Sommers SC. Genesis of endome-
WM. Atypical endometrial changes associated with trial carcinoma. I. Study of prior biopsies. Cancer.
carcinoma. Gynecol Oncol. 1974;2:93–100. 1949;2:946–56.
23. King A, Seraj IM, Wagner RJ. Stromal invasion 37. Kurman RJ, Kaminski PF, Norris HJ. The behav-
in endometrial carcinoma. Am J Obstet Gynecol. ior of endometrial hyperplasia. A long-term study
1984;149:10–4. of “untreated” hyperplasia in 170 patients. Cancer.
24. Mutter GL, Baak JP, Crum CP, Richart RM, Ferenczy 1985;56:403–12.
A, Faquin WC. Endometrial precancer diagnosis by 38. Deligdisch L, Cohen CJ. Histologic correlates and
histopathology, clonal analysis, and computerized virulence implications of endometrial carcinoma
morphometry. J Pathol. 2000;190:462–9. associated with adenomatous hyperplasia. Cancer.
25. Mutter GL, et al. Benign endometrial hyperplasia 1985;56:1452.
sequence and endometrial intraepithelial neoplasia. 39. Buehl IA, Vellios F, Carter JE, Huber CP. Carcinoma
Int J Gynecol Pathol. 2007;26:103–14. in situ of the endometrium. Am J Clin Pathol.
26. Mutter GL. Histopathology of genetically defined 1964;42:594–601.
endometrial precancers. Int J Gynecol Pathol. 40. Campbell PE, Barter RA. The significance of atypi-
2000;19:301–9. cal hyperplasia. J Obstet Gynaecol Br Commonw.
27. Mutter GL. Endometrial intraepithelial neo- 1961;68:668–72.
plasia (EIN): Will it bring order to chaos? The 41. Rakha E, Wong SC, Soomro I, Chaudry Z, Sharma
Endometrial Collaborative Group. Gynecol Oncol. A, Deen S, Chan S, Abu J, Nunns D, Williamson
2000;76:287–90. K, McGregor A, Hammond R, Brown L. Clinical
28. Baak JP. The role of computerized morphometric outcome of atypical endometrial hyperplasia diag-
and cytometric feature analysis in endometrial hyper- nosed on an endometrial biopsy: institutional expe-
plasia and cancer prognosis. J Cell Biochem Suppl. rience and review of literature. Am J Surg Pathol.
1995;23:137–46. 2012;36(11):1683–90.
29. Baak JP, Orbo A, van Diest PJ, Jiwa M, de Bruin P, 42. Winkler B, Alvarez S, Richart RM, Crum CP. Pitfalls
Broeckaert M, et al. Prospective multicenter evalu- in the diagnosis of endometrial neoplasia. Obstet
ation of the morphometric D-score for prediction of Gynecol. 1984;64:185–94.
the outcome of endometrial hyperplasias. Am J Surg 43. Crum CP, Richart RM, Fenoglio CM. Adenoacanthosis
Pathol. 2001;25:930–5. of the endometrium. A clinicopathologic study in pre-
30. Baak JP, Wisse-Brekelmans EC, Fleege JC, van der menopausal women. Am J Surg Pathol. 1981;5:15.
Putten HW, Bezemer PD. Assessment of the risk on 44. Hendrickson MR, Kempson RL. Endometrial epi-
endometrial cancer in hyperplasia, by means of mor- thelial metaplasias: proliferations frequently mis-
phological and morphometrical features. Pathol Res diagnosed as adenocarcinoma. Report of 89 cases
Pract. 1992;188:856–9. and proposed classification. Am J Surg Pathol.
31. Baak JP, Nauta JJ, Wisse-Brekelmans EC, Bezemer 1980;4:525–42.
PD. Architectural and nuclear morphometrical fea- 45. Blaustein A. Morular metaplasia misdiagnosed as
tures together are more important prognosticators in adenoacanthoma in young women with polycystic
endometrial hyperplasias than nuclear morphometri- ovarian disease. Am J Surg Pathol. 1982;6:223.
cal features alone. J Pathol. 1988;154(4):335–41. 46. Ronnett BM, Kurman RJ. Precursor lesions of endo-
32. Baak JP, Mutter GL, Robboy S, et al. The molecu- metrial carcinoma. In: Kurman R, editor. Blaustein’s
lar genetics and morphometry-based endometrial pathology of the female genital tract. 5th ed.
intraepithelial neoplasia classification system predicts New York: Springer-Verlag; 2002. p. 467–500.
disease progression in endometrial hyperplasia more 47. Bahçeci M, Demirel LC. Osseous metaplasia of
accurately than the 1994 World Health Organization the endometrium: a rare cause of infertility and
classification system. Cancer. 2005;103(11):2304–12. its hysteroscopic management. Hum Reprod.
Review 1996;11(11):2537–9.
References 259
48. Gru AA, Hassan A, Pfeifer JD, Huettner cases, including 2 associated with tamoxifen therapy.
PC. Uterine extramedullary hematopoiesis: what Hum Pathol. 2003;34:939–43.
is the clinical significance? Int J Gynecol Pathol. 63. Soslow RA, Pirog E, Isacson C. Endometrial intraepi-
2010;29(4):366–73. thelial carcinoma with associated peritoneal carcino-
49. Stewart CJ, Leake R. Endometrial synovial- matosis. Am J Surg Pathol. 2000;24:726–32.
like metaplasia associated with levonorgestrel- 64. Wheeler DT, Bell KA, Kurman RJ, Sherman
releasing intrauterine system. Int J Gynecol Pathol. ME. Minimal uterine serous carcinoma: diagnosis
2015;34(6):570–5. and clinicopathologic correlation. Am J Surg Pathol.
50. Dutra F. Intraglandular morules of the endometrium. 2000;24:797–806.
Am J Clin Pathol. 1959;31:60–5. 65. Zheng W, Khurana R, Farahmand S, Wang Y, Zhang
51. Andersen WA, Taylor PT, Fechner RE, Pinkerton ZF, Felix JC. p53 immunostaining as a significant
JAV. Endometrial metaplasia associated with endometrial adjunct diagnostic method for uterine surface car-
carcinoma. Am J Obstet Gynecol. 1987;157:597–604. cinoma: precursor of uterine papillary serous carci-
52. Kaku T, Tsukamoto N, Tsuruchi N, Sugihara K, noma. Am J Surg Pathol. 1998;22:1463–73.
Kamura T, Nakano H. Endometrial metaplasia asso- 66. Sherman ME, Bur ME, Kurman RJ. p53 in endo-
ciated with endometrial carcinoma. Obstet Gynecol. metrial cancer and its putative precursors: Evidence
1992;80:812–6. for diverse pathways of tumorigenesis. Hum Pathol.
53. Houghton O, Connolly LE, McCluggage WG. Morules 1995;26:1268–74.
in endometroid proliferations of the uterus and ovary 67. Tashiro H, Isacson C, Levine R, Kurman RJ, Cho KR,
consistently express the intestinal transcription factor Hedrick L. p53 gene mutations are common in uterine
CDX2. Histopathology. 2008;53:156–65. serous carcinoma and occur early in their pathogen-
54. Masterson R, Armstrong EM, More IAR. The cycli- esis. Am J Pathol. 1997;150:177–85.
cal variation in the percentage of ciliated cells in 68. Köbel M, Ronnett BM, Singh N, Soslow RA,
the normal human endometrium. J Reprod Fertil. Gilks CB, McCluggage WG. Interpretation of P53
1975;42:537–40. Immunohistochemistry in endometrial carcinomas:
55. Zaman SS, Mazur MT. Endometrial papillary toward increased reproducibility. Int J Gynecol
syncytial change. A nonspecific alteration asso- Pathol. Mar 6, 2018. [Epub ahead of print].
ciated with active breakdown. Am J Clin Pathol. 69. McCluggage WG. Recent advances in immu-
1993;99:741–5. nohistochemistry in gynaecological pathology.
56. Rorat E, Wallach RC. Papillary metaplasia of the Histopathology. 2002;40:309–26.
endometrium: clinical and histopathologic consider- 70. Zheng W, Xiang L, Fadare O, et al. A proposed model
ations. Obstet Gynecol. 1984;64:90S–2S. for endometrial serous carcinogenesis. Am J Surg
57. Talia KL, Stewart CJR, Howitt BE, Nucci MR, Pathol. 2011;35:e1–14.
McCluggage WG. HPV-negative gastric type adeno- 71. Fadare O, Liang SX, Ulukus EC, Chambers SK,
carcinoma in situ of the cervix: a spectrum of rare Zheng W. Precursors of endometrial clear cell carci-
lesions exhibiting gastric and intestinal differentia- noma. Am J Surg Pathol. 2006;30(12):1519–30.
tion. Am J Surg Pathol. 2017;41(8):1023–33. 72. Fadare O, Zheng W, Crispens MA, et al. Morphologic and
58. Nucci MR, Prasad CJ, Crum CP, Mutter GL. Mucinous other clinicopathologic features of endometrial clear cell
endometrial epithelial proliferations: a morphologic carcinoma: a comprehensive analysis of 50 rigorously
spectrum of changes with diverse clinical signifi- classified cases. Am J Cancer Res. 2013;3(1):70–95.
cance. Mod Pathol. 1999;12:1137–42. 73. Baergen RN, Warren CD, Isacson C, Ellenson LH. Early
59. Vang R, Tavassoli FA. Proliferative mucinous lesions uterine serous carcinoma: clonal origin of extrauterine
of the endometrium: analysis of existing criteria for disease. Int J Gynecol Pathol. 2001;20:214–9.
diagnosing carcinoma in biopsies and curettings. Int J 74. Prevalence of obesity among adults and youth: United
Surg Pathol. 2003;11:261–70. States, 2011–2014, NCHS Data Brief, No 219;
60. Tresserra F, Lopez-Yarto M, Grases PJ, Ubeda November 2015.
A, Pascual MA, Labastida R. Endometrial hyper- 75. https://seer.cancer.gov/statfacts/html/corp.html.
plasia with secretory changes. Gynecol Oncol. 76. American College of Obstetricians and Gynecologists.
2003;88:386–93. Committee Opinion. Number 631; May 2015.
61. Ambros RA, Sherman ME, Zahn CM, Bitter-man 77. Erkanli S, Ayhan A. Fertility-sparing therapy in young
P, Kurman RJ. Endometrial intraepithelial carci- women with endometrial cancer: 2010 update. Int J
noma: a distinctive lesion specifically associated with Gynecol Cancer. 2010;20(7):1170–87. Review.
tumors displaying serous differentiation. Hum Pathol. 78. American College of Obstetricians and Gynecologists.
1995;26:1260–7. ACOG practice bulletin, clinical management guide-
62. McCluggage WG, Sumathi VP, McManus DT. Uterine lines for obstetrician-gynecologists, number 65,
serous carcinoma and endometrial intraepithelial car- August 2005: management of endometrial cancer.
cinoma arising in endometrial polyps: report of 5 Obstet Gynecol. 2005;106(2):413–25.
Endometrial Carcinoma
10
Contents
Classification of Endometrial Carcinoma 262
Important Issues in Interpretation of Biopsies 264
Criteria for the Diagnosis of Well-Differentiated Endometrioid Carcinoma 264
Malignant Neoplasms: Classification, Grading, and Staging of the Tumor 267
Clinically Important Histologic Subtypes 274
Hereditary Syndromes 285
Histologic Effects After Progestin Therapy 289
Serous Carcinoma 290
Clear Cell Carcinoma 295
Rare Histologic Subtypes 297
Carcinosarcoma 301
References 321
Endometrial carcinoma is the 4th most common rate of mortality when compared to white women
cancer for American women and the most com- [2, 3]. Although most endometrial carcinomas
mon malignant tumor of the female genital tract. that are diagnosed at early stages have an excel-
In the United States in 2015, there were 54,644 lent survival rate, this neoplasm represents a bio-
new cases and 10,096 deaths resulting from this logically and morphologically diverse group of
neoplasm [1]. Black women have a much higher tumors, with differing pathogenesis, behavior,

262 10 Endometrial Carcinoma
and prognosis [4–7]. These tumors were grouped on the surface of a polyp. Serous endometrial
into two types based on clinicopathologic fea- intraepithelial carcinoma (SEIC), the putative
tures designated “Type I and Type II” [5]. This precursor lesion, is frequently associated with
grouping was helpful in understanding the patho- serous and some clear cell carcinomas (see
genesis of endometrial cancer, but it is important Chap. 9). Serous carcinoma usually invades the
to understand that this is not a histopathologic myometrium deeply, permeates lymphatic and
classification. The prototypic Type I tumor is the vascular channels, and not infrequently has
usual type of endometrioid carcinoma, and the spread beyond the uterus at the time of
prototypic Type II tumor is serous carcinoma. hysterectomy.
Type I carcinomas are predominantly well to Although most endometrial carcinomas are
moderately differentiated endometrioid carcino- typical endometrioid carcinomas of all grades
mas, accounting for 80–85% of all endometrial and serous carcinoma, there are several other
carcinomas. Typical patients are obese, hyperten- histologic types as well as a few demonstrating
sive, and diabetic perimenopausal or postmeno- ambiguous features that defy straightforward
pausal women. The tumors are often associated classification [11–15]. For instance, in a large
with atypical hyperplasia/EIN, conditions that study involving the review of 131 high-grade
result from unopposed estrogenic stimulation endometrial carcinomas at two institutions,
such as anovulatory cycles at the time of meno- 38% and 17% of cases, respectively, were
pause or in younger women with polycystic ovar- reclassified with mixed endometrioid and
ian disease. The role of unopposed estrogenic serous carcinoma representing the most com-
stimulation in the causation of a subset of endo- monly misclassified histologic types [15].
metrial carcinomas has been supported by vari- Another review of 56 high-grade endometrial
ous studies, one of which showed that women carcinoma cases found consensus in only
using unopposed estrogenic drugs for more than 62.5% [12].
2 years have a two- to threefold increase in the Most clear cell and grade 3 endometrioid
risk of endometrial cancer, whereas women carcinomas do not appear to fit into the Type I
receiving progestins in conjunction with estrogen vs Type II tumor grouping. Clear cell carcino-
have no increased risk [8]. mas, for instance, have been shown by molecu-
In hysterectomy specimens, low-grade endo- lar studies to be heterogeneous with some being
metrioid tumors generally show minimal myo- closely related to serous carcinomas, while oth-
metrial invasion, although deep invasion can ers bearing greater resemblance to endometri-
occur [9, 10]. The prognosis is generally good, oid tumors [16]. Similarly, grade 3 endometrioid
with a 5-year survival of 80% or better [6]. carcinomas have been shown to represent a
Type II neoplasms represent another, very mixture of molecular subtypes of endometrial
different, form of endometrial carcinoma. These cancer, rather than a homogeneous group. Most,
are characteristically high grade and are not however, are best classified as Type II tumors
related to unopposed estrogenic stimulation [4, [17].
5, 7]. Tumors in this group account for 15–20%
of all endometrial carcinomas. Besides uterine
serous carcinoma, the other carcinomas in the Classification of Endometrial
Type II category include a subset of clear cell Carcinoma
carcinomas and other carcinomas that show
high-grade nuclear features. They tend to occur Recently, immunohistochemistry coupled with
in older postmenopausal women and are not molecular (genome-based) studies has been inte-
usually associated with atypical hyperplasia/ grated with the traditional morphologic classifi-
EIN. In contrast, these neoplasms often occur in cation to stratify risk, determine prognosis, and
a background of atrophic endometrium or arise guide treatment [18].
Classification of Endometrial Carcinoma 263
In 2013, The Cancer Genome Atlas (TCGA) stable (MSS) carcinomas. They do not show
genomic-based classification of endometrial car- an increased mutation rate like MSI tumors
cinoma analyzed 373 endometrial carcinomas nor increased copy number alterations, how-
and proposed a molecular classification for endo- ever, other common abnormalities include
metrial carcinomas which encompasses four PTEN, ARID1A and CTNNB1 mutations.
genomic subtypes: (1) hypermutated/microsatel- The vast majority of these tumors are
lite instability, (2) ultramutated polymerase ε endometrioid.
mutated (POLE), (3) low copy number abnor- 4. Copy number high or serous-like carcinomas
malities, and (4) high copy number abnormali- are characterized by large copy number altera-
ties. Subsequently, independent research groups tions, most frequently TP53 mutations
have identified molecular subgroups that parallel (>90%). Histologically, most tumors are
the genomic-based subtypes defined by TCGA serous or grade 3 endometrioid carcinomas.
[19, 20]. The Vancouver group proposed an algo-
rithmic prognostic system termed ProMisE Large-scale studies combining morphologic,
(Proactive Molecular Risk Classifier for immunohistochemical, and molecular tools are
Endometrial Cancer) in an attempt to utilize the necessary to validate this system. Currently, a
TCGA-derived molecular information in a prac- standard H&E morphologic evaluation is the
tical clinical format [21]. Recently this system most important and cost-effective method of
has been validated for use in biopsy/curettage diagnosing endometrial neoplasia coupled with
specimens with high rates of concordance immunohistochemistry in selected cases.
between biopsy and final hysterectomy results. A Most endometrial tumors diagnosed on a
summary of the proposed molecular subtyping biopsy are subsequently treated by total hysterec-
follows [20–22]: tomy and bilateral salpingo-oophorectomy that
allows precise surgical-pathologic staging [23].
1. Hypermutated or microsatellite instability Nonetheless, identification of more aggressive
(MSI) carcinomas are characterized by loss of tumors is important at the time of biopsy, as these
DNA mismatch protein repair and a high rate neoplasms have greater potential for metastatic
of mutations (although not as high as the spread, including involvement of the peritoneal
POLE-ultramutated subtype). This molecular surfaces, and, therefore, merit more extensive
subtype is seen in up to ¼ of all studied endo- surgical staging.
metrial carcinomas, most of which have an The morphologic diversity of endometrial
endometrioid appearance. Microsatellite carcinoma poses challenges in biopsies and
unstable endometrial carcinomas are part of curettings. For low-grade endometrioid carci-
the hereditary nonpolyposis colorectal cancer noma, distinction from other benign lesions that
(HNPCC) or Lynch syndrome. mimic carcinoma is an important issue.
2. POLE-ultramutated carcinomas account for Identifying and properly classifying aggressive,
approximately 6.5% of all investigated clinically significant histologic subtypes of
tumors, most of which demonstrate an endo- endometrial carcinoma is a second important
metrioid morphology, with a subset of undif- area of biopsy interpretation. Another problem
ferentiated and dedifferentiated tumors also in biopsy interpretation is ascertaining whether
noted. POLE-ultramutated tumors have been the tumor originates in the endometrium or the
associated with morphologic poor prognostic endocervix. This chapter addresses the general
features such as high grade or ambiguous his- classification, staging, and grading of endome-
tology, lymphatic invasion, and, paradoxi- trial carcinoma, the differential diagnosis of
cally, a very favorable outcome. benign lesions versus low-grade endometrioid
3. Copy number low endometrial carcinomas are carcinoma, and the classification of different
part of the low copy number or microsatellite types of carcinoma once the diagnosis of carci-
noma has been established. Where relevant, For practical application, specific patterns of
adjunctive immunohistochemistry and molecu- stromal and epithelial alterations have been
lar testing are also included in this discussion. defined that reflect “endometrial stromal inva-
sion” and identify carcinoma [25, 28]. There are
three separate features, any of which indicates
I mportant Issues in Interpretation stromal invasion in low-grade glandular
of Biopsies proliferations:
There are two major concerns in the evaluation of 1. A confluent glandular pattern in which indi-
endometrial biopsies from the standpoint of a vidual glands, uninterrupted by stroma, merge
diagnosis of carcinoma. First, is the lesion benign and create a cribriform arrangement
or malignant? Second, if the biopsy contains a 2. An irregular infiltration of glands associated
malignant tumor, what is the grade, histologic with an altered fibroblastic stroma (desmo-
subtype, and is it a primary endometrial or an plastic response)
endocervical carcinoma? 3. An extensive papillary pattern
Although quantitative features have limited

riteria for the Diagnosis of Well-
C usefulness in endometrial biopsy diagnosis,
Differentiated Endometrioid these specific and objective criteria for invasion
Carcinoma also should be quantitatively significant.
Therefore, the glandular proliferation that ful-
Because a diagnosis of carcinoma will have an fills criteria for well-differentiated endometri-
important impact on clinical management, it is oid carcinoma should be sufficiently extensive
necessary for the pathologist to be familiar with to involve at least one half of a low-power (×4)
the minimal histologic criteria for that diagnosis. field, a distance of about 2.0 mm in most micro-
One of the most problematic areas is the distinc- scopes. This guideline helps mainly to avoid
tion of atypical hyperplasia/EIN (see Chap. 9) the problem of tangential sectioning or other
from well-differentiated endometrioid carcinoma artifacts in establishing the diagnosis of carci-
[24–27]. Separation of these entities is based on noma. This general rule should not be too rig-
identification of specific morphologic criteria idly applied, particularly in scant specimens. If
that establish the diagnosis of low-grade carci- the features of “stromal invasion” are clear, a
noma. Hyperplasia without atypia generally is diagnosis of carcinoma should be made even if
not a problem in the differential diagnosis, as the diagnostic area does not occupy one half of
these forms of hyperplasia do not have nuclear a low-power field.
atypia. It is important, however, to accurately
separate carcinoma from other, benign changes
that mimic neoplasia, including tissue artifacts, Confluent Gland Pattern
pregnancy-related, and epithelial cytoplasmic
changes (see Chaps. 2, 3, and 9). This pattern reflects invasion by showing a com-
The diagnosis of low-grade endometrioid carci- plete absence of stroma between glands. At times
noma can be difficult at times, because these a cribriform bridging pattern with true “gland in
tumors do not always show clear-cut destructive gland” formation is present. With cribriform
stromal invasion. Furthermore, invasion into myo- growth, trabeculae of columnar cells bridge the
metrium is rarely, if ever, demonstrated in biop- lumen, subdividing it into smaller glandular
sies. Nonetheless, invasion is a logical criterion for spaces (Figs. 10.1, 10.2, and 10.3). No stroma
separating frank carcinoma from atypical hyper- supports the bridging cells. A confluent gland
plasia or other lesions that mimic carcinoma. pattern also is represented by large, irregular
Criteria for the Diagnosis of Well-Differentiated Endometrioid Carcinoma 265
Fig. 10.1 Endometrioid carcinoma, FIGO grade 1. The neoplasm shows well-formed glands with a confluent pattern.
The nuclei are grade 1 with minimal pleomorphism, small nucleoli, and low mitotic rate
glands that interconnect continuously throughout stromal cells and gives an eosinophilic appear-
the field, exceeding the outline of any acceptable ance. The fibrous change also leads to retraction
nonneoplastic gland (Fig. 10.4). Confluent gland and distortion of the normal architecture, result-
patterns should be identified in areas free of squa- ing in a haphazard glandular pattern. Dense
mous differentiation, as squamous morules may stroma in polyps, alteration of the stroma associ-
bridge gland lumens, but these do not reflect stro- ated with marked inflammation, the stroma of the
mal invasion. atypical polypoid adenomyoma, and stroma of
the lower uterine segment all may mimic the des-
moplastic response of carcinoma. In these types
ltered Fibrous or Desmoplastic
A of cases, in which desmoplasia is difficult to eval-
Stroma uate, other features of carcinoma, such as a con-
fluent gland pattern, should be used in establishing
With this change, atypical glands are dispersed in an unequivocal diagnosis of malignancy.
a reactive fibroblastic mesenchyme rather than in
endometrial stromal cells (Fig. 10.5). These
fibrous stromal cells are different from normal Extensive Papillary Pattern
endometrial stromal cells, being more spindle-
shaped and having elongate nuclei. The mesen- The extensive papillary growth pattern is charac-
chyme also contains collagen that compresses the terized by delicate, elongate, branching papillary
Fig. 10.2 Endometrioid carcinoma, FIGO grade 1. Confluent gland pattern. Confluent glands with a cribriform bridg-
ing arrangement. There is no stromal support to the epithelium that bridges the glandular lumens
fronds (Fig. 10.6). The fronds have thin, fibrous and not useful for recognizing carcinoma in most
cores. These papillary structures are much more cases. When this pattern is present, other areas
elaborate and branching than the small papillary usually show confluent gland patterns or an
tufts that may occur in the glands of atypical altered (desmoplastic) stroma.
hyperplasia. Papillary tufts also lack fibrovascu- Applying these criteria for invasion to estab-
lar cores. The diffuse papillary pattern distin- lish the diagnosis of well-differentiated carci-
guishes this form of carcinoma from focal noma yields a clinically significant diagnosis of
alterations, such as papillary arrangements in carcinoma, when present. One study found that
eosinophilic syncytial change or hyperplastic in curettage specimens with well-differentiated
papillary proliferations in polyps (see Chap. 9) endometrioid carcinoma, defined by at least one
[29]. Serous carcinoma with papillary features is of these features of invasion, subsequent hyster-
readily separated from this well-differentiated ectomy specimens showed residual carcinoma in
neoplasm by its high-grade nuclear features and one half of the cases [25]. Frequently the residual
extensive papillary tufting (see later). carcinoma was well differentiated but in about
Originally, another pattern comprised of squa- one third of cases the tumor was grade 2 or 3, and
mous masses, in which sheets of bland squamous in one quarter of the cases with tumor, the myo-
cells form irregular coalescent nests throughout metrium was deeply invaded. In another study
the stroma, was proposed as a criterion for inva- that used the same criteria to assess “stromal
sion [25]. This pattern is rare by itself, however, invasion,” 16% of patients without stromal inva-
Fig. 10.3 Endometrioid carcinoma, FIGO grade 1. Confluent glands with a fused or cribriform gland arrangement.
Although the contours of some glands can still be traced, some others have become fused, characterizing stromal breach
sion in endometrial samples had myometrial the neoplasm, it is important to ascertain whether
invasion in the hysterectomy specimens whereas or not the tumor is primarily of endometrial ori-
62.5% of patients with stromal invasion in the gin or whether it arises in the cervix. Finally,
biopsy had myometrial invasion [30]. These stud- especially if the tumor displays an unusual pat-
ies show the utility of these criteria for determin- tern, the possibility of a metastasis from another
ing the presence of well-differentiated site should be considered.
endometrioid carcinoma.
Classification
Malignant Neoplasms:
Classification, Grading, and Staging The current WHO histologic classification recog-
of the Tumor nizes several distinct types of carcinoma that are
important to identify in biopsies (Table 10.1) [31].
Once the diagnosis of carcinoma is established, it Based predominantly on morphologic fea-
is important that the tumor be properly classified tures, the current 2014 WHO classification of
to identify aggressive forms (Table 10.1). Also, epithelial tumors of the uterine corpus recog-
because the biopsy or curettage is generally a nizes the following histologic types of endome-
blind procedure with no direct visualization of trial carcinomas: endometrioid carcinoma with
Fig. 10.4 Endometrioid carcinoma, FIGO grade 1. Confluent gland pattern. Large, irregular glands that interconnect
continuously throughout the field
squamous differentiation, villoglandular and as endometrioid carcinoma and graded in the

secretory variants, mucinous carcinoma, serous microscopic diagnosis.
endometrial intraepithelial carcinoma (SEIC), From 20% to 30% of endometrial carcinomas
serous carcinoma, clear cell carcinoma, neuro- show an “endometrioid” pattern with squamous
endocrine tumors (low-grade neuroendocrine differentiation [36]. Previously, tumors with
tumor/carcinoid and high-grade neuroendocrine squamous differentiation were separated into
carcinoma, small and large cell types), mixed two categories: “adenoacanthoma” denoted
cell adenocarcinoma, and undifferentiated and tumors that had cytologically benign-appearing
dedifferentiated carcinomas. Table 10.1 summa- squamous epithelium (squamous metaplasia),
rizes the current classification of endometrial and “adenosquamous carcinoma” denoted
carcinoma. tumors that had a cytologically malignant squa-
Approximately 70–80% of endometrial carci- mous component. Nevertheless, studies have
nomas have the “typical,” “usual,” or “not other- shown that endometrioid carcinomas with or
wise specified (NOS)” pattern referred to as without squamous epithelium behave in the
“endometrioid” carcinoma [6, 32–35]. The term same fashion when stratified according to the
“endometrioid” provides a specific designation grade of the glandular component [36–38].
for this neoplastic pattern, clearly separating it Accordingly, these tumors are best classified as
from the other histologic types of endometrial carcinomas with squamous differentiation and
carcinoma. Practically, ordinary endometrial car- graded. The terms “adenoacanthoma” and “ade-
cinoma with an endometrioid pattern is classified nosquamous carcinoma” are no longer used. The
Fig. 10.5 Endometrioid carcinoma, FIGO grade 1. Desmoplastic stroma. A fibroblastic mesenchyme encompasses the
neoplastic glands
other histologic types of endometrial carcinoma The traditional grading of endometrial carci-
are much less frequent. Serous tumors account noma standardized by FIGO uses a three-grade
for about 10%, mucinous tumors for roughly system based on architectural features. Tumors
5%, and clear cell carcinoma for 2–5% of all are grade 1 when most (95%) of the tumor forms
endometrial carcinomas [6]. glands (Figs. 10.1, 10.2, 10.4, and 10.5), grade 2
when 6–50% of the tumor exhibits solid growth
(Fig. 10.7), and grade 3 when more than 50% of
Grading the tumor has a solid growth pattern (Fig. 10.8).
Solid growth in this grading system consists of
In addition to identifying specific histologic sub- areas with glandular components but limited
types in biopsies, the histologic grade provides gland formation. It is important to avoid areas of
useful prognostic information and assists in plan- squamous change in making this assessment.
ning treatment. Low-grade tumors usually are Like all morphologic grading systems, this
confined to the corpus at the time of diagnosis, grading scheme is somewhat subjective, espe-
and the overall survival is very good. High-grade cially at the cut points between grades. For exam-
tumors, in contrast, are more aggressive and have ple, determination of more than 5% solid growth
a poor prognosis. This latter group is more likely that establishes grade 2 instead of grade 1 tumor
to spread beyond the corpus and involve the is an estimate and is subject to individual varia-
endocervix or extrauterine sites at the time of tion. In addition, significant variation in architec-
diagnosis. tural grade can be seen within a tumor.
Fig. 10.6 Endometrioid carcinoma, FIGO grade 1. Extensive papillary pattern. This low-grade carcinoma with a vil-
loglandular pattern forms multiple delicate papillae
Table 10.1 Classification of Endometrial Carcinoma

Well-formed glands (grade 1 tumor) may be adja-
Endometrioid carcinoma
cent to solid masses of grade 3 carcinoma. Not
Squamous differentiation
surprisingly, the heterogeneity of tumor grade
Villoglandular
Secretory
can result in discordance between curettage and
Mucinous carcinoma hysterectomy specimens.
Serous endometrial intraepithelial carcinoma Although architectural grading generally cor-
Serous carcinoma relates with prognosis, studies indicate the grad-
Clear cell carcinoma ing system can be improved by incorporating
Neuroendocrine tumors nuclear grading [39–42]. Consequently, the FIGO
Low-grade neuroendocrine tumor (carcinoid) and WHO Histopathologic Classification modi-
High-grade neuroendocrine carcinoma fied the standard architectural grading system to
Small cell
include nuclear grade [23, 43]. Like architectural
Large cell
grading, nuclear grading is somewhat subjective.
Mixed cell adenocarcinomaa
Undifferentiated carcinoma Nuclei with small (2–3 times the diameter of a
Dedifferentiated carcinoma lymphocyte), relatively uniform (little variation in
Tumor with greater than 5% of a second cell type
a size and shape), oval nuclei, evenly dispersed
Fig. 10.7 Endometrioid carcinoma, FIGO grade 2. Although glandular differentiation is readily apparent, a substantial
amount (roughly 30%) of the tumor has a solid growth pattern
chromatin and indistinct nucleoli are grade 1 becomes grade 3). Notable nuclear atypia was
(Fig. 10.9). Nuclei with features between grade 1 not defined, but one study found that for clinical
and grade 3 are grade 2 (more variation in size and significance, nuclear atypia should be grade 3
shape, but less pleomorphism and chromatin [39]. For example, a tumor with grade 1 architec-
clumping than grade 3) (Figs. 10.10 and 10.11). ture but grade 2 nuclei would be given a final
Large (greater than six times the diameter of a FIGO grade of 1, whereas an architectural grade
lymphocyte), highly pleomorphic nuclei with 1 tumor with grade 3 nuclei should be given a
irregular outlines, coarse, clumped, or smudged final FIGO grade of 2. It should be noted that the
chromatin and one or multiple macronucleoli are combination of grade 1 architecture and grade 3
grade 3 (Fig. 10.12 and Table 10.2). Typically, nuclei is quite uncommon in endometrioid carci-
mitotic figures, including atypical forms, increase noma, and therefore it is unusual for nuclear
from grades 1 to 3 [44, 45]. atypia to upgrade architectural grade. In fact, this
FIGO recommends that “notable” nuclear combination should be regarded as a red flag for
atypia, inappropriate for the architectural grade, the diagnosis of serous carcinoma (see later).
raises an architectural grade 1 or grade 2 tumor Other histologic features that influence the
by 1 (i.e., grade 1 becomes grade 2, and grade 2 grade according to the FIGO system are:
Fig. 10.8 Endometrioid carcinoma, FIGO grade 3. A few residual glands are present, but more than 50% of the tumor
showed a solid growth pattern with sheets and nests of malignant epithelial cells
1. Carcinomas with squamous differentiation are A recent study using a molecular (POLE muta-
graded according to the nuclear grade of the tion) and immunohistochemical (TP53 and MSI)
glandular component. algorithmic approach to both the index (biopsy)
2. Serous, clear cell and squamous cell carcino- and final (hysterectomy) diagnoses showed a high
mas should not be FIGO graded. The first two concordance rate of endometrial cancer classifica-
are, by definition, high grade. tion between the two, as well as improved agree-
ment between the index and final diagnoses when
The grade of the tumor from the biopsy speci- compared to histologic evaluation alone [22].
men agrees with the grade in the hysterectomy in Although a surrogate immunohistochemical
fewer than 60% of cases, with both apparent marker for POLE mutation is not yet available for
under-grading and over-grading on biopsies, as use in clinical practice, most laboratories can per-
compared to the hysterectomy specimens [46– form TP53 and MSI immunostains in biopsy
50]. Over-grading may be caused by heterogene- material. This approach may allow for better
ity of the tumor, with the surface component treatment guidance before the final hysterectomy
removed at biopsy showing the higher grade. specimen is evaluated, and further studies may
Under-grading in biopsy specimens is most often validate this methodology.
due to limited sample size. Thus, the biopsy gives Accurate classification of the tumors in biopsy
a general assessment of the degree of differentia- material gives the clinician a reasonable expecta-
tion of a tumor. tion of the degree of malignancy anticipated in
Fig. 10.9 Nuclear grade 1. Nuclei are round to oval with small nucleoli. The mitotic rate is low
the hysterectomy specimen. In occasional cases, ing has not been uniformly applied throughout
biopsy or curettage will remove all or most of the the studies, however, and often has failed to
carcinoma, so there can be insufficient tumor or incorporate all cytologic features in determining
only atypical hyperplasia in the hysterectomy for the final nuclear grade. For example, one study
further evaluation. considered a tumor to have high-grade nuclei if it
The utility of nuclear grade in the assessment showed “large nucleoli” and “coarsely clumped
of endometrioid carcinoma has been examined in chromatin,” but not necessarily nuclear pleomor-
several reports[33, 39, 44, 51–53]. In general, a phism [65]. In our opinion, however, nuclear
correlation between nuclear and architectural grading should consider all features, including
grade has been found [32, 44, 45, 51, 52], pleomorphism, mitotic activity, and abnormal
although some reports found interobserver repro- mitotic figures (Table 10.2). Also, the current
ducibility for nuclear grading was not as good as schemes for nuclear grading are relatively sub-
for architectural grading [39, 54]. Nuclear grad- jective, and in our experience, endometrioid car-
ing in hysterectomy specimens can identify a cinoma with grade 3 nuclei is almost always solid
small subset of nuclear grade 3 tumors that have and would have been graded as 3 based on archi-
a poorer prognosis but do not show grade 3 architecture alone. Before diagnosing tumors with a
tectural patterns. Some studies suggest that predominant grade 1 architectural appearance
nuclear grading is not an independent index of and nuclear grade 3 as endometrioid carcinoma, a
prognostic utility compared to the FIGO archi- serous or clear cell carcinoma should be carefully
tectural grade [53]. The method of nuclear grad- considered. Most endometrioid carcinomas with
Fig. 10.10 Nuclear grade 2. In contrast to grade 1, these nuclei are more pleomorphic and have coarser chromatin with
larger nucleoli
high nuclear grade concomitantly display some groups. This grading system was evaluated only
degree of solid growth and are almost never on hysterectomy specimens, however, and its
exclusively glandular. utility in biopsy material needs further evalua-
Several investigators have proposed a two- tion. Another study suggested that reducing the
grade or binary grading system in place of the FIGO grading system to a two-tiered system by
FIGO three-grade system [54, 55]. One analysis combining FIGO grades 1 and 2 in comparison
used a system with 20% or more solid tumor to with FIGO grade 3 was another reproducible
separate high-grade from low-grade tumors [55]. method of defining tumors with a significant dif-
This two-tiered system found improved interob- ference in prognosis [56].
server agreement and better prediction of uterine
histology at hysterectomy. In another study uti-
lizing a binary grading system, tumors that dis- linically Important Histologic
C
played two or more of the following features: (1) Subtypes
>50% solid growth (glandular or squamous), (2)
tumor cell necrosis, or (3) an infiltrating as Typical (Endometrioid) Carcinoma
opposed to an expansile pattern of invasion were
classified as high grade [54]. This binary grading This type of endometrioid carcinoma typically
was useful in separating tumors into prognostic shows a glandular pattern (Figs. 10.10 and 10.13)
Fig. 10.11 Nuclear grade 2. Although these nuclei show vesicular chromatin and prominent nucleoli, they are not quite
as abnormal as nuclear grade 3
with prominent cribriform bridging, but not glands are lined by cells with cilia along the
infrequently it may display a papillary pattern luminal border [6, 60, 61]. Its significance lies in
that has been referred to as “villoglandular” the recognition that cilia do not always indicate a
(Fig. 10.6) [6, 34, 57, 58]. Glands may contain a benign lesion.
small amount of mucin or necrotic debris In the secretory variant of endometrial carci-
(Fig. 10.14a and b). The cells lining the glands noma, the neoplastic glands are lined by cells
have a moderate amount of eosinophilic to with vacuolated cytoplasm (Fig. 10.15) [6, 62–
amphophilic cytoplasm (Fig. 10.11), [6, 59] and 64]. Cytoplasmic vacuolization is a feature also
the cell membrane at the luminal border of the seen in many clear cell carcinomas (see later,
glands often is ill defined. In the papillary tumors, Clear Cell Carcinoma), but it is important to sep-
the cells are columnar and perpendicular to the arate the low-grade secretory carcinoma from
fibrovascular cores. Nuclei are oval and relatively clear cell carcinoma, which is a high-grade neo-
bland (see later, Serous Carcinoma, for further plasm. In secretory carcinoma, clear vacuoles fill
discussion of papillary patterns). the subnuclear or supranuclear cytoplasm, and
Other variants of endometrioid carcinoma the cells resemble those seen in normal early
include ciliated carcinoma and secretory carci- secretory phase. The nuclei usually show mini-
noma [6, 34, 59]. Ciliated carcinoma is an mal atypia, although the glands fulfill the criteria
extremely rare neoplasm in which the invasive for invasion. These rare tumors have an excellent
Fig. 10.12 Nuclear grade 3. Nuclei are markedly tin. This tumor shows a papillary and glandular morphol-
enlarged and highly pleomorphic. Notice the prominent ogy resembling serous carcinoma
nucleoli, numerous macronucleoli, and vesicular chroma-
Table 10.2 Nuclear grading in endometrial carcinoma glandular and stromal breakdown (see Chap. 5)
Grade 1. Oval/elongated nuclei, fine chromatin, small [66]. Some endometrioid carcinomas show a
nucleoli, few mitoses marked neutrophilic infiltrate. Polymorphonuclear
Grade 2. (Features between grades 1 and 3) leukocytes can be intimately admixed with the
Grade 3. Enlarged/pleomorphic nuclei, coarse tumor, and the tumor cells can show apparent
chromatin, prominent nucleoli, many mitoses phagocytosis of neutrophils (Figs. 10.16 and
10.17). The neutrophilic response has no known
prognosis. Secretory carcinoma can occur in pre- effect on the prognosis.
menopausal or postmenopausal patients and is Most low-grade endometrioid carcinomas
not necessarily related to progestin treatment. demonstrate positive immunolabeling for estro-
Foam cells often are present in the stroma of gen and progesterone hormone receptors, while
endometrioid carcinoma or its variants, especially high-grade endometrioid tumors may progres-
when they are low grade [59, 65]. The presence of sively lose hormone expression (Table 10.4) [67,
foam cells by themselves does not influence the 68]. The vimentin immunostain is consistently
diagnosis or the classification of endometrioid expressed in endometrioid carcinomas, while
carcinoma as foam cells can occur in a variety of CEA can be seen in up to 50% of cases, typically
benign conditions in which there is abnormal in an apical localization, unless there is overt
Fig. 10.13 Typical endometrioid carcinoma pattern. Classic “endometrioid” pattern with the glands lined by cells with
scant, pale cytoplasm and stratified nuclei. The nuclear grade is 1 in this example. Focal squamous change is also seen
a b
Fig. 10.14 Endometrioid carcinoma, FIGO grade 1. (a) standards. (b) Necrotic foci may occasionally indicate the
The glandular cells have mucinous and inflammatory con- presence of a higher-grade neoplasm; however, by itself it
tents within the glandular lumens. This finding does not is not used for tumor grading
affect the classification of the carcinoma based on current
Fig. 10.15 Secretory carcinoma, FIGO grade 1. This phase endometrium. The secretory changes are accompa-
variant shows neoplastic glandular cells with extensive nied by low-grade nuclei
cytoplasmic vacuoles, superficially resembling secretory
mucinous differentiation, in which case CEA can mutant-pattern or “serous-like” immunoprofile

be more diffuse and overlap with the usual-type (Table 10.4) [18, 71]. In such cases an expanded
endocervical adenocarcinoma profile. p16 immu- immunohistochemistry panel to include ER, PR,
nostaining typically shows focal/patchy reactivity and p16 is very helpful for classification.
in endometrioid carcinomas, while strong/diffuse
staining pattern in the vast majority of usual-type
endocervical adenocarcinomas (see Table 10.4) Carcinoma with Squamous
[69, 70]. Accordingly, none of these stains alone Differentiation
are particularly helpful in the diagnosis with the
possible exception of p16 (see below). p53 immu- Squamous differentiation commonly occurs in
nostain generally has a patchy/weak immunopro- tumors with a typical (endometrioid) glandular pat-
file (wild-type TP53 expression) in FIGO grades tern [6]. It is rarely, if ever, associated with serous
1 and 2 endometrioid c arcinomas. About one third or clear cell carcinoma. At least 10% of the tumor
of morphologically typical high-grade endometri- should have squamous features for it to qualify as
oid carcinomas with greater than 50% solid carcinoma with squamous differentiation.
growth and associated squamous differentiation Typically, the squamous epithelium is intimately
will, however, demonstrate an aberrant TP53 admixed with glands (Figs. 10.18 and 10.19).
Fig. 10.16 Endometrioid carcinoma, FIGO grade 1. Frequently, endometrioid carcinomas disclose abundant neutro-
philic infiltrates, both in the glandular epithelium and intraluminally
In the low-grade neoplasms, the squamous malignant squamous component often are com-
changes often include so-called morules, rounded posed of nests of spindle-shaped cells that may
masses of bland squamous cells largely filling the obliterate gland lumens (Figs. 10.20 and 10.21).
lumens of the malignant glands. These squamous Keratinization and squamous pearl formation are
cells are incompletely differentiated and have frequently apparent. Abundant keratin formation
eosinophilic cytoplasm with indistinct cell borders. may even incite a foreign body response. Mitotic
The nuclei are uniform, bland, and lack prominent activity often is brisk in the squamous
nucleoli; they do not palisade. Mitotic figures are component.
infrequent. The squamous cells can show intercel- At times a clear distinction between cytologi-
lular bridges, but this finding is infrequent. Often cally benign and malignant squamous epithelium
these squamous nests are nonkeratinizing, but kera- is not possible. The squamous component may
tinization can, at times, be present. show mild degrees of atypia and scattered mitotic
When the squamous component appears figures. In these cases, the cytologic features
malignant, it is usually associated with a neo- exceed a “benign” appearance but do not have all
plasm that is grade 2 or 3. Portions of the tumor the characteristics of malignancy required for a
may show squamous carcinoma without glandu- diagnosis of adenosquamous carcinoma.
lar differentiation. Tumors with a cytologically Furthermore, this squamous change, because it
Fig. 10.17 Endometrioid carcinoma, FIGO grade 1. The presence of abundant neutrophilic infiltrates bears no clinical
significance and does not affect grading
Table 10.3 Histologic features of serous carcinoma

oid carcinoma lacking squamous differentiation
Complex, coarse papillae
[36–38]. Accordingly, it is the grade of the glan-
Irregular, gaping glands
dular component that has prognostic significance.
Papillary tufting
High nuclear/cytoplasmic ratio
For these reasons, the term “endometrioid carci-
Marked nuclear pleomorphism noma with squamous differentiation” is pre-
Numerous and abnormal mitoses ferred. These tumors should be graded 1, 2, or 3
Macronucleoli based on the architectural and nuclear features of
Clear cell changesa the glandular component.
Psammoma bodiesa Finally, unusual morphologic variants of
a
Nonspecific features found occasionally in serous endometrioid carcinoma may occasionally pres-
carcinoma ent in biopsy and curettage material and cause
diagnostic confusion. For instance, the “corded
occurs in adenocarcinoma, is malignant, regard- and hyalinizing endometrioid carcinoma vari-
less of its histology. When these tumors are strati- ant” having a sex cord-like pattern is an endo-
fied by grade and depth of myometrial invasion, metrioid neoplasm that can be confused with
the presence of squamous epithelium does not carcinosarcomas due to its prominent spindle
alter the prognosis when compared to endometri- cell component (Fig. 10.22) (see below) [72].
Table 10.4 Useful immunohistochemical panel to differentiate subtypes of endometrial carcinoma and endocervical
adenocarcinoma
Usual-type
EMCA Undifferentiated/ endocervical
EMCA (LG) (HG) Serous CA Clear cell CA dedifferentiated CA adenocarcinoma
AE1/ +++ +++ +++ +++ +(focal)e +++
AE3
CAM +++ +++ +++ +++ +(focal)e +++
5.2
EMA +++ +++ +++ +++ +(focal)e +++
ER/PR +++ +++a −/− −/− −/− −/−
or + (weak)
p16 −/+ (patchy/ ++ +++ (strong/ +++ (diffuse) in +++ (50% of +++ (strong/
weak) (patchy) diffuse) 50% of tumors tumors) diffuse)
p53 Wild-type Wild-type +++ (strong/ Wild-type +++ (strong/ −f
pattern patternb diffuse)c patternd diffuse)
HNF-1β −/+ −/+ − +++ (nuclear) − −
Napsin − − − +++ − −
A (cytoplasmic)
PAX-8 ++ ++ +++ (strong/ ++ − −/+g
(patchy) diffuse)
EMCA endometrioid carcinoma, LG low grade, HG high grade, CA carcinoma
a
Approximately half of the high-grade EMCA expresses ER/PR [289]
b
A minority of high-grade EMCA demonstrates strong/diffuse p53 overexpression (>75% of tumor nuclei) [71]
c
Most cases of serous carcinoma demonstrate strong and diffuse (>75% of tumor nuclei) immunolabeling; however, a
small proportion displays other distinct aberrant p53 immunolabeling patterns (see Serous Carcinoma discussion)
d
One third of clear cell carcinomas have an aberrant p53 expression due to TP53 mutation [290]
e
The keratin expression in undifferentiated carcinomas is focal (<10% of tumor cells) but intense, especially for EMA
and CK-18 [87]
f
Although typically shows no expression in the usual-type endocervical adenocarcinomas, p53 has been reported posi-
tive in up to 52% of gastric-type non-HPV-related endocervical adenocarcinomas [291]
g
PAX-8 is reported negative, weak, and strongly positive in 35%, 38%, and 27% of usual-type endocervical adenocar-
cinomas [292]
Mucinous Carcinoma Mucinous carcinomas tend to be well to

moderately differentiated, and they frequently
Mucinous carcinoma of the endometrium has a have a papillary or villous architecture. Usually
glandular architecture resembling endometrioid the mucinous tumor merges with a more typi-
carcinoma but is composed of cells containing cal endometrioid pattern. Foam cells are often
abundant intracytoplasmic mucin (Figs. 10.23 present, and a neutrophilic infiltrate may be
and 10.24) [6, 59, 73–75]. It is cytoplasmic seen.
mucin, not extracellular and luminal that The morphologic features of low-grade
establishes the diagnosis of endometrial muci- mucinous carcinoma overlap with the patterns
nous carcinoma. Furthermore, the presence of of so-called secretory carcinoma; both types of
cytoplasmic mucin should be extensive, involv- tumors have abundant pale cytoplasm and
ing greater than 50% of the cells, for a tumor to basal nuclei with minimal stratification. In
be classified as mucinous carcinoma, as some mucinous carcinoma, the cells contain cyto-
mucin production is present in most endometri- plasmic mucin, whereas in secretory carci-
oid carcinomas [74, 76]. noma, the cytoplasmic vacuoles contain
Fig. 10.18 Endometrioid carcinoma with squamous differentiation. Multiple foci of squamous change blend with this
well-differentiated endometrioid carcinoma (FIGO grade 1)
glycogen. The distinction is largely academic, endometrial origin. In contrast, the presence of
however, because tumor grade rather than cyto- numerous apical mitotic figures or apoptotic
plasmic differentiation determines prognosis. bodies is more typical of usual endocervical
These neoplasms show no difference in behav- adenocarcinoma. Noteworthy, the subtype of
ior from endometrioid tumors of similar grade endocervical adenocarcinoma that more fre-
[74]. Up to 9% of all stage I endometrioid car- quently displays prominent cytoplasmic mucin
cinomas are of the mucinous type when strict (gastric-type endocervical adenocarcinoma)
criteria are applied [74]. lacks association with high-risk human papil-
The differential diagnosis of mucinous carci- loma virus (HPV). Thus, a negative – or weak/
noma of the endometrium includes an endocer- patchy positive – p16 immunolabeling is antici-
vical primary, as the cell population of mucinous pated in this scenario.
carcinoma resembles endocervical epithelium, ER and PR immunostains may be helpful as
with basal nuclei and abundant supranuclear most endometrial mucinous carcinomas are posi-
cytoplasm that contains mucin. The presence of tive for ER and PR. Usual-type endocervical
an associated hyperplastic endometrium or a adenocarcinoma typically displays the opposite
conventional endometrioid carcinoma favors pattern. This differential diagnosis is further dis-
Fig. 10.19 Endometrioid carcinoma with squamous differentiation. Foci of squamous change (top of the image) are
intimately mixed with the glandular component of this well-differentiated endometrioid carcinoma (FIGO grade 1)
cussed in greater detail (see below, Differential Microglandular (Endocervical-Like)

Diagnosis). Endometrial Carcinoma
Finally, occasional endometrial mucinous
glandular proliferations fall short of the estab- Occasional endometrial well-differentiated muci-
lished criteria used to definitively diagnose nous carcinomas have patterns that mimic cervi-
carcinoma, either due to small size of the lesion cal microglandular hyperplasia [78, 79]. These
or presence of only minimal focal confluent/ tumors are typically seen in postmenopausal
cribriform architecture and nuclear atypia. In women taking hormones. They are composed of
such cases, it is justified to apply the term numerous small mucinous glands, intermixed
“atypical mucinous glandular proliferations.” with squamous epithelium and abundant acute
One study has shown that 83% of these cases inflammatory cells in the lumens and the inter-
demonstrated an atypical lesion in the uterus vening stroma (Fig. 10.25). These tumors have
when a hysterectomy was subsequently per- minimal cytologic atypia and inconspicuous
formed [77]. mitotic activity, thus closely resembling an exu-
Fig. 10.20 Endometrioid carcinoma with squamous dif- glandular formation in greater than 50% of the neoplasm.
ferentiation. Poorly differentiated endometrioid carcinoma This tumor is intimately associated with squamous mor-
(FIGO grade 3) is characterized by solid nests lacking ules. Marked nuclear atypia is depicted in this image
berant endocervical microglandular hyperplasia. expression of CD10 supports the diagnosis of

They do typically also predominate in the tumor endometrioid carcinoma with microglandular
surface and, on follow-up hysterectomy, show a hyperplasia-like features over endocervical micro-
transition to a more deeply located ordinary glandular hyperplasia [70, 80, 81].
endometrioid carcinoma. Finally, in biopsy and curettage material, fre-
The criteria outlined above for distinguishing quently a definitive diagnosis cannot be achieved.
endometrial and endocervical primary tumors also In such cases, we have used the term “complex
may help in the recognition of these carcinomas. mucinous glandular proliferation with micro-
There is, however, significant overlap not only in glandular hyperplasia-type features” followed by
the morphology but also in the immunoprofile. an explanatory note recommending further sam-
Nevertheless, a negative immunolabeling for pling to exclude a malignant neoplasm or correla-
PAX2 and p63 coupled with strong positive tion with imaging studies.
Fig. 10.21 Endometrioid carcinoma with squamous dif- which also demonstrate highly atypical nuclear features.
ferentiation. In this tumor, foci of glandular endometrioid The squamous component should be disregarded when
differentiation are seen intermingled with a solid compo- assigning FIGO grade. A collection of histiocytes is pres-
nent difficult to separate from the squamous cell nests, ent at the lower left corner
Hereditary Syndromes senting cancer in greater than half of the affected

women. Accordingly, its identification is of
ereditary Nonpolyposis Colorectal
H utmost importance for the patient’s and her off-
Cancer (Lynch Syndrome) springs’ future surveillance [82]. Hereditary non-
polyposis colorectal cancer (HNPCC) or Lynch
Generally, endometrial carcinomas associated syndrome (LS) is caused by a functional defi-
with hereditary syndromes are endometrioid in ciency of the DNA mismatch repair genes
type. Lynch syndrome is most well known for its (MSH2, MLH1, MSH6, and PMS2) typically due
frequent association with right-sided colorectal to germ line mutations. It accounts for 5% of all
cancers in young patients (≤ 50 years old), but, in endometrial cancer cases [83]. The mutations in
fact, endometrial carcinoma is not only an inte- these DNA mismatch repair (MMR) genes result
gral part of Lynch syndrome but also is the pre- in microsatellite instability (MSI). Microsatellites
Fig. 10.22 Endometrioid carcinoma with sex cord-like “low-grade carcinosarcoma,” the latter is typically bipha-
formations and hyalinization, FIGO grade 1. Although sic with carcinomatous and sarcomatous elements sharply
some overlapping features exist between the “corded and separated
hyalinizing” variant of endometrioid carcinoma and a
are multiple nucleotide repeats that are very 50 years old and, by contemporary Bethesda
prone to errors of replication. The DNA-MMR guidelines and Amsterdam II criteria would be
genes are responsible for correcting these errors; missed in 64% and 42% of cases [88], respec-
however, when deficient, microsatellite instabil- tively, it has become evident that universal
ity occurs [84, 85]. screening should be offered to all women of all
The endometrial carcinomas associated with ages, regardless of family history, when they are
Lynch syndrome have a predilection to involve first diagnosed with endometrial cancer.[89]
the lower uterine segment [86]. Endometrioid Recently, a large single institution study corrobo-
carcinoma variants are reported to be more com- rated that when only the traditional indicators of
mon than nonendometrioid (serous, clear cell) LS (e.g., strong family history, ≤ 50 years of age,
variants in LS. Dedifferentiated and undifferenti- multiple concomitant cancers) are used for
ated histologies (see definition below) are fre- screening, up to 41% of Lynch syndrome patients
quently seen in LS but can also be seen are not identified [89].
sporadically due to epigenetic MLH1 promoter DNA-MMR protein immunohistochemistry
hypermethylation [87]. can be easily implemented in most laboratories
Because most patients with LS that present as a preferred primary screening (not diagnostic)
with a gynecologic malignancy will be older than tool. Nevertheless, interpretation of stains can be
Fig. 10.23 Mucinous carcinoma. In this FIGO grade 1 tumor, the glandular cells have abundant cytoplasmic mucin.
Nuclei are oriented along the basal portion of the cells, resulting in a resemblance to endocervical epithelium
challenging and should follow strict recommen- [91]. If germ line mutations are not identified,
dations [90]. An intact expression refers to intact then somatic mutational analysis should be con-
nuclear immunohistochemical expression, while sidered. Meanwhile, loss of expression of MLH1
a loss of expression refers to complete loss of and/or PMS2 should be tested for MLH1 meth-
staining in all tumor cells when the internal con- ylation as this epigenetic phenomenon is more
trol (e.g., lymphocytes, uninvolved endometrial likely than a germ line mutation. The MLH1
stromal cells) shows preserved nuclear methylation assay can be performed on paraffin
expression. block with tumor tissue to detect hypermethyl-
Interpretation can be challenging, especially ation of the MLH1 gene promoter. Real-time
in situations of heterogeneous expression. PCR quantifies CpG methylation within the pro-
Nevertheless, loss of immunohistochemical moter. If MLH1 methylation is absent, then
expression of the pair MSH2/MSH6 or each of MLH1 sequencing is recommended [92].
these dimers individually may signify a possible Microsatellite instability (MSI) analysis by
MSH2 and/or MSH6 germ line mutation. polymerase chain reaction (PCR) is an alterna-
Alternatively, an EPCAM deletion is a possibility tive primary screening to DNA-MMR immuno-
when both MSH2/MSH6 are unmutated upon histochemistry. It can be the first line test or
sequencing (so-called “Lynch-like” cancers) subsequently ordered when the DNA-MMR pro-
Fig. 10.24 Mucinous carcinoma. Portions of these carcinomas often appear extremely well differentiated, because the
mucinous cytoplasm results in basal alignment of the nuclei with minimal nuclear stratification
tein expressions are intact by immunohisto- Cowden Syndrome

chemistry, but there is a strong clinical suspicion
for LS. This is a costly test performed at a Cowden syndrome (CS) is an autosomal domi-
molecular laboratory, which may be a disadvan- nant disorder frequently associated with germ line
tage compared to immunohistochemistry. mutations in the PTEN tumor suppressor gene
Moreover, this test requires tumor and normal located on chromosome 10 [93]. It is character-
tissue, the latter which frequently is lacking in ized by several hamartomas and malignancies that
biopsy or curettage material that only sampled develop in multiple organs with increased risk for
tumor. When instability of two or more of the the breast and, to a lesser extent, thyroid, kidney,
five tested microsatellite markers is present, the colon, and endometrium [94].
tumor is deemed MSI- high and the patient The lifetime risk of endometrial carcinoma
should be referred to germ line mutational analy- in women with CS is estimated to be between
sis. When instability is found in only one micro- 5% and 10% compared to 2.6% in the general
satellite marker, then the case is considered population. Endometrial carcinoma occurs in
indeterminate and should also be referred to up to 28% of CS patients. In contrast to Lynch
germ line mutational analysis [92]. syndrome, there are no pathologic features that
Histologic Effects After Progestin Therapy 289
Fig. 10.25 Endometrioid carcinoma with endocervical- the confluent, cribriform pattern of carcinoma, however.
like pattern. This FIGO grade 1 carcinoma is composed of Neutrophils are present in the epithelium and the extracel-
anastomosing small glands that have a superficial resem- lular mucin. The nuclei are low grade
blance to microglandular hyperplasia of the cervix. This is
suggest CS. Most tumors are low-grade endo- istologic Effects After Progestin
H
metrioid carcinomas with loss of PTEN by Therapy
immunohistochemistry, the latter which is also
commonly seen in sporadic cases [95]. Pathologists may be asked to evaluate endome-
In view of the increased lifetime risk of devel- trial samples with atypical hyperplasia and/or car-
oping endometrial cancer, specific CS surveil- cinoma following progestin therapy for
lance guidelines currently recommend that assessment of response (see Chap. 6). Histologic
women with Cowden syndrome be screened for changes associated with progestin-treated
endometrial carcinoma with blind biopsies annu- endometria include a decreased gland-to-stroma
ally starting at 35–40 years of age or 5 years ratio, decreased glandular cellularity, decreased to
prior to the earliest diagnosis of endometrial car- absent mitotic activity, loss of cytologic atypia,
cinoma in the family. For postmenopausal and cytoplasmic changes including mucinous,
women, annual screening with ultrasound is secretory, squamous, and eosinophilic metaplasia.
advised [95, 96]. Decreased glandular cellularity throughout the
treatment period was defined as a reduction in the Serous Carcinoma

number of abnormal glandular cells in one quarter
of a high-power field [97, 98]. Although architec- Serous carcinoma accounts for roughly 10% of all
tural changes tend to regress during treatment, endometrial carcinomas; however, it is responsi-
cribriform and papillary architectural patterns ble for almost 40% of uterine corpus cancer
may be induced by progestins and mimic progres- deaths [100]. Generally, it is a combination of a
sion. Decreased glandular cellularity, absence of low-grade architecture (papillary or glandular pat-
posttreatment mitotic index, and decreased tern) and high-grade nuclear atypia that identifies
mitotic index over treatment period were the only serous carcinoma [57, 101–107]. The fibrotic,
histologic parameters significantly associated papillary fronds are lined by a stratified layer of
with complete response. Other histologic findings cells with a high nuclear/cytoplasmic ratio and
such as presence of metaplasia, cribriform or pap- high-grade nuclei (Table 10.3). The malignant
illary glands, and cystic atrophy were not signifi- cells also form complex papillary tufts that may
cantly associated with achievement of complete be present as free-floating clusters of cells
response. Another study found that persistent (Fig. 10.26). These papillary tufts often lack
cytologic atypia and architectural abnormalities stroma and are composed of dense papillary
after 6 months was associated with treatment fail- aggregates of tumor cells. In contrast to low-grade
ure [99]. Presence of higher-grade carcinoma in endometrioid carcinoma with a villoglandular/
subsequent biopsy also characterizes progression papillary pattern, the papillary fronds of serous
and should always be reported [97, 98]. carcinoma usually are coarse, with thick fibrotic
Fig. 10.26 Serous carcinoma. Arborizing papillae with multiple papillary tufts are a frequent characteristic of this
tumor. Many of the papillae appear as free-floating clusters of cells
cores lined by highly atypical epithelial cells diagnosis of serous carcinoma itself establishes
(Fig. 10.27a and b). Because of the papillary the presence of a highly malignant carcinoma.
growth, serous carcinoma can appear deceptively Several other histologic subtypes of endome-
well differentiated in the endometrium, although trial carcinoma, including low-grade
these are high-grade, aggressive neoplasms. endometrioid tumors, also may show papillary
Serous carcinoma was originally considered a growth with a so-called villoglandular pattern [6,
predominantly papillary neoplasm, but studies 45, 57, 59, 107, 108]. The papillary endometrioid
now show that it is morphologically diverse [80]. tumors have low- to moderate-grade nuclei (see
In some cases the papillae are long and slender earlier, Grading) and often grow in long, slender
instead of short and coarse. The tumor may even branching papillary fronds (Fig. 10.6). The lining
be primarily composed of glands with lumens cells are columnar and do not form papillary
(Figs. 10.28, 10.29, 10.30, and 10.31). In the tufts. In contrast, the papillae in serous carcinoma
myoinvasive component of the tumor, the glands tend to be small and coarse, although it is the
have a gaping appearance (Fig. 10.32). Psammoma high nuclear grade that is most useful for identi-
bodies are present in up to one third of cases [57]. fying this neoplasm. An occasional tumor may
The nuclei of serous carcinomas are hyper- show intermediate features between low-grade
chromatic and pleomorphic with macronucleoli villoglandular and serous carcinoma with papil-
and many mitoses (Fig. 10.33). Abnormal mitotic lary architecture showing somewhat coarser
figures are frequent. Some nuclei are lobulated papillae and grade 2 nuclei. If a tumor with
with deep clefts, and, not infrequently, the chro- increased architectural and cytologic atypia does
matin appears smudged. The cells of serous car- not fulfill all the criteria of serous carcinoma,
cinoma tend to be rounded and often have however, it should be classified as a grade 2 endo-
abundant granular, eosinophilic cytoplasm. Areas metrioid carcinoma, with immunohistochemical
displaying the features of serous carcinoma but studies performed to support the diagnosis (see
containing clear cells, that is, clear cell carci- below). To avoid confusion between these papil-
noma, are seen in up to one third of cases (see lary tumors with different cytologic features and
later, Clear Cell Carcinoma). prognosis, we recommend not using the term
Because these tumors always contain high- “papillary” as a diagnostic term for any type of
grade nuclei, nuclear grading is not relevant. The endometrial carcinoma.
a b
Fig. 10.27 Serous carcinoma. (a) A complex papillary growth of a neoplasm composed of fibrovascular cores
neoplastic proliferation comprised of detached papillary covered with high-grade malignant cells having promi-
tufts and high-grade malignant cells is depicted in this nent nucleoli and many mitotic figures
image. (b) Medium magnification shows the papillary
Fig. 10.28 Serous carcinoma, glandular pattern. At plasia. Attention to the nuclear detail which is high grade is
medium magnification, serous carcinoma with a predomi- imperative to make the correct diagnosis. With serous car-
nant glandular pattern may be mistaken for atypical hyper- cinoma, the background endometrium usually is atrophic
Immunohistochemical analysis can assist in absence or null phenotype (complete absence of

the recognition of serous carcinoma and distinc- p53 expression by tumor cells while internal
tion from other histologic types when morpho- control shows moderate to strong but variable
logic features are equivocal. Serous carcinomas staining), (3) p53 cytoplasmic stain only (cyto-
usually show diffuse, intense nuclear overex- plasmic p53 tumor cell expression with internal
pression for p53 protein in at least 75% of the control showing nuclear wild-type pattern), and
neoplastic cells (Fig. 10.31 and Table 10.4) (4) wild-type (variable proportion of tumor cells
which correlates with TP53 gene mutations staining with variable intensity) [113]. To con-
found in more than 90% of such carcinomas firm that a complete absence of p53 staining is
[109–112]. Overall, four distinct p53 immuno- indicative of a TP53 mutation, staining for p16 is
histochemical staining patterns have been helpful since it is typically intense and diffuse
described, all of which (except for the last) are indicating a TP53 mutation is present. In the
surrogates of an underlying TP53 mutation: (1) absence of a TP53 mutation, p16 staining is usu-
p53 overexpression (strong/diffuse nuclear stain- ally patchy. While most of the p53 immunohisto-
ing of ≥75% of tumor cells), (2) complete p53 chemistry studies were largely based on
Fig. 10.29 Serous carcinoma, glandular pattern. At suggest serous carcinoma and should prompt the use of a
higher magnification, the high-grade nuclear features, panel of immunostains if the diagnosis is uncertain
presence of numerous, and abnormal mitoses strongly
observations made on tubo-ovarian high-grade A high proliferation index reflected in the dif-
serous carcinomas, similar patterns have been fuse reactivity for Ki-67 protein in the nuclei is
noted in serous carcinomas of endometrial ori- also typical of endometrial serous carcinoma
gin. Approximately 90% of high-grade tubo- [115]. Generally, serous carcinoma lacks estro-
ovarian serous carcinomas display either nuclear gen and progesterone receptors [115–117]. Low-
overexpression or, less frequently, a null pheno- grade endometrioid carcinomas, in contrast, have
type, but recognition of the other less common relatively low proliferation indices, do not have
phenotypes remains critical for accurate classifi- aberrant p53 (wild-type pattern), and show strong
cation [113]. Importantly, however, the presence reactivity for estrogen and progesterone recep-
of a p53 alteration (or lack of thereof) does not tors (Table 10.4) [116]. WT-1 immunostaining
necessarily categorize or exclude a tumor as can assist in differentiating ovarian from endo-
serous. Combined critical morphologic evalua- metrial serous carcinoma. Typically, the former
tion and identification of additional molecular demonstrates strong and diffuse WT-1 expres-
alterations will play a significant role in proper sion, while endometrial serous carcinoma usually
future classifications [114]. is negative. Some studies, however, showed that a
Fig. 10.30 Serous carcinoma, glandular pattern. A strong and diffuse positive p16 immunolabeling of high-grade
malignant glandular cells is characteristic of serous carcinoma
subset of endometrial serous carcinomas may terectomy [127]. Serous carcinoma with invasion
express WT-1, underscoring the need for critical of less than 1.0 cm has been termed “minimal
interpretation of all histologic and immunohisto- uterine serous carcinoma (MUSC) [128]. Even
chemical features [118, 119]. with no or minimal myometrial invasion, serous
Serous carcinoma is a high-grade variant of carcinoma can disseminate widely [33, 57, 103,
endometrial carcinoma that usually occurs in 107, 108, 123, 128–135].
older women, with a median age in the seventh If after hysterectomy and staging serous car-
decade [57, 105, 107, 108, 120, 121]. Rare cases cinoma is not present in extrauterine sites, the
are seen in younger women, however [122–124]. prognosis is very good [128]. Patients with
These tumors often invade the myometrium serous carcinoma often either have peritoneal
deeply and permeate lymph-vascular spaces. Not spread at the time of hysterectomy or relapse
uncommonly, serous carcinoma is minimally with peritoneal carcinomatosis, so in this regard
invasive and may even be confined to an endome- they behave like their ovarian counterparts.
trial polyp [125, 126]. Serous carcinoma also can Occasional cases also appear to be multifocal,
be limited to curetting specimens only with no with associated ovarian serous carcinoma at the
residual serous carcinoma in the subsequent hys- time of diagnosis. Because of their aggressive
Clear Cell Carcinoma 295
Fig. 10.31 Serous carcinoma, glandular pattern. An aberrant expression (strong and diffuse nuclear staining) for p53
is the most common p53 staining pattern in serous carcinomas
behavior, even when superficial or confined to a solid (Figs. 10.34, 10.35, and 10.36).
polyp, mixed tumors with both endometrioid Occasionally, clear cell carcinomas are more
and serous patterns in which at least 25% of the eosinophilic (pink) than clear and that may pose
tumor contains a serous component should be a diagnostic challenge (Fig. 10.37). The stroma
classified as serous carcinoma. Those tumors and papillary cores of clear cell carcinoma typi-
with less than 25% but more than 5% would be cally have hyalinized stroma (Fig. 10.38). In
mixed carcinomas with a serous component. some cases, the clear cytoplasm is inconspicu-
ous, and the nuclei bulge into the gland lumens,
forming so-called hobnail cells (Fig. 10.39).
Clear Cell Carcinoma These hobnail cells with hyalinized cores must be
differentiated from similar metaplastic changes
Most of the cells in clear cell carcinoma have lacking overt cytologic atypia that occur in a
clear, vacuolated cytoplasm because of the pres- background of abnormal uterine bleeding
ence of glycogen [6, 59, 62, 63, 136–138]. This (Fig. 10.40). Clear cell carcinoma also occurs in
tumor can have a variety of growth patterns, the ovary, cervix, and vagina. In the ovaries, clear
including tubulocystic, glandular, papillary, and cell carcinoma has patterns similar to those found
Fig. 10.32 Serous carcinoma. This tumor shows irregularly shaped, infiltrative glands. There is marked nuclear pleo-
morphism, and the high nuclear grade is a constant feature of serous carcinoma
in the endometrium, but in the vagina and cervix carcinoma, clear cell carcinoma is another
of women exposed in utero to diethylstilbestrol aggressive variant of endometrial cancer, thus
(DES), the tubulocystic pattern predominates. high grade by definition [62, 63, 139]. Like
Many examples of clear cell carcinoma appear serous carcinoma, it tends to occur in older
to be closely related to serous carcinoma, and, in patients and has a high relapse rate [138–140]. A
many instances, they may be admixed with serous recent study has shown that up to 14% of endo-
carcinoma [57]. The similarities between clear metrial serous carcinomas show some areas with
cell and serous carcinomas are reflected in the clear cell change; however, these cases are phe-
high nuclear grade of many clear cell tumors notypically similar to serous carcinoma and dis-
(Fig. 10.41). Especially in the papillary and solid similar to bona fide clear cell carcinomas [141].
patterns, the nuclei generally are pleomorphic Besides serous carcinoma, the differential
with marked atypia, although the hyalinized diagnosis of clear cell carcinoma includes
stroma of clear cell carcinoma often is a distin- endometrioid carcinoma with secretory changes
guishing feature. Macronucleoli and abnormal (Fig. 10.15) and endometrioid carcinoma with
mitotic figures usually are present; however, the prominent clear cell features. These typically are
latter seem to be less abundant than in serous car- low-grade neoplasms that contrast with clear cell
cinoma (personal observation). Occasionally, carcinoma which is, by definition, high grade.
these tumors, like serous carcinoma, contain Immunohistochemically, clear cell carcino-
psammoma bodies. Clinically similar to serous mas are ER and PR negative (Table 10.4). The
Fig. 10.33 Serous carcinoma. At high magnification, the cells show nuclear grade 3. The nuclei are large, with macro-
nucleoli, a high nuclear/cytoplasmic ratio, and numerous mitoses with abnormal forms
opposite pattern is encountered in most low- trioid carcinomas, regardless of the presence of
grade endometrioid carcinomas, including the clear cell features [144–150].
ones with clear cells. P16 immunostain may Practically, a limited panel of immunomarkers
show diffuse immunolabeling in half of clear cell to include p53, p16, Napsin A, ER, and PR should
carcinomas, while is characteristically weak and be differentially expressed in most serous, clear
patchy in low-grade endometrioid tumors [80, cell, and endometrioid carcinomas of endome-
142]. Moreover, clear cell carcinoma demon- trial origin (Table 10.4) [69].
strates an aberrant, diffuse strong reactivity for
p53 protein in approximately one third of cases;
however, the presence or absence of such immu- Rare Histologic Subtypes
noreactivity does not appear to be correlated with
any specific morphologic pattern [115, 143]. Primary squamous carcinoma of the endome-
Napsin A, p504s, and hepatocyte nuclear trium does occur but is rare [151–154]. To diag-
factor-1beta (HNF-1β) are additional markers nose this entity, it is necessary to exclude a
that can help identify clear cell carcinoma since primary cervical carcinoma (see later,
they are often expressed in this neoplasm. Endometrial Versus Endocervical Carcinoma).
However, these antibodies lack specificity and Primary verrucous squamous cell carcinoma also
should be judiciously used in a panel. For has been reported to arise in the endometrium
instance, HNF-1β may label a subset of endome- [155]. There have been several reports of
Fig. 10.34 Clear cell carcinoma, papillary pattern. The clear cytoplasm is less conspicuous, and the nuclei bulge in
a hobnail fashion
transitional cell carcinomas of the endometrium between the tumor in the endometrium and the
that are usually admixed with other patterns of cervical epithelium, and (3) absence of squamous
endometrial carcinoma; however, just like their cell carcinoma in the cervix [159].
ovarian counterparts, they usually represented The rarity of primary squamous cell carcino-
solid variants of serous carcinoma [156, 157]. mas of the endometrium should prompt investi-
Rarely, high-grade squamous intraepithelial gation of secondary involvement. A p16
lesions of cervical origin (HSIL/CIN 3) migrate immunostain, as well as in situ hybridization for
upward, continuing a pattern of intraepithelial human papilloma virus (HPV), may serve as
growth into the endometrial cavity, and may helpful ancillary tools in assigning a cervical ori-
mimic an endometrial primary. By definition, gin since these markers are positive in most cer-
these tumors are not invasive of underlying vical primaries but are negative in primary
stroma. They are strongly reactive for p16 which endometrial squamous carcinomas.
helps to verify their origin in the cervix [158]. Undifferentiated carcinomas are defined by
Primary squamous cell carcinoma of the endo- the current WHO as malignant epithelial neo-
metrium is an extremely rare entity. Nearly a cen- plasms with lack of glandular or squamous dif-
tury ago, Fluhmann reported his criteria for this ferentiation, while dedifferentiated carcinomas
diagnosis which required (1) the absence of a are composed of an undifferentiated carcinoma
coexisting glandular component of the carcinoma coexisting with a more well-differentiated com-
in the endometrium, (2) lack of connection ponent (FIGO 1 or 2 endometrioid carcinoma)
Fig. 10.35 Clear cell carcinoma, glandular pattern. arrangement. This morphology may overlap with endo-
Malignant cells with high-grade nuclei, prominent nucle- metrioid carcinomas with clear cell features
oli, and abundant clear cytoplasm grow in a glandular
(Fig. 10.42) [160, 161]. Undifferentiated carcino- have a moderate amount of cytoplasm and large
mas account for up to 9% of primary endometrial vesicular nuclei with prominent nucleoli.
carcinomas [162]. They are typically solid neo- Notably, undifferentiated carcinomas have
plasms composed of small to intermediate size, occurred in patients with Lynch syndrome, a
dyshesive cells without gland formation [31]. finding that should trigger appropriate testing
Some tumors have features resembling small (see Lynch syndrome discussion) [87, 164].
cell carcinoma of the lung, whereas others are Immunohistochemically, undifferentiated carci-
composed of medium-sized to large cells that nomas typically show at least focal cytokeratins
range from polygonal to spindle in shape. The (AE1/AE3) or epithelial membrane antigen (EMA)
cells are arranged in solid sheets with a monoto- immunoexpression. They are also usually
nous appearance. Mitotic figures and extensive focally positive for cytokeratin 18. They are largely
areas of necrosis are conspicuous. Focal pleo- negative for ER and PR. P16 may be diffusely posi-
morphism and rhabdoid cells may be seen [162]. tive in some cases. They usually show an aber-
With the small cell pattern, the neoplastic cells rant p53 (strong/diffuse) expression (Table 10.4).
have scant cytoplasm and hyperchromatic nuclei These neoplasms also may show neuroendocrine
with indistinct nucleoli [163]. Some tumors with differentiation [165–169]. E-cadherin has been
a small cell component are admixed with typical reported as greatly reduced or entirely absent in
adenocarcinoma [163]. The large-cell variant is 30–70% of the cases, contrasting with preserved
composed of sheets of large epithelial cells that membranous expression seen in low-grade endo-
Fig. 10.36 Clear cell carcinoma, solid pattern. The cells marked enlargement, irregular outlines, smudged chroma-
are organized in solid sheets and demonstrate clear, vacu- tin, and prominent nucleoli. Rare apoptotic bodies are
olated cytoplasm, high-grade nuclei (grade 3) with present
metrioid carcinomas [170]. It is important to recog- histologic types include glassy cell carcinoma
nize undifferentiated carcinoma of the endometrium [178, 179], oxyphilic variant of endometrioid car-
as this is a highly aggressive form of high-grade cinoma [180], signet-ring cell carcinoma
carcinoma frequently mistaken for a FIGO grade 3 (Fig. 10.43) [181], lymphoepithelioma-like carci-
endometrioid carcinoma or a sarcoma. noma [182], and giant cell carcinoma [183]. Most
Undifferentiated carcinomas appear to have a of these are high-grade malignancies that occur in
worse prognosis than FIGO 3 endometrioid carci- older patients. Other rare tumors include primary
nomas [162]. yolk sac tumor [184, 185], α-fetoprotein-secreting
Other rare and unusual patterns of differentia- hepatoid adenocarcinoma associated with endo-
tion, such as osteoclastic-type giant cells or tro- metrioid carcinoma [186], and endometrioid carci-
phoblast, may be found in primary endometrial noma associated with Ewing sarcoma/peripheral
carcinomas which are usually high-grade [58, 171, primitive neuroectodermal tumor [187–191].
172]. Choriocarcinoma of the endometrium rarely Wilms tumor also may rarely occur as a primary
is seen in postmenopausal women, usually result- uterine neoplasm [192, 193]. The rare primary
ing from dedifferentiation of a poorly differenti- endometrial yolk sac tumors may morphologically
ated carcinoma [171, 173–177]. Additional rare overlap with a colonic-derived metastasis [194].
Carcinosarcoma 301
Fig. 10.37 Clear cell carcinoma. Occasionally, clear cell carcinomas appear more “pink” than “clear” because the
cytoplasm is not as glycogen-rich. Other areas of tumor, however, displayed more characteristic clear cell features
Carcinosarcoma polypoid mass that prolapses through the cervical

os. According to some studies, there is an
Carcinosarcoma or malignant mixed Müllerian increased frequency of carcinosarcomas in
tumor (MMMT) is a highly malignant biphasic women previously exposed to tamoxifen, pro-
tumor consisting of both epithelial (carcinoma) longed unopposed estrogen use, or pelvic irradia-
and mesenchymal (sarcoma) components [195– tion [202–205].
199]. The MMMT designation is no longer used, Previously it was thought that these tumors
and the preferred term is “carcinosarcoma.” This behaved like sarcomas and accordingly were clas-
is an uncommon tumor of epithelial derivation sified with the pure uterine sarcomas. Recent clini-
accounting for less than 5% of all uterine malig- copathologic, immunohistochemical, and
nancies and affecting predominantly postmeno- molecular genetic studies have demonstrated the
pausal patients (median age at diagnosis is about two elements are clonal, with TP53 being the most
65 years) [195, 200, 201]. common molecular alteration [206, 207], and these
As with endometrial carcinoma, the patient neoplasms are more closely related to and derived
with carcinosarcoma typically presents with from their epithelial component representing meta-
postmenopausal bleeding. The uterus is almost plastic carcinoma. Accordingly, they are currently
always enlarged, and the tumor may present as a staged like endometrial carcinomas [208].
Fig. 10.38 Clear cell carcinoma typically shows eosinophilic hyalinized cores surrounded by a relatively monotonous
population of malignant clear cells
Some molecular studies suggest carcinosarco- lacking clear features of either carcinoma or sar-
mas demonstrate aberrant expression of fascin, coma (Fig. 10.46).
an actin-bundling protein which is a component The carcinomatous component of carcinosar-
of the epithelial-mesenchymal transition, in both coma typically has features of high-grade carci-
carcinoma and sarcomatous components of the noma. It is frequently composed of serous or
tumor. Such expression has been associated with clear cell carcinoma, although endometrioid pat-
a worse outcome. Likewise, upregulation of terns, including carcinoma with squamous
epithelial-
mesenchymal transition genes has differentiation, may be found. In fact, the finding
been shown in these neoplasms [209–211]. of a high-grade carcinoma with morphologic het-
In biopsy and curettage material, the distinc- erogeneity in biopsy or curettage material should
tion between a carcinosarcoma and a high-grade raise the possibility of an under-sampled carcino-
carcinoma may be difficult. Usually the compo- sarcoma (Fig. 10.47). In curettage specimens,
nents are intimately admixed, with nests and carcinosarcomas are often polypoid and can pro-
masses of carcinoma separated by a malignant duce abundant tissue which can be a clue to the
spindle cell stroma (Figs. 10.44 and 10.45). The diagnosis, yet they may be small and confined to
carcinoma is typically juxtaposed to the sarcoma a portion of a polyp [197].
with the two cell types appearing as separate foci. In our experience, serous carcinoma is one of
However, frequently, carcinoma appears to blend the most frequent epithelial elements. Other epi-
into the sarcoma, with the cells at the interface thelial components such as clear cell, mucinous,
Carcinosarcoma 303
Fig. 10.39 Clear cell carcinoma with detached hyalinized papillary cores with hobnail cells projecting from the
surface
or undifferentiated carcinoma are less common. heterologous components, including liposarcoma

The epithelial component rarely may have a pre- or even glia, may rarely be found [201, 215–218].
dominant pattern of squamous carcinoma. Some carcinosarcomas have a sarcomatous
The sarcomatous component can display a vari- component that contains large cells with abun-
ety of patterns that have been referred to as either dant, eosinophilic cytoplasm. These large cells
homologous or heterologous [195, 199, 201, 212– superficially resemble rhabdomyoblasts, but they
214]. The sarcoma is considered homologous lack sarcomeric filaments and are not immunore-
when it shows differentiation toward mesenchy- active for desmin, myogenin, or myoglobin. Other
mal cells normally present in the uterus. Often the carcinosarcomas may have dense eosinophilic
homologous sarcoma resembles fibrosarcoma or hyaline material between the spindle cells that is
malignant fibrous histiocytoma composed of inter- suggestive, but not diagnostic, of osteosarcoma,
lacing primitive spindle cells, but at times it may lacking unequivocal osteoid and osteoblastic-like
have features of endometrial stromal sarcoma or stromal cells. Neither large, eosinophilic cells nor
leiomyosarcoma. The sarcoma is considered het- hyaline stroma should be construed as evidence of
erologous when it is differentiated into cell types heterologous differentiation unless the pattern
not normally found in the uterus (Fig. 10.48a, b). shows clearly diagnostic features.
The usual heterologous components are rhabdo- Rarely a carcinosarcoma may show well-
myosarcoma (most common heterologous ele- differentiated carcinoma (less than 2% of cases),
ment), chondrosarcoma, or osteosarcoma. Other [219] a relatively low-grade sarcoma, or a combi-
Fig. 10.40 Hobnail metaplastic change mimicking clear encountered in clear cell carcinomas. A limited immunohis-
cell carcinoma. This change typically is seen in a background tochemical stain panel including p53 and ki-67 can be used
of abnormal uterine bleeding and lacks the cytologic atypia to differentiate carcinoma from metaplastic changes
nation of both instead of the high-grade patterns mesenchymal components, while PAX-8 stromal
usually seen in the carcinoma and sarcoma com- (sarcoma) positivity is only seen in the presence
ponents (Fig. 10.49). These tumors nonetheless of epithelial (carcinoma) positivity, according to
show features of malignancy in both the glands a large study [198, 219–225].
and the stroma. The glands show, at a minimum, Biologically, these tumors recur and metasta-
the features of a well-differentiated carcinoma. size like carcinomas rather than sarcomas, and
Low-grade sarcoma lacks high cellularity and the metastases are almost always carcinoma,
nuclear pleomorphism but shows a cellular pro- even if the bulk of the primary tumor is sarcoma-
liferation of enlarged mesenchymal cells and a tous [197, 198]. When serous carcinoma is a
high mitotic rate of four or more mitoses per ten component of the endometrial primary, this
high-power fields (HPFs). Despite their low- should be noted since the tumor may metastasize
grade appearance these latter tumors should be in a pattern of serous carcinoma. The following
diagnosed as carcinosarcoma. features have been associated with a shorter
Immunohistochemically, both the carcinoma- disease-free interval: a sarcomatous component
tous and sarcomatous elements are reactive for on recurrence, sarcoma dominance of the original
cytokeratins and epithelial membrane antigen tumor, tumor size 5 cm or more, outer half myo-
(EMA), although immunostaining is less diffuse invasion, lymph node metastasis, lymphovascu-
in the sarcomatous foci. P16 and p53 are fre- lar invasion, advanced stage (III-IV), residual
quently diffusely expressed in both epithelial and disease at surgery, and positive cytology [201].
Carcinosarcoma 305
Fig. 10.41 Overlapping morphology between serous and growth patterns of each neoplasm usually are present. A
clear cell carcinoma. Both neoplasms display high-grade limited panel of immunostains may be necessary to distin-
nuclear features; however, mitotic indices tend to be lower guish the two neoplasms
in clear cell carcinoma. Areas of more characteristic
Serous histology and rhabdomyoblastic differen- ologous elements is a significant negative prog-
tiation are also associated with a worse 3-year nostic factor [227].
overall survival [201].
Although establishing the diagnosis of carci-
nosarcoma in a limited biopsy material may be Differential Diagnosis
challenging, it must be attempted as these neo- of Carcinosarcoma
plasms are very aggressive and require intraop-
erative staging. Uterine carcinosarcoma is even For most cases of carcinosarcoma, the differen-
more aggressive than other high-grade endome- tial diagnosis centers mainly on carcinoma. The
trioid or serous carcinomas, with more than 35% mesenchymal component may be indistinct or
of affected patients presenting with extrauterine difficult to differentiate from carcinoma with a
disease at diagnosis and an overall median sur- spindle cell component, dedifferentiated carci-
vival of less than 2 years [225, 226]. noma, or high-grade adenocarcinoma with squa-
Finally, the percentage and composition of mous differentiation. In the latter case of
each epithelial (endometrioid, serous, clear cell) adenocarcinoma with squamous differentiation,
and sarcomatous (heterologous or homologous) the squamous component may have a spindle
elements appear to carry prognostic significance. appearance with swirling aggregates of elongate,
In stage I carcinosarcoma, the presence of heter- nonkeratinizing squamous cells. The squamous
Fig. 10.42 Undifferentiated carcinoma. This highly stains, particularly EMA or CK18, which may only show
malignant neoplasm is characterized by a complete lack focal expression, are usually helpful to prove this as an
of glandular or squamous differentiation. Cytokeratin epithelial malignancy
component is within the center of the glands, in cases, the malignant spindle cells have an indis-
contrast to the sarcoma of carcinosarcoma, which tinct interface with the epithelial component, and
surrounds the glands. At times the squamous ele- there is a zone of cells that are indeterminate in
ment may be extensive and efface glandular their differentiation into epithelium or mesen-
structures, thereby simulating a sarcoma. chyme. We have found that this feature, the merg-
Immunohistochemical stains for cytokeratin ing of spindle cells into the clearly carcinomatous
should help resolve the diagnosis, as discussed component, is especially useful for recognizing
below. carcinosarcoma in cases where the sarcoma is
In some cases, the sarcoma is very limited in largely overshadowed by the carcinoma.
the curettage specimen and is eclipsed by a domi- It is important to scrutinize the stroma between
nant carcinomatous element. Careful study of all glands and nests of carcinomatous epithelium to
the material may be necessary to correctly diag- determine whether the mesenchyme has malig-
nose a carcinosarcoma with a subtle sarcomatous nant features, such as increased cellularity, pleo-
component. The sarcomatous areas may blend morphic nuclei, and a high mitotic rate.
imperceptibly into the carcinomatous compo- Nonneoplastic mesenchyme in adenocarcinoma
nent, making the sarcoma difficult to distinguish either resembles normal endometrial stroma or
from poorly differentiated carcinoma. In these shows desmoplasia, with cells developing
Carcinosarcoma 307
Fig. 10.43 Endometrial signet-ring cell carcinoma. These tumors typically blend into a more typical endometrioid
pattern. This can mimic metastatic involvement by a gastrointestinal malignancy
fibroblast-like characteristics. Benign mesenchy- tins, and epithelial membrane antigen (EMA),
mal cells have a low mitotic rate; stromal cellu- p16 and p53 can be useful for demonstrating the
larity may be mildly increased, but the nuclei biphasic pattern of the carcinosarcoma [198, 199,
remain oval and uniform with indistinct chroma- 201, 220–222, 224, 225, 228]. The epithelial com-
tin. Sarcomatous tissue, in contrast, shows ponent shows intense and generalized reactivity
increased cellularity; the nuclei are irregular and for keratin and EMA, and although the epithelial
closely packed; and mitotic figures are numerous, markers are also present in the sarcomatous com-
usually in the range of ≥4 per 10 HPFs. Atypical ponent, this staining is less intense. Extensive
mitoses often are present. staining of p16 in the epithelial component (74%)
In general, immunohistochemistry is of limited and mesenchymal component (71%) has been
value in the diagnosis of carcinosarcoma; both the demonstrated [199]. Immunostaining for p53 in
malignant epithelial and mesenchymal compo- the epithelial and stromal components ranges
nents are usually diagnosed on morphologic from 48–77% to 44–83% of cases, respectively
examination. In questionable cases where the dis- [199, 228]. These immunostains are especially
tinction between carcinosarcoma and poorly dif- useful, as they highlight the mixed population of
ferentiated adenocarcinoma is not clear by routine carcinoma and sarcoma. Mesenchymal markers
histology, immunohistochemical stains for kera- also can help to identify a carcinosarcoma by
Fig. 10.44 Carcinosarcoma. Usually the carcinomatous and sarcomatous components are intimately admixed (bipha-
sic pattern), with nests and masses of carcinoma separated by a malignant spindle cell stroma
highlighting the sarcomatous areas. Desmin, often stain with p16, as they are HPV related
actin, and muscle-specific actin immunostains are [223].
useful for demonstrating muscle differentiation in One other uncommon entity important in the
the spindle cell population. Desmin and espe- differential diagnosis of carcinosarcoma is the
cially myogenin also assist in demonstrating unusual type of endometrioid carcinoma known as
rhabdomyoblasts [199]. “corded and hyalinizing” variant, which has been
Primary carcinosarcomas of the uterine cer- mistaken for a “low-grade” carcinosarcoma
vix are uncommon but may rarely be encoun- because of the frequent association of stromal
tered in an endometrial biopsy sampling. The spindle cells and even osteoid to the more usual
epithelial component of cervical carcinosarco- appearing endometrioid carcinoma. While the
mas is more likely to be of squamous cell carci- “corded and hyalinizing” endometrioid carcinoma
noma, adenoid cystic carcinoma, or adenoid is typically a low-grade carcinoma characterized
basal carcinoma histologic types, but endome- by sex cord-like formations and hyalinization
trioid carcinoma can also occur [223]. The mes- where the various components merge onto each
enchymal component resembles a fibrosarcoma, other (Fig. 10.22), carcinosarcomas are biphasic
or endometrial stromal sarcoma, and heterolo- high-grade malignancies with juxtaposed carcino-
gous elements may be found [229]. The tumors matous and sarcomatous elements [72].
Carcinosarcoma 309
Fig. 10.45 Carcinosarcoma. Endometrioid carcinoma overlies a malignant appearing (sarcomatous) stroma. Notice
the markedly atypical cytologic features of stromal cells and their abundant mitotic activity
Staging tumor is limited to or invades less than one half

of the myometrium and IB when tumor invades
In 1988 FIGO revised its staging system from more than one half of the myometrium. In the
one that was strictly based on clinical evaluation 1988 staging system, endocervical glandular
to one based on combined surgical and histopath- involvement was stage IIA and cervical stromal
ologic findings [29]. Currently, staging of endo- involvement IIB. Endocervical glandular
metrial cancer (carcinomas and carcinosarcomas) involvement was removed from the 2009 stag-
employs a variety of histologic risk factors, such ing system, and now there is a single category of
as grade, depth of myometrial invasion, involve- stage II disease defined by cervical stromal
ment of the cervical stroma, and lymph node sta- involvement.
tus, including the documentation of nodal Although staging is based on surgical-pathologic
micrometastases. See Table 10.5 for details [230]. analysis of the hysterectomy specimen, accurate
Endometrial carcinoma confined to the cor- histologic evaluation of biopsies is important, as
pus is stage I, and about three quarters of all pri- grade and histologic findings influence the plan-
mary endometrial carcinomas are stage I. Based ning for surgery and whether surgical staging and
on the presence and amount of myometrial inva- lymph node sampling are necessary. Furthermore,
sion, carcinomas are subdivided into IA when with a high-grade neoplasm, a gynecologist should
Fig. 10.46 Carcinosarcoma. The carcinoma component (bottom) blends into the sarcoma (homologous elements).
Scattered, ill-defined solid nests are seen in between these two elements, an area which is difficult to classify
request the assistance of a gynecologic oncologist agement of invasive endocervical carcinoma is

ensuring appropriate staging and therapy. either primary radiation therapy or a radical hys-
terectomy and pelvic lymph node dissection, an
operation with potential for greater morbidity.
Differential Diagnosis The tumor type alone can help in the assign-
ment of tumor origin. For instance, serous carcino-
Endometrial Versus Endocervical mas of the cervix are rare and when present in
Carcinoma endocervical curettage more frequently denote
extension from a serous primary of the endome-
In some cases, it is difficult to determine whether trium. Likewise, the endometrioid variant of endo-
carcinoma in an endometrial biopsy primarily cervical carcinoma represents only a very small
involves the endometrium, the endocervix, or subset of all cervical cancers. Thus finding of an
both sites. Accurately assigning a primary site is endometrioid tumor in the cervix likely represents
important for directing further therapy and extension from an endometrial tumor, especially
establishing prognosis. Endometrial carcinoma
in peri- and postmenopausal women [231].
confined to the corpus is managed differently There are a number of features that typically
from an endocervical carcinoma. The usual ther- favor an endometrial origin. These include clini-
apy for endometrial carcinoma is total hysterec- cal or radiographic evidence that the bulk of the
tomy with bilateral salpingo-oophorectomy and lesion is centered in the uterine corpus or lower
lymph node dissection. In contrast, surgical man- uterine segment, background endometrium with
Fig. 10.47 Carcinosarcoma. This image highlights the heterogeneous appearance of the carcinomatous component
with mixed endometrioid, clear cell, and serous features. No sarcoma is seen in this image
atypical hyperplasia, squamous differentiation in If the morphologic features do not clearly estab-
the form of squamous morules, scant intervening lish the primary site, clinical information often
stroma, and the presence of foamy histiocytes. In resolves the problem. The presence or absence of
contrast, cervical adenocarcinomas more fre- gross tumor involving the cervix is important infor-
quently show a desmoplastic intervening stroma, mation. If tumor clinically involves the cervix, then
and the malignant glands display abundant mitotic the issue of primary site is largely irrelevant. The
figures or apoptotic bodies. Squamous differentia- tumor can be managed as a cervical neoplasm [23].
tion, when present, is characteristically in the If the tumor is occult, information on age and
form of an intimately associated adenosquamous menopausal status can be helpful. Endometrial car-
lesion, in contrast to the squamous morules seen cinoma usually occurs in older patients with an
in endometrioid endometrial carcinoma. In ade- average age of 55–60 years. Endometrioid carci-
nosquamous endocervical carcinoma, the squa- noma is unusual before menopause unless the
mous element usually predominates, and the patient has a predisposing condition, such as poly-
glandular differentiation is subtle, while the oppo- cystic ovaries, morbid obesity, or hereditary syn-
site pattern is seen in endometrial carcinomas drome. Endocervical adenocarcinoma, in contrast,
(Fig. 10.50). A background with in situ adenocar- has a much wider age range, with many tumors
cinoma or squamous intraepithelial lesion also occurring in premenopausal women. The average
supports an endocervical origin [142]. age is between 45 and 55 years.
a b
Fig. 10.48 Carcinosarcoma, heterologous elements. (a) philic cytoplasm. Although cross striations are not seen,
Multiple nodules of cartilaginous differentiation are seen immunohistochemical stains for desmin and myogenin
in this field. (b) A focus of rhabdomyosarcoma is present demonstrated cytoplasmic muscle filaments
showing multiple rhabdomyoblasts with abundant eosino-
Fig. 10.49 Carcinosarcoma. In this case the malignant glands appear well differentiated, and the stroma lacks the
marked cellularity often seen in these neoplasms. Nonetheless, this should be diagnosed as carcinosarcoma
Table 10.5 FIGO 2009 and TNM Staging Systems (UICC and AJCC 8th Edition) for Endometrial Carcinomas
TNM FIGO UICC AJCC
Primary tumor
TX – Primary tumor cannot be assessed Primary tumor cannot be assessed
T0 – No evidence of primary tumor No evidence of primary tumor
T1 I Tumor confined to the corpus uteria Tumor confined to corpus uteri, including
endocervical glandular involvement
T1a IA Tumor limited to endometrium or invading less Tumor limited to endometrium or invading less than
than half of myometrium half the myometrium
T1b IB Tumor invades one half or more of myometrium Tumor invading one half or more of the
myometrium
T2 II Tumor invades cervical stroma but does not Tumor invading the stromal connective tissue of the
extend beyond the uterus cervix but not extending beyond the uterus. Does
not include endocervical glandular involvement
T3 III Local and/or regional spread as specified here Tumor involving serosa, adnexa, vagina, or
parametrium
T3a IIIA Tumor invades the serosa of the corpus uteri or Tumor involving the serosa and/or adnexa (direct
adnexa (direct extension or metastasis) extension or metastasis)
T3b IIIB Vaginal or parametrial involvement (direct Vaginal involvement (direct extension or metastasis)
extension or metastasis) or parametrial involvement
T4 IVA Tumor invades bladder/bowel mucosab Tumor invading the bladder mucosa and/or bowel
mucosa (bullous edema is not sufficient to classify a
tumor as T4)
Regional LN
NX – Regional lymph nodes cannot be assessed Regional lymph nodes cannot be assessed
N0 – No regional lymph node metastasis No regional lymph node metastasis
N0(i+) – – Isolated tumor cells in regional lymph node(s) not
>0.2 mm
N1, N2 IIIC Metastasis to pelvic or para-aortic lymph nodesb
N1 IIIC1 Metastasis to pelvic lymph nodes Regional lymph nodes metastasis to pelvic lymph
nodes
NImi IIIC1 – Regional lymph nodes metastasis (>0.2 mm but not
>2.0 mm in diameter) to pelvic lymph nodes
N1a IIIC1 – Regional lymph nodes metastasis (>2.0 mm in
diameter) to pelvic lymph nodes
N2 IIIC2 Metastasis to para-aortic lymph nodes with or Regional lymph nodes metastasis to para-aortic
without metastasis to pelvic lymph nodes lymph nodes, with or without positive pelvic lymph
nodes
N2mi IIIC2 – Regional lymph nodes metastasis (>0.2 mm but not
>2.0 mm in diameter) to para-aortic lymph nodes,
with or without positive pelvic lymph nodes
N2a IIIC2 – Regional lymph nodes metastasis (>2.0 mm in
diameter) to para-aortic lymph nodes, with or
without positive pelvic lymph nodes
Distant metastasis
M0 – No distant metastasis No distant metastasis
M1 IVB Distant metastasis (excluding metastasis to Distant metastasis (includes metastasis to inguinal
vagina, pelvic serosa, or adnexa, including lymph nodes intraperitoneal disease, lung, liver, or
metastasis to inguinal lymph nodes, intra- bone). (It excludes metastasis to pelvic or
abdominal lymph nodes other than para-aortic or para-aortic lymph nodes, vagina, uterine serosa, or
pelvic nodes) adnexa.)
From Ref. [230], with permission
AJCC American Joint Committee on Cancer, FIGO International Federation of Gynecology and Obstetrics, LN lymph
node, UICC Union for International Cancer Control
a
Endocervical glandular involvement only should be considered as stage I
b
The presence of bullous edema is not sufficient evidence to classify as T4
Fig. 10.50 Adenosquamous carcinoma of the endome- usually predominates in tumors of cervical origin, but
trium. These are rare tumors that can occur both in the immunohistochemistry is often required to assist in the
endometrium and endocervix. The squamous element diagnosis
Endometrial carcinomas with extensive muci- In addition to similarities in their morphology,

nous differentiation can be particularly difficult both endocervical carcinoma of the gastric type
to differentiate from an endocervical adenocarci- and mucinous endometrioid carcinomas lack
noma and occasionally from the rare endocervi- association with high-risk human papilloma virus
cal carcinoma, gastric-type (a.k.a. minimal (HR-HPV), limiting the use of immunohisto-
deviation adenocarcinoma or adenoma malig- chemical stains, such as p16, as well as in situ
num). This variant may even ascend into the hybridization and molecular studies.
endometrial cavity mimicking an endometrial In questionable cases, a limited immunohis-
primary (Fig. 10.51) [232, 233]. The presence of tochemical panel is very helpful for separating
abundant mucinous cytoplasm reminiscent of endocervical and endometrial tumors. For
endocervical epithelium is shared by these two instance, p16 tumor suppressor gene
entities. Endocervical carcinoma, gastric-type (CDKN2A) is overexpressed upon integration
may present as a rare sporadic neoplasm, but it is of the HPV virus in most cervical carcinomas.
frequently associated with Peutz-Jeghers syn- Consequently, strong and diffuse p16 immu-
drome [234]. Mucinous endometrial carcinoma, nostaining is usually seen in the nucleus and
although also very uncommon, usually shows cytoplasm of cervical primary neoplasms, both
areas of more characteristic endometrioid mor- in squamous carcinomas and adenocarcinomas
phology with scattered squamous differentiation. [235–246].
Fig. 10.51 Gastric-type endocervical adenocarcinoma. endometrium and malignant cells show abundant cyto-
These tumors are notorious for having inconspicuous plasmic mucin in contrast with the native endometrial
cytologic atypia. In this image, the tumor has spread to the glands
Because of these characteristics, immunohisto- ProExC, another surrogate marker for high-
chemistry for p16 has gained widespread use as a risk HPV, has been used to assist in the distinc-
surrogate marker for high-risk human papilloma tion between endocervical and endometrioid
virus infection in formalin-fixed, paraffin- carcinomas. It shows diffuse and strong expres-
embedded tissues. Usual-type endocervical ade- sion in most endocervical carcinomas, while
nocarcinomas are high-risk HPV-associated and being negative or only focally positive in endo-
characteristically show a strong and diffuse nuclear metrioid carcinomas [250]. Endometrial serous
and cytoplasmic staining for p16 (block- type carcinomas have been reported to be diffusely
staining pattern), while low-grade endometrioid immunoreactive for ProExC.
carcinomas demonstrate only a patchy, focal, and Other immunohistochemical stains also can
weak p16 staining which should be interpreted as be helpful in determining the primary site.
negative for an association with high-risk HPV These include estrogen and progesterone recep-
infection [247, 248]. High-grade endometrial car- tor proteins (ER and PR) as well as vimentin
cinomas, such as serous or clear cell carcinomas, and carcinoembryonic antigen (CEA) [251].
are often p16 positive, although unrelated to HPV Estrogen and progesterone hormone receptor
infection, which emphasizes the importance of proteins are typically strongly reactive in low-
morphology and a multi-marker panel approach in to intermediate-grade endometrioid-type carci-
these tumors [249]. nomas, while endocervical adenocarcinomas
Fig. 10.52 Metastatic carcinoma from upper gastrointes- appears reactive. This patient had a known advanced gas-
tinal origin. Haphazardly arranged glands with scattered tric carcinoma
goblet cells infiltrate the endometrial stroma which
lack hormone receptor staining (Table 10.4) risk HPV [247, 255]. In these cases, the p16
[251–254]. High-grade endometrioid carcino- staining pattern will be the same as in an endo-
mas, however, may lose hormone receptor metrioid carcinoma (entirely negative or weakly/
expression. Vimentin tends to be diffusely pres- patchy positive).
ent in endometrial but not endocervical adeno- Finally, negative HPV in situ hybridization
carcinoma primaries. Conversely, CEA strongly studies in cases highly suspicious for endocervi-
stains most endocervical adenocarcinomas but cal carcinoma should be interpreted cautiously as
shows only limited reactivity in endometrial the assay may have insufficient sensitivity to
primaries [69]. detect low viral copy numbers.
While the limited panel of immunohistochem- Differential cytokeratin staining for cytokera-
ical stains (p16, ER and PR) as well as HPV in tin 7 and cytokeratin 20 is not useful in distin-
situ hybridization studies is very useful in differ- guishing between endocervical and endometrial
entiating endometrial from endocervical adeno- primary sites [256]. Histochemical stains for
carcinomas [251], they are not helpful when the mucin, too, have little value in determining the
differential diagnosis is between the rare muci- primary site, since endometrial carcinoma often
nous endocervical adenocarcinomas, gastric shows at least focal cytoplasmic mucin and some
type, and primary mucinous endometrioid carci- endocervical adenocarcinomas show little cyto-
nomas, both of which lack association with high- plasmic mucin.
Metastatic Carcinoma history of a known extrauterine primary tumor or

a mass lesion in the bowel can be essential for
The most common extrauterine carcinomas that establishing the correct diagnosis. Colon adeno-
metastasize to or extend into the endometrium carcinoma may have a so-called garland-like
arise in the ovary, breast, or gastrointestinal tract arrangement of glands surrounding areas of
(Fig. 10.52), especially the colon [257–259]. “dirty” necrosis composed of cellular debris that
Metastases from other primary sites are rare, but helps in recognizing the tumor. The glands of
on occasion a tumor from the stomach, pancreas, metastatic colon carcinoma have a sharp luminal
or other visceral site metastasizes to the endome- border. Endometrial carcinoma, on the other hand,
trium. It is very unusual for tumors from these typically does not show much necrosis in glandu-
sites to present with abnormal vaginal bleeding lar lumens, and the cells have an ill-defined, fuzzy
and to be diagnosed first in an endometrial biopsy. luminal border in routine sections. Mucin stains
Separating metastatic ovarian carcinoma from have little utility in determining the primary site,
an endometrial primary tumor can be especially as endometrial carcinoma can have abundant
difficult, as virtually all patterns of primary endo- cytoplasmic mucin. An immunohistochemical
metrial carcinoma can occur in primary ovarian panel that includes CEA, CDX2, cytokeratins 7
epithelial carcinoma [260, 261]. Nevertheless, the (CK-7), 20 (CK-20), and ER and PR may be help-
question of ovarian metastasis versus an endome- ful in establishing whether the tumor is metastatic
trial primary site in a biopsy or curetting is infre- from the gastrointestinal tract or is the primary
quent. This differential diagnosis is most likely to tumor in the endometrium. In general, colonic pri-
occur when the endometrial biopsy shows a low- mary tumors are diffusely positive for CEA and
grade endometrioid carcinoma, and the final CDX2, while endometrial carcinomas are not.
resection demonstrates a tumor of similar mor- Also, endometrial carcinomas usually are positive
phology in the ovary. In such cases, it may be for CK-7, ER, and PR and negative for CK-20
impossible to exclude an ovarian primary site. [264], while metastatic colon adenocarcinoma
Recent studies show that the endometrial and exhibits the opposite profile [265].
ovarian tumors often are clonally related indicat- Metastatic breast carcinoma, although rare,
ing that one site was the primary, while the other, is well recognized for its ability to metastasize
metastatic. The findings show that establishment to the endometrium where it can be especially
of primary site required additional studies [262, difficult to diagnose. These tumors typically
263]. Furthermore, despite the presence of meta- infiltrate in solid sheets or small groups in a lin-
static disease from one site to the other, the out- ear pattern, often sparing glands and diffusely
come is generally very favorable suggesting that invading the stroma. The neoplastic cells may
this is an unusual type of metastatic process that is resemble stromal cells or inflammatory cells,
limited in its ability to spread widely [262, 263]. lacking the organoid arrangements seen in most
Metastatic lesions should be considered, how- carcinomas. In such cases, immunohistochemi-
ever, when the amount of tumor is relatively scant cal stains for cytokeratins and histochemical
and is admixed with more abundant fragments of stains for mucin are useful for demonstrating
benign, nonhyperplastic endometrium, and no the epithelial origin of the cells. Also, immunos-
transitions between the carcinoma and benign tis- tains for gross cystic disease fluid protein-15
sue exist. Also, clinical history of an adnexal (GCDFP-15), mammaglobin, or GATA-3 [266]
mass should alert the pathologist to the possibil- help identify the breast as a primary site.
ity of a primary ovarian neoplasm. Differential immunostaining for cytokeratins 7
Metastatic carcinoma from other sites, and 20 has no utility in the differential diagnosis
although rare, can be challenging. Occasionally, of endometrial versus breast carcinoma. Clinical
colon carcinoma may involve the endometrium history can be invaluable, since metastatic breast
and closely simulate a uterine primary tumor, almost always occurs in patients with a known
having an “endometrioid” pattern. In such cases, a history of a breast primary.
Carcinoma Mimics tion, may resemble carcinoma, especially clear

cell carcinoma (see Chap. 3). For a premeno-
Once metastatic tumors or extension from an pausal patient, the possibility of nonneoplastic
endocervical primary has been excluded, the dif- lesions, such as the Arias-Stella reaction, is much
ferential diagnosis of endometrial carcinoma more likely than carcinoma. Thus, the clinical
includes a number of benign cytoplasmic history often clarifies the diagnosis in question-
changes/metaplasia or hyperplasia (Chap. 9), able cases. Features that assist in the recognition
pregnancy-related changes (Chap. 3), and tissue of the Arias-Stella reaction include multifocality
artifacts (Chap. 2) [28, 34, 59, 267–270]. of the changes and intimate association with
For instance, when prominent squamous, secretory glands, lack of a discrete lesion or fea-
mucinous, or eosinophilic cell changes are pres- tures of invasion (e.g., confluent glands, altered
ent, especially in specimens with considerable stroma, or extensive papillary pattern), presence
tissue fragmentation, polyps, hyperplasia, or of decidua, degenerative and optically clear
inflammatory conditions must be excluded before nuclei with vacuolated cytoplasm, and no associ-
a diagnosis of cancer is established [28, 268– ated mitotic figures. In limited samples, Ki-67
275]. Consequently, carcinoma should be diag- immunostaining is useful in the differential diag-
nosed only when there is a glandular proliferation nosis, as the Arias-Stella reaction has a very low
that fulfills, at a minimum, the histologic criteria proliferative index (see Chap. 3).
for well-differentiated endometrioid carcinoma Tissue artifacts can yield worrisome patterns
described previously or if the lesion shows that may mimic carcinoma. For example, the arti-
unequivocal features of malignancy with cyto- factual crowding and distortion of glands that
logic features of grade 2 or 3 carcinoma. Detached occur during biopsy can result in glands becom-
epithelial fragments, especially when they lack ing closely apposed (see Chap. 2). Likewise,
significant nuclear atypia, should not be diag- breakdown and bleeding distort the normal archi-
nosed as carcinoma (see Chap. 9 for details). tecture and present a variety of cytologic
Among the polyps, the atypical polypoid ade- alterations, including eosinophilic syncytial

nomyoma can be confused with carcinoma of the change (see Chap. 9). Cervical contaminants,
endometrium because the lesion shows atypical especially endocervical squamous metaplasia,
glands, usually with squamous morules, in prominent detached fragments of endocervical
smooth muscle that can be confused with carci- epithelium, or microglandular hyperplasia, may
noma invading the myometrium (see Chap. 8) become mixed with endometrial tissue in curet-
[276–278]. Myometrial invasion is rarely seen in ting specimens and yield a complex pattern that
biopsy and curettage specimens, however. can mimic carcinoma at first glance. These tissue
Furthermore, the orderly pattern of the smooth artifacts usually are not a major problem in the
muscle of the atypical polypoid adenomyoma differential diagnosis, as they are generally focal
contrasts with the desmoplasia typically associ- and admixed with normal endometrium. In men-
ated with neoplasia. In those rare cases in which strual endometrium, in which there is more dif-
the differential diagnosis includes the atypical fuse and extensive breakdown, the possibility of
polypoid adenomyoma and adenocarcinoma, it is mistaking the pattern for adenocarcinoma is
important to note that a confluent or cribriform greater. In such cases, it is important to attempt to
pattern does not occur in the atypical polypoid identify secretory glandular changes and intact
adenomyoma, unless there is carcinoma arising endometrium that is not undergoing breakdown.
in it. Immunohistochemical stains for desmin can Furthermore, the clinical history, including the
help to demonstrate the smooth muscle in the patient’s age and menstrual status, can be very
atypical adenomyoma and distinguish it from the helpful in recognizing a menstrual pattern that
fibroblastic desmoplasia of carcinoma. may not be obvious at first inspection.
During pregnancy, crowded secretory glands, Stromal artifacts also can mimic adenocarci-
including those that display the Arias-Stella reac- noma, especially when they develop a signet-ring
cell morphology [279]. Aggregates of signet-ring sue breakdown and necrosis are so ubiquitous in
cells in the stroma can be seen in cases with endometrial tissue, the finding of this pattern of
extensive stromal decidual change, especially necrosis is not helpful in establishing the diagno-
following progestin hormone therapy [279]. sis of neoplasia. Another pattern of necrosis can
Rarely, stromal histiocytes may also have a be seen in the center of nests of squamous change
signet-ring appearance and present as nodules in (morules). Central necrosis in areas of squamous
the endometrial cavity [280]. In both scenarios, change can occur in benign lesions, including
the apparent signet-ring-like cells lack nuclear any form of hyperplasia, polyps, and the atypical
atypia. In cases of extensive stromal decidual polypoid adenomyoma. Consequently, the necro-
change, the cells are negative for broad-spectrum sis associated with squamous change has no sig-
keratin immunostains and are mucin negative by nificance and does not indicate malignancy.
histochemical stains. Signet-ring-like histiocytes
are immunoreactive for CD68. True signet-ring
cells are rare in endometrial biopsy specimens. In Clinical Queries and Reporting
such cases, they may represent a metastasis from
a breast or gastrointestinal primary carcinoma or, It is important to establish the diagnosis of atypi-
alternatively, be part of an unusual signet-ring cal hyperplasia versus well-differentiated endo-
variant of endometrioid adenocarcinoma [181, metrioid carcinoma whenever possible. A biopsy
281–283]. diagnosis of atypical hyperplasia may be man-
Fragmentation and artifacts also occur in curet- aged medically with progestin therapy and peri-
tage specimens containing adenocarcinoma. odic resampling of the endometrium. This
When this occurs, the diagnosis of carcinoma still conservative management is possible because
can be made if the epithelial cells demonstrate many lesions are reversible. A biopsy diagnosis
high nuclear grade. Those areas where glands are of well-differentiated endometrioid carcinoma,
attached to surrounding mesenchymal tissue and in contrast, clearly establishes the presence of
are free of the changes of breakdown and bleeding malignancy. However, selected cases of well-
demonstrate the true relationship of the glands to differentiated endometrioid carcinoma may be
each other. To reliably identify malignancy in treated with progestin therapy with very close
equivocal cases, however, it is best to evaluate the clinical follow up.
features that establish the diagnosis of well-differ- Although the criteria for distinguishing atypical
entiated endometrioid carcinoma in clearly intact hyperplasia from well-differentiated endometrioid
areas with identifiable stroma. carcinoma are well defined, the diagnosis should
Actual tumor necrosis rarely, if ever, occurs in also be placed in the appropriate clinical context in
well-differentiated tumors. This form of necrosis equivocal cases. For example, in younger, pre-
is often seen in high-grade tumors, but in these menopausal patients, a conservative approach is
cases the histologic and nuclear features readily appropriate when the biopsy shows an atypical
identify the lesion as carcinoma. The presence of lesion that may represent well-differentiated endo-
tumor cell necrosis in an otherwise low-grade metrioid carcinoma. Studies show that these
carcinoma could be a red flag indicating a neo- lesions, even if they fulfill criteria for adenocarci-
plasm of more aggressive behavior and therefore noma, are indolent and reversible in up to 75% of
warranting a hysterectomy with lymph node sam- cases with progestin therapy [284]. Thus, in the
pling [54]. Much of the “necrosis” that is com- premenopausal patient younger than 40 years of
monly encountered in biopsy material is unique age, it is especially important to be certain that a
to the endometrium and likely reflects breakdown lesion at least fulfills the minimal criteria for well-
rather than necrotic tumor. The necrosis associ- differentiated endometrioid carcinoma before
ated with bleeding and breakdown is apparent in establishing a diagnosis of malignancy. If the
low-grade carcinoma but often occurs in benign glands are atypical but do not clearly show fea-
conditions too, regardless of cause. Because tis- tures of “stromal invasion” as defined earlier, then
the lesion is best classified as atypical hyperplasia/ Aberrant expression of p53 and strong and diffuse
EIN. These patients can be managed conserva- immunoreactivity for p16 may assist in classifying
tively and re-biopsied, with one study showing a a tumor as serous carcinoma.
90% regression rate [284]. In older postmeno- Serous carcinoma and endometrioid carci-
pausal women, atypical glands should be viewed noma with clear cell/secretory change are the
even more suspiciously for the possibility of main entities in the differential diagnosis of clear
underlying carcinoma, and criteria for the histo- cell carcinoma. Although serous and clear cell
logic diagnosis of carcinoma can be applied more carcinoma share high-grade nuclei, certain mor-
liberally. Most of these patients are best treated by phologic features, such as papillae with hyalin-
a total hysterectomy and bilateral salpingo-oopho- ized cores, abundant clear cytoplasm, absence of
rectomy if they are candidates for surgery. nuclear pseudostratification, absence of diffuse
Once the diagnosis of endometrial carcinoma grade 3 nuclei, absence of long and slender papil-
is made, the gynecologist often needs information lae, and nuclear hyperchromasia in at least one
on several other aspects of the tumor. Whenever third of the studied lesion, tend to favor clear cell
possible, the FIGO grade of endometrioid carci- carcinoma [285].
noma in biopsies should be given. For instance, In contrast to the high nuclear grade of clear
low-grade endometrioid carcinoma frequently is cell carcinoma, other carcinomas with clear or
confined to the endometrium or is only superfi- pale cytoplasm lack high nuclear grade and do
cially invasive of the myometrium; extrauterine not show the consistent growth patterns of clear
spread is unlikely. Conversely, high-grade endo- cell carcinoma including papillary growth with
metrial carcinoma, including aggressive histo- stromal hyalinized cores, hobnail cells, and
logic subtypes of serous and clear cell carcinoma, tubulocystic arrangements. Some of these carci-
has greater potential for extrauterine spread at the nomas with lower-grade nuclei tend to be solid,
time of hysterectomy. If grading is not possible containing cells with clear cytoplasm that may
owing to limited sampling, or to necrosis or other represent squamous differentiation. If an endo-
distortion of the tissue, this should be noted. metrial carcinoma with clear cells does not show
Classifying the tumor according to histologic the typical growth patterns with high nuclear
subtype also is important. When the tumor has an grade, it should be classified as an endometrioid
endometrioid, villoglandular, secretory, or muci- tumor with secretory change [286].
nous pattern, the type of tumor has little clinical A rare case of apparent secretory carcinoma
significance by itself once the tumor has been may show foci with high-grade nuclei, suggest-
assigned the appropriate histologic grade. ing that clear cell carcinoma may be present, so it
Squamous differentiation, too, has little clinical is important to thoroughly sample any low-grade
significance once the tumor is graded but may be tumor with clear cytoplasm to determine if grade
relevant to note for histologic correlation with 3 nuclei and a more aggressive neoplasm may be
any subsequent metastases or recurrences. On the present.
other hand, serous and clear cell carcinomas are Mixed carcinomas have at least 5% of a sec-
aggressive tumors that are important to identify. ond cell type [31]. An example would be a mix-
The diagnosis of these tumors indicates that there ture of endometrioid and serous patterns. If 25%
is an increased risk for deep myometrial invasion or more of such tumor is composed of serous car-
and metastases. cinoma, then it should be classified as such
Endometrial neoplasms may be difficult to clas- because these mixed neoplasms behave the same
sify in small tissue samples or when extensive as pure serous carcinoma. Otherwise it can be
necrosis is present. In such limited materials, immu- termed a “mixed carcinoma,” with a note describ-
nohistochemistry has utility for subclassification of ing the different types of carcinoma present.
primary endometrial carcinomas (Table 10.4). For The amount of tumor in the specimen and
instance, cytokeratin reactivity can help determine associated lesions such as hyperplasia or polyps
whether or not a malignant tumor is a carcinoma. are potentially useful data. For example, a small
References 321
amount of tumor may alert the clinician to diffi- nosarcoma with sarcomatous overgrowth.
culties in accurate grading. The presence of asso- Practically, this distinction is not important for
ciated hyperplasia suggests estrogenic effects directing further therapy. These types of tumors
that may influence the method of therapy. Should are high-grade malignancies with a similar poor
the pattern suggest a metastatic process, the prognosis. Consequently, when the tumor type is
report should clearly indicate this and suggest the not clear-cut, a diagnosis of “malignant neo-
primary sites, if possible. plasm” with a comment indicating the differen-
Carcinosarcoma is an important diagnosis, tial diagnosis should be made. The diagnosis
because the tumor is highly aggressive. Correct usually is clarified by the study of additional tis-
diagnosis prior to hysterectomy alerts the gyne- sue. If the patient is a candidate for surgery, hys-
cologist to the high likelihood of extrauterine terectomy usually is attempted to stage the tumor,
spread at the time of a surgical staging procedure. and the neoplasm can be fully assessed by exam-
High-grade heterologous or high-grade ining the entire uterus.
homologous tumors are associated with decreased Finally, increasing awareness of Lynch syn-
progression-free survival compared to low-grade drome (hereditary nonpolyposis colorectal cancer)
homologous neoplasms [287]. According to one and the understanding of the risk to develop endo-
series, rhabdomyosarcoma portends a worse metrial carcinoma by affected patients have
prognosis compared to other heterologous com- prompted most academic institutions in the country
ponents [201]. When the epithelial component is to accept requests for universal screening in pathol-
serous or clear cell carcinoma, there is a greater ogy samples, a practice which we encourage.
tendency for deep myometrial invasion and meta- In the future, it appears likely that the diag-
static disease [197, 288]. A serous carcinoma epi- nostic approach to endometrial carcinoma will
thelial component has been associated with a be multifaceted, incorporating morphologic,
poor 3-year overall survival [201]. Therefore, immunohistochemical as well as molecular
these histologic types should be noted. information. This additional information will
In some cases, biopsy or curettage may show a help refine classification and provide better
tumor with features that suggest a carcinosarcoma prognostic information for clinicians and
but are not conclusive. This problem is most often patients. Nonetheless, careful evaluation of rou-
seen when the sections show adenocarcinoma tine H&E sections should remain crucial for
with foci of cellular spindle cells around the appropriate triage and therapy.
glands. The spindle cell component may show
increased mitotic activity but is not conclusive for
a sarcomatous element. In such cases, it is best to References
diagnose the carcinoma, clearly describe the other
changes, and to indicate the possibility that the 1. U.S. Cancer Statistics Working Group. United States
cancer statistics: 2015 Incidence and mortality web-
tumor may be a carcinosarcoma. The knowledge based report. Atlanta: Department of Health and Human
that the tumor may be a carcinosarcoma often is Services, Centers for Disease Control and Prevention,
sufficient for clinical management in specimens and National Cancer Institute; 2017. Available at:
with equivocal features, as a hysterectomy will be https://gis.cdc.gov/Cancer/USCS/DataViz.html
2. Farley J, Risinger JI, Rose GS, Maxwell GL. Racial
performed. The entire uterus can then be studied disparities in blacks with gynecologic cancers.
in greater detail to determine whether a sarcoma- Cancer. 2007;110(2):234–43.
tous component is present. 3. Cote ML, Ruterbusch JJ, Olson SH, Lu K, Ali-Fehmi
An occasional case demonstrates only high- R. The growing burden of endometrial cancer: a
major racial disparity affecting black women. Cancer
grade sarcoma with no definite epithelial ele- Epidemiol Biomark Prev. 2015;24(9):1407–15.
ments, despite immunohistochemical analysis. In 4. Smith M, McCartney AJ. Occult, high-risk endome-
these cases, the differential diagnosis includes a trial cancer. Gynecol Oncol. 1985;22:154–61.
pure sarcoma, a carcinosarcoma with predomi- 5. Bokhman JV. Two pathogenetic types of endome-
trial carcinoma. Gynecol Oncol. 1983;15:10–7.
nance of the sarcomatous component, or an ade-
6. Ellenson LH, Ronnett BM, Soslow RA. Endometrial PM, Smit VT, Creutzberg CL. Refining prognosis
carcinoma. In: Kurman RJ, editor. Blaustein’s pathol- and identifying targetable pathways for high-risk
ogy of the female genital tract. 6th ed. New York: endometrial cancer; a TransPORTEC initiative. Mod
Springer; 2011. p. 393–452. Pathol. 2015;28(6):836–44.
7. Sherman ME. Theories of endometrial carcino- 20. Talhouk A, McConechy MK, Leung S, Yang W, Lum
genesis: a multidisciplinary approach. Mod Pathol. A, Senz J, Boyd N, Pike J, Anglesio M, Kwon JS,
2000;13:295–308. Karnezis AN, Huntsman DG, Gilks CB, McAlpine
8. Weiderpass E, Adami HO, Baron JA, Magnusson C, JN. Confirmation of ProMisE: a simple, genomics-
Bergström R, Lindgren A, Correia N, Persson I. Risk based clinical classifier for endometrial cancer.
of endometrial cancer following estrogen replace- Cancer. 2017;123(5):802–13.
ment with and without progestins. J Natl Cancer 21. Kommoss S, McConechy MK, Kommoss F, Leung
Inst. 1999;91(13):1131–7. S, Bunz A, Magrill J, Britton H, Kommoss F,
9. Longacre TA, Hendrickson MR. Diffusely infil- Grevenkamp F, Karnezis A, Yang W, Lum A, Krämer
trative endometrial adenocarcinoma: an adenoma B, Taran F, Staebler A, Lax S, Brucker SY, Huntsman
malignum pattern of myoinvasion. Am J Surg Pathol. DG, Gilks CB, McAlpine JN, Talhouk A. Final vali-
1999;23:69–78. dation of the ProMisE molecular classifier for endo-
10. Landry D, Mai KT, Senterman MK, Perkins DG, metrial carcinoma in a large population-based case
Yazdi HM, Veinot JP, et al. Endometrioid adenocar- series. Ann Oncol. 2018;29(5):1180–8.
cinoma of the uterus with a minimal deviation inva- 22. Talhouk A, Hoang LN, McConechy MK, Nakonechny
sive pattern. Histopathology. 2003;42:77–82. Q, Leo J, Cheng A, Leung S, Yang W, Lum A, Köbel
11. Soslow RA. Endometrial carcinomas with ambiguous M, Lee CH, Soslow RA, Huntsman DG, Gilks CB,
features. Semin Diagn Pathol. 2010;27(4):261–73. McAlpine JN. Molecular c lassification of endome-
12. Gilks CB, Oliva E, Soslow RA. Poor interob- trial carcinoma on diagnostic specimens is highly
server reproducibility in the diagnosis of high- concordant with final hysterectomy: earlier prognos-
grade endometrial carcinoma. Am J Surg Pathol. tic information to guide treatment. Gynecol Oncol.
2013;37(6):874–81. 2016;143(1):46–53.
13. Lomo L, Nucci MR, Lee KR, Lin MC, Hirsch MS, 23. DiSaia PJ, Creasman WT. Clinical gynecologic
Crum CP, Mutter GL. Histologic and immunohisto- oncology. 6th ed. St. Louis: Mosby-Year Book;
chemical decision-making in endometrial adenocar- 2002.
cinoma. Mod Pathol. 2008;21(8):937–42. 24. Hendrickson MR, Ross JC, Kempson RL. Toward
14. Han G, Sidhu D, Duggan MA, Arseneau J, Cesari the development of morphologic criteria for well-
M, Clement PB, Ewanowich CA, Kalloger SE, differentiated adenocarcinoma of the endometrium.
Köbel M. Reproducibility of histological cell type Am J Surg Pathol. 1983;7:819–38.
in high-grade endometrial carcinoma. Mod Pathol. 25. Kurman RJ, Norris HJ. Evaluation of criteria for
2013;26(12):1594–604. distinguishing atypical endometrial hyperpla-
15. Thomas S, Hussein Y, Bandyopadhyay S, Cote M, sia from well-differentiated carcinoma. Cancer.
Hassan O, Abdulfatah E, Alosh B, Guan H, Soslow 1982;49:2547–57.
RA, Ali-Fehmi R. Interobserver variability in the 26. Norris HJ, Tavassoli FA, Kurman RJ. Endometrial
diagnosis of uterine high-grade endometrioid carci- hyperplasia and carcinoma. Diagnostic consider-
noma. Arch Pathol Lab Med. 2016;140(8):836–43. ations. Am J Surg Pathol. 1983;7:839–47.
16. DF DL, Burke KA, Selenica P, Lim RS, Scott SN, 27. Kraus FT. High-risk and premalignant lesions
Middha S, Mohanty AS, Cheng DT, Berger MF, of the endometrium. Am J Surg Pathol.
Soslow RA, Weigelt B. The genetic landscape 1985;9(Suppl):31–40.
of endometrial clear cell carcinomas. J Pathol. 28. Ellenson LH, Ronnett BM, Kurman RJ. Precursor
2017;243(2):230–41. lesions of endometrial carcinoma. In: Kurman R,
17. Bosse T, Nout RA, JN MA, MK MC, Britton H, editor. Blaustein’s pathology of the female genital
Hussein YR, Gonzalez C, Ganesan R, Steele JC, tract. 6th ed. New York: Springer; 2011. p. 359–91.
Harrison BT, Oliva E, Vidal A, Matias-Guiu X, 29. Lehman MB, Hart WR. Simple and complex hyper-
Abu-Rustum NR, Levine DA, Gilks CB, Soslow plastic papillary proliferations of the endometrium:
RA. Molecular classification of grade 3 endometri- a clinicopathologic study of nine cases of appar-
oid endometrial cancers identifies distinct prognostic ently localized papillary lesions with fibrovascular
subgroups. Am J Surg Pathol. 2018;42(5):561–8. stromal cores and epithelial metaplasia. Am J Surg
18. Goebel EA, Vidal A, Matias-Guiu X, Blake Gilks Pathol. 2001;25:1347–54.
C. The evolution of endometrial carcinoma classifi- 30. King A, Seraj IM, Wagner RJ. Stromal invasion
cation through application of immunohistochemistry in endometrial carcinoma. Am J Obstet Gynecol.
and molecular diagnostics: past, present and future. 1984;149:10–4.
Virchows Arch. 2017;472(6):885–96. 31. Zaino R, Carinelli S, Ellenson L. Uterine corpus:
19. Stelloo E, Bosse T, Nout RA, HJ MK, Church DN, epithelial tumours and precursors. In: Kurman R,
Nijman HW, Leary A, Edmondson RJ, Powell ME, Carcangiu M, Herrington C, Young R, editors. World
Crosbie EJ, Kitchener HC, Mileshkin L, Pollock health organization classification of tumours of the
References 323
female reproductive organs. 4th ed. Lyon: IARC; adenocarcinoma: a clinical pathologic study of 46
2014. p. 125–35. cases. Am J Clin Pathol. 1982;77:534–40.
32. Connelly PJ, Alberhasky RC, Christopherson 46. Daniel AG, Peters WA III. Accuracy of office and
WM. Carcinoma of the endometrium. III. Analysis operating room curettage in the grading of endome-
of 865 cases of adenocarcinoma and adenoacan- trial carcinoma. Obstet Gynecol. 1988;71:612–4.
thoma. Obstet Gynecol. 1982;59:569–75. 47. Cowles TA, Magrina JF, Masterson BJ, Capen
33. Hendrickson M, Ross J, Eifel PJ, Cox RJ, Martinez CV. Comparison of clinical and surgical staging
A, Kempson R. Adenocarcinoma of the endome- in patients with endometrial carcinoma. Obstet
trium: An analysis of 256 cases with carcinoma lim- Gynecol. 1985;66:413–6.
ited to the uterine corpus. I. Pathology review and 48. Piver MS, Lele SB, Barlow JJ, Blumenson
analysis of prognostic variables. Gynecol Oncol. L. Paraaortic lymph node evaluation in stage I endo-
1982;13:373–92. metrial cancer. Obstet Gynecol. 1982;59:97–100.
34. Clement PB, Young RH. Endometrioid carcinoma 49. Larson DM, Johnson KK, Broste SK, Krawisz BR,
of the uterine corpus: A review of its pathology Kresl JJ. Comparison of D&C and office endo-
with emphasis on recent advances and problematic metrial biopsy in predicting final histopathologic
aspects. Adv Anat Pathol. 2002;9:145–84. grade in endometrial cancer. Obstet Gynecol.
35. Ferlay J, Shin HR, Bray F, Forman D, et al. Cancer 1995;86:38–42.
incidence and mortality worldwide (2008): IARC 50. Obermair A, Geramou M, Gucer F, Denison U, Graf
CancerBase. IARC. AH, Kapshammer E, et al. Endometrial cancer:
36. Zaino RJ, Kurman R, Herbold D, Gliedman J, Bundy accuracy of the finding of a well differentiated tumor
BN, Voet R, et al. The significance of squamous dif- at dilatation and curettage compared to the findings
ferentiation in endometrial carcinoma—Data from at subsequent hysterectomy. Int J Gynecol Cancer.
a Gynecologic Oncology Group study. Cancer. 1999;9:383–6.
1991;68:2293–302. 51. Mittal KR, Schwartz PE, Barwick KW. Architectural
37. Abeler VM, Kjorstad KE. Endometrial adenocar- (FIGO) grading, nuclear grading, and other prognos-
cinoma with squamous cell differentiation. Cancer. tic indicators in stage I endometrial adenocarcinoma
1992;69:488–95. with identification of high-risk and low-risk groups.
38. Demopoulos RI, Dubin N, Noumoff J, et al. Cancer. 1988;61:538–45.
Prognostic significance of squamous differentia- 52. Nielsen AL, Thomsen HK, Nyholm HCJ. Evaluation
tion in stage I endometrial adenocarcinoma. Obstet of the reproducibility of the revised 1988 International
Gynecol. 1986;68:245–50. Federation of Gynecology and Obstetrics grading
39. Zaino RJ, Kurman RJ, Diana KL, Morrow CP. The system of endometrial cancers with special emphasis
utility of the revised International Federation of on nuclear grading. Cancer. 1991;68:2303–9.
Gynecology and Obstetrics histologic grading of 53. Zaino RJ, Silverberg SG, Norris HJ, Bundy BN,
endometrial adenocarcinoma using a defined nuclear Morrow CP, Okagaki T. The prognostic value of
grading system. A Gynecologic Oncology Group nuclear versus architectural grading in endometrial
study. Cancer. 1995;75:81–6. adenocarcinoma. A Gynecologic Oncology Group
40. Nordstrom B, Strang P, Lindgren A, Bergstrom R, study. Int J Gynecol Pathol. 1994;13:29–36.
Tribukait B. Endometrial carcinoma: the prognos- 54. Lax SF, Kurman RJ, Pizer ES, Wu L, Ronnett
tic impact of papillary serous carcinoma (UPSC) BM. A binary architectural grading system for uter-
in relation to nuclear grade, DNA ploidy and p53 ine endometrial endometrioid carcinoma has supe-
expression. Anticancer Res. 1996;16:899–904. rior reproducibility compared with FIGO grading
41. Ayhan A, Taskiran C, Yuce K, Kucukali T. The and identifies subsets of advance-stage tumors with
prognostic value of nuclear grading and the revised favorable and unfavorable prognosis. Am J Surg
FIGO grading of endometrial adenocarcinoma. Int J Pathol. 2000;24:1201–8.
Gynecol Pathol. 2003;22:71–4. 55. Taylor RR, Zeller J, Lieberman RW, O’Connor
42. Alkushi A, Lim P, Coldman A, Huntsman D, Miller DM. An analysis of two versus three grades for endo-
D, Gilks B. Assessment of a novel grading system for metrial carcinoma. Gynecol Oncol. 1999;74:3–6.
endometrial carcinoma. Mod Pathol. 2003;16:180A. 56. Scholten AN, Smit VTHBM, Beerman H, van Putten
43. Benedet JL, Bender H, Jones H, III NHY, Pecorelli WLJ, Creutzberg CL. Prognostic significance and
S. FIGO staging classifications and clinical practice interobserver variablility of histologic grading systems
guidelines in the management of gynecologic can- form endometrial carcinoma. Cancer. 2004;100:764–72.
cers. FIGO Committee on Gynecologic Oncology. 57. Hendrickson M, Ross J, Eifel P, Martinez A,
Int J Gynaecol Obstet. 2000;70:209–62. Kempson R. Uterine papillary serous carcinoma. A
44. Christopherson WM, Connelly PJ, Alberhasky highly malignant form of endometrial adenocarci-
RC. Carcinoma of the endometrium. V. An analysis noma. Am J Surg Pathol. 1982;6:93–108.
of prognosticators in patients with favorable sub- 58. Clement PB, Scully RE. Endometrial hyperplasia and
types and stage I disease. Cancer. 1983;51:1705–9. carcinoma. In: Clement PB, Young RH, editors. Tumors
45. Christopherson WM, Alberhasky RC, Connelly and tumorlike lesions of the uterine corpus and cervix.
PJ. Carcinoma of the endometrium: II. Papillary New York: Churchill Livingstone; 1993. p. 181–264.
59. Silverberg SG, Kurman RJ. Tumors of the uterine 74. Ross JC, Eifel PJ, Cox RS, Kempson RL, Hendrickson
corpus and gestational trophoblastic disease, Atlas of MR. Primary mucinous adenocarcinoma of the endo-
tumor pathology, 3rd series, Fascicle 3. Washington, metrium. A clinicopathologic and histochemical
DC: Armed Forces Institute of Pathology; 1992. study. Am J Surg Pathol. 1983;7:715–29.
60. Hendrickson MR, Kempson RL. Ciliated carcinoma– 75. Tiltman AJ. Mucinous carcinoma of the endome-
a variant of endometrial adenocarcinoma. A report trium. Obstet Gynecol. 1980;55:244–7.
of 10 cases. Int J Gynecol Pathol. 1983;2:1–12. 76. Czernobilsky B, Katz Z, Lancet M, Gaton
61. Gould PR, Li L, Henderson DW, Barter RA, E. Endocervical type epithelium in endometrial car-
Papadimitriou JM. Cilia and ciliogenesis in endo- cinoma: a report of 10 cases with emphasis on his-
metrial adenocarcinomas. Arch Pathol Lab Med. tochemical methods for differential diagnosis. Am J
1986;110:326–30. Surg Pathol. 1980;4:481–9.
62. Christopherson WM, Alberhasky RC, Connelly 77. Rawish KR, Desouki MM, Fadare O. Atypical
PJ. Carcinoma of the endometrium. I. A clinicopath- mucinous glandular proliferations in endometrial
ologic study of clear cell carcinoma and secretory samplings: follow-up and other clinicopathological
carcinoma. Cancer. 1982;49:1511–23. findings in 41 cases. Hum Pathol. 2017;63:53–62.
63. Kurman RJ, Scully RE. Clear cell carcinoma of 78. Young RH, Scully RE. Uterine carcinomas simu-
the endometrium. An analysis of 21 cases. Cancer. lating microglandular hyperplasia. A report of six
1976;37:872–82. cases. Am J Surg Pathol. 1992;16:1092–7.
64. Tobon H, Watkins GJ. Secretory adenocarci- 79. Zaloudek C, Hayashi GM, Ryan IP, Powell CB,
noma of the endometrium. Int J Gynecol Pathol. Miller TR. Microglandular adenocarcinoma of the
1985;4:328–35. endometrium: a form of mucinous adenocarcinoma
65. Dawagne MP, Silverberg SG. Foam cells in endome- that may be confused with microglandular hyperpla-
trial carcinoma. Gynecol Oncol. 1982;13:67–75. sia of the cervix. Int J Gynecol Pathol. 1997;16:52–9.
66. Silver SA, Sherman ME. Morphologic and immuno- 80. Malpica A. How to approach the many faces of endo-
phenotypic characterization of foam cells in endome- metrioid carcinoma. Mod Pathol. 2016;29(Suppl
trial lesions. Int J Gynecol Pathol. 1998;17:140–5. 1):S29–44.
67. Alkushi A, Clarke BA, Akbari M, Makretsov N, Lim 81. Stewart CJR, Crook ML. PAX2 and cyclin D1
P, Miller D, Magliocco A, Coldman A, van de Rijn expression in the distinction between cervi-
M, Huntsman D, Parker R, Gilks CB. Identification cal microglandular hyperplasia and endometrial
of prognostically relevant and reproducible subsets microglandular-like carcinoma: a comparison with
of endometrial adenocarcinoma based on cluster- p16, vimentin, and Ki-67. Int J Gynecol Pathol.
ing analysis of immunostaining data. Mod Pathol. 2014;34:90–100, 547–554.
2007;20(11):1156–65. 82. Aarnio M, Sankila R, Pukkala E, et al. Cancer risk
68. Wei J-J, Paintal A, Keh P. Histologic and immu- in mutation carriers of DNA-mismatch-repair genes.
nohistochemical analyses of endometrial carci- Int J Cancer. 1999;81:214–8.
nomas: experiences from endometrial biopsies 83. Mills AM, Longacre TA. Lynch syndrome: female
in 358 consultation cases. Arch Pathol Lab Med. genital tract cancer diagnosis and screening. Surg
2013;137(11):1574–83. Pathol Clin. 2016;9(2):201–14.
69. Mittal K, Soslow R, McCluggage WG. Application 84. Peltomaki P, Aaltonen LA, Sistonen P, et al. Genetic
of immunohistochemistry to gynecologic pathology. mapping of a locus predisposing to human colorectal
Arch Pathol Lab Med. 2008;132(3):402–23. cancer. Science. 1993;260:810–2.
70. Chekmareva M, Ellenson LH, Pirog 85. Loeb LA. Microsatellite instability: marker of
EC. Immunohistochemical differences between a mutator phenotype in cancer. Cancer Res.
mucinous and microglandular adenocarcinomas of 1994;54:5059–63.
the endometrium and benign endocervical epithe- 86. Westin SN, Lacour RA, Urbauer DL, et al.
lium. Int J Gynecol Pathol. 2008;27(4):547–54. Carcinoma of the lower uterine segment: a newly
71. Alvarez T, Miller E, Duska L, et al. Molecular pro- described association with Lynch syndrome. J Clin
file of grade 3 endometrioid endometrial carcinoma: Oncol. 2008;26:5965–71.
is it a type I or type II endometrial carcinoma? Am J 87. Tafe LJ, Garg K, Chew I, et al. Endometrial and
Surg Pathol. 2012;36:753–61. ovarian carcinomas with undifferentiated compo-
72. Murray SK, Clement PB, Young RH. Endometrioid nents: clinically aggressive and frequently underrec-
carcinomas of the uterine corpus with sex cord-like ognized neoplasms. Mod Pathol. 2010;23:781–9.
formations, hyalinization, and other unusual mor- 88. Ryan P, Mulligan AM, Aronson M, et al. Comparison
phologic features: a report of 31 cases of a neo- of clinical schemas and morphologic features in
plasm that may be confused with carcinosarcoma predicting Lynch syndrome in mutation-positive
and other uterine neoplasms. Am J Surg Pathol. patients with endometrial cancer encountered in the
2005;29(2):157–66. context of familial gastrointestinal cancer registries.
73. Melhem MF, Tobon H. Mucinous adenocarcinoma Cancer. 2012;118:681–8.
of the endometrium: a clinico-pathological review of 89. Mills AM, Liou S, Ford JM, et al. Lynch syndrome
18 cases. Int J Gynecol Pathol. 1987;6:347–55. screening should be considered for all patients with
References 325
newly diagnosed endometrial cancer. Am J Surg 102. Dunton CJ, Balsara G, McFarland M, Hernandez
Pathol. 2014;38:1501–9. E. Uterine papillary serous carcinoma. A review.
90. Overbeek LI, Ligtenberg MJ, Willems RW, Hermens Obstet Gynecol Surv. 1991;46:97–102.
RP, Blokx WA, Dubois SV, van der Linden H, Meijer 103. Sherman ME, Bitterman P, Rosenshein NB, Delgado
JW, Mlynek-Kersjes ML, Hoogerbrugge N, Hebeda G, Kurman RJ. Uterine serous carcinoma—a morpho-
KM, van Krieken JH. Interpretation of immunohis- logically diverse neoplasm with unifying clinicopath-
tochemistry for mismatch repair proteins is only ologic features. Am J Surg Pathol. 1992;16:600–10.
reliable in a specialized setting. Am J Surg Pathol. 104. Lee KR, Belinson JL. Papillary serous adenocarci-
2008;32(8):1246–51. noma of the endometrium: a clinicopathologic study
91. Mills A, Sloan E, Thomas M, et al. Clinicopathologic of 19 cases. Gynecol Oncol. 1992;46:51–4.
comparison of Lynch syndrome-associated 105. Carcangiu ML, Chambers JT. Uterine papillary
and Lynch-like endometrial carcinomas identi- serous carcinoma—a study on 108 cases with
fied on universal screening. Am J Surg Pathol. emphasis on the prognostic significance of associ-
2015;40:155–65. ated endometrioid carcinoma, absence of invasion,
92. Resnick K, Straughn JM, Backes F, et al. Lynch and concomitant ovarian carcinoma. Gynecol Oncol.
syndrome screening strategies among newly diag- 1992;47:298–305.
nosed endometrial cancer patients. Obstet Gynecol. 106. Demopoulos RI, Genega E, Vamvakas E, Carlson E,
2009;114:530–6. Mittal K. Papillary carcinoma of the endometrium:
93. Nelen MR, GW PA, Peeters EA, et al. Localization morphometric predictors of survival. Int J Gynecol
of the gene for Cowden disease to chromosome Pathol. 1996;15:110–8.
10q22–23. Nat Genet. 1996;13:114–6. 107. Sutton GP, Brill L, Michael H, et al. Malignant pap-
94. Bubien V, Bonnet F, Brouste V, et al. High cumu- illary lesions of the endometrium. Gynecol Oncol.
lative risks of cancer in patients with PTEN 1987;27:294–304.
hamartoma tumour syndrome. J Med Genet. 108. Chen JL, Trost DC, Wilkinson EJ. Endometrial pap-
2013;50:255–63. illary adenocarcinomas: two clinicopathological
95. Garg K, Karnezis AN, Rabban JT. Uncommon types. Int J Gynecol Pathol. 1985;4:279–88.
hereditary gynaecological tumour syndromes: 109. Lax SF, Kendall B, Tashiro H, Slebos RJ, Hedrick
pathological features in tumours that may pre- L. The frequency of p53, K-ras mutations, and mic-
dict risk for a germline mutation. Pathology. rosatellite instability differs in uterine endometrioid
2018;50(2):238–56. and serous carcinoma: evidence of distinct molecu-
96. Daly MB, Pilarski R, Berry M, Buys SS, Farmer M, lar genetic pathways. Cancer. 2000;88:814–24.
Friedman S, Garber JE, Kauff ND, Khan S, Klein 110. Sherman ME, Bur ME, Kurman RJ. p53 in endo-
C, Kohlmann W, Kurian A, Litton JK, Madlensky L, metrial cancer and its putative precursors: evidence
Merajver SD, Offit K, Pal T, Reiser G, Shannon KM, for diverse pathways of tumorigenesis. Hum Pathol.
Swisher E, Vinayak S, Voian NC, Weitzel JN, Wick 1995;26:1268–74.
MJ, Wiesner GL, Dwyer M, Darlow S. NCCN guide- 111. Tashiro H, Isacson C, Levine R, Kurman RJ, Cho
lines insights: genetic/familial high-risk assessment: KR, Hedrick L. p53 gene mutations are common
breast and ovarian, version 2.2017. J Natl Compr in uterine serous carcinoma and occur early in their
Cancer Netw. 2017;15(1):9–20. pathogenesis. Am J Pathol. 1997;150:177–85.
97. Gunderson CC, Dutta S, Fader AN, Maniar KP, 112. Kovalev S, Marchenko ND, Gugliotta BG, Chalas
Nasseri-Nik N, Bristow RE, Diaz-Montes TP, E, Chumas J, Moll UM. Loss of p53 function in
Palermo R, Kurman RJ. Pathologic features associ- uterine papillary serous carcinoma. Hum Pathol.
ated with resolution of complex atypical hyperplasia 1998;29(6):613–9.
and grade 1 endometrial adenocarcinoma after pro- 113. Köbel M, Ronnett BM, Singh N, Soslow RA, Gilks
gestin therapy. Gynecol Oncol. 2014;132(1):33–7. CB, McCluggage WG. Interpretation of P53 immu-
98. Wheeler DT, Bristow RE, Kurman RJ. Histologic nohistochemistry in endometrial carcinomas: toward
alterations in endometrial hyperplasia and well- increased reproducibility. Int J Gynecol Pathol. 2018;
differentiated carcinoma treated with progestins. Am https//doi.org/10.1097/PGP.0000000000000488.
J Surg Pathol. 2007;31(7):988–98. 114. Fadare O, Roma AA, Parkash V, Zheng W, Walavalkar
99. Mentrikoski MJ, Shah AA, Hanley KZ, Atkins V. Does a p53 “wild-type” Immunophenotype
KA. Assessing endometrial hyperplasia and car- exclude a diagnosis of endometrial serous carci-
cinoma treated with progestin therapy. Am J Clin noma? Adv Anat Pathol. 2018;25(1):61–70.
Pathol. 2012;139:524–34. 115. Lax SF, Pizer ES, Ronnett BM, Kurman RJ. Clear cell
100. Hamilton CA, Cheung MK, Osann K, et al. Uterine carcinoma of the endometrium is characterized by a dis-
papillary serous and clear cell carcinomas predict for tinctive profile of p53, Ki-67, estrogen, and progester-
poorer survival compared to grade 3 endometrioid one receptor expression. Hum Pathol. 1998;29:551–8.
corpus cancers. Br J Cancer. 2006;94:642–6. 116. Carcangiu ML, Chambers JT, Voynick IM, Pirro
101. Abeler VM, Kjorstad KE. Serous papillary carci- M, Schwartz PE. Immunohistochemical evalua-
noma of the endometrium: a histopathological study tion of estrogen and progesterone receptor content
of 22 cases. Gynecol Oncol. 1990;39:266–71. in 183 patients with endometrial carcinoma. Part I:
clinical and histologic correlations. Am J Clin Pathol. gin of extrauterine disease. Int J Gynecol Pathol.
1990;94(3):247–54. 2001;20(3):214–9.
117. Kounelis S, Kapranos N, Kouri E, Coppola D, 132. Gehrig PA, Groben PA, Fowler WC Jr, Walton LA,
Papadaki H, Jones MW. Immunohistochemical Van Le L. Noninvasive papillary serous carcinoma of
profile of endometrial adenocarcinoma: a study of the endometrium. Obstet Gynecol. 2001;97:153–7.
61 cases and review of the literature. Mod Pathol. 133. Soslow RA, Pirog E, Isacson C. Endometrial
2000;13:379–88. intraepithelial carcinoma with associated peritoneal
118. Al-Hussaini M, Stockman A, Foster H, McCluggage carcinomatosis. Am J Surg Pathol. 2000;24:726–32.
WG. WT-1 assists in distinguishing ovarian from 134. Chan JK, Loizzi V, Youssef M, Osann K, Rutgers
uterine serous carcinoma and in distinguishing J, Vasilev SA, et al. Significance of comprehensive
between serous and endometrioid ovarian carci- surgical staging in noninvasive papillary serous
noma. Histopathology. 2004;44(2):109–15. carcinoma of the endometrium. Gynecol Oncol.
119. Hedley C, Sriraksa R, Showeil R, Van Noorden S, 2003;90:181–5.
El-Bahrawy M. The frequency and significance of 135. Slomovitz BM, Burke TW, Eifel PJ, Ramondetta
WT-1 expression in serous endometrial carcinoma. LM, Silva EG, Jhingran A, et al. Uterine papillary
Hum Pathol. 2014;45(9):1879–84. serous carcinoma (UPSC): a single institution review
120. Chambers JT, Merino M, Kohorn EI, et al. Uterine of 129 cases. Gynecol Oncol. 2003;91:463–9.
papillary serous carcinoma. Obstet Gynecol. 136. Abeler VM, Kjorstad KE. Clear cell carcinoma of
1987;69:109. the endometrium: a histopathological and clinical
121. Jeffrey JF, Krepart GV, Lotocki RJ. Papillary study of 97 cases. Gynecol Oncol. 1991;40:207–17.
serous adenocarcinoma of the endometrium. Obstet 137. Webb GA, Lagios MD. Clear cell carcinoma
Gynecol. 1986;67:670. of the endometrium. Am J Obstet Gynecol.
122. Woolas RP, Williams K, McCartney AJ. Uterine pap- 1987;156:1486–91.
illary serous carcinoma in a 32-year-old with Turner’s 138. Kanbour-Shakir A, Tobon H. Primary clear cell carci-
syndrome. Gynecol Oncol. 1994;52:395–401. noma of the endometrium. A clinicopathologic study
123. Cirisano FD Jr, Robboy SJ, Dodge RK, Bentley RC, of 20 cases. Int J Gynecol Pathol. 1991;10:67–78.
Krigman HR, Synan IS, et al. Epidemiologic and sur- 139. Sakuragi N, Hareyama H, Todo Y, Yamada H,
gicopathologic findings of papillary serous and clear Yamamoto R, Fujino T, et al. Prognostic significance
cell endometrial cancers when compared to endome- of serous and clear cell adenocarcinoma in surgically
trioid carcinoma. Gynecol Oncol. 1999;74:385–94. staged endometrial carcinoma. Acta Obstet Gynecol
124. Evans-Metcalf ER, Brooks SE, Reale FR, Baker Scand. 2000;79:311–6.
SP. Profile of women 45 years of age and younger 140. Abeler VM, Vergote IB, Kjorstad KE, Trope CG. Clear
with endometrial cancer. Obstet Gynecol. cell carcinoma of the endometrium. Prognosis and
1998;91:349–54. metastatic pattern. Cancer. 1996;78:1740–7.
125. Silva EG, Jenkins R. Serous carcinoma in endome- 141. Hariri N, Qarmali M, Fadare O. Endometrial serous
trial polyps. Mod Pathol. 1990;3:120–8. carcinoma with clear-cell change: frequency and
126. McCluggage WG, Sumathi VP, McManus immunohistochemical analysis. Int J Surg Pathol.
DT. Uterine serous carcinoma and endometrial 2018;26(2):126–34.
intraepithelial carcinoma arising in endometrial pol- 142. Bartosch C, Manuel Lopes J, Oliva E. Endometrial
yps: report of 5 cases, including 2 associated with carcinomas: a review emphasizing overlapping and
tamoxifen therapy. Hum Pathol. 2003;34:939–43. distinctive morphological and immunohistochemical
127. Aquino-Parsons C, Lim P, Wong F, Milden-berger features. Adv Anat Pathol. 2011;18(6):415–37.
M. Papillary serous and clear cell carcinoma limited 143. Fadare O, Gwin K, Desouki MM, Crispens MA,
to endometrial curettings in FIGO stage 1a and 1b Jones HW 3rd, Khabele D, Liang SX, Zheng W,
endometrial adenocarcinoma: treatment implica- Mohammed K, Hecht JL, Parkash V. The clini-
tions. Gynecol Oncol. 1998;71:83–6. copathologic significance of p53 and BAF-250a
128. Wheeler DT, Bell KA, Kurman RJ, Sherman (ARID1A) expression in clear cell carcinoma of the
ME. Minimal uterine serous carcinoma: diagnosis endometrium. Mod Pathol. 2013;26(8):1101–10.
and clinicopathologic correlation. Am J Surg Pathol. 144. Han G, Soslow RA, Wethington S, Levine DA,
2000;24:797–806. Bogomolniy F, Clement PB, Köbel M, Gilks B,
129. Lee KR, Belinson JL. Recurrence in noninvasive DeLair D. Endometrial carcinomas with clear cells:
endometrial carcinoma. Relationship to uterine a study of a heterogeneous group of tumors includ-
papillary serous carcinoma. Am J Surg Pathol. ing interobserver variability, mutation analysis, and
1991;15:965–73. immunohistochemistry with HNF-1β. Int J Gynecol
130. Carcangiu ML, Tan LK, Chambers JT. Stage IA Pathol. 2015;34(4):323–33.
uterine serous carcinoma: a study of 13 cases. Am J 145. Iwamoto M, Nakatani Y, Fugo K, Kishimoto T,
Surg Pathol. 1997;21:1507–14. Kiyokawa T. Napsin a is frequently expressed in
131. Baergen RN, Warren CD, Isacson C, Ellenson clear cell carcinoma of the ovary and endometrium.
LH. Early uterine serous carcinoma: clonal ori- Hum Pathol. 2015;46(7):957–62.
References 327
146. Fadare O, Desouki MM, Gwin K, Hanley KZ, Jarboe case report with literature review. Arch Gynecol
EA, Liang SX, Quick CM, Zheng W, Parkash V, Obstet. 2011;283(2):323–7.
Hecht JL. Frequent expression of napsin A in clear 159. Fluhmann CF. Squamous epithelium in the endo-
cell carcinoma of the endometrium: potential diag- metrium in benign and malignant conditions. Surg
nostic utility. Am J Surg Pathol. 2014;38(2):189–96. Gynecol Obstet. 1928;46:309–16.
147. Yamamoto S, Tsuda H, Aida S, Shimazaki H, 160. Abeler VM, Kjorstad KE, Nesland
Tamai S, Matsubara O. Immunohistochemical JM. Undifferentiated carcinoma of the endome-
detection of hepatocyte nuclear factor 1beta in trium. A histopathologic and clinical study of 31
ovarian and endometrial clear-cell adenocarcino- cases. Cancer. 1991;68:98–105.
mas and nonneoplastic endometrium. Hum Pathol. 161. Altrabulsi B, Malpica A, Deavers MT, Bodurka
2007;38(7):1074–80. DC, Broaddus R, Silva EG. Undifferentiated car-
148. Hoang LN, Han G, McConechy M, Lau S, Chow cinoma of the endometrium. Am J Surg Pathol.
C, Gilks CB, Huntsman DG, Köbel M, Lee 2005;29(10):1316–21.
CH. Immunohistochemical characterization of pro- 162. Silva EG, Deavers MT, Malpica A. Undifferentiated
totypical endometrial clear cell carcinoma—diag- carcinoma of the endometrium: a review. Pathology.
nostic utility of HNF-1β and oestrogen receptor. 2007;39(1):134–8.
Histopathology. 2014;64(4):585–96. 163. Huntsman DG, Clement PB, Gilks CB, Scully
149. Fadare O, Liang SX. Diagnostic utility of hepatocyte RE. Small-cell carcinoma of the endometrium. A
nuclear factor 1-beta immunoreactivity in endome- clinicopathological study of sixteen cases. Am J
trial carcinomas: lack of specificity for endometrial Surg Pathol. 1994;18:364–75.
clear cell carcinoma. Appl Immunohistochem Mol 164. Garg K, Leitao MM Jr, Kauff ND, Hansen J,
Morphol. 2012;20(6):580–7. Kosarin K, Shia J, Soslow RA. Selection of
150. Fadare O, Zhao C, Khabele D, Parkash V, Quick endometrial carcinomas for DNA mismatch repair
CM, Gwin K, Desouki MM. Comparative analysis of protein immunohistochemistry using patient age
Napsin A, alpha-methylacyl-coenzyme A racemase and tumor morphology enhances detection of
(AMACR, P504S), and hepatocyte nuclear factor 1 mismatch repair abnormalities. Am J Surg Pathol.
beta as diagnostic markers of ovarian clear cell car- 2009;33(6):925–33.
cinoma: an immunohistochemical study of 279 ovar- 165. Olson N, Twiggs L, Sibley R. Small-cell carcinoma
ian tumours. Pathology. 2015;47(2):105–11. of the endometrium: light microscopic and ultra-
151. Simon A, Kopolovic J, Beyth Y. Primary squamous structural study of a case. Cancer. 1982;50:760–5.
cell carcinoma of the endometrium. Gynecol Oncol. 166. Paz RA, Frigerio B, Sundblad AS, Eusebi V. Small-
1988;31:454–61. cell (oat cell) carcinoma of the endometrium. Arch
152. Melin JR, Wanner L, Schulz DM, Cassell Pathol Lab Med. 1985;109:270–2.
EE. Primary squamous cell carcinoma of the endo- 167. Manivel C, Wick MR, Sibley TK. Neuroendocrine
metrium. Obstet Gynecol. 1979;53:115–9. differentiation in mullerian neoplasms. An immu-
153. Goodman A, Zukerberg LR, Rice LW, Fuller nohistochemical study of a “pure” endometrial
AF, Young RH, Scully RE. Squamous cell carci- small-cell carcinoma and a mixed mullerian tumor
noma of the endometrium: a report of eight cases containing small-cell carcinoma. Am J Clin Pathol.
and a review of the literature. Gynecol Oncol. 1986;86:438–43.
1996;61:54–60. 168. Meydanli MM, Erguvan R, Altinok MT, Ataoglu
154. Houissa-Vuong S, Catanzano-Laroudie M, Baviera O, Kafkasli A. Rare case of neuroendocrine small
E, Balaton A, Galet B, Gedeon I, et al. Primary cell carcinoma of the endometrium with paraneo-
squamous cell carcinoma of the endometrium: case plastic membranous glomerulonephritis. Tumori.
history, pathologic findings, and discussion. Diagn 2003;89:213–7.
Cytopathol. 2002;27:291–3. 169. Espinosa I, De Leo A, D’Angelo E, Rosa-Rosa
155. Ryder DE. Verrucous carcinoma of the endome- JM, Corominas M, Gonzalez A, Palacios J, Prat
trium—a unique neoplasm with long survival. J. Dedifferentiated endometrial carcinomas with
Obstet Gynecol. 1982;59:78S–80S. neuroendocrine features: a clinicopathologic, immu-
156. Spiegel GW, Austin RM, Gelven PL. Transitional nohistochemical, and molecular genetic study. Hum
cell carcinoma of the endometrium. Gynecol Oncol. Pathol. 2018;72:100–6.
1996;60:325–30. 170. Romero-Pérez L, López-García MÁ, Díaz-Martín
157. Lininger RA, Ashfaq R, Albores-Saavedra J, J, Biscuola M, Castilla MÁ, Tafe LJ, Garg K, Oliva
Tavassoli FA. Transitional cell carcinoma of the E, Matias-Guiu X, Soslow RA, Palacios J. ZEB1
endometrium and endometrial carcinoma with transi- overexpression associated with E-cadherin and
tional cell differentiation. Cancer. 1997;79:1933–43. microRNA-200 downregulation is characteristic
158. Gungor T, Altinkaya SO, Ozat M, Akbay S, of undifferentiated endometrial carcinoma. Mod
Mollamahmutoglu L. Unusual form of superficial Pathol. 2013;26(11):1514–24.
spreading squamous cell carcinoma of cervix involv- 171. Pesce C, Merino MJ, Chambers JT, Nogales
ing the endometrium, bilateral tubes and ovaries: a F. Endometrial carcinoma with trophoblastic
d ifferentiation. An aggressive form of uterine cancer. 187. Daya D, Lukka H, Clement PB. Primitive neuroecto-
Cancer. 1991;68:1799–802. dermal tumors of the uterus—a report of four cases.
172. Rawish KR, Buza N, Zheng W, Fadare Hum Pathol. 1992;23:1120–9.
O. Endometrial carcinoma with trophoblastic com- 188. Hendrickson MR, Scheithauer BW. Primitive neuro-
ponents: clinicopathologic analysis of a rare entity. ectodermal tumor of the endometrium: report of two
Int J Gynecol Pathol. 2018;37(2):174–90. cases, one with electron microscopic observations.
173. Deligdisch L, Kase NG, Bleiweiss IJ. Endometrial Int J Gynecol Pathol. 1986;5:249–59.
cancer in elderly women: a histologic and steroid 189. Sinkre P, Albores-Saavedra J, Miller DS, Copeland
receptor study. Gerontology. 2000;46:17–21. LJ, Hameed A. Endometrial endometrioid carci-
174. Kalir T, Seijo L, Deligdisch L, Cohen C. Endometrial nomas associated with Ewing sarcoma/peripheral
adenocarcinoma with chorio-carcinomatous differ- primitive neuroectodermal tumor. Int J Gynecol
entiation in an elderly virginal woman. Int J Gynecol Pathol. 2000;19:127–32.
Pathol. 1995;14:266–9. 190. Sorensen JB, Schultze HR, Madsen EL, Holund
175. Savage J, Subby W, Okagaki T. Adenocarcinoma of B. Primitive neuroectodermal tumor (PNET) of the
the endometrium with trophoblastic differentiation uterine cavity. Eur J Obstet Gynecol Reprod Biol.
and metastases as choriocarcinoma: a case report. 1998;76:181–4.
Gynecol Oncol. 1987;26:257–62. 191. Fraggetta F, Magro G, Vasquez E. Primitive neu-
176. Tunc M, Simsek T, Trak B, Uner M. Endometrium roectodermal tumour of the uterus with focal
adenocarcinoma with chorio-carcinomatous dif- cartilaginous differentiation. Histopathology.
ferentiation: a case report. Eur J Gynaecol Oncol. 1997;30:483–5.
1998;19:489–91. 192. Bittencourt AL, Britto JF, Fonseca LEJ. Wilms’
177. Black K, Sykes P, Ostor AG. Trophoblastic differ- tumor of the uterus: the first report of the literature.
entiation in an endometrial carcinoma. Aust N Z J Cancer. 1981;47:2496–9.
Obstet Gynaecol. 1998;38:472–3. 193. Muc RS, Grayson W, Grobbelaar JJ. Adult extrare-
178. Arends J, Willebrand D, Gans HJD, Swaen GJV, nal Wilms tumor occurring in the uterus. Arch Pathol
Bosman FT. Adenocarcinoma of the endometrium Lab Med. 2001;125:1081–3.
with glassy cell features—immunohistochemical 194. Damato S, Haldar K, McCluggage WG. Primary
observations. Histopathology. 1984;8:873–9. endometrial yolk sac tumor with endodermal-
179. Christopherson WM, Alberhasky RC, Connelly intestinal differentiation masquerading as metastatic
PJ. Glassy cell carcinoma of the endometrium. Hum colorectal adenocarcinoma. Int J Gynecol Pathol.
Pathol. 1982;13:418–21. 2016;35(4):316–20.
180. Pitman MB, Young RH, Clement PB, Dickersin 195. Zaloudek CJ, Hendrickson MR, Soslow
GR, Scully RE. Endometrioid carcinoma of the RA. Mesenchymal tumors of the uterus. In: Kurman
ovary and endometrium, oxyphilic cell type: RJ, Ellenson LH, Ronnett BM, editors. Blaustein’s
a report of nine cases. Int J Gynecol Pathol. pathology of the female genital tract. 6th ed.
1994;13:290–301. New York: Springer; 2011. p. 453–527.
181. Mooney EE, Robboy SJ, Hammond CB, 196. Clement PB, Scully RE. Tumors with mixed epi-
Berchuck A, Bentley RC. Signet-ring cell carci- thelial and mesenchymal elements. In: Clement PB,
noma of the endometrium: a primary tumor mas- Young RH, editors. Tumors and tumorlike lesions of
querading as a metastasis. Int J Gynecol Pathol. the uterine corpus and cervix. New York: Churchill
1997;16:169–72. Livingstone; 1993. p. 329–70.
182. Vargas MP, Merino MJ. Lymphoepithelioma-like 197. Sreenan JJ, Hart WR. Carcinosarcomas of the female
carcinoma: an unusual variant of endometrial can- genital tract. A pathologic study of 29 metastatic
cer. A report of two cases. Int J Gynecol Pathol. tumors: further evidence for the dominant role of the
1998;17:272–6. epithelial component and the conversion theory of
183. Jones MA, Young RH, Scully RE. Endometrial ade- histogenesis. Am J Surg Pathol. 1995;19(6):666–74.
nocarcinoma with a component of giant cell carci- 198. Geisinger KR, Dabbs DJ, Marshall RB. Malignant
noma. Int J Gynecol Pathol. 1991;10:260–70. mixed mullerian tumors. An ultrastructural and
184. Joseph MG, Fellows FG, Hearn SA. Primary immunohistochemical analysis with histogenetic
endodermal sinus tumor of the endometrium. A considerations. Cancer. 1987;59:1781–90.
clinicopathologic, immuno-cytochemical, and ultra- 199. Cerilli LA, Wick MR. Immunohistology of soft tis-
structural study. Cancer. 1990;63:297–302. sue and osseous neoplasms. In: Dabbs DJ, editor.
185. Ohta M, Sakakibara K, Mizuno K, Kano T, Diagnostic immunohistochemistry. Philadelphia:
Matsuzawa K, Tomoda Y, et al. Successful treatment Churchill Livingstone; 2002. p. 59–112.
of primary endodermal sinus tumor of the endome- 200. Clement PB. Tumorlike lesions of the uterine cor-
trium. Gynecol Oncol. 1988;31:357–64. pus. In: Clement PB, Young RH, editors. Tumors and
186. Hoshida Y, Nagakawa T, Mano S, Taguchi K, Aozasa tumorlike lesions of the uterine corpus and cervix.
K. Hepatoid adenocarcinoma of the endometrium New York: Churchill Livingstone; 1993. p. 137–79.
associated with alpha–fetoprotein production. Int J 201. Wada H, Enomoto T, Fujita M, Yoshino K,
Gynecol Pathol. 1996;15:266–9. Nakashima R, Kurachi H, et al. Molecular evidence
References 329
that most but not all carcinosarcomas of the uterus are pathological analysis of 10 cases. Am J Surg Pathol.
combination tumors. Cancer Res. 1997;57:5379–85. 1989;13:28–38.
202. Segev Y, Arnon E, Siegler E, Gemer O, Goldberg Y, 217. Clement PB, Scully RE. Mullerian adenosarcomas
Auslender R, Kaldawy A, Lavie O. High incidence of the uterus with sex cord–like elements. Am J Clin
of carcinosarcoma among patients previously treated Pathol. 1989;91:664–72.
with tamoxifen. Isr Med Assoc J. 2017;19(3):164–7. 218. Tavassoli FA, Norris HJ. Mesenchymal tumors
203. McCluggage WG, Abdulkader M, Price JH, Kelehan of the uterus. VII. A clinicopathologic study of
P, Hamilton S, Beattie J, Al-Nafussi A. Uterine 60 endometrial stromal nodules. Histopathology.
carcinosarcomas in patients receiving tamoxi- 1981;5:1–10.
fen. A report of 19 cases. Int J Gynecol Cancer. 219. Chen X, Arend R, Hamele-Bena D, Tergas
2000;10(4):280–4. AI, Hawver M, Tong GX, Wright TC, Wright
204. Kloos I, Delaloge S, Pautier P, Di Palma M, JD. Uterine carcinosarcomas: clinical, histopatho-
Goupil A, Duvillard P, Cailleux PE, Lhomme logic and Immunohistochemical characteristics. Int
C. Tamoxifen-related uterine carcinosarcomas occur J Gynecol Pathol. 2017;36(5):412–9.
under/after prolonged treatment: report of five cases 220. Zelmanowicz A, Hildesheim A, Sherman ME,
and review of the literature. Int J Gynecol Cancer. Sturgeon SR, Kurman RJ, Barrett RJ, et al. Evidence
2002;12(5):496–500. for a common etiology for endometrial carcinomas
205. Seidman JD, Kumar D, Cosin JA, et al. Carcinomas and malignant mixed mullerian tumors. Gynecol
of the female genital tract occurring after pelvic irra- Oncol. 1998;69:253–7.
diation: a report of 15 cases. Int J Gynecol Pathol. 221. Auerbach HE, Livolsi VA, Merino MJ. Malignant
2006;25:293–7. mixed mullerian tumors of the uterus. An immu-
206. Meis JM, Lawrence WD. The immunohistochemical nohistochemical study. Int J Gynecol Pathol.
profile of malignant mixed mullerian tumor. Overlap 1988;7:123–30.
with endometrial adenocarcinoma. Am J Clin Pathol. 222. Dinh TV, Slavin RE, Bhagavan BS, Hannigan EV,
1990;94:1–7. Tiamson EM, Yandell RB. Mixed mullerian tumors
207. Colombi RP. Sarcomatoid carcinomas of the female of the uterus. A clinicopathologic study. Obstet
genital tract (malignant mixed mullerian tumors). Gynecol. 1989;74:388–92.
Semin Diagn Pathol. 1993;10:169–75. 223. Gagne H, Tetu B, Blondeau L, Raymond PE, Blais
208. Pecorelli S. Revised FIGO staging for carcinoma of R. Morphologic prognostic factors of malignant
the vulva, cervix, and endometrium. Int J Gynaecol mixed mullerian tumor of the uterus: a clinicopatho-
Obstet. 2009;105:103–4. logic study of 58 cases. Mod Pathol. 1989;2:433–8.
209. Richmond AM, Blake EA, Torkko K, Smith 224. Young RH, Kleinman GM, Scully RE. Glioma of the
EE, Spillman MA, Post MD. Fascin is associ- uterus. Report of a case with comments on histogen-
ated with aggressive behavior and poor outcome esis. Am J Surg Pathol. 1981;5:695–9.
in uterine carcinosarcoma. Int J Gynecol Cancer. 225. Gersell DJ, Duncan DA, Fulling KH. Malignant
2017;27(9):1895–903. mixed mullerian tumor of the uterus with neuro-
210. Castilla MÁ, Moreno-Bueno G, Romero-Pérez ectodermal differentiation. Int J Gynecol Pathol.
L, Van De Vijver K, Biscuola M, López-García 1989;8:169–78.
MÁ, Prat J, Matías-Guiu X, Cano A, Oliva E, 226. Berton-Rigaud D, Devouassoux-Shisheboran M,
Palacios J. Micro-RNA signature of the epithelial- Ledermann JA, Leitao MM, Powell MA, Poveda
mesenchymal transition in endometrial carcinosar- A, Beale P, Glasspool RM, Creutzberg CL, Harter
coma. J Pathol. 2011;223(1):72–80. P, Kim JW, Reed NS, Ray-Coquard I. Gynecologic
211. Stewart CJ, McCluggage WG. Epithelial- Cancer InterGroup (GCIG) consensus review for
mesenchymal transition in carcinomas of the female uterine and ovarian carcinosarcoma. Int J Gynecol
genital tract. Histopathology. 2013;62:31–43. Cancer. 2014;24(9 Suppl 3):S 55–60.
212. Zaloudek CJ, Norris HJ. Adenofibroma and adeno- 227. Ferguson SE, Tornos C, Hummer A, et al. Prognostic
sarcoma of the uterus. A clinicopathologic study of features of surgical stage I uterine carcinosarcoma.
35 cases. Cancer. 1981;48:354–66. Am J Surg Pathol. 2007;31:1653–61.
213. Kaku T, Silverberg SG, Major FJ, Miller A, Fetter B, 228. Azumi N, Czernobilsky B. Immunohistochemistry.
Brady MF. Adenosarcoma of the uterus: a gyneco- In: Kurman RJ, editor. Blaustein’s pathology of the
logic oncology group clinicopathologic study of 31 female genital tract. 5th ed. New York: Springer;
cases. Int J Gynecol Pathol. 1992;11:75–88. 2002. p. 1251–76.
214. Abell MR. Papillary adenofibroma of the uterine 229. Hart WR, Yoonessi M. Endometrial stromatosis of
cervix. Am J Obstet Gynecol. 1971;110:990–3. the uterus. Obstet Gynecol. 1977;49:393–403.
215. Kerner H, Lichtig C. Mullerian adenosarcoma 230. McCluggage WG. Pathologic staging of endometrial
presenting as cervical polyps—a report of seven carcinomas: selected areas of difficulty. Adv Anat
cases and review of the literature. Obstet Gynecol. Pathol. 2018;25(2):71–84.
1993;81:655–9. 231. Young RH, Clement PB. Endocervical adenocarci-
216. Clement PB. Mullerian adenosarcomas of the noma and its variants: their morphology and differ-
uterus with sarcomatous overgrowth. A clinico- ential diagnosis. Histopathology. 2002;41:185–207.
232. Busca A, Mirkovic J, Parra-Herran C. Gastric-type 247. McCluggage WG, Jenkins D. p16 immunoreactiv-
endocervical adenocarcinoma involving the endo- ity may assist in the distinction between endometrial
metrium and clinically mimicking endometrial neo- and endocervical adenocarcinoma. Int J Gynecol
plasia: a diagnostic challenge. Am J Surg Pathol. Pathol. 2003;22:231–5.
2018;42:983. 248. Ansari-Lari MA, Staebler A, Zaino RJ, et al.
233. Fujiwara M, Longacre TA. Low-grade mucinous Distinction of endocervical and endometrial adeno-
adenocarcinoma of the uterine corpus: a rare and carcinomas: immunohistochemical p16 expression
deceptively bland form of endometrial carcinoma. correlated with human papillomavirus (HPV) DNA
Am J Surg Pathol. 2011;35:537–44. detection. Am J Surg Pathol. 2004;28:160–7.
234. Kurakagi C, Enomoto T, Ueno Y, et al. Mutations 249. Yemelyanova A, Ji H, Shih Ie M, et al. Utility of p16
in the STK11 gene characterize minimal deviation expression for distinction of uterine serous carcino-
adenocarcinoma of the uterine cervix. Lab Investig. mas from endometrial endometrioid and endocervi-
83:35–45. cal adenocarcinomas: immunohistochemical analysis
235. Iaconis L, Hyjek E, Ellenson LH, Pirog EC. p16 and of 201 cases. Am J Surg Pathol. 2009;33:1504–14.
Ki-67 immunostaining in atypical immature squa- 250. Kong CS, Beck AH, Longacre TA. A panel of 3
mous metaplasia of the uterine cervix: correlation markers including p16, ProExC, or HPV ISH is opti-
with human papillomavirus detection. Arch Pathol mal for distinguishing between primary endome-
Lab Med. 2007;131:1343–9. trial and endocervical adenocarcinomas. Am J Surg
236. Khleif SN, DeGregori J, Yee CL, et al. Inhibition of Pathol. 2010;34:915–26.
cyclin D-CDK4/ CDK6 activity is associated with 251. Staebler A, Sherman ME, Zaino RJ, Ronnett
an E2F-mediated induction of cyclin kinase inhibitor BM. Hormone receptor immunohistochemis-
activity. Proc Natl Acad Sci U S A. 1996;93:4350–4. try and human papillomavirus in situ hybridiza-
237. Serrano M. The tumor suppressor protein p16INK4a. tion are useful for distinguishing endocervical and
Exp Cell Res. 1997;237:7–13. endometrial adenocarcinomas. Am J Surg Pathol.
238. Kim WY, Sharpless NE. The regulation of INK4/ 2002;26(8):998–1006.
ARF in cancer and aging. Cell. 2006;127:265–75. 252. Kamoi S, AlJuboury MI, Akin MR, Silverberg
239. Murphy N, Heffron CC, King B, et al. p16INK4A SG. Immunohistochemical staining in the distinc-
positivity in benign, premalignant and malignant tion between primary endometrial and endocervical
cervical glandular lesions: a potential diagnostic adenocarcinomas: another viewpoint. Int J Gynecol
problem. Virchows Arch. 2004;445:610–5. Pathol. 2002;21:217–23.
240. Ostor AG. Natural history of cervical intraepithelial 253. McCluggage WG, Sumathi VP, McBride HA,
neoplasia: a critical review. Int J Gynecol Pathol. Patterson A. A panel of immunohistochemical stains,
1993;12:186–92. including carcinoembryonic antigen, vimentin, and
241. Parwani AV, Smith Sehdev AE, Kurman RJ, Ronnett estrogen receptor, aids the distinction between pri-
BM. Cervical adenoid basal tumors comprised of mary endometrial and endocervical adenocarcino-
adenoid basal epithelioma associated with various mas. Int J Gynecol Pathol. 2002;21:11–5.
types of invasive carcinoma: clinicopathologic fea- 254. Zhang X, Lin Z, Lee ES, Kim I. Immunohistochemical
tures, human papillomavirus DNA detection, and profiles of endometrial and endocervical adenocarci-
P16 expression. Hum Pathol. 2005;36:82–90. nomas. Mod Pathol. 2003;16:217A.
242. Sano T, Oyama T, Kashiwabara K, Fukuda T, 255. Ansari-Lari MA, Staebler A, Ronnett BM. Distinction
Nakajima T. Expression status of p16 protein is of endocervical and endometrial adenocarcinomas:
associated with human papillomavirus oncogenic p16 expression correlated with human papillomavi-
potential in cervical and genital lesions. Am J Pathol. rus (HPV) DNA detection by in situ hybridization
1998;153:1741–8. (ISH). Mod Pathol. 2003;16:180A.
243. Keating JT, Ince T, Crum CP. Surrogate biomarkers 256. Castrillon DH, Lee KR, Nucci MR. Distinction
of HPV infection in cervical neoplasia screening and between endometrial and endocervical adenocarci-
diagnosis. Adv Anat Pathol. 2001;8:83–92. noma: an immunohistochemical study. Int J Gynecol
244. Doorbar J. Molecular biology of human papilloma- Pathol. 2002;21:4–10.
virus infection and cervical cancer. Clin Sci (Lond). 257. Kumar A, Schneider V. Metastases to the uterus from
2006;110:525–41. extrapelvic primary tumors. Int J Gynecol Pathol.
245. Durst M, Glitz D, Schneider A, zur Hausen 1983;2:134–40.
H. Human papillomavirus type 16 (HPV 16) gene 258. Kumar NB, Hart WR. Metastases to the uterine cor-
expression and DNA replication in cervical neo- pus from extagenital cances. A clinicopathologic
plasia: analysis by in situ hybridization. Virology. study of 63 cases. Cancer. 1982;50:2163–9.
1992;189:132–40. 259. Mazur MT, Hsueh S, Gersell DJ. Metastases to the
246. Stoler MH, Rhodes CR, Whitbeck A, Wolinsky SM, female genital tract. Analysis of 325 cases. Cancer.
Chow LT, Broker TR. Human papillomavirus type 1984;53:1978–84.
16 and 18 gene expression in cervical neoplasias. 260. Ulbright TM, Roth LM. Metastatic and indepen-
Hum Pathol. 1992;23:117–28. dent cancers of the endometrium and ovary: a
References 331
clinicopathologic study of 34 cases. Hum Pathol. 275. Rorat E, Wallach RC. Papillary metaplasia of the
1985;16:28–34. endometrium: clinical and histopathologic consider-
261. Eifel P, Hendrickson M, Ross J, Ballon S, Martinez ations. Obstet Gynecol. 1984;64:90S–2S.
A, Kempson RL. Simultaneous presentation of 276. Sherman ME, Mazur MT, Kurman RJ. Benign dis-
carcinoma involving the ovary and uterine corpus. eases of the endometrium. In: Kurman RJ, editor.
Cancer. 1982;50:163–70. Blaustein’s pathology of the female genital tract. 5th
262. Schultheis AM, Ng CK, De Filippo MR, Piscuoglio ed. New York: Springer; 2002. p. 421–66.
S, Macedo GS, Gatius S, Perez Mies B, Soslow 277. Mazur MT. Atypical polypoid adenomyomas of the
RA, Lim RS, Viale A, Huberman KH, Palacios JC, endometrium. Am J Surg Pathol. 1981;5:473–82.
Reis-Filho JS, Matias-Guiu X, Weigelt B. Massively 278. Young RH, Treger T, Scully RE. Atypical polypoid
parallel sequencing-based clonality analysis of syn- adenomyoma of the uterus. A report of 27 cases. Am
chronous endometrioid endometrial and ovarian car- J Clin Pathol. 1986;86:139–45.
cinomas. J Natl Cancer Inst. 2016;108(6):djv427. 279. Iezzoni JC, Mills SE. Nonneoplastic endometrial
263. Anglesio MS, Wang YK, Maassen M, Horlings signet-ring cells. Vacuolated decidual cells and stro-
HM, Bashashati A, Senz J, Mackenzie R, Grewal mal histiocytes mimicking adenocarcinoma. Am J
DS, Li-Chang H, Karnezis AN, Sheffield BS, Clin Pathol. 2001;115:249–55.
McConechy MK, Kommoss F, Taran FA, Staebler 280. Fukunaga M, Iwaki S. Nodular histiocytic hyper-
A, Shah SP, Wallwiener D, Brucker S, Gilks CB, plasia of the endometrium. Arch Pathol Lab Med.
Kommoss S, Huntsman DG. Synchronous endome- 2004;128(9):1032–4.
trial and ovarian carcinomas: evidence of Clonality. 281. Chebib I, Chu P, Duggan MA, DiFrancesco
J Natl Cancer Inst. 2016;108(6):djv428. LM. Primary signet-ring cell adenocarcinoma of the
264. Dabbs DJ, Geisinger KR, Norris HT. Intermediate endometrium: case report and review of the litera-
filaments in endometrial and endocervical carcino- ture. Int J Gynecol Pathol. 2010;29(3):269–72.
mas. The diagnostic utility of vimentin patterns. Am 282. Boyd C, Cameron I, McCluggage WG. Endometrial
J Surg Pathol. 1986;10:568–76. adenocarcinoma with signet ring cells: report of
265. Gown AM. Immunohistochemical determination two cases of an extremely rare phenomenon. Int J
of primary sites of carcinomas. J Histotechnol. Gynecol Pathol. 2010;29(6):579–82.
1999;22:209–15. 283. Pusiol T. HPV11 positive endometrioid carcinoma
266. Ni YB, Tsang JYS, Shao MM, Chan SK, Cheung of the endometrium with signet-ring cells: diagnos-
SY, Tong J, To KF, Tse GM. GATA-3 is superior to tic criteria and review of the literature. Case Rep
GCDFP-15 and mammaglobin to identify primary Pathol. 2014;2014:1.
and metastatic breast cancer. Breast Cancer Res 284. Randall TC, Kurman RJ. Progestin treatment of
Treat. 2018;169(1):25–32. atypical hyperplasia and well–differentiated carci-
267. Winkler B, Alvarez S, Richart RM, Crum CP. Pitfalls noma of the endometrium in women under age 40.
in the diagnosis of endometrial neoplasia. Obstet Obstet Gynecol. 1997;90:434–40.
Gynecol. 1984;64:185–94. 285. Fadare O, Parkash V, Dupont WD, Acs G, Atkins
268. Crum CP, Richart RM, Fenoglio CM. Adenoacanthosis KA, Irving JA, Pirog EC, Quade BJ, Quddus MR,
of the endometrium. A clinicopathologic study in pre- Rabban JT 3rd, Vang R, Hecht JL. The diagnosis of
menopausal women. Am J Surg Pathol. 1981;5:15. endometrial carcinomas with clear cells by gyneco-
269. Hendrickson MR, Kempson RL. Endometrial epithelial logic pathologists: an assessment of interobserver
metaplasias: proliferations frequently misdiagnosed as variability and associated morphologic features. Am
adenocarcinoma. Report of 89 cases and proposed clas- J Surg Pathol. 2012;36(8):1107–18.
sification. Am J Surg Pathol. 1980;4:525–42. 286. Silva EG, Young RH. Endometrioid neoplasms
270. Blaustein A. Morular metaplasia misdiagnosed as with clear cells: a report of 21 cases in which the
adenoacanthoma in young women with polycystic alteration is not of typical secretory type. Am J Surg
ovarian disease. Am J Surg Pathol. 1982;6:223. Pathol. 2007;31(8):1203–8.
271. Andersen WA, Taylor PT, Fechner RE, Pinkerton 287. Norris HJ, Taylor HB. Mesenchymal tumors of
JAV. Endometrial metaplasia associated with the uterus. I. A clinical and pathological study
endometrial carcinoma. Am J Obstet Gynecol. of 53 endometrial stromal tumors. Cancer.
1987;157:597–604. 1966;19:755–66.
272. Clement PB. Pathology of the uterine corpus. Hum 288. Yoonessi M, Hart WR. Endometrial stromal sarco-
Pathol. 1991;22:776–91. mas. Cancer. 1977;40:898–906.
273. Zaman SS, Mazur MT. Endometrial papillary syn- 289. Mhawech-Fauceglia P, Yan L, Liu S, Pejovic T. ER+
cytial change. A nonspecific alteration associated /PR+ /TFF3+ /IMP3- immunoprofile distinguishes
with active breakdown. Am J Clin Pathol. 1993; endometrioid from serous and clear cell carcino-
99:741–5. mas of the endometrium: a study of 401 cases.
274. Kaku T, Tsukamoto N, Tsuruchi N, Sugihara K, Histopathology. 2013;62(7):976–85.
Kamura T, Nakano H. Endometrial metaplasia asso- 290. Allo G, Bernardini MQ, Wu RC, Shih IM, Kalloger
ciated with endometrial carcinoma. Obstet Gynecol. S, Pollett A, Gilks CB, Clarke BA. ARID1A loss
1992;80:812–6. correlates with mismatch repair deficiency and intact
p53 expression in high-grade endometrial carcinomas of the endocervix. Am J Surg Pathol. 2018;
mas. Mod Pathol. 2014;27(2):255–61. 42(2):214–26.
291. Stolnicu S, Barsan I, Hoang L, Patel P, Terinte C, 292. Jones MW, Onisko A, Dabbs DJ, Elishaev E, Chiosea
Pesci A, Aviel-Ronen S, Kiyokawa T, Alvarado- S, Bhargava R. Immunohistochemistry and HPV in
Cabrero I, Pike MC, Oliva E, Park KJ, Soslow situ hybridization in pathologic distinction between
RA. International endocervical adenocarcinoma endocervical and endometrial adenocarcinoma: a
criteria and classification (IECC): A new patho- comparative tissue microarray study of 76 tumors.
genetic classification for invasive adenocarcino- Int J Gynecol Cancer. 2013;23(2):380–4.
Mesenchymal Tumors and Other
Rare Neoplasms 11
Contents
Smooth Muscle Tumors 333
Miscellaneous Mesenchymal Tumors 339
Stromal Tumors 340
Mixed Epithelial and Mesenchymal Tumors 347
Rare Neoplasms 351
Other Lesions and Tumor-Like Conditions 353
References 354
Uterine tumors other than benign polyps or carci- Smooth Muscle Tumors
noma are rarely encountered in endometrial biop-
sies and curettings. Even submucosal leiomyomas Leiomyomas
rarely are sampled by an endometrial biopsy or
curettage. The classification in Table 11.1 lists Smooth muscle tumors are rarely encountered in
most of these other neoplasms. It has been endometrial biopsy and curettage specimens and
reported that prior to a hysterectomy, the diagno- more frequently removed via hysteroscopic resec-
sis of endometrial stromal sarcoma is missed in tion. Tissue from a submucosal leiomyoma or a
20% of cases [1]. Similarly for leiomyosarcoma, pedunculated intracavitary leiomyoma is usually
inaccurate or inconclusive diagnosis on curet- obtained rather than intramural leiomyomas.
tings ranges from 20% to 80% [1–4]. Although Those sampled usually are benign leiomyo-
these tumors have morphologic features that mas. These specimens show fragments of smooth
readily assist in their recognition, it is often muscle, although the tissue usually is too frag-
impossible to classify them accurately from mented to allow definitive diagnosis of a leiomy-
biopsy material alone. oma. Leiomyomas usually affect women in their

334 11 Mesenchymal Tumors and Other Rare Neoplasms
fourth and fifth decade, while patients with heredi- effect (see Chap. 6, Effects of Hormones). Often
tary leiomyomatosis and renal cancer syndrome the endometrial tissue from biopsy material shows
(HLRCC) occur at a mean age of 30. In premeno- either no demonstrable effect from the leiomyo-
pausal women, leiomyomas can affect endome- mas or nonspecific glandular and stromal break-
trial development. Endometrium overlying a down, and in these cases, it is not possible to
leiomyoma may show focal hypoplasia or atrophy, diagnose leiomyomas from a biopsy or curettage
being thin with sparse glands that are underdevel- specimen.
oped and small compared to surrounding tissue
(Fig. 11.1) [5]. Endometrium adjacent to leiomyo-
mas also can become distorted [5]. The glands Variants of Leiomyoma
may lose their perpendicular orientation relative to
the surface epithelium. These focal distortions of There are a number of variants of benign leio-
endometrial growth can yield a pattern of irregular myoma characterized either by increased mitotic
maturation, with fragments of normally developed activity or by deviation from the typical pattern
proliferative or secretory endometrium adjacent to of interlacing spindle cells, including mitotically
fragments with irregular or poorly developed active leiomyoma, cellular leiomyoma, leiomy-
glands. Such morphologic abnormalities of the oma with bizarre nuclei, hydropic leiomyoma,
glands generally are nonspecific, however, and are lipoleiomyoma, myxoid leiomyoma, cotyle-
not sufficient to establish the diagnosis of a sub- donoid dissecting leiomyoma, leiomyomatosis
mucosal leiomyoma. Patients are often treated and renal cancer syndrome, leiomyoma with apo-
with gonadotropin-releasing hormone analogues plectic change, epithelioid leiomyoma, and
(GnRHa) or selective progesterone receptor mod- mitotically active leiomyoma [7–27].
ulators (SPRMs) prior to hysterectomy or hystero- Of the leiomyoma variants listed above,
scopic myomectomy for uterine fibroids [6]. This hereditary leiomyomatosis and renal cell carci-
may result in endometrial changes related to drug noma (HLRCC) syndrome is an autosomal domi-
nant disorder associated with a fumarate
hydratase (FH) germline mutation. Therefore,
Table 11.1 Uterine tumors other than carcinoma that diagnosis of a FH-deficient leiomyoma may be a
may be found in biopsies and curettage specimens
sentinel disease manifestation and lead to early
Smooth muscle tumors diagnosis of potentially lethal renal cell carcino-
Leiomyoma
mas. Leiomyomas can also harbor sporadic
Variants of leiomyoma
mutations leading to loss of FH expression, and
Smooth muscle tumor of uncertain malignant
potential (STUMP) not part of the HLRCC syndrome. Leiomyomas
Leiomyosarcoma encountered as part of HLRCC syndrome fre-
Tumorlets quently have increased cellularity and atypical
Miscellaneous mesenchymal tumors nuclei with prominent red macronucleoli sur-
PEComa rounded by a clear halo [27] (Fig. 11.2). One
Endometrial stromal and related tumors study demonstrated morphologic criteria per-
Stromal nodule
formed poorly in identifying HLRCC-associated
Low-grade endometrial stromal sarcoma
leiomyomas, while others demonstrated these
High-grade endometrial stromal sarcoma
Undifferentiated endometrial sarcoma
tumors are best diagnosed by morphologic fea-
Uterine tumors resembling ovarian sex cord tumor tures confirmed by loss of immunohistochemical
Mixed epithelial-stromal tumors staining for FH [28, 29]. The latter group demon-
Adenofibroma strated leiomyomas deficient for FH are rare in
Adenosarcoma routine material (0.4% of cases). They proposed
Other neoplasms targeted FH immunostaining guided by histo-
Inflammatory myofibroblastic tumor morphologic features rather than sporadic FH
Lymphoma/leukemia immunostaining [29]. Loss of immunohisto-
Fig. 11.1 Effect of leiomyoma. A fragment of markedly underlying smooth muscle, is suggestive of a leiomyoma
thinned inactive endometrium overlies smooth muscle. but requires clinical correlation. In this case, subsequent
Other fragments showed a secretory phase pattern. This hysterectomy showed submucosal leiomyomas
pattern of focally hypoplastic endometrium, even without
chemical staining for FH does not definitively have a higher mitotic rate or tumor cell necrosis
diagnose HLRCC as sporadic mutations can also in the portion that is not removed or that more
lead to loss of FH expression [30, 31]. marked nuclear atypia is not present elsewhere.
Leiomyomas that can suggest a malignant Also, other features, such as size of the lesion and
spindle cell process include bizarre, cellular, and interface of the tumor with normal myometrium,
epithelioid leiomyomas, and these have mitotic cannot be assessed. The differential diagnosis of
rates of 4 or fewer per 10 HPFs [9, 10, 15, 17]. stromal tumor and cellular leiomyoma is dis-
Generally the mitotic rate is very low, averaging cussed in the section on stromal tumors.
less than 1 mitotic figure per 10 HPFs. Mitotically
active leiomyomas have high mitotic rates, up to
15 mitoses per 10 HPFs, but lack cytologic atypia mooth Muscle Tumor of Uncertain
S
and tumor cell necrosis [13, 14, 18, 21]. These Malignant Potential
tumors often are submucosal [13]. In the rare
event that any of these lesions are sampled by Smooth muscle tumor of uncertain malignant
curette, they are fragmented and only partially potential (STUMP) is a tumor that falls short of
removed (Fig. 11.3). Sampling is therefore inad- an unequivocal diagnosis of leiomyosarcoma but
equate and precludes an accurate diagnosis. It is has at least one possible concerning feature of
impossible to be certain that the tumor would not leiomyosarcoma and therefore exceeds the diag-
Fig. 11.2 Fumarate hydratase-deficient leiomyoma. cleoli surrounded by a clear halo (center). This finding is
High magnification of a fumarate hydratase-deficient leio- suggestive of, but not specific for, fumarate hydratase-
myoma shows atypical nuclei with prominent macronu- deficient leiomyomas
nosis of leiomyoma. Examples of this include Leiomyosarcoma

increased mitotic activity, but less than that found
in leiomyosarcomas, cytologic atypia, uncertain Leiomyosarcoma typically occurs in patients
type of necrosis, and epithelioid or myxoid older than 40 years of age with a mean age of
change [32]. Most smooth muscle tumors can be 50–55 [34]. Usually the neoplasm is confined to
placed in one of the specific categories. Recent the myometrium, but it may involve the endome-
studies of STUMPs have shown that they gener- trial cavity and be sampled in curettings or
ally have a relatively low risk of aggressive hysteroscopic myomectomies. In a retrospective
behavior, and therefore we prefer the designation review, 40% of leiomyosarcomas were identified
“smooth muscle tumor of low malignant poten- on preoperative fractional curettage or endome-
tial.” Accordingly, the diagnosis of STUMP, trial biopsies [35]. In another retrospective study
which provides little meaningful information for of leiomyosarcoma patients, endometrial sam-
the gynecologist, should be exceedingly rare. pling was significantly more likely to detect
Leiomyoma variants should not be considered malignancy in postmenopausal rather than pre-
STUMPs. The diagnosis of smooth muscle menopausal women with abnormal bleeding [4].
tumors is made on H&E stained slides. There are The frequency of an unexpected uterine malig-
insufficient data on immunohistochemical stain- nancy on hysteroscopic myomectomy was higher
ing of a variety of markers, such as ATRX, (0.86%) compared to hysterectomy (0.19%) [36].
MED12, and DAXX to merit their use [33]. Leiomyosarcomas show features of high-grade
Fig. 11.3 Portion of bizarre leiomyoma in curettings. A the nuclear atypia, the smooth muscle lacked mitotic
fragment of smooth muscle with moderate nuclear pleo- activity. Nonetheless, hysterectomy is necessary to
morphism underlies a fragment of endometrium. Despite exclude a leiomyosarcoma
sarcoma, with closely spaced hyperchromatic Immunohistochemical evaluation of leiomyo-

nuclei showing pleomorphism, a high mitotic sarcoma shows diffuse cytoplasmic reactivity for
rate, and abnormal mitotic figures [7–9, 37]. desmin, smooth muscle actin, and caldesmon.
Coagulative tumor cell necrosis usually is present Calponin and smooth muscle myosin are occa-
(Fig. 11.4) and is a powerful predictor of clinical sionally used as well. Leiomyosarcoma will
behavior [21]. The mitotic rate should be greater occasionally stain with CD10 [40]. Because of
than 5 per 10 HPFs and often is well over 10 per this overlap of immunophenotype with stromal
10 HPFs; the nuclear atypia and coagulative sarcoma, the diagnosis is based primarily on
tumor cell necrosis combined with the mitotic morphologic analysis.
index are used in making the diagnosis. The cells
grow in interlacing fascicles with abundant
eosinophilic fibrillary cytoplasm and elongate Tumorlets
nuclei with blunt ends. Epithelioid variants occur
in which the cells are polygonal rather than spin- Small, circumscribed mesenchymal prolifera-
dle shaped [9]. In contrast to the usual leiomyo- tions called tumorlets occasionally may be found
sarcoma, myxoid leiomyosarcoma has abundant in endometrial curettage specimens. One such
myxoid stroma, relatively bland cytologic fea- lesion is the plexiform tumorlet, a rare small epi-
tures, and infrequent mitoses [38, 39]. thelioid leiomyoma that usually occurs in the
Fig. 11.4 Leiomyosarcoma. Low magnification of a hysterectomy specimen shows coagulative tumor cell necrosis
with an abrupt transition to viable tumor
myometrium but that can involve the endome- Tumorlets, unlike other leiomyoma variants,
trium [41–43]. Often these are multifocal. In the are sufficiently small that they can be confidently
endometrium they typically occur near the recognized in curettage specimens. When they
endometrial-myometrial interface and show tran- occur in the endometrium, they often are com-
sitions to normal endometrial stroma. These pletely removed by curettage. They have no known
lesions form circumscribed microscopic nodules clinical significance. They should be recognized to
and are composed of small, oval to polygonal prevent misclassification of a more serious lesion.
cells in an organoid arrangement, typically
forming small anastomosing trabeculae or cords
with an intervening hyaline matrix. Clinical Queries and Reporting
In another pattern of tumorlet or small leiomy-
oma that may be seen in curettage specimens, the Smooth muscle tumors other than ordinary leio-
cells have a more spindled appearance. In these myomas are so unusual in endometrial biopsies
small proliferations, the cells interlace in a less that they require a comment regarding their
well-defined plexiform pattern, and although they appearance and possible malignant potential. The
are circumscribed, the cells at the periphery blend sensitivity of detecting a leiomyosarcoma on
into surrounding endometrial stroma (Fig. 11.5). biopsy is low and better for postmenopausal
Their intimate relationship to endometrial stroma symptomatic women [4]. Unless the tumor is
is another example of how stromal and smooth obviously malignant, it is best to describe the
muscle cells can coexist with transitional cells that lesion and defer definitive diagnosis based on the
bridge the morphology of these two cell types. hysterectomy specimen.
Miscellaneous Mesenchymal Tumors 339
Fig. 11.5 Tumorlets. These are typically small, circum- bly into adjacent endometrial stroma (right). Tumorlets are
scribed nodules composed of an organoid, plexiform-like invariably benign and of no clinical significance (courtesy
arrangement of ovoid cells. The lesion blends impercepti- of Dr. Margaret Shaffer, Sibley Memorial Hospital)
Leiomyosarcoma usually is readily recog- TSC2 genes is seen in TSC [45], and less than
nized. Often precise classification of the sarcoma 10% of uterine PEComas have been associated
may be difficult from a biopsy specimen, but in with the tuberous sclerosis complex [44, 46–48].
general, recognition of high-grade sarcoma is Histologically, there are two forms. Group A
sufficient for clinical management. A brief micro- PEComas, which are common, have a tonguelike
scopic description with a mitotic count is growth pattern composed of nests of clear cells
helpful. partially encircled by delicate collagenous stroma
or thin-walled vessels. Group B PEComas are
predominantly epithelioid to spindled cells with
Miscellaneous Mesenchymal clear, eosinophilic to granular cytoplasm
Tumors (Figs. 11.6 and 11.7). The nuclei are oval to round
with small nucleoli. Mitotic activity is generally
Perivascular Epithelioid Cell Tumor low, and necrosis may be present. Both groups
may have significant atypia. PEComas typically
Perivascular epithelioid cell tumor (PEComa) of express HMB-45, melan-A, and MITF immuno-
the uterus most commonly occurs in perimeno- histochemical markers [48, 49]. Criteria for
pausal women [44]. The PEComa family of malignant PEComa are difficult to establish
tumors has been demonstrated in tuberous sclero- because of limited numbers of cases. Features
sis complex (TSC). Inactivation of the TSC1 and used in assessment of malignancy have included
Fig. 11.6 Perivascular epithelioid cell tumor. Low magnification demonstrates a perivascular pattern of epithelioid
tumor cells surrounding thin-walled vessels
mitotic activity greater than 1 per 50 HPFs, tumor may sample them. Low-grade endometrial stro-
size 5 cm or greater, lymphovascular invasion, mal sarcoma recurs in up to one half of cases,
high-grade nuclei, and necrosis. Using the crite- often in the pelvis or abdomen following hyster-
ria of Schoolmeester et al., the presence of four ectomy [53–56, 60–62]. Relapses may occur
or more features was associated with higher posi- many years after hysterectomy. This lesion in the
tive predictive value and specificity, but some past was termed “endolymphatic stromal myo-
tumors (15%) classified as benign behaved in a sis” or “stromatosis” [53, 60, 63, 64]. High-grade
malignant fashion [44, 50]. Further studies are endometrial stromal sarcomas are aggressive
clearly needed. neoplasms that spread widely [8, 51, 54, 55, 57].
Most patients with high-grade endometrial stro-
mal sarcoma succumb within 3 years.
Stromal Tumors Stromal tumors occur over a wide age range,
with a mean age in the fifth decade in most stud-
Tumors in this category include the benign endo- ies. The majority of low-grade endometrial stro-
metrial stromal nodule (ESN), low-grade endo- mal sarcomas occur in women under the age of
metrial stromal sarcoma (LGESS), high-grade 50 years [60], whereas high-grade endome-
endometrial stromal sarcoma (HGESS), and trial stromal sarcoma generally occurs in older
undifferentiated uterine sarcoma (USS) [7–9, 32, women [51, 57]. These tumors frequently present
51–59]. These neoplasms frequently involve the with abnormal bleeding, although they may pres-
endometrium, so endometrial biopsy or curettage ent with pelvic pain, with other, nonspecific com-
Stromal Tumors 341
Fig. 11.7 Perivascular epithelioid cell tumor. Relatively monotonous epithelioid cells with clear to granular cytoplasm.
Some nuclei have small, prominent nucleoli
plaints, or even with distant metastases to the in stromal tumors [66, 67]. The smooth muscle
lung or other sites [53–56]. Some cases are not differentiation may result in a “starburst” pattern
detected by curettage and are diagnosed only in of collagen deposition (Fig. 11.9) [68]. Small
hysterectomy specimens [51]. amounts of smooth muscle may be interspersed
between areas with typical stromal differentiation,
but this finding does not alter the diagnosis [40,
Endometrial Stromal Nodule 68]. The mitotic rate is variable, but most tumors
and Low-Grade Endometrial Stromal show no more than 3 mitotic figures per 10 HPFs.
Sarcoma Some tumors can show high mitotic counts, how-
ever, and mitotic count is not a criterion to distin-
The endometrial stromal nodule (ESN) and low- guish stromal nodules, low- grade endometrial
grade endometrial stromal sarcoma (LGESS) are stromal sarcoma, or high-grade endometrial stro-
composed of cells that resemble endometrial mal sarcoma (HGESS) [52]. Endometrial stromal
stroma of the normal proliferative phase [8, 9, 51, nodules and LGESS do not show coagulative-type
52, 54]. Cytologically these lesions are identical necrosis or cellular pleomorphism, but cyst forma-
and are distinguished by their growth patterns (see tion and ischemic-type necrosis may be present
later). The cells are oval and uniform and are sup- [69]. Occasional cases show small aggregates of
ported by a prominent network of small vascular admixed foam cells [70].
spaces (Fig. 11.8). Smooth muscle differentiation Immunohistochemical stains show reactivity
[65] and skeletal muscle differentiation can occur for CD10, WT-1, smooth muscle actin, ER, and
Fig. 11.8 Low-grade endometrial stromal sarcoma. The endometrium in a background of multiple small, thick-walled
tumor is composed of a uniform population of oval mesenchy- blood vessels. The circumferential orientation of the stromal
mal cells that resemble the stromal cells of proliferative phase cells around the blood vessels is a characteristic feature
PR (some tumors may be negative for CD10) [40, central lumen. Immunohistochemical stains for
71–74]. Desmin is positive in areas with smooth actin and desmin show reactivity in the sex-cord-
muscle differentiation and myxoid or fibrous fea- like areas that resemble that found in the spindle
tures [40, 75–79]. These tumors can show focal cell component [77, 79, 90]. Sex-cord-like areas
reactivity for keratin [80] although they generally also may be immunoreactive for keratin [79].
are keratin negative [40]. Stromal tumors demon- Endometrial stromal sarcoma also can show
strating sex-cord-like elements show a variety of smooth and skeletal muscle, focal bizarre nuclei,
expression of inhibin, calretinin, CD99, WT-1, fibrous change, or a component of epithelioid- or
and melan-A [81–87]. Nuclear β-catenin staining rhabdoid-appearing cells with abundant eosino-
has been demonstrated, but no associated muta- philic cytoplasm [67, 81, 91].
tions have been reported [88]. Stromal tumors can also show scattered foci
Endometrial stromal sarcomas may show sex- of benign endometrial glands and may even show
cord-like patterns with nests and trabeculae of extensive endometrial glandular differentiation,
condensed cells forming distinctive epithelial- but the presence of glands does not alter the diag-
like patterns that resemble ovarian sex cord nosis [8, 92]. Tumors with this finding should not
tumors [9, 54, 68, 79, 89, 90]. These epithelioid be classified as carcinosarcomas, as the glands
nests often form anastomosing trabeculae that are benign. Stromal tumors with glandular differ-
yield a plexiform pattern. The nests and cords entiation are distinguished from adenosarcomas
may be solid, or they may form tubules with a by their growth pattern, their lack of leaflike pol-
Stromal Tumors 343
Fig. 11.9 Low-grade endometrial stromal sarcoma. The tion results in a “starburst” pattern of collagen deposition
tumor shows a monotonous population of cells with small (upper left). Mitoses are rare in this field, but can be
nuclei and scant cytoplasm. Smooth muscle differentia- numerous in some cases
ypoid fronds, and the presence of a prominent tion (FISH), cytogenetics, or real-time poly-
vascular framework in the tumor. They resemble merase chain reaction (RT-PCR).
endometrial stromal sarcoma except for the glan- The endometrial stromal nodule is distin-
dular differentiation. guished from the LGESS by the growth pattern, a
Endometrial stromal nodules and LGESS feature that usually cannot be appreciated in
share the same t(7;17) genetic aberration; there- curettage specimens [8, 9, 52, 54, 69]. The pres-
fore the presence of t(7;17) does not distinguish ence of fragments of endometrial-type stroma
between the two entities. Most harbor t(7;17) devoid of glands, specifically glandular fragments
resulting in a fusion between JAZF1 and 5 mm or more, can be a clue to the diagnosis of a
SUZ12(JJAZ1) [93–98]. Several endometrial stromal lesion in a biopsy or curettage specimen
stromal tumor morphologies can be seen with [101]. Endometrial stromal nodule is a circum-
this aberration, including sex-cord-like, smooth scribed tumor ranging from less than 1 to 25 cm
muscle, fibromyxoid change, benign epithelioid with a rounded, pushing interface with the myo-
cells, and conventional morphology of endome- metrium or endometrium. An ESN should have
trial stromal sarcoma [93, 94, 96, 98, 99]. A vari- absent or only minimal myometrial invasion of
ant translocation in t(6;7)(p21;p15) is the second less than 3 mm and less than three protrusions
most frequent aberration resulting in the JAZF1- into the myometrium with no vascular invasion
PHF1 fusion gene [98, 100]. These translocations [52, 69]. Low-grade endometrial stromal sar-
can be detected by fluorescence in situ hybridiza- coma, in contrast, has an infiltrative pattern of
growth, showing irregular invasion of the myo- are tumors with obvious malignant characteris-
metrium (Fig. 11.10) often accompanied by tics, including increased cellularity, pleomor-
endovascular growth, which results in a plug of phism, nuclear atypia, and a high mitotic rate [8,
tumor resulting in a “wormlike” growth pattern. 51, 55, 57].
YWHAE-FAM22 (YWHAE-NUTM2)
igh-Grade Endometrial Stromal
H High-Grade Endometrial Stromal
Sarcoma Sarcomas
This group of HGESS is seen across a wide age
There are two well-defined categories of high- range from 25 to 70 years, and disease may have
grade endometrial stromal sarcomas (HGESS). extended beyond the uterus at the time of presen-
The first is the YWHAE-FAM22 (YWHAE- tation [102]. On microscopic examination, the
NUTM2) high-grade endometrial stromal sar- tumor has an infiltrative growth pattern into myo-
coma, and the second is the ZC3H7B-BCOR metrium and veins. There is typically a diffuse or
high-grade endometrial stromal sarcoma; the lat- vague nested pattern with a delicate capillary net-
ter was described after the 2014 WHO publica- work. The cells are round with irregular nuclear
tion. High-grade endometrial stromal sarcomas contours, large nucleoli, and brisk mitotic activ-
Fig. 11.10 Low-grade endometrial stromal sarcoma. sis of low-grade endometrial stromal sarcoma requires
Low magnification of low-grade endometrial stromal sar- identification of infiltrative or endovascular tumor growth,
coma in a hysterectomy specimen demonstrating a features that usually cannot be assessed in curettage
tonguelike invasive pattern into the myometrium by vari- specimens
ably sized and shaped nests of tumor. A definitive diagno-
Stromal Tumors 345
ity. Atypical mitotic figures and foci of tumor cell subtype is typically negative for CD10 while dif-
necrosis are usually present [102]. An occasional fuse and strong in the ZC3H7B-BCOR tumors
tumor shows some resemblance to low-grade [110–112]. Ancillary testing with FISH or PCR
endometrial stromal sarcoma, with a population can detect the ZC3H7B-BCOR fusions that result
of more uniform ovoid cells but with a high from the t(X;22)(p11q13) rearrangement. Internal
mitotic rate and necrosis. tandem duplications involving exon 15 of BCOR
The HGESS component is typically negative have also been demonstrated [110–112].
for CD10, ER, and PR, while staining for cyclin
D1 can show strong and diffuse positivity (>70%
of cell nuclei); kit and BCOR are frequently posi- Undifferentiated Uterine Sarcoma
tive as well [103–105]. This tumor has a charac-
teristic gene abnormality t(10;17)(q22;p13) Undifferentiated uterine sarcoma (USS) is a rare
associated with a YWHAE-FAM22 (NUTM2AB) tumor in postmenopausal women and displays
fusion and can be associated with a low-grade high-grade cytologic features and no specific
fibromyxoid component [102, 106–109]. type of differentiation. These tumors are difficult
to classify and therefore a diagnosis of exclusion.
C3H7B-BCOR High-Grade Endometrial
Z This tumor can demonstrate destructive myome-
Stromal Sarcomas trial growth, a patternless or fascicular growth
Similar to the YWHAE-FAM22 (NUTM2AB) pattern, and extensive pleomorphism. Necrosis
high-grade endometrial stromal sarcomas, the and lymphovascular invasion are common [74].
ZC3H7B-BCOR group also occurs over a wide The tumors have variable CD10, p53, and cyclin
age range (28–71 years), and extrauterine disease D1 expression, but are usually negative for ER
may be present at the time of diagnosis [110– and PR. Those tumors with a more uniform
112]. One subgroup with BCOR internal tandem appearance had morphologic and immunopheno-
duplications (BCOR-ITD) occurs in younger typic overlap with HGESS, and many, but not all,
patients (18–32 years) [111]. were subsequently shown to have t(10;17) and
Microscopically, this subtype of tumor closely therefore reclassified as HGESS [113]. Tumors in
resembles myxoid leiomyosarcoma. Invasion pat- the pleomorphic category were typically negative
terns vary between tonguelike, broad, and/or for hormone receptors and β-catenin, but often
destructive. The background stroma is either myx- expressed p53. This group likely represents true
oid or collagenous. Sheets and fascicles of densely undifferentiated sarcomas.
packed, ovoid to spindle cells with scant to abun-
dant gray to eosinophilic cytoplasm and large
round nuclei with uniform atypia with clear chro- Differential Diagnosis
matin characterize the tumor. The BCOR- ITD
often demonstrates round epithelioid cells admixed Stromal nodules and endometrial stromal sarco-
with the spindle cells. There is brisk mitotic activ- mas have very characteristic histologic features
ity, and hemangiopericytoma-like vessels may be that limit the differential diagnosis to only a few
present. Necrosis and lymphovascular invasion are other lesions. It is most important, however, to
frequent [110, 112]. Unlike YWHAE-NUTMT2 attempt to distinguish these tumors from each
endometrial stromal sarcomas, a background com- other. The growth pattern determines whether a
ponent of LGESS pattern is not present. tumor is a stromal nodule or an endometrial stro-
Immunohistochemical staining for cyclin D1 mal sarcoma. Usually this feature is not discern-
and BCOR is strong and diffusely positive in ible in biopsy material, as the interface with
nearly all ZC3H7B-BCOR-related tumors. normal tissues is, at best, fragmented.
Hormone receptors are variably positive. There is Endovascular growth and infiltrative growth can-
focal actin expression, while caldesmon and des- not be evaluated in biopsies. For these reasons it
min are negative. Interestingly, the BCOR-ITD may not be possible to distinguish between these
lesions in biopsy material (see later, Clinical cells [70]. Conversely, smooth muscle tumors
Queries and Reporting). have a fascicular pattern with interlacing bundles
Most high-grade endometrial stromal sarco- of elongate cells. Cellular leiomyomas also have
mas are less differentiated and lack a monoto- thicker vessels with muscular walls and artifac-
nous cell population. In these cases, it may be tual, cleft-like spaces separating some of the fas-
difficult to determine whether the tumor is a cicles [55]. Immunohistochemistry also is helpful
HGESS or another type of sarcoma, such as leio- in distinguishing a stromal tumor from a cellular
myosarcoma in biopsy material. Practically, this leiomyoma. Stromal tumors are generally nega-
distinction usually makes little difference, as the tive for desmin and h-caldesmon, whereas cellular
clinical management rests on the recognition of leiomyomas are positive [40, 114]. Although stro-
high-grade sarcoma rather than the histologic mal tumors tend to be diffusely reactive for CD10,
subtype. If the differential includes poorly differ- similar reactivity has also been seen in some cel-
entiated or undifferentiated carcinoma with a lular leiomyomas, so this antigen should be inter-
sheetlike proliferation of small, monotonous preted in the context of a panel of antibodies [40].
cells, immunostains for keratin, EMA, CD10, Occasionally, aggregates of histiocytes from
actin, and desmin should resolve the question. the uterine cavity may present as a sheet of
Stromal tumors show reactivity for CD10, actin, monotonous cells in an endometrial biopsy that
and, focally, desmin, whereas carcinoma does can superficially mimic a proliferation of stromal
not. Keratin may be focally positive in stromal cells. Despite the resemblance at low magnifica-
tumors [80], so focal reactivity for this antigen tion, masses of histiocytes lack the vascular
does not necessarily indicate carcinoma. EMA, framework that uniformly accompanies stromal
however, in our experience, is more specific as a cell proliferations. Immunohistochemical stains
marker for carcinoma and is not found in smooth for histiocytes, such as CD68, are useful for iden-
muscle or stromal cells. tifying these cells.
At times stromal tumors show foci of smooth
muscle, and sometimes there is a prominent mix-
ture of stromal and smooth muscle spindle cells. Clinical Queries and Reporting
For practical purposes these tumors should be clas-
sified as stromal tumors, because their behavior Endometrial stromal cell proliferations are diffi-
will be like that of a pure stromal neoplasm. If cult to classify with precision, on a biopsy or
immunohistochemical stains are employed to help curettage specimen as the interface with myome-
make the distinction, then the diagnosis of a smooth trium, upon which the distinction of an ESN ver-
muscle tumor should be reserved for those cases sus an LGESS is made, cannot be discerned in
that show extensive cytoplasmic staining for des- these specimens. Consequently, when the sec-
min. Tumors that show extensive smooth muscle tions demonstrate an apparent low-grade endo-
differentiation (>30%) can be classified as mixed metrial stromal tumor, the diagnosis of
endometrial stromal-smooth muscle tumors [65]. “endometrial stromal tumor” should be made
Cellular leiomyomas (see above, Smooth with a comment explaining why a definitive diag-
Muscle Tumors) can be difficult to distinguish nosis of an ESN versus a LGESS cannot be made.
from low-grade stromal tumors, as both lesions Except for unusual circumstances when this
show closely spaced, small, and spindle-shaped occurs in a young woman who wishes to preserve
cells. In curettage specimens, cellular leiomyo- her fertility, a hysterectomy is necessary to make
mas are very unusual, however, and stromal a definitive diagnosis. In the case of a young
tumors are more likely to be biopsied. Low-grade woman who wishes to retain fertility, imaging
stromal tumors also show the characteristic pat- studies to exclude intramural involvement may
tern of uniform cells in a network of small vascu- be undertaken followed by hysteroscopy and
lar channels, which is distinctive for stromal dilation and curettage. If all these tests are nega-
tumors but not for smooth muscle neoplasms. tive, careful follow-up and repeat curettage in
Stromal tumors also may have aggregates of foam 3–6 months are appropriate. Continued long-
term follow-up for many years is necessary, as in a cellular mesenchyme. The adenofibroma is
low-grade stromal sarcomas may recur after benign, whereas the adenosarcoma has low-
symptom-free intervals up to 20 years. grade malignant features. These are rare neo-
Occasionally, this may result in definitive treat- plasms that usually arise in the corpus, but about
ment, but more than likely, a hysterectomy must 10% originate in the endocervix [41, 121]. They
eventually be performed. generally occur in postmenopausal patients,
In cases of HGESS where the tumor cells are although approximately 30% of these tumors
poorly differentiated, the separation of stromal are found in patients under the age of 50. Rare
sarcoma from other forms of sarcoma may be cases have occurred in teenagers [125]. A few
extremely difficult in small samples of tissue. cases of adenosarcoma have been associated
Establishing the presence of high-grade sarcoma with estrogen use, polycystic ovary syndrome
is sufficient to guide further therapy, however, (Stein-Leventhal syndrome), or prior pelvic
and the subclassification of the neoplasm on irradiation [121]. Adenosarcoma has been asso-
biopsy material has little clinical relevance. ciated with tamoxifen use [126, 127]. On clini-
cal examination, adenosarcoma typically is a
large, polypoid tumor that expands the endome-
Uterine Tumors Resembling Ovarian trial cavity and often prolapses through the cer-
Sex Cord Tumors vical os [41, 121, 128].
Adenofibroma may recur within the uterus
Uterine tumors resembling ovarian sex cord but does not metastasize or extend beyond the
tumor (UTROSCT) normally occur in perimeno- uterus. Adenosarcoma, however, recurs in the
pausal women with a mean age of 52 [115]. The vagina or pelvis in as many as 25% of cases and
cell of origin is unclear, and endometrial stroma sometimes disseminates widely as a high-grade
or uncommitted cells in the uterus have been pro- sarcoma. Although the terminology suggests that
posed. Most are circumscribed submucosal or there is a clear histologic difference between the
intramural nodules with an average diameter of benign adenofibroma and the malignant adeno-
6–7 cm. Microscopically, the tumor grows in sarcoma, many of these tumors have borderline
sheets, trabeculae, plexiform cords, or nests and features that make separation of the two tumor
may have a glomeruloid or retiform appearance. types difficult. Of the two, adenosarcoma is
The cells are spindled cuboidal or columnar, and much more common, accounting for more than
the amount of eosinophilic or foamy cytoplasm 90% of all tumors in this group of adenofibroma
varies. Nuclei are uniform, small, and bland with and adenosarcoma [120]. Complete surgical
inconspicuous nucleoli. The stroma ranges from resection by hysterectomy remains the only
hyaline to endometrial-like to fibrous. The sex- treatment with well-evidenced overall survival
cord-like elements are usually immunoreactive benefit [129]. Adenofibroma may be managed
for cytokeratin, calretinin, inhibin, CD56, melan- more conservatively with repeat curettage and
A, CD99, WT-1, and smooth muscle actin or des- hysteroscopy.
min [83, 116, 117]. Moderate or strong staining
with SF-1 has been demonstrated [118, 119].
Pathologic Features
ixed Epithelial and Mesenchymal

M Adenofibroma is characterized by benign cyto-
Tumors logic features of the epithelium and stroma [41,
124, 130]. The tumor shows a characteristic pat-
Adenofibroma and Adenosarcoma tern of broad-based papillary fronds covered with
bland epithelium and supported by widely distrib-
Adenofibroma and adenosarcoma are related uted spindle cells. Low columnar epithelium lines
tumors [7, 8, 41, 120–124]. Both are biphasic large polypoid fronds of sparsely cellular mesen-
tumors with benign glands regularly distributed chyme and also forms small glandular infoldings
into the stroma. The epithelium lacks atypia or they may show eosinophilic, mucinous, squa-
mitotic activity. The mesenchyme is composed of mous, or clear cell change.
widely spaced spindle cells without periglandular In adenosarcoma, the sarcomatous component
cuffing and also lacks mitotic activity. usually resembles endometrial stromal sarcoma
Adenosarcoma has a low-power growth pat- with plump spindle cells in a rich vascular back-
tern like that of adenofibroma characterized by ground, but the mesenchyme can have features of
leaflike fronds of mesenchyme covered by cuboi- fibrosarcoma or leiomyosarcoma and may show
dal to columnar epithelium in a phyllodes pat- heterologous elements. The cellularity of the
tern. In adenosarcoma, however, the stroma is stroma is typically accentuated around the glands,
hypercellular, a feature indicating low-grade forming hypercellular cuffs with less cellular
malignant growth (Fig. 11.11) [7, 8, 41, 120, 121, mesenchyme further away from the glands. The
123]. There is hypercellular stromal “cuffing” mitotic rate in the stroma is variable (Fig. 11.12),
around glands. These glands with phyllodes and the reported cases have ranged between 1 and
growth are large and often form gaping cystic 40 mitoses per 10 HPFs. Most cases show 4 or
spaces. In adenosarcoma, glands and surface epi- more mitoses per 10 HPFs, but any convincing
thelium lack cytologic features of malignancy, mitotic activity of as little as 2 or more mitoses
although they often appear hyperplastic, with per 10 HPFs is sufficient to establish a diagnosis
mitotic activity and slight nuclear atypia. Usually of adenosarcoma in a case that shows an other-
the glands have an endometrioid appearance, but wise typical growth pattern [121].
Fig. 11.11 Adenosarcoma. The typical architecture is composed of a biphasic pattern of irregular, phyllodiform glands
with hypercellular stromal cuffing; the latter resembles endometrial stroma
Up to 20% of adenosarcomas have a heterolo- terns also may include cells with abundant clear
gous element, such as cartilage, fat, osteoid, or to foamy cytoplasm.
rhabdomyoblasts. Although it has been reported
that the presence of these components has no
influence on prognosis [121], in our experience Immunohistochemical Analysis
the presence of rhabdomyoblasts is associated
with a poorer outcome (unpublished data). In Periglandular cuffing can be highlighted with
occasional cases adenosarcoma shows over- CD10 and often stains diffusely within the mes-
growth of sarcoma, either stromal sarcoma or, enchymal component. Expression of ER and PR
less commonly, leiomyosarcoma or rhabdomyo- in both the glandular and stromal components is
sarcoma [51, 123]. Adenosarcoma with sarcoma- typical [133]. In high-grade adenosarcomas or
tous overgrowth is important to recognize, as this those with sarcomatous overgrowth, the stromal
tumor has a higher frequency of myometrial inva- component may have aberrant p53 staining (dif-
sion, recurrence, and metastases than the typical fuse strong positive staining in >80% of tumor
adenosarcoma [131]. cell nuclei or complete loss of staining) [134].
Adenosarcoma may also show sex-cord-like Loss of ER, PR, and CD10 expression is com-
elements characterized by ribbons or nests of mon in adenosarcomas with sarcomatous over-
cells, sometimes with tubule formation, resem- growth [133, 135]. Desmin, myogenin, and
bling ovarian sex cord tumors [132]. These pat- myoD1 highlight rhabdomyosarcomatous areas.
Fig. 11.12 Adenosarcoma. The stroma forms a hypercellular cuff around the gland. The gland lacks malignant cyto-
logic features. A mitotic figure is present within the hypercellular stromal cuffing (left of center)
Molecular Analysis increased cellularity of the stroma and periglan-

dular cuffing in adenosarcoma are helpful fea-
Using targeted next-generation sequencing, tures in the differential diagnosis. If a patient has
amplification of MYBL1 and ATRX mutations a history of recurrence of apparent endometrial
was found in 50% of adenosarcomas with sarco- polyps, especially large polyps, the differential
matous overgrowth [136]. MYBL1 is located on diagnosis should include adenosarcoma, as recur-
8q13 and is a transcription factor involved in rence of ordinary polyps is unusual.
regulating cell cycle progression. ATRX regu- The atypical polypoid adenomyoma may
lates DNA methylation and chromatin remodel- resemble adenosarcoma because of its mixture of
ing. Mutations in p53 are uncommon [137, 138], atypical glands in a cellular mesenchyme (see
but other studies found p53 mutations in high- Chap. 8). Unlike adenosarcoma, the mesenchyme
grade adenosarcomas [134], or in those tumors of the atypical polypoid adenomyoma is com-
with sarcomatous overgrowth [136]. Recent posed of short, interlacing fascicles of smooth
karyotype analysis of adenosarcomas, adenosar- muscle cells. The atypical polypoid adenomyoma
comas with sarcomatous overgrowth, and meta- is, in general, a smaller lesion, lacking the leaf-
static adenosarcomas found two cases with like glandular pattern of adenosarcoma. In addi-
rearrangements at 8q13, while three tumors had tion, many of the glands in the atypical polypoid
extra copies of chromosome 8 [139]. adenomyoma show nests of nonkeratinizing
squamous (morular) change.
Another tumor to be considered in the differ-
Differential Diagnosis ential diagnosis of adenosarcoma is embryonal
rhabdomyosarcoma. Cervical embryonal rhabdo-
The distinction of adenofibroma from adenosar- myosarcoma commonly occurs as a polypoid
coma may be difficult, because the mitotic rate is mass protruding from the os of women in late
relatively low in some cases of adenosarcoma. As adolescence and early 20s [140–142]. The tumor
mentioned earlier, in biopsy material virtually has a cambium layer and may surround entrapped
any convincing mitotic activity in an otherwise glands resembling the phyllodes-type pattern
characteristic tumor indicates a diagnosis of ade- found in adenosarcoma. In embryonal rhabdo-
nosarcoma. In this regard, immunostaining for myosarcoma, the epithelium and entrapped
Ki-67 can be helpful by showing increased label- glands are usually closer to the surface; however,
ing in stromal cells that are closely applied to the there may be significant morphologic and immu-
glandular component. Although not necessarily nohistochemical (desmin, p53, and myogenin)
diagnostic, the presence of increased prolifera- overlap with adenosarcoma with sarcomatous
tion in these periglandular areas favors the diag- overgrowth and rhabdomyosarcomatous differ-
nosis of adenosarcoma. In our experience, entiation. Hormone receptors ER and PR will
adenofibroma is an extraordinarily rare diagnosis stain the typical adenosarcoma areas (without
in biopsy and curettage specimens and should be sarcomatous overgrowth) and normally lacking
made only when no mitoses are found in the in embryonal rhabdomyosarcomas. Heterologous
stroma. In occasional cases, however, the distinc- elements may be found in both tumors.
tion between the two lesions may be extremely
difficult, even with repeated evaluations, and hys-
terectomy is necessary to allow thorough sam- Clinical Queries and Reporting
pling and ensure complete removal of the lesion.
Either adenofibroma or adenosarcoma may Because adenofibroma and adenosarcoma are
resemble an endometrial polyp. In contrast to unusual neoplasms, a comment regarding the
polyps, these lesions characteristically have a nature of the tumor is warranted. The comment
phyllodes glandular configuration, and the biopsy can include a brief description with a statement
consists of large fragments of tissue. The regarding the mitotic rate of the stroma. Other
Rare Neoplasms 351
salient features of adenosarcoma, such as heter- ALK rearrangements when the fusion partner is
ologous elements or sarcomatous overgrowth, one of these genes. Therefore, a negative FISH
should be noted in a comment. Sarcomatous result should not exclude the diagnosis of IMT
overgrowth is an especially important feature, as if the tumor displays the morphologic features
it is associated with a higher rate of recurrence and ALK immunohistochemistry is positive.
and poor prognosis. A telephone call can help to Although these tumors normally behave in a
clarify the diagnosis for a gynecologist who is benign fashion, tumor size ≥8 cm was predictive
not familiar with the entity. of aggressive behavior [143].
In small samples it may be difficult to estab-
lish the diagnosis of either adenofibroma or ade-
nosarcoma, as these neoplasms often have Lymphoma and Leukemia
relatively bland cytologic features. If the mor-
phologic features suggest either of these entities Lymphoma or leukemic infiltration of the endo-
but are not clearly diagnostic, the gynecologist metrium is a highly unusual finding in biopsy
should be informed of the differential to allow specimens [146, 147]. When this does occur, it
further evaluation as clinically indicated. Often commonly is found in a patient with known dis-
with this information, the clinician may elect to seminated disease. Primary involvement is rare,
proceed with a hysterectomy in the perimeno- and lymphoma is more common than leukemia.
pausal or postmenopausal patient or repeat tissue In the female reproductive organs, a primary
sampling, preferably hysteroscopy and dilation lymphoma was most commonly ovarian (37%)
and curettage in a premenopausal patient. followed by cervical (21.4%) and uterine (16.5%)
[148]. Most of the lymphomas that involve the
corpus are diffuse large B-cell type or follicular
Rare Neoplasms lymphoma [148]. Other types of lymphoma are
rare. Mucosal-associated lymphoid tissue lym-
Inflammatory Myofibroblastic Tumor phoma (MALToma), Burkitt lymphoma, primary
T-cell lymphoma, and Hodgkin lymphoma all
Inflammatory myofibroblastic tumor (IMT) have been reported, however [147–152].
rarely occurs in the uterus, with a mean age in the Leukemic infiltrates most commonly are attribut-
late fourth decade [143]. Tumors can be polypoid able to myelogenous leukemia and may present
or intramural, with myxoid, hyalinized, or comas granulocytic sarcoma [147].
pact patterns. The polygonal, spindle, or stellate Involvement of the endometrium by lym-
cells rarely have severe nuclear atypia and typi- phoma or leukemia results in extensive infiltra-
cally have low mitotic activity [143]. Pale eosino- tion of the stroma with atypical lymphoid cells
philic cytoplasm is common, with granular or that may envelop a few residual glands
vesicular nuclei. An inflammatory infiltrate is (Figs. 11.13 and 11.14). The process should be
common. Immunohistochemical staining is usu- distinguished from severe endometritis which
ally positive for ALK-1, while desmin and may show scattered lymphoid follicles, trans-
smooth muscle actin are typically negative [143, formed lymphocytes, and immunoblasts [147,
144]. Fluorescence in situ hybridization (FISH) 153]. The mixed inflammatory infiltrate of endo-
detected ALK rearrangements in nearly all IMTs metritis and the frequent presence of a neutro-
tested, while anchored multiplex assays or RNA- philic infiltrate in the glands and surface
sequencing- based fusion assays demonstrated epithelium can be helpful features for separating
THBS1, IGFBP5, FN1, DES, and SEC31 as inflammation from a lymphoid neoplasm.
common fusion partners [143, 145]. IGFBP5, Sheets of malignant lymphoid cells in the
FN1, and DES are located on the same chromo- endometrium also may mimic a high-grade car-
some as ALK and thought to arise via chromo- cinoma or a stromal tumor. Immunohistochemical
somal inversions. FISH testing may not reveal stains for leukocyte common antigen and other
Fig. 11.13 Lymphoma. The endometrial stroma is replaced by a diffuse infiltrate of malignant lymphoid cells sur-
rounding benign endometrial glands
lymphoid markers and keratin can help in estab- carcinomatous elements. All these neoplasms rep-
lishing the correct diagnosis. resent high-grade sarcomas that show aggressive
behavior similar to carcinosarcoma. High-grade
sarcoma that resembles malignant fibrous histio-
Miscellaneous Tumors cytoma or fibrosarcoma is best classified as a vari-
ant of high-grade endometrial stromal sarcoma.
It is extremely unusual to encounter pure mesen- Other sarcomas, such as alveolar soft part sarcoma
chymal tumors other than stromal and smooth and malignant rhabdoid tumor, also have been
muscle neoplasms in endometrial biopsy and reported, with most uterine corpus cases occurring
curettage specimens. Cases of a heterologous sar- in the fourth or fifth decades [166–172].
coma, such as chondrosarcoma, osteosarcoma, Hemangiopericytoma is a diagnosis that should be
angiosarcoma, or pleomorphic rhabdomyosar- used with great caution; most of the reported cases
coma, have been reported in the uterine cavity of so-called hemangiopericytoma of the uterus
[154–165]. Occasionally, these primary uterine actually represent endometrial stromal sarcoma
sarcomas can have pulmonary, liver, bone marrow, [68]. Uterine hemangiopericytoma is vanishingly
or cardiac metastasis [162, 163, 165]. Some of rare, if it exists at all.
these neoplasms may represent carcinosarcoma in Benign mesenchymal tumors other than leio-
which the carcinomatous component is indistinct myomas are extremely rare. Occasional lipomas
or the sarcomatous component has overgrown the and hemangiomas have been encountered, but
Other Lesions and Tumor-Like Conditions 353
Fig. 11.14 Lymphoma. Diffuse large B-cell lymphoma The diffuse infiltrate of highly atypical lymphoid cells
infiltrates the endometrium and encompasses a residual contrasts with the polymorphous lymphoplasmacytic
benign gland. The patient presented with abnormal bleed- inflammatory infiltrate normally seen in chronic
ing and had no known disease prior to endometrial biopsy. endometritis
they are so unusual that other, more common pri- lial tumors [183], paraganglioma [184–187],
mary processes should be excluded before diag- glioma [188], and Wilms tumor [165, 189–191].
nosing these rare lesions in a biopsy specimen. Adenomatoid tumors of the myometrium only
For example, adipose tissue in an endometrial rarely involve the endometrium but may be diag-
biopsy usually represents omental fat indicating a nosed in endometrial samples [192].
uterine perforation rather than a lipomatous
tumor or focal fatty change of the endometrium.
In these instances, the clinician should be con- ther Lesions and Tumor-Like
O
tacted immediately. Likewise, lesions with prom- Conditions
inent vascularity usually represent a low-grade
stromal tumor or vascular ectasia at the surface of On occasion, an endometrial sample may contain
a polyp rather than hemangioma. fragments of the bone or cartilage. These frag-
Other primary tumors of the corpus that may ments may represent residual fetal parts from an
be encountered in endometrial biopsies or curet- earlier abortion or as a result of osseous metaplasia
tings are extremely unusual. These oddities [193–195]. Usually the specimen consists of scat-
include yolk sac tumor [173–176], teratoma [152, tered fragments of ossified material without spe-
177, 178], uterine neuroectodermal tumors [179– cific relation to the other fragments of endometrial
182], central nervous system-type neuroepithe- tissue present. These foci may be accompanied by
an inflammatory response. Calcification also can 8. Zaloudek CJ, Hendrickson MR, Soslow
RA. Mesenchymal tumors of the uterus. In: Kurman
occur in a variety of other circumstances [196– RJ, Hedrick Ellenson L, Ronnett BM, editors.
202]. Sometimes the calcification is extensive, Blaustein’s pathology of the female genital tract. 6th
leading to ossification [199–201]. A few cases of ed. New York: Springer; 2011. p. 453–527.
psammoma body formation in association with 9. Clement P, Young R, editors. Pure mesenchymal
tumors. Tumor and tumorlike lesions of the uterine
exogenous hormone use have been reported [197, corpus and cervix. New York: Churchill Livingstone;
202]. Psammoma bodies without associated epi- 1993. p. 265–328.
thelium may be associated with benign findings 10. Kempson RL, Hendrickson MR. Pure mesenchymal
[203] or may be a sign of uterine or extrauterine neoplasms of the uterine corpus: selected problems.
Semin Diagn Pathol. 1988;5:172–98.
serous carcinoma, however [204]. Rarely, gland 11. Clement PB. Pathology of the uterine corpus. Hum
contents or stroma may show microscopic areas of Pathol. 1991;22:776–91.
dystrophic calcification that have no apparent sig- 12. Evans HL, Chawla SP, Simpson C, Finn KP. Smooth
nificance, possibly representing dystrophic calcifi- muscle neoplasms of the uterus other than ordinary
leiomyoma. A study of 46 cases, with emphasis on
cation of inspissated mucus or cellular debris. diagnostic criteria and prognostic factors. Cancer.
Sampling of a uterine septum in a premeno- 1988;62:2239–47.
pausal patient may yield fragments of poorly 13. Prayson RA, Hart WR. Mitotically active leiomy-
developed endometrium with underlying smooth omas of the uterus. Am J Clin Pathol. 1992;97:
14–20.
muscle. These changes resemble the changes 14. O’Connor DM, Norris HJ. Mitotically active leio-
encountered adjacent to submucosal leiomyo- myomas of the uterus. Hum Pathol. 1990;21:223–7.
mas, and the differential diagnosis requires clini- 15. Kurman RJ, Norris HJ. Mesenchymal tumors of
cal correlation. the uterus. VI. Epithelioid smooth muscle tumors
including leiomyoblastoma and clear-cell leiomy-
oma: a clinical and pathologic analysis of 26 cases.
Cancer. 1976;37:1853–65.
References 16. Mazur MT, Kraus FT. Histogenesis of morphologic
variations in tumors of the uterine wall. Am J Surg
1. Goff BA, Rice LW, Fleischhacker D, Muntz HG, Pathol. 1980;4:59–74.
Falkenberry SS, Nikrui N, Fuller AF. Uterine leio- 17. Hendrickson M, Longacre T, Kempson R. The uter-
myosarcoma and endometrial stromal sarcoma: ine corpus. Sternberg SS, diagnostic surgical pathol-
lymph node metastases and sites of recurrence. ogy. 3rd ed. Philadelphia: Lippincott Williams &
Gynecol Oncol. 1993;50:105–9. Wilkins’; 1999. p. 2203–305.
2. Leibsohn S, d’Ablaing G, Mishell DR, Schlaerth 18. Perrone T, Dehner LP. Prognostically favorable
JB. Leiomyosarcoma in a series of hysterectomies “mitotically active” smooth-muscle tumors of the
performed for presumed uterine leiomyomas. Am uterus. A clinicopathologic study of ten cases. Am J
J Obstet Gynecol. 1990;162:968–74. discussion Surg Pathol. 1988;12:1–8.
974–976 19. Downes KA, Hart WR. Bizarre leiomyomas of the
3. Kahanpää KV, Wahlström T, Gröhn P, Heinonen E, uterus: a comprehensive pathologic study of 24
Nieminen U, Widholm O. Sarcomas of the uterus: cases with long-term follow-up. Am J Surg Pathol.
a clinicopathologic study of 119 patients. Obstet 1997;21:1261–70.
Gynecol. 1986;67:417–24. 20. Ly A, Mills AM, McKenney JK, Balzer BL, Kempson
4. Hinchcliff EM, Esselen KM, Watkins JC, Oduyebo RL, Hendrickson MR, Longacre TA. Atypical leio-
T, Rauh-Hain JA, Del Carmen MG, Quade BJ, Muto myomas of the uterus: a clinicopathologic study of
MG. The role of endometrial biopsy in the preopera- 51 cases. Am J Surg Pathol. 2013;37:643–9.
tive detection of uterine leiomyosarcoma. J Minim 21. Bell SW, Kempson RL, Hendrickson
Invasive Gynecol. 2016;23:567–72. MR. Problematic uterine smooth muscle neoplasms.
5. Deligdish L, Loewenthal M. Endometrial changes A clinicopathologic study of 213 cases. Am J Surg
associated with myomata of the uterus. J Clin Pathol. Pathol. 1994;18:535–58.
1970;23:676–80. 22. Clement PB, Young RH, Scully RE. Diffuse, peri-
6. Lethaby A, Puscasiu L, Vollenhoven B. Preoperative nodular, and other patterns of hydropic degen-
medical therapy before surgery for uterine fibroids. eration within and adjacent to uterine leiomyomas.
Cochrane Database Syst Rev. 2017;11:CD000547. Problems in differential diagnosis. Am J Surg Pathol.
7. Silverberg S, Kurman R. Tumors of the uterine cor- 1992;16:26–32.
pus and gestational trophoblastic disease, Atlas of 23. Boyd C, McCluggage WG. Unusual morphological
tumor pathology 3rd series. Washington, DC: Armed features of uterine leiomyomas treated with proges-
Forces Institute of Pathology; 1992. togens. J Clin Pathol. 2011;64:485–9.
References 355
24. Ferry JA, Harris NL, Scully RE. Uterine leiomyo- a national population-based study. Eur J Obstet
mas with lymphoid infiltration simulating lym- Gynecol Reprod Biol. 2018;224:12–6.
phoma. A report of seven cases. Int J Gynecol 37. Taylor HB, Norris HJ. Mesenchymal tumors of the
Pathol. 1989;8:263–70. uterus. IV. Diagnosis and prognosis of leiomyosar-
25. Prayson RA, Goldblum JR, Hart WR. Epithelioid comas. Arch Pathol. 1966;82:40–4.
smooth-muscle tumors of the uterus: a clinico- 38. Peacock G, Archer S. Myxoid leiomyosarcoma of
pathologic study of 18 patients. Am J Surg Pathol. the uterus: case report and review of the literature.
1997;21:383–91. Am J Obstet Gynecol. 1989;160:1515–8. discussion
26. Roth LM, Reed RJ. Cotyledonoid leiomyoma of 1518–1519
the uterus: report of a case. Int J Gynecol Pathol. 39. Burch DM, Tavassoli FA. Myxoid leiomyosarcoma
2000;19:272–5. of the uterus. Histopathology. 2011;59:1144–55.
27. Sanz-Ortega J, Vocke C, Stratton P, Linehan WM, 40. Oliva E, Young RH, Amin MB, Clement PB. An
Merino MJ. Morphologic and molecular character- immunohistochemical analysis of endometrial
istics of uterine leiomyomas in hereditary leiomyo- stromal and smooth muscle tumors of the uterus:
matosis and renal cancer (HLRCC) syndrome. Am J a study of 54 cases emphasizing the importance of
Surg Pathol. 2013;37:74–80. using a panel because of overlap in immunoreac-
28. Alsolami S, El-Bahrawy M, Kalloger SE, tivity for individual antibodies. Am J Surg Pathol.
AlDaoud N, Pathak TB, Chung CT, Cheung CT, 2002;26:403–12.
Mulligan AM, Tomlinson IP, Pollard PJ, Gilks CB, 41. Zaloudek CJ, Norris HJ. Adenofibroma and adeno-
McCluggage WG, Clarke BA. Current morphologic sarcoma of the uterus: a clinicopathologic study of
criteria perform poorly in identifying hereditary 35 cases. Cancer. 1981;48:354–66.
leiomyomatosis and renal cell carcinoma syndrome- 42. Kaminski PF, Tavassoli FA. Plexiform tumorlet:
associated uterine leiomyomas. Int J Gynecol Pathol. a clinical and pathologic study of 15 cases with
2014;33:560–7. ultrastructural observations. Int J Gynecol Pathol.
29. Siegler L, Erber R, Burghaus S, Brodkorb T, 1984;3:124–34.
Wachter D, Wilkinson N, Bolton J, Stringfellow 43. Seidman JD, Thomas RM. Multiple plexiform
H, Haller F, Beckmann MW, Hartmann A, Agaimy tumorlets of the uterus. Arch Pathol Lab Med.
A. Fumarate hydratase (FH) deficiency in uterine 1993;117:1255–6.
leiomyomas: recognition by histological features 44. Conlon N, Soslow RA, Murali R. Perivascular epi-
versus blind immunoscreening. Virchows Arch. thelioid tumours (PEComas) of the gynaecological
2018;472(5):789–96. tract. J Clin Pathol. 2015;68:418–26.
30. Mäkinen N, Kämpjärvi K, Frizzell N, Bützow R, 45. Martignoni G, Pea M, Reghellin D, Gobbo S,
Vahteristo P. Characterization of MED12, HMGA2, Zamboni G, Chilosi M, Bonetti F. Molecular pathol-
and FH alterations reveals molecular variabil- ogy of lymphangioleiomyomatosis and other peri-
ity in uterine smooth muscle tumors. Mol Cancer. vascular epithelioid cell tumors. Arch Pathol Lab
2017;16:101. Med. 2010;134:33–40.
31. Miettinen M, Felisiak-Golabek A, Wasag B, Chmara 46. Lim GSD, Oliva E. The morphologic spectrum of
M, Wang Z, Butzow R, Lasota J. Fumarase-deficient uterine PEC-cell associated tumors in a patient
uterine leiomyomas: an immunohistochemical, with tuberous sclerosis. Int J Gynecol Pathol.
molecular genetic, and clinicopathologic study of 86 2011;30:121–8.
cases. Am J Surg Pathol. 2016;40:1661–9. 47. Longacre TA, Hendrickson MR, Kapp DS, Teng
32. Kurman RJ, Carcangiu ML, Herrington CS, Young NN. Lymphangioleiomyomatosis of the uterus
RH. WHO classification of tumors of female repro- simulating high-stage endometrial stromal sarcoma.
ductive organs. 4th ed. Lyon: International Agency Gynecol Oncol. 1996;63:404–10.
for Research on Cancer; 2014. 48. Vang R, Kempson RL. Perivascular epithelioid cell
33. Croce S, Chibon F. MED12 and uterine smooth tumor (‘PEComa’) of the uterus: a subset of HMB-
muscle oncogenesis: state of the art and perspec- 45-positive epithelioid mesenchymal neoplasms
tives. Eur J Cancer. 2015;51:1603–10. with an uncertain relationship to pure smooth mus-
34. Abeler VM, Røyne O, Thoresen S, Danielsen HE, cle tumors. Am J Surg Pathol. 2002;26:1–13.
Nesland JM, Kristensen GB. Uterine sarcomas in 49. Folpe AL, Mentzel T, Lehr H-A, Fisher C, Balzer
Norway. A histopathological and prognostic survey BL, Weiss SW. Perivascular epithelioid cell neo-
of a total population from 1970 to 2000 including plasms of soft tissue and gynecologic origin: a clini-
419 patients. Histopathology. 2009;54:355–64. copathologic study of 26 cases and review of the
35. Skorstad M, Kent A, Lieng M. Preoperative evalu- literature. Am J Surg Pathol. 2005;29:1558–75.
ation in women with uterine leiomyosarcoma. 50. Schoolmeester JK, Howitt BE, Hirsch MS, Dal
A nationwide cohort study. Acta Obstet Gynecol Cin P, Quade BJ, Nucci MR. Perivascular epithe-
Scand. 2016;95:1228–34. lioid cell neoplasm (PEComa) of the gynecologic
36. Yuk J-S, Shin J-Y, Moon H-S, Lee JH. The incidence tract: clinicopathologic and immunohistochemi-
of unexpected uterine malignancy in women under- cal characterization of 16 cases. Am J Surg Pathol.
going hysteroscopic myomectomy or polypectomy: 2014;38:176–88.
51. Chang KL, Crabtree GS, Lim-Tan SK, Kempson 69. Dionigi A, Oliva E, Clement PB, Young
RL, Hendrickson MR. Primary uterine endometrial RH. Endometrial stromal nodules and endometrial
stromal neoplasms. A clinicopathologic study of 117 stromal tumors with limited infiltration: a clini-
cases. Am J Surg Pathol. 1990;14:415–38. copathologic study of 50 cases. Am J Surg Pathol.
52. Tavassoli FA, Norris HJ. Mesenchymal tumours 2002;26:567–81.
of the uterus. VII. A clinicopathological study of 70. Oliva E, Young RH, Clement PB, Bhan AK, Scully
60 endometrial stromal nodules. Histopathology. RE. Cellular benign mesenchymal tumors of the
1981;5:1–10. uterus. A comparative morphologic and immuno-
53. Hart WR, Yoonessi M. Endometrial stromatosis of histochemical analysis of 33 highly cellular leio-
the uterus. Obstet Gynecol. 1977;49:393–403. myomas and six endometrial stromal nodules, two
54. Norris HJ, Taylor HB. Mesenchymal tumors of the frequently confused tumors. Am J Surg Pathol.
uterus. I. A clinical and pathological study of 53 endo- 1995;19:757–68.
metrial stromal tumors. Cancer. 1966;19:755–66. 71. McCluggage WG, Sumathi VP, Maxwell P. CD10 is
55. Yoonessi M, Hart WR. Endometrial stromal sarco- a sensitive and diagnostically useful immunohisto-
mas. Cancer. 1977;40:898–906. chemical marker of normal endometrial stroma and
56. Fekete PS, Vellios F. The clinical and histologic of endometrial stromal neoplasms. Histopathology.
spectrum of endometrial stromal neoplasms: a report 2001;39:273–8.
of 41 cases. Int J Gynecol Pathol. 1984;3:198–212. 72. Agoff SN, Grieco VS, Garcia R, Gown
57. Evans HL. Endometrial stromal sarcoma and AM. Immunohistochemical distinction of endome-
poorly differentiated endometrial sarcoma. Cancer. trial stromal sarcoma and cellular leiomyoma. Appl
1982;50:2170–82. Immunohistochem Mol Morphol. 2001;9:164–9.
58. Nucci MR. Practical issues related to uterine pathol- 73. Chu PG, Arber DA, Weiss LM, Chang KL. Utility
ogy: endometrial stromal tumors. Mod Pathol. of CD10 in distinguishing between endometrial stro-
2016;29(Suppl 1):S92–103. mal sarcoma and uterine smooth muscle tumors: an
59. Conklin CMJ, Longacre TA. Endometrial stromal immunohistochemical comparison of 34 cases. Mod
tumors: the new WHO classification. Adv Anat Pathol. 2001;14:465–71.
Pathol. 2014;21:383–93. 74. Bartosch C, Exposito MI, Lopes JM. Low-grade
60. Piver MS, Rutledge FN, Copeland L, Webster K, endometrial stromal sarcoma and undifferentiated
Blumenson L, Suh O. Uterine endolymphatic stro- endometrial sarcoma: a comparative analysis empha-
mal myosis: a collaborative study. Obstet Gynecol. sizing the importance of distinguishing between
1984;64:173–8. these two groups. Int J Surg Pathol. 2010;18:286–91.
61. Mansi JL, Ramachandra S, Wiltshaw E, Fisher 75. Oliva E, Young RH, Clement PB, Scully RE. Myxoid
C. Endometrial stromal sarcomas. Gynecol Oncol. and fibrous endometrial stromal tumors of the
1990;36:113–8. uterus: a report of 10 cases. Int J Gynecol Pathol.
62. Larson B, Silfverswärd C, Nilsson B, Pettersson 1999;18:310–9.
F. Endometrial stromal sarcoma of the uterus. A clini- 76. Devaney K, Tavassoli FA. Immunohistochemistry
cal and histopathological study. The Radiumhemmet as a diagnostic aid in the interpretation of unusual
series 1936–1981. Eur J Obstet Gynecol Reprod mesenchymal tumors of the uterus. Mod Pathol.
Biol. 1990;35:239–49. 1991;4:225–31.
63. Gloor E, Schnyder P, Cikes M, Hofstetter J, Cordey 77. Farhood AI, Abrams J. Immunohistochemistry
R, Burnier F, Knobel P. Endolymphatic stromal of endometrial stromal sarcoma. Hum Pathol.
myosis. Surgical and hormonal treatment of exten- 1991;22:224–30.
sive abdominal recurrence 20 years after hysterec- 78. Franquemont DW, Frierson HF, Mills SE. An immu-
tomy. Cancer. 1982;50:1888–93. nohistochemical study of normal endometrial stroma
64. Thatcher SS, Woodruff JD. Uterine stromatosis: a and endometrial stromal neoplasms. Evidence for
report of 33 cases. Obstet Gynecol. 1982;59:428–34. smooth muscle differentiation. Am J Surg Pathol.
65. Oliva E, Clement PB, Young RH, Scully RE. Mixed 1991;15:861–70.
endometrial stromal and smooth muscle tumors of 79. Lillemoe TJ, Perrone T, Norris HJ, Dehner
the uterus: a clinicopathologic study of 15 cases. Am LP. Myogenous phenotype of epithelial-like areas
J Surg Pathol. 1998;22:997–1005. in endometrial stromal sarcomas. Arch Pathol Lab
66. Lloreta J, Prat J. Endometrial stromal nodule with Med. 1991;115:215–9.
smooth and skeletal muscle components simu- 80. Binder SW, Nieberg RK, Cheng L, al- Jitawi
lating stromal sarcoma. Int J Gynecol Pathol. S. Histologic and immunohistochemical analysis
1992;11:293–8. of nine endometrial stromal tumors: an unexpected
67. Baker PM, Moch H, Oliva E. Unusual morphologic high frequency of keratin protein positivity. Int J
features of endometrial stromal tumors: a report of 2 Gynecol Pathol. 1991;10:191–7.
cases. Am J Surg Pathol. 2005;29:1394–8. 81. McCluggage WG, Date A, Bharucha H, Toner
68. Oliva E, Clement PB, Young RH. Endometrial stro- PG. Endometrial stromal sarcoma with sex cord-
mal tumors: an update on a group of tumors with a like areas and focal rhabdoid differentiation.
protean phenotype. Adv Anat Pathol. 2000;7:257–81. Histopathology. 1996;29:369–74.
References 357
82. Baker RJ, Hildebrandt RH, Rouse RV, Hendrickson ization in endometrial stromal neoplasms. Am J Surg
MR, Longacre TA. Inhibin and CD99 (MIC2) Pathol. 2007;31:65–70.
expression in uterine stromal neoplasms with sex- 96. Oliva E, de Leval L, Soslow RA, Herens C. High fre-
cord-like elements. Hum Pathol. 1999;30:671–9. quency of JAZF1-JJAZ1 gene fusion in endometrial
83. Irving JA, Carinelli S, Prat J. Uterine tumors resem- stromal tumors with smooth muscle differentiation
bling ovarian sex cord tumors are polyphenotypic by interphase FISH detection. Am J Surg Pathol.
neoplasms with true sex cord differentiation. Mod 2007;31:1277–84.
Pathol. 2006;19:17–24. 97. Dal Cin P, Aly MS, De Wever I, Moerman P, Van
84. Fukunaga M, Miyazawa Y, Ushigome S. Endometrial Den Berghe H. Endometrial stromal sarcoma t(7;17)
low-grade stromal sarcoma with ovarian sex cord- (p15–21;q12–21) is a nonrandom chromosome
like differentiation: report of two cases with an change. Cancer Genet Cytogenet. 1992;63:43–6.
immunohistochemical and flow cytometric study. 98. Micci F, Walter CU, Teixeira MR, Panagopoulos I,
Pathol Int. 1997;47:412–5. Bjerkehagen B, Saeter G, Heim S. Cytogenetic and
85. Ohta Y, Suzuki T, Kojima M, Shiokawa A, Mitsuya molecular genetic analyses of endometrial stromal
T. Low-grade endometrial stromal sarcoma with sarcoma: nonrandom involvement of chromosome
an extensive epithelial-like element. Pathol Int. arms 6p and 7p and confirmation of JAZF1/JJAZ1
2003;53:246–51. gene fusion in t(7;17). Cancer Genet Cytogenet.
86. Sumathi VP, Al-Hussaini M, Connolly LE, Fullerton 2003;144:119–24.
L, McCluggage WG. Endometrial stromal neo- 99. Kurihara S, Oda Y, Ohishi Y, Iwasa A, Takahira
plasms are immunoreactive with WT-1 antibody. Int T, Kaneki E, Kobayashi H, Wake N, Tsuneyoshi
J Gynecol Pathol. 2004;23:241–7. M. Endometrial stromal sarcomas and related high-
87. Zámecník M, Michal M. Endometrial stromal nod- grade sarcomas: immunohistochemical and molecu-
ule with retiform sex-cord-like differentiation. lar genetic study of 31 cases. Am J Surg Pathol.
Pathol Res Pract. 1998;194:449–53. 2008;32:1228–38.
88. Jung C-K, Jung J-H, Lee A, Lee Y-S, Choi Y-J, 100. Micci F, Panagopoulos I, Bjerkehagen B, Heim
Yoon S-K, Lee K-Y. Diagnostic use of nuclear beta- S. Consistent rearrangement of chromosomal band
catenin expression for the assessment of endometrial 6p21 with generation of fusion genes JAZF1/PHF1
stromal tumors. Mod Pathol. 2008;21:756–63. and EPC1/PHF1 in endometrial stromal sarcoma.
89. Clement PB, Scully RE. Uterine tumors resem- Cancer Res. 2006;66:107–12.
bling ovarian sex-cord tumors. A clinicopatho- 101. Stemme S, Ghaderi M, Carlson JW. Diagnosis
logic analysis of fourteen cases. Am J Clin Pathol. of endometrial stromal tumors: a clinicopatho-
1976;66:512–25. logic study of 25 biopsy specimens with identi-
90. McCluggage WG, Shah V, Walsh MY, Toner fication of problematic areas. Am J Clin Pathol.
PG. Uterine tumour resembling ovarian sex cord 2014;141:133–9.
tumour: evidence for smooth muscle differentiation. 102. Lee C-H, Mariño-Enriquez A, Ou W, Zhu M, Ali
Histopathology. 1993;23:83–5. RH, Chiang S, Amant F, Gilks CB, van de Rijn M,
91. Oliva E, Clement PB, Young RH. Epithelioid endo- Oliva E, Debiec-Rychter M, Dal Cin P, Fletcher
metrial and endometrioid stromal tumors: a report JA, Nucci MR. The clinicopathologic features of
of four cases emphasizing their distinction from YWHAE-FAM22 endometrial stromal sarcomas: a
epithelioid smooth muscle tumors and other oxy- histologically high-grade and clinically aggressive
philic uterine and extrauterine tumors. Int J Gynecol tumor. Am J Surg Pathol. 2012;36:641–53.
Pathol. 2002;21:48–55. 103. Lee C-H, Ali RH, Rouzbahman M, Marino-Enriquez
92. Clement PB, Scully RE. Endometrial stromal sar- A, Zhu M, Guo X, Brunner AL, Chiang S, Leung S,
comas of the uterus with extensive endometrioid Nelnyk N, Huntsman DG, Blake Gilks C, Nielsen
glandular differentiation: a report of three cases TO, Dal Cin P, van de Rijn M, Oliva E, Fletcher
that caused problems in differential diagnosis. Int J JA, Nucci MR. Cyclin D1 as a diagnostic immu-
Gynecol Pathol. 1992;11:163–73. nomarker for endometrial stromal sarcoma with
93. Huang H-Y, Ladanyi M, Soslow RA. Molecular YWHAE-FAM22 rearrangement. Am J Surg Pathol.
detection of JAZF1-JJAZ1 gene fusion in endo- 2012;36:1562–70.
metrial stromal neoplasms with classic and variant 104. Lee C-H, Hoang LN, Yip S, Reyes C, Marino-
histology: evidence for genetic heterogeneity. Am J Enriquez A, Eilers G, Tao D, Chiang S, Fletcher
Surg Pathol. 2004;28:224–32. JA, Soslow RA, Nucci MR, Oliva E. Frequent
94. Koontz JI, Soreng AL, Nucci M, Kuo FC, Pauwels expression of KIT in endometrial stromal sarcoma
P, van Den Berghe H, Dal Cin P, Fletcher JA, Sklar with YWHAE genetic rearrangement. Mod Pathol.
J. Frequent fusion of the JAZF1 and JJAZ1 genes in 2014;27:751–7.
endometrial stromal tumors. Proc Natl Acad Sci U S 105. Chiang S, Lee C-H, Stewart CJR, Oliva E, Hoang
A. 2001;98:6348–53. LN, Ali RH, Hensley ML, Arias-Stella JA, Frosina
95. Nucci MR, Harburger D, Koontz J, Dal Cin P, Sklar D, Jungbluth AA, Benayed R, Ladanyi M, Hameed
J. Molecular analysis of the JAZF1-JJAZ1 gene M, Wang L, Kao Y-C, Antonescu CR, Soslow
fusion by RT-PCR and fluorescence in situ hybrid- RA. BCOR is a robust diagnostic immunohisto-
chemical marker of genetically diverse high-grade 115. Blake EA, Sheridan TB, Wang KL, Takiuchi T,
endometrial stromal sarcoma, including tumors Kodama M, Sawada K, Matsuo K. Clinical charac-
exhibiting variant morphology. Mod Pathol. teristics and outcomes of uterine tumors resembling
2017;30:1251–61. ovarian sex-cord tumors (UTROSCT): a systematic
106. Lee C-H, Nucci MR. Endometrial stromal sar- review of literature. Eur J Obstet Gynecol Reprod
coma – the new genetic paradigm. Histopathology. Biol. 2014;181:163–70.
2015;67:1–19. 116. de Leval L, Lim GSD, Waltregny D, Oliva E. Diverse
107. Lee C-H, Ou W-B, Mariño-Enriquez A, Zhu M, phenotypic profile of uterine tumors resembling ovar-
Mayeda M, Wang Y, Guo X, Brunner AL, Amant ian sex cord tumors: an immunohistochemical study
F, French CA, West RB, McAlpine JN, Gilks CB, of 12 cases. Am J Surg Pathol. 2010;34:1749–61.
Yaffe MB, Prentice LM, McPherson A, Jones SJM, 117. Hurrell DP, McCluggage WG. Uterine tumour
Marra MA, Shah SP, van de Rijn M, Huntsman DG, resembling ovarian sex cord tumour is an immu-
Dal Cin P, Debiec-Rychter M, Nucci MR, Fletcher nohistochemically polyphenotypic neoplasm which
JA. 14-3-3 fusion oncogenes in high-grade endo- exhibits coexpression of epithelial, myoid and sex
metrial stromal sarcoma. Proc Natl Acad Sci U S A. cord markers. J Clin Pathol. 2007;60:1148–54.
2012;109:929–34. 118. Croce S, de Kock L, Boshari T, Hostein I, Velasco
108. Sciallis AP, Bedroske PP, Schoolmeester JK, Sukov V, Foulkes WD, McCluggage WG. Uterine tumor
WR, Keeney GL, Hodge JC, Bell DA. High-grade endo- resembling ovarian sex cord tumor (UTROSCT)
metrial stromal sarcomas: a clinicopathologic study of a commonly exhibits positivity with sex cord mark-
group of tumors with heterogenous morphologic and ers FOXL2 and SF-1 but lacks FOXL2 and DICER1
genetic features. Am J Surg Pathol. 2014;38:1161–72. mutations. Int J Gynecol Pathol. 2016;35:301–8.
109. Croce S, Hostein I, Ribeiro A, Garbay D, Velasco V, 119. Stewart CJR, Crook M, Tan A. SF1 immunohisto-
Stoeckle E, Guyon F, Floquet A, Neuville A, Coindre chemistry is useful in differentiating uterine tumours
J-M, MacGrogan G, Chibon F. YWHAE rearrange- resembling sex cord-stromal tumours from potential
ment identified by FISH and RT-PCR in endometrial histological mimics. Pathology. 2016;48:434–40.
stromal sarcomas: genetic and pathological correla- 120. Clement P, Young R, editors. Tumors with mixed
tions. Mod Pathol. 2013;26:1390–400. epithelial and mesenchymal elements. Tumor and
110. Hoang LN, Aneja A, Conlon N, Delair DF, Middha tumorlike lesions of the uterine corpus and cervix.
S, Benayed R, Hensley ML, Park KJ, Hollmann TJ, 1st ed. New York: Churchill Livingstone; 1993.
Hameed MR, Antonescu CR, Soslow RA, Chiang p. 329–70.
S. Novel high-grade endometrial stromal sarcoma: a 121. Clement PB, Scully RE. Mullerian adenosarcoma
morphologic mimicker of myxoid leiomyosarcoma. of the uterus: a clinicopathologic analysis of 100
Am J Surg Pathol. 2017;41:12–24. cases with a review of the literature. Hum Pathol.
111. Mariño-Enriquez A, Lauria A, Przybyl J, Ng TL, 1990;21:363–81.
Kowalewska M, Debiec-Rychter M, Ganesan R, 122. Clement PB, Scully RE. Müllerian adenosarcoma
Sumathi V, George S, McCluggage WG, Nucci MR, of the uterus. A clinicopathologic analysis of ten
Lee C-H, Fletcher JA. BCOR internal tandem dupli- cases of a distinctive type of müllerian mixed tumor.
cation in high-grade uterine sarcomas. Am J Surg Cancer. 1974;34:1138–49.
Pathol. 2018;42:335–41. 123. Kaku T, Silverberg SG, Major FJ, Miller A, Fetter B,
112. Lewis N, Soslow RA, Delair DF, Park KJ, Murali R, Brady MF. Adenosarcoma of the uterus: a gyneco-
Hollmann TJ, Davidson B, Micci F, Panagopoulos logic oncology group clinicopathologic study of 31
I, Hoang LN, Arias-Stella JA, Oliva E, Young RH, cases. Int J Gynecol Pathol. 1992;11:75–88.
Hensley ML, Leitao MM, Hameed M, Benayed R, 124. Vellios F, Ng AB, Reagen JW. Papillary adenofi-
Ladanyi M, Frosina D, Jungbluth AA, Antonescu broma of the uterus: a benign mesodermal mixed
CR, Chiang S. ZC3H7B-BCOR high-grade endo- tumor of Müllerian origin. Am J Clin Pathol.
metrial stromal sarcomas: a report of 17 cases of a 1973;60:543–51.
newly defined entity. Mod Pathol. 2018;31:674–84. 125. Shinnick JK, Kumar N, Beffa L, Miller K,
113. Gremel G, Liew M, Hamzei F, Hardell E, Selling Friedman MA, Kalife E, DiSilvestro P, Mathews
J, Ghaderi M, Stemme S, Pontén F, Carlson JW. A C. Management of low-grade Cervical Müllerian
prognosis based classification of undifferentiated Adenosarcoma in a 14-year-old girl. J Pediatr
uterine sarcomas: identification of mitotic index, Adolesc Gynecol. 2017;30:652–4.
hormone receptors and YWHAE-FAM22 translo- 126. Clement PB, Oliva E, Young RH. Mullerian ade-
cation status as predictors of survival. Int J Cancer. nosarcoma of the uterine corpus associated with
2015;136:1608–18. tamoxifen therapy: a report of six cases and a review
114. Nucci MR, O’Connell JT, Huettner PC, Cviko A, of tamoxifen-associated endometrial lesions. Int J
Sun D. Quade BJ: h-Caldesmon expression effec- Gynecol Pathol. 1996;15:222–9.
tively distinguishes endometrial stromal tumors 127. Bocklage T, Lee KR, Belinson JL. Uterine mullerian
from uterine smooth muscle tumors. Am J Surg adenosarcoma following adenomyoma in a woman on
Pathol. 2001;25:455–63. tamoxifen therapy. Gynecol Oncol. 1992;44:104–9.
References 359
128. Kerner H, Lichtig C. Müllerian adenosarcoma 143. Bennett JA, Nardi V, Rouzbahman M, Morales-
presenting as cervical polyps: a report of seven Oyarvide V, Nielsen GP, Oliva E. Inflammatory
cases and review of the literature. Obstet Gynecol. myofibroblastic tumor of the uterus: a clinico-
1993;81:655–9. pathological, immunohistochemical, and molecular
129. Seagle B-LL, Kanis M, Strohl AE, Shahabi analysis of 13 cases highlighting their broad mor-
S. Survival of women with Mullerian adenosarcoma: phologic spectrum. Mod Pathol. 2017;30:1489–503.
a national cancer data base study. Gynecol Oncol. 144. Rabban JT, Zaloudek CJ, Shekitka KM, Tavassoli
2016;143:636–41. FA. Inflammatory myofibroblastic tumor of the
130. Abell MR. Papillary adenofibroma of the uterine uterus: a clinicopathologic study of 6 cases empha-
cervix. Am J Obstet Gynecol. 1971;110:990–3. sizing distinction from aggressive mesenchymal
131. Clement PB. Müllerian adenosarcomas of the tumors. Am J Surg Pathol. 2005;29:1348–55.
uterus with sarcomatous overgrowth. A clinico- 145. Haimes JD, Stewart CJR, Kudlow BA, Culver BP,
pathological analysis of 10 cases. Am J Surg Pathol. Meng B, Koay E, Whitehouse A, Cope N, Lee J-C,
1989;13:28–38. Ng T, McCluggage WG, Lee C-H. Uterine inflam-
132. Clement PB, Scully RE. Müllerian adenosarcomas matory Myofibroblastic tumors frequently harbor
of the uterus with sex cord-like elements. A clinico- ALK fusions with IGFBP5 and THBS1. Am J Surg
pathologic analysis of eight cases. Am J Clin Pathol. Pathol. 2017;41:773–80.
1989;91:664–72. 146. Harris NL, Scully RE. Malignant lymphoma and
133. Soslow RA, Ali A, Oliva E. Mullerian adenosarco- granulocytic sarcoma of the uterus and vagina.
mas: an immunophenotypic analysis of 35 cases. Am A clinicopathologic analysis of 27 cases. Cancer.
J Surg Pathol. 2008;32:1013–21. 1984;53:2530–45.
134. Hodgson A, Amemiya Y, Seth A, Djordjevic B, Parra- 147. Ferry JA, Young RH. Malignant lymphoma, pseudo-
Herran C. High-grade Müllerian Adenosarcoma: lymphoma, and hematopoietic disorders of the female
genomic and clinicopathologic characterization of genital tract. Pathol Annu. 1991;26(Pt 1):227–63.
a distinct neoplasm with prevalent TP53 pathway 148. Nasioudis D, Kampaktsis PN, Frey M, Witkin SS,
alterations and aggressive behavior. Am J Surg Holcomb K. Primary lymphoma of the female geni-
Pathol. 2017;41:1513–22. tal tract: an analysis of 697 cases. Gynecol Oncol.
135. Gallardo A, Prat J. Mullerian adenosarcoma: a clini- 2017;145:305–9.
copathologic and immunohistochemical study of 55 149. Vang R, Medeiros LJ, Ha CS, Deavers M. Non-
cases challenging the existence of adenofibroma. Hodgkin’s lymphomas involving the uterus: a
Am J Surg Pathol. 2009;33:278–88. clinicopathologic analysis of 26 cases. Mod Pathol.
136. Howitt BE, Sholl LM, Dal Cin P, Jia Y, Yuan L, 2000;13:19–28.
MacConaill L, Lindeman N, Kuo F, Garcia E, 150. Masunaga A, Abe M, Tsujii E, Suzuki Y, Ohgida
Nucci MR, Quade BJ. Targeted genomic analysis of T, Toyama M, Nakamura H, Mori S, Sugawara I,
Müllerian adenosarcoma. J Pathol. 2015;235:37–49. Itoyama S. Primary uterine T-cell lymphoma. Int J
137. Blom R, Guerrieri C. Adenosarcoma of the uterus: Gynecol Pathol. 1998;17:376–9.
a clinicopathologic, DNA flow cytometric, p53 and 151. van de Rijn M, Kamel OW, Chang PP, Lee A,
mdm-2 analysis of 11 cases. Int J Gynecol Cancer. Warnke RA, Salhany KE. Primary low-grade
1999;9:37–43. endometrial B-cell lymphoma. Am J Surg Pathol.
138. Swisher EM, Gown AM, Skelly M, Ek M, Tamimi 1997;21:187–94.
HK, Cain JM, Greer BE, Muntz HG, Goff BA. The 152. Clement P. Miscellaneous primary neoplasms and
expression of epidermal growth factor receptor, metastatic neoplasms. In: Clement P, Young RH,
HER-2/Neu, p53, and Ki-67 antigen in uterine editors. Tumors and tumorlike lesions of the uterine
malignant mixed mesodermal tumors and adenosar- corpus and cervix. New York: Churchill Livingstone;
coma. Gynecol Oncol. 1996;60:81–8. 1993. p. 371–418.
139. Howitt BE, Dal Cin P, Nucci MR, Quade 153. Young RH, Harris NL, Scully RE. Lymphoma-like
BJ. Involvement of chromosome 8 in Müllerian lesions of the lower female genital tract: a report of
Adenosarcoma. Int J Gynecol Pathol. 2017;36:24–30. 16 cases. Int J Gynecol Pathol. 1985;4:289–99.
140. Daya DA, Scully RE. Sarcoma botryoides of the uter- 154. Vakiani M, Mawad J, Talerman A. Heterologous
ine cervix in young women: a clinicopathological sarcomas of the uterus. Int J Gynecol Pathol.
study of 13 cases. Gynecol Oncol. 1988;29:290–304. 1982;1:211–9.
141. Houghton JP, McCluggage WG. Embryonal rhabdo- 155. De Young B, Bitterman P, Lack EE. Primary osteo-
myosarcoma of the cervix with focal pleomorphic sarcoma of the uterus: report of a case with immuno-
areas. J Clin Pathol. 2007;60:88–9. histochemical study. Mod Pathol. 1992;5:212–5.
142. McClean GE, Kurian S, Walter N, Kekre A, 156. Clement PB. Chondrosarcoma of the uterus: report
McCluggage WG. Cervical embryonal rhabdomyo- of a case and review of the literature. Hum Pathol.
sarcoma and ovarian Sertoli-Leydig cell tumour: a 1978;9:726–32.
more than coincidental association of two rare neo- 157. Chiarle R, Godio L, Fusi D, Soldati T, Palestro
plasms? J Clin Pathol. 2007;60:326–8. G. Pure alveolar rhabdomyosarcoma of the corpus
uteri: description of a case with increased serum 173. Joseph MG, Fellows FG, Hearn SA. Primary
level of CA-125. Gynecol Oncol. 1997;66:320–3. endodermal sinus tumor of the endometrium. A
158. Holcomb K, Francis M, Ruiz J, Abulafia O, Matthews clinicopathologic, immunocytochemical, and ultra-
RP, Lee YC. Pleomorphic rhabdomyosarcoma of the structural study. Cancer. 1990;65:297–302.
uterus in a postmenopausal woman with elevated 174. Ohta M, Sakakibara K, Mizuno K, Kano T,
serum CA125. Gynecol Oncol. 1999;74:499–501. Matsuzawa K, Tomoda Y, Nakashima N, Ogawa
159. McCluggage WG, Lioe TF, McClelland HR, Lamki T. Successful treatment of primary endodermal
H. Rhabdomyosarcoma of the uterus: report of two sinus tumor of the endometrium. Gynecol Oncol.
cases, including one of the spindle cell variant. Int J 1988;31:357–64.
Gynecol Cancer. 2002;12:128–32. 175. McNamee T, Damato S, McCluggage WG. Yolk
160. Ordi J, Stamatakos MD, Tavassoli FA. Pure pleo- sac tumours of the female genital tract in older
morphic rhabdomyosarcomas of the uterus. Int J adults derive commonly from somatic epithelial
Gynecol Pathol. 1997;16:369–77. neoplasms: somatically derived yolk sac tumours.
161. Schammel DP, Tavassoli FA. Uterine angiosarco- Histopathology. 2016;69:739–51.
mas: a morphologic and immunohistochemical study 176. Ravishankar S, Malpica A, Ramalingam P, Euscher
of four cases. Am J Surg Pathol. 1998;22:246–50. ED. Yolk sac tumor in extragonadal pelvic sites:
162. Zheng G, Richmond A, Liu C, Berrebi A, Pallavajjala still a diagnostic challenge. Am J Surg Pathol.
A, Haley L, Stone RL, McCarthy E, Vang R, Xing 2017;41:1–11.
D. Clinicopathologic features and genetic altera- 177. Martin E, Scholes J, Richart RM, Fenoglio
tions of a primary osteosarcoma of the uterine cor- CM. Benign cystic teratoma of the uterus. Am J
pus. Int J Gynecol Pathol. 2018 Apr 6. doi: 10.1097/ Obstet Gynecol. 1979;135:429–31.
PGP.0000000000000511. [Epub ahead of print] 178. Wang W-C, Lee M-S, Ko J-L, Lai Y-C. Origin of
163. Abraham C, O’Cearbhaill R, Soslow R. Primary uterine teratoma differs from that of ovarian tera-
osteosarcoma of the uterus with cardiac and pulmo- toma: a case of uterine mature cystic teratoma. Int J
nary metastases. Gynecol Oncol Rep. 2015;11:4–6. Gynecol Pathol. 2011;30:544–8.
164. Liu Y, Guo S, Wang L, Suzuki S, Sugimura H, Li 179. Euscher ED, Deavers MT, Lopez-Terrada D, Lazar
Y. Uterine Angiosarcoma: a case report and litera- AJ, Silva EG, Malpica A. Uterine tumors with neu-
ture review. Int J Gynecol Pathol. 2016;35:264–8. roectodermal differentiation: a series of 17 cases
165. Pinto A, Kahn RM, Rosenberg AE, Slomovitz B, and review of the literature. Am J Surg Pathol.
Quick CM, Whisman MK, Huang M. Uterine rhabdo- 2008;32:219–28.
myosarcoma in adults. Hum Pathol. 2018.; (in press). 180. Daya D, Lukka H, Clement PB. Primitive neuroec-
166. Guillou L, Lamoureux E, Masse S, Costa J. Alveolar todermal tumors of the uterus: a report of four cases.
soft-part sarcoma of the uterine corpus: histological, Hum Pathol. 1992;23:1120–9.
immunocytochemical and ultrastructural study of 181. Hendrickson MR, Scheithauer BW. Primitive neuro-
a case. Virchows Arch A Pathol Anat Histopathol. ectodermal tumor of the endometrium: report of two
1991;418:467–71. cases, one with electron microscopic observations.
167. Cho KR, Rosenshein NB, Epstein JI. Malignant Int J Gynecol Pathol. 1986;5:249–59.
rhabdoid tumor of the uterus. Int J Gynecol Pathol. 182. Chiang S, Snuderl M, Kojiro-Sanada S, Quer
1989;8:381–7. Pi-Sunyer A, Daya D, Hayashi T, Bosincu L,
168. Gray GF, Glick AD, Kurtin PJ, Jones HW. Alveolar Ogawa F, Rosenberg AE, Horn L-C, Wang L,
soft part sarcoma of the uterus. Hum Pathol. Iafrate AJ, Oliva E. Primitive neuroectodermal
1986;17:297–300. tumors of the female genital tract: a morphologic,
169. Nolan NP, Gaffney EF. Alveolar soft part sarcoma of Immunohistochemical, and molecular study of 19
the uterus. Histopathology. 1990;16:97–9. cases. Am J Surg Pathol. 2017;41:761–72.
170. Radig K, Buhtz P, Roessner A. Alveolar soft part 183. Murdock T, Orr B, Allen S, Ibrahim J, Sharma R,
sarcoma of the uterine corpus. Report of two cases Ronnett B, Rodriguez F. Central nervous system-
and review of the literature. Pathol Res Pract. type neuroepithelial tumors and tumor-like pro-
1998;194:59–63. liferations developing in the gynecologic tract and
171. Gaertner EM, Farley JH, Taylor RR, Silver pelvis: clinicopathologic analysis of 23 cases. Am J
SA. Collision of uterine rhabdoid tumor and Surg Path. (in press).
endometrioid adenocarcinoma: a case report and 184. Beham A, Schmid C, Fletcher CD, Auböck L, Pickel
review of the literature. Int J Gynecol Pathol. H. Malignant paraganglioma of the uterus. Virchows
1999;18:396–401. Arch A Pathol Anat Histopathol. 1992;420:453–7.
172. Schoolmeester JK, Carlson J, Keeney GL, Fritchie 185. Tavassoli FA. Melanotic paraganglioma of the
KJ, Oliva E, Young RH, Nucci MR. Alveolar soft uterus. Cancer. 1986;58:942–8.
part sarcoma of the female genital tract: a mor- 186. Young TW, Thrasher TV. Nonchromaffin paragan-
phologic, Immunohistochemical, and molecular glioma of the uterus. A case report. Arch Pathol Lab
cytogenetic study of 10 cases with emphasis on its Med. 1982;106:608–9.
distinction from morphologic mimics. Am J Surg 187. van Leeuwen J, van der Putten HWHM, Demeyere
Pathol. 2017;41:622–32. TBJ, Creemers G-J, JMA P. Paraganglioma of
References 361
the uterus. A case report and review of literature. metrium: the bone is not from a fetus. Fertil Steril.
Gynecol Oncol. 2011;121:418–9. 2009;91:1293.e1-4.
188. Young RH, Kleinman GM, Scully RE. Glioma of the 196. Alpert LC, Haufrect EJ, Schwartz MR. Uterine lithi-
uterus. Report of a case with comments on histogen- asis. Am J Surg Pathol. 1990;14:1071–5.
esis. Am J Surg Pathol. 1981;5:695–9. 197. Herbold DR, Magrane DM. Calcifications of
189. Bittencourt AL, Britto JF, Fonseca LE. Wilms’ the benign endometrium. Arch Pathol Lab Med.
tumor of the uterus: the first report of the literature. 1986;110:666–9.
Cancer. 1981;47:2496–9. 198. Untawale VG, Gabriel JB, Chauhan PM. Calcific
190. Muc RS, Grayson W, Grobbelaar JJ. Adult extrare- endometritis. Am J Obstet Gynecol. 1982;144:482–3.
nal Wilms tumor occurring in the uterus. Arch Pathol 199. Ceccacci L, Clancy G. Endometrial ossification:
Lab Med. 2001;125:1081–3. report of an additional case. Am J Obstet Gynecol.
191. McAlpine J, Azodi M, O’Malley D, Kelly M, 1981;141:103–4.
Golenewsky G, Martel M, Rutherford T, Tavassoli 200. Bhatia NN, Hoshiko MG. Uterine osseous metapla-
F. Extrarenal Wilms’ tumor of the uterine corpus. sia. Obstet Gynecol. 1982;60:256–9.
Gynecol Oncol. 2005;96:892–6. 201. Waxman M, Moussouris HF. Endometrial ossifica-
192. Carlier MT, Dardick I, Lagace AF, Sreeram tion following an abortion. Am J Obstet Gynecol.
V. Adenomatoid tumor of uterus: presentation 1978;130:587–8.
in endometrial curettings. Int J Gynecol Pathol. 202. Valicenti JF, Priester SK. Psammoma bodies of
1986;5:69–74. benign endometrial origin in cervicovaginal cytol-
193. McCluggage GW. Benign diseases of the endome- ogy. Acta Cytol. 1977;21:550–2.
trium. In: Kurman RJ, editor. Blaustein’s pathol- 203. Fausett MB, Zahn CM, Kendall BS, Barth WH. The
ogy of the female genital tract. 6th ed. New York: significance of psammoma bodies that are found
Springer; 2011. p. 305–58. incidentally during endometrial biopsy. Am J Obstet
194. Garg D, Bekker G, Akselrod F, Narasimhulu Gynecol. 2002;186:180–3.
DM. Endometrial osseous metaplasia: an unusual 204. Cameron RI, McCluggage WG. Extensive psam-
cause of infertility. BMJ Case Rep. 2015. momatous calcification of the uterus and cervix
195. Cayuela E, Perez-Medina T, Vilanova J, Alejo M, associated with a uterine serous carcinoma. J Clin
Cañadas P. True osseous metaplasia of the endo- Pathol. 2004;57:888–90.
Methods of Endometrial
Evaluation 12
Contents
Endometrial Sampling Techniques 363
Endometrial Imaging Studies 365
References 368
The clinical decision to perform endometrial Endometrial Sampling Techniques

sampling is dependent on medical history, physi-
cal examination, laboratory data, imaging stud- Endometrial Biopsy
ies, and patient age. Abnormal uterine bleeding
(AUB) may be secondary to structural lesions An endometrial biopsy is a limited sampling pro-
including polyps, adenomyosis, leiomyoma, or cedure that does not require endocervical dilation
malignancy and hyperplasia or nonstructural prior to sampling. It is relatively painless, does
lesions including coagulopathies, ovulatory dys- not require anesthesia in contrast to dilation and
function, endometrial, iatrogenic, or not yet spec- curettage (D&C), and therefore is an office-based
ified, which together comprise the acronym procedure [1–3]. Endometrial biopsies are taken
PALM-COEIN (see Chap. 1 for AUB classifica- either with a small sharp curette or with a flexible
tion). A transvaginal ultrasound is often per- plastic cannula that uses suction to aspirate the
formed first to determine the presence of a tissue such as the Pipelle endometrial aspirator.
structural lesion. However, the first-line method Limited sampling techniques are especially use-
of evaluation (imaging versus tissue sampling) ful for obtaining smaller specimens for surveil-
depends on the clinical suspicion for the diagno- ling the endometrium for organic lesions,
ses listed above. including hyperplasia and carcinoma as well as in

364 12 Methods of Endometrial Evaluation
infertility patients, or for evaluating the response tage with sampling of both the endometrial and
of endometrial tissue to steroid hormone therapy the endocervical mucosa.
of various types. Because many patients who undergo D&C do
The Pipelle and related devices have received not come to hysterectomy, the overall sensitivity
widespread clinical usage because they are sim- and specificity of this technique have been diffi-
ple to use, cost-effective, and reliable for giving cult to determine. Several studies report that
adequate tissue samples in most cases [4]. D&C missed a significant number of polyps,
Comparisons of the Pipelle sampling device with hyperplasias, and carcinomas [17, 37–39]. In
other, more traditional sampling mechanisms fact, one study found that D&C performed imme-
show no significant difference in the overall qual- diately before hysterectomy often sampled less
ity of tissue taken for evaluation [5–16]. Although than half of the cavity [38]. In a recent meta-
some studies find that the Pipelle technique sam- analysis, weighted sensitivities were 100%, 82%,
ples much less of the endometrial surface than and 76% for the diagnoses of endometrial cancer,
other biopsy devices, other studies correlating the atypical hyperplasia, or benign disease, respec-
Pipelle to curettage show good levels of concor- tively [31]. Blind curettage was significantly bet-
dance [14]. Limited sampling leading to underdi- ter for the diagnosis of benign disease compared
agnosis is relatively uncommon [17–22]. Pipelle to hysteroscopy alone [31]. Several studies have
biopsies detected endometrial polyps in 54% of found discrepancies between the grade of endo-
premenopausal and 73% of postmenopausal metrial carcinoma in curettage specimens as
women [23]. Hyperplasia and neoplasia can be compared to hysterectomy specimens [17, 34, 36,
accurately diagnosed by the endometrial biopsy 40–43] with clinically significant upgrading in
[5, 6, 18, 19, 24, 25] and comparable to D&C for 8% of cases [17].
diagnosing carcinoma [17]. The Pipelle biopsy Curettage or biopsy before hysterectomy for
performs similarly to D&C for the preoperative leiomyomas has low sensitivity, as only 36–64%
diagnosis of low- and high-grade endometrial of preoperative samples identified a leiomyosar-
tumors (grade 3 endometrioid, serous and clear coma or sarcoma [44–46]. Preoperative endome-
cell carcinomas, and carcinosarcomas) [18, 26]. trial sampling was significantly more likely to
It is possible to perform limited fractional sam- detect malignancy in postmenopausal patients
pling of endocervical as well as endometrial tis- with a pre-sampling clinical diagnosis of leio-
sue using some biopsy devices. myoma and abnormal bleeding [45, 47].
Uterine perforation is a known infrequent
complication of D&C and hysteroscopy. It rarely
occurs with Pipelle [7, 27]. Detection of adipose Hysteroscopy and Curettage
or colonic tissue in endometrial sampling should
prompt a telephone call to the clinician to com- Hysteroscopy in addition to curettage allows for
municate the possibility of uterine perforation. the most thorough sampling of the endometrium.
It is the “gold standard” and found to increase
endometrial sampling sensitivity [17, 28, 48–50].
Dilation and Curettage Hysteroscopy has the advantage of giving
directed biopsy specimens. It is useful for evalu-
The dilation and curettage (D&C) technique is a ation of women with abnormal uterine bleeding
blind procedure that involves dilation of the cer- and can reveal polyps, small submucosal leio-
vix to allow insertion of a curette into the endo- myomas, a uterine septum or other müllerian
metrial cavity [1, 28–30]. This technique is anomalies, hyperplasias, or malignant lesions
generally used when more extensive sampling of while enhancing clinicopathologic correlations
the endometrium is necessary to exclude signifi- [51]. Hysteroscopy with curettage may occasion-
cant pathology [1, 2, 17, 28–36], or to remove ally be of value in reproductive aged women with
tissue. It also readily allows for a fractional curet- acute AUB where intrauterine pathology is
suspected and tissue sampling is desired [37]. Regardless of the sampling technique used,
Surgical management of acute AUB is reserved in postmenopausal women with continued
for those who have failed medical management bleeding, and benign sampling results, further
or are clinically unstable. The technique of hys- investigation is warranted [31, 58]. In women
teroscopic endometrial ablation can be used as a with postmenopausal bleeding, the failure rate
therapy for AUB secondary to ovulatory dysfunc- of endometrial sampling was 11%, and insuffi-
tion or endometrial causes [52, 53]. Hysteroscopy cient samples were found in 31% [31].
is more precise than saline infusion sonohys- Hyperplasia was found in 7% of those failed or
terography (SIS) or transvaginal ultrasound for insufficient samples [31]. In those women with
the diagnosis of intracavitary lesions [54] and PMB, a thickened endometrium on imaging,
also allows for tissue sampling. and benign blind endometrial sampling results
Hysteroscopy, especially with a large- but continued abnormal bleeding, further inves-
diameter hysteroscope, requires dilation of the tigation with hysteroscopy revealed polyps in
cervix and usually requires local or general 51%, of which 3% contained hyperplasia and
anesthesia and therefore is performed in minor 3% contained carcinoma [58].
procedure rooms or in the operating room; how-
ever, 15–25% of gynecologists perform the pro-
cedure in-office [55]. Small, flexible fiber-optic Other Aspiration Devices
hysteroscopes are useful for in-office diagnostic
purposes and have the advantage of not requir- The Vabra aspirator or the Tis-U-Trap uses a
ing prior dilation of the cervix [56]. Hysteroscopy mechanical vacuum to extract tissue into a tissue
is performed by distending the endometrial cav- collection apparatus [3, 7, 8, 14, 29, 59]. The can-
ity with normal saline or glycine to allow visu- nula for these devices is thin, so general anesthe-
alization [28]. The endometrium is then curetted sia is not required. This technique tends to result
under direct visualization. Rigid hysteroscopes in extensive fragmentation of the tissue, but the
permit the passage of instruments such as grasp- overall quality is comparable to that of a D&C
ers, scissors, and biopsy forceps. Other instru- specimen for diagnosis. Another advantage of
ments such as vaporizing electrodes, barrels, this method is that it samples much of the endo-
and roller balls are used for polypectomies, sub- metrium. Endometrial brush biopsy using the Tao
mucosal myomectomies, and removal of adhe- Brush or Uterobrush also has been effective in
sions or septa [56] and occasionally for detecting endometrial abnormalities [60, 61].
endometrial ablation. This technique uses a brush to remove tissue for
The sensitivity of diagnosing endometrial both histology and cytology and requires a spe-
cancer, atypical hyperplasia, and endometrial cial fixative if cytology is desired as well as cen-
disease in two studies was 83–90%, 75–82%, trifugation to prepare the material.
and 39–95%, respectively [31, 49]. In a recent
meta-analysis, the addition of hysteroscopy to
tissue sampling found better agreement between Endometrial Imaging Studies
preoperative sampling and hysterectomy speci-
mens for tumor grade and histologic subtype Ultrasound
when compared to D&C or office biopsy [17].
Although the addition of hysteroscopy increases Transvaginal ultrasound is another in-office
the sensitivity of polyp sampling, in one study, adjunctive technique for examining the endome-
it failed to fully eradicate premalignant or trium [62–79]. Sonography with a transvaginal
malignant endometrial polyps, and in 89% of probe evaluates the thickness and morphology of
cases, residual complex atypical hyperplasia or the endometrium. The technique permits measure-
endometrioid carcinoma was found as multifo- ment of the thickness of the combined anterior and
cal lesions on hysterectomy [57]. posterior endometrium, which is referred to as the
endometrial “stripe.” This parameter can assist in was more accurate than transvaginal ultrasonog-
determining pathologic and physiologic changes in raphy [54, 86].
the endometrium [63, 78]. In postmenopausal Transvaginal ultrasound is useful for assessing
patients, a thin endometrial stripe of less than 4 or the possible presence of an ectopic pregnancy
5 mm indicates that a significant pathologic lesion [87]; however, sensitivity was lower when serum
of the endometrium is unlikely [66, 80], whereas a β-hCG levels were < 2000 mIU/mL or in patients
stripe thicker than 5 mm suggests the presence of with significant bleeding [88]. When ectopic preg-
polyps, hyperplasia, or carcinoma. The above nancy is suspected, sonography can determine
parameters are useful in triggering additional whether a gestation is in utero or tubal [89, 90].
workup with sonohysterography, in-office hyster- With the increased rate of cesarean sections,
oscopy, or tissue sampling in symptomatic women abnormal implantation in the hysterotomy scar has
with postmenopausal bleeding. However, in also increased. Growth of an intramural gesta-
asymptomatic postmenopausal women, an endo- tional sac can lead to uterine rupture. Ultrasound
metrial stripe greater than 4 mm does not routinely plays an important role in the early diagnosis of an
trigger further evaluation, and individualized incomplete healing of the uterine incision or cesar-
assessment is appropriate [81]. One study evalu- ean scar implantation [91, 92]. Ultrasound also is
ated the endometrial stripe in asymptomatic post- used in the diagnosis of gestational trophoblastic
menopausal patients and found the frequency of disease, especially hydatidiform mole [93, 94].
endometrial cancer was low (3%); in this study, the
endometrial stripe thickness indicated for tissue
sampling was therefore increased to >11 mm for Magnetic Resonance Imaging
asymptomatic women [82]. In those asymptomatic
women with an endometrial stripe >11 mm, the Magnetic resonance imaging (MRI) provides a
risk of endometrial cancer was 9% [82]. clear view of the uterine anatomy that is espe-
Ultrasonography also is useful for determin- cially useful in the evaluation of tumors [95–99].
ing the degree of development of the endome- MRI demonstrates the endometrial–myometrial
trium in the secretory phase by determining its interface or “junctional zone,” so it can be used to
thickness and texture [62–65]. This technique assess myometrial invasion by endometrial carci-
cannot replace biopsy for accurate evaluation of noma [100, 101]. It also can demonstrate myo-
endometrial morphology, however [68]. In addi- metrial invasion in gestational trophoblastic
tion, in endometrial carcinomas, ultrasound can disease or rare abdominal ectopic pregnancies
help to determine the presence of an irregular [102, 103]. This imaging modality also may be
endomyometrial border, increased endometrial used for assessment of benign disorders includ-
thickness, a multifocal vessel pattern, and a mod- ing adenomyosis and endometriosis in the recto-
erate or high color score, all of which were asso- vaginal septum or uterosacral ligaments [103].
ciated with high-risk tumors [83]. On occasion this technique is useful for careful
Saline infusion sonohysterography (SIS) follow-up or assessment of other forms of uterine
employs gel or saline inside the uterine cavity to neoplasia, such as stromal tumors or leiomyo-
better visualize intracavitary masses; two- mas. MRI is time-consuming and expensive;
dimensional and three-dimensional methods are however, it is not practical for routine evaluation
available and can be performed in-office. In a of nonneoplastic conditions.
recent meta-analysis, the difference in accuracy
between 2D and 3D SIS was nonsignificant [84].
When using 2D SIS, the sensitivity and specific- Histologic Techniques
ity for diagnosing intracavitary masses were 93%
and 81% for endometrial polyps and 94% and Proper technique is important to ensure adequate
81% for submucosal leiomyomas, respectively histologic evaluation. Biopsy tissue that suffers
[85]. When diagnosing intracavitary masses, SIS from suboptimal fixation, processing, or s ectioning
will have artifacts that hinder microscopic evalua- smaller samples embedded in one or two cas-
tion. Fixation of endometrial tissue often is diffi- settes. Step sections provide the most compre-
cult because of the large amount of blood that is hensive study of the tissue, allowing the
admixed with the tissue fragments. Some patholo- pathologist to assess the three-dimensional
gists advocate special fixatives such as Bouin’s or aspects of the tissue, and are especially useful for
Lillie’s for endometrial biopsies, since they offer endometrial samples, because the tissue tends to
good cytologic detail [104], but acid-containing be highly fragmented and haphazardly oriented.
fixatives such as these degrade DNA, limiting any Furthermore, levels on the block can uncover
type of molecular analysis of the tissues, and we occasional subtle abnormalities that would not be
find them unnecessary. Formalin is the most noticed if only a single section was reviewed. For
widely available and accepted fixative and, in our example, levels may clarify the presence of a
opinion, is the fixative of choice. polyp, or they may reveal that an apparent polyp-
Before processing, tissue fragments should be oid structure simply represents tangential sec-
separated from as much blood as possible. Well- tioning of normal endometrium. Levels also help
fixed tissue can be placed in a strainer and briefly to determine whether apparently disordered
rinsed with water to remove some blood before glands represent a true abnormality or are simply
placing into a cassette. Wrapping the tissue in an artifact of the procedure. Even endometrial
thin, porous paper (tissue wrap or lens paper) or biopsies for infertility patients benefit from mul-
placing tissue in a porous “biopsy bag” or tea bag tiple levels; the secretory phase (early, mid, or
in the cassette prevents loss of small tissue frag- late) may be adjusted because of step sections.
ments. Experience has shown that sponges used Routine hematoxylin and eosin (H&E) stains
to hold small specimens in cassettes cause arti- generally suffice for the diagnosis of most speci-
facts and distort the three-dimensional configura- mens. Other histochemical stains are rarely nec-
tion of the tissue [105] and should not be used. essary. The use of the periodic acid–Schiff (PAS)
During processing, immersion in alcohol–forma- stain to demonstrate glycogen in the early secre-
lin removes some of the blood, which aids in sub- tory phase has no advantage over careful exami-
sequent sectioning. Tissue processors using nation of routine H&E sections for subnuclear
vacuum provide optimal dehydration and pene- vacuoles. Biopsies showing granulomatous
tration of paraffin into the tissue. In our experi- inflammation should be stained for acid-fast and
ence, ethanol is a better dehydrating agent than fungal organisms. Tissue Gram stains are not
denatured alcohol. To achieve optimum process- useful for evaluation of most cases of
ing, it is necessary to change reagents in the pro- endometritis.
cessor daily. Scant specimens can be centrifuged
and held together with HistoGel™ to ensure full
recovery for histologic study. Frozen Section
Specimens from endometrial biopsies and
curettings are among the more difficult tissues to We discourage frozen sections on endometrial
section, because they are highly fragmented and biopsies because significant artifacts occur as the
bloody. The paraffin-embedded tissue tends to be tissue often is edematous and contains consider-
dry, resulting in shatter and a “venetian blind” able blood. These tissues have very different con-
effect. Warming the block in warm water and sistency and water content compared to other
then applying ice to the surface of the block facil- specimens, such as lymph nodes or breast tissue.
itate even sectioning, with decreased fragmenta- Frozen section requested just prior to hyster-
tion and shatter. Specimens should be cut at ectomy in a perimenopausal or postmenopausal
4–6 μm. woman with abnormal uterine bleeding to deter-
The paraffin blocks should be cut at multiple mine whether carcinoma is present is only help-
levels (two or three) in most cases. Multiple lev- ful if the specimen is clearly benign or clearly
els, or step sections, are especially important for malignant. The subtleties of glandular patterns,
which are crucial in distinguishing atypical trial sampling device with the Novak curette. Am J
hyperplasia from well-differentiated adenocarci- Obstet Gynecol. 1991;165:1287–90.
13. Fothergill DJ, Brown VA, Hill AS. Histological sam-
noma, are obscured by artifacts caused by the fro- pling of the endometrium--a comparison between
zen section technique. A better method of formal curettage and the Pipelle sampler. Br J Obstet
assessing the endometrium preoperatively is to Gynaecol. 1992;99:779–80.
obtain an office-based biopsy. In one study a pre- 14. Rodriguez GC, Yaqub N, King ME. A comparison
of the Pipelle device and the Vabra aspirator as mea-
operative curettage predicted the correct final his- sured by endometrial denudation in hysterectomy
tologic type better than frozen section diagnosis specimens: the Pipelle device samples significantly
(93% versus 81%, respectively) [26]. Formalin- less of the endometrial surface than the Vabra aspira-
fixed specimens can be rapidly processed and tor. Am J Obstet Gynecol. 1993;168:55–9.
15. Check JH, Chase JS, Nowroozi K, Wu CH, Chern
reported, offering greater diagnostic accuracy. R. Clinical evaluation of the Pipelle endometrial suc-
tion curette for timed endometrial biopsies. J Reprod
Med. 1989;34:218–20.
References 16. Henig I, Chan P, Tredway DR, Maw GM, Gullett AJ,
Cheatwood M. Evaluation of the Pipelle curette for
endometrial biopsy. J Reprod Med. 1989;34:786–9.
1. Broomfield D, Armstrong A, Carnovale D, Butler 17. Visser NCM, Reijnen C, Massuger LFAG, Nagtegaal
W. Normal and abnormal uterine bleeding. In: Jones ID, Bulten J, Pijnenborg JMA. Accuracy of endo-
III HW, Rock JA, editors. Te Linde’s Operative metrial sampling in endometrial carcinoma: a sys-
Gynecology. 11th ed. Philadelphia: Wolters Kluwer; tematic review and meta-analysis. Obstet Gynecol.
2015. p. 554–76. 2017;130:803–13.
2. Altchek A, Deligdish L, Friedman F, Brodman 18. Dijkhuizen FP, Mol BW, Brölmann HA, Heintz
M. Endometrial sampling techniques. The uterus: AP. The accuracy of endometrial sampling in
pathology, diagnosis and management. New York: the diagnosis of patients with endometrial carci-
Springer-Verlag; 1991. p. 155–62. noma and hyperplasia: a meta-analysis. Cancer.
3. Feldman S, Berkowitz RS, Tosteson AN. Cost- 2000;89:1765–72.
effectiveness of strategies to evaluate postmeno- 19. Guido RS, Kanbour-Shakir A, Rulin MC,
pausal bleeding. Obstet Gynecol. 1993;81:968–75. Christopherson WA. Pipelle endometrial sampling.
4. Spencer CP, Whitehead MI. Endometrial assessment Sensitivity in the detection of endometrial cancer. J
re-visited. Br J Obstet Gynaecol. 1999;106:623–32. Reprod Med. 1995;40:553–5.
5. Ferry J, Farnsworth A, Webster M, Wren B. The effi- 20. Barnhart KT, Gracia CR, Reindl B, Wheeler
cacy of the Pipelle endometrial biopsy in detecting JE. Usefulness of Pipelle endometrial biopsy in the
endometrial carcinoma. Aust N Z J Obstet Gynaecol. diagnosis of women at risk for ectopic pregnancy.
1993;33:76–8. Am J Obstet Gynecol. 2003;188:906–9.
6. Stovall TG, Photopulos GJ, Poston WM, Ling 21. Ong S, Duffy T, Lenehan P, Murphy J. Endometrial
FW, Sandles LG. Pipelle endometrial sampling in Pipelle biopsy compared to conventional dilatation
patients with known endometrial carcinoma. Obstet and curettage. Ir J Med Sci. 1997;166:47–9.
Gynecol. 1991;77:954–6. 22. Ries A, Singson P, Bidus M, Barnes JG. Use of the
7. Koonings PP, Moyer DL, Grimes DA. A randomized endometrial Pipelle in the diagnosis of early abnor-
clinical trial comparing Pipelle and Tis-u-trap for mal gestations. Fertil Steril. 2000;74:593–5.
endometrial biopsy. Obstet Gynecol. 1990;75:293–5. 23. Seto MTY, Ip PPC, Ngu S-F, Cheung ANY, Pun
8. Kaunitz AM, Masciello A, Ostrowski M, Rovira T-C. Positive predictive value of endometrial polyps
EZ. Comparison of endometrial biopsy with the in Pipelle aspiration sampling: a histopathological
endometrial Pipelle and Vabra aspirator. J Reprod study of 195 cases. Eur J Obstet Gynecol Reprod
Med. 1988;33:427–31. Biol. 2016;203:12–5.
9. Hill GA, Herbert CM, Parker RA, Wentz 24. Larson DM, Johnson KK, Broste SK, Krawisz BR,
AC. Comparison of late luteal phase endometrial Kresl JJ. Comparison of D&C and office endometrial
biopsies using the Novak curette or Pipelle endome- biopsy in predicting final histopathologic grade in
trial suction curette. Obstet Gynecol. 1989;73:443–5. endometrial cancer. Obstet Gynecol. 1995;86:38–42.
10. Eddowes HA, Read MD, Codling BW. Pipelle: a 25. Agostini A, Shojaï R, Cravello L, Rojat-Habib MC,
more acceptable technique for outpatient endome- Roger V, Bretelle F, Blanc B. Endometrial biopsy
trial biopsy. Br J Obstet Gynaecol. 1990;97:961–2. during outpatient hysteroscopy: evaluation and com-
11. Silver MM, Miles P, Rosa C. Comparison of Novak parison of two devices. Eur J Obstet Gynecol Reprod
and Pipelle endometrial biopsy instruments. Obstet Biol. 2001;97:220–2.
Gynecol. 1991;78:828–30. 26. Huang GS, Gebb JS, Einstein MH, Shahabi S,
12. Stovall TG, Ling FW, Morgan PL. A prospective, Novetsky AP, Goldberg GL. Accuracy of preop-
randomized comparison of the Pipelle endome- erative endometrial sampling for the detection
References 369
of high-
grade endometrial tumors. Am J Obstet 42. Sant Cassia LJ, Weppelmann B, Shingleton H, Soong
Gynecol. 2007;196:243.e1-5. SJ, Hatch K, Salter MM. Management of early endo-
27. Critchley HOD, Warner P, Lee AJ, Brechin S, Guise metrial carcinoma. Gynecol Oncol. 1989;35:362–6.
J, Graham B. Evaluation of abnormal uterine bleed- 43. Obermair A, Geramou M, Gücer F, Denison U, Graf
ing: comparison of three outpatient procedures AH, Kapshammer E, Medl M, Rosen A, Wierrani
within cohorts defined by age and menopausal sta- F, Neunteufel W, Frech I, Speiser P, Kainz C,
tus. Health Technol Assess. 2004;8:1–139. Breitenecker G. Endometrial cancer: accuracy of the
28. Carlson SM, Goldberg J, Lentz DM. Endoscopy: finding of a well differentiated tumor at dilatation
hysteroscopy and laparoscopy. In: Lobo RA, and curettage compared to the findings at subsequent
Gershenson DM, Lentz GM, Valea FA, editors. hysterectomy. Int J Gynecol Cancer. 1999;9:383–6.
Comprehensive gynecology. 7th ed. Philadelphia: 44. Møller LM, Berget A. Prehysterectomy curettage
Elsevier; 2017. p. 190–204. in women with uterine fibromyomata is not worth-
29. Grimes DA. Diagnostic dilation and curettage: a while. Acta Obstet Gynecol Scand. 1993;72:374–6.
reappraisal. Am J Obstet Gynecol. 1982;142:1–6. 45. Hinchcliff EM, Esselen KM, Watkins JC, Oduyebo
30. Smith JJ, Schulman H. Current dilatation and curet- T, Rauh-Hain JA, Del Carmen MG, Quade BJ, Muto
tage practice: a need for revision. Obstet Gynecol. MG. The role of endometrial biopsy in the preopera-
1985;65:516–8. tive detection of uterine leiomyosarcoma. J Minim
31. van Hanegem N, Prins MMC, Bongers MY, Opmeer Invasive Gynecol. 2016;23:567–72.
BC, Sahota DS, Mol BWJ, Timmermans A. The 46. Bansal N, Herzog TJ, Burke W, Cohen CJ, Wright
accuracy of endometrial sampling in women with JD. The utility of preoperative endometrial sam-
postmenopausal bleeding: a systematic review and pling for the detection of uterine sarcomas. Gynecol
meta-analysis. Eur J Obstet Gynecol Reprod Biol. Oncol. 2008;110:43–8.
2016;197:147–55. 47. Duplantier N, Finan MA, Barbe T. Necessity of
32. Nickelsen C. Diagnostic and curative value of endometrial biopsy in women with enlarged uteri
uterine curettage. Acta Obstet Gynecol Scand. and a preoperative diagnosis of uterine leiomyo-
1986;65:693–7. mata. J Reprod Med. 2003;48:23–7.
33. Schlaerth JB, Morrow CP, Rodriguez M. Diagnostic 48. Clark TJ, Voit D, Gupta JK, Hyde C, Song F, Khan
and therapeutic curettage in gestational trophoblastic KS. Accuracy of hysteroscopy in the diagnosis of
disease. Am J Obstet Gynecol. 1990;162:1465–70. endometrial cancer and hyperplasia: a systematic
discussion 1470–1471 quantitative review. JAMA. 2002;288:1610–21.
34. Daniel AG, Peters WA. Accuracy of office and oper- 49. Gkrozou F, Dimakopoulos G, Vrekoussis T,
ating room curettage in the grading of endometrial Lavasidis L, Koutlas A, Navrozoglou I, Stefos T,
carcinoma. Obstet Gynecol. 1988;71:612–4. Paschopoulos M. Hysteroscopy in women with
35. Berkowitz RS, Desai U, Goldstein DP, Driscoll abnormal uterine bleeding: a meta-analysis on
SG, Marean AR, Bernstein MR. Pretreatment four major endometrial pathologies. Arch Gynecol
curettage-A predictor of chemotherapy response in Obstet. 2015;291:1347–54.
gestational trophoblastic neoplasia. Gynecol Oncol. 50. Svirsky R, Smorgick N, Rozowski U, Sagiv R,
1980;10:39–43. Feingold M, Halperin R, Pansky M. Can we rely
36. Cowles TA, Magrina JF, Masterson BJ, Capen on blind endometrial biopsy for detection of focal
CV. Comparison of clinical and surgical-staging intrauterine pathology? Am J Obstet Gynecol.
in patients with endometrial carcinoma. Obstet 2008;199:115.e1–3.
Gynecol. 1985;66:413–6. 51. Spadoto-Dias D, Bueloni-Dias FN, Elias LV, Leite
37. Bettocchi S, Ceci O, Vicino M, Marello F, Impedovo NJ, Modotti WP, Lasmar RB, Dias R. The value of
L, Selvaggi L. Diagnostic inadequacy of dilatation hysteroscopic biopsy in the diagnosis of endometrial
and curettage. Fertil Steril. 2001;75:803–5. polyps. Womens Health (Lond). 2016;12:412–9.
38. Stock RJ, Kanbour A. Prehysterectomy curettage. 52. Daniell JF, Kurtz BR, Ke RW. Hysteroscopic endo-
Obstet Gynecol. 1975;45:537–41. metrial ablation using the rollerball electrode. Obstet
39. Epstein E, Ramirez A, Skoog L, Valentin Gynecol. 1992;80:329–32.
L. Dilatation and curettage fails to detect most focal 53. Munro MG. Endometrial ablation. Best Pract Res
lesions in the uterine cavity in women with post- Clin Obstet Gynaecol. 2018;46:120–39.
menopausal bleeding. Acta Obstet Gynecol Scand. 54. Grimbizis GF, Tsolakidis D, Mikos T, Anagnostou
2001;80:1131–6. E, Asimakopoulos E, Stamatopoulos P, Tarlatzis
40. Soothill PW, Alcock CJ, MacKenzie IZ. Discrepancy BC. A prospective comparison of transvaginal ultra-
between curettage and hysterectomy histology in sound, saline infusion sonohysterography, and diag-
patients with stage 1 uterine malignancy. Br J Obstet nostic hysteroscopy in the evaluation of endometrial
Gynaecol. 1989;96:478–81. pathology. Fertil Steril. 2010;94:2720–5.
41. Piver MS, Lele SB, Barlow JJ, Blumenson 55. Isaacson K. Office hysteroscopy: a valuable but
L. Paraaortic lymph node evaluation in stage I endo- under-utilized technique. Curr Opin Obstet Gynecol.
metrial carcinoma. Obstet Gynecol. 1982;59:97–100. 2002;14:381–5.
56. Salazar CA, Isaacson K. Office operative hyster- cinoma of the endometrium. Acta Obstet Gynecol
oscopy: an update. J Minim Invasive Gynecol. Scand. 2003;82:378–84.
2017;25:199–208. 70. Giorda G, Crivellari D, Veronesi A, Perin
57. Elyashiv O, Sagiv R, Kerner R, Keidar R, Menczer T, Campagnutta E, Carbone A, Scarabelli
J, Levy T. Hysteroscopic resection of premalignant C. Comparison of ultrasonography, hysteroscopy,
and malignant endometrial polyps: is it a safe alter- and biopsy in the diagnosis of endometrial lesions
native to hysterectomy? J Minim Invasive Gynecol. in postmenopausal tamoxifen-treated patients. Acta
2017;24:1200–3. Obstet Gynecol Scand. 2002;81:975–80.
58. van Hanegem N, Breijer MC, Slockers SA, 71. Panda JK. One-stop clinic for postmenopausal
Zafarmand MH, Geomini P, Catshoek R, Pijnenborg bleeding. J Reprod Med. 2002;47:761–6.
J, van der Voet LF, Dijkhuizen F, van Hoecke G, 72. Gull B, Karlsson B, Milsom I, Granberg S. Can
Reesink-Peters N, Veersema S, van Hooff M, van ultrasound replace dilation and curettage? A longi-
Kesteren P, Huirne JA, Opmeer BC, Bongers MY, tudinal evaluation of postmenopausal bleeding and
Mol B, Timmermans A. Diagnostic workup for post- transvaginal sonographic measurement of the endo-
menopausal bleeding: a randomised controlled trial. metrium as predictors of endometrial cancer. Am J
BJOG. 2017;124:231–40. Obstet Gynecol. 2003;188:401–8.
59. Einerth Y. Vacuum curettage by the Vabra method. 73. Davidson KG, Dubinsky TJ. Ultrasonographic eval-
A simple procedure for endometrial diagnosis. Acta uation of the endometrium in postmenopausal vagi-
Obstet Gynecol Scand. 1982;61:373–6. nal bleeding. Radiol Clin N Am. 2003;41:769–80.
60. Wu HH-J, Casto BD, Elsheikh TM. Endometrial 74. Giusa-Chiferi MG, Gonçalves WJ, Baracat EC,
brush biopsy. An accurate outpatient method of de Albuquerque Neto LC, Bortoletto CC, de Lima
detecting endometrial malignancy. J Reprod Med. GR. Transvaginal ultrasound, uterine biopsy and
2003;48:41–5. hysteroscopy for postmenopausal bleeding. Int J
61. Del Priore G, Williams R, Harbatkin CB, Wan LS, Gynaecol Obstet. 1996;55:39–44.
Mittal K, Yang GC. Endometrial brush biopsy for 75. Langer RD, Pierce JJ, O’Hanlan KA, Johnson SR,
the diagnosis of endometrial cancer. J Reprod Med. Espeland MA, Trabal JF, Barnabei VM, Merino MJ,
2001;46:439–43. Scully RE. Transvaginal ultrasonography compared
62. Rogers PA, Polson D, Murphy CR, Hosie M, Susil with endometrial biopsy for the detection of endo-
B, Leoni M. Correlation of endometrial histology, metrial disease. Postmenopausal estrogen/progestin
morphometry, and ultrasound appearance after dif- interventions trial. N Engl J Med. 1997;337:1792–8.
ferent stimulation protocols for in vitro fertilization. 76. Smith-Bindman R, Kerlikowske K, Feldstein VA,
Fertil Steril. 1991;55:583–7. Subak L, Scheidler J, Segal M, Brand R, Grady
63. Fleischer AC, Gordon AN, Entman SS, Kepple D. Endovaginal ultrasound to exclude endometrial
DM. Transvaginal scanning of the endometrium. J cancer and other endometrial abnormalities. JAMA.
Clin Ultrasound. 1990;18:337–49. 1998;280:1510–7.
64. Khalifa E, Brzyski RG, Oehninger S, Acosta AA, 77. Weber AM, Belinson JL, Bradley LD, Piedmonte
Muasher SJ. Sonographic appearance of the endo- MR. Vaginal ultrasonography versus endometrial
metrium: the predictive value for the outcome of in- biopsy in women with postmenopausal bleeding.
vitro fertilization in stimulated cycles. Hum Reprod. Am J Obstet Gynecol. 1997;177:924–9.
1992;7:677–80. 78. Fleischer AC, Abramowicz J, Goncalves L,
65. Dickey RP, Olar TT, Curole DN, Taylor SN, Rye Manning FA, Monteagudo A, Toy E, Tritsch
PH. Endometrial pattern and thickness associated I. Transvaginal sonography of endometrial disor-
with pregnancy outcome after assisted reproduction ders. In: Sonography in obstetrics & gynecology:
technologies. Hum Reprod. 1992;7:418–21. principles and practice. 8th ed. New York: McGraw
66. Sheth S, Hamper UM, Kurman RJ. Thickened Hill Professional; 2017. p. 995–1014.
endometrium in the postmenopausal woman: 79. Patel V, Wilkinson EJ, Chamala S, Lu X, Castagno
sonographic-pathologic correlation. Radiology. J, Rush D. Endometrial thickness as measured by
1993;187:135–9. transvaginal ultrasound and the corresponding histo-
67. Dørum A, Kristensen GB, Langebrekke A, Sørnes pathologic diagnosis in women with postmenopausal
T, Skaar O. Evaluation of endometrial thickness bleeding. Int J Gynecol Pathol. 2017;36:348–55.
measured by endovaginal ultrasound in women with 80. Goldchmit R, Katz Z, Blickstein I, Caspi B, Dgani
postmenopausal bleeding. Acta Obstet Gynecol R. The accuracy of endometrial Pipelle sampling
Scand. 1993;72:116–9. with and without sonographic measurement of endo-
68. Doherty CM, Silver B, Binor Z, Molo MW, metrial thickness. Obstet Gynecol. 1993;82:727–30.
Radwanska E. Transvaginal ultrasonography and 81. ACOG Committee Opinion No. 734: the role of
the assessment of luteal phase endometrium. Am transvaginal ultrasonography in evaluating the endo-
J Obstet Gynecol. 1993;168:1702–7. discussion metrium of women with postmenopausal bleeding.
1707–1709 Obstet Gynecol. 2018;131:e124–9.
69. Iha T, Shen H, Kanazawa K. Hysteroscopy to detect 82. Laiyemo R, Dudill W, Jones SE, Browne H. Do post-
stage IA well-differentiated endometrioid adenocar- menopausal women with thickened endometrium on
References 371
trans-vaginal ultrasound in the absence of vaginal 92. Vikhareva Osser O, Valentin L. Risk factors for
bleeding need hysteroscopic assessment? A pilot incomplete healing of the uterine incision after cae-
study. J Obstet Gynaecol. 2016;36:223–6. sarean section. BJOG. 2010;117:1119–26.
83. Epstein E, Fischerova D, Valentin L, Testa AC, 93. Fleischer AC, Abramowicz J, Goncalves L, Manning
Franchi D, Sladkevicius P, Frühauf F, Lindqvist FA, Monteagudo A, Toy E, Tritsch I. Sonography of
PG, Mascilini F, Fruscio R, Haak LA, Opolskiene trophoblastic diseases. In: Sonography in obstet-
G, Pascual MA, Alcazar JL, Chiappa V, Guerriero rics & gynecology: principles and practice. 8th
S, Carlson J, Van Holsbeke C, FPG L, De Moor B, ed. New York: McGraw Hill Professional; 2017.
Bourne T, van Calster B, Installe A, Timmerman p. 899–908.
D, Verbakel JY, Van den Bosch T. Ultrasound 94. Fowler DJ, Lindsay I, Seckl MJ, Sebire NJ. Routine
characteristics of endometrial cancer as defined pre-evacuation ultrasound diagnosis of hydatidiform
by the International Endometrial Tumor Analysis mole: experience of more than 1000 cases from a
(IETA) consensus nomenclature - A prospective regional referral center. Ultrasound Obstet Gynecol.
multicenter study. Ultrasound Obstet Gynecol. 2006;27:56–60.
2018;51(6):818–28. 95. Lange RC, Duberg AC, McCarthy SM. An evalu-
84. Nieuwenhuis LL, Hermans FJ, de Vaate AJM B, ation of MRI contrast in the uterus using synthetic
Leeflang MM, Brölmann HA, Hehenkamp WJ, BWJ imaging. Magn Reson Med. 1991;17:279–84.
M, Clark TJ, Huirne JA. Three-dimensional saline 96. Brown HK, Stoll BS, Nicosia SV, Fiorica JV,
infusion sonography compared to two-dimensional Hambley PS, Clarke LP, Silbiger ML. Uterine junc-
saline infusion sonography for the diagnosis of focal tional zone: correlation between histologic findings
intracavitary lesions. Cochrane Database Syst Rev. and MR imaging. Radiology. 1991;179:409–13.
2017;5:CD011126. 97. Scoutt LM, Flynn SD, Luthringer DJ, McCauley
85. Bittencourt CA, Dos Santos SR, Bernardo WM, TR, McCarthy SM. Junctional zone of the uterus:
Fuchs LFP, Soares Júnior JM, Pastore AR, Baracat correlation of MR imaging and histologic exami-
EC. Accuracy of saline contrast sonohysterogra- nation of hysterectomy specimens. Radiology.
phy in detection of endometrial polyps and submu- 1991;179:403–7.
cosal leiomyomas in women of reproductive age 98. Ascher SM, Reinhold C. Imaging of cancer of the
with abnormal uterine bleeding: systematic review endometrium. Radiol Clin N Am. 2002;40:563–76.
and meta-analysis. Ultrasound Obstet Gynecol. 99. Dueholm M, Lundorf E, Olesen F. Imaging tech-
2017;50:32–9. niques for evaluation of the uterine cavity and
86. Maheux-Lacroix S, Li F, Laberge PY, Abbott endometrium in premenopausal patients before
J. Imaging for polyps and leiomyomas in women minimally invasive surgery. Obstet Gynecol Surv.
with abnormal uterine bleeding: a systematic review. 2002;57:388–403.
Obstet Gynecol. 2016;128:1425–36. 100. Frei KA, Kinkel K. Staging endometrial cancer:
87. Phillips CH, Wortman JR, Ginsburg ES, Sodickson role of magnetic resonance imaging. J Magn Reson
AD, Doubilet PM, Khurana B. First-trimester emer- Imaging. 2001;13:850–5.
gencies: a radiologist’s perspective. Emerg Radiol. 101. Kim SH, Kim HD, Song YS, Kang SB, Lee
2018;25:61–72. HP. Detection of deep myometrial invasion in endo-
88. Barnhart KT, Fay CA, Suescum M, Sammel MD, metrial carcinoma: comparison of transvaginal
Appleby D, Shaunik A, Dean AJ. Clinical factors ultrasound, CT, and MRI. J Comput Assist Tomogr.
affecting the accuracy of ultrasonography in symp- 1995;19:766–72.
tomatic first-trimester pregnancy. Obstet Gynecol. 102. Barton JW, McCarthy SM, Kohorn EI, Scoutt LM,
2011;117:299–306. Lange RC. Pelvic MR imaging findings in gesta-
89. Fleischer AC, Abramowicz J, Goncalves L, Manning tional trophoblastic disease, incomplete abortion,
FA, Monteagudo A, Toy E, Tritsch I. Transvaginal and ectopic pregnancy: are they specific? Radiology.
sonography of ectopic pregnancy. In: Sonography 1993;186:163–8.
in obstetrics & gynecology: principles and practice. 103. Valentin L. Imaging in gynecology. Best Pract Res
8th ed. New York: McGraw Hill Professional; 2017. Clin Obstet Gynaecol. 2006;20:881–906.
p. 81–108. 104. Buckley H, Fox H. Biopsy pathology of the endome-
90. Brown DL, Doubilet PM. Transvaginal sonography trium. 2nd ed. London: Arnold; 2002.
for diagnosing ectopic pregnancy: positivity criteria 105. Kepes JJ, Oswald O. Tissue artefacts caused by
and performance characteristics. J Ultrasound Med. sponge in embedding cassettes. Am J Surg Pathol.
1994;13:259–66. 1991;15:810–2.
91. Godin PA, Bassil S, Donnez J. An ectopic pregnancy
developing in a previous caesarian section scar.
Fertil Steril. 1997;67:398–400.
Index
A Atypical/abnormal glandular cells of undetermined

Abnormal, nonmenstrual bleeding patterns, 138, 139 significance (AGUS), 2
Abnormal uterine bleeding (AUB), 2–5, 164 Atypical hyperplasia of endometrium, 229, 231, 232,
abnormal secretory phase, 139 235, 237–241, 244, 248, 254, 255
clinical queries and reporting, 139, 141 Atypical mucinous glandular proliferations, 283
glandular and stromal breakdown, 121 Atypical polyp, 219
irregular shedding with mixed phase pattern, 140 Atypical polypoid adenomyoma (APA), 207–209, 236
non-structural causes
with coagulopathies, 126
cyclic ovulatory menstrual bleeding, 127 B
estrogen-related, 128 Beckwith Wiedemann syndrome, 91
iatrogenic causes, 126 Benign mesenchymal tumors, 352
ovulatory dysfunction, 126, 127 Biopsy-induced artifacts, 1
in perimenopausal and perimenarcheal women, Bizarre leiomyoma in curettings, 337
126 Blastocyst implantation, 45
progesterone-related bleeding, 128 Blaustein’s trophogram, 106
polyps (AUB-P), 2
in postmenopausal years, 4
in reproductive years, 3 C
submucosal leiomyoma and coagulopathy (AUB- Cancer Genome Atlas (TCGA) genomic-based
L(SM)), 2 classification of endometrial carcinoma, 263
Abnormal (nonmolar) villous lesions, 90, 91 Carcinosarcoma, 301–305, 308–312
Abortion, 53 differential diagnosis, 305–308
hydropic change and pathologic changes, 59 Cellular cytoplasmic changes (metaplasia), 256
Actinomycosis-associated inflammation, 182 Cellular leiomyomas, 346
Actinomycotic endometritis, 192, 193 Cervical contaminants, 28, 33
Acute endometrial inflammation, 174 Cervical–vaginal cytologic specimen, 2
Adenocarcinoma, 4 Chemotherapy-related bizarre nuclear atypia, 183
Adenofibroma, 347 Chlamydia endometritis, 189
cytologic features, 347 Chlamydia trachomatis infection, 179
differential diagnosis, 350 Choriocarcinoma, 96–99, 106
Adenomyomatous pattern, 207 diagnosis, 95, 97
Adenosarcoma, 347–349 differential diagnosis, 96
Adenosquamous carcinoma of the endometrium, 314 features, 94
Age and menstrual/menopausal status, 3 H&E diagnosis, 95
Altered fibrous/desmoplastic stroma, 265 hemorrhagic and necrotic, 95
Ancillary testing, 86 intrauterine villi, 96
Anovulatory cycles, 127, 129 non-gestational, 94
Antiprogestin RU486, 162, 163 polymerase chain reaction amplification, 97
Arias-Stella reaction, 45–47 in postmenopausal patient, 100
Artifactually crowded secretory glands, 251 Chorionic villi, 57, 58
Aspiration devices, 365 and villous trophoblast after first trimester, 60
Atrophy, 133, 134, 137 Chromosome in situ hybridization analysis, 94
polyps, 206, 211 Chronic anovulation, 129

https://doi.org/10.1007/978-3-319-98608-1
374 Index
Chronic endometritis, 6 clinical queries and reporting, 319–321

Chronic inflammation with plasma cells, 236 diagnosis, 264
Ciliated carcinoma, 275 with endocervical-like pattern, 289
Ciliated cell change (tubal metaplasia), 243 FIGO staging system, 309, 313
Circumferential trophoblastic hyperplasia, 79 foam cells, 276
Clear cell carcinoma, 262, 296, 297, 301–303 glandular pattern, 274
glandular pattern, 299 grading, 269
hobnail metaplastic change, 304 morphologic diversity, 263
papillary pattern, 298 nuclear grading, 276
solid pattern, 300 pathogenesis, 261
Clear cell endometrial intraepithelial carcinoma, 253 quantitative features, 264
Clomiphene citrate, 160, 161 TNM Staging Systems, 313
Combined estrogen and progestin, replacement therapy type I, 262
for menopausal women, 153–155 type II, 262
Complete hydatidiform mole (CHM), 76, 77, 79–81 villoglandular, 275
vs. PHM, 85–87 WHO histologic classification, 267, 268
Confluent gland pattern, 264 Endometrial dating
Cowden syndrome (CS), 288, 289 early secretory endometrium, 10
Cycle of conception, 40, 41 histologic, 34
Cytogenetic studies, 83, 85 late secretory endometrium, 10
Cytologic atypia and proliferation, 112 mid secretory endometrium, 10
Cytomegalovirus infection, endometritis, 183 Noyes criteria for, 10
Cytoplasmic changes, 249 Endometrial vs. endocervical carcinoma, 310, 314–316
Endometrial glands and stroma in pregnancy, 40, 43, 45
Endometrial hyperplasia, 255
D artifactual changes to glands, 234
Danazol, 161 benign and malignant lesions, 234
Dating, endometrial and carcinoma, 3
early secretory endometrium, 10 secretory change, 250
histologic, 34 WHO classification, 226
late secretory endometrium, 10 with atypia (atypical hyperplasia), 229, 232, 234
mid secretory endometrium, 10 without atypia, 226, 228, 230, 233
Noyes criteria for, 10 with superimposed ciliated cell change, 244
Decidualized stroma, 43 Endometrial mucinous glandular proliferations, 283
Decidual/pregnancy-like pattern, 148, 149 Endometrial polyps, 158, 235
Diffuse hyperplasia, 211 adenomyomatous, 201
Diffuse large B-cell lymphoma, 353 atrophic, 200, 212
Dilation and curettage (D&C) technique, 364 with atypical stromal cells, 205
Disordered proliferative endometrium, 235 benign, 201–204
and persistent proliferative phase, 132, 133 clinical queries and reporting, 216, 217
DNA genotyping, 91 crowded glands, 218
DNA-MMR protein immunohistochemistry, 286 cytologic atypia, 204
Dyssynchronous endometrium, 34 differential diagnosis, 210, 211
with ectatic vessels, 206
focal glandular and stromal breakdown, 204
E forms, 200
Early complete hydatidiform mole, 79 functional polyps, 200–201, 212
Early gestational endometrium, 40–42 histologic features, 201
with decidualized stroma, 43 metastatic tumors, 215
Ectopic pregnancy, 62, 64 microscopic abnormalities, 200
Endocervical and endometrial-type gland development, mixed endometrial–endocervical, 201
206–207 ovulatory dysfunction, 212
Endometrial ablation therapy, 184 pedunculated/sessile growths, 200
Endometrial adhesions, 215 proliferative/hyperplastic, 200, 211
Endometrial basalis, 216 proliferative/secretory, 204
Endometrial biopsies, 363, 364 p16 immunoreactivity, 202
Endometrial carcinoma with squamous change, 207
aggressive forms, 267 surface and glandular epithelium, 204
ciliated carcinoma, 275 Endometrial receptivity array (ERA) test, 35
classification, 270 Endometrial sampling technique, 363–365
Index 375
conception products, 4 proliferative phase endometrium, 13–16

techniques, 1 reporting of, 32–35
Endometrial signet ring cell carcinoma, 307 secretory phase endometrium
Endometrial stromal nodule (ESN), 340, 341, 343 characteristics of, 16
Endometrial stromal sarcomas, 342, 345 early, 17, 18
Endometrioid carcinoma histologic features of, 16
FIGO grade 1, 266–270, 277, 279, 280 interval endometrium, 17
FIGO grade 2, 271 late phase, 17–20, 22
FIGO grade 3, 272 mid phase, 17, 21
with sex cord-like formations and hyalinization, 286 morphological changes, 16
with squamous differentiation, 282–285 undatable endometrium, causes of, 35
Endometrioid intraepithelial neoplasia (EIN), 230–233, weakly proliferative endometrium, 26
239, 240 Eosinophilic cell change, 244–246, 248
Endometritis Eosinophilic syncytial change, 248
abnormal glandular development, 178 Epithelial cytoplasmic alterations (metaplasia), 241–243,
asymptomatic, 173 245, 246, 248
atypia, 238 Epithelioid trophoblastic tumor (ETT), 51, 107–110
clinical queries and reporting, 191, 195 Estrogen hormones, 146
differential diagnosis, 185, 188–191 Estrogen–progestin therapy, 159
diffuse, 174 Estrogen-related bleeding
eosinophils, 176 atrophy, 133, 134
epithelial changes, 178 disordered proliferative, 132
focal glandular and stromal breakdown, 174, 179, estradiol production, 130
235 focal glandular dilation, 132
granulomatous inflammation, 180, 181 glandular and stromal breakdown, 130–131
histiocytes, 177 in perimenarcheal adolescents, 129
histologic findings, 174 European Working Group (EWG)
inflammation plasma cells, 174, 176, 177 classification, 240
morphologic features, 174, 179 Exaggerated placental site (EPS), 50, 84, 111
neutrophils, 175 Extensive papillary growth pattern, 265, 267
nonspecific chronic, 175, 176, 178
oligosymptomatic, 173
pelvic inflammatory disease, 173 F
prevalence, 174 Familial recurrent hydatidiform mole (FRHM), 76
reactive cellular changes, 178 Fertilization, 40
with severe cervical stenosis, 176 Fibrinoid (Rohr’s stria), 63
stromal changes, 177 First trimester placental development, 64
Endometrium Focal papillary proliferation, benign endometrial polyp,
artifacts 237
fragmentation, 27, 31 Foreign body giant cell reaction with polarizable
pseudolipomatosis, 30, 34 material, 192
telescoping, 27, 33 Foreign body granulomatous reaction, 183
atrophic, 6, 26, 29 Fragmentation and artifacts, 319
cervical contaminants, 28, 33 Frozen sections on endometrial biopsies, 367, 368
definition, 9 Fumarate hydratase deficient leiomyoma, 336
ectopic pregnancy, 62, 64 Functional polyp, 206, 213
fallopian tube prolapse, 31
histiocytes, 29
histological structure G
basalis, 11 Gastric-type endocervical adenocarcinoma, 315
functionalis, 11–12 Genotyping analysis, 94
lower uterine segment, 12 Gestational endometrium, 44, 45
surface epithelium, 10, 11 with optically clear nuclei, 48
to hormonal therapy, 1 Gestational trophoblastic disease (GTD)
inactive endometrium, 26 classification, 75
in infertile patients, 1 modified WHO classification, 76
irregular secretory patterns, 31–32 pathologic forms, 75
in later pregnancy, 41, 59 placental development disorders, 75
menstrual endometrium, 20, 25 Gestational trophoblastic neoplasms (GTN), 46, 47, 67,
morphologic changes, 1 75, 94
376 Index
Glandular and stromal breakdown, 122 I

abnormal bleeding pattern, 123 Immature chorionic villi, 64
abnormal secretory phase, 138 with hydropic change, 65, 66
anovulatory bleeding pattern, 133 with trophoblast, 54
with artifactual crowding, 131 with villous fibrosis, 68
and bleeding, 122 Immunohistochemical analysis, 292
disordered proliferative phase pattern, 136 Immunohistochemistry coupled with molecular
with eosinophilic syncytial change, 123, 127–129 (genome-based) studies, 262
fibrin thrombi, 126 Inactive pattern, progestin effect, 150, 151
hemorrhage, stroma, 122 Incidental intraluminal neutrophils, 194
with hobnail change, 130 Indications, endometrial biopsy/curettage, 2
in menstrual and abnormal bleeding, 122 Infarcted endometrial polyp, 209
nuclear debris, 122 Infertility biopsy, 4
proliferative endometrium, 124, 125 Infertility therapy, 165
tissue fragmentation, 124 Inflammatory myofibroblastic tumor (IMT), 351
Glandular nuclear atypia in polyp, 210 Intermediate trophoblastic (IT) cells
Gonadotropin-releasing hormone (GnRH) chorionic-type, 52
agonists, 162 cytokeratin, 52
Gynecologic Pathology Interest Group of the Canadian and decidua, 53, 60
Association of Pathologists (GPIG-CAP) in first trimester, 50
guidelines, 10 in implantation sites, 51
in myometrium, 61
spiral arteries, 55
H International Federation of Gynecology and Obstetrics
Hereditary cancer syndromes, 3 (FIGO), 2
Hereditary nonpolyposis colorectal cancer (Lynch Interval endometrium, 17
syndrome), 285, 286, 288 Intrauterine adhesion, 220
Herpes virus (HSV), 184 Intrauterine pregnancy, 65
High-grade endometrial stromal sarcoma (HGESS), 340, Intrauterine products of conception, 39
344 Irregular secretory endometrium, 136
Histiocytes, 194 Irregular shedding, 137
endometrial, 29–30
Histologic evaluation, 366, 367
Hobnail cell change, 250 L
Hobnail-like artifact of proliferative endometrium, 238 Last menstrual period (LMP), 4
Hormonal contraception, 147 Leiomyomas
Hormone replacement therapy, 200 effect of, 335
Hormone therapies, 4, 5, 163 spindle cell process, 335
Human menopausal gonadotropins (hMG/menotropins), variants of, 334
162 Leiomyosarcoma, 336–338
Hydatidiform mole (HM), 78–83, 93 Levonorgestrel-releasing intrauterine device, 147
algorithmic approach, 90 Localized atypical-appearing endometrial glandular
clinical queries and reporting, 93, 94 proliferations, 48
FIGO/WHO scoring system, 95 Low-grade endometrial stromal sarcoma (LGESS), 340,
vs. hydropic abortus 341, 343, 344
abnormal (nonmolar) villous lesions, 88–90 Luteal phase abnormalities, 31, 136
clinical and cytogenetic comparison, 77 Lymphoma/leukemic infiltration, endometrium, 351, 352
invasive, 92
management, 94
pathologic features, 78 M
persistent trophoblast, 92 Magnetic resonance imaging (MRI), 366
PHM/CHM, 76 Malignant mixed Mullerian tumor (MMMT), 251, 301
subtypes, 76 Menstrual endometrium, 20–21
Hypermutated/microsatellite instability carcinomas, 263 Metabolic syndrome, 199
Hyperplasia, 3–5 Metastatic breast carcinoma, 317
atypical, 200 Metastatic colon adenocarcinoma, 317
progestin therapy, 164 Metastatic ovarian carcinoma, 317
Hyperplasia without atypia, 240 Microglandular (endocervical-like) endometrial
Hyperplastic papillary proliferations, 204 carcinoma, 283, 284
Hysteroscopy and curettage, 364, 365 Microsatellite instability (MSI) analysis, PCR, 287
Index 377
Microsatellite stable (MSS) carcinomas, 263 POLE-ultramutated carcinomas, 263

Microscopic analysis, 2 Polycystic ovarian syndrome (PCOS), 4
Mid-secretory phase endometrium, 17 Polypoid adhesion, 219
Mifepristone, 162, 163 Polypoid endometritis, 217
Minimal uterine serous carcinoma (MUSC), 251, 294 Polypoid exogenous hormonal effect, 218
Missed abortion, 67 Post-intrauterine device endometritis, 191
Mucinous carcinoma of endometrium, 281, 282 Postmenopausal hormone replacement, 163
Mucinous change, 246, 249 Post-partum deciduitis, 181
Mycoplasma, 185 Postprocedural foreign body giant cell reaction, 182
Myometrial invasion, 239 Pregnancy
advancing gestational age, 41
complications, 40
N endometrial glands and stroma, 40, 43, 45
Necrotic debris, 125 histologic changes, 42
Next-generation sequencing, 350 grandular changes, 45
Nodular histiocytic hyperplasia, 29–30 history, 5
Nonatypical hyperplasia, 235, 256 Pregnancy-related acute inflammation, 176
Non-hyperplastic endometrium, 237 Primary squamous cell carcinoma of endometrium, 297,
Nonkeratinizing squamous change, 243 298, 300
Non-neoplastic lesions, 110–112 Progesterone associated endometrial changes (PAEC),
Nonreactive endometrial stroma, 185 156, 163
Nonspecific acute and chronic endometritis, 178, 180 Progesterone-related bleeding, 135, 136
Nonspecific endometritis with reactive epithelial Progestin-like effects with no hormone use,
changes, 186 154, 155
Nonspecific endometritis with squamous epithelial Progestin-only therapy, 146
changes, 187 Progestins, 146, 147
Nontrophoblastic tumors, 98 decidual pattern, 148–150, 157
Notable nuclear atypia, 271–273 inactive pattern, 155, 156
Nuclear pleomorphism, 273 secretory pattern, 151–154
Proliferative endometrium with glandular and stromal
breakdown, 129–132
O Proliferative/hyperplastic polyps, 205
Occasional spontaneous abortion specimens, 65 Pseudoactinomycotic radiate granules, 189, 195
Office-based biopsies, 5, 6 Pseudosulfur granules, 189
Optically clear nuclei, 47
Oral contraceptives, 147, 152–153, 163
Ovarian hormone production, 128 R
Raloxifene, 160
P
Papillary proliferation, 236, 237 S
Partial hydatidiform moles (PHM), 81–83 Saline infusion sonohysterography (SIS), 365, 366
Pathology report, 67 Secretory and clear cell change, 247, 248
Periglandular cuffing, 349 Secretory carcinoma., FIGO grade 1, 278
Perivascular epithelioid cell tumor (PEComa) of uterus, Secretory pattern of progestin, 149, 150
339, 341 Secretory variant of endometrial carcinoma, 275
Persistent postmolar gestational trophoblastic disease, Selective estrogen receptor modulator
91, 92 (SERM), 156
Peutz-Jeghers syndrome, 3 Selective progesterone receptor modulators, 155, 156
Placenta accreta, 63, 71 Serous carcinoma, 290–297
Placental abnormalities, 91 Serous endometrial intraepithelial carcinoma (SEIC), 28,
Placental angiomatous malformation, 91 200, 251, 252, 254–256, 262
Placental implantation site, –57, 53, 58 differential diagnosis, 253, 254
Placental mesenchymal dysplasia, 91 invasive carcinoma, 254
Placental polyps, 63, 64 Sex cord tumor with annular tubules (SCTAT) of the
Placental site nodule (PSN), 51, 107, 111, 112 ovary, 3
Placental site trophoblastic tumor (PSTT), 51, 100–104, Smooth muscle tumor of uncertain malignant potential
106, 107 (STUMP), 335, 336
Plasmacytoid-appearing predecidualized stromal Smooth muscle tumors, 333
cells, 177 Spiral arteries, 55
378 Index
Squamous cell change in hyperplasia, 234 immunohistochemistry, 50–52

Squamous differentiation proliferation, 48
endometrial carcinoma, 278–280 Tuberculous endometritis, 190
metaplasia, 242, 320 Tumorlets (mesenchymal proliferations),
Stromal artifacts, 318 337, 338
Stromal nodules, 345 Two-grade/binary grading system, 274
Subinvoluted post-partum endometrium, 187
Subinvoluted products of conception, 188
Subinvolution, 62 U
Submucosal leiomyoma, 196 Undifferentiated carcinomas, 298, 299, 306
Superimposed ciliated cell change (tubal metaplasia), Undifferentiated uterine sarcoma (USS), 340, 345
244 Unopposed estrogenic stimulation, 146
Suppressive progestin therapy, 256 Ureaplasma urealyticum, 185
Synovial-like metaplasia, 241 Uterine adenomyomas, 207
Uterine growth, 1
Uterine perforation, 30, 34, 364
T Uterine screening with transvaginal ultrasound, 2
Tamoxifen effect with secretory change, 161 Uterine trophoblastic tumors, 99
Tamoxifen-related polyp, 160 Uterine tumor resembling ovarian sex cord tumor
Tamoxifen therapy, 156, 157, 159 (UTROSCT), 347
Tissue artifacts, 318 Uterine tumors, 334
Tissue necrosis, 152
Transvaginal ultrasound, 365, 366
Trophoblast of early pregnancy, 55, 56 V
Trophoblastic cells, 49, 50 Villous trophoblast, 51
clinical queries and reporting, 113 Villus formation, 57

Diagnosis of Endometrial Biopsies and Curettings_ a Practical Approach ( PDFDrive )

Uploaded by

Copyright:

Available Formats

Diagnosis of Endometrial Biopsies and Curettings_ a Practical Approach ( PDFDrive )

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Diagnosis of Endometrial Biopsies and Curettings_ a Practical Approach ( PDFDrive )

Uploaded by

Copyright:

Available Formats

Diagnosis of

Robert J. Kurman Michael T. Mazur

ISBN 978-3-319-98607-4 ISBN 978-3-319-98608-1 (eBook)

Library of Congress Control Number: 2018960871

© Springer Nature Switzerland AG 2019

The third edition of Diagnosis of Endometrial Biopsies and Curettings: A

s­ amples. For example, the classification of endometrial stromal tumors has

Baltimore, MD, USA Tricia A. Murdock

Hydropic Change and Other Pathologic Changes

6 Effects of Hormones������������������������������������������������������������������������ 145

Proliferative/Hyperplastic Pattern������������������������������������������������ 205

© Springer Nature Switzerland AG 2019 1

Table 1.1 Clinical terms for abnormal uterine bleeding (AUB)

reclassified as well, and terms such as menorrha-

Table 1.4 Causes of abnormal uterine bleeding in post- Infertility Biopsy

© Springer Nature Switzerland AG 2019 9

Table 2.1 Secretory phase endometrial changes

 istologic Features of Normal

predecidual transformation, the stroma contains Menstrual Endometrium

become more complex and troublesome in biop- if it is of endometrial or endocervical type.

© Springer Nature Switzerland AG 2019 39

Table 3.1 Complications of pregnancy recrudescence or accentuation of glandular

Table 3.2 Histologic changes of the endometrium in pregnancya

Trophoblast and Villi gestation, implanting trophoblast is the predomi-

smaller than implantation site IT although an

index of 25–50%, which is consistent with its Placental Implantation Site

with an even, delicate chromatin distribution. trated by intermediate trophoblast in curettage

fibrotic and hypovascular, with either no residual

bility of an ectopic pregnancy. In spontaneous endometrial patterns but there is no evidence of

Occasional spontaneous abortion specimens cally relevant. In spontaneous abortions, for

7. Mazur MT, Duncan DA, Younger JB. Endometrial

© Springer Nature Switzerland AG 2019 75

somes (androgenetic diploidy [homozygous]) [22, triploid chromosomal composition is 69XXY;

hydatidiform mole, choriocarcinoma, retained myometrium. Trophoblastic tissue without

 linical Queries and Reporting

Table 4.4 FIGO/WHO scoring system based on prognostic factors

Pathologic Features often contain erythrocytes (Figs. 4.18 and 4.19). In

Differential Diagnosis suspicious for PSTT. The Ki-67 labeling index is

HSD 3B +++ (diffuse)

Lesions of implantation site Lesions of chorionic type

Ki-67 Ki-67 Ki-67 Ki-67

EPS PSTT PSN ETT

Fig. 4.28 Trophogram- an algorithmic approach to trophoblastic proliferations

Obvious cases of either choriocarcinoma or miological and morphological study. Placenta.

site trophoblastic atypia in hydatidiform moles cytometric reevaluation of nontriploid specimens. J

80. Fukunaga M. Immunohistochemical characteriza- maternally transcribed gene p57(KIP2) (CDNK1C)

© Springer Nature Switzerland AG 2019 121

 orphologic Features of Glandular

abnormalities, luminal debris is a nonspecific find-

Estrogen-Related Bleeding 35]. Exceptions to this include obese women,

the hypothalamic-pituitary-ovarian axis such as branes and vasoconstriction with bleeding. In

endometrium, proliferative endometrium with glands because of fragmentation (see

These changes may reflect AUB-O due to Irregular Secretory Endometrium

sions, all can show patterns of abnormal

© Springer Nature Switzerland AG 2019 145

Patterns of Response cases the amount of tissue can be copious, and

c­ollective term “progesterone receptor modula-

Raloxifene Clomiphene Citrate

causes significant alterations in secretory phase Danazol

 uman Menopausal Gonadotropins/

s amples. For example, the classification of endometrial stromal tumors has

6 Effects of Hormones�� 145

Proliferative/Hyperplastic Pattern�� 205

Table 1.4 Causes of abnormal uterine bleeding in post- Infertility Biopsy

istologic Features of Normal

predecidual transformation, the stroma contains Menstrual Endometrium

Trophoblast and Villi gestation, implanting trophoblast is the predomi-

index of 25–50%, which is consistent with its Placental Implantation Site

linical Queries and Reporting

Pathologic Features often contain erythrocytes (Figs. 4.18 and 4.19). In

Differential Diagnosis suspicious for PSTT. The Ki-67 labeling index is

orphologic Features of Glandular

Estrogen-Related Bleeding 35]. Exceptions to this include obese women,

These changes may reflect AUB-O due to Irregular Secretory Endometrium

Patterns of Response cases the amount of tissue can be copious, and

collective term “progesterone receptor modula-

Raloxifene Clomiphene Citrate

causes significant alterations in secretory phase Danazol

uman Menopausal Gonadotropins/

Glandular and Stromal Breakdown Specific Infections

association with poorly formed endometrial Actinomycosis

Mycoplasma tinguished from normal lymphoid tissue of the

Atrophic Pattern functional polyp lose their orientation to surface

These polyps, like the endometrium around them, Mixed Endometrial–Endocervical

cervical and endometrial-type gland develop- typical Polypoid Adenomyoma

Table 9.3 Morphologic features of atypical hyperplasia Differential Diagnosis

Clinical Queries and Reporting

Mucinous Carcinoma Mucinous carcinomas tend to be well to

cussed in greater detail (see below, Differential Microglandular (Endocervical-Like)

Hereditary Syndromes senting cancer in greater than half of the affected

tein expressions are intact by immunohisto- Cowden Syndrome

treatment period was defined as a reduction in the Serous Carcinoma

Carcinosarcoma polypoid mass that prolapses through the cervical

Staging tumor is limited to or invades less than one half

Metastatic Carcinoma history of a known extrauterine primary tumor or

Carcinoma Mimics tion, may resemble carcinoma, especially clear

nosis of leiomyoma. Examples of this include Leiomyosarcoma

ixed Epithelial and Mesenchymal

Molecular Analysis increased cellularity of the stroma and periglan-

The clinical decision to perform endometrial Endometrial Sampling Techniques