1 Comment

Diagnosis of Endometrial Biopsies and Curettings

Diagnosis of
Endometrial Biopsies
and Curettings
Second Edition
Michael T. Mazur, MD
Clinical Professor of Pathology, State University of New York, Upstate
Medical University, Syracuse, New York, and ClearPath Diagnostics,
Syracuse, New York
Robert J. Kurman, MD
Richard W. TeLinde Distinguished Professor, Departments of
Gynecology, Obstetrics and Pathology, The Johns Hopkins Hospital,
and The Johns Hopkins University School of Medicine, Baltimore,
Maryland
Diagnosis of
Endometrial Biopsies
and Curettings
A Practical Approach
Second Edition
With 230 Illustrations, 77 in Full Color
Michael T. Mazur, MD
Clinical Professor of Pathology
State University of New York
Upstate Medical University
Syracuse, NY 13210
and
ClearPath Diagnostics
Syracuse, NY 13202
USA
Robert J. Kurman, MD
Richard W. TeLinde Distinguished
Professor
Departments of Gynecology,
Obstetrics and Pathology
The Johns Hopkins Hospital
and
The Johns Hopkins University
School of Medicine
Baltimore, MD 21231
USA
Library of Congress Cataloging-in-Publication Data
Mazur, Michael T.
Diagnosis of endometrial biopsies and curettings : a practical approach / Michael
Mazur, Robert J. Kurman.
p. ; cm.
Includes bibliographical references and index.
ISBN 0-387-98615-4 (h/c : alk. paper)
1. Endometrium—Biopsy. 2. Endometrium—Diseases—Cytodiagnosis. 3.
Endometrium—Cytopathology. I. Kurman, Robert J. II. Title.
[DNLM: 1. Uterine Diseases—diagnosis. 2. Biopsy—methods. 3. Dilatation
and Curettage—methods. 4. Endometrium—pathology. 5. Pregnancy
Complications—diagnosis. WP 440 M476d 2004]
RG316.M39 2004
618.1¢42—dc22 2004046866
ISBN 0-387-98615-4 Printed on acid-free paper.
© 1995, 2005 Springer Science+Business Media, Inc.
All rights reserved. This work may not be translated or copied in whole or in part without
the written permission of the publisher (Springer Science+Business Media, Inc., 233 Spring
Street, New York, NY 10013, USA), except for brief excerpts in connection with reviews
or scholarly analysis. Use in connection with any form of information storage and retrieval,
electronic adaptation, computer software, or by similar or dissimilar methodology now
known or hereafter developed is forbidden.
The use in this publication of trade names, trademarks, service marks, and similar terms,
even if they are not identified as such, is not to be taken as an expression of opinion as to
whether or not they are subject to proprietary rights.
While the advice and information in this book are believed to be true and accurate at the
date of going to press, neither the authors nor the editors nor the publisher can accept any
legal responsibility for any errors or omissions that may be made. The publisher makes no
warranty, express or implied, with respect to the material contained herein.
Printed in the United States of America. (BS/WW)
987654321 SPIN 10689416
springeronline.com
v
Preface to the Second Edition
This second edition of Diagnosis of Endometrial Biopsies and Curettings:
A Practical Approach follows a number of favorable comments we
received about the first edition. As before, this book is designed to offer
a practical reference for the everyday interpretation of endometrial
biopsies. This edition has been extensively updated to reflect the
advances in our understanding of the pathology and pathophysiology of
the endometrium over the past few years. In addition, a large number of
color illustrations have been added to help the reader understand the
morphologic changes described in the text.
Although the entire book has been revised, several areas received par-ticular attention. Our knowledge of the utility of immunohistochemistry
in the interpretation of these specimens, especially trophoblastic disease
and endometrial neoplasia, has expanded considerably since the first
edition. Accordingly, in this edition the problems, pitfalls, and utility of
this valuable diagnostic adjunct have received greater attention.
While immunohistochemistry is discussed in all the chapters, it is also
summarized in the final chapter, which addresses methods of endome-trial evaluation.
Expanded knowledge of newer entities, such as the epithelioid tro-phoblastic tumor, endometrial intraepithelial carcinoma, and the effects
of tamoxifen on the endometrium have received increased emphasis in
this edition. Because hydatidiform mole is now commonly recognized at
an earlier stage of gestation, the features of these “early moles” are dis-cussed in greater detail. The chapter on polyps was revised to further
clarify the terminology of these common lesions, as they demonstrate a
wide spectrum of morphologic features. Information about the distinc-tion of endometrial carcinoma from endocervical adenocarcinoma also
was significantly revised.
Most importantly, however, the text continues its focus on those
aspects of endometrial biopsy interpretation that can be especially
vexing, such as the diagnosis of atypical hyperplasia, grading of endome-trial carcinoma, and the myriad of benign changes and artifacts that can
be confusing to the pathologist. In addition, a clear understanding of the
terminology that the pathologist uses to communicate diagnostic infor-mation to the clinician is critically important. A diagnosis of carcinoma
is straightforward, but a clear and precise diagnosis of the various benign,
yet abnormal pattens of endometrial development and bleeding can be
a challenge.
Finally, we have tried to not only provide a reference for evaluating
the morphologic details of a wide variety of lesions but also to convey
to the reader the manner by which we approach the evaluation of the
endometrial biopsy. It is not possible to cover every aspect of endome-trial pathology in a single text of this size, but we believe that the book
is a reasonable foundation and starting point for the diagnosis of these
specimens. We hope it will be useful to pathologists and gynecologists.
Michael T. Mazur, MD
Robert J. Kurman, MD
vi Preface to the Second Edition
vii
Preface to the First Edition
The incentive for writing this book came from a short course, “Endome-trial Biopsy Interpretation,” that we presented for five years at the
United States and Canadian Academy of Pathology. The enthusiastic
response we received from this endeavor prompted us to consider
writing a practical text on the histologic interpretation of these speci-mens, which are commonly encountered in the surgical pathology labo-ratory but are given short shrift in standard texts. Several gynecologic
pathology textbooks, such as Blaustein’s Pathology of the Female Genital
Tract, 4th ed. (1994), describe the morphologic features and classification
of benign and malignant endometrial lesions, but little attention is given
to the subtle differences between physiologic changes and pathologic
conditions and the artifacts of biopsy and processing. In addition, micro-scopic findings that can be safely ignored because they have no clinical
bearing are generally not discussed in standard texts. It is our impression
that it is precisely these areas that present most of the difficulties in daily
practice, more so, in fact, that the diagnosis of a malignant tumor.
This text is not a reference or atlas that describes pathologic curiosi-ties that one might never encounter in a lifetime of practice. Instead, we
attempt to provide a logical approach to formulating a pathologic diag-nosis from the diverse array of fragmented, often scant pieces of tissue
and blood received in the laboratory. As such, the material is presented
in a less traditional fashion. Conventional histopathologic classifications
remain an integral part of the text, but the various chapters focus on a
clinically oriented approach to the microscopic diagnosis of common
problems. For example, the individual chapters address the clinical ques-tions and specifics of reporting the findings, aspects that vary according
to the patient’s age and the clinical circumstances.
One important subject is that of changes in the endometrium induced
by breakdown and bleeding, independently of the underlying pathology.
These alterations are highly prevalent in endometrial biopsies and are
often misinterpreted, so they are described in detail. The updated World
Health Organization classification of endometrial hyperplasia is based
on the distinction of atypical and non-atypical hyperplasia. This topic is
especially important, since cytologic atypia is the critical prognostic
feature in their behavior, yet the characteristics of what constitutes
atypia are not well appreciated. Metaplasia and other benign changes
can mimic hyperplasia and carcinoma, so the text focuses on these lesions
in detail. Clinical management of endometrial carcinoma is greatly influ-enced by the histologic evaluation of the curettings. Accordingly, the dis-cussion of endometrial carcinoma considers not only the differential
diagnosis, but also the grading of carcinoma and the distinction of
endometrial from endocervical primary tumors.
Trophoblast presents unique problems in diagnosis. This is largely
because the pathologist lacks experience with the diverse morphologic
array of trophoblastic changes in benign and malignant lesions. Gesta-tional trophoblastic disease is rare in routine practice. Furthermore, tro-phoblast of abortion specimens, including the trophoblast of the
implantation site, usually receives little scrutiny. Two chapters have been
included to cover this complex subject, one devoted to physiologic and
one to neoplastic conditons.
Almost all the illustrations used in this text are from biopsies, and
some show artifact and distortion, as occurs in routine specimens. We
intentionally use this less-than-perfect material, since it better illustrates
the problems that the pathologist faces in the interpretation of these
specimens.
Since this monograph is not a reference text or atlas, we suggest
reading it in its entirety in order to appreciate the clinically oriented
problem-solving approach that we advocate. We hope that the reader
finds this approach informative, useful, and enjoyable.
Michael T. Mazur, MD
Robert J. Kurman, MD
viii Preface to the First Edition
ix
Contents
Preface to the Second Edition  . . . . . . . . . . . . . . . . . . . . . . . . . . . v
Preface to the First Edition  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii
Chapter 1 Introduction  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Chapter 2 Normal Endometrium and Infertility
Evaluation  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
Chapter 3 Pregnancy, Abortion, and Ectopic Pregnancy  . . . . 34
Chapter 4 Gestational Trophoblastic Disease  . . . . . . . . . . . . . 67
Chapter 5 Dysfunctional Uterine Bleeding . . . . . . . . . . . . . . . 100
Chapter 6 Effects of Hormones  . . . . . . . . . . . . . . . . . . . . . . . 121
Chapter 7 Endometritis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147
Chapter 8 Polyps  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163
Chapter 9 Endometrial Hyperplasia, Endometrial
Intraepithelial Carcinoma, and Epithelial
Cytoplasmic Change . . . . . . . . . . . . . . . . . . . . . . . . 178
Chapter 10 Endometrial Carcinoma  . . . . . . . . . . . . . . . . . . . . . 208
Chapter 11 Other Tumors  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 249
Chapter 12 Methods of Endometrial Evaluation  . . . . . . . . . . . 275
Index  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 289
interpretation requires appropriate fixation,
processing, and sectioning of the tissue.
Indications for Biopsy
There are four main indications for endome-trial biopsy or curettage:1–7
1. Determination of the cause of abnormal
uterine bleeding.
2. Evaluation of the status of the
endometrium in infertile patients, including
histologic dating.
3. Evacuation of products of conception,
either spontaneous abortions or termination of
pregnancy.
4. Assessment of the response of the
endometrium to hormonal therapy, especially
estrogen replacement in perimenopausal and
postmenopausal women and Tamoxifen
therapy for breast cancer.
Other indications for biopsy may arise. An
occasional patient will have atypical or abnor-mal glandular cells of undetermined signifi-cance (AGUS) in a cervical–vaginal cytologic
specimen that requires endometrial sampling
to exclude hyperplasia or carcinoma. Uterine
screening with transvaginal ultrasound can
show a thickened endometrial stripe in post-menopausal patients, and a biopsy can be per-formed to exclude significant pathology.8
Some
clinicians sample the endometrium prior to
hysterectomy to exclude significant pathology,
although this procedure reveals little pathology
1
1
Introduction
Endometrial biopsies and curettings are among
the most common tissue specimens received in
the pathology laboratory. In several respects
these specimens present a unique challenge
for the surgical pathologist. The normal
endometrium undergoes a variety of morpho-logic changes, especially during the reproduc-tive years, when cyclical hormonal influences
and pregnancy affect uterine growth. Biopsy-induced artifacts confound this heterogeneous
group of morphologic changes. Whether the
biopsy is limited or a thorough curettage, the
procedure usually is “blind,” with no visualiza-tion of the tissue sampled. The final specimen
contains multiple, irregularly oriented tissue
fragments mixed with blood and contaminating
cervical tissue and mucus.
Interpreting the biopsy material demands a
logical approach that takes into account many
factors, including patient history; the specific
requests of the clinician performing the biopsy;
and an appreciation of the limitations, potential
pitfalls, and complex array of patterns encoun-tered in the microscopic sections. As in evalua-tion of any pathologic specimen, proper
Indications for Biopsy  . . . . . . . . . . . . . . . . 1
Clinical History and Biopsy
Interpretation  . . . . . . . . . . . . . . . . . . . . . . 2
Abnormal Uterine Bleeding  . . . . . . . . . 2
Infertility Biopsy  . . . . . . . . . . . . . . . . . . 3
Products of Conception  . . . . . . . . . . . . . 4
Hormone Therapy  . . . . . . . . . . . . . . . . . 4
Other Considerations  . . . . . . . . . . . . . . 4
Clinical Queries and Reporting  . . . . . . . . 4
in the absence of a history of abnormal bleed-ing.9
Likewise, endometrial biopsy for screen-ing of endometrial cancer or precursor lesions
in asymptomatic perimenopausal and post-menopausal patients has a very low yield of sig-nificant abnormalities and is not cost effective.10
At times these indications for endometrial
sampling overlap. For example, some complica-tions of pregnancy, such as a missed abortion
or trophoblastic disease, are accompanied by
abnormal uterine bleeding. Nonetheless, these
broad categories provide a clinicopathologic
framework for approaching the microscopic
analysis of endometrial biopsy specimens. The
text has therefore been divided into chapters
that correspond to these clinical indications.
Clinical History and Biopsy
Interpretation
Abnormal Uterine Bleeding
The most common reason for performing an
endometrial biopsy is abnormal uterine bleed-ing, a term that refers to any nonphysio-logic uterine bleeding. Age and menstrual/
menopausal status are especially important
data, as causes of abnormal uterine bleeding
vary significantly according to the age and men-strual status of the patient, as discussed later.
Abnormal uterine bleeding is a common sign
of a number of different uterine disorders
ranging from dysfunctional (nonorganic)
abnormalities or complications of pregnancy to
organic lesions such as polyps, hyperplasia,
or carcinoma.5;6;11–16
Several clinical terms are
employed to describe different patterns of
uterine bleeding (Table 1.1).
The prevalence of the various abnormalities
that lead to abnormal bleeding is difficult to
determine precisely, varying with the patient
population and the terms used by investiga-tors.2–4
A practical approach to the possible
diagnoses associated with abnormal bleeding
takes age into account (Tables 1.1 to 1.5). Preg-nancy-related and dysfunctional disorders are
more common in younger patients whereas
atrophy and organic lesions become more fre-quent in older individuals.6
Polyps in peri-menopausal and postmenopausal patients have
been found in 2% to 24% of patients.17–24
Hyperplasia is found in up to 16% of post-menopausal patients undergoing biopsy, and
endometrial carcinoma in fewer than 10% of
patients.17;24
One consistent observation in
studies of postmenopausal patients is the find-ing that atrophy is a common cause of abnor-mal bleeding, being found in 25% or more of
cases.17;18;20;23;25;26
Even among perimenopausal and post-menopausal patients, the proportion of cases
2 1. Introduction
Table 1.1. Clinical terms for abnormal uterine bleeding.
Amenorrhea Absence of menstruation
Hypermenorrhea Uterine bleeding occurring at regular intervals but increased in amount. The period of flow is
normal
Hypomenorrhea Uterine bleeding occurring at regular intervals but decreased in amount. The period of flow
is the same or less than the usual duration.
Menorrhagia Excessive uterine bleeding in both amount and duration of flow occurring at regular intervals
Metrorrhagia Uterine bleeding, usually not heavy, occurring at irregular intervals
Menometrorrhagia Excessive uterine bleeding, usually with prolonged period of flow, occurring at frequent and
irregular intervals
Oligomenorrhea Infrequent or scanty menstruation. Usually at intervals greater than 40 days
Abnormal uterine A term that describes any bleeding from the uterus. Menorrhagia, metrorrhagia,
bleeding (AUB) menometrorrhagia, and postmenopausal bleeding are all forms of AUB.
Dysfunctional uterine Abnormal uterine bleeding with no organic cause. The term implies bleeding caused by
bleeding (AUB) abnormalities in ovulation or follicle development and is a disorder of premenopausal
women.
Postmenopausal Abnormal uterine bleeding that occurs at least 1 year after menopause (the cessation of
menses)
A history of anovulation, obesity, hyperten-sion, diabetes, and exogenous estrogen use
should alert the pathologist that the patient is
at increased risk for hyperplasia and adenocar-cinoma, but this information is rarely included
on the requisition. Typically, there is little
accompanying clinical data except the patient’s
age and a short history of abnormal bleeding.
Consequently, hyperplasia and adenocarci-noma must be diagnostic considerations for
most endometrial specimens received in the
laboratory. On rare occasions, hyperplasia or
even adenocarcinoma is found in biopsies per-formed during an infertility workup where the
clinical question was histologic dating rather
than suspicion of these disorders.27
Infertility Biopsy
When a patient undergoes biopsy for evalua-tion of infertility, the clinical information often
Clinical History and Biopsy Interpretation 3
attributable to any of the aforementioned
conditions is age dependent. Atrophy and
carcinoma occur more frequently in patients
older than 60 years of age, while polyps and
hyperplasia are more common in patients
who are perimenopausal or more recently
postmenopausal. In addition to these uterine
causes of bleeding, other abnormalities, such as
atrophic vaginitis, can cause vaginal bleeding,
and this may be difficult to distinguish from
uterine bleeding until the patient undergoes
thorough clinical evaluation.
Ideally, the clinical history that accompanies
an endometrial sample should include some
description of the pattern and amount of bleed-ing. Often in a patient of reproductive age or
one who is perimenopausal the history is simply
dysfunctional uterine bleeding (DUB). Clini-cally, this term suggests no other causes of
bleeding except ovarian dysfunction. In this
scenario the clinician performs the biopsy to
rule out an organic lesion.
Table 1.2. Causes of abnormal uterine bleeding in
adolescence.
Common Uncommon
Dysfunctional bleeding Endometritis
Anovulatory cycles Clotting disorders
Complications of pregnancya
a
See Chapter 3, Table 3.1 (Complications of pregnancy).
Table 1.3. Causes of abnormal uterine bleeding in
the reproductive years.
Common Uncommon
Complications of pregnancya
Hyperplasia
Endometritis Neoplasia
Dysfunctional bleeding Endometrial carcinoma
Anovulatory cycles Cervical carcinoma
Inadequate luteal phase Clotting disorders
Irregular shedding
Organic lesions
Leiomyomas
Polyps (endometrial, endocervical)
Adenomyosis
Exogenous hormones
Birth control
Progestin therapy
a
See Chapter 3, Table 3.1 (Complications of pregnancy).
Table 1.4. Causes of abnormal uterine bleeding in
perimenopausal years.
Common Uncommon
Dysfunctional bleeding Complications of pregnancya
Anovulatory cycles Endometritis
Organic lesions Adenomyosis
Hyperplasia Neoplasia
Polyps (endometrial, Cervical carcinoma
endocervical) Endometrial carcinoma
Exogenous hormones Sarcoma
Birth control Clotting disorders
Estrogen replacement
Progestin therapy
a
See Chapter 3, Table 3.1 (Complications of pregnancy).
Table 1.5. Causes of abnormal uterine bleeding in
postmenopausal years.
Common Uncommon
Atrophy Endometritis
Organic lesions Sarcoma
Hyperplasia Clotting disorders
Polyps (endometrial)
Neoplasia
Endometrial carcinoma
Exogenous hormones
Estrogen replacement
Progestin therapy (e.g., therapy of breast carcinoma)
is limited, but here, too, the history should
include the date of the last menstrual period
(LMP) to place an approximate time in the
menstrual cycle. This information is useful but
not precise for determining the actual day of
the cycle, as ovulatory frequency and length of
the follicular phase are highly variable among
patients. Usually the main objective of biopsies
for infertility is to determine whether there is
morphologic evidence of ovulation and, if so,
the histologic date (see Chapter 2). The gyne-cologist may seek other specific information,
such as response to hormone therapy, so it is
important that the pathologist be given any
additional history that may be necessary for the
interpretation.
Products of Conception
When endometrial sampling is performed to
remove products of conception, clinical infor-mation often is sparse, as the main goal of the
procedure is simply to remove the placental
and fetal tissue. Significant pathologic changes
are rare. Nonetheless, it is helpful to know if
pregnancy is suspected, and, if so, the approxi-mate gestational age of the pregnancy. If there
is a suspicion of trophoblastic disease, this
should be stated. In such instances the human
chorionic gonadotropin (hCG) titer is relevant.
If an ectopic pregnancy is suspected, alerting
the pathologist can ensure rapid processing and
interpretation of the specimen.
Hormone Therapy
Because the endometrium is responsive to hor-mones, the history of hormone use is important
information. Clinical uses of steroid hormones
(estrogens, progestins, or both) include oral
contraceptive use; postmenopausal replace-ment therapy; and therapy for endometriosis,
hyperplasia, DUB, infertility, and breast carci-noma. As with other facets of the clinical data,
this information may be absent or, if present,
unreadable on the requisition. Consequently,
the pathologist must be prepared to recognize
hormonal effects in the absence of history indi-cating use of hormones.
Other Considerations
Pregnancy history is useful, especially in pre-menopausal patients, regardless of the indica-tion for biopsy, as recent and remote effects
of pregnancy, such as a placental site nodule
or gestational trophoblastic disease, may be
encountered in biopsy material. The history of
recent or past pregnancies is expressed as gra-vidity and parity. The letter G (gravidity) fol-lowed by a number (G1, G2, etc.) indicates the
number of pregnancies, and the letter P (parity)
followed by a number indicates the number of
deliveries. For example, G4, P2 indicates that a
woman has had four pregnancies and two deliv-eries. Further information on parity often is
designated by four numbers indicating full-term pregnancies, premature pregnancies (>20
but <37 weeks’ gestation), abortions (<20
weeks’ gestation), and living children. Thus a
patient who is G5, P3013 is currently pregnant
and has had three previous full-term pregnan-cies and one abortion, and the three children
from the term pregnancies are alive.
The type of procedure, that is, biopsy versus
curettage, is important for deciding whether
focal changes represent significant abnormali-ties and whether or not small specimens are
adequate. Although office-based biopsies gen-erally provide a representative sample, they
may not contain sufficient tissue to ensure the
best diagnosis of some changes. For example,
the irregular glands of hyperplasia may resem-ble patterns seen in some polyps, low-grade
adenocarcinomas, and even artifactually dis-torted normal endometrium. Furthermore,
atypia can be focal in hyperplasia; therefore
biopsy specimens may preclude a definitive
diagnosis. In these cases, a more thorough
biopsy or curettage is necessary to permit the
best diagnosis.
Clinical Queries and Reporting
In general, it is best to avoid diagnostic terms
such as “no pathologic diagnosis” or “no signif-icant pathologic findings,” as there is a wide
range of normal histology. When the tissue
4 1. Introduction
lacks abnormalities, stating the normal phase of
the endometrium, for example, proliferative or
secretory, provides more useful information for
the clinician.
In biopsies for abnormal uterine bleeding,
the pathologic information sought varies with
the patient’s age and clinical history. The gyne-cologist wishes to know the following:
1. Is there an organic lesion, such as a com-plication of pregnancy, inflammation, or a
polyp?
2. Is there evidence of active or old breakdown
and bleeding?
3. Is there evidence to suggest dysfunctional
bleeding?
4. Is there evidence of hyperplasia or
carcinoma?
For example, in the premenopausal patient
the possibility of pregnancy and related bleed-ing is a frequent question. In a perimenopausal
patient the concern shifts to hyperplasia and
carcinoma, and in postmenopausal patients the
importance of ruling out carcinoma becomes
paramount. In any of these conditions, glandu-lar and stromal breakdown may be present
either focally or diffusely. It is the underlying
disorder that is most important to report. The
changes of breakdown and bleeding are sec-ondary and do not indicate a primary disorder
by themselves. Nonetheless, when there is a
history of abnormal bleeding, it can be helpful
to note whether or not there is histologic
evidence of glandular and stromal breakdown
(see Chapter 5), especially if the tissue lacks
evidence of an organic process such as hyper-plasia or carcinoma. This information serves to
document to the gynecologist that bleeding is,
in fact, endometrial in origin. Even when there
is no evidence of active bleeding, foci of stromal
foam cells or hemosiderin, sometimes with
fibrosis, indicate that abnormal bleeding has
taken place and deserve comment.
Besides reporting the morphologic changes
present, noting significant negative findings
can be helpful to the clinician. As an example,
the diagnosis of chronic endometritis is more
helpful if it includes a comment regarding the
presence or absence of specific etiologic factors
such as evidence of a recent pregnancy. Like-wise, if an organic lesion such as a polyp is
present, it is helpful to indicate whether nonin-volved tissue is present and, if so, its appear-ance. In perimenopausal and postmenopausal
patients, if the gynecologist indicates a specific
concern regarding the presence of hyperplasia,
atypia, or carcinoma, then a statement noting
the absence of these lesions is reassuring.
For all cases, specimen adequacy is a con-sideration, but this needs to be specifically
addressed only in limited samples in which the
diagnosis is not clear-cut. Scant tissue obtained
by an office-based biopsy may be insufficient to
allow thorough assessment of the status of the
endometrium. In these cases the pathologist
should indicate in the report that the specimen
is scant. For instance, small samples may reveal
hyperplastic glands, but it may be difficult to
determine whether the abnormality represents
a localized polyp with a proliferative/hyper-plastic pattern (see Chapter 8) or a diffuse
hyperplasia. Some assessment of the
endometrium can be done even on very limited
specimens. For example, histologic dating is pos-sible on small samples. Atrophic endometrium
typically yields a very small amount of tissue, yet
these specimens should not be regarded as inad-equate (see Chapter 5). The subsequent chap-ters consider in greater detail the queries likely
to arise in various circumstances and the infor-mation that the pathologist should incorporate
in the final report.
References
(1) Noyes RW, Hertig AT, Rock J. Dating the
endometrial biopsy. Fertil Steril 1950; 1:3–25.
(2) Baitlon O, Hadley JO. Endometrial biopsy.
Pathologic findings in 3,600 biopsies from
selected patients. Am J Clin Pathol 1975; 63:
9–15.
(3) Nickelsen C. Diagnostic and curative value of
uterine curettage. Acta Obstet Gynecol Scand
1980; 65:693–697.
(4) Van Bogaert L-J, Maldague P, Staquet J-P.
Endometrial biopsy interpretation. Shortcom-ings and problems in current gynecologic prac-tice. Obstet Gynecol 1978; 51:25–28.
References 5
(5) Galle PC, McRae MA. Abnormal uterine
bleeding. Finding and treating the cause. Post-grad Med 1993; 93:73–81.
(6) Speroff L, Glass RH, Kase NG. Clinical gyne-cologic endocrinology and infertility. 6th ed.
Baltimore: Lippincott Williams & Wilkins; 1999.
(7) Merrill JA. The interpretation of endometrial
biopsies. Clin Obstet Gynecol 1991; 34:211–221.
(8) Fleischer AC, Jones HW, III. Early detection of
ovarian and endometrial cancer with transvagi-nal and color Doppler sonography. In: Fleischer
AC, Manning FA, Jeanty P, Romero R, eds.
Sonography in obstetrics and gynecology. Prin-ciples and practice. 6th ed. New York: McGraw-Hill; 2001: 1001–1017.
(9) Stovall TG, Solomon SK, Ling FW. Endometrial
sampling prior to hysterectomy. Obstet
Gynecol 1989; 73:405–409.
(10) Archer DF, McIntyre-Seltman K, Wilborn WW,
Dowling EA, Cone F, Creasy GW, et al.
Endometrial morphology in asymptomatic
postmenopausal women. Am J Obstet Gynecol
1991; 165:317–322.
(11) Goldfarb JM, Little AB. Abnormal vaginal
bleeding. N Engl J Med 1980; 302:666–669.
(12) Povey WG. Abnormal uterine bleeding at
puberty and climacteric. Clin Obstet Gynecol
1970; 13:474–488.
(13) Stenchever MA. Differential diagnosis of major
gynecologic problems by age groups. In:
Stenchever MA, Droegemueller W, Herbst AL,
Mishell DR, Jr, eds. Comprehensive gynecol-ogy. 4th ed. St. Louis: Mosby; 2001: 155–177.
(14) Butler WJ. Normal and abnormal uterine
bleeding. In: Rock JA, Jones HW, III, eds. Te
Linde’s operative gynecology. 9th ed. Philadel-phia: Lippincott Williams & Wilkins; 2003:
457–481.
(15) Kilbourn CL, Richards CS. Abnormal uterine
bleeding. Diagnostic considerations, manage-ment options. Postgrad Med 2001; 109:137–4,
147.
(16) Wren BG. Dysfunctional uterine bleeding. Aust
Fam Phys 1998; 27:371–377.
(17) Rubin SC. Postmenopausal bleeding: Etiology,
evaluation, and management. Med Clin N Am
1987; 71:59–69.
(18) Schindler AE, Schmidt G. Post-menopausal
bleeding: A study of more than 1000 cases.
Maturitas 1980; 2:269–274.
(19) Van Bogaert L-J. Clinicopathologic findings in
endometrial polyps. Obstet Gynecol 1988;
71:771–773.
(20) Choo YC, Mak KC, Hsu C, Wong TS, Ma HK.
Postmenopausal uterine bleeding of nonor-ganic cause. Obstet Gynecol 1985; 66:225–228.
(21) Mencaglia L, Perino A, Hamou J. Hysteroscopy
in perimenopausal and postmenopausal
women with abnormal uterine bleeding. J
Reprod Med 1987; 32:577–582.
(22) Pacheco JC, Kempers RD. Etiology of post-menopausal bleeding. Obstet Gynecol 1968;
32:40–46.
(23) Lidor A, Ismajovich B, Confino E, David MP.
Histopathological findings in 226 women with
post-menopausal uterine bleeding. Acta Obstet
Gynecol Scand 1983; 65:41–43.
(24) Moghal N. Diagnostic value of endometrial
curettage in abnormal uterine bleeding—a
histopathological study. J Pak Med Assoc 1997;
47:295–299.
(25) Meyer WC, Malkasian GD, Dockerty MB,
Decker DG. Postmenopausal bleeding from
atrophic endometrium. Obstet Gynecol 1971;
38:731–738.
(26) Gambrell RD. Postmenopausal bleeding. J Am
Geriatr Soc 1974; 22:337–343.
(27) Scurry J, Brand A, Sheehan P, Planner R. High-grade endometrial carcinoma in secretory
endometrium in young women. a report of five
cases. Gynecol Oncol 1996; 60:224–227.
6 1. Introduction
biopsy or curettage is part of a comprehensive
workup of the patient in the operating room
that includes laparoscopy, hysteroscopy, or
hysterosalpingography to assess the presence
or absence of uterine or tubal lesions that con-tribute to infertility. In these cases, the endome-trial sampling may not be timed as precisely for
the mid- to late luteal phase. Nonetheless, his-tologic evaluation provides the gynecologist
with information regarding the response of the
endometrium to hormonal stimulation, includ-ing indirect evidence of ovulatory function.
The secretory phase is constant in the normal
cycle, lasting 14 days from the time of ovula-tion to the onset of menstruation.1
Variations in
cycle length occur because the proliferative
phase of the cycle varies, both between cycles
and between women. Accordingly, the gynecol-ogist correlates the cycle date by histology with
the woman’s cycle date based on the time of
onset of the upcoming menstrual period, not
the last menstrual period. Ovulation with se-cretory endometrial changes ceases in most
women by age 53, although rarely ovulation
with secretory endometrium and a confirmed
corpus luteum of the ovary has been seen up to
a least age 56 ( personal observation).
The biopsy findings help confirm that ovula-tion occurred, and indicate whether there was
sufficient secretory effect, mediated by proges-terone, during the luteal phase. To utilize fully
the morphologic interpretation, the gynecolo-gist compares the histologic date to the clinical
data, including the date of the rise in the basal
body temperature, the time of the serum
7
2
Normal Endometrium and
Infertility Evaluation
The histologic features of what constitutes
“normal” endometrium change with a woman’s
age, through the premenarchal, reproductive,
perimenopausal, and postmenopausal years.1–3
During the reproductive years, the cyclical hor-monal changes of the menstrual cycle provide
a continuously changing morphologic pheno-type that is “normal.” In biopsy specimens, the
combination of these cyclical changes along
with artifacts and limited sampling can make
normal patterns difficult to interpret. During
the reproductive years, deviations from normal,
either in histologic pattern or in temporal rela-tionship to ovulation, often indicate underlying
abnormalities that may cause female infertility.
The endometrial biopsy is an important part
of the evaluation of the woman with infertil-ity.4;5
Biopsies for the evaluation of infertility
often are performed in the office using a small
curette or a Pipelle aspirator and therefore
the specimens tend to be small.6
Occasionally,
General Considerations in Histologic
Evaluation  . . . . . . . . . . . . . . . . . . . . . . . . . 8
Histologic Dating of the Normal,
Cycling Endometrium  . . . . . . . . . . . . . . . . 10
Proliferative Phase Endometrium  . . . . . 11
Secretory Phase Endometrium  . . . . . . . 12
Menstrual Endometrium  . . . . . . . . . . . . 19
Pitfalls in Dating  . . . . . . . . . . . . . . . . . . . . 21
Artifacts and Contaminants  . . . . . . . . . . . 23
Luteal Phase Defect and Abnormal
Secretory Phase Patterns . . . . . . . . . . . . . . 26
Clinical Queries and Reporting  . . . . . . . . 29
luteinizing hormone (LH) surge, transvaginal
ultrasound evaluation of follicular or corpus
luteum development, serum progesterone
level, or subtraction of 14 days from the onset
of menses.4;7–9
Consequently, the biopsy typi-cally is timed to coincide with the luteal (secre-tory) phase of the cycle. In addition to defining
the precise histologic date, an endometrial
biopsy is part of the infertility workup to
exclude other organic uterine abnormalities.
This chapter reviews the morphologic varia-tions caused by ovarian hormonal stimulation
that provide a background for the interpreta-tion of endometrial biopsies in infertility
patients. These patterns include changes result-ing from normal hormonal fluctuations during
the menstrual cycle and variations in normal
development that are caused by abnormalities
in the endogenous ovarian hormonal levels
during the reproductive years. The latter repre-sent the so-called dysfunctional abnormali-ties that are, for the most part, due to abnor-malities in ovarian follicular development or in
hormone production by the corpus luteum.
Ovarian dysfunction also can result in abnormal
bleeding, and Chapter 5 reviews dysfunctional
uterine bleeding caused by ovulatory abnor-malities. During gestation the endometrium
undergoes other “normal,” that is, physiologic,
alterations as discussed in Chapter 3.
General Considerations in
Histologic Evaluation
Histologic evaluation begins with identification
of surface epithelium, a prerequisite for orient-ing the underlying glands and stroma. The sur-face epithelium is less responsive to sex steroid
hormones than the underlying glands, but it
often shows alteration in pathologic conditions,
especially when the abnormalities are subtle or
focal. For example, during the proliferative
phase, estrogenic stimulation induces develop-ment of ciliated cells along the surface.10
In con-trast, ciliated surface epithelial cells are far
more frequent in pathologic conditions, par-ticularly those associated with unopposed
estrogen stimulation, such as hyperplasia and
metaplasia.2;3;11–13
The subsurface endometrium is divided into
two regions, the functionalis (stratum spongio-sum) and the basalis (stratum basale) (Fig. 2.1).
The functionalis, situated between the surface
epithelium and the basalis, is important to
evaluate because it shows the greatest degree
of hormonal responsiveness. The size and
distribution of glands as well as the cytologic
features of the glandular epithelial cells
are important features in the histologic evalua-tion. Under normal conditions, the glands
should be regularly spaced and have a per-pendicular arrangement from the basalis to
the surface epithelium. In the secretory phase,
the endometrium also shows a stratum com-pactum, a thin region beneath the surface
epithelium. In the stratum compactum the
stroma is dense and the glands are straight
and narrow, even when the glands in the
functionalis are tortuous. The basalis adjoins
the myometrium, serving to regenerate the
functionalis and surface epithelium following
shedding during menses. The endometrium of
the basalis is less responsive to steroid hor-mones, and typically shows irregularly shaped,
inactive appearing glands, dense stroma, and
aggregates of spiral arteries. The spiral arteries
of the basalis (basal arteries) have thicker mus-cular walls than those in the functionalis. In
biopsies, tissue fragments that contain basalis
often do not have surface epithelium. The
glands and stroma of the basalis cannot be
dated, as they are unresponsive to steroid
hormones. A specimen consisting solely of
endometrium from the basalis is therefore
inadequate for dating.
Tissue from the lower uterine segment or
isthmus is another region of the endometrium
that is less responsive to steroid hormones. In
the lower uterine segment the endometrium
has shorter, poorly developed, inactive glands
dispersed in a distinctive stroma (Fig. 2.2). The
columnar cells lining the glands resemble those
of the corpus. Some glands near the junction
with the endocervix show a transition to muci-nous endocervical-type epithelium. The stromal
cells in the lower uterine segment are elongate
and resemble fibroblasts with more abundant
eosinophilic cytoplasm, in contrast to the oval
to rounded stromal cells with minimum cyto-plasm seen in the corpus.
8 2. Normal Endometrium and Infertility Evaluation
The tangential orientation of the functionalis
in biopsies and the tortuosity of the glands, par-ticularly in the late proliferative phase, often
lead to irregular cross sections of the tissue. In
this instance, gland development can be diffi-cult to assess. Furthermore, not all fragments of
tissue in a biopsy or curettage include surface
epithelium, which helps to orient the glands.
General Considerations in Histologic Evaluation 9
Figure 2.1. Normal secretory phase endometrium.
Surface epithelium orients the tissue. The midportion
of the tissue consists of functionalis where glands,
stroma, and blood vessels demonstrate the typical
patterns of maturation through the menstrual cycle.
The basalis in the lower portion of the illustration
consists of irregular, closely spaced glands, dense
stroma, and aggregates of arteries. The stratum com-pactum is composed of the surface epithelium and a
subjacent thin layer of dense stroma.
Nonetheless, at least focally, portions of better-oriented glands usually can be traced through
the functionalis to the surface epithelium, and
these foci are critical for assessing appropriate
glandular and stromal development.
Histologic Dating of the Normal,
Cycling Endometrium
In the ovulatory patient, normal endometrium
has two phases. The first is the proliferative (fol-licular or preovulatory) phase characterized by
growth of glands, stroma, and vessels that is
influenced by estradiol produced mainly by
granulosa cells in the ovarian follicles. Fol-lowing ovulation, the secretory (luteal or
postovulatory) phase reflects the effect of
the combined production of progesterone and
estradiol by luteinized granulosa and theca cells
of the corpus luteum.4
The regular sequence of
morphologic changes determined by the fluctu-ating levels of ovarian steroid hormones forms
the basis for histologic dating.
Dating uses an arbitrarily defined “normal”
cycle of 28 days, with day 1 the first day of men-strual bleeding.1
Histologic dating is most
precise in the postovulatory secretory phase,
as the follicular phase can be highly variable
in length. Furthermore, proliferative phase
changes are not as discrete as those in the secre-tory phase. The date of the secretory phase is
expressed either as the specific day of the 28-day menstrual cycle, assuming ovulation occurs
on day 14, or is stated as the postovulatory day
(e.g., secretory day 21 or postovulatory [P.O.]
day 7). Local custom often determines the pre-ferred method of stating the histologic date.
There are nine histologic features of the
glands and stroma that determine the phase of
10 2. Normal Endometrium and Infertility Evaluation
Figure 2.2. Lower uterine segment. Small, poorly developed glands are seen in nonreactive stroma that is
composed of widely spaced spindle cells. Tissue from the lower uterine segment cannot be dated.
the cycle and the histologic date (Table 2.1).1
Five of these features affect glands: (1) tortu-osity, (2) gland mitoses, (3) orientation of nuclei
(pseudostratified versus basal), (4) basal sub-nuclear cytoplasmic vacuoles, and (5) luminal
secretions with secretory exhaustion. Four
features relate to the stromal: (6) edema, (7)
mitoses, (8) predecidual change, and (9) infil-tration of granular lymphocytes. Practically, the
most important glandular features are orienta-tion of nuclei, subnuclear cytoplasmic vacuoles,
and luminal secretions with secretory exhaus-tion (3, 4, and 5), and the most important
stromal features are edema, predecidual
change, and granular lymphocytic infiltration
(6, 8, and 9). These salient features are usually
readily apparent when present, allowing the
pathologist to assign a histologic date.
Proliferative Phase Endometrium
During the proliferative phase, the endo-metrium grows from about 0.5 mm up to 4.0 to
5.0 mm in thickness, so by the late proliferative
phase, a biopsy obtains a moderate amount of
tissue. Proliferative endometrium has three
stages: early, mid, and late (Table 2.2).2
These
divisions are seldom used in dating biopsies,
however. Usually the diagnosis of proliferative
phase alone is sufficient, indicating that the
endometrium is growing, shows a normal glan-dular distribution, and evidence of ovulation is
not present.
Growth of endometrium is the main charac-teristic of the proliferative phase (Figs. 2.3 and
2.4). Glands and stroma show brisk mitotic
activity. In early proliferative phase endo-metrium, the functionalis contains small,
tubular glands. The glands progressively elon-gate and become tortuous from the mid- to
the late proliferative phase because the gland
growth is disproportionate to the stromal
growth. Despite the tortuosity, the glands main-tain a relatively regular spacing between each
other. Throughout the proliferative phase, the
epithelium lining the glands has pseudostrati-fied, oval nuclei with small nucleoli and dense
basophilic cytoplasm. The pseudostratified
nuclei remain oriented to the basement mem-brane, but some nuclei are raised above the
basement membrane, giving a two-dimensional
layering of the nuclei. The pseudostratification
of the nuclei and the presence of mitotic activ-ity in the glands and stroma are two constant
features of the proliferative phase.
In the proliferative phase, the stromal cells
are widely separated in the functionalis (Fig.
2.4). They are small and oval, with dense nuclei,
scant wisps of cytoplasm, and ill-defined cell
borders. Some stromal edema is normal at mid-proliferative phase. A few lymphocytes also are
scattered throughout the stroma, being most
prominent around the vessels. Small spiral
arteries and thin-walled venules are present.
The orientation and outline of proliferative
phase glands and their relationship to intact
stroma are important features for recognizing
Histologic Dating of the Normal, Cycling Endometrium 11
Table 2.1. Morphologic features used in endome-trial dating.
Glandular changes
1. Tortuosity
2. Mitoses
3. Orientation of nuclei (pseudostratified or basal)a
4. Subnuclear cytoplasmic vacuolesa
5. Secretory exhaustion (luminal secretions)a
Stromal changes
1. Edemaa
2. Mitoses
3. Predeciduaa
4. Granular lymphocyte infiltratea
a
Salient features used in dating the secretory phase.
Table 2.2. Proliferative phase changes.a
Early (4–7 days)
Thin regenerating epithelium
Short narrow glands with epithelial mitoses
Stroma compact with mitoses (cells stellate or spindle
shaped)
Mid (8–10 days)
Long, curving glands
Columnar surface epithelium
Stroma variably edematous, mitoses frequent
Late (11–14 days)
Tortuous glands
Pseudostratified nuclei
Moderately dense, actively growing stroma
a
These changes are subtle. They are rarely used for actual
dating.
Figure 2.3. Proliferative endometrium. Glands are
tubular and regularly spaced in abundant stroma.
The stroma contains small vessels with thin walls.
Both the glandular and the stromal cells show
this normal pattern, as hyperplastic glands or
glands in a polyp can have cytologic features
identical to those of glands in the proliferative
phase. The regular spacing and uniform shape
of the glands are characteristics of normal
proliferative endometrium. Assessing gland
orientation can be complicated, however, by
biopsy-induced fragmentation, an especially
common artifact in early to mid-proliferative
phase biopsies when the mucosa is still thin.
Detached and disrupted glands may appear
abnormally crowded or irregular. To separate
fragmentation artifact from true abnormalities,
it is important to assess the integrity of the
stroma as well as the glands and to use surface
epithelium to help orient the tissue fragments.
Detached and poorly oriented glands that show
pseudostratified nuclei and mitotic activity
usually represent proliferative endometrium
unless better-oriented tissue suggests another
diagnosis. Also, proliferative phase glands fre-quently show the telescoping artifact (see
later).
Secretory Phase Endometrium
In the secretory phase, the glands and stroma
develop in an orderly sequence and display spe-cific histologic features of secretory activity
from histologic day 16 through day 28. The
endometrium attains a thickness of up to 7.0 to
8.0 mm. Unlike in the proliferative phase, the
changes in the glands and stroma are relatively
discrete, varying sharply from one day to the
next, thus permitting accurate dating. Dating
of the first half of the secretory phase is based
primarily on glandular changes whereas dating
of the second half is based mainly on stromal
alterations (Table 2.3).
12 2. Normal Endometrium and Infertility Evaluation
mitotic activity. Focal hemorrhage beneath the
surface epithelium is a result of the biopsy and does
not represent a pathologic change.
Histologic Dating of the Normal, Cycling Endometrium 13
Figure 2.4. Proliferative endometrium. The prolif-erative phase gland shows pseudostratified nuclei
with mitotic activity. The stromal cells have oval
Table 2.3. Endometrial dating, secretory phase.
Interval phase, 14–15 d.a
No datable changes for 36–48
hours after ovulation
Early secretory phase, 16–20 d. Glandular changes
predominate
16 d. Subnuclear vacuoles (Note: Scattered small
irregular vacuoles can be caused by
estrogen alone.)
17 d. Regular vacuolation—nuclei lined up with
subnuclear vacuoles
18 d. Vacuoles decreased in size
Early secretions in lumen
Nucleus approaches base of cell.
19 d. Few vacuoles remain.
Intraluminal secretion
No pseudostratification, no mitoses
20 d. Peak of intraluminal secretions
Mid- to late secretory phase, 21–27 d. Stromal changes
predominate, variable secretory exhaustion
21 d. Marked stromal edema
22 d. Peak of stromal edema—cells have “naked
nuclei”
23 d. Periarteriolar predecidual change
Spiral arteries prominent
24 d. More prominent predecidual change
Stromal mitoses recur.
25 d. Predecidual differentiation begins under
surface epithelium.
Increased numbers of granular lymphocytes
26 d. Predecidua starts to become confluent.
27 d. Granular lymphocytes more numerous
Confluent sheets of predecidua
Focal necrosis
24–27 d. Secretory exhaustion of glands—tortuous with
intraluminal tufts (saw-toothed), ragged
luminal borders, variable cytoplasmic
vacuolization, and lumenal secretions
nuclei and indistinct cytoplasm. Scattered lympho-cytes are normally present.
a
d. = day of ideal 28-day menstrual cycle. To state as postovulatory day, subtract 14
The morphologic changes of the secretory
phase begin 36 to 48 hours after ovulation.
There is an interval phase of 36 to 48 hours
between ovulation and the first recognizable
histologic changes of the endometrium attrib-utable to ovulation. During the interval phase,
the glands become more tortuous and begin to
show subnuclear vacuoles (Fig. 2.5). The first
diagnostic evidence of ovulation, however, is
the presence of abundant subnuclear glycogen
vacuoles in the undulating, tortuous glands
(Fig. 2.6). At this time the stroma is indistin-guishable from that of the late proliferative
phase. Because focal subnuclear vacuolization
may occur in the proliferative phase, at least
50% of the glands should contain vacuoles to
confirm ovulation. In addition, at least 50% of
the cells in a gland should contain vacuoles. If
the 50% rule is not fully met, but the clinical
history and morphology suggest recent ovula-tion, the endometrium may be in the interval
phase. Special stains for glycogen add little to
routine histologic evaluation for establishing
the presence of secretory changes.
Subnuclear vacuoles are abundant by day 17,
and by day 18 the vacuoles begin to move from
the basal to the supranuclear cytoplasm (Fig.
2.7). Concurrently, the nuclei become basally
oriented and line up in a single layer perpen-dicular to the basement membrane. The cyto-plasmic contents then form mucin that is
expelled into the gland lumen. Luminal secre-tions peak at day 20 (Fig. 2.8).
After day 20, the stromal changes are more
important for dating than the glandular
changes. Nonetheless, the glands continue to
show increasing tortuosity, and variable
amounts of luminal secretions persist until just
before menses. From days 20 to 22 the glands
in the functionalis begin to show secretory
exhaustion, a change that becomes more promi-nent by days 24 to 25 (Fig. 2.9). Secretory
14 2. Normal Endometrium and Infertility Evaluation
Figure 2.5. Interval endometrium. The glands main-tain proliferative phase characteristics and show
scattered subnuclear vacuoles. The extent of cyto-plasmic vacuolization is not sufficient to be certain
ovulation has occurred.
Figure 2.6. Early secretory endometrium, days
16–17. Postovulatory changes are clearly present
with a regular distribution of subnuclear vacuoles in
the serpiginous glands. The stroma shows no changes
Figure 2.7. Early secretory endometrium, days 17–18. Glandular cell vacuoles remain prominent but begin
to migrate to the supranuclear cytoplasm. A portion of the stroma shows mild edema.
compared to the late proliferative phase. Inset: Every
gland cell contains a vacuole, resulting in a uniform
alignment of nuclei away from the basement mem-brane.
Figure 2.8. Mid-secretory endometrium, days 20–21. Glands are distended with secretions. The stroma shows
edema and there is no predecidual change.
Figure 2.9. Late secretory endometrium, days 23–24. Predecidual stromal change is evident around spiral
arteries with intervening zones showing edema. The glands are tortuous and show secretory exhaustion.
Histologic Dating of the Normal, Cycling Endometrium 17
exhaustion is characterized by the presence of
a single layer of cells that lie in disarray with
loss of orientation. The cytoplasmic border
along the luminal surface becomes ragged, and
luminal secretions are usually, although not
invariably, present. By days 24 to 25 the glands
often develop a serrated, “saw-toothed”
luminal border (Fig. 2.10). The glandular cells
may continue to show a variable degree of vac-uolization throughout the remainder of the
secretory phase. Cytoplasmic vacuolization is a
physiologic change as long as the glands other-wise have appropriate tortuosity; the cytoplas-mic changes from vacuolization to complete
secretory exhaustion with no vacuoles repre-sent a continuum of normal development. By
day 27, cellular necrosis (apoptosis) becomes
evident with accumulation of nuclear debris in
the basal cytoplasm of the glandular epithelial
cells. Throughout the secretory phase, the
glands in the stratum compactum immediately
beneath the surface epithelium remain small
and tubular despite their increasing tortuosity
in the functionalis.
As the glandular cells develop cytoplasmic
vacuoles and produce luminal secretions,
edema, the first stromal change, begins and
peaks quickly at days 21 to 22 (Table 2.3). Once
stromal changes begin, the glandular changes
are less important for dating. Because of the
edema, the stromal cells take on the so-called
naked nucleus appearance at days 21 to 22.
With this change the stromal cells are widely
dispersed and have small nuclei with scant,
imperceptible cytoplasm (Fig. 2.8). This phase
of pure stromal edema is brief, and the subse-quent predecidual transformation of the stroma
becomes the main feature in dating the late
secretory phase. Although stromal edema is
maximal at days 21 to 22, edema begins in a
Figure 2.10. Late secretory endometrium. Stromal cells around spiral arteries show predecidual change with
increased cytoplasm. The gland shows secretory exhaustion with patchy cytoplasmic vacuolization.
patchy distribution in the early secretory phase
at day 17 to 18. Therefore, some edema in the
earlier portion of the secretory phase does not
represent an irregularity of maturation.
Predecidual change characterizes the late
secretory phase (days 23 to 28). With the
appearance of predecidua (not “pseudode-cidua”), the cells gain identifiable cytoplasm
(Fig. 2.9). These cells become oval to polygonal
shaped in the functionalis and show a moder-ate amount of eosinophilic to amphophilic
cytoplasm (Figs. 2.10 and 2.11). Just below the
surface epithelium they can be spindle shaped.
Cell borders of predecidual cells often are
indistinct in formalin-fixed specimens. Prede-cidual transformation begins on day 23 around
spiral arteries, making the walls appear thicker
and leading to prominence of the vessels (Fig.
2.9). In the predecidua there is a resurgence of
stromal mitotic activity at day 24, while the
glandular epithelium lacks mitoses. Predecidual
change expands, extending to the subsurface
stroma on day 25. The predecidual change
around vessels and beneath the surface epithe-lium becomes confluent, forming larger sheets
by day 26 (Fig. 2.12). Predecidua is easy to rec-ognize when advanced, but this change can be
subtle when it is early and not confluent. Inter-vening stroma often shows some edema, and
dating remains based on the most advanced
changes. By day 27, predecidual change is
extensive.
With predecidual transformation, the stroma
shows a gradually increasing number of smaller
18 2. Normal Endometrium and Infertility Evaluation
Figure 2.11. Predecidua and granular lymphocytes
in late secretory endometrium. Predecidualized
stromal cells in the late secretory phase appear oval
to polygonal with a moderate amount of pale cyto-plasm. At this time in the secretory phase, mitotic
activity recurs in the stromal cells. Stromal granular
lymphocytes are scattered throughout the stroma.
These cells have dark, often lobulated nuclei.
mononuclear and bilobate cells with faintly
granular cytoplasm (Fig. 2.11). Others have
small round and dense nuclei. Variably termed
“stromal granulocytes,” “granulated lympho-cytes,” “K cells,” “leukocytes,” or “neutrophils,”
these cells are not polymorphonuclear leuko-cytes (neutrophils).2;3;14;15
The latter occur nor-mally only in menstrual endometrium. Recent
immunohistochemical studies reveal that most
of these cells are T lymphocytes.16–18
We there-fore prefer the designation “granular lympho-cytes.” They are normally present in small
numbers earlier in the cycle but become promi-nent by the late secretory phase.
At day 27 the endometrium is premenstrual.
Predecidua is present in sheets with many inter-spersed granular lymphocytes. The glands are
highly convoluted and saw-toothed. The glands
begin to show apoptosis with nuclear dust at
their base. On day 28 fibrin thrombi begin to
form in small vessels, and hemorrhage follows
with extravasation of erythrocytes into the
stroma.
Menstrual Endometrium
Menstrual endometrium shows glandular and
stromal breakdown that rapidly affects all the
functionalis by the end of day 28. This stage
shows fibrin thrombi in small vessels, con-densed and collapsed stroma, and necrotic
debris (Figs. 2.13 and 2.14). With this necrosis,
a true neutrophilic infiltrate becomes a part of
the physiologic process.19
When the bleeding is
extensive, it may not be possible to assess the
development of the glands or stroma or the
“normality” of the tissue. Once breakdown
starts, the stromal cells coalesce into aggregates
Histologic Dating of the Normal, Cycling Endometrium 19
Figure 2.12. Late secretory endometrium, days
26–27. The stroma consists of sheets of predecidual-ized cells with a heavy infiltrate of granular lympho-cytes. The glands in the functionalis remain tortuous
and show a variable amount of intraluminal secre-tion. In the stratum compactum beneath the surface
epithelium the glands are small and tubular.
Figure 2.13. Menstrual endometrium. Hemorrhage into the stroma forms lakes of erythrocytes. The
hemorrhage disrupts the glands and stroma, although the tortuosity of the glands persists.
Figure 2.14. Menstrual endometrium. With stromal hemorrhage, the predecidual cells collapse and they lose
their abundant cytoplasm.
and clusters that often show little cytoplasm.
With extensive stromal collapse during men-struation, the predecidual change in the stromal
cells becomes indistinct (Figs. 2.14 and 2.15).
The extensive breakdown also can result in
striking morphologic alterations with artifac-tual glandular crowding. As a result, menstrual
endometrium can be confused with hyperplasia
or even carcinoma if the background bleeding
pattern is not recognized. Conversely, hyper-plasia and carcinoma are proliferative pro-cesses that rarely show extensive breakdown
of the type displayed by menstrual endo-metrium. Because of the artifacts induced
by the breakdown and bleeding of the men-strual phase, this tissue is not suitable for eval-uation of glandular and stromal development.
Some advocate biopsy at the onset of bleeding
to be certain that the procedure does not inter-rupt an early pregnancy, but this tissue is
not optimal unless obtained very early in the
menstrual phase before breakdown becomes
extensive.4
Pitfalls in Dating
The preceding description summarizes the
basic histologic changes of endometrial devel-opment. In addition to understanding the
normal morphology in ideal situations, one
needs to consider a number of practical points
when interpreting the endometrial biopsy.
There are several caveats and potential pitfalls,
knowledge of which assists in accurate diagno-sis of normal endometrium and helps avoid
errors in dating. The following, in our opinion,
are especially important aspects to consider in
evaluating this biopsy material:
1. Endometrium with surface epithelium is
best for interpretation. Absence of surface
epithelium compromises the interpretation.
Pitfalls in Dating 21
Figure 2.15. Menstrual endometrium. Glands and
stroma near the basalis undergo collapse as the
superficial tissue sloughs. The glands retain tortuous
shapes but show nuclear dust accumulating in the
subnuclear cytoplasm  (arrowheads). Predecidual
change in the stroma has become indistinct.
2. Tissue from the lower uterine segment
or basalis is not satisfactory for dating.
Endometrium from these regions does not
respond fully to hormones.
3. Straight, tubular glands beneath the sur-face are normal and not a sign of irregularity in
maturation in the late secretory phase.
4. Scattered subnuclear vacuoles in glands
are not sufficient evidence of ovulation. To be
certain that ovulation has occurred, more than
50% of the glands must show subnuclear
vacuoles.
5. The presence of secretions in the glandu-lar lumen does not indicate secretory endome-trium. Proliferative, hyperplastic, and neoplastic
glands can contain luminal secretions. It is the
glandular cytoplasm and nuclear changes that
are most important for determining the pres-ence or absence of secretory changes.
6. Focal glandular crowding caused by tan-gential sectioning can occur in proliferative or
secretory endometrium (Figs. 2.16 and 2.17).
This artifact can result in back-to-back glands
that do not represent hyperplasia.
7. Focal cystic glands or nonreactive glands
can occur in normal endometrium and have no
significance by themselves.
8. Patchy stromal edema is normal by days
17 to 18 of the secretory phase and does not
signify irregular maturation.
9. Identifying very early pregnancy based on
endometrial changes alone is very difficult.
Apparent “hypersecretory” late secretory
phase glands with vacuolated cytoplasm usually
are a variation of normal development and do
not, by themselves, indicate early pregnancy
(see Chapter 3).
10. Compact predecidua with spindle-shaped stromal cells may not be appreciated as
a true predecidual reaction. Directing attention
to stromal changes around spiral arteries assists
in the identification of predecidua.
11. Lymphocytes and granular lymphocytes
normally become prominent in the stroma of
22 2. Normal Endometrium and Infertility Evaluation
Figure 2.16. Artifactual crowding of late secretory
endometrium. Tangential sectioning of normal late
secretory endometrium near the basalis yields a
pattern of focal glandular crowding. This artifact has
no significance and should not be misinterpreted as
hyperplasia.
the late secretory phase. These do not represent
inflammation.
12. If the tissue is difficult to date because of
apparent discordance in features, the possibil-ity of chronic endometritis or a polyp should be
considered.
13. The endometrium cannot be dated
accurately when polyps, inflammation, or other
abnormalities are present.
Artifacts and Contaminants
Besides variations in the normal anatomy, such
as the basalis and lower uterine segment,
several artifacts of the biopsy often complicate
the histologic patterns. One frequent artifact is
tissue fragmentation caused by mechanical dis-ruption of the tissue. As a result, glands are
detached from the surrounding stroma, and
fragmented glands become randomly oriented,
often appearing closely spaced (Fig. 2.18). This
artifact should not be mistaken for real crowd-ing that occurs in hyperplasia or carcinoma.
Fragmentation and close apposition of dis-parate tissues such as cervical epithelium and
functionalis also lead to confusing patterns.
Artifactually crowded glands lack a continuous
investment of tissue and are not connected by
intervening stroma. These latter features help
in recognition of the artifact. Fragmentation
also is a common feature of atrophy (see
Chapter 5).
Another frequent change is so-called tele-scoping of glands.20
Telescoping may occur
in either proliferative or secretory phase
endometrium (Fig. 2.19), but it also complicates
many nonphysiologic conditions. Telescoping
Artifacts and Contaminants 23
Figure 2.17. Artifactual crowding of late secretory
endometrium. This is another example of late se-cretory phase glands that appear crowded due to
tangential sectioning. This pattern is neither
hyperplastic nor “hypersecretory,” however. Identi-fying surface epithelium elsewhere in the sections
often helps to avoid misinterpretation of these
normal glands.
Figure 2.18. Artifactual fragmentation. Normal pro-liferative phase endometrium is fragmented as a
result of the procedure. The glands have a haphaz-Figure 2.19. Telescoping artifact. Normal secretory
endometrium shows telescoping artifact with a
gland-in-gland appearance. This common alteration
ard arrangement that should not be confused with a
significant abnormality. Focal “telescoping” artifact
also is present (arrow).
is an apparent result of the biopsy procedure and has
no significance.
results in a pattern of an apparent gland within
the lumen of another gland and can mimic
hyperplasia or neoplasia (Fig. 2.20). This arti-fact seems to be a result of mechanical disrup-tion and “snap back” of the gland during
curettage, resulting in intussusception which
rarely occurs in hysterectomy specimens. Tan-gential sectioning of tortuous glands also con-tributes to this phenomenon. Fortunately,
telescoping rarely presents difficulty in inter-pretation once the observer understands the
phenomenon. In questionable cases, the cytol-ogy of the glandular cells and comparison with
surrounding tissue establishes this change as an
artifact.
Endometrial biopsies also often contain con-taminants from the cervix. Most of these conta-minants are obvious. Strips of bland squamous
or mucinous epithelium and irregular pools of
extracellular mucin are common (Fig. 2.21). The
extracellular mucin may contain neutrophils,
cell debris, macrophages, or giant cells that are
normal components with no pathologic signifi-cance in the absence of inflammation in the
endometrial stroma. Occasionally, benign cervi-cal contaminants become more complex and
troublesome in biopsies. Endocervical glands
with squamous metaplasia or microglandular
hyperplasia yield complex patterns, but these
elements are cytologically bland and usually
blend into more typical cervical epithelium (Fig.
2.22). In questionable cases, continuity with
endometrial surface epithelium may help to
establish origin in the corpus. It is also helpful to
look at the surrounding stroma and see if it is of
endometrial or endocervical type. Rarely, an
endometrial biopsy also may reveal an admix-ture of fragments of tissue from cervical dyspla-Artifacts and Contaminants 25
Figure 2.20. Telescoping with focal artifactual gland
crowding. Focus of disrupted secretory glands shows
apparent gland crowding caused by telescoping and
fragmentation. This is a common artifact seen in
normal proliferative and secretory endometrium
that should not be mistaken for hyperplasia.
Figure 2.21. Cervical contaminants. Left: Fragmen-tation has resulted in a strip of mucinous glandular
epithelium adjacent to detached proliferative phase
sia, squamous carcinoma, or adenocarcinoma.
Chapter 10 addresses the differential diagnosis
of endocervical versus endometrial carcinoma.
Occasionally, curettage yields sheets of histi-ocytes with no associated mucin or other tissue
(Fig. 2.23). These histiocytes apparently reside
in the endometrial cavity, and show the typical
histiocyte cytology with a lobulated nucleus and
amphophilic cytoplasm. We have seen them in
association with hydrometra and with benign
bleeding patterns. They apparently represent a
response to intracavitary debris and have been
referred to as “nodular histiocytic hyperplasia”
because they can represent a nodular accumu-lation of histiocytes in sections.21
They are
benign but can mimic endometrial stromal cells
or stromal cell lesions. Immunohistochemical
stains for histiocyte markers, such as lysozyme
or KP 1, can facilitate their recognition. Stromal
foam cells, in contrast, represent stromal cells
and macrophages that are filled with lipid from
erythrocytes in areas of chronic nonphysiologic
bleeding.22
Separate fragments of adipose tissue
with clearly identifiable fat cells in an endome-trial biopsy almost always represent omentum
or extrauterine pelvic soft tissue and indicate
perforation of the uterus. We have also occa-sionally seen colonic mucosa in endometrial
biopsies. In these circumstances the clinician
should be notified immediately.
Luteal Phase Defect and
Abnormal Secretory
Phase Patterns
Luteal phase defect (LPD), or inadequate
luteal phase, is a recognized cause of infertility,
so-called ovulatory infertility.4;23–25
This disorder
glands. Right: Amorphous endocervical mucus with
a few macrophages may be admixed with endome-trial tissues. This finding has no clinical significance.
26 2. Normal Endometrium and Infertility Evaluation
is sporadic and relatively common, but it is a
significant factor for infertility in fewer than 5%
of patients.4;25
LPD also has been implicated as
a factor in early habitual spontaneous abortion
and in abnormal uterine bleeding.26
The etiol-ogy of LPD is obscure. Usually the abnormal-ity appears to arise as a result of hypothalamic
or pituitary dysfunction that causes decreased
levels of follicle-stimulating hormone (FSH)
in the follicular phase, abnormal luteinizing
hormone (LH) secretion, decreased levels of
LH and FSH at the time of ovulation, or ele-vated prolactin levels.4;23;27
Hypothyroidism also
may be a factor in LPD.4
Women with apparent
LPD often have relatively low progesterone
levels during the luteal phase,7;28–32
and moni-toring mid-luteal phase progesterone levels can
be effective for diagnosing this abnormality.33
Endometrial biopsy plays an important role in
the diagnosis and management of this disor-der,27;30;34;35
although some authors suggest that
endometrial biopsy with histologic dating has
limited utility in luteal phase evaluation.9;36;37
In LPD, ovulation occurs, but the subsequent
luteal phase does not develop appropriately.
There is insufficient progesterone production
to support development of the endometrium to
histologic day 28 of the cycle. Usually, this
abnormality is recognized clinically when the
histologic date lags more than 2 days from the
actual postovulatory date.23;34
In this circum-stance LPD is clinically significant only if the
abnormal lag in maturation occurs in at least
two consecutive biopsies.34;35;38
Using the crite-rion for a lag in secretory phase development
by dates, there usually is no morphologic abnor-mality. Clinical evaluation, including the basal
body temperature, the time of the LH surge, or
the onset of menses after biopsy establishes the
diagnosis in this situation.
Luteal Phase Defect and Abnormal Secretory Phase Patterns 27
Figure 2.22. Cervical contaminant. Fragmented
endocervical epithelium with microglandular hyper-plasia lies adjacent to a fragment of endometrium
showing breakdown. Other fields in the section
showed proliferative endometrium.
Because of inadequate progesterone produc-tion, LPD may also cause abnormalities in the
development of secretory endometrium.35;39
In
fact, experimental evidence shows that varia-tions in the relative amounts of the sex steroid
hormones, estradiol and progesterone, affect
endometrial development.40
In these experi-mental conditions, relatively low doses of estro-gen and progesterone result in glandular and
stromal hypoplasia. Higher doses of estrogen
but low doses of progesterone result in stromal
inadequacy, while high levels of progesterone
and low doses of estrogen lead to glandular
inadequacy. These data suggest that variations
in follicle and corpus luteum development with
decreases in hormone production alter the
development of secretory endometrium follow-ing ovulation.
Morphologic features of LPD other than a
lag in the histologic date are poorly character-ized. LPD may cause discordance in the devel-opment of the glands and the stroma.35;39;41–43
The resulting pattern is that of irregular matu-ration, with different areas showing a marked
(greater than 4 days) variation in development.
Although LPD can be reflected in endometrial
morphologic abnormalities such as irregular
maturation, there are no large-scale studies that
have clearly identified specific pathologic fea-tures of this condition.
On occasion, endometrial biopsies show
abnormal secretory phase patterns.35;39;41;43–45
In
such cases the endometrium typically shows
secretory changes that cannot be assigned to
any day of the normal cycle (Figs. 2.24 and
2.25). The pattern may show true irregular
maturation with a large variation in the pattern
of endometrial development from field to field.
For example, some areas may show early secre-tory changes characteristic of about days 18 to
28 2. Normal Endometrium and Infertility Evaluation
Figure 2.23. Histiocytes. A detached aggregate of
histiocytes superficially resembles endometrial
stroma. Inset: The cells have cytologic features of his-tiocytes with oval, folded nuclei and faintly vacuo-lated cytoplasm. A lack of intrinsic vasculature helps
to indicate that this is a contaminant.
Clinical Queries and Reporting 29
Figure 2.24. Abnormal secretory phase pattern.
Endometrial biopsy in a premenopausal woman
being evaluated for infertility shows secretory
changes that cannot be assigned to a specific histo-logic date of the normal menstrual cycle. The glands
19, while other areas show foci of predecidua
consistent with at least day 24. Pathologic
processes other than LPD can cause abnormal
secretory phase development, however (Table
2.4). One study of “deficient secretory phase”
endometrium found the changes of elongated,
hyperchromatic gland nuclei, diminished secre-tory activity, and poorly developed stroma.45
The etiology of these “deficient” secretory
pattens was not known although immunohisto-chemical expression of estrogen and proges-terone receptor content of gland cells appeared
decreased.
When abnormal secretory phase patterns
are the result of specific abnormalities such as
inflammation or polyps, the primary abnormal-ity may be evident. In other cases, there may be
no identifiable etiology for the abnormally
show secretory changes but have tubular outlines
and lack appropriate tortuosity. Further, some glands
lack secretory vacuoles. The stroma shows edema in
this field but other areas lack this stromal response
(see Fig. 2.25).
developed secretory pattern. For this latter
group of cases the abnormality may be dys-functional, related to abnormal development of
the corpus luteum, or the aberration may be
secondary to underlying pathology that is not
adequately sampled. Such cases can be classi-fied only as abnormal secretory phase pattern
(see later).
Clinical Queries and Reporting
Using endometrial biopsy or curettage in the
infertility workup, the gynecologist seeks the
following information: (1) histologic evidence
of ovulation, (2) histologic date of secretory
phase specimens, and (3) presence or absence
of endometrial abnormalities that may be
responsible for infertility. Secretory phase
changes indicate that ovulation has occurred.
Dating the secretory phase gives a general
assessment of progesterone production by the
corpus luteum and the ability of the endo-metrium to respond to progesterone.
In practice, a span of up to 3 days in dates
from field to field is acceptable as normal, espe-cially in mid- to late secretory endometrium
where areas of edematous stroma alternate
with areas of predecidualized stroma. Also, the
glandular changes in the late secretory phase
can be highly variable. The pathologist should
date the tissue on the basis of the most
advanced changes using a 2-day span (e.g., days
24 to 25). Dating the secretory phase is some-what subjective and neither completely exact
nor reproducible, and therefore it is also impor-tant that everyone involved in the interpreta-30 2. Normal Endometrium and Infertility Evaluation
Figure 2.25. Abnormal secretory phase pattern.
Another area from the specimen shown in Fig. 2.24
shows poorly developed secretory glands and dense
stroma that lacks edema. The cause of this type of
abnormality cannot be determined by morphology
alone.
Table 2.4. Causes of undatable endometrium.
Hormonal effects
Anovulation
Luteal phase defect
Persistent corpus luteum
Exogenous hormones
Pregnancy
Organic lesions
Polyps
Leiomyomas
Chronic inflammation
Hyperplasia
Carcinoma
Atrophy
Sampling problems
Fragmentation
Lack of surface epithelium
tion and clinical application of histologic dating
understand the limitations of this morphologic
assessment.46;47
Studies of interobserver varia-tion show that 60% to 80% or more of endome-trial biopsies for dating are within 2 days of
each other when evaluated by experienced
pathologists.47
Furthermore, 80% of the time
the dates are within 2 days of the expected day
compared with basal body temperature and
menstrual dates.48
Some investigators find that
the criteria for histologic dating do not have
sufficient precision to determine the degree of
corpus luteum function in the late secretory
phase.49
The field-to-field variations in normal late
secretory endometrium probably lead to some
of the interobserver and intraobserver varia-tions found in dating. With experience, a pathol-ogist should be able to provide a reasonable
assessment of the endometrial development in
the secretory phase.
Obviously for the infertility patient every
attempt should be made to provide an accurate
date, but there are cases in which precise histo-logic dating is not possible. When accurate
dating cannot be done, it is important, if pos-sible, to indicate why (Table 2.4). Hormonal
effects, various organic lesions, and sampling
problems all can make dating difficult or impos-sible. Other organic factors such as inflamma-tion, adhesions, or polyps may interfere with
pregnancy (see Chapter 7). These abnormalities
affect fertility by altering the development of
the glands and stroma, thereby preventing
normal implantation or mechanically disrupt-ing the early implanting placenta. Therefore,
when present these abnormalities should be
reported. If factors preclude reporting a specific
date, an attempt should be made to decide
whether the endometrium is proliferative or
secretory, because secretory phase develop-ment generally indicates that ovulation has
occurred.
On occasion the biopsy shows an abnormal
secretory pattern that cannot be histologically
dated. In such cases, the abnormal pattern may
be due to an LPD or to some other pathologic
factor that is not identifiable in the specimen.
In practice, a descriptive diagnosis of the
changes (e.g., “secretory with irregular matura-tion” or “abnormal secretory phase pattern”)
with a description of the abnormality is suffi-cient to indicate that the secretory develop-ment is not normal. The gynecologist can use
this information in combination with other clin-ical observations to determine its significance
and possible cause. Descriptive diagnoses
should be used carefully, however. The term
“dyssynchronous endometrium” has been used
to describe apparent alterations in secretory
phase development. “Dyssynchronous” is not a
word with a specific connotation and therefore
its use can be confusing unless there is clear
communication between the pathologist and
the gynecologist regarding its meaning.
Recently, morphometric analysis has been
attempted to increase the accuracy of endome-trial histologic dating.50
It was found that five
morphometric measurements, including mitotic
rate in gland cells, amount of luminal secretion,
volume fraction of gland occupied by gland cell,
amount of pseudostratification of gland cells,
and amount of predecidual reaction, added
precision to histologic dating.50
Furthermore,
immunohistochemical analysis for specific
secretory products of the endometrium and
integrins may help to identify LPDs.51–54
Other
cell products such as cyclin E and p27 also may
be useful in assessing secretory phase develop-ment.55
These data indicate that further refine-ments in evaluation of normal endometrium
may evolve that have clinical utility. At present,
however, routine histologic evaluation remains
a cost-effective method of determining the
relative degree of endometrial development
through the menstrual cycle.
References
(1) Noyes RW, Hertig AT, Rock J. Dating the
endometrial biopsy. Fertil Steril 1950; 1:3–25.
(2) Dallenbach-Hellweg G. Histopathology of
the endometrium. 4th ed. New York: Springer-Verlag; 1987.
(3) Hendrickson MR, Kempson RL. Surgical
pathology of the uterine corpus (Major prob-lems in pathology series). Volume 12. Philadel-phia: W.B. Saunders; 1980.
References 31
(4) Speroff L, Glass RH, Kase NG. Clinical gyne-cologic endocrinology and infertility. 6th ed.
Baltimore: Lippincott Williams & Wilkins;
1999.
(5) Merrill JA. The interpretation of endometrial
biopsies. Clin Obstet Gynecol 1991; 34:211–221.
(6) Hill GA, Herbert CMI, Parker RA, Wentz AC.
Comparison of late luteal phase endometrial
biopsies using the Novak curette or Pipelle
endometrial suction curette. Obstet Gynecol
1989; 73:443–446.
(7) Hecht BR, Bardawil WA, Khan-Dawood FS,
Dawood MY. Luteal insufficiency: Correlation
between endometrial dating and integrated
progesterone output in clomiphene citrate-induced cycles. Am J Obstet Gynecol 1990;
163:1986–1991.
(8) Shoupe D, Mishell DRJ, Lacarra M, Lobo RA,
Horenstein J, d’Ablaing G et al. Correlation of
endometrial maturation with four methods of
estimating day of ovulation. Obstet Gynecol
1989; 73:88–92.
(9) Peters AJ, Lloyd RP, Coulam CB. Prevalence of
out-of-phase endometrial biopsy specimens.
Am J Obstet Gynecol 1992; 166:1738–1746.
(10) Masterson R, Armstrong EM, More IAR. The
cyclical variation in the percentage of ciliated
cells in the normal human endometrium.
J Reprod Fertil 1975; 42:537–540.
(11) Hendrickson MR, Kempson RL. Ciliated carci-noma-a variant of endometrial adenocarci-noma. A report of 10 cases. Int J Gynecol
Pathol 1983; 2:1–12.
(12) Ronnett BM, Kurman RJ. Precursor lesions of
endometrial carcinoma. In: Kurman R, ed.
Blaustein’s pathology of the female genital
tract. 5th ed. New York: Springer-Verlag; 2002:
467–500.
(13) Silverberg SG, Kurman RJ. Tumors of the
uterine corpus and gestational trophoblastic
disease. Atlas of tumor pathology, 3rd series,
Fascicle 3. Washington, DC: Armed Forces
Institute of Pathology; 1992.
(14) Buckley CH, Fox H. Biopsy pathology of the
endometrium. 2nd ed. London: Arnold, 2002.
(15) Mutter GL, Ferenczy A. Anatomy and histol-ogy of the uterine corpus. In: Kurman RJ, ed.
Blaustein’s pathology of the female genital
tract. 5th ed. New York: Springer-Verlag; 2002.
(16) Bulmer JN, Lunny DP, Hagin SV. Immunohis-tochemical characterization of stromal leuco-cytes in nonpregnant human endometrium.
Am J Reprod Immmunol Microbiol 1988;
17:83–90.
(17) King A, Wellings V, Gardner L, Loke YW.
Immunocytochemical characterization of the
unusual large granular lymphocytes in human
endometrium throughout the menstrual cycle.
Hum Immunol 1989; 24:195–205.
(18) Bulmer JN, Hollings D, Ritson A. Immunocy-tochemical evidence that endometrial stromal
granulocytes are granulated lymphocytes.
J Pathol 1987; 153:281–288.
(19) Poropatich C, Rojas M, Silverberg SG. Poly-morphonuclear leukocytes in the endometrium
during the normal menstrual cycle. Int J
Gynecol Pathol 1987; 6:230–234.
(20) Numers CV. On the traumatic effect of curet-tage on the endometrial biopsy, with special ref-erence to so-called invagination pictures and
the “crumbling endometrium.” Acta Obstet
Gynecol Scand 1942; 28:305–313.
(21) Kim KR, Lee YH, Ro JY. Nodular histiocytic
hyperplasia of the endometrium. Int J Gynecol
Pathol 2002; 21:141–146.
(22) Silver SA, Sherman ME. Morphologic and
immunophenotypic characterization of foam
cells in endometrial lesions. Int J Gynecol
Pathol 1998; 17:140–145.
(23) Jones GS. Luteal phase insufficiency. Clin
Obstet Gynecol 1972; 16:255–273.
(24) Wentz AC. Diagnosing luteal phase inadequacy.
Fertil Steril 1982; 37:334.
(25) Wentz AC, Kossoy LR, Parker RA. The impact
of luteal phase indaequacy in an infertile
population. Am J Obstet Gynecol 1990;
162:937–945.
(26) Sherman ME, Mazur MT, Kurman RJ. Benign
diseases of the endometrium. In: Kurman RJ,
ed. Blaustein’s pathology of the female genital
tract. 5th ed. New York: Springer-Verlag; 2002:
421–466.
(27) Daly DC, Walters CA, Soto-Albors CE,
Riddick DH. Endometrial biopsy during treat-ment of luteal phase defects is predictive of
therapeutic outcome. Fertil Steril 1983; 40:
305–310.
(28) Rosenfeld DL, Garcia C. A comparison of
endometrial histology with simultaneous
plasma progesterone determinations in infer-tile women. Fertil Steril 1976; 27:1256–1266.
(29) Annos T, Thompson IE, Taymor ML. Luteal
phase deficiency and infertility: Difficulties
encountered in diagnosis and treatment. Obstet
Gynecol 1980; 55:705–710.
(30) Rosenfeld DL, Chudow S, Bronson RA. Diag-nosis of luteal phase inadequacy. Obstet
Gynecol 1980; 56:193–196.
32 2. Normal Endometrium and Infertility Evaluation
(31) Cooke ID, Morgan CA, Parry TE. Correlation
of endometrial biopsy and plasma proges-terone levels in infertile women. J Obstet
Gynaecol Br Commonw 1972; 79:647–650.
(32) Cumming DC, Honore LH, Scott JZ, Williams
KP. The late luteal phase in infertile women:
Comparison of simultaneous endometrial
biopsy and progesterone levels. Fertil Steril
1985; 43:715–719.
(33) Jordan J, Craig K, Clifton DK, Soules MR.
Luteal phase defect: The sensitivity and speci-ficity of diagnostic methods in common clinical
use. Fertil Steril 1994; 62:54–62.
(34) Wentz AC. Endometrial biopsy in the evalua-tion of infertility. Fertil Steril 1980; 33:121–
124.
(35) Witten BI, Martin SA. The endometrial biopsy
as a guide to the management of luteal phase
defect. Fertil Steril 1985; 44:460–465.
(36) Balasch J, Fabregues F, Creus M, Vanrell JA.
The usefulness of endometrial biopsy for luteal
phase evaluation in infertility. Hum Reprod
1992; 7:973–977.
(37) Scott RT, Snyder RR, Strickland DM, Tyburski
CC, Bagnall JA, Reed KR, et al. The effect of
interobserver variation in dating endometrial
histology on the diagnosis of luteal phase
defects. Fertil Steril 1988; 50:888–892.
(38) Downs KA, Gibson M. Basal body temperature
graph and the luteal phase defect. Fertil Steril
1983; 40:466–468.
(39) Dallenbach-Hellweg G. The endometrium of
infertility. A review. Pathol Res Pract 1984;
178:527–537.
(40) Good RG, Moyer DL. Estrogen-progesterone
relationships in the development of secretory
endometrium. Fertil Steril 1968; 19:37–45.
(41) Gillam JS. Study of the inadequate secretion
phase endometrium. Fertil Steril 1955; 6:
18–36.
(42) Deligdisch L. Hormonal pathology of the
endometrium. Mod Pathol 2000; 13:285–294.
(43) Dallenbach-Hellweg G. Changes in the
endometrium caused by endogenous hormonal
dysfunction. Verh Dtsch Ges Pathol 1997; 81:
213–218.
(44) Noyes RW. The underdeveloped secretory
endometrium. Am J Obstet Gynecol 1959; 77:
929–945.
(45) Thornburgh I, Anderson MC. The endometrial
deficient secretory phase. Histopathology 1997;
30:11–15.
(46) Zaino R, Novotney D, Ireland D, Meyer W,
Fritz M, Lessey B, et al. Dating the endometrial
biopsy: A re-evaluation. Mod Pathol 2003; 16:
216A.
(47) Duggan MA, Brashert P, Ostor A, Scurry J,
Billson V, Kneafsey P, et al. The accuracy and
interobserver reproducibility of endometrial
dating. Pathology 2001; 33:292–297.
(48) Noyes RW, Haman JO. The accuracy of
endometrial dating. Fertil Steril 1953;
4:504–517.
(49) Li T-C, Dockery P, Rogers AW, Cooke ID. How
precise is histologic dating of endometrium
using the standard dating criteria? Fertil Steril
1989; 51:759–763.
(50) Li TC, Rogers AW, Dockery P, Lenton EA,
Cooke ID. A new method of histologic dating
of human endometrium in the luteal phase.
Fertil Steril 1988; 50:52–60.
(51) Graham RA, Seif MW, Aplin JD, Li TC, Cooke
ID, Rogers AW, et al. An endometrial factor
in unexplained infertility. Br Med J 1990;
300:1428–1431.
(52) Seif MW, Aplin JD, Buckley CH. Luteal phase
defect: The possibility of an immunohistochem-ical diagnosis. Fertil Steril 1989; 51:273–279.
(53) Lessey BA, Castelbaum AJ, Wolf L, Greene W,
Paulson M, Meyer WR, et al. Use of integrins
to date the endometrium. Fertil Steril 2000;
73:779–787.
(54) Ordi J, Creus M, Ferrer B, Fabregues F,
Carmona F, Casamitjana R, et al. Midluteal
endometrial biopsy and alphavbeta3 integrin
expression in the evaluation of the endo-metrium in infertility: implications for fecun-dity. Int J Gynecol Pathol 2002; 21:231–238.
(55) Dubowy RL, Feinberg RF, Keefe DL, Doncel
GF, Williams SC, McSweet JC, et al. Improved
endometrial assessment using cyclin E and p27.
Fertil Steril 2003; 80:146–156.
References 33
34
pathologic states such as abnormal bleeding or
chronic endometritis.
An abortion before the 16th week of preg-nancy is the usual source of endometrial tissue
specimens that show apparent gestational
changes. The different types of abortions are
defined as follows. “Spontaneous abortions” are
unexpected and unplanned interruptions of
pregnancy that present with bleeding and
passage of tissue. Approximately 15% to 20%
of early pregnancies end in a spontaneous abor-tion.1–4
In addition, many other early pregnan-cies spontaneously abort before pregnancy is
recognized by the woman and are occult.5;6
Most
spontaneous abortions occur before 12 weeks of
pregnancy, and at least half of these are attrib-utable to a genetic (karyotypic) anomaly. An
“incomplete abortion” is a spontaneous abor-tion in which the conceptus and decidua are
incompletely passed, thus requiring curettage.
A “missed abortion” refers to an abortion with
retained products of conception but no abnor-mal bleeding for 5 to 8 weeks after death of the
embryo or fetus. The criteria for diagnosis of a
missed abortion vary among practitioners and
institutions. “Therapeutic abortions” are those
in which the pregnancy is electively terminated.
Besides abortions, several other complica-tions of pregnancy, such as retained placenta or
placental implantation site, ectopic pregnancy,
or gestational trophoblastic disease, lead to the
need for endometrial curettage (Table 3.1).
Specimens from patients with these conditions
show either trophoblastic tissue, the effects of
trophoblastic tissue on the endometrium, or a
3
Pregnancy, Abortion, and
Ectopic Pregnancy
Recognition of the features of gestational
endometrium, trophoblast, and villi, as well as
the pathologic changes in chorionic tissues, is an
important part of endometrial biopsy interpre-tation. The presence of intrauterine products of
conception generally excludes the diagnosis of
ectopic pregnancy and can help explain other
Endometrial Glands and Stroma in
Pregnancy  . . . . . . . . . . . . . . . . . . . . . . . . . 35
Early Gestational Endometrium (1 to 3
Weeks Postfertilization) . . . . . . . . . . . . . 35
Endometrium in Later Pregnancy (4 or
More Weeks Postfertilization)  . . . . . . . . 37
Arias-Stella Reaction . . . . . . . . . . . . . . . 39
Other Glandular Changes in
Pregnancy  . . . . . . . . . . . . . . . . . . . . . . . 41
Trophoblast and Villi . . . . . . . . . . . . . . . . . 43
Trophoblastic Cells  . . . . . . . . . . . . . . . . 44
Immunohistochemistry of Trophoblastic
Cells  . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
Placental Implantation Site  . . . . . . . . . . 48
Exaggerated Placental Implantation
Site . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
Placental Site Nodules  . . . . . . . . . . . . . . 54
Chorionic Villi and Villous Trophoblast
in the First Trimester  . . . . . . . . . . . . . . . 57
Hydropic Change and Other
Pathologic Changes in Abortions  . . . . . 58
Chorionic Villi and Villous Trophoblast
After the First Trimester  . . . . . . . . . . . . 59
Placental Polyps . . . . . . . . . . . . . . . . . . . 61
Placenta Accreta  . . . . . . . . . . . . . . . . . . 62
Endometrium Associated with Ectopic
Pregnancy  . . . . . . . . . . . . . . . . . . . . . . . . . 62
Clinical Queries and Reporting  . . . . . . . . 63
combination of trophoblastic tissue and its
effects.
This chapter first reviews the physiologic
changes of the endometrium in pregnancy,
especially early pregnancy. This is followed by
a discussion of normal placental implantation
and growth, as well as benign and pathologic
trophoblast conditions. Chapter 4 reviews the
closely related topic of gestational trophoblas-tic disease.
Endometrial Glands and Stroma
in Pregnancy
Early Gestational Endometrium
(1 to 3 Weeks Postfertilization)
Fertilization occurs in the fallopian tube soon
after ovulation, and implantation (nidation) of
the developing blastocyst takes place on day
20 or 21 (postovulatory day 6 or 7). Implanta-tion occurs on the surface of the endometrium,
usually on the midportion of the posterior wall.
The ovulatory cycle, during which fertilization
and implantation take place, is called the cycle
of conception. Immediately after implantation,
subtle changes begin to appear in the glands
and stroma, although the tissue retains the
overall characteristics of the mid- to late secre-tory phase for several days. An endometrial
biopsy or curettage performed inadvertently at
this time, usually during infertility evaluation,
may not include trophoblast or disrupt the
early gestation, yet will show very early preg-nancy-related changes. These changes include
recrudescence or accentuation of glandular
secretions, distension of the glands, edema, and
an extensive predecidual reaction.7–10
The
coiled glands show secretory activity and a
serrated lumen, but they appear distended or
wider than those in the late secretory phase of
a menstrual cycle (Fig. 3.1). Vascular promi-nence with engorgement and dilation of super-ficial veins and capillaries also occurs,8
and the
spiral arteries develop thicker walls.
Other reported changes in the cycle of con-ception include persistent basal cytoplasmic
vacuoles in late secretory phase glands,9
a
disparity between development of the glands
and stroma,11
or a histologic date earlier than
the cycle date as determined by basal body
temperature or luteinizing hormone (LH)
surge.11–14
One study reported that in the cycle
of conception pronounced stromal edema and
vascular engorgement of capillaries and small
veins correlated better with pregnancy than did
the glandular changes.7
Because menstruation
does not occur, there is no substantial change
in the number of true granular lymphocytes at
this time. Practically, however, the morphologic
changes during the first 1 to 2 weeks after con-ception are subtle. It is difficult to decide
whether the vacuolated cytoplasm within glan-dular epithelium reflects normal persistence of
vacuoles in the late secretory phase or the
changes of pregnancy. It usually is not possible
to be certain of pregnancy-related changes
until 2 weeks or more after conception,
when decidua, as opposed to predecidua, is
fully developed (see later).
Within 10 to 15 days of fertilization, the
endometrium gradually begins to show more
characteristic changes of pregnancy as differ-entiation of stromal cells into decidua pro-gresses (Table 3.2). As compared to predecidual
cells, decidual cells are larger and contain more
abundant eosinophilic to amphophilic cyto-plasm that may contain faint vacuoles (Figs. 3.2
and 3.3). These cells become more clearly
polyhedral with well-defined cell membranes.
Nuclei of the decidualized stromal cells are
round to oval and uniform, with smooth out-lines, finely dispersed chromatin, and indistinct
nucleoli (compare with Fig. 3.15). Occasional
decidualized stromal cells are binucleate.
Stromal granular lymphocytes persist in early
pregnancy and are clearly evident among
decidual cells. The presence of granular lym-phocytes may suggest a chronic inflammatory
infiltrate, but the granular lymphocytes, in con-trast to inflammatory cells, have characteristic
lobated nuclei and plasma cells are not present.
Endometrial Glands and Stroma in Pregnancy 35
Table 3.1. Complications of pregnancy.
Spontaneous abortion
Missed abortion
Retained placental tissue/implantation site
Ectopic pregnancy
Placenta accreta, increta, and percreta
Gestational trophoblastic disease
36 3. Pregnancy, Abortion, and Ectopic Pregnancy
Figure 3.1. Early gestational endometrium, cycle of
conception. Inadvertent endometrial biopsy in the
cycle of conception shows distended, coiled glands
and engorged vessels. Early decidual reaction is
present around thickened spiral arteries. In the
Table 3.2. Histologic changes of the endometrium in pregnancy.a
Duration of
pregnancyb
Stroma Glands Vessels
1–3 weeks Edema, then progressive Hypersecretory with cytoplasmic Spiral arteries begin to thicken
decidual change vacuoles, luminal secretions Superficial venules congested,
Saw-toothed, tortuous, distended dilated
Rare Arias-Stella reaction
4 or more weeks Marked decidual change Irregular with marked atrophy Spiral arteries thickened
Variable Arias-Stella reaction, clear Superficial venules dilated
cytoplasm, optically clear nuclei
a
Changes of endometrial glands, and stroma and vessels only.
b
Duration from time of fertilization (day 15–16 of menstrual cycle). Add 2 weeks for time from last menstrual period.
absence of trophoblast or chorionic villi, these fea-tures are too subtle to be diagnostic of early preg-nancy until the stroma shows more advanced
decidual change.
Figure 3.2. Early gestational endometrium with
decidualized stroma. Endometrial stroma from a
first-trimester abortion shows prominent decidual
transformation. The decidualized cells have abun-Along with progressive decidual transforma-tion of the stromal cells, the glands and the
vessels undergo pronounced alterations. Secre-tory changes in the glands become more promi-nent with increased cytoplasmic vacuolization
and augmented luminal secretions that distend
the glands (Fig. 3.4). In addition to this hyper-secretory activity, the glands become highly
coiled with prominent serrations and papillary
folds of epithelium projecting into the lumen.
The epithelial cells become stratified. Concur-rently, spiral arteries are more prominent
(Fig. 3.2).
Endometrium in Later Pregnancy
(4 or More Weeks Postfertilization)
With advancing gestational age, pregnancy-related patterns become more pronounced
and distinctive (Table 3.2). The decidualized
stromal cells are widespread and prominent,
especially as the cell borders become better
defined, and they develop an epithelioid
appearance. Decidual cell nuclei become some-what larger and may appear vesicular, but they
maintain their uniform contours. The decidua
shows small foci of physiologic necrosis during
pregnancy, as it remodels during growth of the
fetus and placenta and as the decidua capsularis
fuses with the decidua parietalis. These small
foci of necrosis, with a localized neutrophilic
response, are physiologic. They do not reflect an
infectious or septic process and do not indicate
a significant abnormality. The decidua contin-ues to contain a sprinkling of granular lympho-cytes that remain throughout gestation. The
hypersecretory pattern of the glands begins to
regress early in pregnancy, and with increasing
decidualization the glands become atrophic
(Fig. 3.5) Conversely, in areas where the glands
Endometrial Glands and Stroma in Pregnancy 37
dant, pale cytoplasm, distinct cell borders, and
uniform round to oval nuclei. Granular lymphocytes
are numerous. Spiral artery walls are thicker than in
nongestational endometrium.
Figure 3.3. Decidualized stroma. Decidual cells have prominent cell borders and uniform, round to oval
nuclei. Their cytoplasm has small vacuoles. A portion of an atrophic gland is present in the left upper corner.
Figure 3.4. Gestational endometrium. Hypersecretory pattern of endometrial glands in early pregnancy with
extensive cytoplasmic vacuolization. Decidualized stroma was present in other areas of the sections.
appear hypersecretory, the stroma often is not
decidualized (Fig. 3.4) Usually a mixture of
hypersecretory and atrophic glands is present.
By the end of the first trimester, the glands for
the most part are atrophic and have lost their
luminal secretions. In fact, as they form irregu-lar, dilated spaces with indistinct epithelium,
they may be difficult to distinguish from vascu-lar channels.
As pregnancy advances, the spiral arteries
maintain thick walls, a feature that persists to
term and is helpful in recognizing gestational
changes. Some authors suggest that in the first
trimester the arteries develop a characteristic
atherosclerosis-like change when an intrauter-ine pregnancy is present, characterized by
subintimal proliferation of myofibroblasts with
foam cells.15
In addition, the venules beneath
the surface epithelium dilate. Dilated superfi-cial venules are not a specific change of preg-nancy, however, as they may also be observed
as a result of progestin stimulation, hyper-plasia, and occasionally in polyps when the
endometrium grows but does not undergo
cyclical shedding.
Arias–Stella Reaction
At 4 to 8 weeks after blastocyst implantation,
the endometrium often shows at least a focal
Arias–Stella reaction in the glands.16–20
This
glandular change is a physiologic response to
the presence of chorionic tissue either in the
uterus or at an ectopic site. The morphologic
features of the Arias–Stella reaction include
nuclear enlargement up to three times normal
size and nuclear hyperchromasia, often accom-panied by abundant vacuolated cytoplasm
(Figs. 3.6 and 3.7). The cells typically are strati-fied and the nuclei hobnail-shaped, bulging
Endometrial Glands and Stroma in Pregnancy 39
Figure 3.5. Gestational endometrium. Endo-metrium from spontaneous abortion shows diffuse
decidual reaction of stroma. The glands are dilated
and lined by atrophic, indistinct epithelial cells.
Therapy with high-dose progestins could induce a
similar pattern.
Figure 3.6. Arias-Stella reaction. Glands from an
abortion specimen show prominent Arias-Stella
reaction with hyperchromatic nuclei that bulge into
Figure 3.7. Arias-Stella reaction. Left. Arias-Stella
reaction with a “hypersecretory” pattern shows strat-ified nuclei and vacuolated cytoplasm. Right. Arias-the glandular lumen. The glands also show marked
cytoplasmic vacuolization. Identical changes can be
seen in an ectopic pregnancy.
Stella reaction with “regenerative” pattern shows
hobnail cells with dense cytoplasm. Several nuclei
show prominent cytoplasmic invaginations.
into the gland lumen. These large nuclei may
contain prominent cytoplasmic invaginations.21
Mitotic figures are rarely present, and Ki-67
immunostains demonstrate a very low pro-liferative index. This process may be extensive,
involving many glands, or the reaction can be
focal, involving only a few glands (Fig. 3.8). The
change can even be limited to part of a gland,
leaving the remaining nuclei unaffected.
The Arias–Stella reaction has two histologic
patterns (Fig. 3.7).18
One is a “hypersecretory”
change characterized by highly convoluted
glands lined by cells with stratified nuclei and
abundant clear to foamy cytoplasm. The other
pattern has been termed “regenerative,” al-though this hypothesized etiology for the change
remains unsubstantiated. This pattern is charac-terized by glands lined by enlarged hobnail
cells with little cytoplasmic secretory activity. In
fact, the two patterns are not very distinct and
there is frequent overlap between them.
The degree and extent of the Arias–Stella
reaction are highly variable in normal and
abnormal intrauterine gestation, in ectopic
pregnancy, and in gestational trophoblastic
disease.18;22–25
This change occurs as early as 4
days after implantation,17
although it generally
is seen after about 14 days.18
The Arias–Stella
reaction persists up to at least 8 weeks follow-ing delivery. There is no apparent relation-ship between the presence and extent of the
Arias–Stella reaction and the status of the fetus.
The Arias–Stella reaction is almost unique to
pregnancy or gestational trophoblastic disease.
Similar phenomena are rarely produced by
administration of exogenous progestins.26
Other Glandular Changes
in Pregnancy
Besides the Arias-Stella reaction, the endome-trial gland cells may undergo other specific
Endometrial Glands and Stroma in Pregnancy 41
Figure 3.8. Gestational endometrium with Arias-Stella reaction. A dilated gland in decidualized
stroma shows Arias-Stella reaction with stratified,
enlarged, and hyperchromatic nuclei (arrow). A
gland in the upper portion of the field does not show
Arias-Stella change. A dilated venule is present
beneath surface epithelium on the right.
changes in the presence of trophoblastic tissue.
One such change is abundant clear cytoplasm.16
This phenomenon overlaps with the Arias–
Stella reaction yet does not show the nuclear
enlargement of the latter. With this change the
gland cells accumulate abundant amounts of
clear, glycogen-rich cytoplasm (Fig. 3.9). The
nuclei in areas of clear cell change can become
stratified, which, combined with the abundant
clear cytoplasm, can result in apparent oblitera-tion of the gland lumens.
Another pregnancy-related change is opti-cally clear nuclei of gland cells (Fig. 3.10).27;28
This alteration is also often associated with the
Arias–Stella reaction but can occur indepen-dently. Optically clear nuclei usually are focal.
They have a clear to glassy appearance that is
caused by accumulation of a filamentous mate-rial in the nuclei. A recent study indicates that
the change is related to the intranuclear accu-mulation of biotin.27
Clear nuclei can mimic the
changes seen in herpesvirus infection, although
the optically clear nuclei associated with
pregnancy lack the Cowdry type A eosinophilic
nuclear inclusions, nuclear molding, and associ-ated necrosis seen in the virus infection. This
alteration is infrequent, occurring in fewer than
10% of first-trimester abortion specimens.
These changes may persist until term, however.
As in the Arias–Stella reaction, optically clear
nuclei simply reflect the presence of chorionic
tissue.
Localized atypical-appearing endometrial
glandular proliferations rarely may be found
during gestation.29
These focal abnormalities
show glandular expansion with nuclear stratifi-cation and cribriform change. Mitotic activity
is present, but nuclear cytology is bland.
Intraglandular calcifications frequently are
present. The few lesions studied have been
42 3. Pregnancy, Abortion, and Ectopic Pregnancy
Figure 3.9. Gestational endometrium with clear cell
change. Glands in early gestational endometrium are
lined by cells with abundant clear cytoplasm. The
cells lack the nuclear enlargement of the Arias-Stella
reaction.
benign, and some patients have had subsequent
pregnancies, suggesting these proliferative foci
represent another unusual gland response to a
concurrent pregnancy.29
Rarely, frank endome-trial adenocarcinoma also may be associated
with intrauterine gestation.30;31
In early pregnancy, endometrial glands
become strongly immunoreactive for S-100
protein.32;33
This immunoreactivity rapidly dis-appears after the 12th week of gestation.
Normal proliferative and secretory endome-trium, as well as glands in patients with hyper-plasia and neoplasia, do not stain for S-100
protein. There are no other markers of the
glands that are generally practical for identify-ing pregnancy-related changes except for the
low Ki-67 index of the Arias–Stella reaction,
which helps to indicate a benign glandular
change.
Trophoblast and Villi
In early pregnancy trophoblastic proliferation
begins with the development of the blastocyst,
the outer layer of which is termed the tro-phoblastic shell. Villous formation does not
begin until about 7 days after implantation of
the blastocyst (13 days following conception).4
For morphologic identification, the products of
conception are divided into three components:
(1) the villi and their trophoblast (“villous” tro-phoblast), (2) the implantation site (“extravil-lous” trophoblast), and (3) fetal tissues. Usually
these tissues are easy to recognize.
Identifying trophoblast and villi is essential
for confirming the diagnosis of an abortion.
Also, the presence of placental and fetal tissue
in curettage samples, for all practical purposes,
rules out an ectopic pregnancy. The morpho-Trophoblast and Villi 43
Figure 3.10. Gestational endometrium with opti-cally clear nuclei. Crowded glands show promi-nent optically clear nuclei. This gland cell change is
infrequent and usually is focal. It may be seen
throughout pregnancy, however.
logic features of these abortion specimens can
be highly varied. Occasionally the diagnostic
features of the products of conception are dif-ficult to identify, especially in early pregnancy,
when the placental component is very small
and often missed in small biopsy specimens, or
if most of the products of conception were
expelled prior to the curettage. When villi and
fetal tissue are not present in curettage samples,
trophoblastic cells should be searched for to
confirm the diagnosis of intrauterine preg-nancy. Recognizing the full morphologic spec-trum of normal trophoblastic cells is important,
not only for establishing the presence of an
intrauterine pregnancy, but also for distinguish-ing exaggerated but physiologic changes from
gestational trophoblastic disease.
Trophoblastic Cells
The trophoblast is extraembryonic but fetal in
origin, growing in intimate association with
host maternal tissues. Very early in pregnancy
trophoblastic cells differentiate and invade
decidua, even before villi form.4
At this stage
of early gestation, implanting trophoblast is
the predominant component of placental tissue.
The trophoblast continues to grow along this
interface of maternal and placental tissue
throughout pregnancy. The decidua basalis
where trophoblast interfaces with the endome-trium and myometrium becomes the placental
implantation site. The trophoblastic cells are
the epithelial component of the placenta and
are divided into three cytologically and
functionally distinct populations: cytotro-phoblastic (CT) cells, syncytiotrophoblastic
(ST) cells, and intermediate trophoblastic (IT)
cells (Table 3.3).34–38
Trophoblastic cells can
also be classified according to their anatomic
location as “villous” and “extravillous”
trophoblast.39
CT cells are the germinative cells from which
other trophoblastic cells differentiate. Accord-ingly, they are mitotically active. They are
uniform cells about the size of a decidualized
stroma cell, with a single nucleus, one or two
nucleoli, pale to faintly granular cytoplasm,
and prominent cell borders (Fig. 3.11). ST cells,
in contrast, are larger and multinucleate with
dense amphophilic to basophilic cytoplasm. The
nuclei of ST cells are dark and often appear
pyknotic; they do not contain mitoses. The cyto-plasm also typically contains small vacuoles and
larger lacunae in which maternal erythrocytes
can be identified. A microvillous brush border
44 3. Pregnancy, Abortion, and Ectopic Pregnancy
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, eosinophilic Pleomorphic and large; abundant, Round to polyhedral, regular
to clear cytoplasm; prominent eosinophilic cytoplasm; occasional abundant eosinophilic and
cell borders multinucleated cells clear cytoplasm
Immunostaining
Cytokeratina
++++ b
++++ ++++
hCG –; + in multinucleated IT -hPL -/+ ++++ ++
Mel-CAM -/++++ c
++++ ++
PLAP – – +++
EMA – – +++
Inhibin-a- +/-++
Ki-67 index >90%d
0 3%–10%
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 proportion of cells positive.
c
Mel-CAM staining increases from the base to the tip of the trophoblastic column.
d
Ki-67 staining decreases from the base to the tip of the trophoblastic column.
sometimes lines the lacunae of the ST cells. CT
and ST cells typically display a dimorphic
growth pattern, with the two cell types growing
in close proximity. In early abortions, the CT
and ST cells are quite prominent compared
with the amount of villi present. In very early,
unanticipated abortions, the entire products of
conception consist of previllous trophoblast
that can be easily confused with choriocarci-noma (Fig. 3.12) (see Chapter 4). In curettage
for suspected abortions, sometimes the only
evidence of an intrauterine pregnancy is the
presence of a few isolated trophoblastic cells
mixed with blood, and these may be necrotic
(Fig. 3.13). Careful scrutiny may be necessary to
identify these diagnostic cells.
The intermediate trophoblast develops from
cytotrophoblast on the villous surface, and in
early pregnancy is manifested as sprouts and
columns that extend to and extensively infil-trate the underlying decidua at the implanta-tion site (see later). In fact, the predominant
location of the IT is at the implantation site,
which explains why it is often called “extravil-lous cytotrophoblast.”39
This latter term is less
precise, however, as these cells also occur in
association with villi (Fig. 3.11), and they are
immunohistochemially and physiologically
different from cytotrophoblast. Another older
term for IT is “X cells.”4
The IT actually
represent a heterogeneous population of
trophoblastic cells: the villous IT, the implanta-tion site IT, and the chorionic-type IT.34;40
The morphologic and immunohistochemical
features of these IT subtypes depend on their
differentiation status and their anatomic
location.
The IT that extends from the trophoblastic
column of the anchoring villi is designated
“villous” IT.40
These cells are mononucleate and
Trophoblast and Villi 45
Figure 3.11. Immature chorionic villi with tro-phoblast. Syncytiotrophoblast and cytotrophoblast
grow from the surface of immature first-trimester
villi. The villi are lined by an inner layer of mononu-cleated CT cells capped by ST cells. To the right of
the villus a dimorphic mixture of ST and villous IT
cells proliferates toward the implanting margin of
the placental tissue.
46 3. Pregnancy, Abortion, and Ectopic Pregnancy
Figure 3.12. Trophoblast of early pregnancy. A
prominent network of ST and CT encountered in
inadvertent endometrial biopsy during early preg-nancy. The cells with pleomorphic, hyperchromatic
nuclei in the lower left part of the field are IT cells
that are infiltrating the decidua. No villi are present.
larger than CT cells. They have pale cytoplasm
and large, round nuclei. The implantation site
IT cells infiltrate the decidua and the myome-trium and have a heterogeneous appearance, as
discussed in the following paragraphs. The
chorionic-type IT constitute the cells of the
chorion laeve where they form a cohesive
layer of epithelium. These latter IT are com-posed of relatively unform cells with
eosinophilic to clear (glycogen-rich) cytoplasm.
These cells are smaller than implantation site
IT although an occasional cell is multinucleated
(Table 3.3).
The villous and the implantation site IT con-stitute the two forms of IT usually seen in
biopsy specimens from early pregnancy. The
villous IT cells are readily recognized because
they are associated with anchoring villi, but the
implantation site IT may pose a greater chal-lenge in recognition because the implantation
site may be seen with no associated villi or the
implantation site can appear prominent or
“exaggerated” (see later). Implantation site IT
also are the primary cell type of the placental
site trophoblastic tumor discussed in Chapter 4.
The chorionic-type IT constitute the cell
population seen in the placental site nodule
discussed later in this chapter and in the
epithelioid trophoblastic tumor discussed in
Chapter 4.
Immunohistochemistry of
Trophoblastic Cells
Trophoblastic cells express a number of pro-teins. The  b-subunit of human chorionic
gonadotropin (hCG), human placental lactogen
(hPL), placental alkaline phosphatase (PLAP),
Despite the biopsy, the pregnancy continued to term
and uneventful delivery. In the absence of villi, this
pattern resembles choriocarcinoma, and clinical
history may be required to determine the signifi-cance of the finding.
and cytokeratin all react with trophoblastic
cells, but the degree of reactivity varies among
the cell types. Cytokeratin is the most ubiqui-tous cell product.34
Broad-spectrum antibodies
such as AE1/AE3 diffusely stain all tro-phoblastic cells. Trophoblast also is reactive for
simple epithelium-type cytokeratins, and
cytokeratins 7 and 18 are especially notable for
their reactivity in trophoblast. Cytokeratin 20,
in contrast, is variably reactive in trophoblast,
and high molecular weight cytokeratin is nega-tive in trophoblastic cells. Other than cytoker-atin, CT cells show limited immunoreactivity to
commonly available antibodies. The CT cells do
not stain for hCG, hPL, PLAP, and other anti-gens associated with specialized trophoblast
such as Mel-CAM (melanoma cell adhesion
molecule [CD 146]) and inhibin-a . The ST cells
stain for the common trophoblastic markers
including hCG, hPL, PLAP, and inhibin-a as
well as cytokeratin.41
The ST do not react with
antibodies to Mel-CAM.42
The IT cells vary in their staining pattern
depending on the subtype (Table 3.3).40
All IT
cells react strongly for cytokeratin. Mel-CAM
(CD 146) also is seen in all types of IT, although
in the villous IT the staining increases from the
base to the tip of the trophoblastic columns. The
implantation site IT and chorionic-type IT are
reactive for hPL and PLAP but the villous IT
cells generally are not. Inhibin-a is seen only in
chorionic-type IT. Chorionic-type IT cells also
stain well for epithelial membrane antigen
(EMA); it is the only type of trophoblastic cell
that is EMA reactive. Ki-67 immunostaining
also shows a variable proliferation index. CT
shows a Ki-67 index of 25% to 50%, which is
consistent with its role as the germinative tro-Trophoblast and Villi 47
Figure 3.13. Isolated ST cells in abortion specimen.
A few ST cells mixed with fibrin and blood are the
only evidence of intrauterine pregnancy in this curet-tage specimen from a patient with a spontaneous
abortion. Isolated ST cells such as this are often
found in areas of hemorrhage.
phoblastic cell. The Ki-67 proliferation index in
villous trophoblast is high at the junction with
the CT at the base of the villus but decreases
progressively toward the tip (distal end) of the
trophoblastic column. The Ki-67 proliferation
index is zero at the junction of the column with
the decidua of the basal plate and is zero in the
implantation-type IT cells as they “drop off”
the column and infiltrate the endomyometrium.
The Ki-67 proliferation index is low (<5%) in
chorionic type IT. The Ki-67 index in ST cells is
zero, which is consistent with that of terminally
differentiated cells.
Placental Implantation Site
The placental implantation site is composed
largely of IT cells that infiltrate the decidua
basalis, mixing with decidualized stromal cells,
glands, and vessels.4;34;36;39;43–45
This site begins as
a microscopic focus where the blastocyst
implants into secretory endometrium, and it
expands with the growing placenta to cover
the entire area of decidua and superficial
myometrium to which the placenta is attached.
It can be diffuse or focal in endometrial biopsy
specimens. Previous terminology for the tro-phoblastic infiltrate in decidual and myometrial
tissue included “syncytial endometritis” and
“syncytial endomyometritis” or “placental giant
cell reaction.” None of these terms is correct,
however, because the process is physiologic, not
inflammatory; the majority of the cells are
not giant cells; and the change is not confined
to the endometrium but also involves the
myometrium.
Placental site intermediate trophoblast
characteristically diffusely permeates the
decidua and implantation site (Figs. 3.14 to
3.18). Because they closely resemble decidual
48 3. Pregnancy, Abortion, and Ectopic Pregnancy
Figure 3.14. Placental implantation site. Fibrin and
implantation site IT with irregular, hyperchromatic
nuclei and cytoplasmic vacuoles are interspersed
among decidual cells. In the right lower corner the
IT cells are partially replacing the endothelial cells
of a blood vessel. Chorionic villi are not seen, but the
presence of intermediate trophoblast in decidua
establishes the diagnosis of intrauterine pregnancy.
Figure 3.15. IT cells and decidua. Placental implan-tation site in an abortion specimen contains numer-ous implantation site IT that infiltrate the decidua.
Figure 3.16. IT cells and decidua. Implantation site
IT in decidua are prominent and hyperchromatic
compared to the decidual cells on the right side of
IT have characteristic large, irregular, hyperchro-matic nuclei. Some of the IT cells have prominent
nucleoli.
the figure. A spiral artery in the center shows the
wall replaced by IT and fibrinoid material.
Figure 3.17. IT cells and decidua. Implantation site
IT with enlarged, hyperchromatic nuclei infiltrate
decidua in an abortion specimen. Several dilated
Figure 3.18. IT cells and decidua. The implantation site IT (arrows) have irregular, hyperchromatic nuclei
that contrast with the round to oval, uniform nuclei of the surrounding decidualized stromal cells.
spiral arteries are infiltrated by the IT, replacing the
endothelium.
cells, they are often difficult to recognize. In
fact, these IT cells have been misinterpreted as
degenerating decidual cells because of their
intimate association with the latter (Figs. 3.16
and 3.18). The implantation site IT cells have
variable size and shape ranging from polygonal
to round to spindle-shaped, with a moderate
amount of eosinophilic to amphophilic cyto-plasm.44
They are larger than decidualized
stromal cells, with which they are intimately
admixed. They may have sharply outlined cyto-plasmic vacuoles. The nuclear morphology of
implantation site IT, however, is the most
important feature that distinguishes these cells
from decidua. They are enlarged, lobated, and
hyperchromatic with irregular nuclear mem-branes. Sometimes they have deep clefts, and
some nuclei appear smudged.46
Most of these
IT cells contain a single nucleus, but bi- or mul-tinucleate cells with similar nuclear and cyto-plasmic features occur as well. The dark and
irregular implantation site IT nuclei, which
often contain a prominent nucleolus, contrast
with the nuclei of decidualized stromal cells,
which are uniform and round to oval with an
even, delicate chromatin distribution. Immuno-histochemical stains for keratin and human
placental lactogen (hPL) help identify inter-
mediate trophoblast cells and distinguish them
from decidual cells which are negative (see
later).35;47
At the implantation site intermediate tro-phoblastic cells also infiltrate into myometrium,
where they often have a spindled appearance
(Fig. 3.19). The implantation site IT cells are
often more conspicuous here, especially when
some of them become multinucleated. In the
myometrium they infiltrate between muscle
bundles and fibers, often with no evidence of a
tissue reaction. Many are spindle shaped and
can closely resemble smooth muscle cells. It is
very common to see fragments of myometrium
infiltrated by intermediate trophoblast in curet-tage samples from abortions.
Trophoblast and Villi 51
Figure 3.19. IT cells in myometrium. Curetted pla-cental site from an abortion includes a fragment
of myometrial smooth muscle infiltrated by large
implantation site IT. These cells often become mult-inucleated when they invade the myometrium.
In addition to infiltrating the decidua and
myometrium, the implantation site IT cells
invade spiral arteries, extensively replacing the
wall and endothelium while maintaining the
integrity of the lumen (Figs. 3.16, 3.17, and
3.20). When vascular infiltration is extensive,
the IT cells may partially fill the lumen of the
vessel. This infiltration of vessels contributes to
the enlargement of the spiral arteries.
Besides their characteristic growth pattern
and cytologic features, implantation site IT
cells can be recognized because they typically
are associated with patches of extracellular
eosinophilic fibrinoid material (Fig. 3.21). This
fibrinoid matrix in the placental bed eventually
becomes the so-called Nitabuch’s layer. When
this fibrinoid material becomes disrupted in
abortions, it forms hyaline, eosinophilic strands
termed Rohr’s stria. The origin of this fibrinoid,
hyaline material is not fully understood,
although it appears to be partly composed of
fibronectin, laminen, type IV collagen, and a
small amount of fibrin. In any event, fibrinoid is
a distinctive part of the implantation site.
Placental site fibrinoid can resemble fibrin
thrombi that occur as a result of chronic bleed-ing from a variety of causes. Also, endometrial
stroma can become hyalinized and fibrotic
in areas of continued breakdown such as the
surface of polyps, and this change, too, can
mimic the fibrinoid deposition of the implanta-tion site. Fibrin thrombi and hyalinized
endometrial stroma are distinguished from
fibrinoid by the absence of interspersed IT cells
and the lack of the linear deposits of
eosinophilic implantation site fibrinoid.
Changes in the spiral arteries of the placen-tal site are significant. The presence of enlarged,
hyalinized spiral arteries in the decidua of
curettage specimens is a valuable adjunct in
diagnosing intrauterine gestation.47
These ves-sels have thickened hyalinized walls that are
52 3. Pregnancy, Abortion, and Ectopic Pregnancy
Figure 3.20. IT cells. Prominent infiltration of blood vessels in the decidua by implantation site IT in an
abortion specimen. The cells infiltrate and replace the wall but preserve the lumen.
partially infiltrated by IT and show an in-creased luminal diameter (Figs. 3.16, 3.17, and
3.20). Often several implantation site vessels
are seen in cross section forming a prominent
cluster. These vascular changes, like the pres-ence of IT in the decidua, are characteristic
of the implantation site and are not found
in endometrium associated with ectopic
pregnancy.
Histologic recognition of the placental
implantation site usually is straightforward.
Sometimes IT cells are indistinct or are difficult
to distinguish from degenerating decidual cells
that develop dark nuclei. In these instances,
ancillary immunohistochemical techniques are
useful for the detection of IT. Broad-spectrum
keratin antibodies are very useful for demon-strating IT.35;47–49
Endometrial glands also
stain with keratin, however, and therefore all
keratin-positive cells do not represent IT cells.
It is the growth pattern along with immunore-activity for keratin that identifies IT cells. With
the keratin immunostain, the IT cells appear as
intensely staining single cells or irregular clus-ters of cells with intervening decidua or smooth
muscle that is nonreactive for keratin (Fig.
3.22). The IT cells, unlike decidual cells, also
express hPL, Mel-CAM (CD 146), and, to a
lesser degree, hCG.34
Immunostains, especially
hPL, are useful for detecting IT cells including
the multinucleated IT in the myometrium, as
smooth muscle cells do not express hPL.35;47–49
The implantation site IT shows no proliferative
activity and should not be reactive with Ki-67.50
Exaggerated Placental
Implantation Site
An exaggerated placental site represents one
end of the spectrum of the morphologic fea-Trophoblast and Villi 53
Figure 3.21. Fibrinoid (Rohr’s stria). A strand of
densely eosinophilic fibrinoid material of the pla-cental implantation site overlies the decidua in curet-tings from an abortion specimen. A few IT are
enmeshed in the fibrinoid (arrows).
tures of the normal implantation site.43;44
It is
not a tumor. It is an unusually prominent but
physiologic placental site that may be difficult
to distinguish from a placental site trophoblas-tic tumor (PSTT) (see Chapter 4). In complete
molar pregnancy the placental site is typically
exaggerated, but exaggerated placental site can
occur in association with a normal gestation as
well. The exaggerated placental site is charac-terized by an increase in the number and size
of individual IT cells. In addition, widely dis-persed multinucleated IT cells are a compo-nent of the trophoblastic infiltrate. Often
several fragments of tissue in curettage
samples contain portions of the lesion, and this
process can extensively infiltrate fragments of
myometrium. A few chorionic villi may be
present. In the exaggerated placental site, IT
cells appear larger and more hyperchromatic
than normal. Despite their apparent promi-nence, these IT cells show no mitotic activity,
and Ki-67 immunostaining is zero when the
exaggerated implantation site is associated with
an abortus.50
The Ki-67 index can be slightly
elevated in exaggerated implantation sites asso-ciated with a complete mole. It is important to
note that lymphocytes that typically are present
in the implantation site often express Ki-67.
These lymphocytes should not be misinter-preted as IT cells. Double staining with Mel-CAM and Ki-67 can be very helpful in localiz-ing Ki-67 to Mel-CAM–positive IT cells.
Necrosis is not a feature of the exaggerated pla-cental site, although the surrounding decidua
often shows degeneration and necrosis typical
of spontaneous abortions. PSTT is an important
consideration in the differential diagnosis of
this lesion. The distinction is largely a matter of
degree, as discussed in Chapter 4.
Placental Site Nodules
Placental site nodules are small, circumscribed
foci of hyalinized implantation site with IT cells
that occasionally present in an endometrial
biopsy or curettage.40;43;44;51–53
These benign
lesions occur in women of reproductive age,
although often the pregnancy history is
remote.51;54
Usually they are incidental findings,
although they may be associated with abnormal
uterine bleeding. The nodules may be present
in biopsies taken several years after tubal
ligation, suggesting that they are retained in
the endometrium for extended periods of
time.40;51;54
The known antecedent pregnancy
dates back 2 to 108 months, demonstrating
the long duration of some lesions. They show a
propensity for the lower uterine segment and
54 3. Pregnancy, Abortion, and Ectopic Pregnancy
Figure 3.22. Keratin immunoreac-tivity of intermediate trophoblast.
Scattered IT in decidua show
cytoplasmic staining for keratin
(arrows).
cervix. The surrounding endometrium often is
proliferative or secretory, and usually is not
decidualized.
Generally these lesions are microscopic,
although hysterectomies may yield gross lesions
as large as 1 or 2 cm in diameter. Occasionally,
multiple nodules are present. In the past these
nodules and plaques were considered hyalin-ized decidua, but they are now recognized as
a distinctive benign lesion of IT. The lesion
itself is circumscribed, nodular, or plaque-like
with densely eosinophilic, hyalinized stroma
containing aggregates of IT cells (Fig. 3.23).
Often focal chronic inflammation including
plasma cells surrounds the nodule, while the rest
of the endometrium shows no inflammation.
The trophoblastic cells in these nodules resem-ble chorionic-type IT. In the placental site
nodule the cells vary in size; many have small,
uniform nuclei and some larger cells show irreg-ular, hyperchromatic nuclei. Occasional multi-nucleated cells are present. (Fig. 3.24). Mitoses
are rare or absent, and the cells show a low
Ki-67 labeling index of <10%.
The IT of the placental site nodule demon-strate immunoreactivity that is similar to that
seen in chorionic-type IT. The cells are strongly
reactive for keratin and epithelial membrane
antigen as well as PLAP, inhibin-a , and
pregnancy-specific SP-1 (Fig. 3.25).51;54;55
Other
trophoblastic markers such as hPL and Mel-CAM (CD 146) may be positive but only in a
few cells, and hCG reactivity is usually absent.
The small size, circumscription, and extensive
hyalinization are consistent features of the
lesions that help to separate them from the pla-cental site trophoblastic tumor and epithelioid
trophoblastic tumor discussed in Chapter 4.
Trophoblast and Villi 55
Figure 3.23. Placental site nodule. Well-circum-scribed fragment of placental site nodule is present
in endometrial curettings. This microscopic focus is
composed of hyaline material with entrapped,
degenerate IT cells.
Figure 3.24. Placental site nodule. Placental site
nodule shows degenerate, vacuolated IT cells with
smudged nuclear chromatin surrounded by dense,
Figure 3.25. Placental site nodule. Immunohistochemical stain for keratin shows strong reactivity of the
trophoblast.
hyaline stroma. A chronic inflammatory infiltrate is
present at the periphery of the lesion.
Chorionic Villi and Villous
Trophoblast in the First Trimester
Villus formation in very early pregnancy
depends on the existence of the embryonic disc.
Villi begin to develop on the 12th to 13th day
postfertilization, and by days 12 to 15 the pla-centa can develop for a while independently,
without the presence of an embryo.4;56
In fact,
in spontaneous abortions the placenta can
persist for several weeks after the death of the
embryo. In the early stages of pregnancy, villi
have a loose, edematous stroma with few well-developed capillaries (Figs. 3.11 and 3.26). Once
the yolk sac and embryo develop, vascular cir-culation is established in the villous stroma and
these vessels contain nucleated red blood cells.
During this early period of placental develop-ment, the trophoblastic covering of the villi
consists of an inner layer of CT cells and an
outer layer of ST cells. The CT and IT cells also
proliferate at the implanting end of the anchor-ing villi that grow along the basal plate of the
developing placenta.
A few histologic changes in the placenta help
to determine the age of the developing con-ceptus (Table 3.4), although the length of ges-tation is an infrequent clinical question. These
developmental intervals are stated in relation
to the time of fertilization, also known as the
postcoital or postconception date. Taking into
account the time from the last known menstrual
period adds 2 weeks to these figures. The con-ceptus is traditionally called an embryo in the
first 2 months of development; thereafter it is
called a fetus.
As placental development proceeds, the
presence of nucleated erythrocytes produced in
the yolk sac helps determine the approximate
Trophoblast and Villi 57
Figure 3.26. Immature chorionic villi. Immature
villi from an abortion specimen show loose, edema-tous stroma containing a few capillaries with nucle-ated erythrocytes. Trophoblast emanates from one
pole of several villi, which is the implanting portion
of the anchoring villi.
gestational age.56
The nucleated erythrocytes
from the yolk sac appear in the villous circula-tion at 4.5 weeks. By 5 to 6 weeks, non-nucleated erythrocytes from the embryonic
liver also begin to appear in the villi, and from
this point on there is a shift in the proportion
of nucleated to non-nucleated erythrocytes. By
9 weeks postfertilization the percentage of
nucleated erythrocytes in the villi decreases
from 100% to only 10%. Consequently, if the
embryo dies before 4.5 weeks postfertilization,
the villi contain no red blood cells. Death of the
embryo between 4.5 and 10 weeks leaves a
mixture of nucleated and non-nucleated ery-throcytes in villous capillaries, and later death
of the embryo usually leaves non-nucleated red
cells in the stroma.
Other features also help to determine the
relative length of gestation. For example, nor-mally developing immature villi are relatively
larger than those of later pregnancy. They have
a loose, myxoid stroma with widely spaced
capillaries. As pregnancy progresses, the villi
become smaller but more vascular and their
stroma loses its edematous appearance. The
morphologic features of the trophoblast cover-ing the villi change along with growth of the
placenta. The bilayered CT and ST covering of
the villi persists to some extent throughout
gestation, but a visible inner layer of CT cells
starts to disappear at about 14 weeks of gesta-tion. Between week 14 and week 18, the
percentage of villi showing an inner layer of CT
cells decreases from 80% to 60%.4
From that
point on there is a continuing gradual decrease
in the CT layer. The large decrease in identifi-able CT cells by week 18 is especially useful
for determining the relative duration of early
pregnancy.
Hydropic Change and Other
Pathologic Changes in Abortions
The microscopic features of the decidua and
the products of conception in curettage samples
vary depending on the type of abortion.4;57;58
Villi are usually normal in therapeutic abor-tions, whereas they tend to reflect early death
of the embryo in spontaneous or missed
abortions. Therapeutic abortion specimens may
show pathologic changes in the villi, however.59
In spontaneous or missed abortions, placental
morphology is influenced by gestational age,
karyotype, and regressive changes.4;57;60–63
With
the death of the embryo, the villi often show
hydropic change because of loss of the villous
vascular supply, especially if embryonic death
occurs very early, often before 4.5 weeks post-fertilization age.56
The avascular villi are mildly
distended with fluid and the curettage samples
do not contain fetal tissue giving the changes
of the so-called blighted ovum (Fig. 3.27). This
pattern of mild villous edema and no evidence
of fetal development indicates that the embryo
either never developed or ceased development
at a very early stage of gestation. Microscopi-cally, villous edema in a hydropic abortion
can appear especially prominent at first
glance. Hydropic change affects most villi but is
minimal and microscopic; cistern formation in
the villi is rare but does occur (Fig. 3.28). There
is no associated trophoblastic hyperplasia
except the normal growth at one pole of the
anchoring villi. Usually these microscopic
abnormalities are less impressive when the
gross, quantitative aspects are considered also.
Hydropic abortions usually consist of one or
two cassettes of tissue with villi whereas moles
typically yield multiple cassettes. It is important
to recognize the changes of the blighted ovum
in order to separate an abortion with hydropic
changes from a hydatidiform mole (see
Chapter 4). Hydropic change also may be
58 3. Pregnancy, Abortion, and Ectopic Pregnancy
Table 3.4. Events in first trimester placental
development.a
Development of placenta Time after fertilization
Blastocyst implantation 6–7 days
Villus formation beginsb
12 days
Nucleated RBCs from yolk sac 4.5 weeks
appear in villi
Non-nucleated RBCs from liver 5–6 weeks
appear in villi
Proportion of nucleated RBCs 4.5–9 weeks
decreases from 100% to 10%
in villi
Decrease in prominence of 16–18 weeks
inner cytotrophoblast layer
RBCs, Red blood cells.
a
From time of ovulation. For menstrual age add 2 weeks.
b
After 12–15 days, placental development proceeds even
if embryo dies.
focally present in therapeutic abortions,59;64
especially if the tissue blocks sample villi of the
chorion leave, where the villi normally degen-erate, and mild hydropic change often is diffi-cult to distinguish from the loose, myxoid
stroma of the normal early placenta.
Other morphologic changes in chorionic villi
from first-trimester abortion specimens may
be found.4
Irregular outlines of villi yielding
a scalloped appearance and trophoblastic
invagination into the villous stroma forming
pseudoinclusions often are associated with
abnormal karyotypes of the conceptus, particu-larly triploidy, but the findings are not suffi-ciently specific by themselves to be diagnostic
of a chromosomal abnormality. Karyotyping is
necessary to determine whether a chromoso-mal abnormality is present.60;61;65
In incomplete abortions, the amount of
villous tissue may be greatly reduced or even
absent if portions of the placenta sponta-neously pass before curettage. The implantation
site, however, with its characteristic features,
usually is present. In missed abortions the villi
often are necrotic or hyalinized and the decidua
is necrotic. Another change in villi associated
with death of the embryo is loss of villous vas-cularity and fibrosis of the villous stroma (Fig.
3.29).62;66;67
This change occurs more frequently
in missed abortions. Some or many villi may be
necrotic. These villous changes have little if any
clinical significance.
Chorionic Villi and Villous
Trophoblast After the First Trimester
After the first trimester or early second
trimester, the placenta is sufficiently large that
it is delivered spontaneously or by induction,
and curettage specimens are rare. The villi are
Trophoblast and Villi 59
Figure 3.27. Immature chorionic villi with hydropic
change. Abortion specimen shows avascular, edema-tous chorionic villi. This change is associated with the
so-called blighted ovum and usually indicates very
early demise of the embryo. Although the villi are
edematous, the change is microscopic and not asso-ciated with hyperplasia of trophoblast. Furthermore,
there is a polar distribution of the trophoblast
toward the placental implantation site at the bottom
of the field. These features distinguish this hydropic
abortus from a partial mole.
60 3. Pregnancy, Abortion, and Ectopic Pregnancy
Figure 3.28. Immature chorionic villi with hydropic
change. This hydropic abortus shows scattered
cisterns. Despite the edema, this is a microscopic
Figure 3.29. Immature chorionic villi with villous
fibrosis. Abortion specimen showing immature villi
with absence of vessels and stromal fibrosis. Hyper-finding. There is no associated hyperplasia of the tro-phoblast, and this does not represent a hydatidiform
mole.
plastic trophoblast remains confined to one pole of
the villi.
more numerous and become more complex.
By this time there is a mixture of larger stem
villi with prominent vessels and central stromal
fibrosis, and many smaller tertiary villi contain-ing numerous capillaries.4
By late in the second
trimester, the inner CT cells are indistinct over
most villi and the trophoblastic proliferation
along the anchoring villi has largely ceased.
Some abortion specimens present early in the
second trimester following intrauterine fetal
death. In these cases the villous stroma often is
fibrotic and hypovascular, with either no resid-ual erythrocytes or a few residual degenerating
red blood cells present.
Following delivery of the term or near-term
placenta, abnormal bleeding may require curet-tage. The histologic findings in the postpartum
endometrium vary. Failure of the implantation
site to resolve quickly is called subinvolution.
With subinvolution the uterus contains rem-nants of necrotic decidua and the placental site
with eosinophilic fibrinoid, trophoblastic cells
and enlarged vessels that often are infiltrated by
the intermediate trophoblast. Retained placen-tal site also serves as a nidus for inflammation,
yielding inflamed and necrotic endometrium
and implantation site. Immunohistochemical
stains for keratin help to demonstrate the
residual intermediate trophoblastic cells.
Placental Polyps
Placental polyps are a form of retained product
of conception that represent polypoid portions
of chorionic villi from an incomplete abortion
or a term gestation retained in the uterine
cavity.4
The villi may be necrotic, hyalinized, or
partially calcified (Fig. 3.30). These polyps are
not tumors, but they do form a nidus for inflam-mation and bleeding. Often these pedunculated
masses of villi are found within days to weeks
following abortion or delivery of a term pla-centa. Rarely, they persist for months or years
after pregnancy.
Trophoblast and Villi 61
Figure 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.
Placenta Accreta
Placenta accreta is a form of abnormal implan-tation in which the placenta implants directly
onto the myometrium with no intervening
decidua; it may grow into or through the
myometrium (placenta increta and percreta,
respectively).4;62
Usually placenta accreta pre-sents immediately following delivery of a term
pregnancy when the placenta or portions of the
placenta cannot be delivered. The placental
tissue fails to detach from the implantation site
in the myometrium and cannot be manually
removed. The usual management of extensive
placenta accreta is hysterectomy, but on occa-sion focal placenta accreta is encountered in
curettage for postpartum hemorrhage. The
diagnostic feature is villi in direct apposition to
myometrium without intervening decidua (Fig.
3.31). Hyaline fibrinoid material with scattered
IT cells is interposed between the villi and
myometrium but decidual cells are absent.
Some placental polyps may represent focal
placenta accreta, although the latter are diag-nosed only when villi are contiguous with
myometrium. Scattered IT cells without villi in
the myometrium are a physiologic phenome-non and not placenta accreta.
Endometrium Associated with
Ectopic Pregnancy
In most circumstances, the endometrium asso-ciated with an ectopic pregnancy shows the
typical features of early gestation, yet tro-phoblastic tissue is not present. A decidualized
stroma, hypersecretory to atrophic glands,
and thick-walled spiral arterioles usually
are present.18;19;24;68;69
Often the Arias-Stella
reaction is present, at least focally. The
endometrium associated with ectopic gestation
can be highly variable, however, depending on
62 3. Pregnancy, Abortion, and Ectopic Pregnancy
Figure 3.31. Placenta accreta. Curetting for bleeding
following delivery of a term gestation shows mature
chorionic villi at the top of the field. Implantation site
IT and fibrinoid material but no intervening decidua
are all that separate the villi from the myometrial
smooth muscle at the lower portion of the figure.
the status of the trophoblastic tissue. If
ectopic trophoblast is actively proliferating, the
endometrium continues to show the changes of
pregnancy. If the trophoblast begins to regress,
the endometrium can display a variety of pat-terns ranging from proliferative to secretory
changes. The endometrium can show features
seen in dysfunctional bleeding, including
anovulatory bleeding patterns, abnormal secre-tory patterns, or progestin effects. Subtle clues,
such as a focal Arias-Stella reaction or a small
aggregate of gland cells with clear cytoplasm,
can suggest an ectopic pregnancy.
Establishing the presence of an intrauterine
pregnancy effectively rules out an ectopic
pregnancy. Evidence of an intrauterine preg-nancy includes chorionic villi, trophoblastic
cells, or the placental implantation site.
Occasionally individual ST giant cells may
be detected enmeshed in blood or fibrin. In
cases where there are no villi or ST cells, an
attempt should be made to identify IT cells
scattered in partially necrotic decidua, often
around spiral arterioles. Identification of
IT cells of the placental site is crucial, and in
such cases immunostaining for keratin can
help show scattered trophoblast, usually IT
cells in the decidua. Immunostains for hPL,
hCG, inhibin-a, and Mel-CAM specifically
identify trophoblast, but the reactivity is less
sensitive than that seen with keratin antibodies.
A panel using several of these latter immuno-stains can be helpful for identifying tropho-blastic tissue if the keratin immunostain is
inconclusive.
Clearly, adequate sampling of endometrial
tissue is important to ensure recognition of
chorionic tissue. We have empirically found that
three cassettes from abortion specimens are
usually sufficient to establish the presence of
chorionic tissue. If no trophoblast or villi are
present in the first three tissue blocks, all of the
residual tissue should be processed.
Clinical Queries and Reporting
For most endometrial biopsies or curettings
related to pregnancy, three clinical questions
need to be answered by pathologic examina-tion: (1) Does the endometrium show features
of pregnancy? (2) If the changes indicate preg-nancy, are chorionic villi and/or trophoblast
present? (3) If villi and/or trophoblast are
present, do they appear normal? For example,
endometrial changes of pregnancy without
the presence of chorionic villi or trophoblast
suggest the possibility of an ectopic pregnancy.
In spontaneous abortions the villi may reflect
pathologic development of the embryo, a
feature that helps explain the occurrence of the
abortion. In other cases, edematous villi or pro-liferative trophoblast can raise the question of
a hydatidiform mole, choriocarcinoma, or pla-cental site trophoblastic tumor (see Chapter 4).
The pathology report should consider the
clinical questions asked regarding pregnancy
as well as the pathologic findings. Most cases
require only documentation of the presence of
placental or fetal tissue. The most urgent ques-tion is that of an ectopic pregnancy. If preg-nancy is suspected, or if the morphology shows
pregnancy-induced endometrial patterns but
there is no evidence of chorionic villi, placental
site trophoblast, or fetal tissue, then an ectopic
pregnancy must be considered. Occasionally
the entire placenta with implantation site is
expelled before the curettage, so lack of
identifiable products of conception does not
unequivocally indicate ectopic pregnancy.
Nonetheless, an ectopic pregnancy can result in
sudden, life-threatening intra-abdominal hem-orrhage, so immediate notification of the clini-cian managing the patient is imperative. Also,
all the residual tissue should be processed. The
clinician should be informed if immunohisto-chemical stains to identify trophoblast are
pending or if residual tissue is being processed.
Another call should follow as soon as the
results are available.
Occasional spontaneous abortion specimens
show only a small amount of early placental site
with intermediate trophoblast and no chorionic
villi. This finding is sufficient to establish the
diagnosis of an intrauterine pregnancy. The
specimen represents products of conception,
and a descriptive diagnosis such as “implanta-tion site and decidua” or “intermediate tro-phoblast” serves to verify the presence of an
intrauterine gestation.
Clinical Queries and Reporting 63
When villi are present in first-trimester spec-imens, some morphologic findings may be clini-cally relevant. In spontaneous abortions, for
example, mild villous edema (hydropic change)
and absence of fetal tissues including erythro-cytes in villous vessels indicates that the gesta-tion was abnormal and may help the clinician in
the counseling of a patient. Although the micro-scopic findings can help indicate early death of
the embryo, cytogenetic analysis of tissue has
more value for assessing significant abnormali-ties that may lead to recurrent abortion. Some-times the term “hydropic villi” raises the specter
of hydatidiform mole, so this term should be
used cautiously unless the clinician fully under-stands the significance. If hydropic change is
diagnosed, it may be useful to add a comment to
indicate that this does not represent a mole.
Because pregnancy can be complicated by
gestational trophoblastic disease, the status of
the trophoblast, especially any abnormal pro-liferative activity, deserves comment. A speci-men containing more than a small amount of
trophoblast without villi, or an exaggerated pla-cental site without villi or unusually hydropic
villi that are not clearly molar (see Chapter 4)
should be reported. In such cases, a comment
regarding the uncertainty of the finding and re-commendation for follow-up with serum hCG
titers helps in the management of the patient.
The placental site nodule can be a confusing
diagnosis to the gynecologist. The terminology
for this lesion is relatively new. Although these
lesions are almost always microscopic and inci-dental, confusion with PSTT may arise. It is
therefore important to indicate clearly the
small and benign nature of the lesion. This diag-nosis of placental site nodule may be perplex-ing to the gynecologist because the patient
often has no recent history of pregnancy and
may even have had a tubal ligation, so a
comment regarding the fact that the gestation
may have been remote helps the clinician
understand the lesion.
References
(1) Mishell DRJ. Spontaneous and recurrent abor-tion. In: Stenchever MA, Droegemueller W,
Herbst AL, Mishell DR, Jr, eds. 2nd ed. Com-prehensive gynecology. St. Louis: Mosby; 2001:
413–441.
(2) Kalousek DK. Pathology of abortion: Chromo-somal and genetic correlations. In: Kraus FT,
Damjanov I, Kaufman N, eds. Pathology of
reproductive failure. Baltimore: Williams &
Wilkins; 1991: 228–256.
(3) Miller JF, Williamson E, Glue J, Gordon YB,
Grudzinskas JG, Sykes A. Fetal loss after
implantation. A prospective study. Lancet 1980;
2:554–556.
(4) Benirschke K, Kaufmann P. Pathology of the
human placenta, 4th ed. New York: Springer-Verlag, 2000.
(5) Wilcox AJ, Weinberg CR, O’Conner JF, Baird
DD, Schlatterer JP, Camfield RE et al. Inci-dence of early loss of pregnancy. N Engl J Med
1988; 319:189–194.
(6) Simpson JL. Incidence and timing of pregnancy
losses. Am J Med Genet 1990; 35:165–173.
(7) Mazur MT, Duncan DA, Younger JB. Endome-trial biopsy in the cycle of conception. Histo-logic and lectin histochemical evaluation. Fertil
Steril 1989; 51:764–767.
(8) Foss BA, Horne HW, Hertig AT. The
endometrium and sterility. Fertil Steril 1958;
9:193–206.
(9) Karow WG, Gentry WC, Skeels RF, Payne SA.
Endometrial biopsy in the luteal phase of
the cycle of conception. Fertil Steril 1971;
22:482–495.
(10) Hertig AT. Gestational hyperplasia of
endometrium. A morphologic correlation of
ova, endometrium, and corpora lutea during
early pregnancy. Lab Invest 1964; 13:1153–1191.
(11) Wentz AC, Herbert CMI, Maxson WS, Hill GA,
Pittaway DE. Cycle of conception endometrial
biopsy. Fertil Steril 1986; 46:196–199.
(12) Andoh K, Yamada K, Mizunuma H, Michishita
M, Nakazato Y, Ibuki Y. Endometrial dating
in the conception cycle. Fertil Steril 1992;
58:1127–1130.
(13) Arronet GH, Bergquist CA, Parekh MC,
Latour JPA, Marshall KG. Evaluation of
endometrial biopsy in the cycle of conception.
Int J Fertil 1973; 18:220–225.
(14) Sulewski JM, Ward SP, McGaffic W. Endome-trial biopsy during a cycle of conception. Fertil
Steril 1980; 34:548–551.
(15) Lichtig C, Korat A, Deutch M, Brandes JM.
Decidual vascular changes in early pregnancy
as a marker for intrauterine pregnancy. Am J
Clin Pathol 1988; 90:284–288.
(16) Clement PB, Young RH, Scully RE. Nontro-phoblastic pathology of the female genital tract
64 3. Pregnancy, Abortion, and Ectopic Pregnancy
and peritoneum associated with pregnancy.
Semin Diagn Pathol 1989; 6:372–406.
(17) Holmes EJ, Lyle WH. How early in pregnancy
does the Arias–Stella reaction occur? Arch
Pathol 1973; 95:302–303.
(18) Arias-Stella J. Atypical endometrial changes
produced by chorionic tissue. Hum Pathol 1972;
3:450–453.
(19) Arias-Stella J. Atypical endometrial changes
associated with the presence of chorionic tissue.
AMA Arch Pathol 1954; 58:112–128.
(20) Arias-Stella J. The Arias-Stella reaction: Facts
and fancies four decades after. Adv Anat Pathol
2002; 9:12–23.
(21) Dardi LE, Ariano L, Ariano MC, Gould
VE. Arias-Stella reaction with prominent
nuclear pseudoinclusions simulating herpetic
endometritis. Diagn Gynecol Obstet 1982;
4:127–132.
(22) Beswick IP, Gregory MM. The Arias-Stella
phenomenon and the diagnosis of pregnancy. J
Obstet Gynaecol 1971; 78:143–148.
(23) Bernhardt RN, Bruns PD, Drose VE. Atypical
endometrium associated with ectopic preg-nancy. The Arias-Stella phenomenon. Obstet
Gynecol 1966; 23:849–853.
(24) LaIuppa MA, Cavanagh D. The endometrium
in ectopic pregnancy. A study based on 35
patients treated by hysterectomy. Obstet
Gynecol 1963; 21:155–164.
(25) Roach WR, Guderian AM, Brewer JI. Endome-trial gland cell atypism in the presence of tro-phoblast. Am J Obstet Gynecol 1960;
79:680–691.
(26) Huettner PC, Gersell DJ. Arias-Stella reaction
in nonpregnant women: A clinicopathologic
study of nine cases. Int J Gynecol Pathol 1994;
13:241–247.
(27) Yokoyama S, Kashima K, Inoue S, Daa T,
Nakayama I, Moriuchi A. Biotin-containing
intranuclear inclusions in endometrial glands
during gestation and puerperium. Am J Clin
Pathol 1993; 99:13–17.
(28) Mazur MT, Hendrickson MR, Kempson
RL. Optically clear nuclei. An alteration of
endometrial epithelium in the presence of
trophoblast. Am J Surg Pathol 1983; 7:415–
423.
(29) Genest DR, Brodsky G, Lage JA. Localized
endometrial proliferations associated with
pregnancy: Clinical and histopathologic fea-tures of 11 cases. Hum Pathol 1995; 26(11):
1233–1240.
(30) Schammel DP, Mittal KR, Kaplan K, Deligdisch
L, Tavassoli FA. Endometrial adenocarcinoma
associated with intrauterine pregnancy. A
report of five cases and a review of the litera-ture. Int J Gynecol Pathol 1998; 17:327–335.
(31) Kovacs AG, Cserni G. Endometrial adenocarci-noma in early pregnancy. Gynecol Obstet
Invest 1996; 41:70–72.
(32) Nakamura Y, Moritsuka Y, Ohta Y, Itoh S,
Haratake A, Kage M et al. S-100 protein in
glands within decidua and cervical glands
during early pregnancy. Hum Pathol 1989;
20:1204–1209.
(33) Agarwal S, Singh UR. Immunoreactivity with
S100 protein as an indicator of pregnancy.
Indian J Med Res 1992; 96:24–26.
(34) Shih I-M, Mazur MT, Kurman RJ. Gestational
trophoblastic disease and related lesions. In:
Kurman RJ, ed. Blaustein’s pathology of the
female genital tract. 5th ed. New York:
Springer-Verlag; 2002: 1193–1247.
(35) Kurman RJ, Main CS, Chen HC. Intermediate
trophoblast: A distinctive form of tro-phoblast with specific morphological, biochem-ical and functional features. Placenta 1984; 5:
349–370.
(36) Kurman RJ, Young RH, Norris HJ, Main CS,
Lawrence WD, Scully RE. Immunocytochemi-cal localization of placental lactogen and chori-onic gonadotropin in the normal placenta and
trophoblastic tumors, with emphasis on inter-mediate trophoblast and the placental site tro-phoblastic tumor. Int J Gynecol Pathol 1984;
3:101–121.
(37) Yeh IT, O’Connor DM, Kurman RJ. Intermedi-ate trophoblast: further immunocytochemical
characterization. Mod Pathol 1990; 3:282–287.
(38) Yeh I-T, Kurman RJ. Functional and morpho-logic expressions of trophoblast. Lab Invest
1989; 61:1–4.
(39) Wells M, Bulmer JN. The human placental bed:
Histology, immunohistochemistry and pathol-ogy. Histopathology 1988; 13:483–498.
(40) Shih IM, Seidman JD, Kurman RJ. Placental
site nodule and characterization of distinctive
types of intermediate trophoblast. Hum Pathol
1999; 30:687–694.
(41) Shih IM, Kurman RJ. Immunohistochemical
localization of inhibin-alpha in the placenta and
gestational trophoblastic lesions. Int J Gynecol
Pathol 1999; 18:144–150.
(42) Shih IM, Kurman RJ. Expression of melanoma
cell adhesion molecule in intermediate tro-phoblast. Lab Invest 1996; 75:377–388.
(43) Silverberg SG, Kurman RJ. Tumors of the
uterine corpus and gestational trophoblastic
disease. Atlas of tumor pathology, 3rd series,
References 65
Fascicle 3. Washington, DC: Armed Forces
Institute of Pathology, 1992.
(44) Kurman RJ. The morphology, biology, and
pathology of intermediate trophoblast—a look
back to the present. Hum Pathol 1991;
22:847–855.
(45) Young RE, Kurman RJ, Scully RE. Prolifera-tions and tumors of intermediate trophoblast of
the placental site. Semin Diagn Pathol 1988;
5:223–237.
(46) Wan SK, Lam PWY, Pau MY, Chan JKC.
Multiclefted nuclei. A helpful feature for iden-tification of intermediate trophoblastic cells in
uterine curetting specimens. Am J Surg Pathol
1992; 16:1226–1232.
(47) O’Connor DM, Kurman RJ. Intermediate tro-phoblast in uterine curettings in the diagnosis
of ectopic pregnancy. Obstet Gynecol 1988;
72:665–670.
(48) Daya D, Sabet L. The use of cytokeratin as a
sensitive and reliable marker for trophoblastic
tissue. Am J Clin Pathol 1991; 95:137–141.
(49) Angel E, Davis JR, Nagle RB. Immunohisto-chemical demonstration of placental hormones
in the diagnosis of uterine versus ectopic preg-nancy. Am J Clin Pathol 1985; 84:705–709.
(50) Shih IM, Kurman RJ. Ki-67 labeling index in
the differential diagnosis of exaggerated pla-cental site, placental site trophoblastic tumor,
and choriocarcinoma: a double immunohisto-chemical staining technique using Ki-67 and
Mel-CAM antibodies. Hum Pathol 1998;
29:27–33.
(51) Young RH, Kurman RJ, Scully RE. Placental
site nodules and plaques. A clinicopathologic
analysis of 20 cases. Am J Surg Pathol 1990;
14:1001–1009.
(52) Santos LD, Fernando SS, Yong JL,
Killingsworth MC, Wu XJ, Kennerson AR.
Placental site nodules and plaques: A clinico-pathological and immunohistochemical study
of 25 cases with ultrastructural findings.
Pathology 1999; 31:328–336.
(53) Shih IM, Kurman RJ. The pathology of inter-mediate trophoblastic tumors and tumor–like
lesions. Int J Gynecol Pathol 2001; 20:31–47.
(54) Huettner PC, Gersell DJ. Placental site nodule:
A clinicopathologic study of 38 cases. Int J
Gynecol Pathol 1994; 13:191–198.
(55) Shitabata PK, Rutgers JL. The placental site
nodule: An immunohistochemical study. Hum
Pathol 1994; 25:1295–1301.
(56) Szulman AE. Examination of the early concep-tus. Arch Pathol Lab Med 1991; 115:696–700.
(57) Rushton DI. Examination of products of con-ception from previable human pregnancies. J
Clin Pathol 1981; 34:819–835.
(58) Redline RW, Zaragoza M, Hassold T.
Prevalence of developmental and inflammatory
lesions in nonmolar first-trimester spontaneous
abortions. Hum Pathol 1999; 30(1):93–100.
(59) Jurkovic I, Muzelak R. Frequency of pathologic
changes in the young human chorion in thera-peutic abortions of normal pregnancies: A
report of 500 cases studied histologically. Am J
Obstet Gynecol 1970; 108:382–386.
(60) Vanlijnschoten G, Arends JW, Delafuente AA,
Schouten HJA, Geraedts JPM. Intra-observer
and inter-observer variation in the interpreta-tion of histological features suggesting chromo-somal abnormality in early abortion specimens.
Histopathology 1993; 22:25–29.
(61) Novak R, Agamanolis D, Dasu S, Igel H, Platt
M, Robinson H, et al. Histologic analysis of pla-cental tissue in first trimester abortions. Pediatr
Pathol 1988; 8:477–482.
(62) Gersell DJ, Kraus FT. Diseases of the placenta.
In: Kurman RJ, ed. Blaustein’s pathology of
the female genital tract. 5th ed. New York:
Springer-Verlag 2002: 1103–1191.
(63) Genest DR. Estimating the time of death in
stillborn fetuses: II. Histologic evaluation of
the placenta; a study of 71 stillborns. Obstet
Gynecol 1992; 80:585–592.
(64) Ladefoged C. Hydrop degeneration: A
histopathological investigation of 260 early
abortions. Acta Obstet Gynecol Scand 1980;
59:509–512.
(65) Rehder H, Coerdt W, Eggers R, Klink F,
Schwinger E. Is there a correlation between
morphological and cytogenetic findings in
placental tissue from early missed abortions?
Hum Genet 1989; 82:377–385.
(66) Fox H. Morphological changes in the human
placenta following fetal death. J Obstet
Gynaecol Br Commonw 1968; 75:839–843.
(67) Hustin J, Gaspard U. Comparison of histologi-cal changes seen in placental tissue cultures and
in placentae obtained after fetal death. Br J
Obstet Gynaecol 1977; 84:210–215.
(68) Buckley CH, Fox H. Biopsy pathology of the
endometrium. 2nd ed. London: Arnold, 2002.
(69) Romney SL, Hertig AT, Reid DE. The endome-tria associated with ectopic pregnancy. A
study of 115 cases. Surg Gynecol Obstet 1950;
91:605–611.
66 3. Pregnancy, Abortion, and Ectopic Pregnancy
disease are important, as they have different
clinical presentations and behavior.
Hydatidiform mole, either partial or com-plete, is the most common form of GTD, and
this also is the trophoblastic lesion most com-monly encountered in endometrial curettings.2
Choriocarcinoma, PSTT, and epithelioid tro-phoblastic tumor are infrequent. Recognition
of any of these lesions can be difficult, however,
because the morphologic features of all forms
of GTD overlap with the features of placental
and trophoblastic growth encountered in early
pregnancy, abortion, and persistent placental
implantation sites.
Hydatidiform Mole
General Features
Hydatidiform mole, either partial or complete,
is infrequent in the United States and Europe,
occurring in about one in 1000 to one in 2000
pregnancies,4–6
although some studies have sug-gested that partial mole may be even more fre-quent, occurring in up to 1 in 700 pregnancies.7
In other parts of the world, including Asia and
Latin America, these disorders are more com-mon, although problems in methodology often
complicate studies of their frequency when
deliveries take place at home.2;8;9
The separation of hydatidiform mole into two
subtypes, complete and partial, represents a sig-nificant advance in our understanding of molar
pregnancy. These two forms of hydatidiform
67
4
Gestational Trophoblastic Disease
Gestational trophoblastic disease (GTD)
includes disorders of placental development
(hydatidiform mole) and neoplasms of the
trophoblast (choriocarcinoma, placental site
trophoblastic tumor [PSTT], and epithelioid
trophoblastic tumor).1;2
The recent classifica-tion of these lesions by the World Health
Organization (WHO) clearly defines the differ-ent histologic forms of GTD (Table 4.1).3
A
common feature of all of these trophoblastic
lesions is that they produce human chorionic
gonadotropin (hCG), which serves as a marker
for the presence of persistent or progressive
trophoblastic disease. Because these lesions,
especially postmolar trophoblastic disease, are
often treated in the absence of a histologic diag-nosis, they may be clinically classified as GTD
without designation of the morphologic
subtype. Nonetheless, identification and separa-tion of the different pathologic forms of the
Hydatidiform Mole  . . . . . . . . . . . . . . . . . . 67
Persistent Postmolar Gestational
Trophoblastic Disease and Invasive
Hydatidiform Mole  . . . . . . . . . . . . . . . . 79
Clinical Queries and Reporting of
Hydatidiform Mole  . . . . . . . . . . . . . . . . 80
Trophoblastic Neoplasms  . . . . . . . . . . . . . 81
Choriocarcinoma  . . . . . . . . . . . . . . . . . . 81
Placental Site Trophoblastic
Tumor  . . . . . . . . . . . . . . . . . . . . . . . . . . 85
Epithelioid Trophoblastic Tumor  . . . . . . 91
Clinical Queries and Reporting of
Trophoblastic Neoplasms . . . . . . . . . . . . 94
mole have different cytogenetic patterns that
are accompanied by different clinicopathologic
profiles and different degrees of risk for the
development of persistent GTD (Table 4.2).
1;2;7;10–13
Both forms of mole typically present in
the first trimester, often as an abortion.
Complete mole, also known as “classic”
hydatidiform mole, has been recognized and
studied for many years.4
The clinical presenta-tion of complete mole has changed in recent
years. In the past typical complete moles pre-sented at approximately the 16th week of
pregnancy, but with the widespread use of
ultrasound in prenatal assessment, many moles
are now detected earlier in gestation.14;15
Cur-rently, most cases of complete mole seen in
North America and Europe present as a spon-taneous or missed abortion between 6 and 18
weeks of gestation with a mean of about 11 to
12 weeks. Complete moles commonly present
with uterine enlargement greater than that
expected for the gestational age, and the patient
may have signs or symptoms of toxemia of
pregnancy. Abortion with abnormal bleeding
and passage of molar tissue is a frequent pre-sentation. Although historically many complete
moles were diagnosed before curettage, in
current practice complete mole may be unde-tected clinically and diagnosed only when the
pathology is reviewed. The serum  b-human
chorionic gonadotropin (b-hCG) titers may be
markedly elevated, but levels are often unreli-able for establishing the diagnosis. However, a
b-hCG titer above 82,350 mIU/ml, coupled with
absence of fetal heart movement, is correlated
with the presence of hydatidiform mole.16
Partial moles tend to present slightly later in
gestation, occurring between 8 and 22 weeks of
gestation with a mean of about 14 weeks. These
moles also may be clinically unsuspected, pre-senting as a spontaneous or missed abortion
(Table 4.2).10;11
The uterus often is small for ges-tational age. Serum b-hCG titers are in the low
or normal range for that time in pregnancy, and
toxemia is less frequent than in the case of com-plete moles. The subtle clinical presentation of
most moles, either complete or partial, under-scores the need for careful pathologic evalua-tion of abortion specimens.17
Persistent GTD following a complete or
partial mole is detected by serum hCG titers
that fail to return to normal. The risk of persis-tent GTD is greater with complete mole. Up to
20% or more of patients with complete mole
require further therapy, usually chemotherapy,
for a plateau or increase in the serum hCG titer
after evacuation.2;18–20
About 2% to 3% of pati-ents with complete mole will develop chorio-carcinoma. Among patients with partial mole,
68 4. Gestational Trophoblastic Disease
Table 4.1. Modified World Health Organization
classification of gestational trophoblastic disease.
Molar lesions
Hydatidiform mole
Complete
Partial
Invasive mole
Nonmolar lesions
Choriocarcinoma
Placental site trophoblastic tumor
Epithelioid trophoblastic tumor
Miscellaneous trophoblastic lesions
Exaggerated placental sitea
Placental site nodulea
Unclassified trophoblastic lesion
a
See Chapter 3 for a description of these lesions.
Table 4.2. Clinical and cytogenetic comparison of hydatidiform moles and hydropic abortus.
Complete mole Partial mole Hydropic abortus
Presentation Spontaneous abortion Missed abortion Spontaneous or missed abortion
Gestational age 8–18 weeks 8–22 weeks 6–14 weeks
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 tissue Variable, may be increased Variable, usually decreased Usually decreased
Karyotype 46XX (all paternal) 69XXY or XXX (2 : 1, Variable, often abnormal
paternal : maternal)
Persistent GTD 15%–20% 0.5%–5% None
hCG, Human chorionic gonadotropin.
the risk of persistent GTD is much lower, as it
occurs in from 0.5% to 5% of cases.2;10;11;21–25
Most cases of persistent GTD represent persis-tent mole in the uterine cavity or invasive mole
in the myometrium. Less often a patient has
invasive mole with villi and trophoblast that
migrates to the lungs, vulva, or vagina, or
lung metastases that are not biopsied for
classification.26
Development of choriocarci-noma following a partial mole is a very rare
sequela.25;27;28
Genotyping and chromosome in
situ hybridization analysis has shown that some
cases of genetically confirmed partial mole do
have metastatic trophoblastic disease with lung
and liver lesions.28a
Cytogenetics
Cytogenetic studies show that complete mole
has a normal diploid DNA content, usually
46XX. The entire chromosomal complement,
however, is paternal, resulting from duplication
of a haploid paternal genome (23X); the com-plete mole lacks maternal DNA.29
More than
90% of complete moles contain this composi-tion of paternal chromosomes.12;30–32
The
remaining complete moles also are androge-netic but are 46XY and formed by dispermy, that
is, fertilization by two spermatozoa of an ovum
lacking functional maternal chromosomes.31;33;34
Cytogenetically, partial mole usually is
triploid (69 chromosomes) with two sets of
chromosomes of paternal origin (diandric) and
a haploid maternal set.12;34–36
In more than two
thirds of cases, the triploid chromosomal com-position is 69XXY; less often it is XXX, and
rarely it is XYY.34
Cytogenetic distinctions
between complete and partial moles have not
been absolute, however. There have been
reports of partial moles that are diploid,37–39
and
occasional complete moles that are triploid.37;40
Furthermore, both complete and partial moles
have been reported to show marked hetero-geneity in ploidy patterns. Haploid, aneuploid,
and tetraploid moles, both partial and com-plete, have been reported.40;41
DNA ploidy
analysis can be helpful in classifying cases that
lack clear-cut morphologic features to identify
the mole as either complete or partial.38;40–43
A
review of apparent nontriploid partial moles
found that errors in pathologic or ploidy classi-fication were common and suggested that non-triploid partial moles may not exist.44
Recently, immunohistochemical analysis of
the paternally imprinted gene product p57KIP2
has been used in the differential diagnosis of
moles.45–48
The p57KIP2
gene is expressed pre-dominantly from the maternal allele in most
tissues, and, because complete moles contain
only paternal DNA, p57KIP2
is underexpressed.
As a consequence, antibodies against the p57KIP2
protein generally show little to no expression in
cytotrophoblastic cells and villous stromal cells
of complete mole. In contrast, the cytotro-phoblastic cells and villous mesenchyme of
partial moles as well as spontaneous abortions
show strong immunoreactivity for p57KIP2
.47;48
A
rare case of complete mole has shown cytotro-phoblast and villous stromal cell immunoreac-tivity for p57KIP2
, however.47
In addition, p57KIP2
immunostaining can be seen in villous interme-diate trophoblastic cells48
and in intervillous
trophoblast islands composed of intermediate
trophoblast within complete moles.47
Syncy-tiotrophoblastic cells are nonreactive for p57KIP2
and maternal decidualized stromal cells are
strongly reactive for p57KIP2
in all types of ges-tations. In our experience, this antibody has not
proven sufficiently specific to be useful for the
subclassification of hydatidiform mole. Accord-ingly, the p57KIP2
immunohistochemical findings
should be interpreted with caution when using
this antibody to assist in the distinction of com-plete versus partial moles.
To date, most studies find no consistent asso-ciation between DNA ploidy and the subse-quent clinical course of either partial or
complete mole.40
One report suggested that
aneuploidy predicts persistence in complete
moles,49
but another study found that aneuploid
complete mole is associated with less risk for
progressive disease than diploid or tetraploid
complete moles.50
Pathologic Features
Hydatidiform mole is one of the few curettage
specimens that can have distinctive gross fea-tures, namely large, translucent villi.2
Tissue
recovered from molar pregnancies in these cases
is voluminous, especially in complete mole. The
grapelike villi typically range from several mil-Hydatidiform Mole 69
limeters to 2.0 cm or more in the largest dimen-sion (Table 4.3). Often, however, villi are not
grossly visible. This is attributable to early ges-tational age,14;17
villi being passed spontaneously
before curettage, or villi collapsing during the
suction curettage. Besides the presence of
grossly edematous villi, specimens from partial
moles may contain remnants of a fetus.
Microscopically, the villi in both complete
mole and partial mole show circumferential
hyperplasia of trophoblast and cistern forma-tion.1;2
In partial mole, the villous abnormalities
affect only a portion of the placenta, resulting
in two populations of villi.1;2;10;13
In complete
mole, edema affects all villi, although the
degree of enlargement caused by the edema is
variable. Sometimes one of the two features,
circumferential hyperplasia or edema, predom-inates, but both features should be present to
establish the diagnosis. When several hydropic
villi are sufficiently large to fill a microscopic
field under a ¥10 objective lens, the diagnosis of
hydatidiform mole is probably established.
Complete Mole
The morphologic features of complete mole
differ to some extent depending on the gesta-tional age. Consequently, the diagnosis of com-plete mole requires recognition of both classic
changes and the more subtle changes of very
early complete mole. In fact, classic complete
mole is now relatively uncommon, as many
moles are detected earlier in pregnancy
because of the routine use of ultrasound per-formed in early pregnancy. Whether a complete
mole is detected very early in gestation or at a
more advanced stage, both diffuse villous
edema and irregular hyperplasia of trophoblast
from the surface of some villi should be seen.
The features of classic mole are considered first,
followed by a description of the early complete
mole, defined as a complete mole detected prior
to 12 weeks of gestation. Clearly there are tran-sitions between these patterns.
Classic features of a complete mole obtained
at 16 to 18 weeks of gestation include a volu-minous gross specimen showing many large,
grapelike translucent villi. Microscopically,
these advanced complete moles show marked
villous edema with cistern formation (Figs. 4.1
and 4.2). A cistern is a completely acellular
central cavity within a villus that is filled with
edema fluid and surrounded by a sharply
demarcated stromal border. In complete mole,
many, but not all, of the villi show cistern for-mation, although  all the villi are edematous
(Fig. 4.3). Scattered small villi without cisterns
often are admixed (Fig. 4.1). In addition to
these changes, some villi may be necrotic and
occasional villi can show partial calcification.
Because fetal development ends very early in
placentation, the villi usually do not have
visible blood vessels. Although stromal vessels
are indistinct, occasional vessels in the villous
stroma can be found. In fact, CD34 immunos-taining will show numerous blood vessels; these
may contain cellular debris.51
Circumferential trophoblastic hyperplasia
seen in complete mole is characterized by
masses of trophoblast, often confluent, that
70 4. Gestational Trophoblastic Disease
Table 4.3. Comparison of pathologic features of complete mole, partial mole, and hydropic abortus.
Complete mole Partial mole Hydropic abortus
Villous hydrops (swelling) Generalized, often grossly visible Partial, often grossly visible Microscopic, limited
Villous shape Round to bulbous Irregular, scalloped Round, small
Cisterns Present Present Usually absent
Trophoblastic inclusions Rare Common Rare
Fetal tissue Rare Common Rare
Trophoblast
Distribution Circumferential, multifocal Circumferential, multifocal Polar at anchoring villi only
Proliferation Variable, may be marked Focal, minimal, with ST sprouts Limited to anchoring villi
Atypia Often present Rare Absent
Implantation site Exaggerated Normal or exaggerated Normal
ST, Syncytiotrophoblastic.
Figure 4.1. Complete hydatidiform mole. Low-magnification view shows generalized villous edema
and marked enlargement of many villi. The massively
distended villus to the right of center has a central
cistern. Irregular, haphazard hyperplasia of tro-phoblast is present along the surface of several villi.
Figure 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.
project randomly along the surface of
edematous villi (Figs. 4.1 and 4.2).1;2;4;52;53
The
trophoblastic hyperplasia can involve smaller
villi as well as the large villi with cisterns. This
haphazard trophoblastic proliferation contrasts
with the normal trophoblastic proliferation in
an immature placenta that maintains polar ori-entation at the anchoring villi and is not present
on other villi. The degree of trophoblastic
hyperplasia in a complete mole is highly vari-able. Often the hyperplasia is moderate to
marked with large masses of trophoblast that
may be confluent, extending from the surface of
the villi. Occasional cases, however, have only
minimal trophoblastic hyperplasia.
In early complete mole, detected at a younger
gestational age (<12 weeks), the villous features
are subtle and well-formed cisterns may be
absent.14;15;54–56
These early complete moles
demonstrate less advanced degrees of villous
abnormalities at both the gross and microscopic
levels. The gross specimen may not show
obvious hydropic villi. Microscopically, early
complete moles show bulbous, cauliflower-like
terminal villi with hypercellular villous stroma
and karyorrhexis (Figs. 4.4 and 4.5).14;54;57
At
this stage, villi show relatively smooth contours.
Even these small villi show edema, however,
with sparsely cellular stroma showing widely
separated fibroblasts. In very early complete
mole the trophoblastic hyperplasia tends to
be focal but definite, with proliferation of
both cytotrophoblastic (CT) and syncytiotro-phoblastic (ST) cells from the villous surface. If
present, the ST along the chorionic plate also is
hyperplastic.
The trophoblast that accompanies complete
moles also often shows atypia, with enlarged,
pleomorphic, and hyperchromatic nuclei.
Mitotic activity may be brisk. These findings
contrast with those found in the trophoblast
of an abortus that does not display marked
nuclear atypia. In addition, “layering” of atypi-cal implantation type trophoblast on fibrin is a
72 4. Gestational Trophoblastic Disease
Figure 4.3. Complete hydatidiform mole. Portions
of edematous, avascular villi in a complete mole
show hyperplasia of the trophoblastic covering. A
portion of a cistern is present in the villus to the left
of center. Note the smaller villus with edema in the
right upper corner. Occasional smaller villi such as
these are commonly found in complete mole
specimens.
Figure 4.4. Early complete hydatidiform mole. The
villi are bulbous with mildly hypercellular stroma
that shows minimal edema with no well-formed cis-Figure 4.5. Early complete hydatidiform mole.
Portion of a villus from early complete mole shows
minimal edema and karyorrhexis of the stroma. The
terns. The trophoblast show circumferential hyper-plasia, however, which distinguishes this early com-plete mole from a hydropic abortus.
trophoblast demonstrates slight but definite circum-ferential hyperplasia with minimal atypia.
very characteristic feature of complete moles.
The amount of trophoblast and the degree of
atypia present in moles have no apparent
bearing on the subsequent clinical course, so
grading the trophoblast is not helpful.53
With hydatidiform mole, especially complete
mole, the trophoblastic infiltration of the pla-cental implantation site typically is exagger-ated, even in early complete mole (see Chapter
3).14;58;59
Curettage, especially sharp curettage
after suction extraction, can yield abundant,
atypical trophoblast including many intermedi-ate trophoblastic cells from the implantation
site (Fig. 4.6). This trophoblastic proliferation is
a standard feature of hydatidiform moles, and
should not be misinterpreted as a coexisting
placental site trophoblastic tumor. As with the
trophoblastic proliferation that covers the villi,
the amount of trophoblast and the atypia in the
implantation site do not influence the diagnosis
or prognosis of these lesions.
Partial Mole
Partial mole, as the name implies, shows only
partial involvement of villi by edema and tro-phoblastic hyperplasia (Table 4.3).1;2;10;13;60
The
result is two populations of villi, one composed
of enlarged and hydropic and one of small, non-molar villi that do not show edema (Figs. 4.7 to
4.9). Frequently the nonedematous villi are
fibrotic, especially in partial moles that are
greater than 12 weeks.17
Typically the enlarged
villi have irregular, scalloped borders with deep
infoldings (Figs. 4.7 and 4.8) that contrast to the
smooth or rounded contour of villi in complete
mole. Transverse sectioning of the invaginations
yields trophoblastic “inclusions” in the villous
74 4. Gestational Trophoblastic Disease
Figure 4.6. Exaggerated placental implantation site
of complete mole. Abundant, atypical trophoblastic
cells are present at the implantation site of a com-plete mole. The trophoblastic cells in this field are not
associated with villi, but other fields showed
markedly edematous villi of a complete mole. This
exaggerated placental site has no significance by
itself.
Figure 4.7. Partial hydatidiform mole. A mixture of
large, edematous villi and small, fibrotic villi charac-terizes partial mole. A markedly enlarged villus to
Figure 4.8. Partial hydatidiform mole. Many of the
enlarged villi show irregular outlines. Note tro-phoblastic inclusion at the right of the field formed
the right contains a central cistern. Irregular, patchy
hyperplasia of trophoblast is present along the
surface of the larger villi.
by invagination of the surface of the villi into the
stroma (arrow).
stroma. Trophoblastic hyperplasia usually is
limited, with only small foci of syncytiotro-phoblast projecting randomly from the surface
of the affected villi (Fig. 4.9). Another frequent
finding in partial moles is microscopic evidence
of fetal development, such as fetal tissue, ery-throcytes in villous capillaries, or fetal mem-branes. Fetal tissue is not invariably present in
partial mole, however, and in some studies this
feature has been found in fewer than one half
of cases.61
The implantation site trophoblast
usually shows only focal and mild atypia com-pared to the implantation site seen in complete
mole.59
Differential Diagnosis
Complete Versus Partial Mole
The distinction between complete and partial
mole is straightforward when the characteristic
features of either entity are pronounced. Mul-tiple edematous villi with large cisterns and
diffuse trophoblastic hyperplasia of complete
moles contrast with the more limited villous
edema and focal trophoblastic hyperplasia of
partial moles. Another feature that separates
complete from partial mole is the generalized
edema of the villi in the complete mole, which
contrasts with the mixture of two populations
of villi, with fibrosis of some of the villi in
partial mole. Partial moles often show evidence
of development of a fetus or embryo that gen-erally is not seen in complete mole.
Some cases have obvious features of a mole,
and the distinction between a complete mole
and a partial mole is not clear cut. In these cases
the morphologic features that allow separation
of complete and partial mole are ambiguous,
even after extensive histologic sampling. Diag-nosis of a mole early in pregnancy can be par-ticularly difficult, and more moles are now
detected as early as the 7th or the 8th week of
pregnancy.14;15;17;56
In such cases the differential
diagnosis is especially challenging, because the
76 4. Gestational Trophoblastic Disease
Figure 4.9. Partial hydatidiform mole. Portion of an edematous villus shows haphazard foci of trophoblas-tic hyperplasia along the surface and infoldings that form inclusions in the stroma.
cistern formation and trophoblastic hyperplasia
are less pronounced than those found several
weeks later in gestation.
Rarely a complete mole may occur as a twin
gestation in conjunction with a normal pla-centa.40;60;62–65
In these cases the curettage spec-imens contain a mixture of normal sized and
molar villi, which mimics a partial mole. Thus, it
may not always be possible to classify a speci-men accurately as a complete or a partial mole
by morphology alone. Management of either
type of mole requires monitoring of serum
hCG titers after evacuation. It is important to
characterize the type of mole whenever pos-sible, however, as partial mole more frequently
resolves spontaneously and is only very rarely
complicated by choriocarcinoma. Because of
the relatively low rate of persistent GTD asso-ciated with partial mole, further management
of this lesion requires shorter-term  b -hCG
follow-up than does complete mole.
Tissue from moles should be generously
sampled (at least 4 cassettes) to ensure accurate
diagnosis. The histologic features that permit
the distinction of complete and partial mole as
well as a hydropic abortus may not be present
in all sections and therefore adequate sam-pling is essential. Flow cytometric analysis of
paraffin blocks for ploidy is rarely used in diag-nosis because of the cost and the occasional
overlap in karyotypes between the two types of
mole. Nonetheless, many studies suggest that
ploidy is useful in difficult cases,43;44;46;52;57;60;66
especially if the morphologic features are
ambiguous and there is a pressing clinical need
to determine whether the mole is complete or
partial. In addition, the use of immunohisto-chemistry for p57KIP2
may be useful in distin-guishing the two forms of hydatidiform
mole,46–48
although this ancillary technique
requires further validation.
Hydatidiform Mole Versus Hydropic Abortus
Another frequent consideration in the differen-tial diagnosis of hydatidiform mole is the non-molar hydropic abortion with villous edema
(see Chapter 3).7;52;67;68
Microscopically, the
edema of the hydropic abortus can be striking.
Gross specimens from hydropic abortions gen-erally are smaller, however, and villous enlarge-ment is not seen by either the clinician or the
pathologist on gross examination (Table 4.3). It
is important to keep microscopic observations
in context with the gross findings. Most hydropic
abortions yield only one or a few cassettes of
tissue, whereas moles tend to be voluminous.
These generalizations usually hold, but in some
cases of hydatidiform mole the villi are also not
grossly visible. This is especially true if part of
the molar tissue was spontaneously aborted
prior to curettage, if the mole is evacuated very
early in gestation, if there is collapse of villi sec-ondary to suction curettage, or if the specimen
is a partial mole with limited tissue.
Several microscopic features distinguish
hydropic abortus from mole.1;2
In hydropic
abortion the villi are edematous and avascular.
Some also may show trophoblastic inclusions.69
Occasional small cisterns occur in hydropic
abortuses, but they are infrequent and do not
cause gross villous enlargement.The most useful
feature for separating a mole from a hydropic
abortus is the distribution of the villous
trophoblast. In a mole at least occasional villi
show circumferential hyperplasia of trophoblast
along their surface, whereas in the hydropic
abortus the proliferating trophoblast has a polar
distribution, projecting only from one surface
of the anchoring villi. Because trophoblastic
hyperplasia may be limited and focal, especially
in a partial hydatidiform mole, thorough sam-pling may be needed to establish the diagnosis.
In questionable cases it often is best to process
multiple blocks to assess the overall edema
and trophoblastic growth pattern. Nonmolar
hydropic abortus specimens may be diploid,
triploid, or aneuploid,40;69
so DNA ploidy analy-sis is useful only for the separation of mole from
hydropic abortus when it is combined with
careful pathologic evaluation to identify the key
morphologic features of a molar gestation.60;70
Villous abnormalities associated with
trisomy include scalloped villous outlines and
pseudoinclusion into the stroma. These features
can closely simulate partial mole. The tro-phoblastic cells usually show a polar orienta-tion in these cases, which can help in the
diagnosis. In some cases, however, the features
are not sufficiently clear to allow a definitive
interpretation, and a descriptive diagnosis is
appropriate.
Hydatidiform Mole 77
Other placental abnormalities such as
Beckwith Wiedemann syndrome or placental
angiomatous malformation that can mimic
partial mole are typically seen in more
advanced gestations, in the second half of preg-nancy, and are not a significant issue in the
differential diagnosis of early abortion
specimens.60;71;72
These latter abnormalities are
also referred to as “placental mesenchymal dys-plasia” and may show marked hydrops of pla-cental stem villi that can mimic a partial mole
in later pregnancy. In contrast to a partial mole,
however, no abnormal trophoblast prolifera-tion or trophoblastic “inclusions” are seen in
the enlarged villi.71
Hydatidiform Mole Versus Choriocarcinoma
The marked trophoblastic hyperplasia and
cytologic atypia found in some cases of com-plete moles closely resemble the patterns found
in choriocarcinoma. These cases can show large
sheets of trophoblast with an alternating
arrangement of cytotrophoblast and syncy-tiotrophoblast mixed with hemorrhage. Tro-phoblast may be prominent in the original
curettage samples (Fig. 4.10) or in subsequent
curettings done for abnormal elevation of hCG
titers (Fig. 4.11). As long as edematous chori-onic villi are present, no matter how much tro-phoblastic proliferation is present, the lesion is
a hydatidiform mole. In addition to the absence
of chorionic villi, the diagnosis of choriocarci-noma requires the presence of necrosis and
destructive infiltrative growth of trophoblast
into the myometrium.
Rarely, choriocarcinoma may arise in a
developing placenta, and this usually is seen in
a placenta from a second- or third-trimester
gestation.73
We have seen choriocarcinoma
arising in association with retained intrauterine
villi from either an abortion or a term preg-78 4. Gestational Trophoblastic Disease
Figure 4.10. Hydatidiform mole. Prominent hyperplasia of the trophoblast from the surface of a villus of a
complete mole.
nancy. In these very unusual cases, curettage
specimens show a few villi that tend to be
hyalinized or fibrosed mixed with fragments of
tissue showing choriocarcinoma. Presumably in
these cases the neoplasm arises from the
retained trophoblastic cells.
Persistent Postmolar Gestational
Trophoblastic Disease and Invasive
Hydatidiform Mole
After a hydatidiform mole has been evacuated,
a subsequent curettage may be done for persis-tence or elevation of follow-up hCG titers or
for significant uterine bleeding.74–76
The repeat
curettage may show persistent hydatidiform
mole, choriocarcinoma, retained implantation
site, no evidence of trophoblastic tissue, or,
rarely, a new pregnancy.74
If the specimen con-tains persistent hydatidiform mole, it will show
residual molar villi mixed with trophoblast (Fig.
4.11). Usually the amount of tissue and villi are
greatly reduced compared to the original curet-tage specimen, but as long as villi are present,
the diagnosis remains that of persistent intra-cavitary mole. Choriocarcinoma is diagnosed
when there is abundant trophoblast without
villi that shows a dimorphic arrangement of
syncytiotrophoblast and mononucleate tro-phoblast. In addition, it is necessary to see
tumor necrosis or destructive infiltrative
growth by trophoblast into the myometrium.
Scant trophoblastic tissue without villi is not
choriocarcinoma but persistent trophoblast
(Fig. 4.12).77
With invasive hydatidiform mole, hydropic
molar villi and hyperplastic trophoblast either
invade myometrium or are present at other
sites, usually the vulva, vagina, or lungs.1;2;78
To establish the diagnosis, it is necessary to
clearly identify molar villi beyond the
endometrium. In curettings this requires find-Hydatidiform Mole 79
Figure 4.11. Persistent complete mole. Persistent
mole obtained by curettage several weeks after initial
evacuation of a complete hydatidiform mole.The tro-phoblastic proliferation is striking, but the presence
of a villus to the left of center indicates that the diag-nosis is persistent mole, not choriocarcinoma.
ing the villi within myometrial smooth muscle,
an extremely rare event. Consequently, invasive
mole is almost never diagnosed by endometrial
biopsy or curettage. It is important to remem-ber that the presence of residual mole in a
recurettage specimen does not represent inva-sive mole in the absence of demonstrable
myometrial invasion.
Clinical Queries and Reporting of
Hydatidiform Mole
One of the most important clinical questions
in the evaluation of an abortion specimen is
whether gestational trophoblastic disease is
present, as hydatidiform mole, choriocarci-noma, or PSTT can present as a spontaneous or
missed abortion. Even a therapeutic abortion
for an apparently normal gestation may reveal
an unsuspected hydatidiform mole.
Often the clinician suspects a hydatidiform
mole from the clinical history and from findings
such as rapid uterine enlargement, abnormally
high hCG titers, or toxemia in the first
trimester. Ultrasound findings may support the
clinical impression of a mole. At other times
hydatidiform mole is clinically suspected when
visibly enlarged, edematous villi are encoun-tered at curettage. In such cases the logical
questions are whether a mole is present, and, if
so, whether it is a complete or partial mole. In
the distinction between a complete versus a
partial mole, the pathologist should err on the
side of a complete mole, as the partial moles
have few significant clinical consequences. With
either type of mole, the grading of the tro-phoblast does not have clinical significance in
terms of the overall risk for persistent GTD.53
If there is any doubt about whether a specimen
should be classified as an abortus with hydropic
villi or as a hydatidiform mole, more tissue
should be submitted if available. If a case
remains equivocal, the gynecologist should be
cautioned to follow the patient to be certain
that hCG titers return to normal before preg-nancy is attempted again.
80 4. Gestational Trophoblastic Disease
Figure 4.12. Persistent trophoblast following hyda-tidiform mole. A cluster of trophoblastic cells found
in curettage specimen after evacuation of a mole.
Although no villi are present, the amount of tro-phoblast is scant and shows no evidence of invasion.
The quantity of tissue is not sufficient for a conclu-sive diagnosis, and this tissue should be diagnosed as
“persistent trophoblast.”
Patients with known hydatidiform mole who
are being followed may undergo repeat curet-tage for continued bleeding or for abnormally
persistent or elevated hCG titers. If the biopsy
shows trophoblastic tissue, then the presence or
absence of chorionic villi is important and
should be clearly reported. Molar villi generally
indicate persistent intrauterine mole rather
than invasive mole or development of chorio-carcinoma.77
Immature but normal villi indicate
a new pregnancy unrelated to GTD.
Trophoblastic Neoplasms
Choriocarcinoma
General Features
Gestational choriocarcinoma can occur in the
uterine cavity following any type of preg-nancy.1;2;4
As a rule, the risk of choriocarcinoma
increases with the abnormality of the ante-cedent gestation. Complete hydatidiform
mole is a major predisposing factor, and about
half the cases of choriocarcinoma follow a com-plete mole. Choriocarcinoma also can arise
from the trophoblast of an abortion or a term
pregnancy. Consequently, this lesion may be
present whenever abnormal vaginal bleeding
occurs during the postpartum period in a young
woman who has had a pregnancy of any type.
Patients with choriocarcinoma also can present
with metastatic disease without uterine signs or
symptoms. Typically, the patient with chorio-carcinoma has markedly elevated serum hCG
titers.
Pathologic Features
Choriocarcinoma is hemorrhagic and necrotic,
composed of trophoblastic cells without villi
that invade normal tissues (Fig. 4.13). The two
Trophoblastic Neoplasms 81
Figure 4.13. Choriocarcinoma. A mass of tro-phoblast with prominent syncytiotrophoblast in a
curettage specimen of a young woman with abnor-mal uterine bleeding. Abundant trophoblastic tissue
is present with no associated villi. The trophoblast
shows invasion of the myometrium.
Figure 4.14. Choriocarcinoma. Dimorphic population of ST and CT cells in choriocarcinoma. Large ST cells
with multiple nuclei and abundant, vacuolated cytoplasm are interspersed among CT and IT.
main diagnostic features are an absence of
chorionic villi and a dimorphic population of
trophoblast cells (Figs. 4.13 to 4.15).The first cri-terion, absence of villi, is important, as the pro-liferative trophoblast of hydatidiform moles or
of early normal pregnancy can closely simulate
the trophoblast of choriocarcinoma. The second
criterion of choriocarcinoma, a dimorphic
pattern of syncytiotrophoblastic (ST) cells alter-nating with nests or sheets of mononucleate
trophoblast (cytotrophoblastic [CT] or inter-mediate trophoblastic [IT]) cells should be
found, at least focally, to establish a histologic
diagnosis of choriocarcinoma. Often, the char-acteristic pattern of choriocarcinoma is readily
apparent, but at times the dimorphic population
of trophoblast may be difficult to recognize (Fig.
4.16). The admixture of ST with CT or IT cells
yields a plexiform pattern. In these cases identi-fication of ST cells is an important diagnostic
feature. These cells contain multiple nuclei,
ranging from 3 to more than 20 per cell, which
are variable in size. Often the nuclei are
pyknotic but they can be vesicular with promi-nent nucleoli. ST cells have dense eosinophilic
to amphophilic cytoplasm with small vacuoles or
large lacunae that often contain erythrocytes
(Fig. 4.14 and 4.15). In contrast to ST, CT cells
are small (about the size of a decidualized
stromal cell) and uniform. They have a single
nucleus with a prominent nucleolus, pale to clear
cytoplasm, and distinct cell borders. Large IT
cells with polygonal shapes and one or two large,
hyperchromatic nuclei also occur in choriocar-cinoma (Figs. 4.14 and 4.15). The percentage of
IT cells in choriocarcinoma is highly variable,
ranging from rare cells to the large majority
of mononucleate trophoblastic cells. Typically,
82 4. Gestational Trophoblastic Disease
Figure 4.15. Choriocarcinoma. Endometrial curet-tings following a complete mole show ST cells con-taining multiple nuclei and cytoplasmic vacuoles and
numerous IT interspersed. Although intermediate
Figure 4.16. Choriocarcinoma. In this field ST cells are indistinct and CT cells predominate, yielding a
pattern resembling poorly differentiated nontrophoblastic carcinoma.
trophoblast is prominent, the juxtaposition of syncy-tiotrophoblast forming a dimorphic population
establishes the diagnosis of choriocarcinoma.
there is generalized enlargement of the tro-phoblastic cells with increased nuclear atypia in
choriocarcinoma compared to normal tro-phoblastic cell populations in early pregnancy.
The amount of trophoblast in cases of chori-ocarcinoma is highly variable. There may be
abundant neoplastic tissue, but frequently
only a small amount of viable tumor associated
with extensive hemorrhage is present. Small
amounts of tumor may pose problems in diag-nosis. While ST cells generally are prominent in
uterine gestational choriocarcinoma, in occa-sional cases these cells are indistinct and CT
and IT cells predominate. Immunohistochemi-cal stains for hCG can be very helpful for
demonstrating ST cells in such cases. The ST
cells stain intensely for hCG, whereas CT and
IT cells are generally nonreactive. The staining
pattern will clearly demonstrate the plexiform
pattern of ST cells. The ST cells of choriocarci-noma also show strong immunoreactivity for
inhibin-a (Table 4.4).
Differential Diagnosis
The differential diagnosis of choriocarcinoma
includes physiologic trophoblastic prolifera-tions associated with normal pregnancies and
the trophoblast associated with hydatidiform
moles, as well as PSTT and nontrophoblastic
tumors. Trophoblast of normal pregnancy can
appear highly proliferative, especially at the
implantation site of anchoring villi. In contrast
to choriocarcinoma, the trophoblast in normal
pregnancy usually is associated with small,
immature chorionic villi and decidua. Occa-sionally, only trophoblast without villi is present
in an abortion specimen. In such cases the tro-phoblast is small in quantity and lacks necrosis,
significant hemorrhage, and atypical nuclear
features. The finding of a significant amount of
secretory endometrium, decidua, or placental
implantation site favors the presence of normal
trophoblast and not choriocarcinoma. The pres-ence of atypia, including nuclear pleomor-phism, macronucleoli, and abnormal mitotic
figures, strongly suggests choriocarcinoma.
The marked trophoblastic proliferation that
may accompany some hydatidiform moles also
resembles choriocarcinoma (Figs. 4.10 and
4.11). This feature can be particularly problem-atic in persistent hydatidiform mole that is
found in a repeat curettage, since villi often are
sparse in these specimens. Nonetheless, as long
as any villi are present the diagnosis is that of
hydatidiform mole. Even without villi, a diag-nosis of choriocarcinoma should be made only
when atypical, dimorphic trophoblast without
villi are present along with hemorrhage and
tumor necrosis or unequivocal invasion of the
myometrium by a dimorphic population of
trophoblastic cells. In the absence of these
features, the diagnosis should be persistent
trophoblast.
Nontrophoblastic tumors may mimic chorio-carcinoma when they show a large component
84 4. Gestational Trophoblastic Disease
Table 4.4. Immunohistochemistry of uterine trophoblastic tumors compared with nontrophoblastic tumors.
Choriocarcinoma PSTT Epithelioid trophoblastic tumor Nontrophoblastic uterine tumors
hCG +++ +/-+/-+/-Inhibin-a++ ++ ++ -hPL + +++ + /–PLAP +/– ++ +/-Mel-CAM ++ +++ +/–Keratina
+++ +++ +++ +++ b
EMA +++ ++ ++ b
p63 + – +++ +
PSTT, Placental site trophoblastic tumor; 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.
of giant cells. Anaplastic carcinomas and sarco-mas with tumor giant cells may simulate chori-ocarcinoma, at least focally, and may rarely
show choriocarcinomatous differentiation
although this occurs in older women. Usually
the clinical history helps resolve this question,
as high-grade carcinomas and sarcomas gener-ally occur in older postmenopausal patients,
whereas trophoblastic tumors occur in women
of reproductive age. Furthermore, many
patients with choriocarcinoma have a history of
a prior hydatidiform mole. In equivocal cases,
immunohistochemical stains for  b -hCG are
useful for demonstrating syncytiotrophoblast.
Inhibin-a also is very useful for demonstrating
trophoblast tissue (Table 4.4). Because both
choriocarcinoma and anaplastic carcinoma are
epithelial tumors, immunostains for cytokeratin
are less useful in establishing the diagnosis.
Rarely choriocarcinoma may be found in a
postmenopausal patient. In such cases the
tumor may represent gestational choriocarci-noma with a long latent period or it may
represent somatic carcinoma with choriocarci-nomatous transformation.79;80
PSTT and epithe-lioid trophoblastic tumor are discussed and
compared with choriocarcinoma in the follow-ing section.
Placental Site Trophoblastic Tumor
General Features
The placental site trophoblastic tumor (PSTT)
is a rare form of trophoblastic neoplasia
composed predominantly of implantation site
IT.1;2;58;81–84
Like other forms of GTD, it almost
always occurs during the reproductive years.
PSTT is typically diagnosed at the time a curet-tage is performed in a woman who is thought
to be pregnant, usually with a preoperative
diagnosis of a missed abortion; but in contrast
to choriocarcinoma, this tumor rarely is directly
associated with a recent pregnancy.85–89
Molec-ular genetic analysis has established the tro-phoblastic origin of these tumors.90
The hCG
titer is generally low and may not be noticeably
elevated if a sensitive assay method is not
used. Because PSTTs extensively infiltrate the
myometrium, the uterus can be perforated
during curettage. These neoplasms usually are
benign, despite destructive growth in the
myometrium. About 15% of reported tumors
have shown aggressive malignant behavior with
disseminated metastases.
Pathologic Features
PSTT typically produces a mass lesion. These
tumors range from focal lesions 1 to 2 cm in
diameter to large masses that replace much of
the corpus. As a consequence, curettings of
PSTT typically yield multiple fragments of neo-plastic tissue. Microscopically PSTT is com-posed predominantly of implantation site IT
cells that invade normal tissues (Fig. 4.17).91
These cells generally are polyhedral and grow
in cohesive masses that often show areas of
necrosis (Figs. 4.18 and 4.19). The curettage
specimens usually include fragments of
myometrium infiltrated by IT cells. The IT cell
cytoplasm is generally amphophilic with occa-sional clear vacuoles and distinct cell borders.
Some parts of the tumor, especially in areas of
myometrial invasion, are composed of spindle-shaped cells. Most cells have a single irregular
and hyperchromatic nucleus, but binucleate
and multinucleate IT cells are also present.
Marked variation in nuclear size and shape is
often a feature of the tumor (Figs. 4.18 and
4.20). Some nuclei have deep folds or grooves,
and others may have pseudoinclusions as a
result of large cytoplasmic invaginations.
Scattered ST cells with several nuclei and
vacuolated cytoplasm are also present. Mitotic
activity is usually low but can be brisk includ-ing occasional abnormal forms.
Besides the characteristic cytologic features
of the individual cells, the growth pattern of
intermediate trophoblast in PSTT is an impor-tant diagnostic feature. When myometrium is
present in the biopsy, the confluent masses of
cells infiltrate and dissect between smooth
muscle fibers (Fig. 4.21). Furthermore, in PSTT
there is a characteristic pattern of vascular inva-sion in which the intermediate trophoblast
surrounds and replaces the vessel wall
while retaining its overall structural integrity
Trophoblastic Neoplasms 85
Figure 4.18. Placental site trophoblastic tumor. The
cells are pleomorphic and generally contain a single
hyperchromatic nucleus. There is moderate variation
in nuclear size. In the right lower corner the IT infil-Figure 4.17. Placental site trophoblastic tumor. Low magnification shows PSTT invading endometrium and
myometrium. Residual endometrium is present on the left side of the micrograph.
trate and replace the wall of a blood vessel while pre-serving the lumen. Fibrinoid material is present on
the left side.
Figure 4.19. Placental site trophoblastic tumor. Left:
Fragments of a PSTT in a curettage specimen show
sheets of intermediate trophoblast with associated
necrosis. The presence of necrosis is a helpful feature
for distinguishing this tumor from an exaggerated
placental site. Right: Higher magnification of the
tumor shows characteristic IT cells.
Figure 4.20. Placental site trophoblastic tumor. The
tumor is composed of a monomorphic population of
IT cells with hyperchromatic, irregular nuclei and a
moderate amount of eosinophilic cytoplasm. The
nuclei vary in size.
Figure 4.21. Placental site trophoblastic tumor. IT cells in PSTT infiltrate a fragment of myometrium. The
neoplastic cells dissect between the smooth muscle fibers. A mitotic figure is present (arrow).
(Figs. 4.22 and 4.23). One other constant find-ing in this tumor is patchy deposition of
eosinophilic fibrinoid material (Fig. 4.18). The
hyaline, amorphous deposits of fibrinoid occur
randomly throughout the tumor, often entrap-ping individual cells. Fibrinoid also accumulates
in the walls of vessels invaded by IT (Fig. 4.22).
PSTT is diffusely immunoreactive for human
placental lactogen (hPL) and Mel-CAM (CD
146), as well as inhibin-a, epithelial membrane
antigen, and cytokeratins (AE1/AE3 and
cytokeratin 18).2;92
Immunohistochemical stains
for these antigens can be helpful in the differ-ential diagnosis (Table 4.4). In contrast, the
other trophoblastic markers, hCG and PLAP,
show limited immunostaining, with only focal
reactivity.
It is difficult to reliably predict the behavior
of PSTT based on the microscopic features, and
therefore this neoplasm is not divided into
benign and malignant categories. The reported
malignant cases of PSTT generally show some
features that predict aggressive behavior. These
clinically malignant tumors are composed of
larger sheets and masses of cells with more
extensive necrosis than benign tumors.1;2;58;83;84
In malignant PSTT the cells also tend to have
clear instead of amphophilic cytoplasm. Finally,
the mitotic rate usually is higher in the malig-nant tumors, with more than 5 mitoses per 10
high-power fields (HPFs) in most malignant
cases.2;93
In contrast, the benign tumors usually
show a mitotic rate of about 2 mitoses per 10
HPFs, with the highest reported rate being 5
mitoses per 10 HPFs. In several clinically malig-nant PSTTs, the mitotic rate was only 2 per 10
HPFs,94;95
so it appears that some overlap exists
in the mitotic rates of malignant and benign
88 4. Gestational Trophoblastic Disease
Figure 4.22. Placental site trophoblastic tumor. In
this field the IT cells invade blood vessel walls in a
characteristic fashion, preserving the lumen. In addi-Figure 4.23. Placental site trophoblastic tumor. In this field the IT cells are relatively uniform with dense
cytoplasm. IT cells embedded in fibrinoid material replace the wall of a large blood vessel.
tion, amorphous fibrinoid material is deposited in the
walls and in the interstitium. Several of the cells have
vacuolated cytoplasm.
PSTT. Abnormal mitotic figures occur in benign
as well as malignant PSTT. Preliminary findings
suggest that Ki-67 labeling index may be a sig-nificant prognostic indicator as it is usually
greater than 50% in malignant tumors but only
about 14% in benign PSTT.2
Differential Diagnosis
The differential diagnosis of PSTT includes
exaggerated implantation site, choriocarci-noma, placental site nodule, and other, nontro-phoblastic tumors. An exaggerated placental
implantation site is one of the most important
considerations in the differential diagnosis, as it
can have features that are very similar to those
of PSTT (see Chapter 3). The distinction is
largely one of degree (Table 4.5). The exagger-ated placental site usually is a focal finding,
maintaining the overall architecture, and in
other portions of the tissue there is decidua
and/or chorionic villi. PSTT, in contrast, is com-posed of sheets and masses of cells typically
accompanied by necrosis with little normal
tissue in the sections. In addition, the exagger-ated placental site tends to have more ST giant
cells, the nuclei tend to have smudged (degen-erative) chromatin, and mitotic activity is
absent. Any evidence of unequivocal mitotic
figures is suspicious for PSTT. The Ki-67 label-ing index is very useful in the differential diag-nosis because the index is near zero in the
normal and exaggerated implantation site but
14% ± 6.9% in PSTT.96
It may be slightly ele-vated (<5%) in exaggerated implantation sites
associated with a complete mole.96
Separation of PSTT from choriocarcinoma is
important, as these two tumors behave differ-ently and are treated differently. Choriocarci-noma may have a monomorphic appearance in
some areas and can have large numbers of IT.
In contrast to PSTT, however, a network of syn-cytiotrophoblast in choriocarcinoma results in
a dimorphic population, at least focally (Table
4.5). The syncytiotrophoblast in the PSTT is
composed of isolated giant cells that do not
show the dimorphic pattern found in chorio-carcinoma. Immunohistochemical stains for  b-hCG can be especially helpful in highlighting
the network of syncytiotrophoblast in chorio-carcinoma. Although both PSTT and chorio-carcinoma show immunostaining for hPL and
hCG, the ratio of the number of immunoreac-tive cells for each marker differs in the two
tumors. In PSTT the hPL/hCG ratio is typically
3 : 1, whereas in choriocarcinoma it is 1 : 3. Occa-sionally, however, PSTT, especially one that
behaves in a malignant fashion, shows a ratio of
hPL and hCG staining that more closely resem-bles that of choriocarcinoma. The Ki-67 prolif-erative index can also assist in the differential
diagnosis, as it is very high (>50%) in chorio-carcinoma and significantly lower in PSTT
(15% to 20%).96
Rarely, a trophoblastic tumor may show fea-tures of both choriocarcinoma and PSTT. This
90 4. Gestational Trophoblastic Disease
Table 4.5. Comparison of microscopic features. Choriocarcinoma, PSTT, and exaggerated placental site.
Choriocarcinoma PSTT Exaggerated placental site
Amount of lesional tissue Variable, often abundant Variable, often abundant Usually limited
Villi Absent Absenta
Usually present, focal
Trophoblast growth pattern Dimorphic: ST with CT Monomorphic: IT cells Monomorphic: IT cells
and cell types or IT cells
Mitoses Present, usually high rate Present, usually low rate Absent or rare
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, syncytiotrophoblastic; CT, cytotrophoblastic; IT, intermediate trophoblastic,
hCG, human chorionic gonadotropin; hPL, human placental lactogen.
a
Rarely present and, if so, very focal.
is called a mixed PSTT–choriocarcinoma. There
is insufficient experience with these tumors to
predict their behavior accurately.
Placental site nodules also are focal abnor-malities, usually associated with proliferative or
secretory endometrium elsewhere in the sec-tions (see Chapter 3).91
In comparison to PSTT,
these lesions tend to be circumscribed and
small. The IT cells in placental site nodules are
bland and are widely spaced in a hyalinized
stroma. This distribution of IT contrasts with
the sheetlike growth of IT in the PSTT. In con-trast to PSTT, the placental site nodule is only
focally reactive for hPL and Mel-CAM. Also,
the placental site nodule is reactive for PLAP
whereas the PSTT is not.
Now that the histologic features of interme-diate trophoblast are better recognized, the
problem of distinguishing PSTT from other
forms of malignancy has decreased. PSTT is a
tumor of the reproductive years, whereas many
of the malignant tumors that enter into the dif-ferential diagnoses tend to occur at a more
advanced age. Cytologic features of intermedi-ate trophoblast combined with the typical pat-terns of vascular invasion and fibrinoid
deposition usually allow differentiation of PSTT
from other neoplasms. In biopsies, PSTT may be
confused with keratinizing squamous cell carci-noma when the keratin has an amorphous
eosinophilic appearance that superficially
resembles the fibrinoid of PSTT. Squamous cell
carcinoma usually arises in the cervix, however,
whereas the PSTT occurs in the corpus; clinical
features can help to distinguish the neoplasms
(see Chapter 10). Furthermore, with squamous
cell carcinoma a transition to a more obvious
squamous pattern or normal endocervix often is
found. Conversely, in the case of PSTT the curet-tage samples often contain fragments of
endometrium or myometrium that have been
invaded by the tumor.
On occasion PSTT can mimic leiomyosar-coma, especially in areas where intermediate
trophoblast invades myometrium and becomes
intimately admixed with smooth muscle cells.
The trophoblastic cells of the PSTT are strongly
immunoreactive for cytokeratin, inhibin-a ,
hPL, and Mel-CAM whereas leiomyosarcomas,
except for occasional staining with keratin, are
negative for these other markers. (Table 4.4).
The epithelioid appearance of intermediate
trophoblast can also resemble high-grade non-trophoblastic carcinomas. In questionable
cases, immunohistochemical stains for inhibin-a, hPL, Mel-CAM, and hCG are very helpful in
distinguishing PSTT from a nontrophoblastic
neoplasm.
Epithelioid Trophoblastic Tumor
General Features
Epithelioid trophoblastic tumor is a rare form
of trophoblastic tumor that has only recently
become recognized.2;90;97–100
This trophoblastic
neoplasm is distinct from choriocarcinoma and
PSTT with features resembling those of
somatic carcinomas. This lesion was initially
observed in a few patients with persistent lung
metastases following intensive chemotherapy
for documented choriocarcinoma.101
Similar
lesions were reported as multiple nodules of
intermediate trophoblast in the uteri of patients
following evacuation of hydatidiform moles.102
These tumors have also been seen in patients
without a history of antecedent GTD. We have
observed similar tumors that merged imper-ceptibly with typical choriocarcinoma and
PSTT. This lesion also has been found in the
uterus adjacent to placental site nodules fol-lowing hydatidiform mole.
The epithelioid trophoblastic tumor is pre-ceded by a term gestation in two thirds of cases,
with spontaneous abortions and hydatidiform
moles being the antecedent gestation in the
remaining cases. Usually there is a long interval
following the gestation and the diagnosis of this
tumor with a range of 1 to 18 years (average,
6.2 years). A case has been reported in a
postmenopausal woman.103
Serum hCG levels
are usually elevated at the time of diag-nosis, although the levels are generally low
(< 2500 mIU/ml).
Pathologic Features
Epithelioid trophoblastic tumor lacks the
dimorphic pattern of classical choriocarcinoma
and is composed of chorionic-type IT (Figs. 4.24
and 4.25).2;91;97
ST cells are indistinct. The tumor
Trophoblastic Neoplasms 91
Figure 4.24. Epithelioid trophoblastic tumor. Cords and nests of trophoblastic cells with indistinct ST are
separated by hyaline stroma. The epithelioid cells have pale to vacuolated cytoplasm.
Figure 4.25. Epithelioid trophoblastic tumor. The
tumor lacks a dimorphic pattern of CT and ST but
is composed of a population of relatively uniform,
polygonal, mononucleate trophoblastic cells with
prominent cellular membranes. A transition to chori-ocarcinoma with a typical biphasic pattern was iden-tified in other areas of the tumor.
displays a nodular growth pattern and has a
striking epithelioid appearance, both in its cyto-logic features and in its pattern of invasion. The
neoplasm is composed of small nests and cords
of cells. The nests often contain dense central
hyaline material and necrotic debris, and the
cords are encompassed by a hyaline matrix
(Fig. 4.26). The predominant cells are relatively
uniform in size and are mononucleate with
round, uniform nuclei and eosinophilic or clear
cytoplasm. They are larger than CT cells but
smaller than implantation site IT cells. Rarely,
larger cells resembling implantation site IT
cells are admixed or scattered within the extra-cellular hyaline material. Apoptotic cells and
islands of necrotic debris are abundant in most
tumors. The mitotic index varies from 0 to 9 per
10 HPF with an average of 2 per 10 HPF. Focal
areas resembling placental site nodule, placen-tal site trophoblastic tumor, or choriocarcinoma
may be seen within these tumors.
These tumors have immunohistochemical
profiles similar to that seen in normal chori-onic-type IT. They are diffusely reactive for
cytokeratin (AE1/AE3 and cytokeratin 18) as
well as epithelial membrane antigen. In at least
a portion of the cells the tumors are positive
for inhibin-a .92
Other trophoblastic markers
including hCG, hPL, and Mel-CAM (CD 146)
are only focally expressed (Table 4.4).2;97
This
immunophenotype contrasts with the placental
site trophoblastic tumor, which is diffusely pos-itive for hPL and Mel-CAM. The epithelioid
trophoblastic tumor has a Ki-67 proliferative
index of 10% to 25% with a mean of about
20%.2
Differential Diagnosis
The differential diagnosis includes placental
site nodule, placental site trophoblastic tumor,
choriocarcinoma, and keratinizing squamous
cell carcinoma. The distinction from a placental
site nodule is usually not difficult because the
placental site nodule is typically a microscopic,
well-circumscribed nodule with low cellularity,
while epithelioid trophoblastic tumors are
larger, cellular neoplasms that show necrosis.
Trophoblastic Neoplasms 93
Figure 4.26. Epithelioid trophoblastic tumor. In this field the tumor is composed of cords and nests of
monotonous cells in a hyaline matrix.
The placental site nodule has little to no mitotic
activity and a Ki-67 index of less than 10%
while the epithelioid trophoblastic tumor is
more cellular with a Ki-67 index of greater than
10%.2
The nodular growth pattern with the distinc-tive hyaline matrix contrasts with the diffuse,
infiltrative pattern of PSTT. In addition, the
cells in epithelioid trophoblastic tumor are
smaller than the IT cells of the PSTT and tend
to grow in nests and cords. Immunohistochem-istry can be very useful in distinguishing epithe-lioid trophoblastic tumor (ETT) from PSTT as
the antibody against p63 reacts with chorionic
type trophoblast that comprises ETT but not
with implantation type trophoblast of PSTT.103a
In contrast to choriocarcinoma, the epithelioid
trophoblastic tumor does not have a dimorphic
pattern with interspersed ST cells. Further-more, the epithelioid trophoblastic tumor is not
as hemorrhagic as choriocarcinoma. Immuno-histochemical profiles are very helpful, as
epithelioid trophoblastic tumor is positive for
inhibin-a but shows only limited reactivity for
hPL and Mel-CAM, in contrast to the PSTT,
and shows only random  b-hCG reactivity, in
contrast to the abundant staining of ST cells in
choriocarcinoma.91
ETT also can resemble a nontrophoblastic
tumor, especially squamous cell carcinoma of
the cervix, because of their epithelioid appear-ance, resemblance of the hyaline and necrotic
debris to keratin, and propensity for the cells of
ETT to grow along the surface of the cervix. In
addition, about 50% of these tumors present in
the cervix. Immunostains for inhibin-a and
cytokeratin 18 can be very useful, as these are
positive in epithelioid trophoblastic tumor but
are negative in squamous cell carcinoma (Table
4.4).97;104
Furthermore, the Ki-67 proliferative
index is lower in epithelioid trophoblastic
tumor (10% to 25%) compared to squamous
cell carcinomas, which have Ki-67 labeling
indices of greater than 50%.
Clinical Queries and Reporting of
Trophoblastic Neoplasms
In curettings, choriocarcinoma is most com-monly found in a repeat curettage during
follow-up of a hydatidiform mole. Sometimes,
however, this neoplasm is an unsuspected
finding in a reproductive-age patient with
abnormal uterine bleeding. PSTT, because of its
rarity, often is not clinically suspected, and the
patient presents with amenorrhea or an appar-ent missed abortion.
Obvious cases of either choriocarcinoma or
PSTT do not require more than a concise diag-nosis. Because of its rarity, the diagnosis of
PSTT often is most useful if accompanied by an
explanatory comment that describes this as a
form of GTD derived from intermediate tro-phoblast. Furthermore, for PSTT the mitotic
count should be stated, as this feature may help
to predict aggressive behavior. All PSTTs
should be considered potentially malignant,
even those with a low mitotic index. Because of
the rarity and the potential for highly aggres-sive growth of any trophoblastic tumor, but
especially choriocarcinoma, we also recom-mend oral communication with the gynecolo-gist whenever possible.
At times a specific diagnosis of a trophoblas-tic tumor may not be possible, yet the lesion is
suspicious for neoplasia. WHO does, in fact,
have a category of “unclassified” trophoblastic
lesion (Table 4.1), a term reserved for those
unusual cases that cannot clearly be placed in
one of the defined subgroups of the disease. An
example of such a case would be a small
amount of proliferative trophoblast without
villi, which can lead to a difficult differential
diagnosis of choriocarcinoma versus tro-phoblast of a normal pregnancy. Also, promi-nent IT within decidua and myometrium may
lead to a differential diagnosis of exaggerated
placental site versus PSTT. When the diagnosis
is not straightforward, a descriptive diagnosis of
atypical trophoblast is best. With this type of
diagnosis, the clinician is alerted to the possi-bility of trophoblastic disease. Then the patient
can be followed with hCG titers or rebiopsied
if symptoms persist.
References
(1) Silverberg SG, Kurman RJ. Tumors of the
uterine corpus and gestational trophoblastic
disease. Atlas of tumor pathology, 3rd series,
94 4. Gestational Trophoblastic Disease
Fascicle 3. Washington, DC: Armed Forces
Institute of Pathology, 1992.
(2) Shih I-M, Mazur MT, Kurman RJ. Gesta-tional trophoblastic disease and related
lesions. In: Kurman RJ, editor. Blaustein’s
pathology of the female genital tract. 5th
ed.
New York: Springer-Verlag, 2002: 1193–1247.
(3) Scully RE, Poulson H, Sobin LH. Interna-tional Histological Classification and Typing
of Female Genital Tract Tumours. Berlin:
Springer-Verlag, 1994.
(4) Hertig AT, Mansell H. Tumors of the female
sex organs. Part I. Hydatidiform mole and
choriocarcinoma. Atlas of Tumor Pathology,
section 9, Fascicle 33. Washington, DC:
Armed Forces Institute of Pathology, 1956.
(5) Craighill MC, Cramer DW. Epidemiology of
complete molar pregnancy. J Reprod Med
1984; 29:784–787.
(6) Womack C, Elston W. Hydatidiform mole in
Nottingham: A 12-year retrospective epi-demiological and morphological study. Pla-centa 1985; 6:93–106.
(7) Jeffers MD, O’Dwyer P, Curran B, Leader M,
Gillan JE. Partial hydatidiform mole: A
common but underdiagnosed condition. A
3-year retrospective clinicopathological and
DNA flow cytometric analysis. Int J Gynecol
Pathol 1993; 12:315–323.
(8) Bracken MB, Brinton LA, Hayashi K. Epi-demiology of hydatidiform mole and chorio-carcinoma. Epidemiol Rev 1984; 6:52–75.
(9) Grimes DA. Epidemiology of gestational tro-phoblastic disease. Am J Obstet Gynecol
1984; 150:309–318.
(10) Czernobilsky B, Barash A, Lancet M. Partial
moles: A clinicopathologic study of 25 cases.
Obstet Gynecol 1982; 59:75–77.
(11) Szulman AE, Surti U. The clinicopathologic
profile of the partial hydatidiform mole.
Obstet Gynecol 1982; 59:597–602.
(12) Szulman AE, Surti U. The syndromes of
hydatidiform mole. I. Cytogenetic and mor-phologic correlations. Am J Obstet Gynecol
1978; 131:665–671.
(13) Szulman AE, Surti U. The syndromes of
hydatidiform mole. II. Morphologic evolu-tion of the complete and partial mole. Am J
Obstet Gynecol 1978; 132:20–27.
(14) Keep D, Zaragoza MV, Hassold T, Redline
RW. Very early complete hydatidiform mole.
Hum Pathol 1996; 27:708–713.
(15) Sebire NJ, Rees H, Paradinas F, Seckl M,
Newlands E. The diagnostic implications of
routine ultrasound examination in histologi-cally confirmed early molar pregnancies.
Ultrasound Obstet Gynecol 2001; 18:662–
665.
(16) Romero R, Horgan JG, Kohorn EI, Kadar N,
Taylor KJW, Hobbins JC. New criteria for the
diagnosis of gestational trophoblastic
disease. Obstet Gynecol 1985; 66:553–558.
(17) Fukunaga M. Early partial hydatidiform
mole: Prevalence, histopathology, DNA
ploidy, and persistence rate. Virchows Arch
2000; 437:180–184.
(18) Bagshawe KD, Wilson H, Dublon P, Smith A,
Baldwin M, Kardana A. Follow-up after
hydatidiform mole: Studies using radio-immunoassay for urinary human chorionic
gonadotrophin (hCG). J Obstet Gynaecol Br
Commonw 1973; 80:461–468.
(19) Curry SL, Hammond CB, Tyrey L, Creasman
WT, Parker RT. Hydatidiform mole: diagno-sis, management, and long–term follow-up of
347 patients. Obstet Gynecol 1975; 45:1–8.
(20) Lurain JR, Brewer JI, Torok EE, Halpern B.
Natural history of hydatidiform mole after
primary evacuation. Am J Obstet Gynecol
1983; 145:591–595.
(21) Goto S, Yamada A, Ishizuka T, Tomoda Y.
Development of postmolar trophoblastic dis-ease after partial molar pregnancy. Gynecol
Oncol 1993; 48:165–170.
(22) Berkowitz RS, Goldstein DP, Bernstein MR.
Natural history of partial molar pregnancy.
Obstet Gynecol 1983; 66:677–681.
(23) Rice LW, Berkowitz RS, Lage JM, Goldstein
DP, Bernstein MR. Persistent gestational
trophoblastic tumor after partial hydatidi-form mole. Gynecol Oncol 1990; 36:358–362.
(24) Zalel Y, Dgani R. Gestational trophoblastic
disease following the evacuation of partial
hydatidiform mole: A review of 66 cases.
Eur J Obstet Gynecol Reprod Biol 1997;
71(1):67–71.
(25) Seckl MJ, Fisher RA, Salerno G, Rees H,
Paradinas FJ, Foskett M et al. Choriocarci-noma and partial hydatidiform moles. Lancet
2000; 356(9223):36–39.
(26) Menczer J, Girtler O, Zajdel L, Glezerman
M. Metastatic trophoblastic disease following
partial hydatidiform mole: Case report and
literature review. Gynecol Oncol 1999;
74:304–307.
(27) Gardner HAR, Lage JM. Choriocarcinoma
following a partial hydatidiform mole—A
case report. Hum Pathol 1992; 23:468–471.
(28) Matsui H, Iizuka Y, Sekiya S. Incidence of
invasive mole and choriocarcinoma fol-References 95
lowing partial hydatidiform mole. Int J
Gynaecol Obstet 1996; 53:63–64.
(28a) Cheung AN, Khoo US, Lai CY, Chan KYK,
Cheng DKL, Chiu PM, et al. Metastatic tro-phoblastic disease after an initial diagnosis of
partial hydatidiform mole. Genotyping and
in situ hybridization analysis. Cancer 2004;
100:1211–1417.
(29) Kajii T, Ohama K. Androgenetic origin of
hydatidiform mole. Nature (London) 1977;
268:633–634.
(30) Wake N, Takagi N, Sasaki M. Androgenesis as
a cause of hydatidiform mole. J Natl Cancer
Inst 1978; 60:51–57.
(31) Kajii T, Kurashige H, Ohama K, Uchino F.
XY and XX complete moles: Clinical and
morphologic correlations. Am J Obstet
Gynecol 1984; 150:57–64.
(32) Vassilakos P, Riotton G, Kajii T. Hydatidi-form mole: Two entities. A morphologic and
cytogenetic study with some clinical consid-erations. Am J Obstet Gynecol 1977; 127:
167–170.
(33) Surti U, Szulman AE, O’Brien S. Dispermic
origin and clinical outcome of three complete
hydatidiform moles with 46, XY karyotype.
Am J Obstet Gynecol 1982; 144:84–87.
(34) Surti U. Genetic concepts and techniques. In:
Szulman AE, Buchsbaum HJ, eds. gestational
Trophoblastic disease. New York: Springer-Verlag; 1987: 111–121.
(35) Jacobs PA, Szulman AE, Funkhouser J,
Matsuura JS, Wilson CC. Human triploidy:
relationship between parental origin of the
additional haploid complement and develop-ment of partial hydatidiform mole. Ann Hum
Genet 1982; 46:223–231.
(36) Redline RW, Hassold T, Zaragoza MV. Pre-valence of the partial molar phenotype in
triploidy of maternal and paternal origin.
Hum Pathol 1998; 29:505–511.
(37) Hemming JD, Quirke P, Womack C, Wells M,
Elston CW, Bird CC. Diagnosis of molar
pregnancy and persistent trophoblastic
disease by flow cytometry. J Clin Pathol 1987;
40:615–620.
(38) Davis JR, Kerrigan DP, Way DL, Weiner SA.
Partial hydatidiform moles: Deoxyribonu-cleic acid content and course. Am J Obstet
Gynecol 1987; 157:969–973.
(39) Lage JM, Berkowitz RS, Rice LW, Goldstein
DP, Bernstein MR, Weinberg DS. Flow cyto-metric analysis of DNA content in partial
hydatidiform moles with persistent gesta-tional trophoblastic tumor. Obstet Gynecol
1991; 77:111–115.
(40) Lage JM, Mark SD, Roberts DJ, Goldstein
DP, Bernstein MR, Berkowitz RS. A flow
cytometric study of 137 fresh hydropic pla-centas: Correlation between types of hyda-tidiform moles and nuclear DNA ploidy.
Obstet Gynecol 1992; 79:403–410.
(41) Lawler SD, Fisher RA, Dent J. A prospective
genetic study of complete and partial hyda-tidiform moles. Am J Obstet Gynecol 1991;
164:1270–1277.
(42) Lage JM, Weinberg DS, Yavner DL, Bieber
FR. The biology of tetraploid hydatidi-form moles: histopathology, cytogenetics, and
flow cytometry. Hum Pathol 1989; 20:
419–425.
(43) Fukunaga M, Ushigome S, Sugishita M.
Application of flow cytometry in diagnosis of
hydatidiform moles. Mod Pathol 1993; 6:353–
359.
(44) Genest DR, Ruiz RE, Weremowicz S,
Berkowitz RS, Goldstein DP, Dorfman DM.
Do nontriploid partial hydatidiform moles
exist? A histologic and flow cytometric
reevaluation of nontriploid specimens. J
Reprod Med 2002; 47:363–368.
(45) Fisher RA, Hodges MD, Rees HC, Sebire NJ,
Seckl MJ, Newlands ES, et al. The maternally
transcribed gene  p57 (KIP2) (CDNK1C) is
abnormally expressed in both androgenetic
and biparental complete hydatidiform moles.
Hum Mol Genet 2002; 11:3267–3272.
(46) Genest DR, Dorfman DM, Castrillon DH.
Ploidy and imprinting in hydatidiform moles.
Complementary use of flow cytometry and
immunohistochemistry of the imprinted gene
product p57KIP2 to assist molar classifica-tion. J Reprod Med 2002; 47:342–346.
(47) Castrillon DH, Sun D, Weremowicz S, Fisher
RA, Crum CP, Genest DR. Discrimination of
complete hydatidiform mole from its mimics
by immunohistochemistry of the paternally
imprinted gene product p57KIP2. Am J Surg
Pathol 2001; 25:1225–1230.
(48) Fukunaga M. Immunohistochemical charac-terization of p57(KIP2) expression in early
hydatidiform moles. Hum Pathol 2002; 33:
1188–1192.
(49) Martin DA, Sutton GP, Ulbright TM, Sledge
GW, Stehman FB, Ehrlich CE. DNA content
as a prognostic index in gestational tro-phoblastic neoplasia. Gynecol Oncol 1989;
34:383–388.
96 4. Gestational Trophoblastic Disease
(50) Fukunaga M. Flow cytometric and clinico-pathologic study of complete hydatidiform
moles with special reference to the signifi-cance of cytometric aneuploidy. Gynecol
Oncol 2001; 81:67–70.
(51) Qiao S, Nagasaka T, Nakashima N. Numerous
vessels detected by CD34 in the villous
stroma of complete hydatidiform moles. Int J
Gynecol Pathol 1997; 16:233–238.
(52) Conran RM, Hitchcock CL, Popek EJ, Norris
HJ, Griffin JL, Geissel A, et al. Diagnostic
considerations in molar gestations. Hum
Pathol 1993; 24:41–48.
(53) Genest DR, Laborde O, Berkowitz RS, Gold-stein DP, Bernstein MR, Lage J. A clinico-pathologic study of 153 cases of complete
hydatidiform mole (1980–1990)—Histologic
grade lacks prognostic significance. Obstet
Gynecol 1991; 78:402–409.
(54) Mosher R, Goldstein DP, Berkowitz R, Bern-stein M, Genest DR. Complete hydatidiform
mole. Comparison of clinicopathologic fea-tures, current and past. J Reprod Med 1998;
43:21–27.
(55) Paradinas FJ. The diagnosis and prognosis
of molar pregnancy: The experience of
the National Referral Centre in London.
Int J Gynaecol Obstet 1998; 60 Suppl 1:S57–
S64.
(56) Coukos G, Makrigiannakis A, Chung J,
Randall TC, Rubin SC, Benjamin I. Com-plete hydatidiform mole. A disease with a
changing profile. J Reprod Med 1999;
44(8):698–704.
(57) Paradinas FJ, Browne P, Fisher RA, Foskett
M, Bagshawe KD, Newlands E. A clinical,
histopathological and flow cytometric study
of 149 complete moles, 146 partial moles and
107 non-molar hydropic abortions.
Histopathology 1996; 28:101–110.
(58) Kurman RJ. The morphology, biology, and
pathology of intermediate trophoblast—a
look back to the present. Hum Pathol 1991;
22:847–855.
(59) Montes M, Roberts D, Berkowitz RS, Genest
DR. Prevalence and significance of implan-tation site trophoblastic atypia in hydatidi-form moles and spontaneous abortions. Am
J Clin Pathol 1996; 105:411–416.
(60) Genest DR. Partial hydatidiform mole: Clin-icopathological features, differential diagno-sis, ploidy and molecular studies, and gold
standards for diagnosis. Int J Gynecol Pathol
2001; 20:315–322.
(61) Iyengar V, Roberts D, Lage J. Histopatho-logic study of triploid partial moles. Mod
Pathol 1993; 6:74A.
(62) Vejerslev LD, Dueholm M, Hassing N. Hyda-tidiform mole: cytogenetic marker analysis in
twin gestation. Report of two cases. Am J
Obstet Gynecol 1986; 155:614.
(63) Steller MA, Genest DR, Bernstein MR, Lage
JM, Goldstein DP, Berkowitz RS. Clinical
features of multiple conception with partial
or complete molar pregnancy and coexisting
fetuses. J Reprod Med 1994; 39:147–154.
(64) Steller MA, Genest DR, Bernstein MR, Lage
JM, Goldstein DP, Berkowitz RS. Natural
history of twin pregnancy with complete
hydatidiform mole and coexisting fetus.
Obstet Gynecol 1994; 83:35–42.
(65) Choi-Hong SR, Genest DR, Crum CP,
Berkowitz R, Goldstein DP, Schofield DE.
Twin pregnancies with complete hydatidi-form mole and coexisting fetus: use of fluo-rescent in situ hybridization to evaluate
placental X- and Y-chromosomal content.
Hum Pathol 1995; 26:1175–1180.
(66) Lage JM, Driscoll SG, Yavner DL, Olivier
AP, Mark SD, Weinberg DS. Hydatidiform
moles. Application of flow cytometry in diag-nosis. Am J Clin Pathol 1988; 89:596–600.
(67) Messerli ML, Parmley T, Woodruff JD, et al.
Inter- and intra-pathologist variability in the
diagnosis of gestational trophoblastic neo-plasia. Obstet Gynecol 1987; 69:622–626.
(68) Javey H, Borazjani G, Behmard S, Langley
FA. Discrepancies in the histological diagno-sis of hydatidiform mole. Br J Obstet
Gynaecol 1979; 86:480–483.
(69) Koenig C, Demopoulos RI, Vamvakas EC,
Mittal KR, Feiner HD, Espiritu EC. Flow
cytometric DNA ploidy and quantitative
histopathology in partial moles. Int J
Gynecol Pathol 1993; 12:235–240.
(70) Chew SH, Perlman EJ, Williams R, Kurman
RJ, Ronnett BM. Morphology and DNA
content analysis in the evaluation of first
trimester placentas for partial hydatidiform
mole (PHM). Hum Pathol 2000; 31:914–924.
(71) Paradinas FJ, Sebire NJ, Fisher RA, Rees HC,
Foskett M, Seckl MJ, et al. Pseudo–partial
moles: Placental stem vessel hydrops and the
association with Beckwith–Wiedemann syn-drome and complete moles. Histopathology
2001; 39(5):447–454.
(72) Sander CM. Angiomatous malformation of
placental chorionic stem vessels and pseudo–
References 97
partial molar placentas: Report of five cases.
Pediatr Pathol 1993; 13:621–633.
(73) Zanetta G, Maggi R, Colombo M, Bratina G,
Mangioni C. Choriocarcinoma coexistent
with intrauterine pregnancy: Two additional
cases and a review of the literature. Int J
Gynecol Cancer 1997; 7:66–77.
(74) Schlaerth JB, Morrow CP, Rodriguez M.
Diagnostic and therapeutic curettage in ges-tational trophoblastic disease. Am J Obstet
Gynecol 1990; 162:1465–1471.
(75) Berkowitz RS, Desai U, Goldstein DP,
Driscoll SG, Marean AR, Berstein MR.
Pretreatment curettage—a predictor of
chemotherapy response in gestational tro-phoblastic disease. Gynecol Oncol 1980;
10:39–43.
(76) Berkowitz RS, Birnholz J, Goldstein DP,
Bernstein MR. Pelvic ultrasonography and
the management of gestational trophoblastic
disease. Gynecol Oncol 1983; 15:403–412.
(77) Elston CW, Bagshawe KD. The diagnosis of
trophoblastic tumours from uterine curet-tings. J Clin Pathol 1972; 25:111–118.
(78) Takeuchi S. Nature of invasive mole and its
rational treatment. Semin Oncol 1982; 9:
181–186.
(79) Savage J, Subby W, Okagaki T. Adenocarci-noma of the endometrium with trophoblastic
differentiation and metastases as choriocar-cinoma: A case report. Gynecol Oncol 1987;
26:257–262.
(80) Pesce C, Merino MJ, Chambers JT, Nogales
F. Endometrial carcinoma with trophoblastic
differentiation. An aggressive form of uterine
cancer. Cancer 1991; 68:1799–1802.
(81) Eckstein RP, Paradinas FJ, Bagshawe KD.
Placental site trophoblastic tumour (tropho-blastic pseudotumour): A study of four cases
requiring hysterectomy, including one fatal
case. Histopathology 1982; 6:211–226.
(82) Kurman RJ, Scully RE, Norris HJ. Tro-phoblastic pseudotumor of the uterus. An
exaggerated form of “syncytial endometritis”
simulating a malignant tumor. Cancer 1976;
38:1214–1226.
(83) Young RH, Scully RE. Placental site tro-phoblastic tumor: Current status. Clin Obstet
Gynecol 1984; 27:248–258.
(84) Young RE, Kurman RJ, Scully RE. Prolifer-ations and tumors of intermediate tro-phoblast of the placental site. Semin Diagn
Pathol 1988; 5:223–237.
(85) Bower M, Paradinas FJ, Fisher RA, Nichol-son SK, Rustin GJ, Begent RH, et al. Placen-tal site trophoblastic tumor: Molecular
analysis and clinical experience. Clin Cancer
Res 1996; 2:897–902.
(86) Chang YL, Chang TC, Hsueh S, Huang KG,
Wang PN, Liu HP, et al. Prognostic factors
and treatment for placental site trophoblas-tic tumor–report of 3 cases and analysis of 88
cases. Gynecol Oncol 1999; 73:216–222.
(87) Feltmate CM, Genest DR, Goldstein DP,
Berkowitz RS. Advances in the understand-ing of placental site trophoblastic tumor. J
Reprod Med 2002; 47(5):337–341.
(88) Papadopoulos AJ, Foskett M, Seckl MJ,
McNeish I, Paradinas FJ, Rees H, et al.
Twenty–five years’ clinical experience with
placental site trophoblastic tumors. J Reprod
Med 2002; 47:460–464.
(89) Wright JD, Powell MA, Horowitz NS, Huet-tner PC, White F, Herzog TJ. Placental site
trophoblastic tumor presenting with a pneu-mothorax during pregnancy. Obstet Gynecol
2002; 100(5 Pt 2):1141–1144.
(90) Oldt RJ, III, Kurman RJ, Shih I. Molecular
genetic analysis of placental site trophoblas-tic tumors and epithelioid trophoblastic
tumors confirms their trophoblastic origin.
Am J Pathol 2002; 161:1033–1037.
(91) Shih IM, Kurman RJ. The pathology of
intermediate trophoblastic tumors and
tumor–like lesions. Int J Gynecol Pathol
2001; 20:31–47.
(92) Shih IM, Kurman RJ. Immunohistochemical
localization of inhibin-alpha in the placenta
and gestational trophoblastic lesions. Int J
Gynecol Pathol 1999; 18:144–150.
(93) Feltmate CM, Genest DR, Wise L, Bernstein
MR, Goldstein DP, Berkowitz RS. Placental
site trophoblastic tumor: A 17-year experi-ence at the New England Trophoblastic
Disease Center. Gynecol Oncol 2001;
82:415–419.
(94) Gloor E, Dialdas J, Hurlimann J, Ribolzi J,
Barrelet L. Placental site trophoblastic tumor
(trophoblastic pseudotumor) of the uterus
with metastases and fatal outcome. Am J
Surg Pathol 1983; 7:483–486.
(95) Fukunaga M, Ushigome S. Metastasizing
placental site trophoblastic tumor: An im-munohistochemical and flow cytometric
study of two cases. Am J Surg Pathol 1993;
17:1003–1010.
98 4. Gestational Trophoblastic Disease
(96) Shih IM, Kurman RJ. Ki-67 labeling index in
the differential diagnosis of exaggerated pla-cental site, placental site trophoblastic tumor,
and choriocarcinoma: A double immunohis-tochemical staining technique using Ki-67
and Mel-CAM antibodies. Hum Pathol 1998;
29:27–33.
(97) Shih IM, Kurman RJ. Epithelioid tro-phoblastic tumor: A neoplasm distinct from
choriocarcinoma and placental site tro-phoblastic tumor simulating carcinoma. Am
J Surg Pathol 1998; 22:1393–1403.
(98) Hamazaki S, Nakamoto S, Okino T,
Tsukayama C, Mori M, Taguchi K, et al.
Epithelioid trophoblastic tumor: Morpholog-ical and immunohistochemical study of three
lung lesions. Hum Pathol 1999; 30:1321–1327.
(99) Ohira S, Yamazaki T, Hatano H, Harada O,
Toki T, Konishi I. Epithelioid trophoblastic
tumor metastatic to the vagina: an immuno-histochemical and ultrastructural study. Int J
Gynecol Pathol 2000; 19:381–386.
(100) Shen DH, Khoo US, Ngan HY, Ng TY, Chau
MT, Xue WC, et al. Coexisting epithelioid
trophoblastic tumor and choriocarcinoma of
the uterus following a chemoresistant hyda-tidiform mole. Arch Pathol Lab Med 2003;
127:e291–e293.
(101) Mazur MT. Metastatic gestational choriocar-cinoma. Unusual pathologic variant follow-ing therapy. Cancer 1989; 63:1370–1377.
(102) Silva EG, Tornos C, Lage J, Ordonez NG,
Morris M, Kavanagh J. Multiple nodules of
intermediate trophoblast following hydatidi-form moles. Int J Gynecol Pathol 1993; 12:
324–332.
(103) Coulson LE, Kong CS, Zaloudek C. Epithe-lioid trophoblastic tumor of the uterus in a
postmenopausal woman: A case report and
review of the literature. Am J Surg Pathol
2000; 24:1558–1562.
(103a) Shih IM, Kurman RJ. p63 expression is useful
in the distinction of epithelioid trophoblastic
and placental site trophoblastic tumors by
profiling trophobastic subpopulations. Am J
Surg Pathol (In press).
(104) Shih IM, Seidman JD, Kurman RJ. Placental
site nodule and characterization of distinc-tive types of intermediate trophoblast. Hum
Pathol 1999; 30:687–694.
References 99
100
phologic changes of endometrial glandular and
stromal breakdown. The changes of glandular
and stromal breakdown are not unique to DUB;
they may be found in a variety of organic disor-ders, as well. Conversely, not all biopsy speci-mens of patients with a history of bleeding show
evidence of breakdown. Nonetheless, endome-trial breakdown and bleeding is commonly
encountered, and the morphologic features of
bleeding need to be recognized in order to allow
clear separation of these nonspecific artifacts
and degenerative/regenerative changes from
other, more specific histologic changes associ-ated with hyperplasia or carcinoma.
Because glandular and stromal breakdown
tends to be most extensive when associated
with DUB, the morphologic features of this
peculiar form of early tissue breakdown are
first presented. It is important to recognize that
glandular and stromal breakdown by itself is
a nonspecific finding; changes in the intact,
nonbleeding endometrial glands allow the
assessment of the presence or absence of other
specific organic abnormalities.
Morphologic Features of
Glandular and Stromal
Breakdown
Breakdown and bleeding patterns have been
referred to by a number of different descriptive
terms, such as “lytic,” “shedding,” “slough,” or
even “menstrual” endometrium. The histologic
5
Dysfunctional Uterine Bleeding
This chapter specifically addresses uterine
bleeding resulting from alterations in the
normal cyclical changes of the endometrium.
This type of bleeding is commonly referred to
as dysfunctional uterine bleeding (DUB).1–5
Clinically, DUB indicates ovulatory dysfunc-tion. By definition, DUB excludes post-menopausal bleeding or bleeding caused by
the presence of specific pathologic processes
such as inflammation, polyps, hyperplasia, car-cinoma, exogenous hormones, and complica-tions of pregnancy. It is important to recognize
the endometrial changes associated with DUB,
because they may be confused with more
serious lesions such as hyperplasia.
Although DUB denotes abnormal bleeding
with no underlying organic disorder, one
common finding in these biopsies is the mor-Morphologic Features of Glandular and
Stromal Breakdown  . . . . . . . . . . . . . . . . . 100
Dysfunctional Uterine Bleeding  . . . . . . . . 108
Estrogen-Related Bleeding  . . . . . . . . . . . . 109
Proliferative with Glandular and
Stromal Breakdown . . . . . . . . . . . . . . . . 109
Disordered Proliferative Phase and
Persistent Proliferative Phase  . . . . . . . . 112
Atrophy  . . . . . . . . . . . . . . . . . . . . . . . . . 113
Progesterone-Related Bleeding  . . . . . . . . 115
Luteal Phase Defects  . . . . . . . . . . . . . . . 117
Irregular Shedding . . . . . . . . . . . . . . . . . 117
Abnormal Secretory Endometrium with
Breakdown of Unknown Etiology . . . . . 117
Clinical Queries and Reporting  . . . . . . . . 119
First, the operative procedure itself causes
tissue fragmentation and hemorrhage. Second,
when breakdown does occur, the pattern of
tissue necrosis is unlike that seen in other
organs. This unique morphologic expression of
necrosis is due to the rapid expulsion of tissue
into the uterine cavity. Consequently, extensive
necrosis and autolysis often do not occur,
and the tissue shows only early signs of
degeneration.
Among the distinctive features of acute
breakdown and bleeding is a particular pattern
of hemorrhage in the stroma. This process is
characterized by collapse of the stroma and
coalescence of stromal cells into small aggre-gates and clusters separated by lakes of blood
(Fig. 5.1). As stromal condensation becomes
more advanced, small, rounded clusters of
stromal cells, sometimes called “stromal blue
balls,” become detached from the surrounding
tissue. These cellular clusters are characterized
by tightly packed aggregates of cells with
hyperchromatic small nuclei and scant cyto-plasm mixed with karyorrhectic debris (Fig.
5.2). With loss of cytoplasm, the collapsed
stromal cells in these aggregates can show
nuclear molding. They often are capped by
attenuated surface epithelium or by eosino-philic epithelial cells forming a syncytium (see
later). Sometimes these condensed aggregates
of cells form small polypoid extrusions along
the endometrial surface (Fig. 5.3) as they detach
from the intact stroma below. These are not
true polyps.
Another characteristic feature of early
breakdown, occurring before actual stromal
collapse, is the accumulation of nuclear debris
in the basal cytoplasm of glandular cells. This
feature is prominent in premenstrual and men-strual endometrium but also occurs, to a lesser
degree, with abnormal bleeding patterns (Fig.
5.2). The debris appears to represent nuclear
karyorrhexis from apoptosis of individual cells
within the glands.6
Along with stromal collapse, fibrin thrombi
typically form in small vessels, representing
another sentinel of abnormal bleeding.7;8
The
thrombi form either in superficial portions of
the spiral arteries or in ectatic venules in the
stroma beneath the surface epithelium (Fig.
Morphologic Features of Glandular and Stromal Breakdown 101
changes seen in abnormal endometrial bleeding
are somewhat different from the changes seen
in menstrual endometrium, however, as they
occur usually on a nonsecretory background
and are focal rather than diffuse. “Glandular
and stromal breakdown” is the best term to
describe the morphologic features without
attempting to assign an etiology or a prognosis.
Irregular endometrial glandular and stromal
breakdown, the “breakdown and bleeding pat-tern,” has certain unique features (Table 5.1)
that should be recognized, as it commonly
occurs with dysfunctional bleeding, especially
anovulatory cycles. These changes also occur
with organic lesions such as inflammation,
polyps, hyperplasia, and carcinoma, in which the
change may be limited or extensive. Regardless
of the extent of the morphologic changes,
abnormal glandular and stromal breakdown
usually does not occur uniformly throughout
the uterine cavity. As a result, abnormal
bleeding typically leads to a heterogeneous
pattern with fragments of intact, nonshedding
endometrium admixed with endometrium
showing the morphologic changes of abnormal
bleeding. Menstrual endometrium also shows
breakdown, but the changes affect all the tissue
and occur on a background of late secretory
phase glands (see Chapter 2). Furthermore, in
menstrual endometrium the breakdown is acute
and lacks the changes of chronic bleeding, such
as hemosiderin deposition, eosinophilic syncy-tial change, or foam cell accumulation, seen in
abnormal bleeding patterns.
Two factors complicate recognition of break-down and bleeding in endometrial biopsies.
Table 5.1. Features of endometrial breakdown and
bleeding.
Stromal “collapse”
Stromal cell clusters
Fibrin thrombi
Nuclear debris at base of gland cells
Nuclear debris in stroma
Eosinophilic syncytial change
Hemosiderin
Foam cells
Stromal fibrosis and hyalinization
102 5. Dysfunctional Uterine Bleeding
Figure 5.1. Glandular and stromal breakdown.
Abnormal bleeding pattern associated with chronic
inflammation secondary to retained placental site
shows extensive glandular and stromal breakdown.
5.4). These ectatic vessels develop in association
with any condition in which prolonged
endometrial growth is not followed by physio-logic shedding. Examples include anovulatory
cycles, hyperplasia, polyps, and progestin effect.
Thrombi form and extend into the stroma.
Fibrin by itself does not always signify true
bleeding; some fibrin may be a result of the
mechanical disruption that occurs during the
biopsy procedure.
Another feature of active bleeding is
eosinophilic syncytial change of the surface
epithelium. This lesion, previously termed
“papillary syncytial metaplasia,”9
“papillary
syncytial change,”10
or “surface syncytial
change,”11
is a result of focal, irregular break-down.10
Because of its consistent association
with breakdown, eosinophilic syncytial change
appears to be a degenerative/regenerative
alteration and is not a metaplastic transforma-tion.9;10
In eosinophilic syncytial change, por-tions of the surface epithelium coalesce into
syncytial aggregates of eosinophilic cells (Figs.
5.5 to 5.8). The cells often randomly pile up into
small papillae infiltrated by neutrophils, and
these papillae may contain microcystic spaces
(Figs. 5.5 to 5.7). In this change the pink cells
have sparsely vacuolated cytoplasm and indis-tinct cell borders, leading to the appearance of
a syncytium (Fig. 5.8). Nuclei vary from oval
to rounded and enlarged but have a uniform
chromatin distribution (Figs. 5.7 and 5.8). At
times the nuclei may be hyperchromatic with
irregular borders. Occasional nucleoli and
rare mitotic figures are present. These changes
should not be interpreted as atypical. The
underlying stromal collapse and condensation
helps to identify eosinophilic syncytial change
as a component of a bleeding pattern.
Because of stromal collapse and tissue
fragmentation associated with bleeding, the
endometrial glands and surface epithelium
Stromal cells are aggregated into rounded clusters
(arrows) and are surrounded by blood. The glands
show poorly developed secretory changes in this
specimen. Other fields showed residual placental site.
Morphologic Features of Glandular and Stromal Breakdown 103
become disrupted and crowded, yielding a
pattern that can mimic complex hyperplasia or
even adenocarcinoma (Fig. 5.9). Glandular and
stromal breakdown is recognized by the pattern
of stromal collapse as well as the disrupted
glands. With breakdown and bleeding, these
fragmented glands lose the stroma that should
be present between intact glands, even in
conditions such as hyperplasia, in which intact
glands show marked crowding.
Other epithelial changes associated with
breakdown are highly variable. As bleeding
becomes chronic, variable amounts of hemo-siderin deposits12
and foam cells13
appear in the
stroma. Neither hemosiderin nor foam cells
occur in normal cycling endometrium during
the reproductive years, apparently because the
tissue sloughs during menstruation. Foam cells,
originally thought to represent macrophages,
are endometrial stromal cells that become dis-tended with lipid following erythrocyte break-down in areas of nonphysiologic hemorrhage.13
These cells have abundant pale, faintly granu-lar cytoplasm and small oval nuclei. Foam cells
usually are associated with endometrial carci-noma (see Chapter 10), but they can occur in
hyperplasia or other benign pathologic condi-tions and therefore are a nonspecific finding.
The architecture of the glands and the cytologic
features of the epithelial cells, not the foam
cells, are features that determine the pathologic
diagnosis. Chronic bleeding occasionally results
in patches of stromal hyalinization and fibrosis
when bleeding is more severe and prolonged,
and this appears to be related to fibrin deposi-tion (Fig. 5.10).
Necrotic debris or old blood is sometimes
present in the lumens of otherwise normal
Figure 5.2. Glandular and stromal breakdown.
Focus of glandular and stromal breakdown in prolif-erative endometrium due to anovulatory cycles. Col-lapse of the tissue results in distortion of the glands
and condensation of stromal cells into tight clusters,
so-called blue balls (right upper corner). Nuclear
dust is present at the base of glandular cells.
Figure 5.3. Glandular and stromal breakdown. Pro-liferative endometrium with glandular and stromal
breakdown shows a focus of breakdown just beneath
Figure 5.4. Fibrin thrombus with glandular and stromal breakdown. Proliferative endometrium shows an
ectatic venule that contains a fibrin thrombus. There is early associated glandular and stromal breakdown.
the surface with collapsed stroma that is beginning
to form characteristic clusters.
Figure 5.5. Glandular and stromal breakdown with
eosinophilic syncytial change. A focus of prominent
eosinophilic syncytial change in an area of break-down in proliferative endometrium. The surface
Figure 5.6. Glandular and stromal breakdown with
eosinophilic syncytial change. At low magnification
the endometrium shows extensive breakdown. The
detached epithelium demonstrates marked
eosinophilic syncytial change that forms pseudopap-epithelium is heaped into a syncytium of eosinophilic
cells with indistinct cell borders. A few clusters of
stromal cells are entrapped in the syncytium.
illary structures. The stroma is collapsed into dense
aggregates. Other areas showed underdeveloped
secretory changes. The cause of the breakdown was
not determined but the pattern is benign.
Figure 5.7. Glandular and stromal breakdown with
eosinophilic syncytial change. Left: Proliferative
endometrium with breakdown shows prominent
eosinophilic syncytial change in a papillary configu-Figure 5.8. Glandular and stromal breakdown with
eosinophilic syncytial change. A syncytium of surface
epithelial cells overlying a focus of glandular and
ration. The eosinophilic epithelium overlies several
clusters of condensed stromal cells. Right: The
eosinophilic cells in the papillae show little atypia.
stromal breakdown shows haphazard nuclei and
pale, sparsely vacuolated cytoplasm.
Figure 5.9. Glandular and stromal breakdown with
artifactual crowding. Nonmenstrual secretory phase
endometrium undergoing extensive breakdown
Figure 5.10. Hyalinized stroma secondary to glan-dular and stromal breakdown. A focus of hyalinized
stroma in a sample that showed proliferative
endometrium with glandular and stromal break-shows artifactual crowding of the glands. The dis-continuous, collapsed stroma around the glands
helps to identify this pattern as an artifact.
down. This appearance is due to extravasation of
fibrin into the stroma. Inset: Focus of breakdown
with detached stromal clusters capped by
eosinophilic epithelium.
endometrial glands (Fig. 5.11). This debris
seems to result from abnormal breakdown with
entrapment of the debris within glands. The
association with abnormal endometrial break-down and bleeding is poorly defined, however,
and often no definite abnormalities are present
when luminal debris is seen. Without other
morphologic abnormalities, luminal debris is
a nonspecific finding with no known clinical
significance.
Dysfunctional Uterine Bleeding
Dysfunctional abnormalities are frequent
causes of uterine bleeding in perimenopausal
and perimenarcheal women, and they occur
to a lesser extent in women of reproductive
age.1;2;14–16
DUB occurs either because of lack
of ovulation following follicular development
(anovulatory cycles) or because of luteal phase
abnormalities.5
The latter include luteal phase
defects (LPDs) and abnormal persistence of
the corpus luteum (irregular shedding). Often,
before a biopsy is performed, DUB is managed
by hormonal therapy. When bleeding persists,
curettage often becomes necessary to control
bleeding and exclude organic lesions. Clini-cal–pathologic correlations of the full spectrum
of morphologic changes associated with DUB
are not known and may never be known.
Nonetheless, certain endometrial alterations
can be correlated with abnormalities in the
pattern of sex steroid hormone production.
Anovulatory cycles are, by far, the most
common cause of DUB,5
and in some classifi-cations DUB refers only to anovulatory bleed-ing. The prevalence of postovulatory LPDs as
etiologies for DUB is not known. Ovarian dys-function with anovulatory cycles or luteal phase
abnormalities also may present with infertility
(see Chapter 3) rather than DUB.
108 5. Dysfunctional Uterine Bleeding
Figure 5.11. Debris in glandular lumens. Several glands in midsecretory phase endometrium contain cellu-lar debris in their lumen. By itself, this finding has no known significance.
To place dysfunctional abnormalities in the
appropriate pathophysiologic context, these
disorders can be grouped into two broad
categories: estrogen-related and progesterone-related bleeding. The more common is estrogen-related DUB, which refers to episodes of
bleeding that are related to lack of ovulation
with alterations in endogenous estrogen levels.
Although not really a manifestation of DUB,
atrophy is included as a form of estrogen-related bleeding because it occurs when the
endometrium is deprived of estrogen for a rel-atively long period of time. In clinical classifica-tions of bleeding disorders, atrophy is not
regarded as a form of dysfunctional uterine
bleeding, yet it is a significant cause of abnor-mal bleeding.The second, less frequent category
of DUB is progesterone related and reflects
abnormal endogenous progesterone levels.
All these disorders, classified here as DUB,
reflect variations in ovarian hormone pro-duction. Exogenous hormones may produce
endometrial patterns that are indistinguishable
from the patterns seen in DUB caused by
endogenous hormone fluctuations. Although
this chapter specifically addresses DUB, similar
morphologic changes may be attributable to
exogenous hormones; these effects are dis-cussed further in Chapter 6.
Estrogen-Related Bleeding
Proliferative with Glandular and
Stromal Breakdown
This term describes the endometrial changes
resulting from anovulatory cycles. It is probably
the most common abnormality found in biop-sies performed for abnormal bleeding in peri-menopausal women. Anovulatory cycles with
bleeding also occur in perimenarcheal adoles-cents in whom regular ovulatory cycles are not
established. Anovulatory bleeding even occurs
sporadically in women throughout the repro-ductive years. Usually, this bleeding does not
lead to the need for biopsy in younger patients,
as the risk of other lesions, especially hyperpla-sia and carcinoma, is low in individuals of this
age.14;15
An exception is women with chronic
anovulation associated with the Stein–
Leventhal syndrome (polycystic ovarian dis-ease), because these women have an increased
risk of development of hyperplasia or
carcinoma.
Anovulatory cycles result when a cohort of
ovarian follicles begins to develop but ovulation
does not occur. Chronic anovulation may be the
result of a variety of disorders, including
hypothalamic dysfunction and obesity, because
of peripheral conversion of androgens to
estrogen in adipose tissue, as well as increased
androgen production by the adrenal glands or
the ovaries.1;2
Causes of anovulation follow-ing recruitment of follicles are complex. They
include defects in the hypothalamic–pituitary–
ovarian axis such as hyperprolactinemia, abnor-mal feedback mechanisms of hormonal control,
and local ovarian factors that interfere with
appropriate follicular development.1
Whatever
the pathogenesis, if ovulation does not occur, a
corpus luteum does not develop, and proges-terone is not produced. The follicles produce
estradiol, which stimulates endometrial growth.
The developing follicles may persist for variable
periods of time before undergoing atresia. As
long as the follicles persist, estradiol is produced
and the endometrium proliferates.
When these follicles undergo atresia, estra-diol production falls precipitously and estrogen
withdrawal bleeding occurs. In this instance, the
loss of estrogenic support of endometrial pro-liferation results in destabilization of lysosome
membranes and vasoconstriction with bleeding.
In contrast to estrogen withdrawal bleeding,
estrogen breakthrough bleeding results from
persisting follicles that produce estradiol; the
proliferating endometrium becomes thicker
and outgrows its structural support. Focal vaso-constriction and thrombosis of dilated capillar-ies follow. In either event, the result is irregular
breakdown and bleeding of the endometrium.
Although the terms “withdrawal” and
“breakthrough” help to describe the mecha-nisms for estrogen-related bleeding, various
authors have defined these terms in different
ways. The lack of uniform interpretation of
these terms limits their usefulness for report-ing the pathologic changes associated with
endometrial bleeding.
Estrogen-Related Bleeding 109
In most cases of anovulatory DUB, the
endometrium shows a proliferative phase
pattern with glandular and stromal breakdown
(Figs. 5.12 and 5.13) (see also Figs. 5.2 to 5.5 and
5.7). The amount of tissue and the architectural
pattern of the glands depend on the duration of
unopposed estrogenic stimulation, not neces-sarily the level of estrogen. In addition, the
extent of breakdown can be highly variable,
ranging from minute areas to extensive involve-ment of the specimen. Sporadic anovulation
often results in rapid atresia of follicles with
estrogen withdrawal bleeding. This results in
minimal endometrial proliferation. A small
amount of endometrium with poorly devel-oped, weakly proliferative glands and stroma
develops (Fig. 5.14). Chronic anovulation re-sults in persistence of follicles and sustained
unopposed estrogen stimulation. A greater
quantity of endometrial tissue develops with
actively proliferating glands and augmented
glandular tortuosity. Dilated venules appear in
the subepithelial stroma and often thrombose
(Fig. 5.15). Because of continuous estrogenic
stimulation, the tissue often shows estrogen-induced epithelial changes (“metaplasia”),
especially ciliated cell and eosinophilic cell
change (see Chapter 9). The glands also may
show focal subnuclear vacuolization as a
response to estrogen stimulation, but the extent
and uniformity of the vacuolization are less
than that seen in normal early secretory glands.
The cytoplasmic changes and subnuclear vac-uoles complicate the interpretation of the his-tologic pattern but do not change the diagnosis.
Prolonged, unopposed estrogenic stimulation
110 5. Dysfunctional Uterine Bleeding
Figure 5.12. Proliferative with glandular and
stromal breakdown. Anovulatory bleeding pattern in
perimenopausal women shows proliferative phase
pattern with foci of breakdown along surface. This
patient was not on hormones, but estrogen replace-ment therapy can present a similar appearance. Inset:
Area of fragmentation in biopsy shows clusters of
collapsed endometrium with stromal condensation
and eosinophilic syncytial change on surface adja-cent to detached proliferative glands.
Figure 5.13. Proliferative with glandular and
stromal breakdown. The tissue is extensively frag-mented secondary to the breakdown and the proce-dure. A few detached glands continue to show
proliferative features with tubular outlines. Other
portions of the epithelium show breakdown with
eosinophilic change and underlying stromal collapse.
Figure 5.14. Weakly proliferative with glandular
and stromal breakdown. Fragmented endometrium
with isolated glands, extensive blood, and scant
stroma from perimenopausal patient with anovula-tory bleeding. The glandular epithelium is minimally
stratified and shows only rare mitotic activity; it is
therefore designated weakly proliferative. Inset:
Focus of glandular and stromal breakdown with
stromal cluster, nuclear debris, and eosinophilic
syncytial change.
112 5. Dysfunctional Uterine Bleeding
Figure 5.15. Ectatic venules with thrombi. Speci-men from a perimenopausal patient with DUB
attributable to apparent anovulatory cycles shows
proliferative endometrium with a thrombus in super-ficial ectatic venule (arrow) and another dilated
venule to the left of center. Glandular and stromal
breakdown was present in other areas.
also can lead to the development of varying
degrees of hyperplasia, atypical hyperplasia,
and even well differentiated adenocarcinoma,
but these organic lesions are not functional dis-orders and, as such, are not considered causes
of DUB.
When proliferative endometrium shows
breakdown and bleeding, the pattern strongly
suggests anovulatory cycles. Exogenous estro-gens can cause similar patterns, and therefore a
complete clinical history is needed to be certain
that the bleeding pattern is truly due to anovu-lation. The differential diagnosis of prolifera-tive phase endometrium with glandular and
stromal breakdown also includes inflammation,
polyps, and leiomyomas. In such cases, the pres-ence of other features, such as plasma cells in
chronic endometritis or the dense stroma and
thick-walled vessels of polyps, establishes the
proper diagnosis.
Disordered Proliferative Phase and
Persistent Proliferative Phase
When chronic anovulatory cycles result in
abundant proliferative tissue, mild degrees
of disorganization characterized by focal
glandular dilation may occur. Usually these are
regarded as variants of normal proliferative
endometrium. Sometimes more sustained
estrogen stimulation may result in the focal
branching and some dilation of glands, yielding
a proliferative phase pattern that is neither
normal nor hyperplastic (Fig. 5.16). The terms
“disordered proliferative phase pattern” and
“persistent proliferative phase” have been
applied to describe this pattern of proliferative
endometrium with tortuous and mildly disor-ganized glands. A designation of a “disordered”
or “persistent proliferative phase” has utility in
correlating the morphologic findings with the
apparent pathophysiology. When used, these
terms should be clarified in a note so that the
gynecologist understands the clinical signifi-cance of the change.
In our experience, “disordered proliferative”
often is inappropriately applied to a variety
of patterns, including normal proliferative
endometrium, proliferative endometrium with
breakdown, artifactual crowding, basalis, and
simple hyperplasia. The diagnosis of disordered
proliferative phase should be reserved for cases
in which assessment is based on intact, well-oriented fragments of tissue. In these areas the
abnormal glands should be focal. These glands
are qualitatively similar to those seen in simple
hyperplasia, but they are limited in extent and
interspersed among glands with a normal pro-liferative phase pattern. This criterion helps to
separate the focal disordered proliferative
phase pattern from simple hyperplasia, a more
diffuse abnormality. If the tissue is extensively
fragmented or disrupted by the procedure and
contains mainly detached proliferative glands,
it is best to diagnose the change only as prolif-erative. Extensive breakdown in proliferative
endometrium can also display a disorganized
appearance to the glands because of fragmen-tation (see Fig. 5.13), but again this change is
not that of a true disordered proliferative phase
pattern.
Atrophy
As previously noted, atrophy is an important
cause of abnormal uterine bleeding in post-menopausal patients, found in 25% or more of
Figure 5.16. Disordered proliferative phase pattern.
Portion of endometrium from a patient with appar-ent anovulatory bleeding shows disorganized prolif-erative phase glands with focal branching and
glandular dilatation.
Estrogen-Related Bleeding 113
cases coming to biopsy.17
The percentage of
patients with atrophy varies greatly among
studies, probably reflecting different patient
populations as well as variations in indication
for biopsy and criteria for diagnosis of atrophy
among different pathologists.18–23
In many labo-ratories, atrophy is found in up to 50% of biopsy
specimens taken for postmenopausal bleeding,
and in one study 82% of cases of post-menopausal bleeding were attributable to
atrophy.20
Besides being common in post-menopausal patients, atrophic endometrium
can occur in reproductive-age patients with pre-mature ovarian failure, either idiopathic or due
to radiation or chemotherapy for malignancies.
With atrophy, tissue obtained at biopsy is
typically scant, often consisting only of a small
amount of mucoid material. Characteristically,
atrophic endometrium is composed of tiny
strips and wisps of surface endometrium and
detached, fragmented endometrial glands
(Fig. 5.17). The epithelium is low columnar to
cuboidal with small, dark nuclei and minimal
cytoplasm. Stroma is scant or absent, consisting
of a few clusters of small spindle cells. Mitotic
activity is absent. The cystic change seen in
atrophic glands in hysterectomy specimens is
not observed in biopsies because tissue frag-mentation from the procedure disrupts the
glands. Breakdown and bleeding may be super-imposed on the features of atrophy, although
often, even when there is a history of abnormal
uterine bleeding, the sections show no evidence
of glandular and stromal breakdown.
Although there is a paucity of tissue in biopsy
specimens of atrophic endometrium, these
specimens are not insufficient or inadequate.
The scant tissue may be all that is present and
therefore is completely representative of
the lining of the uterine cavity. The minimal
114 5. Dysfunctional Uterine Bleeding
Figure 5.17. Atrophy. Atrophic endometrium con-sists of scant tissue that shows extensive fragmenta-tion. The specimen consists largely of detached wisps
and strips of epithelium with almost no stroma. Inset:
The epithelial cells have small, dark nuclei, scant
cytoplasm, and no mitotic activity.
quantity of tissue should serve as a clue to the
diagnosis; it does not represent an insufficient
specimen (see Clinical Queries and Reporting).
Progesterone-Related Bleeding
Biopsy specimens from reproductive-age and
perimenopausal women occasionally show ab-normal secretory phase patterns with associated
nonmenstrual breakdown and bleeding. In
such cases the pattern is secretory owing to
ovarian progesterone production, but the glan-dular and stromal changes usually are less
advanced than those seen in normal late secre-tory endometrium.5;24–28
The endometrial
pattern does not correlate with any date of the
normal luteal phase. The glands may show
secretory changes yet lack marked tortuosity
and secretory exhaustion, while the stroma
lacks extensive predecidual change (Figs. 5.18
and 5.19). In other cases the glands appear to
show a “hypersecretory” pattern, with vacuo-lated cytoplasm, marked tortuosity, and luminal
secretion, while the stroma lacks predecidual
change. In addition, the tissue shows foci of
breakdown with characteristic changes such as
nuclear dust, fibrin thrombi, and dense cell clus-ters, similar to that which occurs in the prolifer-ative endometrium with glandular and stromal
breakdown (Figs. 5.1 to 5.3). Often in abnormal
secretory bleeding patterns, the glands show
stellate shapes as they involute (Fig. 5.20). This
latter pattern of collapsing, star-shaped secre-tory glands is nonspecific, however, and simply
shows secretory gland regression that could be
due to a variety of factors.
These changes may reflect DUB due to
luteal phase abnormalities that include LPDs
and irregular shedding (see Chapter 2). The
Progesterone-Related Bleeding 115
Figure 5.18. Abnormal secretory phase with glan-dular and stromal breakdown. In this field the glands
are intact. The stroma shows evidence of breakdown
with early collapse. Inset: The glands are tortuous
with vacuolated cytoplasm and basally oriented
nuclei corresponding to secretory days 18 to 19.
Figure 5.19. Abnormal secretory phase. The glands
show secretory changes with basal nuclei and vacuo-lated cytoplasm but are small and tubular rather than
Figure 5.20. Abnormal secretory phase with glan-dular and stromal breakdown. Early collapse of
endometrium with a poorly developed secretory
phase pattern shows involuting, star-shaped glands.
tortuous. The stroma is dense. Other areas in the
sections showed glandular and stromal breakdown.
This pattern of gland involution is often seen focally
in secretory phase endometrium that is undergoing
nonmenstrual glandular and stromal breakdown.
etiology and frequency of dysfunctional bleed-ing caused by luteal phase abnormalities are not
known, however, as these disorders appear to
be sporadic and do not persist long enough to
permit clinical–pathologic correlations. Altera-tions in the morphology of the endometrium
due to changes in the absolute or relative levels
of estrogen and progesterone have been well
established in experimental systems,29
so it is
likely that abnormal secretory bleeding patterns
are, at least in part, the result of ovulatory abnor-malities that involve the luteal phase. Nonethe-less, the secretory patterns with nonmenstrual
bleeding are not well characterized. When
a pattern of nonmenstrual phase secretory
endometrium with breakdown is present, the
abnormality may be due to defined or undefined
luteal phase abnormalities or other causes that
are not evident from the sections.
Luteal Phase Defects
Of the two defined luteal phase abnormalities
that may cause abnormal bleeding, LPD prob-ably occurs more frequently. As discussed in
Chapter 2, the corpus luteum is “insufficient” in
LPD, either regressing prematurely or failing to
produce an adequate amount of progesterone
to sustain normal secretory phase develop-ment. This is a sporadic disorder of the repro-ductive and perimenopausal years. With LPD,
ovulation occurs, so secretory changes develop.
If abnormal bleeding is the result, the appear-ance is that of breakdown and bleeding in a
nonmenstrual secretory phase pattern. Glands
with secretory changes, including basally ori-ented nuclei and vacuolated cytoplasm, but
lacking the tortuosity of late secretory phase
glands, characterize the pattern. Focal break-down is present with “stromal blue balls” and
karyorrhectic debris. This pattern is nonspecific,
however, and may be attributable to other
factors discussed in the following.
Irregular Shedding
Irregular shedding is attributed to a persistent
corpus luteum with prolonged progesterone
production.15;24;26;27
This is the least studied and
consequently the most poorly understood form
of dysfunctional patterns. Some authors have
used this term to refer to irregular secretory
phase bleeding from a variety of causes, includ-ing early pregnancy, inflammation, and exoge-nous hormone effects, but we prefer to restrict
this term to bleeding caused by true ovulatory
dysfunction. One pattern of irregular shedding
yields a mixed phase pattern composed of
secretory and proliferative endometrium (Fig.
5.21). The diagnosis is reserved for those spec-imens in which there is a mixed pattern of
secretory and proliferative glands at least 5
days after the onset of bleeding. Irregular shed-ding is also manifested by irregular secretory
phase development in which different foci show
more than 4 days’ difference in the morpho-logic date. Breakdown and bleeding with glan-dular and stromal collapse is present, usually
focally, but occasionally in a diffuse pattern.
Although the frequency of irregular shedding
as a cause of DUB is not known, it is an unusual
event in our experience.
Abnormal Secretory
Endometrium with Breakdown
of Unknown Etiology
Some examples of secretory endometrium with
abnormal, nonmenstrual bleeding patterns pre-sumably reflect the specific luteal phase abnor-malities described in the preceding. A variety of
other factors may also be associated with a
pattern of aberrant secretory phase develop-ment with superimposed bleeding (Table 5.2).
For example, endometrial changes associated
with abortions or ectopic pregnancies, response
to exogenous progestins, tissue near a polyp,
endometrium overlying leiomyomas, and
endometrium involved with inflammation or
adhesions all can show patterns of abnormal
secretory development and bleeding. Other
poorly understood ovarian disorders, such as
a luteinized unruptured follicle, presumably
could result in abnormal secretory endometrial
changes. With this latter entity, developing fol-licles are believed to undergo luteinization of
the granulosa and theca cells with progesterone
production in the absence of ovulation. In addi-tion to these considerations, management of
DUB often involves progestational therapy. If
Progesterone-Related Bleeding 117
bleeding is not controlled, curettage is per-formed. Accordingly, progestin effects super-imposed on the underlying abnormality may
complicate the histology.
These patterns show glands with secretory
changes such as basally oriented nuclei and
diffuse cytoplasmic vacuolization and absence
of mitotic activity (Figs. 5.18 to 5.20). The glands
may be tortuous. Often the stroma is dense,
lacking edema or predecidua.The endometrium
in such cases cannot be assigned to any histo-logic day of the normal secretory phase of the
menstrual cycle. The changes of glandular and
stromal breakdown are similar to those found
in any bleeding phase endometrium with glan-dular and stromal collapse, “stromal blue balls,”
and eosinophilic syncytial change. With early
breakdown, tortuous secretory glands often
show star-shaped outlines (Fig. 5.20).
There are no clear-cut clinical correlations
for many abnormal secretory phase patterns
with breakdown, so the alterations at times
defy precise pathologic diagnosis. Nonetheless,
recognition of the general category of ab-normal secretory bleeding patterns helps to
exclude other specific organic lesions and
suggests the possibility of a luteal phase dys-functional abnormality if no other pathologic
condition is clinically manifest.
118 5. Dysfunctional Uterine Bleeding
Table 5.2. Possible causes of nonmenstrual secre-tory phase bleeding.
Luteal phase defect
Persistent corpus luteum (irregular shedding)
Organic lesions
Submucosal leiomyomas
Intrauterine adhesions
Inflammation
Complications of pregnancya
Progestin effects
a
See Chapter 3, Table 3.1.
Figure 5.21. Irregular shedding with mixed phase
pattern. Two fields from the same sample show a
mixed phase pattern with proliferative (left) and
secretory (right) changes. This mixed phase pattern
may be due to a persistent corpus luteum, but the
etiology cannot be determined by morphology alone,
and clinical correlation is required.
Clinical Queries and Reporting
When a biopsy is performed for DUB, the
report should address the presence or absence
of morphologic changes of breakdown and
bleeding as well as any specific lesions. If the
pattern is that of proliferative endometrium
with breakdown and if the clinical history is
appropriate, the changes can accurately be
attributed to anovulatory cycles. A descriptive
diagnosis such as “proliferative endometrium
with glandular and stromal breakdown” offers
a precise morphologic interpretation of the
anovulatory bleeding pattern that often is suf-ficient for clinical management. An additional
comment indicating that the change is compat-ible with anovulatory cycles helps to clarify the
diagnosis. If the changes show nonmenstrual
secretory endometrium with breakdown but
these are not diagnostic of a defined luteal
phase abnormality, descriptive terms such as
“abnormal secretory phase pattern with break-down” communicate the observation of an
abnormal yet benign appearance while not
assigning definite morphologic etiology. In
general, a comment regarding the absence
of other possible causes of bleeding, such as
hyperplasia, inflammation, pregnancy, or
polyps, is most useful in addressing specific clin-ical concerns.
Because atrophy is one of the most frequent
causes of abnormal bleeding in post-menopausal patients, it is important to recog-nize the morphologic features of atrophy and
correctly report the findings. A scant amount of
endometrium consisting of detached strips of
atrophic endometrial epithelium with little
stroma should be regarded as consistent with
atrophy and not “insufficient for diagnosis.” A
brief comment or description of the findings
helps the clinician understand the basis of the
diagnosis while providing reassurance that the
endometrium has, in fact, been sampled.
Occasional biopsies show extensive break-down and bleeding that largely obscures the
cytologic details of the glands and stroma.
Although it is usually possible to exclude neo-plastic processes in such cases, detailed assess-ment of the endometrium to determine the
underlying pathologic process becomes diffi-cult. Unless the breakdown is clearly menstrual,
i.e., reflecting the shedding at the end of a
normal ovulatory cycle, breakdown patterns
should not be diagnosed as “menstrual.”
Instead, it is better to use descriptive diagnoses
that reflect the morphologic changes. On
rare occasions endometrial biopsy in a post-menopausal patient may show only evidence of
abnormal bleeding such as foam cells. In such
cases the subtle evidence of abnormal bleeding
should be recognized and reported; further
sampling may be needed to determine the
cause, as endometrial carcinoma becomes a
more likely cause of abnormal bleeding in this
age group.
It is imperative that the terms used in
descriptive pathologic diagnoses be clearly
understood by the clinician. The terms
employed can be self-explanatory, especially
when a brief microscopic description accompa-nies the report to clarify the histologic finding
and to exclude more specific organic lesions.
The diagnosis of “proliferative endometrium
with glandular and stromal breakdown” is an
example of a clinically relevant term that can
be applied to apparent anovulatory bleeding.
Diagnoses such as “disordered prolifera-tive phase pattern” or “persistent prolifera-tive phase” applied to the proliferative
endometrium of anovulatory cycles are most
useful when there is a clear understanding of
the microscopic findings and the physiologic
and clinical significance of the pattern. There-fore an explanatory note added to the diagnosis
is helpful to describe the changes and indicate
that the lesion lacks atypia or clear-cut criteria
for hyperplasia. The terms “withdrawal” and
“breakthrough” should be avoided in patho-logic diagnoses because they lack clear defini-tions in the clinical literature regarding
endometrial bleeding.
References
(1) Speroff L, Glass RH, Kase NG. Clinical gyne-cologic endocrinology and infertility, 6th ed.
Baltimore: Lippincott Williams & Wilkins, 1999.
(2) Bayer SR, DeCherney AH. Clinical manifesta-tions and treatment of dysfunctional uterine
bleeding. JAMA 1993; 269:1823–1828.
References 119
(3) Galle PC, McRae MA. Abnormal uterine
bleeding. Finding and treating the cause. Post-grad Med 1993; 93:73–81.
(4) Mishell DR, Jr. Abnormal uterine bleeding.
In: Stenchever MA, Droegemueller W, Herbst
AL, Mishell DR, Jr, eds. Comprehensive gyne-cology. 4th ed. St. Louis: Mosby; 2001:
1079–1097.
(5) Vakiani M, Vavilis D, Agorastos T, Stam-atopoulos P, Assimaki A, Bontis J. Histopatho-logical findings of the endometrium in patients
with dysfunctional uterine bleeding. Clin Exp
Obstet Gynecol 1996; 23:236–239.
(6) Stewart CJ, Campbell-Brown M, Critchley HO,
Farquharson MA. Endometrial apoptosis in
patients with dysfunctional uterine bleeding.
Histopathology 1999; 34:99–105.
(7) Picoff RC, Luginbuhl WH. Fibrin in the
endometrial stroma: Its relation to uterine
bleeding. Am J Obstet Gynecol 1964;
88:642–646.
(8) Ferenczy A. Pathophysiology of endometrial
bleeding. Maturitas 2003; 45:1–14.
(9) Clement PB. Pathology of the uterine corpus.
Hum Pathol 1991; 22:776–791.
(10) Zaman SS, Mazur MT. Endometrial papillary
syncytial change. A nonspecific alteration asso-ciated with active breakdown. Am J Clin Pathol
1993; 99:741–745.
(11) Silverberg SG, Kurman RJ. Tumors of the
uterine corpus and gestational trophoblastic
disease. Atlas of tumor pathology, 3rd series,
Fascicle 3. Washington, DC: Armed Forces
Institute of Pathology, 1992.
(12) Reeves G, Sommers SC. Endometrial hemo-siderin as evidence of metrorrhagia. Obstet
Gynecol 1962; 19:790–792.
(13) Fechner RE, Bossart MI, Spjut HJ. Ultrastruc-ture of endometrial stromal foam cells. Am J
Clin Pathol 1979; 72:628–633.
(14) Altchek A. Dysfunctional uterine bleeding in
adolescence. Clin Obstet Gynecol 1977; 20:
633–650.
(15) Scommegna A, Dmowski WP. Dysfunctional
uterine bleeding. Clin Obstet Gynecol 1973; 16:
221–254.
(16) Aksel S, Jones GS. Etiology and treatment
of dysfunctional uterine bleeding. J Obstet
Gynecol 1974; 44:1–13.
(17) Gredmark T, Kvint S, Havel G, Mattsson LA.
Histopathological findings in women with post-menopausal bleeding. Br J Obstet Gynaecol
1995; 102:133–136.
(18) Rubin SC. Postmenopausal bleeding: etiology,
evaluation, and management. Med Clin N Am
1987; 71:59–69.
(19) Schindler AE, Schmidt G. Post-menopausal
bleeding: A study of more than 1000 cases.
Maturitas 1980; 2:269–274.
(20) Choo YC, Mak KC, Hsu C, Wong TS, Ma HK.
Postmenopausal uterine bleeding of nonor-ganic cause. Obstet Gynecol 1985; 66:225–228.
(21) Meyer WC, Malkasian GD, Dockerty MB,
Decker DG. Postmenopausal bleeding from
atrophic endometrium. Obstet Gynecol 1971;
38:731–738.
(22) Gambrell RD. Postmenopausal bleeding. J Am
Geriatr Soc 1974; 22:337–343.
(23) Lidor A, Ismajovich B, Confino E, David MP.
Histopathological findings in 226 women with
post-menopausal uterine bleeding. Acta Obstet
Gynecol Scand 1983; 65:41–43.
(24) Sherman ME, Mazur MT, Kurman RJ. Benign
diseases of the endometrium. In: Kurman RJ,
ed. Blaustein’s pathology of the female genital
tract. 5th ed. New York: Springer-Verlag; 2002:
421–466.
(25) Dallenbach-Hellweg G. The endometrium of
infertility. A review. Pathol Res Pract 1984;
178:527–537.
(26) Dallenbach-Hellweg G. Histopathology of the
endometrium, 4th ed. New York: Springer-Verlag; 1987.
(27) Hendrickson MR, Kempson RL. Surgical
pathology of the uterine corpus (Major prob-lems in pathology series), Vol. 12. Philadelphia:
W.B. Saunders; 1980.
(28) Buckley CH, Fox H. Biopsy pathology of the
endometrium 2nd ed. London: Arnold; 2002.
(29) Good RG, Moyer DL. Estrogen–progesterone
relationships in the development of secretory
endometrium. Fertil Steril 1968; 19:37–45.
120 5. Dysfunctional Uterine Bleeding
for some disorders. The endometrium shows
the effects of these hormones.
An endometrial biopsy or curettage may
be performed when abnormal bleeding occurs
or when hormone therapy does not correct
abnormal bleeding that is thought to be
dysfunctional. Sometimes, however, the biopsy
is intended to evaluate the status of the
endometrium following hormonal therapy, as in
the case of hyperplasia managed with progestin
therapy or routine follow-up of patients on
hormone replacement therapy. In other cir-cumstances the endometrial sampling is coin-cidental with another procedure, such as tubal
ligation, where the patient has received
hormone therapy to ensure no interval preg-nancy. The hormone, the dosage, and the dura-tion of therapy influence the appearance of the
endometrium. Clinical information regarding
hormone use helps in the pathologic interpre-tation but this history sometimes is incomplete
when the specimen is received by the pathology
laboratory. Consequently, the possibility of
exogenous hormonal effects should always be
kept in mind.
This chapter reviews the effects of different
types of hormones on the endometrium: (1)
hormones used in women of reproductive age
that clearly have estrogenic or progestogenic
effects, such as oral contraceptives; (2) estro-gen–progestin hormone replacement therapy
in postmenopausal women; (3) tamoxifen
therapy for breast cancer; and (4) other hor-mones with less well established effects on the
endometrium.
121
6
Effects of Hormones
Women receive hormone preparations for a
variety of reasons, including birth control and
treatment for dysfunctional uterine bleeding,
perimenopausal and postmenopausal symp-toms, endometriosis, endometrial hyperplasia
and carcinoma, breast carcinoma, and certain
types of infertility. Usually the exogenous
hormone is some form of progestin, but estro-genic and even androgenic hormones are used
Estrogenic Hormones  . . . . . . . . . . . . . . . . 122
Progestins and Oral Contraceptives  . . . . . 122
Patterns of Response  . . . . . . . . . . . . . . . 123
Combined Estrogen and Progestin as
Replacement Therapy for Menopausal
Women  . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
Progestin-Like Effects with No Hormone
Use  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132
Effects of Other Hormones . . . . . . . . . . . . 133
Tamoxifen  . . . . . . . . . . . . . . . . . . . . . . . 133
Raloxifene  . . . . . . . . . . . . . . . . . . . . . . . 136
Clomiphene Citrate  . . . . . . . . . . . . . . . . 136
Danazol  . . . . . . . . . . . . . . . . . . . . . . . . . 137
Human Menopausal Gonadotropins/
Human Chorionic Gonadotropin  . . . . . 137
Gonadotropin-Releasing Hormone
Agonists . . . . . . . . . . . . . . . . . . . . . . . . . 138
Antiprogestin RU 486  . . . . . . . . . . . . . . 138
Clinical Queries and Reporting  . . . . . . . . 139
Postmenopausal Hormone
Replacement  . . . . . . . . . . . . . . . . . . . . . 139
Abnormal Uterine Bleeding  . . . . . . . . . 139
Treatment of Hyperplasia  . . . . . . . . . . . 140
Infertility Therapy  . . . . . . . . . . . . . . . . . 140
Estrogenic Hormones
Estrogen therapy is largely used in peri-menopausal or postmenopausal women to treat
symptoms of the menopause, such as vasomo-tor instability, atrophic vaginitis, and osteo-porosis.1;2
Estrogenic substances include
conjugated estrogens, such as Premarin (Wyeth
Pharmaceuticals, Philadelphia, PA), and other
synthetic estrogens, such as ethinyl estradiol or
diethylstilbestrol (DES). Use of estrogenic
hormones by themselves is associated with
an increased risk of developing endometrial
adenocarcinoma, so the use of these hormones
alone is now unusual in patients with a uterus.
Consequently, the effects of unopposed exoge-nous estrogen are seen less frequently in biopsy
specimens than the effects of combined estro-gen–progestin compounds, as progestins abro-gate the effect of estrogen stimulation on the
uterus. Nonetheless, some patients do receive
estrogen replacement only.
Unopposed estrogenic stimulation causes
the endometrium to proliferate.3–5
The result is
variable, depending on the dose and duration
of use. Often the pattern is that of proliferative
phase endometrium, showing tubular to tortu-ous glands and abundant stroma. The patterns
can be identical to those seen with anovulatory
cycles and may include superimposed break-down and bleeding (see Chapter 5). Continued,
prolonged estrogenic stimulation can lead to
disordered proliferative phase patterns and
hyperplasia. Estrogen-related epithelial cyto-plasmic changes, especially squamous differ-entiation and ciliated cell change, also often
occur.
In some patients continued estrogen use
leads to atypical hyperplasia and adenocarci-noma.6;7
The risk of malignancy increases with
the duration of therapy. Usually unopposed
estrogen use for at least 2 to 3 years is found
in patients who develop adenocarcinoma,
and the highest risk is in patients who have
taken estrogens for 10 years or longer. The
duration of therapy generally is more impor-tant than the dose of the estrogen. When carci-noma develops, it usually is low grade and
superficially invasive, but high-grade lesions
may occur.
All of the estrogen-related changes are re-viewed elsewhere in the text, including normal
proliferative phase patterns (Chapter 2), prolif-erative with glandular and stromal breakdown
(Chapter 5), hyperplasia and cytoplasmic
change (Chapter 9), and carcinoma (Chapter
10). The reader should refer to those chapters
for detailed morphologic descriptions of the
specific entities.
Progestins and Oral
Contraceptives
Although progestin effects are common, the
subject of progestin-related changes is com-plex. Various forms of these synthetic analogues
of progesterone, also termed “progestogens” or
“progestagens,” are widely used, either alone or
in combination with an estrogen. Progestin-only therapy is useful in the empirical medical
management of abnormal uterine bleeding
that clinically appears to be dysfunctional.
These hormones, such as medroxyprogesterone
acetate or norethindrone acetate, suppress ovu-lation and endometrial growth. They also lead
to secretory maturation and progesterone with-drawal bleeding, effecting a medical “curet-tage.” Consequently, progestins are especially
helpful in managing ovulatory disorders where
irregular, noncyclical endometrial growth
results in abnormal bleeding. Often a trial of
progestin is given in an attempt to alleviate
apparent dysfunctional bleeding, and if the
bleeding does not resolve, biopsy or curettage
follows to exclude other organic pathology.
Some progestins, such as oral or injectable
medroxyprogesterone acetate, may also be
used to treat neoplasia of the breast or
endometrium. Still others, such as norgestrel,
are used for contraception. Progestins, espe-cially oral contraceptives, also are used to treat
endometriosis.
Progestins, usually given in combination with
estrogens, are the basis for the oral contra-ceptive or “birth control pill.” Most of the oral
contraceptives are used in a fixed-dose formu-lation, with small doses of both the estrogen
and the progestin taken daily. Some oral
122 6. Effects of Hormones
contraceptives use a “phasic” combination with
increasing amounts of progestin over a 21-day
medication period. In either case the combina-tion estrogen and progestin is administered
over 3 weeks, and no medication is given in the
4th week to allow withdrawal bleeding to occur.
Some oral contraceptives contain only prog-estin. The dose of progestin and estrogen used
in modern oral contraceptives is much lower
than that used in the initial formulations of oral
contraceptives 20 to 30 years ago. Conse-quently, the pharmacologic effects of these
steroid contraceptives are somewhat different
than those originally described.
The morphologic appearance of the
endometrium following progestin therapy is
variable and depends on the underlying status
of the endometrium as well as the dose,
potency, and duration of progestin therapy.7–13
In fact, brief use of oral contraceptives for
emergency contraception (Yuzpe regimen,14
or
“morning after pill”) shows no significant his-tologic effects on the endometrium.15;16
With
more prolonged use of progestin, however, its
effects can persist for several weeks to months
following cessation of their use. To help simplify
this complex subject, the effects of the prog-estins can be placed into three general mor-phologic patterns that form the basis for
understanding the entire spectrum of prog-estin-mediated changes. These patterns include:
(1) decidual (pregnancy-like) changes, (2)
secretory changes, and (3) atrophic changes
(Table 6.1). The pattern encountered depends
on the degree of estrogen “priming” of the
endometrium and the dose and duration of
administration of the progestin. In reality, there
often is overlap between the various patterns of
progestin effect in the endometrium.
Patterns of Response
Decidual Pattern
The decidual or pregnancy-like pattern, as
the term implies, features differentiation of
endometrial glands and stroma to a point
where they resemble the endometrium in
pregnancy with decidual transformation of the
stroma. Although the term “decidua” applies
most strictly to the endometrium of pregnancy,
this term also is useful for describing this prog-estin-induced pattern. This exaggerated effect
typically occurs in endometrium, which is influ-enced by high estrogen levels and therefore is
actively growing and proliferating. This mor-phology is most common following high-dose
progestin therapy for anovulatory cycles or for
hyperplasia. In these cases the amount of tissue
can be copious, and the biopsy or curettage can
yield large polypoid tissue fragments (Fig. 6.1).
Although the tissue is polypoid, this finding
does not indicate the presence of true polyps.
With marked progestin effect, the stromal cells
become enlarged and show abundant cyto-plasm and prominent cell borders, resembling
the decidua of pregnancy. The stroma can show
occasional mitotic figures. In this form of
marked progestin effect, the glands develop a
hypersecretory pattern with vacuolated cyto-plasm and abundant luminal secretions. Some
glands are dilated. The Arias-Stella reaction
with nuclear enlargement and hyperchromasia
may occur in glands, but this is very rare. The
spiral arteries also can show marked thickening
with endothelial and smooth muscle hyper-plasia.13
The venules in the superficial portion
of the endometrium become ectatic (Fig. 6.2).
Occasional cases of decidua-like progestin
Progestins and Oral Contraceptives 123
Table 6.1. Morphologic features of progestin
effects.
Decidual (pregnancy-like) effects
Abundant tissue, often polypoid
Glands show marked secretory activity
Stroma appears decidualized with lymphoid infiltrate
Vascular ectasia
Secretory effects
Moderate to sparse amount of tissue
Mildly tortuous secretory glands lined by columnar
cells
Stromal cells plump, oval (predecidual)
Vascular ectasia
Atrophic effects
Sparse tissue
Glands small and atrophic, not coiled
Variable amount of stroma with plump to spindle-shaped cells
Figure 6.1. Progestin effect, decidual pattern.
Marked decidual reaction following progestin
therapy of anovulatory bleeding. The changes resem-ble the endometrial decidual transformation in preg-Figure 6.2. Progestin effect, decidual pattern.
Marked decidual change in the stroma and small
inactive glands lined by a single layer of epithelium
nancy. Glands with secretory exhaustion are sur-rounded by abundant decidualized stroma. Small
ectatic vessels are present in the superficial portions
of the endometrium.
characterize the decidual pattern of progestin effect.
An ectatic venule is present in the lower left part of
the field.
Progestins and Oral Contraceptives 125
effect show prominent squamous change
(metaplasia) within glands. This change is
usually seen in cases in which the biopsy results
that led to therapy demonstrate hyperplasia,
often with no squamous differentiation.11;17
Cases with advanced decidual changes often
show areas of breakdown and bleeding, espe-cially as the dilated venules thrombose (Fig.
6.3). As a result, many of the features of break-down and bleeding described in Chapter 5 are
superimposed on the progestin effect. With
breakdown, the collapse of the stroma and
glands significantly alters their appearance, par-tially masking the patterns of development.
In these areas of breakdown, the decidua-like
character of the stromal cells is lost as the cells
degenerate and lose cytoplasm. The glands frag-ment and become haphazardly oriented. Con-sequently, it remains important to avoid areas
of active bleeding and find intact tissue in order
to accurately assess the changes associated with
the progestin effect.
Secretory Pattern
The secretory pattern of progestin effect
mimics the glandular and stromal changes seen
in the luteal (secretory) phase of the menstrual
cycle. This pattern is closely related to the preg-nancy-like changes described in the preceding,
but neither the glands nor the stroma show such
an exaggerated response. With the secretory
pattern, the glands are tortuous and the glan-dular cells have basally oriented nuclei (Figs.
6.4 and 6.5). These low columnar cells typically
have a small amount of pale-staining supranu-clear cytoplasm and may show small, randomly
distributed vacuoles. The apical border often
Figure 6.3. Progestin effect, decidua pattern. The
glands show marked secretory changes and the
stroma is transformed into decidua-like cells. A fibrin
thrombus fills a dilated venule in the left lower
corner. Thrombi such as this result in bleeding that
frequently leads to biopsy.
Figure 6.4. Progestin effect, secretory pattern. The
endometrium has some resemblance to secretory
endometrium of the normal luteal phase with tor-tuous glands and abundant, predecidualized stroma.
Figure 6.5. Progestin effect, secretory pattern. High-magnification view of pattern resembling secretory
phase endometrium shows a tortuous gland with
The amount of stroma is increased relative to the
normal secretory phase and lacks edema. The glands,
while slightly tortuous, are markedly underdevel-oped relative to glands in the normal luteal phase.
secretions. The stromal cells are plump, having a
moderate amount of cytoplasm.
becomes smooth and well defined, unlike the
ragged luminal border in normal secretory
endometrium.13
The lumens may have a small
amount of dense, eosinophilic secretions. The
stromal cells show weak predecidual change
as they gain cytoplasm and become mildly
enlarged. These predecidualized cells are ovoid
with identifiable pale cytoplasm but are not as
large or polygonal as fully decidualized stromal
cells. Although the glandular and stromal
changes superficially resemble the secretory
phase endometrium of a menstrual cycle,
neither the glands nor the stroma are appro-priately developed for any day of the normal
cycle (Fig. 6.6). Usually the glands appear
to be underdeveloped, lacking tortuosity.
Stromal predecidual change tends to be con-fluent, lacking the intermittent edema that
characterizes most of the normal secretory
phase. Scattered mitotic figures can be found
in the stroma. As in other progestin-related pat-terns, the superficial stroma contains ectatic
venules.
Atrophic Pattern
The atrophic pattern represents the other end
of the spectrum of progestin effect, in which
the endometrium is hypoplastic. This pattern
evolves following prolonged progestin therapy
or with continued use of contraceptive hor-mones. The glands atrophy, although they con-tinue to show weak secretory changes. The
glands lose their tortuosity and are small and
tubular with scant to absent luminal secretions
(Fig. 6.7). The epithelium is low columnar with
basal nuclei and a small amount of pale cyto-plasm.
Figure 6.6. Progestin effect, secretory pattern.
Underdeveloped secretory glands in abundant
stroma that do not show decidual change. Several
ectatic venules are present in the superficial stroma.
Neither the glands nor the stroma are appropriately
developed for a normal postovulatory secretory
phase. This patient was taking oral contraceptives.
Progestins and Oral Contraceptives 127
When the progestin dose is low, the stromal
cells remain mildly enlarged with a small
amount of discernible cytoplasm but lose their
decidua-like appearance. Instead, they are
plump and ovoid with only a moderate amount
of cytoplasm. Cell borders become indistinct,
and stromal mitoses are not found. Vascular
channels beneath the surface epithelium
become ectatic. In contrast to the physiologic
atrophy pattern of the postmenopausal
endometrium, progestin-induced atrophy often
has more abundant stroma while the glands
become tiny and indistinct.
Other Stromal Changes
Whereas the glands atrophy with prolonged
progestin effect, with high doses of progestin
the stroma retains a decidua-like appearance. In
such cases the stroma can show alterations that
can be confusing or alarming. For instance, the
stroma can appear hyperplastic and pseudosar-comatous with increased cellularity as well as
nuclear hyperchromasia, enlarged nucleoli, and
variation in cell and nuclear size.13;18;19
A pseu-dosarcomatous change is rare with modern
progestin therapy, however, and is infrequent in
our experience. The stroma can show other
peculiar alterations. One change occasionally
seen is clustering of groups of enlarged stromal
cells with intervening areas of myxoid change
or edema (Fig. 6.8). This change can impart an
epithelioid appearance to some of the decidu-alized cells, especially when the cells are
enlarged with prominent cell borders. The
decidualized stromal cells can develop other
epithelioid features such as eccentric nuclei and
vacuolated cytoplasm (Fig. 6.9).20;21
In such
cases the decidualized stroma can mimic signet-ring cells of metastatic carcinoma.21
128 6. Effects of Hormones
Figure 6.7. Progestin effect, atrophic pattern. Small
atrophic glands in spindle-cell stroma show scant
secretory changes. The abundant stroma composed
of plump cells with a moderate amount of cytoplasm
distinguishes this pattern from atrophy due to lack
of estrogen. This pattern is commonly seen in women
on continuous oral contraceptives.
Figure 6.8. Progestin effect, atrophic pattern. An
aggregate of tightly packed, plump stromal cells in
an edematous, myxoid background. Clustering of
stromal cells such as this is occasionally seen with
progestin effect. The two glands present are atrophic.
The patient was on long-term oral contraceptives.
Figure 6.9. Progestin effect, decidual pattern. The
decidual cells have vacuolated cytoplasm and eccen-tric nuclei resulting in a signet-ring cell appearance
(inset). The cells are clustered with intervening
myxoid zones. A markedly atrophic gland is present
to the left of center and a dilated venule is seen to
the right. The patient was on high-dose megestrol
acetate for breast cancer.
In some cases of progestin effect, infiltrates of
lymphocytes or neutrophils yield patterns that
can suggest endometritis. For instance, with pro-longed progestin effect the stroma often con-tains a moderate infiltrate of stromal granular
lymphocytes and mononuclear cells (Fig. 6.10).
These are the normal lymphoid cells of
the endometrium that appear exaggerated
owing to the relative atrophy of the other
components. This striking infiltrate can mimic
chronic inflammation. An absence of plasma
cells and no evidence of gland infiltration by
inflammatory cells are helpful features to sepa-rate this progestin effect from true inflammation
(see Chapter 7). Also, especially in pregnancy-like patterns of progestin effect, multiple small
foci of breakdown are accompanied by a neu-trophilic response. These neutrophils, however,
are a localized response to tissue necrosis and do
not represent an infectious process.
Often there is overlap between the various
patterns of progestin effect, which depend on
the duration of progestin use, the dose of the
progestin, and underlying endogenous estrogen
levels. In some cases different fields from the
same specimen show different patterns of
progestin effect that can range from decidual-ized stroma to a secretory or atrophic change.
Consequently, morphologic identification of
progestin effects requires recognition of the
spectrum of changes that may be found. Fur-thermore, some patients on progestins, espe-cially those on oral contraceptives, can even
show proliferative phase patterns when the
estrogen influence is present but the progestin
influence is temporarily decreased or absent.
Long-term use of oral contraceptives rarely
may result in permanent endometrial atrophy
after the agent is discontinued.22
130 6. Effects of Hormones
Figure 6.10. Progestin effect, decidual pattern. Left:
Abundant stroma containing a rich infiltrate of
stromal granular lymphocytes mimicking chronic
inflammation. No glands are present in this field.
Right: High magnification shows decidualized stroma
and numerous granular lymphocytes but no plasma
cells. A stromal mitosis (arrow) is present.
Combined Estrogen and Progestin as Replacement Therapy for Menopausal Women 131
Figure 6.11. Estrogen–progestin therapy. End-ometrium from a postmenopausal woman receiving
estrogen–progestin replacement therapy shows small
secretory glands with extensive subnuclear vacuoles.
This progestin effect resembles days 16–17 of the
secretory phase of the normal cycle.
Combined Estrogen and
Progestin as Replacement
Therapy for Menopausal Women
Because of the possible deleterious conse-quences of unopposed estrogen therapy on the
endometrium, estrogen replacement is nearly
always given with a progestin in perimenopausal
and postmenopausal patients with a uterus.
Combined estrogen–progestin hormonal re-placement can be given either sequentially or in
combination.23;24
Sequential medication uses
daily estrogen for the first 21 to 25 days of the
month and daily progestin added for the last 10
to 13 days. This regimen results in withdrawal
bleeding. The continuous regimen uses both
estrogen and progestin daily. In the continuous
regimen, breakthrough bleeding may occur dur-ing the first 6 months but then bleeding usually
stops. Patients receiving either the sequential or
the combined regimen may undergo biopsy as
part of the routine surveillance to ensure that no
neoplasm develops.
With the sequential estrogen–progestin
regimen, the endometrium often shows a
weakly proliferative pattern with small, tubular
glands in scant stroma.25
The epithelium can
have occasional mitotic figures. The pattern
is identical to the weakly proliferative phase
pattern seen in association with anovulatory
dysfunctional bleeding caused by estrogen
withdrawal (see Chapter 5). Sometimes the
tissue shows a superimposed progestin effect
with poorly developed secretory changes in the
glands (Fig. 6.11).4;26–28
This latter pattern of
secretory changes is especially likely to be seen
if the biopsy is taken during the period of prog-
estin administration. In these cases the glandu-lar cells show mild tortuosity, basal nuclei, some
cytoplasmic vacuoles, and scant luminal secre-tions. In biopsy material, focal glandular and
stromal breakdown also may be seen (Fig.
6.12). With the combination regimens, the
endometrium usually is atrophic,11;25;29–33
and
often an atrophic pattern is seen with sequen-tial therapy also.26
Secretory changes can be
seen, however, especially if higher doses of
estrogen and progestin are used.34
Occasionally
a patient receiving either the sequential or the
combination regimen may have a more signifi-cant lesion in the biopsy specimen.33;35;36
Polyps, hyperplasia, and carcinoma are lesions
that have been found in a few cases. In
general, however, estrogen–progestin replace-ment therapy controls endometrial prolifera-tion, and significant proliferative and neoplastic
lesions are less common than in women not
receiving this therapy.3;25;37–41
Progestin-Like Effects with No
Hormone Use
On rare occasions endometrial tissue will show
morphologic features of a progestin effect even
though there is clearly no history of exogenous
hormone use. These changes may be seen in
both premenopausal and postmenopausal
women, and their etiology is poorly under-stood. In premenopausal women these patterns
can be decidua-like42
or can resemble “pill
effect” changes, with hypoplastic secretory
glands and plump stromal cells (Fig. 6.13). It is
possible that this alteration is due either to
abnormal persistence of a functioning corpus
132 6. Effects of Hormones
Figure 6.12. Estrogen–progestin therapy. Frag-mented endometrial sample from a postmenopausal
woman on estrogen–progestin replacement therapy.
Small glands show secretory changes with vacuo-lated cytoplasm and basally oriented nuclei. Focal
glandular and stromal breakdown is seen in the left
upper corner.
luteum or to the so-called luteinized unrup-tured follicle. This latter entity, as the name
implies, occurs when a follicle develops, does
not rupture (ovulate), and persists with
luteinization of the granulosa and theca cells. If
progesterone is produced by the unruptured
follicle, then the result could be a progestin
effect from the endogenous source.
There have been a few examples of idio-pathic endometrial decidual reaction in post-menopausal women who are not taking
hormones (Fig. 6.14).20
These patients have
tended to present with abundant polypoid
tissue. The etiology of the change is not known,
but it may be the result of local mechanical
factors rather than a response to progesterone-like hormones. Mechanical stimulation, includ-ing biopsy, can cause increased decidual
changes in the progesterone-primed end-ometrium.9;43
Also, an intrauterine device
(IUD) may lead to an enhanced decidual reac-tion in the endometrium.9;44;45
Effects of Other Hormones
Tamoxifen
Tamoxifen is a nonsteroidal antiestrogen that is
widely used in the hormonal therapy of breast
carcinoma. This drug is a selective estrogen
receptor modulator (SERM) with its action
mediated through the estrogen receptor. The
effect of tamoxifen on the endometrium ap-pears to depend on the menopausal status and
the dose and duration of tamoxifen use.46–48
Current data suggest it can act as both an estro-gen antagonist and an agonist. Most normally
Effects of Other Hormones 133
Figure 6.13. Progestin-like effect with no hormone
use. Endometrium from premenopausal women
shows small, atrophic glands in abundant predecidu-alized stroma resembling progestin effect. This
patient was not taking hormones but did have a
small, palpable ovarian cyst that may have been a
source of endogenous progesterone production, such
as a persistent corpus luteum or a luteinized unrup-tured follicle.
cycling premenopausal patients taking tamox-ifen continue to have regular menstrual cycles,
but some develop amenorrhea. With continued
use, serum estrogen and progesterone levels
often are increased to two or three times the
normal levels. In postmenopausal women,
tamoxifen has estrogenic effects on vaginal
epithelium.
A large volume of literature has accumulated
on the effects of tamoxifen, and the results vary
from study to study. Several conclusions appear
to be emerging from the data. Both transvagi-nal ultrasound and endometrial biopsy have
been used to monitor patients on tamoxifen,
and in the absence of symptoms few pathologic
lesions are found.46;49
In asymptomatic post-menopausal patients on tamoxifen, atrophy is
the most common finding.50;51
Endometrial
abnormalities are more commonly found in
symptomatic patients,52
although some women
with no symptoms or ultrasound abnormalities
will be found to have a pathologic lesion.50;53
Patients receiving a progestin after initial
tamoxifen therapy can show decidual reaction
of the stroma.54;55
Both endometrial hyperplasia and carcinoma
occasionally occur in patients on tamoxifen,
and some studies suggest an increased inci-dence of both these disorders in patients re-ceiving tamoxifen.47;56–62
The relative risk of
developing endometrial carcinoma appears to
be within the same range as reported with
unopposed estrogen use.56
The apparent in-crease may be the result of increased rate of
detection of otherwise asymptomatic, “silent”
tumors, however. Some patients receiving
tamoxifen develop high-grade carcinoma,58;63–65
but in general those carcinomas associated with
tamoxifen use do not differ in grade from car-cinomas that occur in patients not receiving this
134 6. Effects of Hormones
Figure 6.14. Decidua-like stroma in an 82-year-old
patient. Endometrium shows decidua-like stroma,
dilated atrophic glands, and ectatic venules beneath
surface epithelium. The patient presented with post-menopausal bleeding and had not been on hor-mones. The specimen also showed fragments of a
large, benign polyp.
hormone.58;66
Most endometrial cancers are
low stage and well differentiated.67
Carcinosar-comas and other sarcomas including end-ometrial stromal sarcoma and adenosarcoma
also have been reported in patients on
tamoxifen.68–70
Endometrial polyps appear to be one of the
most common pathologic findings in patients
on tamoxifen.46;50;63;65;71–76
These patients are
postmenopausal and have received long-term
tamoxifen therapy for metastatic breast carci-noma. The polyps tend to be large and multi-ple,77
and they may be recurrent.78
The stroma
is variably edematous, myxoid, or fibrous. Often
these polyps show mildly hyperplastic changes
(Fig. 6.15) (see Chapter 8). Various types of
cytoplasmic change or metaplasia have been
described in the glands, especially mucinous
and clear cell change.74;77
Occasionally endome-trial polyps in patients receiving tamoxifen
show foci of secretory changes in the glands
with clear to vacuolated cytoplasm (Fig. 6.16).
The mechanism of the secretory effects is not
known. Some authors have found that the
polyps show glands polarized along the long
axis of the polyp with staghorn-shaped glands
and a cambium layer of stromal condensation
around glands.74
In at least one reported polyp,
the stroma also showed decidual changes that
could not be attributed to any exogenous prog-estin use.72
Some polyps, however, show
markedly hyperplastic glands, and others show
cystic glands with focal atypia.71;72
A few exam-ples of atypical hyperplasia and carcinoma
arising within these polyps also have been
reported.65;77;79;80
Metastatic breast carcinoma to
the polyps has been reported as well.81
With the
expanding role of tamoxifen in the treatment
and possible prevention of breast carcinoma,
more clinical information will be accumulated
regarding the effect of this drug on the
endometrium.
Effects of Other Hormones 135
Figure 6.15. Tamoxifen-related polyp. Portion of a
large polyp removed by curette in a postmenopausal
patient on tamoxifen for breast carcinoma. The
glands have irregular, staghorn shapes and showed
weak proliferative activity. The stroma varies from
fibrous to edematous.
Raloxifene
Raloxifene is another selective estrogen recep-tor modulator (SERM) and represents a newer
anti-estrogen used for therapy of breast cancer.
It may also be useful in preventing post-menopausal osteoporosis. This drug appears to
be a more pure estrogen antagonist, lacking the
weak estrogen agonist effects of tamoxifen.82;83
Although only a relatively small number of
patients have been followed for endometrial
side effects in comparison to those receiving
tamoxifen, there appears to be no increase in
endometrial pathologic lesions with ralox-ifene.84–87
Usually the endometrium is atrophic.
Clomiphene Citrate
Clomiphene citrate is another antiestrogen that
is used to induce ovulation in the treatment of
infertile patients who are anovulatory.1
It also
may be used to treat luteal phase defects
(LPDs). This hormone stimulates multiple
follicles to develop, and ovulation follows. It
is thought to act by competitively binding
to estrogen receptors in the hypothalamus,
causing increased levels of follicle-stimulating
hormone (FSH) and luteinizing hormone (LH)
that induce ovulation. Like tamoxifen,
clomiphene citrate has been found to have
estrogenic as well as antiestrogenic activity.
Morphologic effects of clomiphene on the
endometrium are difficult to assess. Biopsies
often are performed in the luteal phase of
clomiphene-induced cycles to assess the
endometrial development. Usually the pattern
is that of normally developing secretory phase
endometrium, which can be histologically
dated, and this has been our experience. Often
the histologic date correlates with the chrono-136 6. Effects of Hormones
Figure 6.16. Tamoxifen effect with secretory
change. Endometrial biopsy from woman with breast
cancer shows weak secretory effect manifested by
small subnuclear vacuoles. The patient had under-gone bilateral oophorectomy several years previ-ously. Subsequent curettage showed a polyp with
focal secretory changes.
logical postovulatory date, but sometimes the
endometrium shows a significant lag in devel-opment. It is postulated that clomiphene citrate
may cause LPDs (see Chapter 2).88–92
Some
investigators do not find a significant asso-ciation of LPDs with clomiphene citrate use,
however.93;94
One study of the morphologic effects of
clomiphene citrate on the endometrium sug-gested that the drug causes significant alter-ations in secretory phase development.95
Decreased gland tortuosity with scant secretions
in clomiphene-treated cases was described. The
gland-to-stroma ratio was reportedly decreased
relative to secretory endometrium in untreated
cases. In early secretory endometrium, the sub-nuclear vacuoles appeared to be larger and more
sharply defined, and later in the secretory phase,
the luminal secretions appeared to be hyalinized
and inspissated. The stroma also showed
decreased predecidual change compared to
untreated endometrium. This report suggests
that clomiphene may cause reproducible mor-phologic changes in secretory phase endo-metrium. Another study reported advanced
secretory activity in clomiphene citrate–
induced cycles.96
Other investigators have found
no changes in endometrial morphology
following clomiphene citrate administration.97;98
In our opinion, any morphologic changes
associated with clomiphene citrate are subtle
and difficult to appreciate by routine
microscopy. Further evaluation, including
double-blinded studies, is necessary to establish
whether or not clomiphene citrate yields
consistent abnormalities in gland and stromal
development.
Danazol
Danazol is structurally related to testosterone
and is a weak androgen.99
Its main metabolite,
ethisterone, is a weak progestin, however.100
This steroid is used for the treatment of
endometriosis.1
Because it suppresses endome-trial growth, it also may be used to treat men-orrhagia101;102
and endometrial hyperplasia.103;104
The few studies of the effects of danazol on
the endometrium show changes similar to those
found in progestin therapy (Fig. 6.17).100;105;106
Within a few months of use, the amount of
tissue is reduced.107
Glands show weak and
irregular secretory changes with mild tortuos-ity, basal nuclei, and some cytoplasmic vac-uolization. The stroma is hypercellular. With
prolonged therapy, the glands show atrophy
with scant to no secretory activity.99;106–108
Vascular ectasia also can occur.109
Occasional
patients will show some proliferative activity
with stromal and glandular mitoses.
Human Menopausal
Gonadotropins/Human
Chorionic Gonadotropin
Human menopausal gonadotropins (hMG or
menotropins) are extracted from the urine of
postmenopausal women and consist of FSH
and LH. They are used to induce ovulation in
the treatment of infertility due to anovulation,
such as polycystic ovarian disease. Human
chorionic gonadotropin (hCG) has structural
and biologic similarities to LH and is used to
simulate and improve on the midcycle LH
surge associated with ovulation. This hormone
is used in conjunction with hMG and also can
be used along with clomiphene citrate-induced
ovulation.
The effects of hMG and hCG on endometrial
morphology are difficult to define. Some studies
suggest that the main change caused by admin-istration of hMG/hCG is endometrial “inade-quacy” with retarded development of more
than 2 days when the histologic date is com-pared to the chronologic date.110–113
These
changes are similar to those commonly found
in LPD (see Chapter 2). In one study, 27%
of patients treated with hMG/hCG showed
inadequacy in development with out-of-phase
histologic dates.110
Other studies, however, have
reported “advanced” histology with more
highly developed secretory changes than
expected for the cycle day, gland-stromal
dyssynchrony, or normal glandular develop-ment.114;115
Despite the apparent discrepancies
in the findings between studies, there are no
specific morphologic alterations that can be
definitely correlated with the effects of these
hormones. For the pathologist interpreting
biopsies from women who receive these hor-Effects of Other Hormones 137
mones, accurate histologic dating is the most
important consideration.
Gonadotropin-Releasing
Hormone Agonists
Gonadotropin-releasing hormone (GnRH)
agonists include leuprolide acetate, buserelin
acetate, and goserelin acetate. They are also
referred to as “luteinizing hormone-releasing
hormone agonists.” These preparations are
used to suppress the endometrium prior to
resectoscopic ablation and to decrease the size
of leiomyomas before surgical removal.116
They
are also used to prevent spontaneous LH
surges before oocyte retrieval and to improve
follicular development for ovulation induction
during in vitro fertilization (IVF) and gamete
intrafallopian transfer (GIFT).1
In conjunc-tion with progestins, these compounds may
also have utility for contraception. When they
are used to suppress endometrial growth,
GnRH agonists cause marked atrophy of the
endometrium.109
When GnRH agonists are
used in conjunction with a progestin, the endo-metrium shows apparent secretory changes
consistent with progestin effect.117
Antiprogestin RU 486
The synthetic progestogenic steroid RU 486, or
mifepristone, has high affinity for progesterone
receptors in the endometrium, which causes its
antiprogesterone action.118
Its main use is for
spontaneous termination of early pregnancy,
although currently it is not available for this
purpose in the United States. The drug is used
in high dose to induce early abortion. This drug
also has been assessed for possible contracep-tive use at much lower doses, and preliminary
studies suggest that with longer term use it
retards the secretory phase.118–126
Long-term,
138 6. Effects of Hormones
Figure 6.17. Danazol effect. Small glands in abundant stroma resemble progestin effect. The patient was on
danazol therapy for endometriosis.
low-dose treatment inhibits ovulation and
menstruation.126
Thus, mefepristone results in
inhibition of glandular secretory activity with
degenerative and vascular changes. Several
reports have indicated that the glands may
become irregular in size and shape, with dilated
glands.127
The stromal response appears to be
variable, but some reports indicate that the
stroma remains dense with increased mitotic
activity.122;128;129
One case report of an adoles-cent female with morbid osteoporosis treated
with high-dose mifepristone for its antigluco-corticoid effect describes recurrent diffuse
simple hyperplasia of the endometrium.130
In
postmenopausal women receiving estrogen
alone, this drug has effects similar to those of
progesterone, suggesting it acts as a proges-terone agonist.118–120;131
Clinical Queries and Reporting
A wide variety of indications for hormone use
all influence the reasons for biopsy and the
clinical questions posed to the pathologist.
Most biopsies that show hormone effects are
performed for one of the following reasons: (1)
surveillance during postmenopausal hormone
replacement therapy, (2) evaluation of abnor-mal uterine bleeding related to hormone
therapy, (3) assessment of treatment of hyper-plasia, and (4) evaluation of the status of the
endometrium following hormone induction of
ovulation or endometrial growth in infertility
therapy. For each of these indications, there are
specific considerations in interpretation and
reporting.
Postmenopausal Hormone
Replacement
Surveillance biopsy of asymptomatic post-menopausal patients receiving estrogen or
estrogen–progestin replacement therapy usu-ally is an office-based procedure intended to
provide a small representative sample. The
primary concern is whether or not the
endometrium shows evidence of hyperplasia or
neoplasia that would require cessation or
change of the hormone therapy.The presence or
absence of these changes should be explicitly
stated in the report.This information can appear
either in the diagnosis or as a comment. In these
cases accurate reporting of the degree of prolif-erative activity, if any, also is important.
Abnormal Uterine Bleeding
There are several situations in which hormone
therapy is related to abnormal bleeding. Some-times patients on oral contraceptives or other
progestins experience “breakthrough bleed-ing,” which indicates bleeding of a noncyclical
type. This may occur as irregular bleeding
during the first few months of oral contracep-tive use or later after many months of oral con-traceptive use. Often the bleeding is caused by
atrophy and increased “fragility” of the tissue
with focal glandular and stromal breakdown.
In other situations, hormone therapy may be
used as primary therapy for abnormal bleeding,
especially dysfunctional bleeding (see Chapter
5). Usually these patients have anovulatory
cycles, and a progestin is given to suppress pro-liferation. If the bleeding is not controlled by
progestin, then biopsy is performed.
Other hormone therapies, too, may lead to
abnormal bleeding. For example, tamoxifen
therapy for breast cancer, or estrogen or estro-gen replacement therapy for menopausal symp-toms, may be associated with bleeding. Biopsy
in such cases is performed to rule out signifi-cant organic lesions. Also, megestrol acetate
(Megace) therapy for breast cancer may lead to
decidual transformation of the endometrium
with focal glandular and stromal breakdown. In
any of these cases in which hormone therapy is
associated with abnormal bleeding, the biopsy
is performed to assess the status of the
endometrium and to rule out possible underly-ing organic disorders such as polyps, inflam-mation, hyperplasia, or neoplasia. In addition,
the biopsy may be therapeutic, removing
the abnormal tissue that is bleeding. Most of
these biopsies show progestin effects that range
from decidual patterns to atrophic patterns.
When the patient has received unopposed
estrogen, however, the pattern usually is prolif-erative. A specific diagnosis should be ren-dered. If this is not possible, a descriptive
Clinical Queries and Reporting 139
diagnosis should be made. When glandular and
stromal breakdown also is present, this should
be reported. In all these cases it is important to
exclude other lesions such as hyperplasia or
carcinoma, so a comment regarding the absence
of these lesions is generally indicated.
Treatment of Hyperplasia
Hyperplasia, either with or without atypia, may
be managed by progestin therapy to suppress
gland growth and then re-biopsied to deter-mine the efficacy of the therapy. In these cases
it is important to note whether or not there
is evidence of continued persistent hyper-plasia, and whether or not atypia is found. In
progestin-treated hyperplasia, the underlying
pattern of gland growth and cellular differenti-ation often becomes distorted. Frequently the
tissue resembles that seen in early pregnancy.
When there is a well-developed decidua-like
response, it is difficult to fully assess the degree
of gland complexity, and the secretory changes
of the glands make evaluation of cytologic
atypia difficult or impossible. Generally, cyto-logic atypia and mitotic activity are suppressed
by progestin treatment. It is important to state
clearly that progestin effect is present and that
it may not be possible to assess fully the degree
of the glandular abnormality. An effort also
should be made to review the previous material
to assess the effect of the therapy on the
hyperplasia.
Infertility Therapy
When biopsy is done during hormone therapy
in the infertile patient, the clinical questions
depend on the hormone used and the indica-tion for biopsy. Often during treatment with a
drug such as clomiphene citrate the gynecolo-gist wishes to know whether secretory phase
development is normal and, if so, the precise
histologic date (see Chapter 2). When ovula-tion-inducing medication is used, one obvious
question is whether the endometrium shows
secretory changes that reflect ovulation or if the
pattern is proliferative. The gynecologist also is
concerned with the possibility of an LPD that
will be reflected in delayed maturation when
compared to the time of ovulation (see Chapter
2). Usually this determination requires only
accurate histologic dating of secretory
endometrium. However, on occasion abnor-malities in secretory phase development are
found, and these should be clearly noted. For
example, when a disparity between the glands
and stroma is seen, the pathologist should
clearly indicate that irregular maturation is
present. A comment regarding the approximate
histologic date for both the glands and the
stroma further clarifies the pathologic findings.
In addition to the specific situations for
hormone use given in the preceding, hormone
effects, especially progestin-related changes,
can be found in other biopsy specimens when
the hormone use is incidental. These cases
include biopsies performed during evaluation
of pelvic pain, pre-hysterectomy sampling, or
curettage at the time of tubal ligation. Further-more, progestin effects can persist for a variable
period of 1 or more months after the drug is
discontinued. In addition, some of these cases
may represent an endogenous progestin effect
with no exogenous hormone use. Consequently,
the finding of the histologic changes of appar-ent progestin effect is not negated by the
clinical history. When the biopsy clearly
shows a progestin effect, we recommend using
the term “progestin effect” as the diagnosis
with an accompanying brief description of the
changes.
In some cases the glands and stroma show
some changes that suggest progestin effect but
the features are not diagnostic. Usually these
cases show glands with poorly developed secre-tory changes and plump stromal cells that
appear to be partly decidualized. If the clinical
history does not establish progestin use and the
changes by themselves are not diagnostic of
progestin effect, then it is best to give a descrip-tive diagnosis. This diagnosis can reflect the fact
that irregular secretory changes are present and
suggest the possibility of progestin effect. A
descriptive diagnosis would serve to indicate
that the changes are benign and caused by an
imbalance in the amounts of sex steroid hor-mones, regardless of their source.
140 6. Effects of Hormones
References
(1) Speroff L, Glass RH, Kase NG. Clinical gyne-cologic endocrinology and infertility, 6th ed.
Baltimore: Lippincott Williams & Wilkins,
1999.
(2) Schiff I. Menopause. In: Kistner RW, ed.
Gynecology: Principles and practice. Chicago:
Year Book Medical Publishers; 1986: 565–582.
(3) Effects of hormone replacement therapy on
endometrial histology in postmenopausal
women. The Postmenopausal Estrogen/Prog-estin Interventions (PEPI) Trial. The Writing
Group for the PEPI Trial. JAMA 1996;
275:370–375.
(4) Ferenczy A, Gelfand MM. Endometrial
histology and bleeding patterns in post-menopausal women taking sequential, com-bined estradiol and dydrogesterone. Maturitas
1997; 26:219–226.
(5) Gomes AM, Baracat EC, Simoes MJ, Haidar
MA, Focchi GR, Evencio-Neto J, et al. Mor-phologic and morphometric aspects of the
endometrium of postmenopausal women
before and after cyclic oestrogen replacement
treatment. Eur J Obstet Gynecol Reprod Biol
1997; 74:79–82.
(6) Grady D, Rubin SM, Petitti DB, Fox CS, Black
D, Ettinger B, et al. Hormone therapy to
prevent disease and prolong life in post-menopausal women. Ann Intern Med 1992;
117:1016–1037.
(7) Whitehead MI, Fraser D. The effects of estro-gens and progestogens on the endometrium.
Obstet Gynecol Clin North Am 1987; 14:
299–320.
(8) Ober W. Effects of oral and intrauterine
administration of contraceptives on the
uterus. Hum Pathol 1977; 8:513–527.
(9) Dallenbach-Hellweg G. Histopathology of the
endometrium, 4th ed. New York: Springer-Verlag; 1987.
(10) Sherman ME, Mazur MT, Kurman RJ. Benign
diseases of the endometrium. In: Kurman RJ,
ed. Blaustein’s pathology of the female genital
tract. 5th
ed. New York: Springer-Verlag; 2002:
421–466.
(11) Deligdisch L. Effects of hormone therapy on
the endometrium. Mod Pathol 1993; 6:94–106.
(12) Hesla J, Kurman R, Rock J. Histologic effects
of oral contraceptives on the uterine corpus
and cervix. Sem Reprod Endocrinol 1989; 7:
213–219.
(13) Deligdisch L. Hormonal pathology of the
endometrium. Mod Pathol 2000; 13:285–294.
(14) Yuzpe AA, Thurlow HJ, Ramzy I, Leyshon JI.
Post coital contraception—A pilot study. J
Reprod Med 1974; 13:53–58.
(15) Raymond EG, Lovely LP, Chen-Mok M,
Seppala M, Kurman RJ, Lessey BA. Effect of
the Yuzpe regimen of emergency contracep-tion on markers of endometrial receptivity.
Hum Reprod 2000; 15:2351–2355.
(16) Swahn ML, Westlund P, Johannisson E,
Bygdeman M. Effect of post-coital contracep-tive methods on the endometrium and the
menstrual cycle. Acta Obstet Gynecol Scand
1996; 75:738–744.
(17) Miranda MC, Mazur MT. Endometrial squa-mous metaplasia. An unusual response to
progestin therapy of hyperplasia. Arch Pathol
Lab Med 1995; 119:458–460.
(18) Dockerty MB, Smith RA, Symmonds RE.
Pseudomalignant endometrial changes in-duced by administration of new synthetic
progestins. Staff Meetings of the Mayo Clinic
1959; 34:321–328.
(19) Cruz-Aquino M, Shenker L, Blaustein A.
Pseudosarcoma of the endometrium. Obstet
Gynecol 1967; 29:93–96.
(20) Clement PB, Scully RE. Idiopathic post-menopausal decidual reaction of the
endometrium. Int J Gynecol Pathol 1988;
7:152–161.
(21) Iezzoni JC, Mills SE. Nonneoplastic endome-trial signet-ring cells. Vacuolated decidual cells
and stromal histiocytes mimicking adenocar-cinoma. Am J Clin Pathol 2001; 115:249–255.
(22) Bernardini L, Araujo FE, Balmaceda JP. A
case of permanent endometrial hypotrophy
after long-term use of oral contraceptives.
Hum Reprod 1993; 8:543–546.
(23) Gelfand MM, Ferenczy A. A prospective 1-year study of estrogen and progestin in
postmenopausal women: Effects on the
endometrium. Obstet Gynecol 1989; 74:
398–402.
(24) Mishell DRJ. Menopause. In: Stenchever MA,
Droegemueller W, Herbst AL, Mishell DRJ,
eds. Comprehensive gynecology. 4th
ed. St.
Louis: Mosby: 2001: 1217–1258.
(25) Feeley KM, Wells M. Hormone replacement
therapy and the endometrium. J Clin Pathol
2001; 54:435–440.
(26) Byrjalsen I, Thormann L, Meinecke B, Riis BJ,
Christiansen C. Sequential estrogen and
References 141
progestogen therapy—Assessment of pro-gestational effects on the postmenopausal
endometrium. Obstet Gynecol 1992; 79: 523–
528.
(27) Carranza-Lira S, Martinez-Chequer JC, Santa
Rita MT, Ortiz dlP, Perez Y, Fernandez RL.
Endometrial changes according to hormone
replacement therapy schedule. Menopause
1998; 5:86–89.
(28) Habiba MA, Bell SC, Al Azzawi F. Endome-trial responses to hormone replacement
therapy: histological features compared with
those of late luteal phase endometrium. Hum
Reprod 1998; 13:1674–1682.
(29) Magos AL, Brewster E, Singh R, O’Dowd T,
Brincat M. Amenorrhea and endometrial
atrophy with continuous oral estrogens and
progestogen therapy in post-menopausal
women. Obstet Gynecol 1985; 65:496–499.
(30) Hillard TC, Siddle NC, Whitehead MI, Fraser
DI, Prysedavies J. Continuous combined
conjugated equine estrogen–progestogen
therapy—effects of medroxyprogesterone
acetate and norethindrone acetate on bleed-ing patterns and endometrial histologic diag-nosis. Am J Obstet Gynecol 1992; 167:1–7.
(31) Piegsa K, Calder A, Davis JA, McKay-Hart D,
Wells M, Bryden F. Endometrial status in post-menopausal women on long-term continuous
combined hormone replacement therapy
(Kliofem). A comparative study of endome-trial biopsy, outpatient hysteroscopy and
transvaginal ultrasound. Eur J Obstet Gynecol
Reprod Biol 1997; 72:175–180.
(32) Sturdee DW. Endometrial morphology and
bleeding patterns as a function of progestogen
supplementation. Int J Fertil Menopausal Stud
1996; 41:22–28.
(33) Udoff L, Langenberg P, Adashi EY. Combined
continuous hormone replacement therapy:
A critical review. Obstet Gynecol 1995; 86:
306–316.
(34) Moyer DL, Delignieres B, Driguez P, Pez JP.
Prevention of endometrial hyperplasia by
progesterone during long-term estradiol
replacement—influence of bleeding pattern
and secretory changes. Fertil Steril 1993; 59:
992–997.
(35) Bjarnason K, Cerin A, Lindgren R, Weber T.
Adverse endometrial effects during long cycle
hormone replacement therapy. Scandinavian
Long Cycle Study Group. Maturitas 1999;
32:161–170.
(36) Sturdee DW, Ulrich LG, Barlow DH, Wells M,
Campbell MJ, Vessey MP, et al. The endome-trial response to sequential and continuous
combined oestrogen–progestogen replace-ment therapy. Br J Obstet Gynaecol 2000;
107:1392–1400.
(37) Heikkinen J, Vaheri R, Kainulainen P,
Timonen U. Long-term continuous combined
hormone replacement therapy in the pre-vention of postmenopausal bone loss: a com-parison of high- and low-dose estrogen–
progestin regimens. Osteoporos Int 2000; 11:
929–937.
(38) Nand SL, Webster MA, Baber R, O’Connor V.
Bleeding pattern and endometrial changes
during continuous combined hormone
replacement therapy. The Ogen/Provera Study
Group. Obstet Gynecol 1998; 91(5 Pt
1):678–684.
(39) Oosterbaan HP, van Buuren AH, Schram JH,
van Kempen PJ, Ubachs JM, van Leusden HA,
et al. The effects of continuous combined
transdermal oestrogen–progestogen treat-ment on bleeding patterns and the
endometrium in postmenopausal women.
Maturitas 1995; 21:211–219.
(40) Varila E, Wahlstrom T, Rauramo I. A 5-year
follow-up study on the use of a levonorgestrel
intrauterine system in women receiving
hormone replacement therapy. Fertil Steril
2001; 76:969–973.
(41) Kurman RJ, Felix JC, Archer DF, Nanavati N,
Arce J, Moyer DL. Norethindrone acetate and
estradiol-induced endometrial hyperplasia.
Obstet Gynecol 2000; 96:373–379.
(42) Te Linde RW, Henriksen E. Decidualike
changes in the endometrium without preg-nancy. Am J Obstet Gynecol 1940;39:733–749.
(43) Dallenbach-Hellweg G, Hohagen F. On the
problem of premature decidualization
induced by the embryo transfer. Lab Investi
1985; 52:17A (Abstr).
(44) Tamada T, Okagaki T, Maruyama M,
Matsumoto S. Endometrial histology asso-ciated with an intrauterine contraceptive
device. Am J Obstet Gynecol 1967; 98:
811–817.
(45) Bonney WA, Glasser SR, Clewe TH, Noyes
RW, Cooper CL. Endometrial response to the
intrauterine device. Am J Obstet Gynecol
1966; 96:101–113.
(46) Vosse M, Renard F, Coibion M, Neven P,
Nogaret JM, Hertens D. Endometrial disor-ders in 406 breast cancer patients on tamox-ifen: the case for less intensive monitoring. Eur
J Obstet Gynecol Reprod Biol 2002;
101:58–63.
142 6. Effects of Hormones
(47) Goldstein SR. The effect of SERMs on the
endometrium. Ann NY Acad Sci 2001; 949:
237–242.
(48) Cohen I, Altaras MM, Shapira J, Tepper R,
Rosen DJD, Cordoba M, et al. Time-dependent effect of tamoxifen therapy on
endometrial pathology in asymptomatic
postmenopausal breast cancer patients. Int J
Gynecol Pathol 1996; 15:152–157.
(49) Barakat RR, Gilewski TA, Almadrones L,
Saigo PE, Venkatraman E, Hudis C, et
al. Effect of adjuvant tamoxifen on the
endometrium in women with breast cancer:
A prospective study using office endome-trial biopsy. J Clin Oncol 2000; 18:3459–
3463.
(50) Cohen I, Perel E, Flex D, Tepper R, Altaras
MM, Cordoba M, et al. Endometrial pathol-ogy in postmenopausal tamoxifen treatment:
Comparison between gynaecologically symp-tomatic and asymptomatic breast cancer
patients. J Clin Pathol 1999; 52:278–282.
(51) Marchesoni D, Driul L, Fabiani G, Di Loreto
C, Cataldi P, Mozzanega B. Endometrial his-tologic changes in post-menopausal breast
cancer patients using tamoxifen. Int J
Gynaecol Obstet 2001; 75:257–262.
(52) Seoud M, Shamseddine A, Khalil A, Salem Z,
Saghir N, Bikhazi K, et al. Tamoxifen and
endometrial pathologies: A prospective study.
Gynecol Oncol 1999; 75:15–19.
(53) Ozsener S, Ozaran A, Itil I, Dikmen Y.
Endometrial pathology of 104 post-menopausal breast cancer patients treated
with tamoxifen. Eur J Gynaecol Oncol 1998;
19:580–583.
(54) Cohen I, Figer A, Altaras MM, Tepper R,
Shapira J, Cordoba M, et al. Common
endometrial decidual reaction in post-menopausal breast cancer patients treated
with tamoxifen and progestogens. Int J
Gynecol Pathol 1996; 15:17–22.
(55) Cohen I, Shapira J, Altaras M, Cordoba M,
Rosen D, Beyth Y. Endometrial decidual
changes in a postmenopausal woman treated
with tamoxifen and megestrol acetate. Br J
Obstet Gynaecol 1992; 99:773–774.
(56) Andersson M, Storm HH, Mouridsen HT.
Incidence of new primary cancers after
adjuvant therapy and radiotherapy for early
breast cancer. J Natl Cancer Inst 1991; 83:
1013–1017.
(57) Segna RA, Dottino RR, Deligdisch L, Cohen
CJ. Tamoxifen and endometrial cancer. Mt
Sinai J Med 1992; 59:416–418.
(58) Magriples U, Naftolin F, Schwartz PE,
Carcangiu ML. High-grade endometrial car-cinoma in tamoxifen-treated breast cancer
patients. J Clin Oncol 1993; 11:485–490.
(59) Gusberg SB. Tamoxifen for breast cancer:
Associated endometrial cancer. Cancer 1990;
65:1463–1464.
(60) Fornander T, Cedermark B, Mattsson A,
Skoog L, Theve T, Askevgren J, et al. Adjuvant
tamoxifen in early breast cancer: Occurrence
of new primary cancers. Lancet 1989;
1(8636):117–120.
(61) Seoud MA-F, Johns J, Weed JCJ. Gynecologic
tumors in tamoxifen-treated women with
breast cancer. Obstet Gynecol 1993; 82:
165–169.
(62) Carcangiu ML. Uterine pathology in tamox-ifen-treated patients with breast cancer. Anat
Pathol 1997; 2:53–70.
(63) Dallenbach-Hellweg G, Schmidt D, Hellberg
P, Bourne T, Kreuzwieser E, Doren M, et al.
The endometrium in breast cancer patients on
tamoxifen. Arch Gynecol Obstet 2000; 263:
170–177.
(64) Silva EG, Tornos CS, Follen-Mitchell M.
Malignant neoplasms of the uterine corpus
in patients treated for breast carcinoma: The
effects of tamoxifen. Int J Gynecol Pathol
1994; 13:248–258.
(65) Deligdisch L, Kalir T, Cohen CJ, de Latour M,
Le Bouedec G, Penault-Llorca F. Endometrial
histopathology in 700 patients treated with
tamoxifen for breast cancer. Gynecol Oncol
2000; 78:181–186.
(66) Cohen I, Azaria R, Fishman A, Tepper R,
Shapira J, Beyth Y. Endometrial cancers in
postmenopausal breast cancer patients with
tamoxifen treatment. Int J Gynecol Pathol
1999; 18:304–309.
(67) Fisher B, Costantino JP, Wickerham DL,
Redmond CK, Kavanah M, Cronin WM, et al.
Tamoxifen for prevention of breast cancer:
Report of the National Surgical Adjuvant
Breast and Bowel Project P-1 Study. J Natl
Cancer Inst 1998; 90:1371–1388.
(68) Wickerham DL, Fisher B, Wolmark N, Bryant
J, Costantino J, Bernstein L, et al. Association
of tamoxifen and uterine sarcoma. J Clin
Oncol 2002; 20:2758–2760.
(69) Clement PB, Oliva E, Young RH. Mullerian
adenosarcoma of the uterine corpus associ-ated with tamoxifen therapy: A report of six
cases and a review of tamoxifen–associated
endometrial lesions. Int J Gynecol Pathol
1996; 15:222–229.
References 143
(70) McCluggage WG, Abdulkader M, Price JH,
Kelehan P, Hamilton S, Beattie J, et al. Uterine
carcinosarcomas in patients receiving tamox-ifen. A report of 19 cases. Int J Gynecol
Cancer 2000; 10:280–284.
(71) Nuovo MA, Nuovo GJ, McCaffrey FM, Levine
RU, Barron B, Winkler B. Endometrial
polyps in postmenopausal patients receiving
tamoxifen. Int J Gynecol Pathol 1989; 8:
125–131.
(72) Corley D, Rowe J, Curtis MT, Hogan WM,
Noumoff JS, Livolsi VA. Postmenopausal
bleeding from unusual endometrial polyps in
women on chronic tamoxifen therapy. Obstet
Gynecol 1992; 79:111–116.
(73) Lahti E, Blanco G, Kauppila A, Apajasarkki-nen M, Taskinen PJ, Laatikainen T. Endome-trial changes in postmenopausal breast cancer
patients receiving tamoxifen. Obstet Gynecol
1993; 81:660–664.
(74) Kennedy MM, Baigrie CF, Manek S. Tamox-ifen and the endometrium: Review of 102
cases and comparison with HRT-related and
non-HRT-related endometrial pathology. Int J
Gynecol Pathol 1999; 18:130–137.
(75) Cohen I, Azaria R, Bernheim J, Shapira J,
Beyth Y. Risk factors of endometrial polyps
resected from postmenopausal patients with
breast carcinoma treated with tamoxifen.
Cancer 2001; 92:1151–1155.
(76) Liedman R, Lindahl B, Andolf E, Willen R,
Ingvar C, Ranstam J. Disaccordance between
estimation of endometrial thickness as mea-sured by transvaginal ultrasound compared
with hysteroscopy and directed biopsy in
breast cancer patients treated with tamoxifen.
Anticancer Res 2000; 20:4889–4891.
(77) Schlesinger C, Kamoi S, Ascher SM, Kendell
M, Lage JM, Silverberg SG. Endometrial
polyps: A comparison study of patients receiv-ing tamoxifen with two control groups. Int J
Gynecol Pathol 1998; 17:302–311.
(78) Biron-Shental T, Tepper R, Fishman A,
Shapira J, Cohen I. Recurrent endometrial
polyps in postmenopausal breast cancer pa-tients on tamoxifen. Gynecol Oncol 2003;
90:382–386.
(79) Cohen I, Bernheim J, Azaria R, Tepper R,
Sharony R, Beyth Y. Malignant endometrial
polyps in postmenopausal breast cancer
tamoxifen-treated patients. Gynecol Oncol
1999; 75:136–141.
(80) McCluggage WG, Sumathi VP, McManus DT.
Uterine serous carcinoma and endometrial
intraepithelial carcinoma arising in endome-trial polyps: Report of 5 cases, including 2
associated with tamoxifen therapy. Hum
Pathol 2003; 34:939–943.
(81) Houghton JP, Ioffe OB, Silverberg SG,
McGrady B, McCluggage WG. Metastatic
breast lobular carcinoma involving tamoxifen-associated endometrial polyps: Report of two
cases and review of tamoxifen-associated
polypoid uterine lesions. Mod Pathol 2003;
16:395–398.
(82) Cummings SR, Eckert S, Krueger KA, Grady
D, Powles TJ, Cauley JA, et al. The effect
of raloxifene on risk of breast cancer in
postmenopausal women: Results from the
MORE randomized trial. Multiple Outcomes
of Raloxifene Evaluation. JAMA 1999; 281:
2189–2197.
(83) Goldstein SR. Update on raloxifene to pre-vent endometrial–breast cancer. Eur J Cancer
2000; 36 Suppl 4:S54–S56.
(84) Boss SM, Huster WJ, Neild JA, Glant MD,
Eisenhut CC, Draper MW. Effects of ralox-ifene hydrochloride on the endometrium
of postmenopausal women. Am J Obstet
Gynecol 1997; 177:1458–1464.
(85) Davies GC, Huster WJ, Shen W, Mitlak B,
Plouffe L, Jr, Shah A, et al. Endometrial
response to raloxifene compared with
placebo, cyclical hormone replacement
therapy, and unopposed estrogen in post-menopausal women. Menopause 1999; 6:
188–195.
(86) Delmas PD, Bjarnason NH, Mitlak BH,
Ravoux AC, Shah AS, Huster WJ, et al.
Effects of raloxifene on bone mineral
density, serum cholesterol concentrations, and
uterine endometrium in postmenopausal
women. N Engl J Med 1997; 337:1641–
1647.
(87) Goldstein SR, Scheele WH, Rajagopalan SK,
Wilkie JL, Walsh BW, Parsons AK. A 12-month comparative study of raloxifene, estro-gen, and placebo on the postmenopausal
endometrium. Obstet Gynecol 2000; 95:
95–103.
(88) Wentz AC. Endometrial biopsy in the evalua-tion of infertility. Fertil Steril 1980; 33:
121–124.
(89) Keenan JA, Herbert CM, Bush JR, Wentz AC.
Diagnosis and management of out-of-phase
endometrial biopsies among patients receiv-ing clomiphene citrate for ovulation induc-tion. Fertil Steril 1989; 51:964–967.
(90) Cook CL, Schroeder JA, Yussman MA,
Sanfilippo JS. Induction of luteal phase defect
144 6. Effects of Hormones
with clomiphene citrate. Am J Obstet Gynecol
1984; 149:613–616.
(91) Van Hall EV, Mastboom JL. Luteal phase
insufficiency in patients treated with
clomiphene. Am J Obstet Gynecol 1969;
103:165–171.
(92) Yeko TR, Bardawil WA, Nicosia SM, Dawood
MY, Maroulis GB. Histology of midluteal
corpus luteum and endometrium from
clomiphene citrate-induced cycles. Fertil Steril
1992; 57:28–32.
(93) Hecht BR, Bardawil WA, Khan-Dawood FS,
Dawood MY. Luteal insufficiency: correlation
between endometrial dating and integrated
progesterone output in clomiphene citrate-induced cycles. Am J Obstet Gynecol 1990;
163:1986–1991.
(94) Lamb EJ, Colliflower WW, Williams JW.
Endometrial histology and conception rates
after clomiphene citrate. Obstet Gynecol
1972; 39:389–396.
(95) Benda JA. Clomiphene’s effect on
endometrium in infertility. Int J Gynecol
Pathol 1992; 11:273–282.
(96) Birkenfeld A, Navot D, Levij IS, Laufer N,
Beier-Hellwig K, Goecke C, et al. Advanced
secretory changes in the proliferative human
endometrial epithelium following clomiphene
citrate treatment. Fertil Steril 1986;
45:462–468.
(97) Thatcher SS, Donachie KM, Glasier A, Hillier
SG, Baird DT. The effects of clomiphene
citrate on the histology of human
endometrium in regularly cycling women
undergoing in vitro fertilization. Fertil Steril
1988; 49:296–301.
(98) Li TC, Warren MA, Murphy C, Sargeant S,
Cooke ID. A prospective, randomised,
cross–over study comparing the effects of
clomiphene citrate and cyclofenil on endome-trial morphology in the luteal phase of normal,
fertile women. Br J Obstet Gynaecol 1992;
99:1008–1013.
(99) Floyd WS. Danazol: Endocrine and endome-trial effects. Int J Fertil 1980; 25:75–80.
(100) Fedele L, Marchini M, Bianchi S, Baglioni A,
Bocciolone L, Nava S. Endometrial patterns
during danazol and buserelin therapy for
endometriosis: Comparative structural and
ultrastructural study. Obstet Gynecol 1990;
76:79–84.
(101) Beaumont H, Augood C, Duckitt K, Lethaby
A. Danazol for heavy menstrual bleeding.
Cochrane Database Syst Rev 2002;(2):
CD001017.
(102) Robins JC. Therapies for the treatment of
abnormal uterine bleeding. Curr Womens
Hlth Rep 2001; 1:196–201.
(103) Soh E, Sato K. Clinical effects of danazol on
endometrial hyperplasia in menopausal and
postmenopausal women. Cancer 1990;
66:983–988.
(104) Tamaoka Y, Orikasa H, Sumi Y, Sakakura K,
Kamei K, Nagatani M, et al. Direct effect of
danazol on endometrial hyperplasia in adeno-myotic women: Treatment with danazol con-taining intrauterine device. Hum Cell 2000;
13:127–133.
(105) Wentz AC, Jones GS, Sapp KC, King TM.
Progestational activity of danazol in the
human female subject. Am J Obstet Gynecol
1976; 126:378–384.
(106) Marchini M, Fedele L, Bianchi S, Dinola G,
Nava S, Vercellini P. Endometrial patterns
during therapy with danazol or gestrinone for
endometriosis—structural and ultrastructural
study. Hum Pathol 1992; 23:51–56.
(107) Jeppsson S, Mellquist P, Rannevik G. Short-term effects of danazol on endometrial histol-ogy. Acta Obstet Gynecol Scand Suppl 1984;
123:41–44.
(108) Fedele L, Bianchi S, Marchini M, Di Nola G.
Histological impact of medical therapy—
clinical implications. Br J Obstet Gynaecol
1995; 102(Suppl 12):8–11.
(109) Brooks PG, Serden SP, Davos I. Hormonal
inhibition of the endometrium for resecto-scopic endometrial ablation. Am J Obstet
Gynecol 1991; 164:1601–1608.
(110) Reshef E, Segars JH, Hill GA, Pridham DD,
Yussman MA, Wentz AC. Endometrial
inadequacy after treatment with human
menopausal gonadotropin/human chorionic
gonadotropin. Fertil Steril 1990; 54:
1012–1016.
(111) Bonhoff A, Naether O, Johannisson E, Bohnet
HG. Morphometric characteristics of endome-trial biopsies after different types of ovarian
stimulation for infertility treatment. Fertil
Steril 1993; 59:560–566.
(112) Campbell BF, Phipps WR, Nagel TC. Endome-trial biopsies during treatment with sub-cutaneous pulsatile gonadotrophin releasing
hormone and luteal phase human chorionic
gonadotropins. Int J Fertil 1988; 33:329.
(113) Sterzik K, Dallenbach C, Schneider V, Sasse V,
Dallenbach-Hellweg G. In vitro fertilization:
The degree of endometrial insufficiency varies
with the type of ovarian stimulation. Fertil
Steril 1988; 50:457–462.
References 145
(114) Garcia JE, Acosta AA, Hsiu J-G, Jones HWJ.
Advanced endometrial maturation after
ovulation induction with human menopausal
gonadotropin/human chorionic gonadotropin
for in vitro fertilization. Fertil Steril 1984;
41:31–35.
(115) Sharma V, Whitehead M, Mason B, Pryse-Davies J, Ryder T, Dowsett M, et al. Influence
of superovulation on endometrial and embry-onic development. Fertil Steril 1990; 53:
822–829.
(116) Casper RF. Clinical uses of gonadotro-pin-releasing hormone analogues. Can Med
Assoc J 1991; 144:153–158.
(117) Lemay A, Jean C, Faure N. Endometrial his-tology during intermittent intranasal luteiniz-ing hormone-releasing hormone (LH–RH)
agonist sequentially combined with an oral
progestogen as an antiovulatory contraceptive
approach. Fertil Steril 1987; 48:775–782.
(118) Spitz IM, Bardin CW. Mifepristone (RU
486)—a modulator of progestin and glucocor-ticoid action. N Engl J Med 1993; 329:
404–412.
(119) Swahn ML, Bygdeman M, Cekan S, Xing S,
Masironi B, Johannisson E. The effect of RU
486 administered during the early luteal phase
on bleeding pattern, hormonal parameters
and endometrium. Hum Reprod 1990;
5:402–408.
(120) Li T-C, Dockery P, Thomas P, Rogers AW,
Lenton EA, Cooke ID. The effects of proges-terone receptor blockade in the luteal phase
of normal fertile women. Fertil Steril 1988;
50:732–742.
(121) Batista MC, Cartledge TP, Zellmer AW,
Merino MJ, Axiotis C, Loriaux DL, et al.
Delayed endometrial maturation induced
by daily administration of the antiprogestin
RU–486—A potential new contraceptive
strategy. Am J Obstet Gynecol 1992; 167:
60–65.
(122) Cameron ST, Critchley HO, Thong KJ,
Buckley CH, Williams AR, Baird DT. Effects
of daily low dose mifepristone on endometrial
maturation and proliferation. Hum Reprod
1996; 11:2518–2526.
(123) Heikinheimo O, Hsiu JG, Gordon K, Kim S,
Williams RF, Gibbons WE, et al. Endometrial
effects of RU486 in primates—antiprolifera-tive action despite signs of estrogen action and
increased cyclin-B expression. J Steroid
Biochem Mol Biol 1996; 59:179–190.
(124) Gemzell-Danielsson K, Westlund P, Johannis-son E, Swahn ML, Bygdeman M, Seppala M.
Effect of low weekly doses of mifepristone on
ovarian function and endometrial develop-ment. Hum Reprod 1996; 11:256–264.
(125) Jimenez-Moreno V, Billeter M, Liu CQ,
Gordon K, Mahony M. Local effects of
mifepristone on the nonhuman primate
endometrium. Fertil Steril 2000; 74:134–140.
(126) Baird DT, Brown A, Critchley HO, Williams
AR, Lin S, Cheng L. Effect of long–term treat-ment with low-dose mifepristone on the
endometrium. Hum Reprod 2003; 18:61–68.
(127) Croxatto HB, Kovacs L, Massai R, Resch BA,
Fuentealba B, Salvatierra AM, et al. Effects of
long-term low-dose mifepristone on repro-ductive function in women. Hum Reprod
1998; 13:793–798.
(128) Greb RR, Kiesel L, Selbmann AK, Wehrmann
M, Hodgen GD, Goodman AL, et al.
Disparate actions of mifepristone (RU 486)
on glands and stroma in the primate
endometrium. Hum Reprod 1999; 14:
198–206.
(129) Murphy AA, Kettel LM, Morales AJ, Roberts
V, Parmley T, Yen SS. Endometrial effects of
long-term low-dose administration of RU486.
Fertil Steril 1995; 63:761–766.
(130) Newfield RS, Spitz IM, Isacson C, New MI.
Long-term mifepristone (RU 486) therapy
resulting in massive benign endometrial
hyperplasia. Clin Endocrinol (Oxf) 2001; 54:
399–404.
(131) Koering MJ, Healy DL, Hodgen GD. Mor-phologic response of endometrium to a prog-esterone receptor antagonist, RU486, in
monkeys. Fertil Steril 1986; 45:280–287.
146 6. Effects of Hormones
endometritis.9
In our experience, endometritis is
an infrequent diagnosis, however, and the preva-lence of this disorder appears to vary greatly
depending on the practice setting.
Endometrial inflammation often is nonspe-cific and rarely has morphologic features that
indicate a definite etiology. The nonspecific
forms of endometritis traditionally have been
separated into chronic and acute forms,
depending on the type of inflammatory infil-trate; most are referred to as chronic nonspe-cific endometritis. The inflammatory infiltrate
often is a mixed acute and chronic inflamma-tory process, however, and neutrophils as well
as plasma cells and lymphocytes can be present.
Rigorous separation of the type of inflamma-tory process is less important than recogni-tion of the presence of inflammation. Acute
endometrial inflammation is relatively infre-quent except for puerperal-related infections,
and these latter cases rarely come to biopsy or
curettage.
Nonspecific Endometritis
Endometritis may be diffuse or focal and can
range from a subtle inflammatory infiltrate
to a pronounced inflammatory reaction.
Endometritis typically shows a pattern of a
mixed inflammatory infiltrate containing
plasma cells and lymphocytes, and, not in-frequently, neutrophils (polymorphonuclear
leukocytes) and eosinophils. In addition to
inflammatory cells, there is a constellation of
147
7
Endometritis
Endometritis usually is a disorder of the re-productive years, although it may occur in
postmenopausal patients. Endometrial inflam-mation typically accompanies pelvic inflamma-tory disease of the upper genital tract.1;2
It may
also be associated with a recent pregnancy,
either an abortion or a term pregnancy.3–5
Other possible causes include instrumentation,
such as a prior biopsy, an intrauterine contra-ceptive device, cervical stenosis, or the presence
of an organic lesion such as a polyp, leiomyoma,
hyperplasia, or carcinoma.5
Endometritis typi-cally presents with intermenstrual vaginal
bleeding, and sometimes it causes menorrhagia.
This disorder also may be associated with infer-tility,6;7
although some studies find no associa-tion between endometritis and decreased
fertility.8
In one study, 8% of outpatient
endometrial biopsies, most of which were
done for abnormal bleeding, showed chronic
Nonspecific Endometritis  . . . . . . . . . . . . . 147
Inflammatory Cells  . . . . . . . . . . . . . . . . 148
Stromal Changes  . . . . . . . . . . . . . . . . . . 151
Abnormal Glandular Development  . . . 152
Glandular and Stromal Breakdown . . . . 154
Specific Infections  . . . . . . . . . . . . . . . . . . . 154
Granulomatous Inflammation  . . . . . . . . 154
Actinomycosis  . . . . . . . . . . . . . . . . . . . . 155
Cytomegalovirus  . . . . . . . . . . . . . . . . . . 155
Herpesvirus  . . . . . . . . . . . . . . . . . . . . . . 156
Mycoplasma . . . . . . . . . . . . . . . . . . . . . . 156
Differential Diagnosis  . . . . . . . . . . . . . . . . 158
Clinical Queries and Reporting  . . . . . . . . 160
histologic findings that facilitate recognition
of endometrial inflammation (Table 7.1).9
The
other morphologic changes include reactive
stroma, epithelial changes, abnormal glandular
development, and evidence of glandular and
stromal breakdown.9;10
Inflammatory Cells
Plasma cells are the most important histologic
feature for the diagnosis of endometritis.4;5;11;12
Their presence is required to establish the
diagnosis of chronic endometritis, because, in
contrast to lymphocytes, they are not present in
normal endometrium. Plasma cells may be
diffuse and easily recognizable but more com-monly are focal and widely dispersed. Plasma
cells generally are most numerous in the
periglandular and subepithelial stroma and
around lymphoid aggregates (Fig. 7.1). Plasma
cells should be readily identifiable and numer-ous before a diagnosis of chronic endometritis
is established unless associated features of
inflammation are clearly present (see later).9
The diagnosis of endometritis should not rest
on the finding of an apparent plasma cell,
however, in endometrium that otherwise
appears normal. In cases in which the plasma
cell infiltrate appears subtle or equivocal, the
background pattern is as important as the
quantity of plasma cells for establishing the
diagnosis of endometritis. The number of
plasma cells does not appear to correlate with
the severity of the lesion.3;9
148 7. Endometritis
Table 7.1. Morphologic features of nonspecific
endometritis.
Plasma cell infiltrate
Increased number of lymphocytes and lymphoid follicles
Variable presence of neutrophils in surface epithelium
and glands
Reactive stromal response
Altered gland development
Breakdown and bleeding
Figure 7.1. Nonspecific chronic endometritis. Scat-tered plasma cells and lymphocytes surround a small
proliferative gland. The stromal cells are reactive,
with elongated, spindle-shaped nuclei, and they swirl
around the gland.
Normal endometrial stromal cells, especially
predecidualized cells in the late secretory
phase, can resemble plasma cells, having ec-centric nuclei and a pale perinuclear zone.
The plasma cell, however, is identified by its
distinctive, clumped chromatin arrangement
yielding a clock-face pattern. A methyl green
pyronin histochemical stain,4
immunohisto-chemistry for immunoglobulin G13
or synde-can,14
and in situ hybridization for kappa and
lambda immunoglobulin light chains15
can
help demonstrate plasma cells when the cyto-logic features are not diagnostic by routine
histology.
Whereas plasma cells may be the predomi-nant inflammatory component of endometritis,
more severe inflammation commonly shows
a mixed inflammatory infiltrate. Often the
inflammatory infiltrate includes numerous lym-phocytes that tend to concentrate in the
subepithelial stroma (Fig. 7.2). Lymphoid folli-cles become prominent (Fig. 7.3) and may show
germinal centers; larger transformed lympho-cytes and immunoblasts also may be inter-spersed (Fig. 7.4).16–18
Neutrophils as a part of the inflammatory
infiltrate indicate an acute process (Figs. 7.2 and
7.5). This neutrophilic inflammatory infiltrate
typically infiltrates the surface epithelium and
extends into gland lumen, sometimes forming
microabscesses in the glands.9–12
Neutrophils,
however, also can be present in menstrual
endometrium, where foci of glandular and
stromal breakdown are present, too, without
signifying an infectious process. Therefore, like
the presence of lymphocytes, the presence of
neutrophils alone is not sufficient to indicate
inflammation. The pattern of distribution of
these cells in the endometrium and the accom-panying cellular infiltrate, usually including
Nonspecific Endometritis 149
Figure 7.2. Nonspecific acute and chronic
endometritis. A dense inflammatory infiltrate com-posed of plasma cells and lymphocytes is present in
the stroma. A few neutrophils and lymphocytes infil-trate the subepithelial stroma and extend into the
surface epithelium. The stroma has a reactive
pattern, with spindle-shaped cells.
Figure 7.3. Nonspecific chronic endometritis. Left:
Proliferative glands surrounded by spindle-shaped
stromal cells.A lymphoid follicle is present. Right:The
lymphoid follicle is composed predominantly of lym-Figure 7.4. Nonspecific chronic endometritis. A mixed inflammatory infiltrate with scattered transformed
lymphocytes and immunoblasts. Transformed lymphocytes are often associated with chlamydia infection.
phocytes, but scattered plasma cells also are present.
The presence of plasma cells distinguishes inflamma-tory lymphoid follicles from normal lymphoid folli-cles that occur in noninflamed endometrium.
plasma cells, must be considered before making
the diagnosis of endometritis.
Acute endometritis without a chronic
(plasma cell) component is extremely unusual
and occurs most frequently in the postpartum
or postabortal patient. Patients with pregnancy-related acute inflammation rarely come to
biopsy or curettage, however. When acute
endometritis is present, there is a neutrophilic
infiltrate in the glands with microabscess for-mation and infiltration of neutrophils into the
surface epithelium. Marked inflammation also
will result in formation of granulation tissue
with a network of small vessels in a fibroblastic
stroma (Fig. 7.6).
On occasion eosinophils may be present as a
part of the inflammatory infiltrate.9
They are
not normally present in the endometrium. Like
lymphocytes or neutrophils, eosinophils should
be present in a background of inflammatory
changes to be a component of endometritis.
Eosinophilic infiltrates also can occur following
curettage, apparently as a result of the instru-mentation,19
and in this case they represent a
nonspecific response to the procedure.
Endometritis also can have a component
of histiocytes. Usually these cells are widely
distributed in a mixed inflammatory infiltrate.
Hemosiderin-laden stromal cells and histio-cytes often are interspersed.3;5
Sometimes
histiocytes can be prominent, with large
aggregates of these cells in the stroma sur-rounded by plasma cells and lymphocytes.
When the histiocytes develop abundant,
foamy cytoplasm, the process becomes
xanthogranulomatous.12;20–23
Stromal Changes
With endometritis the stroma typically shows
reactive changes.3;9;12
Stromal cells become
spindle-shaped, resembling fibroblasts, and
are elongate and bipolar, in contrast to the
rounded, ovoid shape of the nonreactive
Nonspecific Endometritis 151
Figure 7.5. Nonspecific acute and chronic endometritis. Neutrophils infiltrate the surface epithelium, and
the deeper stroma contains numerous lymphocytes and plasma cells.
stromal cell (see Figs. 7.1 and 7.3). The reactive
process is also characterized by a swirling, inter-lacing pattern of the spindle cells, which may
form radial, “pinwheel” arrangements (Fig. 7.7).
Plasma cells usually are interspersed in the
reactive stroma. Superficial stroma may be-come edematous.9
Abnormal Glandular Development
In cycling patients the endometrial response
to hormones is often diminished. Usually the
endometrium has proliferative phase charac-teristics, with tubular glands showing mitotic
activity. In the secretory phase the glands may
lose their normal pattern of reactivity. Secre-tory changes occur in ovulating women, but
they often show abnormal development with
less gland tortuosity and distension than is seen
in a normal, noninflamed secretory phase. The
changes can include irregular or retarded
maturation of secretory phase endometrium.
Glands may appear underdeveloped, lacking
tortuosity and lumenal secretions. Plasma cells
may rarely be seen in histologically normal
secretory phase endometrium, however.
Epithelial Changes
Reactive cellular changes also affect the
endometrial surface and glandular epithelium.
The epithelium may show squamous and
eosinophilic cell change (see Chapter 9), espe-cially when the inflammation is long standing
and intense.9;10
The reactive epithelial cells
may become stratified, with prominent nucleoli,
cleared chromatin, and increased mitotic
activity (Fig. 7.8).
152 7. Endometritis
Figure 7.6. Acute endometritis with granulation
tissue. Normal endometrium is replaced by granula-tion tissue composed of numerous capillaries in a
dense background of inflammatory cells. The surface
is ulcerated. Inset: The inflammation is characterized
by numerous plasma cells, neutrophils, and lympho-cytes.
Figure 7.7. Nonspecific endometritis with reactive
stroma. Chronically inflamed endometrium shows
reactive stroma with interlacing, elongate spindle
Figure 7.8. Nonspecific endometritis with reactive
epithelial changes. The reactive glandular cells have
abundant, eosinophilic cytoplasm and enlarged
cells that resemble fibroblasts. Plasma cells and lym-phocytes are interspersed.
nuclei. A dense chronic inflammatory infiltrate with
many plasma cells surrounds the glands.
Glandular and Stromal Breakdown
Endometritis also results in focal glandular and
stromal breakdown. With severe and prolonged
chronic inflammation, the changes of irregular
bleeding with breakdown and regeneration
become prominent. This pattern of haphazard
bleeding leads to a corrugated surface with foci
of regenerating and shedding endometrium
interspersed. The inflamed stroma becomes
dense and less responsive to hormonal changes.
Because of the irregular growth, the tissue may
become polypoid, and the resemblance to a
polyp is accentuated by the dense stroma that
accompanies the inflammation. A mixture of
acute and chronic bleeding patterns also may
be present, with areas of stromal collapse,
glandular breakdown, stromal fibrosis,
macrophages, and hemosiderin deposition (see
Chapter 5).
Specific Infections
Although the etiology of chronic endometritis
usually is not apparent in biopsy specimens,
some cases show morphologic features that
offer clues to the etiology. For example, the
endometrium may harbor other changes, such
as a retained subinvoluted implantation site
following a pregnancy or abortion, a placental
site nodule from a remote pregnancy, or other
lesions, such as a polyp.
The inflammatory response associated with
Chlamydia trachomatis infection is usually
marked. The inflammatory infiltrate tends to
be diffuse, with plasma cells, lymphocytes,
and lymphoid follicles with transformed lym-phocytes.1;10;24–26
The inflammatory response to
chlamydia also may be mixed, with an infiltrate
of acute as well as chronic inflammatory cells.
Stromal necrosis and reactive atypia of the
epithelium also can be present.10
These marked
inflammatory changes are not specific for
chlamydia, however, but appear to reflect the
presence of upper genital tract infection and
acute salpingitis.1
One study showed that neu-trophils in the endometrial surface epithelium
and in gland lumens, along with a dense subep-ithelial stromal lymphocytic infiltrate, plasma
cells, and germinal centers containing trans-formed lymphocytes, are features that are pre-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
stroma and five or more neutrophils per  ¥400
field in surface epithelium was strongly asso-ciated with upper genital tract infection and
salpingitis.1
Chlamydia trachomatis or Neisseria
gonorrheae are most frequently associated with
these findings of a marked acute and chronic
inflammatory infiltrate, although infections
with C. trachomatis produce a greater concen-tration of plasma cells and more lymphoid
follicles than  N. gonorrheae.1
In cases with
plasma cell endometritis, the presence of  C.
trachomatis infection can be established by
using plasmid-based polymerase chain reaction
(PCR) or immunohistochemistry of paraffin-embedded sections.25;26
Granulomatous Inflammation
Granulomatous inflammation of the endo-metrium is infrequent. Often the process is
caused by mycobacterium, especially Mycobac-terium tuberculosis, and the infection usually
indicates advanced disease.Although rare in the
United States, in some countries tuberculous
endometritis is more commonly encountered in
endometrial biopsies undertaken during assess-ment of primary or secondary female infertility,
as endometrial involvement is a reflection of
more widespread disease that also affects the
fallopian tubes in most cases.27;28
Tuberculous
endometritis also can cause abnormal uterine
bleeding in postmenopausal patients.29
In tuber-culous infection the granulomatous response
is variable. Often the granulomas are non-necrotizing.28
Well-formed granulomas may be
difficult to identify unless the endometrium
is biopsied in the late secretory phase when
the granulomas have had sufficient time to
develop.5;11;30
The surrounding stroma can show
a lymphocytic infiltrate. As with any form of
inflammation, gland development may be
altered, lacking an appropriate secretory
response if the biopsy is taken in the luteal
phase.30
Acid-fast stains rarely demonstrate the
characteristic organism in endometrial infec-154 7. Endometritis
tions, and culture of fresh tissue or PCR of
paraffin-embedded tissue may be needed to
establish the diagnosis.28
Fungal infections,
including cryptococcosis, coccidioidomycosis,
and blastomycosis, rarely involve the endome-trium, resulting in granulomatous inflamma-tion.4
Cytomegalovirus infection has been seen
in association with poorly formed endometrial
granulomas.31
Sarcoidosis, too, may rarely lead
to non-necrotizing granuloma formation in
the endometrium.32
Necrotizing granulomatous
inflammation has been seen following endome-trial hysteroscopic ablation therapy.33–35
A
foreign body reaction often is present in addi-tion to the granulomas (Fig. 7.9).35
Actinomycosis
Infection by Actinomyces israelii is another rare
cause of endometritis. This organism typically is
found in endometritis associated with use of the
intrauterine device (IUD). Use of the IUD has
declined in the United States, so actinomycotic
endometritis is also infrequent. When actino-mycosis-associated inflammation is present,
the inflammatory response usually is intense,
with many plasma cells, lymphocytes, and neu-trophils present throughout the tissue. The
organisms show the typical sulfur granule
morphology and can be stained by Gram and
methenamine-silver stains.12
Cytomegalovirus
Rarely, endometrial biopsy will show evidence
of cytomegalovirus infection. This may occur in
immunosuppressed patients or it may be found
in women with no known underlying disor-der.36–39
Regardless of immunologic status, the
tissue shows the characteristic nuclear and
cytoplasmic inclusions in epithelial cells and
occasional endothelial cells (Fig. 7.10). The
stroma may show a sparse plasma cell infiltrate,
but other changes associated with inflamma-Specific Infections 155
Figure 7.9. Foreign body granulomatous response
following endometrial ablation therapy. Chronic
inflammation with foreign body giant cells surround
an area of necrosis.The necrotic center contains amor-phous debris and carbon deposits that accumulated
following thermal ablation of benign endometrium.
tion, such as a spindle-cell reactive stroma, may
not be found. One reported case had small,
ill-defined, non-necrotizing granulomas in
the endometrium but no visible inclusions,
although the presence of the virus was demon-strated by PCR for viral DNA.31
Herpesvirus
Herpesvirus rarely infects the endometrium,
but it may occur, usually as an ascending
process associated with cervical infection.40–43
When present in the endometrium, it can cause
patchy necrosis of the glands and stroma (Fig.
7.11).43
The diagnosis is established by identify-ing cells that show typical herpesvirus cyto-pathic effect. Cowdry type A inclusion and
multinucleate cells with molded ground-glass
nuclei can be found in the glandular epithelium
or the stroma in areas of necrosis (Fig. 7.12).
Several nonviral alterations, including optically
clear nuclei associated with the presence of tro-phoblast44
and cytoplasmic nuclear invagina-tions in the Arias-Stella reaction45
(see Chapter
3), may superficially resemble the herpesvirus
effect. Immunohistochemical stains for her-pesvirus antigens can be helpful in document-ing the presence of the virus.
Mycoplasma
The morphologic changes associated with myco-plasma, especially  Ureaplasma urealyticum,
have been described.46;47
The inflammatory
pattern, termed “subacute focal inflammation,”
is subtle but distinctive. In this condition the
inflammatory infiltrate is patchy, focal, and
difficult to discern. It is best seen in the mid-secretory phase from days 20 to 23 when stromal
edema accentuates the inflammatory foci
156 7. Endometritis
Figure 7.10. Cytomegalovirus endometritis. Pos-tabortal curettage specimen shows cytomegalovirus
endometritis. A single glandular cell contains a
prominent dark nuclear inclusion and granular
cytoplasmic inclusions.
Figure 7.11. Herpesvirus endometritis. Area of
necrosis in secretory endometrium due to her-pesvirus infection. Multinucleate cells showing viral
Figure 7.12. Herpesvirus endometritis. The nuclei of infected cells have a ground-glass appearance and
contain inclusions of herpesvirus.
cytopathic effect are present. The patient was not
immunocompromised, but herpesvirus was also
present in the endocervix.
(Fig. 7.13). The areas of inflammation consist
mainly of lymphocytes with macrophages and
only rare plasma cells or neutrophils.46–48
These
small inflammatory foci tend to be located
beneath surface epithelium, around spiral arte-rioles, or adjacent to glands.46
Chronic lesions
may appear granulomatous.46
Biopsies that are
not timed for the edematous portion of the
secretory phase may miss these abnormalities
because the inflammatory infiltrate cannot be
distinguished from normal lymphoid tissue of
the endometrium. Subacute focal inflammation
also has been linked with the presence of pelvic
adhesions.47
Differential Diagnosis
One of the most difficult problems in the dif-ferential diagnosis of endometritis is to decide
whether apparent inflammatory cells represent
true inflammation or whether they are a part of
the normal cellular infiltrate of the endo-metrium. Normal stromal cells can resemble
plasma cells, having the same size and an ec-centric nucleus. The cytologic features of the
nuclei distinguish stromal cells from plasma
cells, as the latter have a characteristic clock-face chromatin pattern.
Normal lymphoid aggregates and stromal
granular lymphocytes can be especially difficult
to distinguish from an inflammatory infiltrate,
especially if the tissue is poorly preserved.
Normal lymphoid aggregates, however, typi-cally are widely spaced, located near the basalis,
and do not include plasma cells. Stromal gran-ular lymphocytes are uniformly distributed
throughout the stroma and are most prominent
in the late secretory phase. These cells have
dark, irregular nuclei that often appear bilobed;
their cytoplasm is faintly granular. Stromal
granular lymphocytes normally occur in tissue
158 7. Endometritis
Figure 7.13. Subacute focal inflammation. Patchy
chronic inflammatory infiltrate adjacent to glands
and vessels in mid secretory endometrium. Plasma
cells are sparse. This pattern is associated with
mycoplasma infection.
that lacks other features of inflammation,
including plasma cells, reactive stroma, and
glands that appear out of phase. Granular
lymphocytes become especially prominent in
decidualized gestational endometrium or in
endometrium that shows a progestin effect,
especially when the stroma shows a decidua-like reaction (see Chapter 6). In these cases
the infiltrate can be so marked that at casual
inspection it resembles the lymphoid response
seen with endometritis. In such cases the
absence of plasma cells is especially helpful in
distinguishing this pattern from a true inflam-matory response. The presence of decidualized
stromal cells, usually containing atrophic glands
showing faint secretory activity, indicates that
the process is an effect of progesterone or a
synthetic progestin.
Neutrophils, like lymphocytes, can be present
without indicating an infectious process. They
normally occur in areas of stromal necrosis
associated with bleeding and in necrotic tissues
such as decidua or degenerating polyps. Men-strual endometrium also shows neutrophilic
infiltrates as part of the physiologic tissue
breakdown. These neutrophilic infiltrates sec-ondary to necrosis and breakdown do not
represent infection. They are recognized and
separated from infection-related inflammation
by their lack of infiltration into the epithelium
as well as their association with glandular and
stromal breakdown. On occasion neutrophils
also may be found only in gland lumens but not
infiltrating the epithelium. This phenomenon
apparently is caused by entrapment of cellular
debris from a previous cycle and is a finding of
no known significance.
Inflamed endocervix may also be sampled
during endometrial biopsy or curettage. This
tissue is a contaminant that has no relevance as
long as the endometrium itself is free of an
inflammatory infiltrate. Foci of nonspecific cer-vical inflammation usually are a minor com-ponent of the tissue in endometrial samples.
Endocervical epithelium that shows squamous
metaplasia or microglandular hyperplasia often
is present and helps to identify these foci (see
Chapter 2).
Another possible contaminant in biopsy
specimens consists of aggregates of free-float-ing histiocytes that do not infiltrate endometrial
glands or stroma. These histiocytes are a
response to extracellular mucin in the endocer-vical canal or cellular debris in the cavity (see
Chapter 2). Large sheets of histiocytes may be
found in curettings following cervical stenosis
with obstruction of the os. In the absence of
an inflammatory infiltrate in the endometrial
stroma that includes plasma cells, these free-floating histiocytes do not indicate inflam-mation.49
Immunostains for histiocytes, such
as lysozyme or KP-1, can be very useful for
demonstrating these cells.
Pseudosulfur granules, also known as
“pseudoactinomycotic radiate granules,” occa-sionally occur in the lower female genital tract,
especially in association with an IUD.50;51
These
radiate structures mimic the appearance of true
actinomycotic organisms but actually represent
an unusual response (Splendore–Hoeppli
phenomenon) to foreign bodies or bacteria
(Fig. 7.14). Pseudosulfur granules do not stain
with tissue Gram stains or with methenamine-silver, whereas true actinomyces do. These
peculiar structures are not associated with
other endometrial abnormalities, including
endometritis, and should not be mistaken for
actinomycetes.
Inflamed glands may show reactive changes,
with nuclear enlargement and prominent nucle-oli, and the cytologic features may suggest
hyperplasia or neoplasia (see Fig. 7.8). In addi-tion, the spindle-shaped, reactive stromal cells
of endometritis may be difficult to distinguish
from the fibrous stroma of a polyp or from the
desmoplasia of carcinoma. The architecture of
tubular and uniform glands is usually preserved
during inflammation, and it is important to
evaluate glands in areas that do not show frag-mentation or breakdown. Other underlying
abnormalities, such as hyperplasia and carci-noma, may become secondarily inflamed, espe-cially if the lesion or associated bleeding results
in loss of the cervical mucus barrier or dilation
of the internal os with secondary infection.
These lesions should retain the typical mor-phology of the glands and stroma, which would
allow their diagnosis in the absence of an
inflammatory response. There are no data to
suggest that chronic inflammation has a signif-Differential Diagnosis 159
icant relationship to the genesis of either hyper-plasia or carcinoma of the endometrium. Polyps
should be localized abnormalities with at least
a partial lining of surface epithelium, and these
generally do not contain plasma cells unless
they are secondarily inflamed as a result of
extension into the endocervix.
Severe chronic endometritis can produce an
intense lymphoid infiltrate with large lymphoid
cell immunoblasts that can resemble signs of
malignant lymphoma or a leukemic infil-trate.16–18
Usually with severe inflammation, the
cellular infiltrate is mixed, with a combination
of plasma cells, neutrophils, and lymphoid cells
with follicle formation, whereas lymphoma or
leukemic infiltrates usually are composed of a
relatively monotonous cell population. Involve-ment of the endometrium by malignant lym-phoma is rare in the absence of disseminated
disease. The most common hematologic malig-nancy to involve the endometrium is non-Hodgkin lymphoma, usually the diffuse large-cell type, but this is a rare finding in biopsy
specimens (see Chapter 11).
Clinical Queries and Reporting
Accurate diagnosis of endometritis is impor-tant, as the presence of inflammation can
establish a cause for abnormal bleeding or
unexplained infertility. Most cases of
endometritis have no specific etiology that can
be determined by histologic study of the biopsy
specimens. When endometritis is present,
however, a statement to indicate whether the
sections also show a demonstrable cause of the
inflammation, such as evidence of a recent
pregnancy or an organic lesion, such as a polyp,
may be helpful for subsequent clinical manage-160 7. Endometritis
Figure 7.14. Pseudoactinomycotic radiate granules.
These granular structures superficially resemble
sulfur granules of actinomycosis. The filamentous
structures are thicker than actinomycosis filaments,
however, and do not stain for organisms with Gram
or silver stains. The patient had an IUD in place at
the time of the biopsy. The endometrium was not
inflamed.
ment. The presence of such lesions helps to
establish the clinical cause and significance of
the inflammation.
The intensity of the inflammation also should
be noted, especially as severe inflammation
with neutrophils in the surface epithelium, a
dense subepithelial stromal lymphocytic infil-trate, and lymphoid follicles with transformed
lymphocytes raises the possibility of upper
genital tract inflammation and salpingitis.
Specific infections, such as tuberculosis or
cytomegalovirus, should clearly be indicated
when they are present. If special stains for
organisms are performed, the results should be
given in the report.
References
(1) Kiviat NB, Wolner-Hanssen P, Eschenbach DA,
Wasserheirt JN, Paavonen JA, Bell TA, et al.
Endometrial histopathology in patients with
culture-proved upper genital tract infection and
laparoscopically diagnosed acute salpingitis.
Am J Surg Pathol 1990; 14:167–175.
(2) Paavonen J, Aine R, Teisala K, Heinonen PK,
Punnonen R. Comparison of endometrial
biopsy and peritoneal fluid cytologic testing
with laparoscopy in the diagnosis of acute
pelvic inflammatory disease. Am J Obstet
Gynecol 1985; 151:645–650.
(3) Cadena D, Cavanzo FJ, Leone CL, Taylor HB.
Chronic endometritis. A comparative clinico-pathologic study. Obstet Gynecol 1973; 41:
733–738.
(4) Dallenbach-Hellweg G. Histopathology of the
endometrium, 4th ed. New York: Springer-Verlag; 1987.
(5) Rotterdam H. Chronic endometritis. A clinico-pathologic study. Pathol Annu 1978; 13:209–
231.
(6) Czernobilsky B. Endometritis and infertility.
Fertil Steril 1978; 30:119–130.
(7) Wallach EE. The uterine factor in infertility.
Fertil Steril 1972; 23:138–158.
(8) Haggerty CL, Ness RB, Amortegui A, Hendrix
SL, Hillier SL, Holley RL, et al. Endometritis
does not predict reproductive morbidity after
pelvic inflammatory disease. Am J Obstet
Gynecol 2003; 188:141–148.
(9) Greenwood SM, Moran JJ. Chronic endometri-tis. Morphologic and clinical observations.
Obstet Gynecol 1981; 58:176–184.
(10) Winkler B, Reumann W, Mitao M, Gallo L,
Richart RM, Crum CP. Chlamydial endometri-tis. A histological and immunohistochemical
analysis. Am J Surg Pathol 1984; 8:771–778.
(11) Buckley CH, Fox H. Biopsy pathology of the
endometrium. 2nd ed. London: Arnold; 2002.
(12) Sherman ME, Mazur MT, Kurman RJ. Benign
diseases of the endometrium. In: Kurman RJ,
ed. Blaustein’s pathology of the female genital
tract. 5 ed. New York: Springer-Verlag; 2002:
421–466.
(13) Crum CP, Egawa K, Fenoglio CM, Richart RM.
Chronic endometritis. The role of immunohis-tochemistry in the detection of plasma cells. Am
J Obstet Gynecol 1983; 147:812–815.
(14) Bayer-Garner IB, Korourian S. Plasma cells in
chronic endometritis are easily identified when
stained with syndecan–1. Mod Pathol 2001; 14:
877–879.
(15) Euscher E, Nuovo GJ. Detection of kappa- and
lambda-expressing cells in the endometrium by
in situ hybridization. Int J Gynecol Pathol 2002;
21:383–390.
(16) Young RH, Harris NL, Scully RE. Lymphoma-like lesions of the lower female genital tract. A
report of 16 cases. Int J Gynecol Pathol 1985;
4:289–299.
(17) Ferry JA, Young RH. Malignant lymphoma,
pseudolymphoma, and hematopoietic disorders
of the female genital tract. Pathol Annu 1991;
26:227–263.
(18) Skensved H, Hansen A, Vetner M. Immunore-active endometritis. Br J Obstet Gynaecol 1991;
98:578–582.
(19) Miko TL, Lampe LG, Thomazy VA, Molnar P,
Endes P. Eosinophilic endomyometritis asso-ciated with diagnostic curettage. Int J Gynecol
Pathol 1988; 7:162–172.
(20) Ladefoged C, Lorentzen M. Xanthogranuloma-tous inflammation of the female genital tract.
Histopathology 1988; 13:541–551.
(21) Russack V, Lammers RJ. Xanthogranuloma-tous endometritis. Arch Pathol Lab Med 1990;
114:929–932.
(22) Lopez JI, Nevado M. Exuberant xanthogranu-lomatous-like reaction following endometrial
curettage. Histopathology 1989; 15:315–322.
(23) Shintaku M, Sasaki M, Baba Y. Ceroid-contain-ing histiocytic granuloma of the endometrium.
Histopathology 1991; 18:169–172.
(24) Paavonen J, Aine R, Teisala K, Heinonen PK,
Punonen R, Lehtinen M, et al. Chlamydial
endometritis. J Clin Pathol 1985; 38:726–732.
(25) Mount S, Mead P, Cooper K. Chlamydia tra-chomatis in the endometrium: Can surgical
pathologists identify plasma cells? Adv Anat
Pathol 2001; 8:327–329.
References 161
(26) Paukku M, Puolakkainen M, Paavonen T,
Paavonen J. Plasma cell endometritis is associ-ated with Chlamydia trachomatis infection. Am
J Clin Pathol 1999; 112:211–215.
(27) Bazaz–Malik G, Maheshwari B, Lal N. Tuber-culous endometritis. A clinicopathological
study of 1000 cases. Br J Obstet Gynaecol 1983;
90:84–86.
(28) Nogales-Ortiz F, Tarancon I, Nogales FF. The
pathology of female genital tuberculosis. A 31-year study of 1436 cases. Obstet Gynecol 1979;
53:422–428.
(29) Schaefer G, Marcus RS, Kramer EE. Post-menopausal endometrial tuberculosis. Am J
Obstet Gynecol 1972; 112:681–687.
(30) Govan ADT. Tuberculous endometritis. J
Pathol Bacteriol 1962; 83:363–372.
(31) Frank TS, Himebaugh KS, Wilson MD. Granu-lomatous endometritis associated with histo-logically occult cytomegalovirus in a healthy
patient. Am J Surg Pathol 1992; 16:716–720.
(32) Ho K-L. Sarcoidosis of the uterus. Hum Pathol
1979; 10:219–222.
(33) Ferryman SR, Stephens M, Gough D. Necrotis-ing granulomatous endometritis following
endometrial ablation therapy. Br J Obstet
Gynaecol 1992; 99:928–930.
(34) Ashworth MT, Moss CI, Kenyon WE. Granu-lomatous endometritis following hysteroscopic
resection of the endometrium. Histopathology
1991; 18:185–187.
(35) Colgan TJ, Shah R, Leyland N. Post-hystero-scopic ablation reaction: A histopathologic
study of the effects of electrosurgical ablation.
Int J Gynecol Pathol 1999; 18:325–331.
(36) Sayage L, Gunby R, Gonwa T, Husberg B,
Goldstein R, Klintmalm G. Cytomegalovirus
endometritis after liver transplantation. Trans-plantation 1990; 49:815–817.
(37) Brodman M, Deligdisch L. Cytomegalovirus
endometritis in a patient with AIDS. Mt Sinai J
Med (NY) 1986; 53:673–675.
(38) Dehner LP, Askin FB. Cytomegalovirus
endometritis: Report of a case associated with
spontaneous abortion. Obstet Gynecol 1975;
45:211–214.
(39) McCracken AW, A’Agostino AN, Brucks AB,
Kingsley WB. Acquired cytomegalovirus infec-tion presenting as viral endometritis. Am J Clin
Pathol 1974; 61:556–560.
(40) Goldman RL. Herpetic inclusions in the
endometrium. Obstet Gynecol 1970; 36:
603–605.
(41) Abraham AA. Herpes virus hominis endo-metritis in a young woman wearing an intrau-terine contraceptive device. Am J Obstet
Gynecol 1978; 131:340–342.
(42) Schneider V, Behm FG, Mumau VR. Ascending
herpetic endometritis. Obstet Gynecol 1982;
58:259–262.
(43) Duncan DA, Varner RE, Mazur MT. Uterine
herpes virus infection with mutifocal necrotiz-ing endometritis. Hum Pathol 1989; 20:
1021–1024.
(44) Mazur MT, Hendrickson MR, Kempson RL.
Optically clear nuclei. An alteration of
endometrial epithelium in the presence of tro-phoblast. Am J Surg Pathol 1983; 7:415–423.
(45) Dardi LE, Ariano L, Ariano MC, Gould VE.
Arias-Stella reaction with prominent nuclear
pseudoinclusions simulating herpetic
endometritis. Diagn Gynecol Obstet 1982;
4:127–132.
(46) Horne HW, Hertig AT, Kundsin RB, Kosasa TS.
Sub-clinical endometrial inflammation and T-mycoplasma. A possible cause of human repro-ductive failure. Int J Fertil 1973; 18:226–231.
(47) Burke RK, Hertig AT, Miele CA. Prognostic
value of subacute focal inflammation of the
endometrium, with special reference to pelvic
adhesions as observed on laparoscopic exami-nation. An eight-year review. J Reprod Med
1985; 30:646–650.
(48) Khatamee MA, Sommers SC. Clinicopatho-logic diagnosis of mycoplasma endometritis. Int
J Fertil 1989; 34:52–59.
(49) Kim KR, Lee YH, Ro JY. Nodular histiocytic
hyperplasia of the endometrium. Int J Gynecol
Pathol 2002; 21:141–146.
(50) O’Brien PK, Roth-Moyo LA, Davis BA.
Pseudo-sulfur granules associated with intra-uterine contraceptive devices. Am J Clin Pathol
1981; 75:822–825.
(51) Bhagavan BS, Ruffier J, Shinn B. Pseudoac-tinomycotic radiate granules in the lower
female genital tract. Relationship to the
Splendore–Hoeppli phenomenon. Hum Pathol
1982; 13:898–904.
162 7. Endometritis
In general, polyps are benign growths with
no malignant potential.4
Atypical hyperplasia
and occasional cases of carcinoma, including
serous carcinoma and even mixed mesodermal
tumors, can be confined to a polyp.3;20–26
Fur-thermore, polyps have been associated with the
occurrence of carcinoma in several studies; they
are often found concurrently with carcinoma
3;27;28
or are seen more frequently in benign
endometrial biopsies taken prior to a diagnosis
of carcinoma.29
Nonetheless, polyps are not
regarded as a major risk factor for the devel-opment of carcinoma. Some studies have found
that hyperplasia, but not carcinoma, was more
frequently found in biopsy specimens that also
contained polyps.30;31
Most lesions are success-fully treated with curettage or hysteroscopic
excision.
The large variation in the reported preva-lence of polyps reflects the difficulties in estab-lishing the histologic diagnosis. Polyps often are
fragmented and removed piecemeal at curet-tage and therefore they can be difficult to rec-ognize. Hysteroscopy can be useful to confirm
the diagnosis of polyp,6;32
although in one study
as many as 13% of polyp-like structures seen by
hysteroscopy were not confirmed histologi-cally.2
Given the difficulty in the histologic diag-nosis of polyps, it is conceivable that true polyps
detected by hysteroscopy were not properly
classified on microscopic examination. Focal
lesions such as polyps are often difficult to diag-nose accurately even by dilation and curettage
(D&C), and up to 58% of polyps were missed
by D&C in one study.33
In our experience we
163
8
Polyps
Most endometrial polyps appear to originate
from localized hyperplasia of the basalis,
although their pathogenesis is not well under-stood. Polyps occur over a wide age range, but
are most common in women in the fourth
and fifth decades, becoming less frequent after
age 60.1–3
Usually they present with abnormal
uterine bleeding,4;5
and have been implicated as
a cause of abnormal bleeding in between 2%
and 23% of patients coming to biopsy.2;6–8
They
also have been implicated as a possible cause
of infertility, either by physically interfering
with blastocyst implantation or by altering the
development of secretory phase endometrium,
making it less receptive to the implanting
embryo.9–12
Tamoxifen therapy is a risk factor
for endometrial polyps,5;13
and polyps are the
most frequent pathologic lesion found in pa-tients receiving tamoxifen therapy for breast
carcinoma (see Fig. 6.15).14–16
Tamoxifen-related polyps may be multiple and large
(Chapter 6).17;18
Hormone replacement therapy
is not a risk factor for polyps.13;19
Large polyps
that extend into the endocervix and dilate the
internal os can cause endometritis.
Classification and Histologic Features  . . . 164
Common Polyps . . . . . . . . . . . . . . . . . . . 167
Atypical Polypoid Adenomyoma . . . . . . 170
Differential Diagnosis  . . . . . . . . . . . . . . . . 173
Adhesions  . . . . . . . . . . . . . . . . . . . . . . . 174
Clinical Queries and Reporting  . . . . . . . . 175
have seen a number of cases in which limited
biopsies reveal only small amounts of prolifer-ative or atrophic endometrium, but on follow-up, a more thorough sampling, often with
hysteroscopy, reveals a polyp.
Classification and
Histologic Features
Polyps vary greatly in size, ranging from micro-scopic abnormalities that are only a few mil-limeters across to huge lesions that can fill the
uterine cavity or prolapse through the endo-cervical canal. Usually polyps are small and
solitary, but they can be large or multiple. They
may be sessile or pedunculated. The glands and
stroma of endometrial polyps can show diverse
histologic patterns.1;3;4;34–36
For practical pur-poses polyps are divided into benign polyps and
the atypical polypoid adenomyoma. There are
five morphologic forms of common benign
endometrial polyps: proliferative/hyperplastic,
atrophic, functional, mixed endometrial–endo-cervical, and adenomyomatous (Table 8.1).
These patterns often overlap, and have little
clinical significance, serving only to demon-strate the spectrum of morphologic forms that
may be seen. Common polyps also can contain
focal lesions such as complex hyperplasia, atyp-ical hyperplasia (see Chapter 9), intraepithelial
carcinoma (see Chapter 9), or carcinoma (see
Chapter 10). The atypical polypoid adenomy-oma represents a distinctive form of polyp that
should be segregated from the variants of
common polyps.1
Despite their diverse growth patterns, all
polyps show several histologic features that
facilitate their diagnosis (Table 8.2).1;4
One
important feature of polyps in biopsy speci-mens is the presence of large, polypoid tissue
fragments (Figs. 8.1 and 8.2). These large frag-ments tend to be lined on three sides by surface
epithelium. The polypoid fragments appear dis-tinctly different from fragments of normal
endometrium, which often are admixed. The
stroma in polyps can be highly variable. Many
polyps have dense stroma with tightly packed
spindle cells that generally resemble the stroma
of proliferative endometrium (Fig. 8.3). This
stroma can appear mildly hypercellular and
mitotic activity can be increased.37
Rare polyps
show enlarged, atypical stromal cells that
resemble benign atypical stromal cells seen at
other sites in the female genital tract such as the
vagina (Fig. 8.4).38
Variable amounts of edema
and myxoid change also can be found, and
these changes are especially notable in tamox-ifen-related polyps (see Chapter 6).39
In addi-tion, thick-walled vessels may be prominent in
the stroma, especially when the polyp is large.
Small veins in the superficial stroma can
become ectatic.
The glands in polyps are irregular in shape
and have a highly variable architecture. They
may be focally crowded. The glands lack the
uniform orientation of normal endometrial
glands and may lose their perpendicular orien-tation to surface epithelium and course parallel
to the surface. In secretory endometrium the
irregularly shaped glands often lack normal
development and appear to be “out of phase.”2
The surface and glandular epithelium in
polyps often shows epithelial cytoplasmic
changes. These include ciliated cell,
164 8. Polyps
Table 8.1. Histologic patterns of endometrial
polyps.
Common polyps
Proliferative/hyperplastic
Atrophic
Functional
Mixed endometrial–endocervical
Adenomyomatous
Atypical polypoid adenomyoma
Table 8.2. Histologic features of polyps.
Larger tissue fragments
Polypoid shape
Surface epithelium on three sides
Dense stroma, may be fibrous
Thick-walled arteries
Glands more irregular, tortuous and dilated than normal
glands
Glands appear “out of phase” or hyperplastic
Distinct fragments that “stand apart” from the remaining
endometrial tissue
Figure 8.1. Benign polyp. The glands in this prolif-erative/hyperplastic type polyp are irregular and the
stroma is dense. In contrast to the other tissue
fragments that showed atrophic endometrium, this
Figure 8.2. Benign polyp. The irregularly shaped
glands resemble those in hyperplasia including
cystic and irregular glands. In contrast to the diffuse
large polypoid fragment of tissue represents a polyp.
The irregular gland pattern can mimic simple
hyperplasia.
process that characterizes hyperplasia, this was a
focal abnormality. The remaining endometrium was
proliferative.
Figure 8.3. Benign polyp. Higher magnification of a
polyp shows irregular glands with pseudostratified
nuclei. The dense stroma is a feature of polyps. The
Figure 8.4. Benign polyp with atypical stromal cells.
Portion of a benign polyp with proliferative/hyper-plastic gland changes shows increased stromal cellu-presence of proliferative endometrium elsewhere in
the curettings helped to distinguish this polyp from
hyperplasia, which tends to be diffuse.
larity with enlarged stromal cell nuclei. In the
absence of increased mitotic activity or other fea-tures of malignancy this finding has no significance.
eosinophilic, mucinous, and squamous changes
(Figs. 8.5 and 8.6). Cytoplasmic changes usually
occur in the glands, but sometimes they occur
along the surface, where they can be focal or
extensive. Hyperplastic papillary proliferations
(see Chapter 9) have been described as a focal
component of polyps, usually along the
surface.40
In addition to these characteristic features,
polyps may show evidence of focal glandular
and stromal breakdown, if they outgrow their
vascular supply or undergo thrombosis of
dilated superficial veins (see Chapter 5).
Chronic bleeding leads to hemosiderin deposi-tion or foam cells in the stroma. Occasionally a
larger polyp will show extensive ischemic
necrosis of the distal portion. The specific fea-tures of each type of polyp that assist in the
pathologic classification are given in the sec-tions that follow.
Common Polyps
Proliferative/Hyperplastic Pattern
Proliferative/hyperplastic polyps are most
common. They are highly variable in size and
may measure up to several centimeters in great-est dimension (Figs. 8.1 and 8.2). They often are
diagnosed only on microscopic examination.
Regardless of size, they show irregular, prolif-erating glands with pseudostratified nuclei and
mitotic activity (Fig. 8.3). These polyps usually
have a moderate amount of intervening stroma
between the glands, but sometimes the glands
are closely packed. The glands generally
resemble those in a disordered proliferative
pattern or hyperplasia without atypia, either
simple or complex, showing irregular gland
sizes and shapes (see Chapter 9). These irregu-lar glands can also appear as focal hyperplasia
in a background of proliferative or atrophic
Classification and Histologic Features 167
Figure 8.5. Benign polyp. Polyp with irregular, crowded glands and extensive squamous change (arrows).
Benign endometrium separate from the polyp is present in the left upper corner.
endometrium. Identifying endometrium not
involved by the polyp is especially helpful for
accurate diagnosis of a proliferative/hyperplas-tic polyp, as the polyp may otherwise resemble
a diffuse hyperplasia. The endometrium that is
not involved by the polyp is usually prolifera-tive or atrophic, but sometimes it is secretory in
premenopausal women.
Atrophic Pattern
Atrophic polyps are usually seen in post-menopausal women. These polyps contain
atrophic glands lined by low columnar epithe-lium showing no mitotic activity. The glands
often are dilated and cystic, and the stroma
appears dense and fibrotic (Fig. 8.7). Many of
these polyps apparently represent prolifera-tive/hyperplastic polyps that no longer show
proliferative activity.
Functional Pattern
These polyps, like the endometrium around
them, are hormonally responsive and show pro-liferative or secretory changes (Fig. 8.8). They
occur in premenopausal patients and can be
difficult to diagnose. Unlike normal glands, the
glands in the functional polyp lose their orien-tation to surface epithelium and have a hap-hazard distribution. In well-oriented fragments,
the glands may appear to branch toward the
surface like veins on a leaf. The stroma in a
polyp may show edema or predecidual change
but often is dense and inactive. When polyps
have secretory changes, the glands often are not
as well developed as those in the surrounding
endometrium (Fig. 8.9). The result is that they
appear to be “out of phase” and the stroma
shows decreased edema and predecidual
formation. Some examples of secretory
168 8. Polyps
Figure 8.6. Squamous change in polyp. High-magnification view of area in a polyp shows nonkeratinizing
squamous change.
Figure 8.7. Atrophic polyp with cystic glands lined by atrophic epithelium.
Figure 8.8. Functional polyp. Irregular glands with early secretory changes surrounded by edematous
stroma. The stalk of the polyp has dense stroma and thick-walled vessels (arrows).
endometrium with irregular maturation proba-bly represent functional polyps that are too
fragmented for accurate diagnosis.
Mixed Endometrial–Endocervical Pattern
Some polyps originate in the upper endocervix
and lower uterine segment and show both
endocervical and endometrial-type gland
development (Fig. 8.10). Such polyps tend to
have a fibrous stroma resembling the stroma of
the lower uterine segment.
Adenomyomatous Pattern
These polyps have smooth muscle in their
stroma, usually as irregular bundles and strands
in proximity to thick-walled vessels. Although
smooth muscle is present, the glands are
invested by stroma. These polyps usually have
proliferative/hyperplastic or functional gland
changes.
Atypical Polypoid Adenomyoma
This is an unusual and distinctive polyp charac-terized by glands that are lined by atypical
epithelium and surrounded by cellular smooth
muscle and variable amounts of fibrous tissue
(Fig. 8.11).41–43
The atypical polypoid adenomy-oma typically occurs in premenopausal or per-imenopausal women, with a mean age of about
40 years. Premenopausal women often are nul-liparous.43
We have seen these lesions in women
as old as 56, however, and there is a report of a
case in an 81-year-old woman.44
A few cases
have been associated with Turner syndrome
and appear to be a complication of long-term
estrogenic stimulation of the endometrium.45
Often these lesions arise in the lower uterine
segment.42
The glands of the atypical polypoid adeno-myoma are haphazardly arranged but generally
are not markedly crowded or back-to-back.
170 8. Polyps
Figure 8.9. Functional polyp. Polypoid tissue frag-ment shows irregular secretory glands with dense
central stroma and superficial edema. Endometrium
elsewhere in the sections showed a normal mid
secretory phase pattern. Inset: The glands show
secretory exhaustion and the stroma is dense.
Figure 8.10. Mixed endometrial–endocervical po-lyp.This polyp occurred in a 60-year-old and is charac-terized by irregular, dilated glands in fibrous stroma.
Figure 8.11. Atypical polypoid adenomyoma. Irregular, atypical glands are haphazardly distributed in
smooth muscle. The smooth muscle has a characteristic pattern of short, interlacing fascicles.
Inset: The epithelium is partially endocervical–type
with columnar mucinous cells.
They resemble the glands in simple atypical
hyperplasia. The glandular cells have enlarged,
stratified, and rounded nuclei with a vesicular
chromatin pattern and prominent nucleoli (Fig.
8.12). The cytoplasm is eosinophilic and the
glands resemble those found in atypical hyper-plasia (see Chapter 9). A very characteristic
though not specific feature is the presence
of squamous change (metaplasia), containing
central nonkeratinizing nests of squamous cells
(Fig. 8.13). Central necrosis may occur in the
squamous nests.
Smooth muscle encompasses the glands, and
endometrial stroma is largely absent. The
smooth muscle is arranged in short interlacing
fascicles that contrast with the elongate bundles
of smooth muscle found in normal
myometrium or in adenomyomatous polyps
(Figs. 8.11 and 8.13). The smooth muscle com-ponent can show increased mitotic activity, with
up to two mitoses per 10 high-power fields
(HPFs), but there is no evidence of cytologic
atypia. The smooth muscle is diffusely reactive
for desmin.
As with any type of polyp, the endometrium
not involved by the atypical polypoid adeno-myoma can be highly variable and can show
proliferative, secretory, gestational, or hyper-plastic changes. The atypical polypoid adeno-myoma often presents in curettage specimens
as large polypoid tissue fragments admixed
with small fragments of noninvolved
endometrium.
Endometrial adenocarcinoma rarely may
develop in association with an atypical poly-poid adenomyoma.43;46–48
Usually these carcino-mas are not aggressive and curettage may be
curative. The atypical polypoid adenomyoma
without carcinoma requires follow-up, which
can include hysteroscopy, repeat curettage, and
imaging studies to be certain that the entire
lesion has been removed.
172 8. Polyps
Figure 8.12. Atypical polypoid adenomyoma. The glandular epithelium is atypical with enlarged, stratified
nuclei that have clumped chromatin and prominent nucleoli.
Differential Diagnosis
Endometrial polyps may be difficult to identify
in curettage specimens and are often missed.
When the specimen is large and polypoid-shaped, has surface epithelium covering three
sides, and contains dense stroma with thick-walled vessels, the diagnosis is straightforward.
Often polyps are microscopic abnormalities or
they are highly fragmented by the biopsy pro-cedure. Because of fragmentation and lack of
clear-cut diagnostic features, the differential
diagnosis often includes normal proliferative or
secretory endometrium. Subtle features, includ-ing irregular gland shape and distribution as
well as a somewhat denser stroma, are helpful
in the recognition of a small polyp that is
admixed with normal proliferative or secretory
endometrium.
Tangential sectioning of normal endo-metrium is a frequent consideration in the dif-ferential diagnosis of small and fragmented
polyps. Basalis, especially, can resemble a polyp
because it has irregular glands, dense stroma,
and prominent arteries. Basalis, however, lacks
a lining of surface epithelium and does not have
a polypoid shape. The vessels in basalis consist
of small aggregates of basal arteries, usually
numbering six or more in cross section. Levels
through the tissue block often show transitions
to normal functionalis if the area in question
represents basalis. Conversely, if the tissue rep-resents a true polyp, the step sections often
reveal surface epithelium, at least focally.
The fibroblastic spindle-cell stromal response
that occurs in endometritis may resemble the
fibrous stroma of a polyp. However, with
inflammation, there is a plasma cell infiltrate in
Differential Diagnosis 173
Figure 8.13. Atypical polypoid adenomyoma. One
of the cytologically atypical glands shows squamous
change with a nest of cytologically benign squamous
cells. The mesenchymal component is composed of
short interlacing fascicles of smooth muscle cells.
Squamous change can be extensive in the atypical
polypoid adenomyoma and may even show focal
necrosis. Squamous change, even with necrosis, has
no prognostic significance.
the stroma, and the epithelium also may be
infiltrated by neutrophils.
Irregular polypoid tissue due to progestin
effect also may superficially resemble a polyp
in a biopsy or curettage specimen. These speci-mens usually contain atrophic glands and a
decidua-like stroma. The finding of normal
endometrium admixed with polypoid tissue
fragments is helpful for identifying polyps in
questionable cases.
Diffuse hyperplasia (see Chapter 9) is also
included in the differential diagnosis of polyp,
especially polyps with the proliferative/hyper-plastic pattern. Both polyps and hyperplasia
have an irregular proliferation of glands with a
variable amount of stroma. Hyperplasia, fur-thermore, often becomes polypoid. Hyperplasia
is a diffuse process, however, usually affecting
all or most of the endometrium, while polyps
are focal and admixed with normal, nonhyper-plastic endometrium. Most polyps have charac-teristically dense stroma and appear clearly
different from the surrounding uninvolved
endometrium. In small biopsies, however, the
distinction may not be possible, and the ques-tion of whether a small polypoid fragment of
tissue represents a portion of a polyp or a
hyperplasia cannot be resolved. Dilation and
curettage or hysteroscopy may be necessary to
establish the correct diagnosis.
A large polyp with proliferative/hyperplastic
features also may resemble adenosarcoma (see
Chapter 11), as both lesions are polypoid with
hyperplastic glands. Adenosarcoma typically
has a leaf-like pattern with broad-based papil-lae lined by surface epithelium. In contrast, a
polyp has a smooth outline. In addition,
adenosarcoma, unlike benign polyps, has a more
cellular stroma with increased mitotic activity
that aggregates in cuffs around the glands.
Polyps generally have a more dense or fibrous
stroma that is less cellular and has a low mitotic
rate. Large polyps typically have thick-walled
vessels, which are lacking in adenosarcoma.
The differential diagnosis for functional
polyps includes changes due to ovulatory dys-function, such as luteal phase defects (LPDs) or
exogenous hormone effects. These polyps show
secretory changes that are out of phase with the
surrounding endometrium. Functional polyps
are composed of polypoid tissue fragments
covered with surface epithelium and containing
dense stroma and thick-walled vessels. Sur-rounding fragments of endometrium that show
normal secretory changes are especially helpful
for establishing the diagnosis of a functional
polyp.
The atypical polypoid adenomyoma may
resemble well-differentiated carcinoma invad-ing the myometrium because it shows atypical
glands in smooth muscle. The glands in the
atypical polypoid adenomyoma, however, do
not fulfill criteria for well-differentiated adeno-carcinoma, lacking features of stromal invasion
such as cribriform or confluent growth (see
Chapter 10). Also, the smooth muscle of the
atypical polypoid adenomyoma grows in short,
interlacing fascicles in contrast to the elongated
fibers seen in normal myometrium. It is
extremely unusual to find well-differentiated or
moderately differentiated adenocarcinoma
invading myometrial smooth muscle in
endometrial curettings. Accordingly, in curet-tings atypical glands surrounded by smooth
muscle represent an atypical polypoid adeno-myoma until proven otherwise. Squamous
change is common in the atypical polypoid ade-nomyoma but also can occur in adenocarci-noma or hyperplasia. In atypical polypoid
adenomyoma with squamous change, however,
the affected glands are invested by smooth
muscle, and this is an important feature that
helps to identify the lesion (Fig. 8.13). The rare
examples of adenocarcinoma occurring in asso-ciation with atypical polypoid adenomyoma
show typical endometrioid features without
smooth muscle investing the malignant
glands.46;47
Adhesions
Endometrial adhesions are bands of tissue that
bridge the endometrial cavity. These may
consist of fibrous tissue that forms as a result of
inflammation associated with an abortion or
pregnancy, and their presence is known as
Asherman syndrome.1;10;49
In these cases the
endometrium is largely replaced by a prolifer-ation of fibroblasts, but this is an extraordinar-ily rare finding in biopsy material.
174 8. Polyps
We have observed, on occasion, another type
of intrauterine adhesion formed by a band of
endometrial tissue with dense stroma and
small, nonreactive glands. This form of adhesion
typically is found in premenopausal women and
its etiology is not known. Because such adhe-sions consist of dense stroma and poorly devel-oped glands, they may be indistinguishable
from fragments of a polyp (Fig. 8.14). Correla-tion of histologic features with other informa-tion, such as hysteroscopic findings, may be
necessary to establish the diagnosis.
Clinical Queries and Reporting
Diagnosis of a polyp is important because its
presence often offers an explanation for abnor-mal bleeding or accounts for a thickened stripe
on vaginal ultrasound. On the other hand, clas-sification of polyps is of importance to the
pathologist as an aid for recognition, but is of
little relevance to the gynecologist for thera-peutic decisions. In fact, subclassification of
polyps can unnecessarily complicate the patho-logic report. Usually it is only important to indi-cate that a benign polyp is present. Often a
comment regarding the status of endometrium
not involved by the polyp also can be helpful
for the gynecologist. For instance, if curettings
in a postmenopausal patient show a polyp and
separate fragments of atrophic endometrium, a
comment or diagnosis that reflects all of these
observations is warranted, since it also indicates
the status of the endometrium not involved by
the primary lesion. This type of interpretive
comment helps the gynecologist understand
that the endometrial cavity has been sampled
beyond the focal polyp.
When the atypical polypoid adenomyoma is
diagnosed, a comment should be included to
describe the lesion, because most gynecologists
Clinical Queries and Reporting 175
Figure 8.14. Adhesion. Intrauterine adhesion shows
irregular glands in dense stroma. This lesion was
identified by hysteroscopy in a 23-year-old patient
who was being evaluated for infertility. She had no
known risk factors for the development of adhesions.
are not familiar with it. It is necessary to indi-cate that the lesion is benign and that it does not
represent atypical hyperplasia or carcinoma.
At times endometrial sampling may yield
tissue suggestive, but not diagnostic, of a polyp.
Small samples and markedly fragmented tissue
may show features suggestive of a polyp but can
be especially difficult to accurately diagnose. In
limited samples, small fragments of polypoid
tissue with irregular glands may represent por-tions of a polyp or they may represent part of
a hyperplasia. In such cases, a descriptive diag-nosis to indicate the presence of an abnormal-ity, coupled with a microscopic description of
the findings, is most helpful to the clinician. If
the diagnosis is not straightforward, further
evaluation, including a dilation and curettage or
hysteroscopy, may be necessary to establish the
correct diagnosis.
References
(1) Sherman ME, Mazur MT, Kurman RJ. Benign
diseases of the endometrium. In: Kurman RJ,
ed. Blaustein’s pathology of the female genital
tract. 5th ed. New York: Springer-Verlag; 2002:
421–466.
(2) Van Bogaert L-J. Clinicopathologic findings
in endometrial polyps. Obstet Gynecol 1988;
71:771–773.
(3) Peterson WF, Novak ER. Endometrial polyps.
Obstet Gynecol 1956; 8:40–49.
(4) Schlaen I, Bergeron C, Ferenczy A, Wong P,
Naves A, Bertoli R. Endometrial polyps: A
study of 204 cases. Surg Pathol 1988; 1:375–382.
(5) Reslova T, Tosner J, Resl M, Kugler R, Vavrova
I. Endometrial polyps. A clinical study of 245
cases. Arch Gynecol Obstet 1999; 262:133–139.
(6) Loffer FD. Hysteroscopy with selective
endometrial sampling compared with D&C for
abnormal uterine bleeding. The value of a neg-ative hysteroscopic view. Obstet Gynecol 1989;
73:16–20.
(7) Schindler AE, Schmidt G. Post-menopausal
bleeding: A study of more than 1000 cases.
Maturitas 1980; 2:269–274.
(8) Choo YC, Mak KC, Hsu C, Wong TS, Ma HK.
Postmenopausal uterine bleeding of nonor-ganic cause. Obstet Gynecol 1985; 66:225–228.
(9) Dallenbach-Hellweg G. The endometrium of
infertility. A review. Pathol Res Pract 1984; 178:
527–537.
(10) Wallach EE. The uterine factor in infertility.
Fertil Steril 1972; 23:138–158.
(11) Foss BA, Horne HW, Hertig AT. The
endometrium and sterility. Fertil Steril 1958;
9:193–206.
(12) Sillo-Seidl G. The analysis of the endometrium
of 1,000 sterile women. Hormones 1971;
2:70–75.
(13) Bakour SH, Gupta JK, Khan KS. Risk factors
associated with endometrial polyps in abnor-mal uterine bleeding. Int J Gynaecol Obstet
2002; 76:165–168.
(14) Cohen I, Azaria R, Bernheim J, Shapira J, Beyth
Y. Risk factors of endometrial polyps resected
from postmenopausal patients with breast car-cinoma treated with tamoxifen. Cancer 2001;
92:1151–1155.
(15) Deligdisch L, Kalir T, Cohen CJ, de Latour M,
Le Bouedec G, Penault-Llorca F. Endometrial
histopathology in 700 patients treated with
tamoxifen for breast cancer. Gynecol Oncol
2000; 78:181–186.
(16) Vosse M, Renard F, Coibion M, Neven P,
Nogaret JM, Hertens D. Endometrial disorders
in 406 breast cancer patients on tamoxifen: The
case for less intensive monitoring. Eur J Obstet
Gynecol Reprod Biol 2002; 101:58–63.
(17) Nuovo MA, Nuovo GJ, McCaffrey FM, Levine
RU, Barron B, Winkler B. Endometrial polyps
in postmenopausal patients receiving tamox-ifen. Int J Gynecol Pathol 1989; 8:125–131.
(18) Corley D, Rowe J, Curtis MT, Hogan WM,
Noumoff JS, Livolsi VA. Postmenopausal
bleeding from unusual endometrial polyps in
women on chronic tamoxifen therapy. Obstet
Gynecol 1992; 79:111–116.
(19) Perrone G, DeAngelis C, Critelli C, Capri O,
Galoppi P, Santoro G, et al. Hysteroscopic find-ings in postmenopausal abnormal uterine
bleeding: a comparison between HRT users
and non-users. Maturitas 2002; 43:251–255.
(20) Dinh TV, Slavin RE, Bhagavan BS, Hannigan
EV, Tiamson EM, Yandell RB. Mixed mullerian
tumors of the uterus. A clinicopathologic study.
Obstet Gynecol 1989; 74:388–392.
(21) Salm R. The incidence and significance of early
carcinomas in endometrial polyps. J Pathol
1972; 108:47–53.
(22) Wolfe SA, Mackles A. Malignant lesions arising
from benign endometrial polyps. Obstet
Gynecol 1962; 20:542–550.
(23) Silva EG, Jenkins R. Serous carcinoma in
endometrial polyps. Mod Pathol 1990; 3:120–
128.
176 8. Polyps
(24) Kahner S, Ferenczy A, Richart RM. Homolo-gous mixed mullerian tumors (carcinosarcoma)
confined to endometrial polyps. Am J Obstet
Gynecol 1975; 121:278–279.
(25) Barwick KW, Livolsi VA. Heterologous mixed
mullerian tumor confined to an endometrial
polyp. Obstet Gynecol 1979; 53:512–514.
(26) Silverberg SG, Major FJ, Blessing JA, Fetter B,
Askin FB, Liao S-Y, et al. Carcinosarcoma
(malignant mixed mesodermal tumor) of the
uterus. A Gynecologic Oncology Group patho-logic study of 203 cases. Int J Gynecol Pathol
1990; 9:1–19.
(27) Armenia CS. Sequential relationship between
endometrial polyps and carcinoma of the
endometrium. Obstet Gynecol 1967; 30:
524–529.
(28) Gray LA, Robertson RWJ, Christopherson
WM. Atypical endometrial changes associated
with carcinoma. Gynecol Oncol 1974; 2:93–100.
(29) Pettersson B, Adami H-O, Lindgren A, Hes-selius I. Endometrial polyps and hyperplasia as
risk factors for endometrial carcinoma. A case-control study of curettage specimens. Acta
Obstet Gynecol Scand 1985; 64:653–659.
(30) Bakour SH, Khan KS, Gupta JK. The risk of
premalignant and malignant pathology in
endometrial polyps. Acta Obstet Gynecol
Scand 2000; 79:317–320.
(31) Orvieto R, Bar-Hava I, Dicker D, Bar J, Ben
Rafael Z, Neri A. Endometrial polyps during
menopause: Characterization and significance.
Acta Obstet Gynecol Scand 1999; 78:883–886.
(32) Fraser IS. Hysteroscopy and laparoscopy in
women with menorrhagia. Am J Obstet
Gynecol 1990; 162:1264–1269.
(33) Epstein E, Ramirez A, Skoog L, Valentin L.
Dilatation and curettage fails to detect most
focal lesions in the uterine cavity in women
with postmenopausal bleeding. Acta Obstet
Gynecol Scand 2001; 80:1131–1136.
(34) Dallenbach-Hellweg G. Histopathology of the
endometrium, 4th ed. New York: Springer-Verlag;1987.
(35) Silverberg SG, Kurman RJ. Tumors of the
uterine corpus and gestational trophoblastic
disease. Atlas of tumor pathology, 3rd series,
Fascicle 3. Washington, DC: Armed Forces
Institute of Pathology, 1992.
(36) Buckley CH, Fox H. Biopsy pathology of the
endometrium. 2nd ed. London: Arnold; 2002.
(37) Hattab EM, Allam–Nandyala P, Rhatigan RM.
The stromal component of large endometrial
polyps. Int J Gynecol Pathol 1999; 18:332–337.
(38) Tai LH, Tavassoli FA. Endometrial polyps with
atypical (bizarre) stromal cells. Am J Surg
Pathol 2002; 26:505–509.
(39) Kennedy MM, Baigrie CF, Manek S. Tamoxifen
and the endometrium: review of 102 cases and
comparison with HRT-related and non-HRT-related endometrial pathology. Int J Gynecol
Pathol 1999; 18:130–137.
(40) Lehman MB, Hart WR. Simple and complex
hyperplastic papillary proliferations of the
endometrium: A clinicopathologic study of nine
cases of apparently localized papillary lesions
with fibrovascular stromal cores and epithelial
metaplasia. Am J Surg Pathol 2001; 25:
1347–1354.
(41) Mazur MT. Atypical polypoid adenomyomas of
the endometrium. Am J Surg Pathol 1981;
5:473–482.
(42) Young RH, Treger T, Scully RE. Atypical poly-poid adenomyoma of the uterus. A report of 27
cases. Am J Clin Pathol 1986; 86:139–145.
(43) Fukunaga M, Endo Y, Ushigome S, Ishikawa E.
Atypical polypoid adenomyomas of the uterus.
Histopathology 1995; 27:35–42.
(44) Kuwashima Y, Uehara T, Kurosumi M, Kishi K,
Ohno H, Maruyama M, et al. Atypical polypoid
adenomyoma of the uterus in a very old
woman. Report of a case with immunohisto-chemical characterization of its stromal
components and proliferative status. Eur J
Gynaecol Oncol 1995; 16:115–119.
(45) Clement PB, Young RH. Atypical polypoid ade-nomyoma of the uterus associated with
Turner’s syndrome. Int J Gynecol Pathol 1987;
6:104–113.
(46) Lee KR. Atypical polypoid adenomyoma of the
endometrium associated with adenomyomato-sis and adenocarcinoma. Gynecol Oncol 1993;
51:416–418.
(47) Staros EB, Shilkitus WF. Atypical polypoid ade-nomyoma with carcinomatous transformation:
A case report. Surg Pathol 1991; 4:157–166.
(48) Sugiyama T, Ohta S, Nishida T, Okura N, Tanabe
K, Yakushiji M. Two cases of endometrial
adenocarcinoma arising from atypical polypoid
adenomyoma. Gynecol Oncol 1998; 71:
141–144.
(49) Friedler S, Margalioth EJ, Kafka I, Yaffe H.
Incidence of post-abortion intra-uterine adhe-sions evaluated by hysteroscopy. A prospective
study. Hum Reprod 1993; 8:442–444.
References 177
178
Hyperplasia occurs most frequently in peri-menopausal women, as they frequently have
anovulatory cycles, but it also occurs in post-menopausal women who either have excess
endogenous estrogen levels or are receiving
exogenous estrogen.1;3
On occasion hyperpla-sia may arise in younger women, including
teenagers, because sporadic anovulation occurs
during the reproductive years and anovulatory
cycles are frequent in adolescents.8–10
In the
reproductive years, women with chronic anovu-lation associated with the Stein–Leventhal syn-drome (polycystic ovaries) are especially prone
to develop hyperplasia. Sometimes hyperplasia
occurs when there is no apparent underlying
endocrinologic disorder. Recent studies have
shown that there are two forms of hyperplasia,
one (atypical) that is a precursor lesion to ade-nocarcinoma and another form (without atypia)
that is largely self limited with little apparent
direct relationship to carcinoma.1;3;5;9;11–18
The subject of epithelial metaplasia is closely
linked to the topic of hyperplasia, because so-called metaplasia occurs frequently in hyper-plasia.1;4;19–23
Most metaplasias represent
alterations of the epithelium that are either a
degenerative or regenerative “change” or a
form of cytoplasmic differentiation and not
truly metaplastic. These cellular changes are not
unique to hyperplasia, however, and occur in a
variety of other conditions. These alterations
often mimic the cellular features of hyperplasia
and therefore complicate the interpretation.
Refinements in the classification of endome-trial hyperplasia and related cellular changes
9
Endometrial Hyperplasia, Endometrial
Intraepithelial Carcinoma, and
Epithelial Cytoplasmic Change
Endometrial hyperplasia is a noninvasive pro-liferation of the endometrium that results in a
morphologic pattern of glands with irregular
shapes and varying size.1–7
This disorder results
from sustained, unopposed estrogen stimula-tion and presents clinically as abnormal uterine
bleeding. Sometimes hyperplasia is encoun-tered incidentally in a biopsy performed for
other reasons, such as infertility workup prior
to or during hormone replacement therapy.
Hyperplasia can mimic a wide variety of normal
physiologic changes, artifacts resulting from
tissue sampling and processing, benign organic
disorders, and well-differentiated adenocarci-noma. Because management of these condi-tions and the different forms of hyperplasia can
range from no treatment to hysterectomy,
correct diagnosis is essential.
Terminology and Classification of
Hyperplasia  . . . . . . . . . . . . . . . . . . . . . . . . 179
Endometrial Hyperplasia  . . . . . . . . . . . . . 179
Hyperplasia Without Atypia  . . . . . . . . . 180
Atypical Hyperplasia  . . . . . . . . . . . . . . . 182
Differential Diagnosis  . . . . . . . . . . . . . . 188
Behavior  . . . . . . . . . . . . . . . . . . . . . . . . 192
Endometrial Intraepithelial
Carcinoma  . . . . . . . . . . . . . . . . . . . . . . . . . 193
Differential Diagnosis  . . . . . . . . . . . . . . 193
Behavior  . . . . . . . . . . . . . . . . . . . . . . . . 195
Epithelial Cytoplasmic Change
(Metaplasia)  . . . . . . . . . . . . . . . . . . . . . . . 195
Differential Diagnosis  . . . . . . . . . . . . . . 202
Clinical Queries and Reporting  . . . . . . . . 204
reflect our increased understanding of endome-trial pathology. For example, a recently
described entity termed “endometrial intra-epithelial carcinoma (EIC),” which has no rela-tionship to endometrial hyperplasia, appears to
be a precursor of serous carcinoma.24
Introduc-tion of new terminology and classification,
however, is always met with resistance, and
statements such as “I used to understand hyper-plasia” express the frustration of some pathol-ogists and gynecologists who are reluctant to
adopt new terminology. The current classifica-tion and recognition of superimposed non-hyperplastic epithelial changes actually assist in
correctly diagnosing endometrial cancer pre-cursors. This chapter reviews the current classi-fication and morphologic features of these
lesions as well as their differential diagnosis.
Terminology and Classification
of Hyperplasia
The diagnosis and management of endometrial
hyperplasia have been unnecessarily compli-cated by the use of a wide variety of terms and
histologic classifications. Terms such as “adeno-matous hyperplasia,” “atypical hyperplasia,”
and “carcinoma in situ” have been used by
different authors for the same lesions, and,
conversely, different investigators have used
the same term to describe different
lesions.3;6;11;16–19;25
The distinction of atypical
hyperplasia from well-differentiated adenocar-cinoma has been further clouded by the term
“carcinoma in situ.”16;18;25
The confusion result-ing from the use of different classifications
often precluded comparison of data between
institutions and created problems in communi-cation between the gynecologist and the
pathologist. The World Health Organization
(WHO) and the International Society of Gyne-cologic Pathologists (ISGYP) have promoted
one classification of endometrial hyperplasia
that has gained widespread acceptance (Table
9.1).1;4;26;26a
This classification subdivides hyper-plasias into two categories, those without cyto-logic atypia and those with cytologic atypia
(atypical hyperplasia). The glandular complex-ity has secondary importance. Thus, hyperplasia
without atypia and atypical hyperplasia are
both divided into simple and complex cate-gories. These latter terms give a general assess-ment of the degree of gland crowding and
irregularity.
Lesions previously classified as “adenoma-tous hyperplasia” and “carcinoma in situ” have
been abandoned. Borderline lesions are classi-fied as either atypical hyperplasia or invasive
well-differentiated adenocarcinoma. With the
use of a single classification, refined morpho-logic criteria, and better understanding of the
behavior of these lesions, the diagnosis of
hyperplasia in a biopsy specimen allows the
gynecologist to individualize patient manage-ment. Recent studies have shown that some
degree of interobserver and intraobserver vari-ation in the diagnosis of hyperplasia does
occur.27–29
In addition, morphometric studies30;31
and molecular genetic analysis of clonality32;33
have led some investigators to propose a new
classification scheme for hyperplasia. In this
classification, proliferations with little to no risk
of evolution to carcinoma are termed “endome-trial hyperplasia (EH),” and a new term,
“endometrial intraepithelial neoplasia (EIN),”
is used to describe true carcinoma precursor
lesions.34;35
Approximately 30% of EIN cases
will progress to carcinoma.36;37
At present, how-ever, the WHO classification seems to be the
most practical and generally accepted.
Endometrial Hyperplasia
All types of hyperplasia are characterized by an
increase in the gland-to-stroma ratio, irregular-ities in gland shape, and variation in gland
size.1;4;5;38;39
In addition, mitotic activity is evi-dent, although the level is variable and may be
less than that in proliferative endometrium.
Endometrial Hyperplasia 179
Table 9.1. World Health Organization classification
of endometrial hyperplasia.
Hyperplasia (without atypia)
Simple
Complex
Atypical hyperplasia
Simple
Complex
Tissue obtained at curettage may be consider-able, sometimes yielding enough to fill three or
more tissue cassettes. On the other hand, office-based biopsies may yield only a limited volume
of tissue.
The amount of stroma separating the glands
distinguishes simple and complex forms of
hyperplasia, regardless of the presence of
atypia. Usually there also is increased glandu-lar complexity. Hyperplasia is generally a
diffuse abnormality but may be a focal abnor-mality, possibly because of regional differences
in estrogen and progesterone receptor content
in the endometrium.
Hyperplasia Without Atypia
Simple Hyperplasia
In simple hyperplasia, many, but not necessar-ily all, of the proliferating glands are dilated
and cystic, with irregular size and shape and
varying degrees of irregular branching with
infoldings and outpouchings of the glands, yet
separated by abundant stroma (Fig. 9.1 and
Table 9.2).
Cytologically, the glandular epithelium
resembles proliferative endometrium (Fig. 9.2).
The cells are columnar with amphophilic cyto-plasm and have pseudostratified nuclei that
maintain their orientation to the underlying
basement membrane. Nuclei are oval with
smooth contours, evenly dispersed chromatin,
and small, inconspicuous nucleoli. Mitotic activ-ity can be quite variable, but the mitotic rate
has no influence on the diagnosis of simple
hyperplasia. Cilia (ciliated cell change) often
are seen along the luminal border of glands as
well as along the surface epithelium. Squamous
metaplasia may also be present, although this
change is relatively infrequent in hyperplasia
without atypia.
180 9. EH, EIC, and Epithelial Cytoplasmic Change
Figure 9.1. Simple hyperplasia. Irregular glands showing marked variation in size and shape are separated
by abundant stroma. Several cystic glands are present in this field.
By definition, considerable stroma is present
in simple hyperplasia. The stroma resembles
that seen in the normal proliferative phase, con-sisting of small, oval cells with scant cytoplasm.
Like the glands, the stroma shows mitotic
activity. When the hyperplasia is polypoid,
the stroma may contain thick-walled arteries
similar to those seen in polyps. In simple hyper-plasia, ectatic vascular channels (venules) are
often present in the superficial stroma beneath
the surface epithelium. The pathogenesis of this
change is not well understood but appears to be
associated with nonphysiologic, noncyclical
endometrial growth. Morphologic evidence of
active breakdown and bleeding (see Chapter 5)
also may be present around thrombosed ectatic
venules.
Complex Hyperplasia
In contrast to simple hyperplasia, complex
hyperplasia shows more densely crowded
Endometrial Hyperplasia 181
Table 9.2. Morphologic features of hyperplasia
without atypia.
Cytologic features
Nuclei
Pseudostratified
Cigar-shaped to oval with smooth contours
Uniform chromatin distribution
Small to indistinct nucleoli
Mitotic activity, variable amount
Cytoplasm
Variable, often amphophilic
Glands
Irregular, variable size, some dilated
Branching, infolding and outpouching
Simple hyperplasia
Haphazardly spaced in abundant stroma
Complex hyperplasia
Closely spaced with decreased stroma
Highly irregular outlines
Frequent associated features
Polypoid growth
Ciliated cells
Ectatic venules
Breakdown and bleeding
Figure 9.2. Simple hyperplasia. The glandular cell
nuclei are oval and pseudostratified with uniform
outlines, lacking cytologic atypia. Nucleoli are indis-tinct. Both glandular and stromal cells are cytologi-cally similar to those of proliferative phase
endometrium. Note mitotic figures (arrows).
glands. In addition, the glands may demonstrate
increased structural complexity with more out-pouchings and infoldings (Fig. 9.3). Usually the
glands are closely apposed and often back-to-back, although a small amount of intervening
stroma always is present (Fig. 9.4). It is the
degree of glandular crowding, however, that
separates complex from simple hyperplasia.
Cystic glands can involve a portion of the
endometrium in complex hyperplasia, and mix-tures of simple and complex hyperplasia often
occur.
Cytologically the glands in complex hyper-plasia are identical to those in simple hyperpla-sia (Table 9.2). The cells are pseudostratified,
with oval nuclei, small and inconspicuous nucle-oli, and a variable amount of mitotic activity
(Fig. 9.5). Thus, architecture alone separates
simple and complex hyperplasia.
Sometimes the glands are somewhat crowded
and irregular but not densely packed, and it is
not clear whether the process should be termed
simple or complex hyperplasia. When the dis-tinction between complex and simple hyperpla-sia is not clear, we recommend classifying the
lesions as simple hyperplasia.
Atypical Hyperplasia
The diagnosis of atypical hyperplasia is based
on the presence of nuclear atypia. Architec-turally, atypical hyperplasia can have simple
or complex patterns.1;13
In contrast to non-atypical hyperplasia, however, most cases of
atypical hyperplasia have a complex architec-ture with closely packed glands (complex atyp-ical hyperplasia) (Figs. 9.6 to 9.10). The glands
tend to be highly irregular in size and shape
182 9. EH, EIC, and Epithelial Cytoplasmic Change
Figure 9.3. Complex hyperplasia. The glands are closely packed, lacking the abundant stroma seen in a
simple hyperplasia. There is no atypia.
Figure 9.4. Complex 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.
Figure 9.5. Complex 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 prolif-erative phase. Nucleoli are indistinct.
Figure 9.6. Complex atypical hyperplasia. Left: The
glands are closely spaced, with little intervening
stroma. Right: The lining epithelium shows atypia,
characterized by rounded, stratified nuclei. The cyto-plasm is eosinophilic.
Figure 9.7. Complex atypical hyperplasia. The
glands are highly irregular, but a small amount of
stroma encompasses each gland. The cells show
atypia with stratified nuclei that have prominent
nucleoli. Inset: The nuclei are vesicular with chro-matin clumped along the nuclear membrane. The
cells also have abundant, eosinophilic cytoplasm.
Figure 9.8. Complex atypical hyperplasia. The com-plex glands are closely spaced but each gland has
a stromal investment with basement membrane. Pap-illary tufts of eosinophilic cells project into the lumen
of many glands. The glandular cells have eosinophilic
cytoplasm and are stratified.
Figure 9.9. Complex atypical hyperplasia. The glands are not highly convoluted but are closely spaced and
vary in size and shape. Nuclei are stratified.
(Table 9.3). Papillary infolding or tufts lacking
a fibrovascular core may project into the lumen
(Figs. 9.7 and 9.8). Even with apparent back-to-back glandular crowding, each gland has a base-ment membrane with a thin rim of stroma
separating it from adjacent glands. In some
cases, however, the glands are widely dispersed
(simple atypical hyperplasia) (Fig. 9.11). Glands
displaying no cytologic atypia may be admixed
with those showing cytologic atypia.
The specific nuclear features of atypical
hyperplasia include stratification, nuclear
enlargement with altered chromatin, and nucle-oli (Table 9.3). The nuclei characteristically
show true stratification ranging from two to
four layers, with loss of polarity in relation to
the basement membrane, giving an appearance
of disarray to the nuclei that contrasts with the
pseudostratification in nonatypical hyperplasia.
The nuclei are enlarged and rounded rather
than oval and may have irregular nuclear mem-branes (Figs. 9.7, 9.10, and 9.11). The chromatin
186 9. EH, EIC, and Epithelial Cytoplasmic Change
Figure 9.10. Complex atypical hyperplasia. High
magnification shows glandular nuclei with features
of atypia. The nuclei are rounded and vesicular, with
prominent nucleoli. They have a haphazard distribu-tion, losing their orientation to the underlying base-ment membrane. The cells contain a moderate
amount of pale pink cytoplasm.
Table 9.3. Morphologic features of atypical
hyperplasia.
Cytologic featuresa
Nuclei
Stratification with loss of polarity
Enlarged, rounded with irregular shapes
Coarsening of chromatin creating a vesicular
appearance
Prominent nucleoli
Mitotic activity, variable amount
Cytoplasm
Eosinophilia, diffuse or focal
Glands
Irregular, variable size, some dilated
Simple atypical hyperplasia
Haphazardly spaced in abundant stroma
Complex atypical hyperplasia
Closely spaced with decreased stroma
Highly irregular outlines
Frequent associated features
Papillary infoldings into glands (no bridging)
Decreased stroma
Ciliated cells
Squamous metaplasia
a
Atypical nuclei should be readily apparent, involving
most of the cells lining affected glands.
Figure 9.11. Simple atypical hyperplasia. Several of
the glands are cystic, and there is a moderate amount
of intervening stroma. Some of the cystic glands lack
nuclear atypia, but the lesion still represents atypical
hyperplasia. Inset: The nuclei show atypia. They are
round, vesicular, and stratified. In this case the cells
lack abundant eosinophilic cytoplasm.
is centrally dispersed and forms clumps along
the nuclear membrane, resulting in a distinctive
vesicular appearance that is highly characteris-tic of endometrial atypia. Nucleoli may be
prominent.
Epithelial cellular changes (metaplasia)
often are found in atypical hyperplasia (see
later, Epithelial Cytoplasmic Change). The
cytoplasm of the atypical glandular cells often
is abundant and eosinophilic. This eosinophilia,
a helpful feature when present, is not specific
for atypical hyperplasia. Sometimes the cells in
atypical hyperplasia are highly stratified, yet
the eosinophilia of the cytoplasm is less pro-nounced. Ciliated cells frequently are seen, at
least focally, and other epithelial changes, such
as secretory or mucinous change, occasionally
are present. Squamous metaplasia can be focal
or extensive in atypical hyperplasia. When pre-sent, squamous metaplasia sometimes fills and
expands the glands, accentuating the crowded
appearance and leaving only a partial rim of
columnar gland cells (Fig. 9.12). Often the
squamous cells partially bridge the lumen,
yielding an apparent cribriform pattern. The
squamous epithelium by itself has no influence
on prognosis, however. The cytologic features
and architecture of the glands, not the pre-sence of squamous epithelium, determine the
diagnosis.
Atypical hyperplasia can be focally present
in tissue along with nonatypical hyperplasia.14
The minimal criteria for the diagnosis of focal
atypia have not been defined. Nonetheless, for
focal atypia to be a significant finding, it should
be readily discernible in a background of
clearly hyperplastic glands. In equivocal cases
where there is a question of focal atypia in a
background of simple or complex hyperplasia,
there often are atypical nuclei focally distrib-uted in many glands. In other cases the appar-ent atypia is confined to only a few glands. In
Endometrial Hyperplasia 187
either case a diagnosis of atypical hyperplasia
is best limited to those cases in which clearly
atypical nuclei are readily identified without
diligent searching. In equivocal cases, we rec-ommend that atypia not be diagnosed unless
clearly atypical nuclei involve most of the
epithelium, lining several well-visualized glands
in cross section. Surface epithelium should be
avoided in assessing the presence of atypia.
Atypia cannot be reproducibly subdivided or
graded into categories such as mild, moderate,
and severe.
In an attempt to improve on diagnostic accu-racy, a morphometric approach to carcinoma
precursors based on computerized morphome-try has been proposed.31;37
Most often quoted is
a “D-Score” for “multivariate discriminant
score.” This “D-Score” considers volume per-centage stroma (VPS), gland branching/convo-lution, and nuclear variation. A low D-Score is
associated with clonality. Interestingly, VPS
appears to be a potentially useful parameter,
with a VPS below 55% likely to be present in
lesions that are precursors to carcinoma. Cur-rently, it appears that the finding of an altered
gland/stroma ratio as defined by a volume per-centage stroma (VPS) of less than 55% also can
help in identifying atypical hyperplasia in
equivocal cases.33
Thus, although morphometry
is not the standard used to assess hyperplasia in
most laboratories, the general concept of a low
VPS of less than 55% can be a useful parame-ter to include in assessing some cases.
Differential Diagnosis
A number of artifacts, as well as a variety of
benign and malignant lesions, can be confused
with hyperplasia, especially in endometrial
biopsies. Artifactual changes to glands include
fragmentation during biopsy or curettage,
active bleeding with stromal collapse, and poor
188 9. EH, EIC, and Epithelial Cytoplasmic Change
Figure 9.12. Complex atypical hyperplasia with
squamous change. Many of the glands are partially
filled with nonkeratinizing squamous epithelium that
bridges the lumen. There is central necrosis of the
squamous change in the gland at the lower right of
the field, a finding that has no significance in the
diagnosis. Each gland is separated by a thin rim of
stroma.
orientation. With any of these artifacts, the
glands can appear irregular and crowded on
casual inspection. Fragmented proliferative or
normal late secretory glands may become
closely positioned, giving the illusion of
crowded, disorganized glands with irregular
shapes and sizes. The artifact of “telescoping,”
resulting in a “gland within a gland” appear-ance, frequently occurs in association with frag-mentation (see Chapter 2). This change can be
mistaken for hyperplasia, as it often occurs in
proliferative endometrium. Glandular and
stromal breakdown and bleeding distorts the
tissue, causing irregular crowding of glands
around areas of collapse that can be mistaken
for hyperplasia (see Chapter 5). Likewise,
basalis has irregular glands that focally resem-ble the glands in hyperplasia. These potential
pitfalls of interpretation are avoided by ensur-ing that the tissue on which the diagnosis is
based has intact glands and stroma without
areas of breakdown. Surface epithelium is a
very important anatomic landmark to orient
the tissue and can help steer one away from
artifactual changes (see Chapter 2).
Disordered proliferative endometrium fre-quently enters into the differential diagnosis
of hyperplasia, especially simple hyperplasia.
Some pathologists use the term “disordered
proliferative” to avoid assigning the term
“hyperplasia” to a patient’s case. Disordered
proliferative endometrium has mild irregulari-ties in gland patterns that do not fulfill the
quantitative criteria for simple hyperplasia.
Often this change results from estrogen stimu-lation that leads to focal glandular irregularities
(see Chapter 5). Sometimes it might also be
applied to mildly irregular proliferative
endometrium that is difficult to classify in a
small sample but may reflect the presence of a
polyp or other focal benign abnormality that is
not fully represented in the sections. We use the
term when only a few glands are dilated or
branched, being confined to no more than scat-tered foci within the functionalis. Other areas
show tubular to tortuous proliferative glands.
With diffuse glandular irregularities, the
process is better classified as hyperplasia.
Polyps are another frequent source of confu-sion in the differential diagnosis of hyperplasia.
Many polyps represent focal hyperplasia of the
basalis, which contains irregular glands (see
Chapter 8). In addition, they may show ciliated
cell or squamous change. Nonetheless, they
are separated from hyperplasia because they
are generally not estrogen-related abnorma-lities. In general, polyps are focal lesions,
and the surrounding endometrium is normal.
The polypoid shape, dense stroma, and thick-walled vessels are helpful features in re-cognizing the ordinary polyp. In summary,
it is the focal nature of the polyp that sepa-rates this lesion from the more diffuse hyper-plasia. Sometimes, however, it is difficult to
make this distinction with certainty in a
small biopsy. Repeat curettage and hys-teroscopy may be necessary to establish the
correct diagnosis.
In atypical hyperplasia the differential diag-nosis is broader than in nonatypical hyperpla-sia. At one end of the spectrum, the cytologic
changes of atypia must be distinguished from
benign abnormalities, such as eosinophilic syn-cytial change. At the other end of the spectrum,
the differential diagnosis includes well-differentiated adenocarcinoma, as both lesions
can be composed of closely packed glands with
cytologic atypia. Typically, the benign cellular
changes that mimic atypia are those that result
in cytoplasmic eosinophilia, as the cells of atyp-ical hyperplasia also frequently have
eosinophilic cytoplasm [see later, Epithelial
Cytoplasmic Change (Metaplasia)].
Endometritis may at times result in glandular
changes that mimic hyperplasia with atypia (see
Chapter 7). In cases with marked inflammation,
especially those with acute and chronic inflam-mation, the glands will show reactive changes
with an irregular distribution in a reactive,
spindle stroma. The reactive process includes
cytologic changes with enlarged, stratified
nuclei, but these are generally limited findings.
With endometritis the glands are not irregular
and crowded unless there is fragmentation arti-fact. Usually these reactive changes associated
with inflammation occur in premenopausal
patients.
Chronic inflammation with plasma cells can
be seen in hyperplasia, and sometimes the
inflammatory infiltrate is striking. This inflam-Endometrial Hyperplasia 189
mation in hyperplasia probably occurs sec-ondary to the constant abnormal bleeding or
the polypoid growth of the tissue, as either of
these processes can dilate the internal os. When
the internal os is dilated, the endometrium is
susceptible to infection with the inflammatory
response. Consequently, it is important to rec-ognize the architectural and cytologic charac-teristics of hyperplasia and not dismiss the
changes as only chronic endometritis.
The atypical polypoid adenomyoma (see
Chapter 8) also enters into the differential diag-nosis, because in this lesion the glands are
highly irregular and the cytologic changes of
the epithelium are similar to those of atypical
hyperplasia. In atypical adenomyoma the
smooth muscle cells around the glands set this
lesion apart from atypical hyperplasia. Instead
of endometrial stroma, as is seen in hyper-plasia, the glands are more widely separated by
smooth muscle in short, interlacing fascicles.
Immunohistochemical stains for desmin or
actin can assist in the diagnosis of this lesion by
demonstrating the smooth muscle.
Rarely, the endometrium may show papillary
proliferations composed of papillary processes
with fibrovascular cores and a lining of cuboidal
to low columnar epithelium with no atypia and
little, if any, mitotic activity (Fig. 9.13).4
This
change, termed “papillary proliferation”4
or
“papillary hyperplasia,”40
is extremely unusual
and has received little clinicopathologic study
with long-term follow-up. One small study
found that these lesions involved endometrial
polyps in two thirds of cases.40
Both simple and
complex papillary patterns were identified and
other cytoplasmic changes were invariably
present. These papillary proliferations appear
190 9. EH, EIC, and Epithelial Cytoplasmic Change
Figure 9.13. Focal papillary proliferation in a
benign endometrial polyp. A focus of glands near the
surface of a polyp show small papillary tufts pro-jecting into the lumens. Inset: At high magnification
the papillae have fibrovascular cores and show no
cytologic atypia. This “simple papillary” pattern,
when focal in a polyp, is benign and has no known
clinical consequence.
to be benign alterations that may be adequately
treated by curettage or polypectomy.
In addition to these specific alterations that
may be confused with atypical hyperplasia,
there are occasional situations in which normal,
nonhyperplastic endometrium can be confused
with an atypical proliferation. Often in such
cases, the cells lining glands of proliferative
endometrium can appear atypical at high mag-nification; the nuclei seem to be stratified and
rounded, and show a coarse chromatin distrib-ution. Artifactual distortion of the tissue also
may yield changes that superficially resemble
atypical proliferations. For example, there is an
infrequent but peculiar artifact of biopsies in
which the glandular cells appear to be stratified,
with a hobnail-like pattern (Fig. 9.14). In
addition, mitotically active cells protrude into
the lumen. This hobnail-like artifact usually
involves only a few glands, generally occur-ring at the edge of tissue fragments while the
remainder of the tissue is free of the abnor-mality. It usually is found in fragmented prolif-erative endometrium and can be mistaken for
atypia unless the overall pattern of normal
gland architecture is recognized and frag-mented areas are avoided.
When atypia is suspected, it is important to
identify areas with intact glands and surround-ing stroma. If atypical hyperplasia is present,
the glands should have the architectural as well
as the cytologic features that establish the diag-nosis. If the glands in these foci are tubular and
lack the irregular outlines and altered gland-to-stroma ratio of hyperplastic glands, the appar-ent atypia probably has no clinical significance.
Also, if the nuclei in areas with intact glands
and stroma lack atypical features, then the diag-Endometrial Hyperplasia 191
Figure 9.14. Hobnail-like artifact of proliferative
endometrium. Curetting specimen of proliferative
endometrium shows a focus where glandular epithe-lium appears stratified and disorderly. Although
appearing worrisome at first glance, this artifact typ-ically occurs as a focal finding at the edge of tissue
fragments in curettings. This change does not repre-sent atypia. Inset: The glandular cells have a hobnail
appearance as they become detached. A few of the
cells show mitotic figures.
nosis of atypical hyperplasia is suspect. An
occasional case of proliferative endometrium
can show foci of apparent crowded glands in
which the nuclei look enlarged and vesicular,
that is, “atypical.” In such cases areas of obvious
proliferative endometrium show nuclei that
look identical, indicating that this does not rep-resent atypical hyperplasia, and showing the
importance of comparing the “atypical” cells to
normal proliferative endometrium in the same
specimen. Several levels through the tissue
block can help resolve equivocal cases. If step
sections do not resolve the question, then addi-tional sampling such as a dilation and curettage
may be necessary, especially if the first speci-men was from an office-based biopsy.
Finally, 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 un-usual in premenopausal women unless they
have a history of anovulation associated with
obesity or polycystic ovaries.8–10
On the other
hand, atypical glandular proliferations become
more common in perimenopausal or post-menopausal patients. Consequently, foci of
apparent gland cell atypia in premenopausal
women should be viewed very conservatively,
considering the possibility of artifact. In the
postmenopausal patient, critical study also is
necessary, but with the consideration that
subtle gland cell changes may actually repre-sent atypia.
Once true atypia is identified and benign
lesions that mimic hyperplasia are excluded, the
differential diagnosis includes well-differenti-ated adenocarcinoma. A diagnosis of well-
differentiated carcinoma is established easily
when there is myometrial invasion, but this is a
very rare finding in curettings. Thus, a diagnosis
of carcinoma is based on identifying invasion of
endometrial stroma that can be a subtle change
in well-differentiated neoplasms. There are
three criteria, any of which identifies endome-trial stromal invasion: (1) an irregular infiltra-tion of glands associated with an altered
fibroblastic stroma (desmoplastic response); (2)
a confluent glandular pattern in which individ-ual glands, uninterrupted by stroma, merge and
create a cribriform pattern; and (3) an exten-sive papillary pattern.1;2;41
Increasing degrees of
nuclear atypia, mitotic activity, and stratifica-tion in curettage specimens also are associated
with a higher frequency of carcinoma but are of
limited value compared to the main criterion of
stromal invasion. These three criteria for deter-mining the presence of invasion allow the diag-nosis of atypical hyperplasia or carcinoma to
be made more objectively and are discussed
further and illustrated in the next chapter.
Endometrial intraepithelial carcinoma, dis-cussed later in this chapter, also must be distin-guished from atypical hyperplasia as it is a
different type of precursor lesion.
Behavior
Hyperplasia without atypia, simple or complex,
usually is a self-limited lesion that will regress.
Atypical hyperplasia, however, is associated
with a high risk for the development of adeno-carcinoma.5;11–17;42–44
In one study that examined
untreated hyperplasia in detail, 80% of both
simple and complex hyperplasia without atypia
regressed (Table 9.4).13
Furthermore, the risk
of progression to carcinoma was slight, 1% in
192 9. EH, EIC, and Epithelial Cytoplasmic Change
Table 9.4. Follow-up comparing cytologic and architectural abnormalities in 170 patientsa
Progressed to
No. of
Regressed Persisted carcinoma
Type of hyperplasia 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)
a
Adapted with permission of Kurman et al. 1985.
simple hyperplasia and 3% in complex hyper-plasia. Approximately 60% of the cases of atyp-ical hyperplasia also regressed, but the risk of
progression to carcinoma was significantly
greater compared to hyperplasia without atypia
(Table 9.4). In the same study, 8% of cases
of simple atypical hyperplasia and 29% of cases
of complex atypical hyperplasia progressed to
carcinoma. These differences in progression
rates between simple and complex atypical
hyperplasia did not achieve statistical signifi-cance, but overall there was a 23% rate of pro-gression to carcinoma of atypical hyperplasia
compared to 2% in hyperplasia without atypia,
and this was a statistically significant difference.
Accordingly, the presence of atypia is the most
important prognostic feature for endometrial
hyperplasia. Other studies also find that 17% to
25% of patients who undergo a hysterectomy
soon after the diagnosis of atypical hyperplasia
at biopsy have well-differentiated adenocarci-noma in the uterus. When adenocarcinoma is
present following a biopsy diagnosis of atypical
hyperplasia, the neoplasm is almost always well
differentiated, focal, and either confined to the
endometrium or minimally invasive into the
myometrium.
Endometrial Intraepithelial
Carcinoma
Endometrial intraepithelial carcinoma (EIC) is
the precursor lesion of serous carcinoma.45
The
lesion is characterized by epithelial cells that
show marked nuclear abnormalities identical to
that seen in serous carcinoma of the
endometrium, yet lacking an invasive com-ponent. EIC is commonly seen in endometrium
adjacent to serous carcinoma and also is
observed adjacent to malignant mixed meso-dermal tumors (MMMTs).24
Occasionally, EIC
is seen without an invasive carcinoma present.
The changes often occur within a polyp.46
This lesion also has been termed “carcinoma
in situ” and “uterine surface carcinoma,” but
the term “endometrial intraepithelial carci-noma” is preferred because these lesions may
rarely metastasize, even in the absence of inva-sive carcinoma in the endometrium.47
At times
it may be difficult to determine whether crowed
glands involved by EIC are invasive. In these
cases, if the lesion measures less than 1.0 cm, it
is designated minimal uterine serous carcinoma
(MUSC).48
These small, equivocally invasive
lesions seem to behave the same as EIC.
In contrast to atypical hyperplasia, EIC is
not associated with hyperplasia and does
not typically occur in the clinical setting of
increased estrogen exposure but instead is
seen in atrophic endometrium of older post-menopausal women.45
EIC can show flat or
stratified cells with very high grade nuclei,
sometimes with papillary tufts. The lesion
involves surface epithelium and may extend
into atrophic glands, showing an abrupt transi-tion to normal epithelium (Figs. 9.15 and 9.16).
The cells are polygonal and hobnail with highly
irregular nuclei having coarse to smudged
chromatin. Mitoses, including atypical mitoses,
are readily seen. The cells of EIC are strongly
reactive for p53 by immunohistochemistry,
reflecting overexpression of p53 protein (Fig.
9.17).48;49
This finding correlates closely with
p53 gene mutations, which are seen in more
than 80% of cases of EIC and serous carci-noma.45;50;51
The immunohistochemical prolifer-ation marker, MIB-1, which stains the Ki-67
nuclear protein, shows a very high proliferative
index in EIC and therefore is another useful
diagnostic tool.1;48;52
Differential Diagnosis
The differential diagnosis of EIC in biopsy
specimens includes invasive serous carcinoma
and atypical hyperplasia. EIC is separated from
invasive serous carcinoma by its pattern of
growth, with preservation of normal surface
and glandular architecture as the lesion extends
along the epithelial surfaces. This pattern is in
contrast to the irregular papillary and confluent
gland patterns of frank serous carcinoma (see
Chapter 10). As noted earlier, in small lesions
(<1.0 cm) where it is difficult to determine the
presence of invasion, the term MUSC is
applied.48
EIC also should not be confused with atypi-cal hyperplasia even though both lesions are
composed of nonconfluent glands lined by
Endometrial Intraepithelial Carcinoma 193
Figure 9.15. Endometrial intraepithelial carcinoma.
The highly malignant cells line the surface epithe-lium and involve a gland in atrophic endometrium.
The degree of nuclear atypia far exceeds that seen in
atypical hyperplasia and resembles that seen in
serous carcinoma.
Figure 9.16. 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. Inset:
The cells have markedly enlarged and irregular
nuclei with smudged chromatin and large nucleoli.
atypical epithelium. EIC, in contrast to atypical
hyperplasia, features a much greater degree of
nuclear abnormality with cells showing clearly
malignant features and numerous mitotic
figures including abnormal mitotic figures.
Atypical hyperplasia, in contrast, shows lesser
degrees of nuclear abnormalities. In addition,
EIC often occurs in a background of atrophic
endometrium while atrophy is unusual in asso-ciation with atypical hyperplasia. Immunostains
for p53 and Ki-67 protein are helpful as EIC,
but not atypical hyperplasia, is typically
strongly and diffusely positive for both p53 and
Ki-67.
Behavior
EIC is often found in the presence of frank
invasive carcinoma, typically serous type. EIC
or serous carcinoma confined to the
endometrium has a generally very good prog-nosis.48
However, a few reported cases of pure
EIC without an invasive component have been
reported in which the patient had metastases
involving extrauterine organs.47;48;53
Even
microscopic disease was associated with recur-rence and death despite aggressive chemother-apy. Consequently, the diagnosis of EIC in an
endometrial biopsy is an indication for further
evaluation of the uterus and surgical staging at
the time of hysterectomy.
Epithelial Cytoplasmic Change
(Metaplasia)
Epithelial cytoplasmic alterations, commonly
designated metaplasia, often occur in the
endometrium. The term “metaplasia” refers to
transformation of cells to a type not normally
found in an organ. By this definition, most of
the alterations commonly classified as endome-trial metaplasia do not qualify as such. Conse-Epithelial Cytoplasmic Change (Metaplasia) 195
Figure 9.17. Endometrial intraepithelial carcinoma, p53 immunostaining. The glands lined by intraepithe-lial carcinoma show diffuse and strong reactivity for p53 protein in the high-grade nuclei.
quently, some of the cytologic transformations
of the epithelium previously referred to as
endometrial metaplasia are better classified as
a “change.” The latter term has the advantage
of offering a descriptive designation without
implying a specific mechanism of development.
Because these “changes” are especially com-mon in hyperplasia, it is important that they
be recognized and clearly separated from more
significant glandular abnormalities. “Changes”
vary from squamous differentiation to other
benign cytoplasmic transformations, such as
secretory-like vacuolization, to degenerative
or reparative processes. Eosinophilic syncytial
change is an example of this latter group.54
This change, discussed in greater detail in
Chapter 5, appears to be a degenerative/
regenerative process related to endometrial
breakdown.
There are five general types of cytoplasmic
transformation that occur in the endometrium.
These are squamous, ciliated cell, eosinophilic,
mucinous, and secretory (clear cell and hobnail
cell) change.1;21;39
The terminology for these
changes continues to evolve as greater experi-ence with them is gained. For example, ciliated
cell change has also been termed tubal meta-plasia, and the terms eosinophilic and pink cell
change are synonymous. The relative frequen-cies of these cytoplasmic changes are difficult to
determine. Ciliated and squamous changes are
the most widely recognized, but in our experi-ence eosinophilic change is the most common
of the nonspecific cellular changes.
Squamous and ciliated cells are generally
found in endometria that show signs of estro-genic stimulation, especially hyperplasia.20–22
They are also found in low-grade endometrial
carcinoma,55;56
although squamous differentia-tion can occur in association with all grades
of endometrial carcinoma. The association of
cytoplasmic change with hyperplasia and carci-noma suggests that many forms of cytoplasmic
differentiation or transformation are induced
by chronic estrogen stimulation. Cytoplasmic
changes also may be associated with trauma,
polyps, or inflammation, however. Occasionally
these cellular changes may be found in atrophy
with no other known underlying pathology. It is
important to recognize the various types of
cellular change and determine whether they
accompany hyperplasia or not, as these changes
by themselves have no neoplastic potential.
Squamous differentiation (squamous meta-plasia) often is nonkeratinizing, forming
so-called morules because of their three-dimensional resemblance to mulberries.57
The
squamous epithelium is rarely keratinized in
hyperplasia, keratinization occurring more
frequently in adenocarcinoma with squamous
differentiation (see Chapter 10). The nonkera-tinizing morules have a characteristic appear-ance, forming solid nests of bland eosinophilic
cells that fill gland lumens (Fig. 9.12) (see also
Chapter 8, Figs. 8.5, 8.6, and 8.13, and Chapter
10, Fig. 10.16). The cells have uniform, round to
oval nuclei with small nucleoli and rare or
absent mitoses. The nuclei are centrally placed
in dense, eosinophilic cytoplasm. When the
squamous change forms morules, the gland is
largely filled with a round to oval mass of
uniform cells with indistinct cell borders. The
intraglandular nests of squamous epithelium
may show central necrosis, but this feature has
no effect on the diagnosis or prognosis of the
lesion. Squamous change predominantly occurs
within gland lumens, and in most cases sur-face epithelium shows minimal involvement.
Surface squamous change is occasionally
observed secondary to inflammation.
Ciliated cell change (tubal metaplasia) is
arguably not a true metaplasia, as ciliated cells
are normally present along the surface epithe-lium, being most numerous in proliferative
endometrium.58
Glands lined by ciliated cells
are not normal, however. Ciliated cells usually
are prominent in endometrium stimulated by
unopposed estrogen. Hyperplasia may or may
not be present. These cells often are inter-spersed in small groups among nonciliated
columnar cells, but sometimes they are exten-sive and line most of the gland. Ciliated cells
have pale to eosinophilic cytoplasm (Fig. 9.18).
The luminal border of these cells may show a
cuticle of dense cytoplasm formed by the ciliary
basal bodies (Fig. 9.19). Often the nuclei are
mildly stratified, yet they remain cytologically
bland with round to oval shapes, an even chro-matin distribution, and small nucleoli. The
rounding and slight nuclear enlargement that
196 9. EH, EIC, and Epithelial Cytoplasmic Change
Figure 9.18. Ciliated cell change. A gland in simple
hyperplasia is lined by ciliated cells. The cytoplasm
is eosinophilic and some nuclei are enlarged and
rounded. The nuclei lack features of atypia, and they
have smooth, uniform contours, a delicate chromatin
pattern, and tiny nucleoli.
Figure 9.19. Ciliated cell change. The epithelial
lining of this endometrial gland resembles that of the
fallopian tube. The luminal border is sharply demar-cated where some of the cells have a dense cyto-plasmic cuticle. Some nuclei are slightly enlarged
and rounded, but they lack the coarsely clumped
chromatin and irregular nuclear contours seen with
atypia.
characteristically occurs should not be consid-ered as evidence of atypia. Mitoses generally do
not occur in ciliated cells.
Eosinophilic (pink) cell change also is
common. This change actually represents
several types of cytoplasmic transformation.
Eosinophilic cells may be a variant of ciliated
cells, squamous cells, or oncocytes as well as
eosinophilic syncytial change.1;4
All of these
cytoplasmic transformations are without clini-cal consequence, per se. Eosinophilic cytoplasm
also is a frequent feature of glands in atypical
hyperplasia and low-grade adenocarcinoma,
so it is important to determine if there is a
coexisting neoplastic process.
Eosinophilic cell change that resembles cili-ated cell change is common. In this situation the
cells are columnar or slightly rounded and have
a moderate amount of pale pink cytoplasm,
resembling the cytoplasm of ciliated cells but
lacking luminal cilia (Fig. 9.20). Eosinophilic
cell change also merges with squamous change
in some cases; in these instances the cells are
rounded to polygonal and pavement-like,
resembling cells seen in squamous differentia-tion but lacking the solid, morule-like growth
pattern. In other cases eosinophilic cells contain
abundant, granular cytoplasm resembling onco-cytes or Hurthle cells seen in other organs
(Fig. 9.21). Eosinophilic cell change may even
show interspersed cells with a small amount of
cytoplasmic mucin, suggesting overlap with
mucinous cell change. In all these forms of
eosinophilic cell change, the nuclei are often
round rather than oval and somewhat stratified.
Luminal cell borders are sharply demarcated.
The nuclei are smaller and more uniform and
lack the irregular nuclear membrane, chromatin
condensation along the membrane, and promi-nent nucleoli that characterize cells with true
cytologic atypia. As in other forms of cytoplas-mic change, mitoses are extremely rare. Occa-Figure 9.20. Eosinophilic cell change. A gland in a
benign polyp has cells with abundant eosinophilic
cytoplasm and small, round to oval nuclei. The cells
are similar to those seen in ciliated cell change, but
they lack visible cilia. Scattered cytoplasmic vacuoles
contain mucin.
198 9. EH, EIC, and Epithelial Cytoplasmic Change
Epithelial Cytoplasmic Change (Metaplasia) 199
Figure 9.21. Eosinophilic cell change. Endometrium
in postmenopausal patient shows partial replacement
of atrophic epithelium by cells with abundant granu-lar eosinophilic cytoplasm, resembling oncocytes.The
nuclei lack atypical features, and mitotic figures are
absent. This alteration, by itself, has no significance.
sionally, eosinophilic cell change occurs in non-estrogenic patterns such as atrophy (Fig. 21).
As noted in the preceding, eosinophilic syn-cytial change is not a metaplastic transforma-tion, yet it has been commonly described as
such. In several studies this cellular alteration
has been termed “papillary syncytial metapla-sia,” “surface syncytial change,” or an “early”
form of squamous metaplasia.4;39;54;59
The classi-fication of these eosinophilic cells as a meta-plastic phenomenon is attributable to the fact
that the syncytial aggregation of eosinophilic
cells in this change superficially resembles the
cells in squamous metaplasia. Eosinophilic syn-cytial change should not be mistaken for squa-mous metaplasia or interpreted as an “early”
form of squamous differentiation, however. The
constant association of eosinophilic syncytial
change with breakdown and bleeding indicates
that this change is degenerative and regenera-tive rather than metaplastic.54
Syncytial change
is recognized by its prominent localization
along surface epithelium, although it may also
occur in glands (see Chapter 5, Figs. 5.5 to 5.8).
Eosinophilic syncytial change usually is accom-panied by karyorrhectic debris, neutrophils, and
adjacent glandular and stromal breakdown
with stromal collapse. Furthermore, in this
change, nuclei have a haphazard distribution,
whereas with the other cytoplasmic changes,
nuclei generally have a uniform distribution.
Mucinous change is characterized by the
presence of abundant mucinous cytoplasm,
resembling normal endocervical glandular cells
(Fig. 9.22). Often with this change the epithe-lium is also thrown into small papillary projec-tions. This pattern is not as common as the
other cytoplasmic changes and is seen most
often in association with atypical hyperplasia
(Fig. 9.23) or carcinoma. These cells are colum-nar, with basal nuclei and abundant pale
supranuclear cytoplasm that contains mucin.
Histochemical stains, such as mucicarmine or
periodic–acid Schiff with diastase digestion,
Figure 9.23. Mucinous change in atypical hyper-plasia. The cells have a moderate amount of pale
supranuclear cytoplasm containing mucin. The
glands are irregular, closely spaced, and show nuclear
atypia, features of complex atypical hyperplasia.
Figure 9.22. Mucinous change. The glandular epithelial cells contain vacuoles of mucin in the supranuclear
cytoplasm, resembling endocervical cells.
Figure 9.24. Secretory change. The glandular cells
have abundant vacuolated cytoplasm in this
endometrium with glandular and stromal break-down. The patient was not pregnant. In contrast to
mucinous change, cytoplasmic vacuoles are irregular
and the cells have ragged luminal borders.
demonstrate the abundant cytoplasmic mucin.
As in ciliated cell change and eosinophilic
change, the nuclei remain small and uniform,
although they may contain small nucleoli.
Mitotic figures are infrequent. Very rarely
mucinous change can include transformation
into goblet cells, and the change has been des-ignated “intestinal metaplasia.” This goblet cell
change should be differentiated from true
colonic epithelium that rarely is inadvertently
obtained if the uterus is perforated during
biopsy or curettage.
Several studies of various endometrial
mucinous proliferations found that mucinous
change occurs across a morphologic spectrum
from bland cytoplasmic change in simple muci-nous proliferations to mucinous carcinoma.60:60a
The absence of cytologic atypia and architec-tural complexity in simple mucinous prolifera-tions was associated with a low risk of
neoplasia. Alternatively, lesions showing archi-tectural complexity or cytologic atypia but
not clearly neoplastic were associated with the
concurrent or subsequent presence of well-differentiated adenocarcinoma. Consequently,
it is important to identify the background
cytologic and architectural features when
confronted with mucinous change in a biopsy.
When a lesion is encountered that shows
worrisome architectural or cytologic features,
the abnormality could be termed “complex
atypical endometrial mucinous prolifera-tion” with a comment that there is a substantial
risk of well-differentiated carcinoma in the
uterus.
Secretory and clear cell change is very infre-quent once progestin-related effects are
excluded. This is usually a focal alteration,
limited to scattered glands. As the names imply,
the cells contain clear, glycogen-rich cytoplasm
and resemble those found in secretory or ges-tational endometrium (Fig. 9.24). Rarely the
Epithelial Cytoplasmic Change (Metaplasia) 201
cells develop a hobnail pattern with nuclei that
protrude into the gland lumen, resembling the
Arias-Stella reaction (Fig. 9.25). The secre-tory/clear cell change usually occurs in
endometrium that shows estrogenic effects
that range from a proliferative pattern to car-cinoma. Sometimes secretory endometrium
shows extensive cytoplasmic clear cell change
that exceeds the amount of vacuolization seen
during normal luteal phase of the menstrual
cycle, and this, too, can be considered a form of
clear cell change.
Diffuse secretory changes sometimes occur
in hyperplasia,61
and this has been called
“secretory hyperplasia” (Fig. 9.26). This process
can be seen in the premenopausal or peri-menopausal patient with hyperplasia who has
sporadic ovulation or who has been treated
with progestins.61
However, some examples are
found with no evidence of either ovulation or
exogenous progestin use. Regardless of the
cause, in secretory hyperplasia the glands main-tain the disordered architecture of hyperplasia,
but they also show secretory changes with vari-ably vacuolated cytoplasm and luminal secre-tions. Atypia is difficult to recognize in these
cases, because the secretory changes result in
differentiation of the gland cells that creates
a rather bland appearance. In addition,
superimposed secretory changes can increase
the tortuosity of the glands, complicating the
interpretation of gland crowding. In such cases,
rebiopsy may be necessary to assess the
endometrium after the secretory change has
resolved.
Differential Diagnosis
Squamous, ciliated cell, and the various types of
eosinophilic cell change all may superficially
resemble the epithelium in atypical hyper-plasia, or even well-differentiated adenocar-cinoma, because they have pale, often pink
cytoplasm and nuclei that appear to be
stratified.19;55;56
To complicate matters further,
these changes often occur in hyperplastic
202 9. EH, EIC, and Epithelial Cytoplasmic Change
Figure 9.25. Hobnail secretory change. The cells
show a hobnail pattern that resembles the pattern
of secretory exhaustion or Arias-Stella reaction in
which secretory vacuoles are absent and the irregu-lar nuclei bulge into the glandular lumen.
Epithelial Cytoplasmic Change (Metaplasia) 203
Figure 9.26. Secretory change in hyperplasia.
Glands in complex atypical hyperplasia show exten-sively vacuolated cytoplasm indicating secretory
change. The secretory change masks some of the fea-tures of atypia as the cells lose their pseudostratified
nuclei and eosinophilic cytoplasm secondary to cyto-plasmic vacuoles. In this case, nuclear atypia was
better seen in other areas. This pattern often is the
result of progestin therapy prior to biopsy but may
occur in the absence of this history.
endometrium. In fact, metaplasia in the absence
of an underlying proliferative process is quite
uncommon except for surface ciliated cell
change. When these cellular changes occur in
hyperplasia, their recognition and separation
from true glandular atypia require attention to
the nuclear features. With atypia, the nuclei are
enlarged and rounded with central vesicular
chromatin and irregular nuclear membranes.
Usually the nuclei are stratified. These cytologic
findings contrast with the relatively bland
nuclear features of the various cellular cyto-plasmic changes. Mitoses are very infrequent in
the latter, another helpful feature in the differ-ential diagnosis.
Cytoplasmic change may occur in benign,
nonhyperplastic endometrium including prolif-erative or secretory endometrium or in other
conditions such as endometritis or polyps.
Eosinophilic cell change also can occur focally
in otherwise atrophic endometrium. In these
situations, the lack of hyperplastic glandular
architecture is helpful in recognizing these
alterations as incidental processes with no bio-logic significance. Therefore, it is important to
assess the overall configuration of the glands
and to be certain that intact endometrium
without glandular and stromal breakdown is
studied to determine whether or not variations
in cytoplasmic features represent nonspecific
“change” or a more significant lesion.
Sometimes in biopsy specimens, small,
detached fragments of tissue may contain areas
of squamous, eosinophilic, or mucinous change
that are suggestive of a more significant abnor-mality, even in the absence of glands with iden-
tifiable atypia. Such foci are especially worri-some when they occur in postmenopausal
patients, as their endometria should be
atrophic. Detached fragments of squamous or
mucinous epithelium may reflect the presence
of more significant glandular abnormalities,
including atypical hyperplasia or even adeno-carcinoma that has not been adequately
sampled. Further sampling, usually by dilation
and curettage, may be necessary to determine
the significance of the focal alteration.
Secretory changes in hyperplasia must be
distinguished from normal secretory phase
patterns that appear abnormal because of
fragmentation, haphazard orientation, or
crowding (artifactual or real), as these are much
more common than this rare form of hyper-plasia. Orientation of the tissue with regard to
surface epithelium, surrounding stroma, and
basalis is necessary to avoid this pitfall (see
Chapter 2).
Clinical Queries and Reporting
Patients with hyperplasia typically present with
abnormal uterine bleeding, and the diagnosis
establishes a cause for the bleeding. The diag-nosis of hyperplasia should be qualified as to
whether or not atypia is present. Although the
WHO terminology has achieved widespread
use, some pathologists prefer using terminology
to which they have become accustomed. Like-wise, gynecologists who are not familiar with
the current classification may need clarification
of the terminology and correlation with the
terms for hyperplasia that they have used
previously. We recommend using the WHO
terminology primarily and appending other
terminology parenthetically.
If atypical hyperplasia is present, the gyne-cologist will be concerned about the possibility
of adenocarcinoma, as this lesion is a recog-nized risk factor for adenocarcinoma. Even
focal atypia carries a greater risk for the pres-ence or subsequent development of adenocar-cinoma. Conversely, hyperplasia without atypia
has little neoplastic potential, and often it is
useful to append the statement “no atypia
seen” to these diagnoses. Hyperplasia, espe-cially without atypia, can be managed conserv-atively, as these lesions are self limited. Patients
with polycystic ovaries or postmenopausal
patients on hormone replacement therapy do
warrant especially close follow up. Atypical
hyperplasia does not necessarily require hys-terectomy either. This lesion can be managed
medically with suppressive progestin therapy in
young women who wish to retain their fertility
and in older women in whom surgery is con-traindicated. Close follow-up with endometrial
biopsies is necessary to monitor the response
to therapy in patients managed medically,
however.
Separating cases of hyperplasia according to
whether they are simple or complex is of lesser
importance than determining the presence or
absence of atypia. Often areas of simple hyper-plasia and complex hyperplasia are admixed,
and these lesions should be classified as mixed
simple and complex hyperplasia. In contrast,
almost all atypical hyperplasias are complex.
They may coexist with nonatypical hyperplasia,
but the diagnosis should be based on the worst
lesion.
Endometrial intraepithelial carcinoma is a
relatively unusual diagnosis, and clinicians may
not be familiar with this entity. Accordingly, it
is important to indicate the significance of this
diagnosis. The gynecologist should understand
that this lesion is distinctly different from atyp-ical hyperplasia, and, despite lack of invasion,
may be associated with metastatic disease out-side the uterus. Because of the association with
serous carcinoma or, possibly, MMMTs with
serous features, these patients should undergo
a hysterectomy and careful staging. Cellular
cytoplasmic changes (metaplasia) should be
clearly separated from hyperplasia, as they
have no effect on prognosis. Changes such as
ciliated cell change and eosinophilic cell change
usually do not need to be reported as long as
the other underlying conditions are evident.
Eosinophilic syncytial change also does not
need to be reported, as it is simply a marker of
breakdown and bleeding. The importance of
these changes lies in their recognition as benign
cytoplasmic transformations. Generally it is
best that these changes not be specified in the
report, but if they are, it is important to add a
comment to explain that the alteration has no
clinical significance.
204 9. EH, EIC, and Epithelial Cytoplasmic Change
Finally, there are situations in which the
biopsy sample may not be sufficient to com-pletely determine the full extent of the under-lying abnormality. For example, architectural
features may be suggestive, but not conclusive,
of hyperplasia, especially when the specimen is
small or shows extensive fragmentation. Like-wise, nuclear changes may be present that
suggest atypia but are inconclusive. In such
cases it is best to describe the abnormality as
thoroughly as possible, indicating that artifact
or limited sampling precludes a definitive diag-nosis. A descriptive diagnosis that indicates the
uncertainty of the findings is appropriate in
these cases.
At other times the distinction between atyp-ical hyperplasia and well-differentiated adeno-carcinoma may be difficult because of small
amounts of tissue. When the differential diag-nosis is between these two lesions, it is best to
issue a diagnosis of atypical hyperplasia and
indicate that the findings strongly suggest that
a presence of well-differentiated adenocarci-noma is present but that a definitive diagnosis
could not be made based on the submitted
sample. Follow-up and re-biopsy may be
needed to clarify the true nature of the lesion.
References
(1) Ronnett BM, Kurman RJ. Precursor lesions of
endometrial carcinoma. In: Kurman R, ed.
Blaustein’s pathology of the female genital
tract. 5th ed. New York: Springer-Verlag; 2002:
467–500.
(2) Norris HJ, Tavassoli FA, Kurman RJ. Endome-trial hyperplasia and carcinoma. Diagnostic
considerations. Am J Surg Pathol 1983; 7:
839–847.
(3) Silverberg SG. Hyperplasia and carcinoma of
the endometrium. Semin Diagn Pathol 1988;
5:135–153.
(4) Silverberg SG, Kurman RJ. Tumors of the
uterine corpus and gestational trophoblastic
disease. Atlas of tumor pathology, 3rd series,
Fascicle 3. Washington, DC: Armed Forces
Institute of Pathology, 1992.
(5) Kraus FT. High–risk and premalignant lesions
of the endometrium. Am J Surg Pathol 1985;
9 (Suppl):31–40.
(6) Welch WR, Scully RE. Precancerous lesions
of the endometrium. Hum Pathol 1977; 8:
503–512.
(7) Silverberg SG. Problems in the differential
diagnosis of endometrial hyperplasia and car-cinoma. Mod Pathol 2000; 13:309–327.
(8) Chamlian DL, Taylor HB. Endometrial hyper-plasia in young women. Obstet Gynecol 1970;
36:659–660.
(9) Lee KR, Scully RE. Complex endometrial
hyperplasia and carcinoma in adolescents and
young women 15 to 20 years of age: A report
of 10 cases. Int J Gynecol Pathol 1989;8:
201–213.
(10) McBride JM. Pre-menopausal cystic hyper-plasia and endometrial carcinoma. J Obstet
Gynaecol Br Emp 1959; 66:288–296.
(11) Gusberg SB, Kaplan AL. Precursors of corpus
cancer. IV. Adenomatous hyperplasia as stage
0 carcinoma of the endometrium. Am J Obstet
Gynecol 1963; 87:662–677.
(12) Hertig AT, Sommers SC. Genesis of endome-trial carcinoma. I. Study of prior biopsies.
Cancer 1949; 2:946–956.
(13) Kurman RJ, Kaminski PF, Norris HJ. The
behavior of endometrial hyperplasia. A long-term study of “untreated” hyperplasia in 170
patients. Cancer 1985; 56:403–412.
(14) Tavassoli F, Kraus FT. Endometrial lesion in
uteri resected for atypical endometrial hyper-plasia. Am J Clin Pathol 1978; 70:772–777.
(15) Deligdisch L, Cohen CJ. Histologic correlates
and virulence implications of endometrial car-cinoma associated with adenomatous hyper-plasia. Cancer 1985; 56:1452.
(16) Buehl IA, Vellios F, Carter JE, Huber CP. Car-cinoma in situ of the endometrium. Am J Clin
Pathol 1964; 42:594–601.
(17) Campbell PE, Barter RA. The significance of
atypical hyperplasia. J Obstet Gynaecol Br
Commonw 1961; 68:668–672.
(18) Hertig AT, Sommers SC, Bengaloff H. Genesis
of endometrial carcinoma. III. Carcinoma in
situ. Cancer 1949; 2:964–971.
(19) Winkler B, Alvarez S, Richart RM, Crum CP.
Pitfalls in the diagnosis of endometrial neo-plasia. Obstet Gynecol 1984; 64:185–194.
(20) Crum CP, Richart RM, Fenoglio CM. Ade-noacanthosis of the endometrium. A clinico-pathologic study in premenopausal women.
Am J Surg Pathol 1981; 5:15.
(21) Hendrickson MR, Kempson RL. Endometrial
epithelial metaplasias: Proliferations fre-quently misdiagnosed as adenocarcinoma.
Report of 89 cases and proposed classification.
Am J Surg Pathol 1980; 4:525–542.
(22) Blaustein A. Morular metaplasia misdiag-nosed as adenoacanthoma in young women
References 205
with polycystic ovarian disease. Am J Surg
Pathol 1982; 6:223.
(23) Clement PB. Pathology of the uterine corpus.
Hum Pathol 1991; 22:776–791.
(24) Ambros RA, Sherman ME, Zahn CM, Bitter-man P, Kurman RJ. Endometrial intraepithelial
carcinoma: A distinctive lesion specifically
associated with tumors displaying serous dif-ferentiation. Hum Pathol 1995; 26:1260–1267.
(25) Vellios F. Endometrial hyperplasia and carci-noma in situ. Gynecol Oncol 1974; 2:152–161.
(26) Scully RE, Poulson H, Sobin LH. International
histological classification and typing of female
genital tract tumours. Berlin: Springer-Verlag,
1994.
(26a) Silverberg SG, Kurman RJ, Nogales F, Matter
GL, Kubik-Nutch RA, Tavassoli FA. Tumors
of the uterine corpus. Epithelial tumours and
related lesions. In: World Health Organization
Classification of tumours. Pathology and
genetics of tumours of the breast and female
genital organs. Travassoli FA, Devilee P, eds.
Lyon: IARC Press; 2003: 221–232.
(27) Kendall BS, Ronnett BM, Isacson C, Cho KR,
Hedrick L, Diener-West M, et al. Repro-ducibility of the diagnosis of endometrial
hyperplasia, atypical hyperplasia, and well–
differentiated carcinoma. Am J Surg Pathol
1998; 22:1012–1019.
(28) Bergeron C, Nogales FF, Masseroli M, Abeler
V, Duvillard P, Muller-Holzner E, et al. A
multicentric European study testing the re-producibility of the WHO classification of
endometrial hyperplasia with a proposal of a
simplified working classification for biopsy
and curettage specimens. Am J Surg Pathol
1999; 23:1102–1108.
(29) Wright TC, Holinka CF, Ferenczy A, Gatsonis
CA, Mutter GL, Nicosia S, et al. Estradiol-induced hyperplasia in endometrial biopsies
from women on hormone replacement
therapy. Am J Surg Pathol 2002; 26:1269–1275.
(30) Baak JP. The role of computerized morpho-metric and cytometric feature analysis in
endometrial hyperplasia and cancer prognosis.
J Cell Biochem Suppl 1995; 23:137–146.
(31) Baak JP, Orbo A, van Diest PJ, Jiwa M, de
Bruin P, Broeckaert M, et al. Prospective mul-ticenter evaluation of the morphometric D-score for prediction of the outcome of
endometrial hyperplasias. Am J Surg Pathol
2001; 25:930–935.
(32) Mutter GL, Baak JP, Crum CP, Richart RM,
Ferenczy A, Faquin WC. Endometrial precan-cer diagnosis by histopathology, clonal analy-sis, and computerized morphometry. J Pathol
2000; 190:462–469.
(33) Mutter GL. Histopathology of genetically
defined endometrial precancers. Int J Gynecol
Pathol 2000; 19:301–309.
(34) Mutter GL. Endometrial intraepithelial neo-plasia (EIN): Will it bring order to chaos? The
Endometrial Collaborative Group. Gynecol
Oncol 2000; 76:287–290.
(35) Dietel M. The histological diagnosis of
endometrial hyperplasia. Is there a need to
simplify? Virchows Arch 2001; 439:604–608.
(36) Dunton CJ, Baak JP, Palazzo JP, van Diest PJ,
McHugh M, Widra EA. Use of computer-ized morphometric analyses of endometrial
hyperplasias in the prediction of coexistent
cancer. Am J Obstet Gynecol 1996; 174:
1518–1521.
(37) Baak JP, Wisse-Brekelmans EC, Fleege JC,
van der Putten HW, Bezemer PD. Assessment
of the risk on endometrial cancer in hyperpla-sia, by means of morphological and morpho-metrical features. Pathol Res Pract 1992; 188:
856–859.
(38) Hendrickson MR, Kempson RL. Surgical
pathology of the uterine corpus (Major prob-lems in pathology series), Vol. 12. Philadel-phia: W.B. Saunders; 1980.
(39) Buckley CH, Fox H. Biopsy pathology of the
endometrium. 2nd ed. London: Arnold; 2002.
(40) Lehman MB, Hart WR. Simple and complex
hyperplastic papillary proliferations of the
endometrium: A clinicopathologic study of
nine cases of apparently localized papillary
lesions with fibrovascular stromal cores and
epithelial metaplasia. Am J Surg Pathol 2001;
25:1347–1354.
(41) Kurman RJ, Norris HJ. Evaluation of criteria
for distinguishing atypical endometrial hyper-plasia from well–differentiated carcinoma.
Cancer 1982; 49:2547–2557.
(42) Gray LA, Robertson RWJ, Christopherson
WM. Atypical endometrial changes associated
with carcinoma. Gynecol Oncol 1974;2:93–
100.
(43) King A, Seraj IM, Wagner RJ. Stromal inva-sion in endometrial carcinoma. Am J Obstet
Gynecol 1984; 149:10–14.
(44) Terakawa N, Kigawa J, Taketani Y, Yoshikawa
H, Yajima A, Noda K, et al. The behavior of
endometrial hyperplasia: A prospective study.
Endometrial Hyperplasia Study Group. J
Obstet Gynaecol Res 1997; 23:223–230.
206 9. EH, EIC, and Epithelial Cytoplasmic Change
(45) Sherman ME. Theories of endometrial car-cinogenesis: A multidisciplinary approach.
Mod Pathol 2000; 13:295–308.
(46) McCluggage WG, Sumathi VP, McManus DT.
Uterine serous carcinoma and endometrial
intraepithelial carcinoma arising in endome-trial polyps: Report of 5 cases, including 2
associated with tamoxifen therapy. Hum
Pathol 2003; 34:939–943.
(47) Soslow RA, Pirog E, Isacson C. Endometrial
intraepithelial carcinoma with associated peri-toneal carcinomatosis. Am J Surg Pathol 2000;
24:726–732.
(48) Wheeler DT, Bell KA, Kurman RJ, Sherman
ME. Minimal uterine serous carcinoma: Diag-nosis and clinicopathologic correlation. Am J
Surg Pathol 2000; 24:797–806.
(49) Zheng W, Khurana R, Farahmand S, Wang Y,
Zhang ZF, Felix JC. p53 immunostaining as
a significant adjunct diagnostic method for
uterine surface carcinoma: precursor of
uterine papillary serous carcinoma. Am J Surg
Pathol 1998; 22:1463–1473.
(50) Sherman ME, Bur ME, Kurman RJ. p53 in
endometrial cancer and its putative precur-sors: Evidence for diverse pathways of tumori-genesis. Hum Pathol 1995; 26:1268–1274.
(51) Tashiro H, Isacson C, Levine R, Kurman RJ,
Cho KR, Hedrick L. p53 gene mutations are
common in uterine serous carcinoma and
occur early in their pathogenesis. Am J Pathol
1997; 150:177–185.
(52) McCluggage WG. Recent advances in immu-nohistochemistry in gynaecological pathology.
Histopathology 2002; 40:309–326.
(53) Baergen RN, Warren CD, Isacson C, Ellenson
LH. Early uterine serous carcinoma: Clonal
origin of extrauterine disease. Int J Gynecol
Pathol 2001; 20:214–219.
(54) Zaman SS, Mazur MT. Endometrial papillary
syncytial change. A nonspecific alteration
associated with active breakdown. Am J Clin
Pathol 1993; 99:741–745.
(55) Andersen WA, Taylor PT, Fechner RE,
Pinkerton JAV. Endometrial metaplasia asso-ciated with endometrial carcinoma. Am J
Obstet Gynecol 1987; 157:597–604.
(56) Kaku T, Tsukamoto N, Tsuruchi N, Sugihara K,
Kamura T, Nakano H. Endometrial metapla-sia associated with endometrial carcinoma.
Obstet Gynecol 1992; 80:812–816.
(57) Dutra F. Intraglandular morules of the
endometrium. Am J Clin Pathol 1959; 31:
60–65.
(58) Masterson R, Armstrong EM, More IAR. The
cyclical variation in the percentage of ciliated
cells in the normal human endometrium.
J Reprod Fertil 1975; 42:537–540.
(59) Rorat E, Wallach RC. Papillary metaplasia of
the endometrium: Clinical and histopathologic
considerations. Obstet Gynecol 1984; 64:90S–
92S.
(60) Nucci MR, Prasad CJ, Crum CP, Mutter GL.
Mucinous endometrial epithelial prolifera-tions: A morphologic spectrum of changes
with diverse clinical significance. Mod Pathol
1999; 12:1137–1142.
(60a) Vang R, Tavassoli FA. Proliferative mucinous
lesions of the endometrium: Analysis of exist-ing criteria for diagnosing carcinoma in biop-sies and curettings. Int J Surg Pathol 2003;
11:261–270.
(61) Tresserra F, Lopez-Yarto M, Grases PJ, Ubeda
A, Pascual MA, Labastida R. Endometrial
hyperplasia with secretory changes. Gynecol
Oncol 2003; 88:386–393.
References 207
208
The tumors are often associated with hyper-plasia and atypical hyperplasia, conditions that
result from unopposed estrogenic stimulation
such as anovulatory cycles that normally occur
at the time of menopause or in younger women
with the Stein–Leventhal syndrome (polycystic
ovarian disease). Unopposed exogenous estro-gen use as hormone replacement therapy in
older women also predisposes to endometrial
carcinoma that tends to be low-grade. In hys-terectomy specimens these low-grade tumors
generally show minimal myometrial invasion,
although deep invasion can occur in some
cases.5;6
The prognosis is generally good, with a
5-year survival of 80% or better.3
Type II neoplasms represent another, very
different, form of endometrial carcinoma. They
are high-grade neoplasms that do not appear to
be related to sustained estrogen stimulation.1;2;4
Tumors in this group account for 15% to 20%
of all endometrial carcinomas. The prototypical
type II neoplasm is serous carcinoma, but other
histologic subtypes include clear cell carcino-mas and other carcinomas that show high-grade
nuclear features. They tend to occur in older
postmenopausal women, are not associated
with atypical hyperplasia, and often occur in
atrophic endometrium. Endometrial intra-epithelial carcinoma (EIC), the putative pre-cursor lesion, is frequently associated with
serous carcinoma (see Chapter 9). Serous
carcinoma usually invades the myometrium
deeply, permeates lymphatic and vascular chan-nels, and not infrequently has spread beyond
the uterus at the time of hysterectomy.
10
Endometrial Carcinoma
Endometrial adenocarcinoma is the most com-mon malignant tumor of the female genital
tract in the United States. This neoplasm rep-resents a biologically and morphologically
diverse group of tumors, with differing patho-genesis.1–4
These tumors have two basic clinico-pathologic forms, type I and type II. These two
forms of endometrial carcinoma display differ-ent clinicopathologic, immunohistochemical,
and molecular biologic features suggesting two
pathways of carcinogenesis.
Type I carcinomas are generally well to mod-erately differentiated and account for 80% to
85% of all endometrial carcinomas. The typical
patient is an obese, hypertensive, and diabetic
perimenopausal or postmenopausal woman.
Important Issues in Interpretation of
Biopsies  . . . . . . . . . . . . . . . . . . . . . . . . . . . 209
Criteria for the Diagnosis of
Well-Differentiated Endometrial
Adenocarcinoma  . . . . . . . . . . . . . . . . . . 209
Benign Changes that Mimic
Carcinoma  . . . . . . . . . . . . . . . . . . . . . . . 214
Malignant Neoplasms—Classification,
Grading, and Staging of the Tumor  . . . . . . 216
Classification  . . . . . . . . . . . . . . . . . . . . . 216
Grading  . . . . . . . . . . . . . . . . . . . . . . . . . 216
Clinically Important Histologic
Subtypes  . . . . . . . . . . . . . . . . . . . . . . . . 221
Staging  . . . . . . . . . . . . . . . . . . . . . . . . . . 236
Endometrial Versus Endocervical
Carcinoma  . . . . . . . . . . . . . . . . . . . . . . . 237
Metastatic Carcinoma  . . . . . . . . . . . . . . 238
Clinical Queries and Reporting  . . . . . . . . 239
Although most endometrial carcinomas tend
to be either type I or type II, there are cases
that do not fall neatly into either category. For
instance, some endometrioid carcinomas are
very high grade. Also, some endometrioid car-cinomas can show foci of serous or clear cell
carcinoma.
Clinicopathologic studies over the past
several years have shown the importance of
recognizing specific histologic subtypes and
accurately grading carcinomas to help pre-dict outcome and direct treatment.7–26
Most
endometrial tumors diagnosed on a biopsy
are subsequently treated by extrafascial total
abdominal hysterectomy and bilateral salp-ingo-oophorectomy that allows precise
surgical–pathologic staging.27
Nonetheless,
identification of more aggressive tumors is
important at the time of biopsy, as these
neoplasms have greater potential for metasta-tic spread, including involvement of the peri-toneal surfaces. The aggressive neoplasms are
high-grade carcinomas that merit surgical
staging. The clinical relevance of the histologic
classification and grading of endometrial carci-noma is reflected in the revised World Health
Organization (WHO) Histologic Classifica-tion28;28a
and the International Federation of
Gynecology and Obstetrics (FIGO) staging
system.29
The morphologic diversity of endometrial
carcinoma can lead to problems in the diagno-sis of carcinoma in biopsies and curettings. For
low-grade carcinoma, distinction from atypical
hyperplasia and other benign lesions that
mimic carcinoma is an important issue. Identi-fying and properly classifying aggressive,
clinically significant histologic subtypes of
endometrial carcinoma is a second important
area of biopsy interpretation. Another problem
in biopsy interpretation is ascertaining whether
the tumor originates in the endometrium or the
endocervix. This chapter addresses the general
classification, staging, and grading of endome-trial carcinoma, the differential diagnosis of
benign lesions versus low-grade carcinoma, and
the classification of different types of carci-noma once the diagnosis of carcinoma has been
established.
Important Issues in
Interpretation of Biopsies
There are two major concerns in the evaluation
of endometrial biopsies from the standpoint of
a diagnosis of carcinoma. First, is the lesion
benign or malignant? Second, if the biopsy con-tains a malignant tumor, what is the grade, his-tologic subtype, and is it a primary endometrial
or an endocervical carcinoma?
Criteria for the Diagnosis of
Well-Differentiated Endometrial
Adenocarcinoma
Because a diagnosis of carcinoma will have an
important impact on clinical management, it is
necessary for the pathologist to be familiar with
the minimal histologic criteria for that diagno-sis. One of the most problematic areas is the dis-tinction of atypical hyperplasia (see Chapter 9)
from well-differentiated adenocarcinoma.30–33
Separation of these entities is based on identi-fication of specific morphologic criteria that
establish the diagnosis of low-grade carcinoma.
Hyperplasia without atypia generally is not a
problem in the differential diagnosis, as these
forms of hyperplasia do not have nuclear
atypia. It is important, however, to accurately
separate carcinoma from other, benign changes
that mimic neoplasia, including tissue artifacts
and pregnancy-related changes.
Diagnosis of low-grade adenocarcinoma can
be difficult at times, because these tumors do not
always show clear-cut destructive stromal inva-sion. Furthermore, invasion into myometrium
is rarely, if ever, demonstrated in biopsies.
Nonetheless, invasion is a logical criterion for
separating frank adenocarcinoma from atypical
hyperplasia or other lesions that mimic
adenocarcinoma.
For practical application, specific patterns of
stromal and epithelial alterations have been
defined that reflect “endometrial stromal inva-sion” and identify carcinoma.31;34
There are
three separate features, any of which indicates
stromal invasion in low-grade glandular
proliferations:
Important Issues in Interpretation of Biopsies 209
1. A confluent glandular pattern in which
individual glands, uninterrupted by stroma,
merge and create a cribriform arrangement;
2. An irregular infiltration of glands associated
with an altered fibroblastic stroma (desmo-plastic response); and
3. An extensive papillary pattern.
Although quantitative features have limited
usefulness in endometrial biopsy diagnosis,
these specific and objective criteria for invasion
also should be quantitatively significant. There-fore, the glandular proliferation that fulfills
criteria for well-differentiated adenocarcinoma
should be sufficiently extensive to involve at
least one half of a low-power (¥4) field, a dis-tance of 2.0 mm. This guideline helps mainly to
avoid the problem of tangential sectioning or
other artifacts in establishing the diagnosis of
carcinoma. This general rule should not be too
rigidly applied, particularly in scant specimens.
If the features of “stromal invasion” are clearly
evident, a diagnosis of carcinoma should be
made even if the diagnostic area does not
occupy one-half of a low-power field.
Confluent Gland Pattern
This pattern reflects invasion by showing a
complete absence of stroma between glands. At
times a cribriform bridging pattern with true
“gland in gland” formation is present. With
cribriform growth, trabeculae of columnar cells
bridge the lumen, subdividing the lumen into
smaller glandular spaces (Figs. 10.1 and 10.2).
No stroma supports the bridging cells. A con-fluent gland pattern also is represented by large,
irregular glands that interconnect continuously
throughout the field, exceeding the outline of
any acceptable non-neoplastic gland (Fig. 10.3).
Confluent gland patterns should be identified in
areas free of squamous differentiation, as squa-mous morules may bridge gland lumens, but
these do not reflect stromal invasion.
210 10. Endometrial Carcinoma
Figure 10.1. Adenocarcinoma, FIGO grade 1. The
neoplasm shows well-formed glands with a confluent
pattern. The stroma is desmoplastic. The nuclei are
grade 1 with minimal pleomorphism, small nucleoli,
and low mitotic rate.
Figure 10.2. Adenocarcinoma, FIGO grade 1. Confluent gland pattern. Confluent glands with a cribriform
bridging arrangement. There is no stromal support to the epithelium that bridges the glandular lumens.
Figure 10.3. Adenocarcinoma, FIGO grade 1. Confluent gland pattern. Large irregular and branching glands
interconnect continuously throughout the field.
Altered Fibrous or Desmoplastic Stroma
With this change, atypical glands are dispersed
in a reactive fibroblastic mesenchyme rather
than in endometrial stromal cells (Fig. 10.4).
These fibrous stromal cells are different from
normal endometrial stromal cells, being more
spindle-shaped and having elongate nuclei.
The mesenchyme also contains collagen that
compresses the stromal cells and gives an
eosinophilic appearance (Fig. 10.5). The fibrous
change also leads to retraction and distortion of
the normal architecture, resulting in a haphaz-ard glandular pattern. Dense stroma in polyps,
alteration of the stroma associated with marked
inflammation, the stroma of the atypical poly-poid adenomyoma, and stroma of the lower
uterine segment all may mimic the desmoplas-tic response of carcinoma. In these types of
cases, in which desmoplasia is difficult to
evaluate, other features of carcinoma, such
as a confluent gland pattern, should be used
in establishing an unequivocal diagnosis of
malignancy.
Extensive Papillary Pattern
The extensive papillary growth pattern is char-acterized by delicate, elongate, branching
papillary fronds (Fig. 10.6). The fronds have
thin, fibrous cores. These papillary structures
are much more elaborate and branching than
the small papillary tufts that may occur in the
glands of atypical hyperplasia. Papillary tufts
also lack fibrovascular cores. The diffuse
papillary pattern distinguishes this form of
adenocarcinoma from focal alterations, such as
papillary arrangements in eosinophilic syncytial
change (see Chapter 5) or hyperplastic papil-lary proliferations in polyps (see Chapter 9).35
Serous papillary adenocarcinoma is readily sep-arated from this well-differentiated neoplasm
by its high-grade nuclear features and extensive
papillary tufting (see later).
212 10. Endometrial Carcinoma
Figure 10.4. Adenocarcinoma, FIGO grade 1. Desmoplastic stroma. A fibroblastic mesenchyme encom-passes the neoplastic glands.
Figure 10.5. Adenocarcinoma, FIGO grade 1.
Desmoplastic stroma. The stroma supporting the
malignant glands is composed of elongated fibrob-lasts and collagen. Squamous change is present in the
lower portion of the field.
Figure 10.6. Adenocarcinoma, FIGO grade 1. Extensive papillary pattern. This low-grade adenocarcinoma
with a villoglandular pattern forms multiple delicate papillae.
Originally, another pattern, squamous
masses, in which sheets of bland squamous cells
form irregular coalescent nests throughout the
stroma, was proposed as a criterion for inva-sion.31
This pattern is rare by itself, however,
and not useful for recognizing carcinoma in
most cases. When this pattern is present, other
areas usually show confluent gland patterns or
an altered (desmoplastic) stroma.
Applying these criteria for invasion to estab-lish the diagnosis of well-differentiated adeno-carcinoma yields a clinically significant
diagnosis of carcinoma, when present. One
study found that in curettage specimens with
well-differentiated adenocarcinoma, defined by
at least one of these features of invasion, sub-sequent hysterectomy specimens showed resid-ual adenocarcinoma in one half of the cases.31
Frequently the residual carcinoma was well dif-ferentiated but in about one third of cases the
tumor was grade 2 or 3, and in one quarter of
the cases with tumor, the myometrium was
deeply invaded. In another study that used the
same criteria to assess “stromal invasion,” 16%
of patients without stromal invasion in
endometrial samples had myometrial invasion
in the hysterectomy specimens whereas 62.5%
of patients with invasion in the biopsy had
myometrial invasion.36
These studies show the
utility of these criteria for determining the pres-ence of well-differentiated adenocarcinoma.
Benign Changes that Mimic
Carcinoma
Epithelial Cytoplasmic Change
Benign cytoplasmic changes and metaplasia
can be confused with endometrial carcinoma
(see Chapter 9).34;37–42
This differential diagno-sis is most likely to occur in the presence of
prominent squamous or eosinophilic cell
change, especially in specimens with consider-able tissue fragmentation. These cytoplasmic
changes can occur in a variety of conditions,
including polyps, hyperplasia, inflammation and
nonspecific glandular and stromal breakdown,
as well as in carcinoma.34;39–41;43–47
Consequently,
carcinoma should be diagnosed only when
there is a glandular proliferation that fulfills, at
a minimum, the histologic criteria for well-differentiated carcinoma described previously
or if the lesion shows unequivocal features of
malignancy with cytologic features of grade 2
or 3 carcinoma. Detached epithelial fragments,
especially when they lack significant nuclear
atypia, should not be diagnosed as carcinoma.
These epithelial changes may occur in carci-noma, however, and equivocal cases require
processing of additional tissue. Cytoplasmic
change is discussed in greater detail in
Chapter 9.
Atypical Polypoid Adenomyoma
The atypical polypoid adenomyoma can be
confused with carcinoma of the endometrium
because the lesion shows atypical glands,
usually with squamous morules, in smooth
muscle that can be confused with carcinoma
invading the myometrium (see Chapter 8).48–50
Myometrial invasion is rarely seen in biopsy
and curettage specimens, however. Further-more, the orderly pattern of the smooth muscle
of the atypical polypoid adenomyoma contrasts
with the desmoplasia typically associated with
neoplasia. In those rare cases in which the dif-ferential diagnosis includes the atypical poly-poid adenomyoma and adenocarcinoma, it is
important to note that a confluent or cribriform
pattern does not occur in the atypical polypoid
adenomyoma. Immunohistochemical stains for
desmin can help to demonstrate the smooth
muscle in the atypical adenomyoma and distin-guish it from the fibroblastic desmoplasia of
carcinoma.
Pregnancy and the Arias-Stella Reaction
In pregnancy, crowded secretory glands, includ-ing those that display the Arias-Stella reaction,
may resemble carcinoma, especially clear cell
carcinoma. For a premenopausal patient, the
possibility of non-neoplastic lesions, such
as the Arias-Stella reaction, is much more
likely than carcinoma. Thus, the clinical history
often clarifies the diagnosis in questionable
cases.
Several microscopic features also help in the
recognition of the Arias-Stella reaction (see
Chapter 3). This lesion tends to be multifocal,
214 10. Endometrial Carcinoma
admixed with secretory glands, and does not
form a discrete lesion. Usually decidua is
present, too. In addition, this change lacks the
features of invasion, such as confluent glands,
an extensively papillary pattern, or altered
stroma. The nuclei in the Arias-Stella reaction
generally appear to be degenerated, with
smudged chromatin, and mitotic figures almost
never are present. Other pregnancy-related
cytologic changes, such as vacuolated cytoplasm
and optically clear nuclei, also may be found in
the epithelial cells when the Arias-Stella reac-tion is present. In addition, Ki-67 immuno-staining is useful in the differential diagnosis,
as the Arias-Stella reaction has a very low pro-liferative index.
Tissue Artifacts, Contaminants,
and Necrosis
Tissue artifacts can yield worrisome patterns
that may mimic carcinoma. For example, the
artifactual crowding and distortion of glands
that occur during biopsy can result in glands
becoming closely apposed (see Chapter 2).
Likewise, breakdown and bleeding distort the
normal architecture and present a variety of
cytologic alterations, including eosinophilic
syncytial change (see earlier). Cervical contam-inants, especially endocervical squamous meta-plasia, prominent detached fragments of
endocervical epithelium, or microglandular
hyperplasia, may become mixed with endome-trial tissue in curetting specimens and yield a
complex pattern that can mimic carcinoma at
first glance. These tissue artifacts usually are not
a major problem in the differential diagnosis,
as they are generally focal and admixed
with normal endometrium. In menstrual
endometrium, in which there is more diffuse
and extensive breakdown, the possibility of
mistaking the pattern for adenocarcinoma is
greater. In such cases it is important to attempt
to identify secretory glandular changes and
intact endometrium that is not undergoing
breakdown. Furthermore, the clinical history,
including the patient’s age and menstrual
status, can be very helpful in recognizing a men-strual pattern that may not be obvious at first
inspection.
Stromal artifacts also can mimic adeno-carcinoma, especially when they develop a
signet-ring cell morphology.51
Aggregates of
signet-ring cells in the stroma can be seen in
cases with extensive stromal decidual change,
especially following progestin hormone
therapy.51
Rarely, stromal histiocytes may also
have a signet-ring appearance. In these cases
the apparent signet-ring cells lack nuclear
atypia. In addition, these signet-ring cells are
negative for broad-spectrum keratin immuno-stains and are mucin negative by histochemical
stains.
Fragmentation and artifacts also occur in
curettage specimens containing adenocarci-noma. When this occurs, the diagnosis of carci-noma still can be made if the epithelial cells
demonstrate high nuclear grade. Those areas
where glands are attached to surrounding mes-enchymal tissue and are free of the changes of
breakdown and bleeding demonstrate the true
relationship of the glands to each other. To reli-ably identify malignancy in equivocal cases,
however, it is best to evaluate the features that
establish the diagnosis of well-differentiated
adenocarcinoma in clearly intact areas.
Actual tumor necrosis rarely, if ever, occurs
in well-differentiated tumors. This form of
necrosis is often seen in high-grade tumors, but
in these cases the histologic and nuclear fea-tures readily identify the lesion as carcinoma.
The presence of tumor cell necrosis in what
appears to be a low-grade carcinoma is a red
flag that the tumor is, in fact, a high-grade
carcinoma and therefore further sampling is
warranted. Much of the “necrosis” that is
commonly encountered in biopsy material is
unique to the endometrium and actually
reflects breakdown rather than necrotic tumor.
The necrosis associated with bleeding and
breakdown is apparent in low-grade carcinoma
but often occurs in benign conditions, too,
regardless of cause. Because tissue breakdown
and necrosis are so ubiquitous in endometrial
tissue, the finding of this pattern of necrosis is
not helpful in establishing the diagnosis of
neoplasia. Another pattern of necrosis can
be seen in the center of nests of squamous
change (morules). Central necrosis in areas of
squamous change can occur in benign lesions,
Important Issues in Interpretation of Biopsies 215
including any form of hyperplasia, polyps, and
the atypical polypoid adenomyoma. Conse-quently, the necrosis associated with squamous
change has no significance and does not indi-cate malignancy.
Malignant Neoplasms—
Classification, Grading, and
Staging of the Tumor
Once the diagnosis of carcinoma is established,
it is important that the tumor be properly
classified in order to identify aggressive forms
of carcinoma (Table 10.1). Also, because the
biopsy or curettage is generally a blind proce-dure with no direct visualization of the neo-plasm, it is important to ascertain whether or
not the carcinoma is primarily in the
endometrium or whether it arises in the cervix.
Finally, especially if the tumor displays an
unusual pattern, the possibility of a metastasis
from another site should be considered.
Classification
The current WHO histologic classification rec-ognizes several distinct types of carcinoma that
are important to identify in biopsies (Table
10.1).52
Many examples of endometrial adeno-carcinoma have the “typical,” “usual,” or “not
otherwise specified (NOS)” pattern referred to
as “endometrioid” carcinoma. More than one
half of all endometrial carcinomas have this
pattern.3;10;11;42
The term “endometrioid” pro-vides a specific designation for this neoplastic
pattern, clearly separating it from the other
histologic types of endometrial carcinoma.
Because “endometrioid carcinoma” is more
widely applied to primary ovarian cancer,
however, this terminology for primary uterine
neoplasia is potentially confusing to clinicians.
Practically, ordinary adenocarcinoma with this
pattern is frequently classified as endometrial
carcinoma and graded in the microscopic
diagnosis.
From 20% to 30% of endometrial carcino-mas show an “endometrioid” pattern with squa-mous differentiation.18
Previously, these tumors
with squamous differentiation were separated
into two categories: “adenoacanthoma”
denoted tumors that had cytologically benign-appearing squamous epithelium (squamous
metaplasia), and “adenosquamous carcinoma”
denoted tumors that had a cytologically malig-nant squamous component. More recently,
studies have shown that endometrioid carcino-mas with or without squamous epithelium
behave in the same fashion when stratified
according to the grade of the glandular compo-nent.18;19;53
Accordingly, these tumors are best
classified as adenocarcinomas with squamous
differentiation and graded. The terms “adenoa-canthoma” and “adenosquamous carcinoma”
are no longer used for endometrial carcinoma.
The other histologic types of endometrial
carcinoma are relatively infrequent. Serous and
mucinous tumors both account for 10% to 15%
of endometrial primary tumors in most series,
and clear cell adenocarcinoma occurs in no
more than 5% of all cases.3
Grading
In addition to identifying specific histologic
subtypes in biopsies, the histologic grade pro-vides useful prognostic information and assists
in planning treatment. Low-grade tumors
usually are confined to the corpus at the time
of diagnosis, and the overall survival is very
216 10. Endometrial Carcinoma
Table 10.1. World Health Organization (WHO)
classification of endometrial adenocarcinoma.
Endometrioid adenocarcinoma, NOS
Variants
Villoglandular
Ciliated cell
Secretory
Adenocarcinoma, NOS, with squamous differentiation
Mucinous adenocarcinoma
Serous adenocarcinoma
Clear cell adenocarcinoma
Squamous carcinoma
Undifferentiated carcinoma
Mixed carcinomaa
Metastatic carcinoma
NOS, Not otherwise specified.
a
Tumor with greater than 10% of a second cell type.
good. High-grade tumors, in contrast, are more
aggressive and have a poor prognosis. This
latter group is more likely to spread beyond the
corpus and involve the endocervix or extrauter-ine sites at the time of diagnosis.
The traditional grading of endometrial ade-nocarcinoma standardized by FIGO uses a
three-grade system based on architectural fea-tures. Tumors are grade 1 when most (95%) of
the tumor forms glands (Figs. 10.1 to 10.4),
grade 2 when 6–50% of the tumor exhibits solid
growth (Fig. 10.7), and grade 3 when more than
50% of the tumor has a solid growth pattern
(Fig. 10.8). Solid growth in this grading system
consists of areas with glandular components
but limited gland formation. It is important to
avoid areas of squamous change in making this
assessment.
Like all morphologic grading systems, this
grading scheme is somewhat subjective, espe-cially at the cut points between grades. For
example, determination of more than 5% solid
growth that determines grade 2 instead of grade
1 tumor is an estimate and is subject to indi-vidual variation. In addition, significant varia-tion in architectural grade can be seen within
a tumor. Well-formed glands (grade 1 tumor)
may be adjacent to solid masses of grade 3 car-cinoma. Not surprisingly, the heterogeneity of
tumor grade can result in discordance between
curettage and hysterectomy specimens.
Although architectural grading generally
correlates with prognosis, studies suggest the
grading system can be improved by incorporat-ing nuclear grading.54–57
Consequently, FIGO
and WHO Histopathologic Classification modi-fied the standard architectural grading system to
include nuclear grade.27;29
Like architectural
grading, nuclear grading is somewhat subjective.
Nuclei with small, relatively uniform, oval nuclei
and low mitotic activity are grade 1 (Fig. 10.9).
Nuclei with features between grade 1 and grade
Malignant Neoplasms—Classification, Grading, and Staging of the Tumor 217
Figure 10.7. Adenocarcinoma, FIGO grade 2. Although glandular differentiation is readily apparent, a sub-stantial amount (approximately 20%) of the tumor has a solid growth pattern.
Figure 10.8. Adenocarcinoma, FIGO grade 3. A few residual glands are present, but more than 50% of the
tumor shows a solid growth pattern with sheets and nests of malignant epithelial cells.
Figure 10.9. Nuclear grade 1. Nuclei are round to oval with small nucleoli. The mitotic rate is low.
Figure 10.10. Nuclear grade 2. In contrast to grade
1, these nuclei are more pleomorphic and have
coarser chromatin and larger nucleoli, but they are
not as abnormal as grade 3 nuclei. Mitoses are
readily identified.
3 are grade 2 (Fig. 10.10). Highly pleomorphic
nuclei with irregular outlines, macronucleoli,
and numerous, often abnormal mitotic figures
are grade 3 (Fig. 10.11 and Table 10.2).9;58
FIGO recommends that “notable” nuclear
atypia, inappropriate for the architectural
grade, raises an architectural grade 1 or grade
2 tumor by 1 (i.e., grade 1 becomes grade 2, and
grade 2 becomes grade 3). Notable nuclear
atypia was not defined, but one study found
that for clinical significance, nuclear atypia
should be grade 3.54
For example, a tumor with
grade 1 architecture but grade 2 nuclei would
be given a final FIGO grade of 1, whereas an
architectural grade 1 tumor with grade 3 nuclei
should be given a final FIGO grade of 2. It
should be noted that the combination of grade
1 architecture and grade 3 nuclei is quite
uncommon in endometrioid carcinoma, and
therefore it is unusual for nuclear atypia to
upgrade architectural grade. In fact, this com-bination should be regarded as a red flag for the
diagnosis of serous carcinoma (see later).
Other histologic features that influence the
grade according to the FIGO system are:
1. Nuclear grading takes precedence in serous,
clear cell, and squamous cell carcinoma.
2. Adenocarcinomas with squamous differenti-ation are graded according to the nuclear
grade of the glandular component.
Although these guidelines provide a reason-able baseline for grading endometrial carci-noma, they require further evaluation. For
example, serous carcinoma, by definition, has a
high nuclear grade, so the concept of “prece-dence” of nuclear grading for serous carcinoma
has no relevance, in our opinion.
The grade of the tumor from the biopsy spec-imen agrees with the grade in the hysterectomy
Malignant Neoplasms—Classification, Grading, and Staging of the Tumor 219
in fewer than 60% of cases, with both apparent
under grading and over grading of biopsies, as
compared to the hysterectomy specimens.59–63
Over-grading may be caused by heterogeneity
of the tumor, with the surface component
removed at biopsy showing the higher grade.
Under grading in biopsy specimens is most
often due to limited sample size. Thus, the
biopsy gives a general assessment of the
degree of differentiation of a tumor. Treatment
decisions are generally based on the grade of
the tumors in the hysterectomy specimen.
Nonetheless, grading of the biopsy gives the
clinician a reasonable expectation of the degree
of malignancy to be expected in the hysterec-tomy specimen. Furthermore, in occasional
cases, biopsy or curettage will remove all or
most of the carcinoma, so there can be insuffi-cient tumor in the hysterectomy for further
evaluation.
The utility of nuclear grade in the assessment
of endometrial carcinoma has been examined
in several reports.9;11;16;54;64;65
In general, a corre-lation between nuclear and architectural grade
has been found,9;10;16;58;64
although some reports
found interobserver reproducibility for nuclear
grading was not as good as for architectural
grading.54;66
Nuclear grading in hysterectomy
specimens can identify a small subset of nuclear
grade 3 tumors that have a poorer prognosis
but do not show grade 3 architectural patterns.
Some studies suggest that nuclear grading is not
an independent index of prognostic utility com-pared to the FIGO architectural grade.65
The
220 10. Endometrial Carcinoma
Table 10.2. Nuclear grading.
Grade 1. Oval/elongated nuclei, fine chromatin, small
nucleoli, few mitoses
Grade 2. (Features between grades 1 and 3)
Grade 3. Enlarged/pleomorphic nuclei, coarse chromatin,
prominent nucleoli, many mitoses
Figure 10.11. Nuclear grade 3. Nuclei are markedly enlarged and highly /pleomorphic. Although glands are
well formed in this field, this tumor had overall features of serous carcinoma.
method of nuclear grading has not been
uniform throughout the studies, however, and
often has failed to incorporate all cytologic fea-tures in determining the final nuclear grade. For
example, one study considered a tumor to have
high-grade nuclei if it showed “large nucleoli”
and “coarsely clumped chromatin,” but not
necessarily nuclear pleomorphism.65
In our
opinion, however, nuclear grading should con-sider all features, including pleomorphism,
mitotic activity, and abnormal mitotic figures
(Table 10.2). Also, the current schemes for
nuclear grading are relatively subjective, and in
our experience endometrioid carcinoma with
grade 3 nuclei are almost always solid and
would have been graded as 3 based on archi-tecture alone. For apparent large discrepancies
with high nuclear grade and apparent grade 1
architecture, it is important to make sure that
one is not dealing with a serous or clear cell car-cinoma. Further study will be needed to deter-mine the ultimate role of nuclear grading in the
assessment of endometrial carcinoma.
Recently, several investigators have pro-posed a two-grade or binary grading system in
place of the FIGO three-grade system.66;67
One
analysis used a system with 20% or more solid
tumor to separate high-grade from low-grade
tumors.67
This two-tiered system found im-proved interobserver agreement and better
prediction of uterine histology at hysterectomy.
In another study utilizing a binary grading
system, tumors that displayed two or more of
the following features: (1)  >50% solid growth
(glandular or squamous), (2) tumor cell necro-sis, or (3) an infiltrating as opposed to an expan-sile pattern of invasion were classified as high
grade.66
This binary grading was useful in
separating tumors into prognostic groups. In
particular it identified a subset of patients
with low-grade endometrioid carcinomas with
metastatic disease who had a relatively better
prognosis that was comparable to that of high-grade carcinomas confined to the uterus. This
grading system was evaluated only on hys-terectomy specimens, and, therefore, its utility
in biopsy material needs further testing.
Another study suggested that reducing the
FIGO grading system to a two-tiered system by
combining FIGO grades 1 and 2 in comparison
with FIGO grade 3 was another reproducible
method of defining tumors with a significant
difference in prognosis.67a
Clinically Important
Histologic Subtypes
Typical (Endometrioid) Carcinoma
This type of adenocarcinoma typically shows a
glandular pattern with prominent cribriform
bridging, but not infrequently it may display
a papillary pattern that has been referred to
as “villoglandular” (Figs.10.6 and 10.12).3;42;68;69
Glands may contain a small amount of mucin
or necrotic debris (Fig. 10.13). The cells lining
the glands have a moderate amount of
eosinophilic to amphophilic cytoplasm (Figs.
10.9 and 10.12),3;37
and the cell membrane at the
luminal border of the glands often is ill defined.
In the papillary tumors, the cells are columnar
and perpendicular to the fibrovascular core.
Nuclei are oval and relatively bland (see later,
Serous Carcinoma, for further discussion of
papillary patterns).
Other variants of endometrioid carcinoma
include ciliated carcinoma and secretory carci-noma.3;37;42
Ciliated carcinoma is an extremely
rare neoplasm in which the invasive glands
are lined by cells with cilia along the luminal
border.3;70;71
Its significance lies in the recogni-tion that cilia do not always indicate a benign
lesion.
In the secretory variant of endometrial ade-nocarcinoma, the neoplastic glands are lined
by cells with vacuolated cytoplasm (Fig.
10.14).3;8;12;72
Cytoplasmic vacuolization is a
feature also seen in many clear cell carcinomas
(see later, Clear Cell Carcinoma), but it is
important to separate the low-grade secretory
carcinoma from clear cell carcinoma, which is
generally a high-grade neoplasm. In secretory
carcinoma, clear vacuoles fill the subnuclear or
supranuclear cytoplasm, and the cells resemble
those seen in normal early secretory phase. The
nuclei usually show minimal atypia, although
the glands fulfill the criteria for invasion. These
rare tumors have an excellent prognosis. Secre-tory carcinoma can occur in premenopausal or
postmenopausal patients and are not necessar-ily related to progestin treatment.
Malignant Neoplasms—Classification, Grading, and Staging of the Tumor 221
Figure 10.12. Adenocarcinoma. Typical (endometri-oid) pattern. Classic “endometrioid” pattern with the
glands lined by cells with pale cytoplasm and strati-fied nuclei. This well-differentiated, FIGO grade 1
tumor also shows villoglandular papillary features.
Figure 10.13. Adenocarcinoma. Typical (endometri-oid) pattern, FIGO grade 1. The glandular cells have
a moderate amount of eosinophilic cytoplasm. The
mucoid contents within the glandular lumens do not
affect the classification of the carcinoma.
Foam cells often are present in the stroma of
endometrioid carcinoma or its variants, espe-cially when they are low grade.37;73
The presence
of foam cells by themselves does not influence
the diagnosis or the classification of endome-trial carcinoma as foam cells can occur in a
variety of benign conditions in which there is
abnormal glandular and stromal breakdown
(see Chapter 5).74
Some endometrial carcino-mas show a marked neutrophilic infiltrate. Poly-morphonuclear leukocytes can be intimately
admixed with the tumor, and the tumor cells
can show apparent phagocytosis of neutrophils.
The neutrophilic response has no known effect
on the prognosis.
Adenocarcinoma with Squamous
Differentiation
Squamous differentiation commonly occurs in
tumors with a typical (endometrioid) glandular
pattern.3
It is rarely, if ever, associated with
serous or clear cell carcinoma. At least 10% of
the tumor should have squamous features for it
to qualify as adenocarcinoma with squamous
differentiation. Typically, the squamous epithe-lium is intimately admixed with glands (Fig.
10.15).
In the low-grade neoplasms, the squamous
changes often include so-called morules,
rounded masses of bland squamous cells largely
filling the lumens of the malignant glands (Fig.
10.16). These squamous cells are incompletely
differentiated and have eosinophilic cytoplasm
and indistinct cell borders. The nuclei are
uniform, bland, and lack prominent nucleoli;
they do not palisade. Mitotic figures are infre-quent. The squamous cells can show intercellu-lar bridges, but this finding is infrequent. Often
these squamous nests are nonkeratinizing, but
keratinization can, at times, be present.
When the squamous component appears
malignant, it is usually associated with a neo-plasm that is grade 2 or 3. Portions of the tumor
Malignant Neoplasms—Classification, Grading, and Staging of the Tumor 223
Figure 10.14. Secretory carcinoma, FIGO grade 1.
This variant shows neoplastic glandular cells with
extensive cytoplasmic vacuoles, superficially resem-bling secretory phase endometrium. The secretory
changes are accompanied by low-grade nuclei.
Figure 10.15. Adenocarcinoma with squamous differentiation. Multiple foci of squamous change are inter-spersed in this well–differentiated (FIGO grade 1) adenocarcinoma.
Figure 10.16. Adenocarcinoma with squamous dif-ferentiation. Central nests of squamous epithelium
in FIGO grade 1 adenocarcinoma. The nuclear grade
of the glandular and the squamous element is the
same (grade 1).
may show squamous carcinoma without glan-dular differentiation. Tumors with a cytologi-cally malignant squamous component often are
composed of nests of spindle-shaped cells that
obliterate gland lumens (Fig. 10.17). Kera-tinization and squamous pearl formation are
frequently apparent. Abundant keratin forma-tion may even incite a foreign body response.
Mitotic activity often is brisk in the squamous
component.
At times a clear distinction between cytolog-ically benign and malignant squamous epithe-lium is not possible. The squamous component
may show mild degrees of atypia and scattered
mitotic figures (Fig. 10.18). In these cases the
cytologic features exceed the “benign” appear-ance required for a diagnosis of adenoacan-thoma but do not have all the characteristics
of malignancy required for a diagnosis of
adenosquamous carcinoma. Furthermore, this
squamous change, because it occurs in adeno-carcinoma, is malignant, regardless of its histol-ogy. When these tumors are stratified by grade
and depth of myometrial invasion, the presence
of squamous epithelium does not alter the
prognosis when compared to endometrioid
carcinoma lacking squamous epithelium.18;19;53
Accordingly, it is the grade of the glandular
component that has prognostic significance. For
these reasons, the term “adenocarcinoma with
squamous differentiation” is preferred. These
tumors should be graded 1, 2, or 3 based on the
architectural and nuclear features of the glan-dular component.
Mucinous Carcinoma
Mucinous carcinoma of the endometrium has a
glandular architecture resembling endometri-oid carcinoma but is composed of cells con-Malignant Neoplasms—Classification, Grading, and Staging of the Tumor 225
Figure 10.17. Adenocarcinoma with squamous dif-ferentiation. Poorly differentiated, FIGO grade 3
adenocarcinoma shows nests of spindle–shaped
squamous cells filling glands. The squamous compo-nent is cytologically malignant, with multiple mitoses
and no keratinization. The spindle pattern should not
be mistaken for a sarcomatous component.
Figure 10.18. Adenocarcinoma with squamous dif-ferentiation. In this tumor the squamous cell nuclei
are intermediate between grade 1 and grade 3
(compare to Figs. 10.16 and 10.17). The glandular
component showed features of moderately differen-tiated (FIGO grade 2) adenocarcinoma.
taining abundant intracytoplasmic mucin (Fig.
10.19).3;13;14;37;75
It is cytoplasmic mucin, not
extracellular and luminal that establishes the
diagnosis of endometrial mucinous carcinoma.
Furthermore, the presence of cytoplasmic
mucin should be extensive, involving greater
than 50% of the cells, for a tumor to be classi-fied as mucinous carcinoma, as some mucin
production is present in most endometrial car-cinomas.14;76
Special stains for epithelial mucin,
such as mucicarmine or the periodic acid–Schiff
(PAS) with diastase digestion, can be helpful to
demonstrate the mucin. Up to 9% of all stage
1 endometrial carcinomas are of the mucinous
type according to these criteria.14
Mucinous carcinomas tend to be well to mod-erately differentiated, and they frequently have
a papillary or villous architecture. Portions of
these carcinomas often appear extremely well
differentiated, because the mucinous cytoplasm
results in basal alignment of the nuclei with
minimal nuclear stratification, but usually the
tumor merges with a more typical endometri-oid pattern. Foam cells are often present, and a
neutrophilic infiltrate may be seen.
The morphologic features of low-grade muci-nous carcinoma overlap with the patterns of
so-called secretory carcinoma; both types of
tumors have abundant pale cytoplasm and
basal nuclei with minimal stratification. In
mucinous carcinoma the cells contain cytoplas-mic mucin, whereas in secretory carcinoma the
cytoplasmic vacuoles contain glycogen. The dis-tinction is largely academic, however, because
tumor grade rather than cytoplasmic differenti-ation determines prognosis. These neoplasms
show no difference in behavior from
endometrioid tumors of similar grade.14
Because the cell population of mucinous
carcinoma resembles endocervical epithelium,
226 10. Endometrial Carcinoma
with basal nuclei and abundant supranuclear
cytoplasm that contains mucin, the differential
diagnosis often includes endocervical adeno-carcinoma. This differential is discussed in
greater detail below.
Endocervical-Like Endometrial
Carcinoma
Occasional well-differentiated mucinous
endometrial carcinomas have patterns that
mimic cervical microglandular hyperplasia.77;78
These tumors are composed of numerous small
mucinous glands (Fig. 10.20) and some also
contain acute inflammatory cells in the lumens
and the intervening stroma. These tumors have
minimal cytologic atypia and mitotic activity,
thus closely resembling endocervical micro-glandular hyperplasia. The endometrial neo-plasms typically show a transition to ordinary
endometrioid adenocarcinoma, which facili-tates the diagnosis. In addition, the criteria
outlined in a later section for distinguish-ing endometrial and endocervical primary
tumors also help in the recognition of these
carcinomas.
Serous Carcinoma
Serous carcinoma is recognized by its marked
nuclear atypia and its resemblance to ovarian
serous carcinoma.20;23;68;79–8< F2
Like their ovarian
counterparts, these tumors often have a papil-lary growth pattern (Fig. 10.21). The fibrotic,
papillary fronds are lined by a stratified layer of
cells with a high nuclear/cytoplasmic ratio and
high-grade nuclei (Table 10.3). The malignant
cells also form complex papillary tufts that may
be present as free-floating clusters of cells.
These papillary tufts often lack stroma and are
Malignant Neoplasms—Classification, Grading, and Staging of the Tumor 227
Figure 10.19. 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.
Figure 10.20. Adenocarcinoma with endocervical-like pattern. This FIGO grade 1 adenocarcinoma is
composed of anastomosing small glands that have a
superficial resemblance to microglandular hyperpla-sia of the cervix. This is the confluent, cribriform
pattern of adenocarcinoma, however. Neutrophils
are present in the epithelium and the extracellular
mucin. Inset: The nuclei are low grade.
Figure 10.21. Serous carcinoma. Arborizing papillae with multiple papillary tufts are a frequent character-istic of this tumor. Many of the papillae appear as free-floating clusters of cells.
composed of dense papillary aggregates of
tumor cells. In contrast to low-grade papillary
carcinoma, the papillary fronds usually are
coarse, with thick fibrotic cores lined by highly
epithelial atypical cells (Fig. 10.22). Because of
the papillary growth, serous carcinoma can
appear deceptively well differentiated in the
endometrium, although these are high-grade,
aggressive neoplasms. Serous carcinoma was
originally considered a predominantly papillary
neoplasm, but studies now show that it is mor-phologically diverse.80
In some cases the papil-lae are long and slender instead of short and
coarse. The tumor may even be primarily com-posed of glands with lumens (Fig. 10.11). In the
myoinvasive component of the tumor the
glands have a gaping appearance (Fig. 10.23).
Psammoma bodies are present in up to one
third of cases.68
It is the combination of a low-grade archi-tecture (papillary or glandular pattern) and
high-grade nuclear atypia that identifies serous
carcinoma.83
The nuclei are hyperchromatic
and pleomorphic with macronucleoli and many
mitoses (Figs. 10.11 and 10.24). Abnormal
mitotic figures are frequent. Some nuclei are
lobulated with deep clefts, and, not infre-quently, the chromatin appears smudged. The
cells of serous carcinoma tend to be rounded
Malignant Neoplasms—Classification, Grading, and Staging of the Tumor 229
Table 10.3. Histologic features of serous carcinoma.
Complex, coarse papillae
Irregular, gaping glands
Papillary tufting
High nuclear/cytoplasmic ratio
Marked nuclear pleomorphism
Numerous and abnormal mitoses
Macronucleoli
Clear cell componenta
Psammoma bodiesa
a
Nonspecific features found occasionally in serous
carcinoma.
Figure 10.22. Serous carcinoma. In this field the tumor forms multiple coarse papillae.
230 10. Endometrial Carcinoma
Figure 10.23. Serous carcinoma. This tumor shows gaping glands and papillae with papillary tufts. There is
marked nuclear pleomorphism, and the high nuclear grade is a constant feature of serous carcinoma.
Figure 10.24. Serous carcinoma. At high magnification the cells show nuclear grade 3 changes. The nuclei
are large, with macronucleoli, a high nuclear/cytoplasmic ratio, and numerous mitoses with abnormal forms.
and often have abundant granular, eosinophilic
cytoplasm. Areas displaying the features of
serous carcinoma but containing clear cells, that
is, clear cell carcinoma, are seen in up to one
third of cases (see later, Clear Cell Carcinoma).
Because these tumors always contain high-grade nuclei, nuclear grading as recommended
by FIGO is not relevant. The diagnosis of
serous carcinoma itself establishes the presence
of a highly malignant carcinoma.
Several other histologic subtypes of endome-trial carcinoma, including low-grade endometri-oid tumors, also may show papillary growth,
so-called villoglandular pattern.3;15;37;58;68;84
The
papillary endometrioid tumors have low- to
moderate-grade nuclei (see earlier, Grading)
and often grow in long, slender branching pap-illary fronds (Figs. 10.12, 10.25, and 10.26). The
lining cells are columnar and do not form pap-illary tufts. In contrast, the papillae in serous
carcinoma tend to be small and coarse, although
it is the high nuclear grade that is most useful
for identifying this neoplasm. An occasional
tumor may show intermediate features between
low-grade villoglandular and serous carcinoma
with papillary architecture showing somewhat
coarser papillae and grade 2 nuclei (Fig. 10.26).
If a tumor with increased architectural and cyto-logic atypia does not fulfill all the criteria of
serous carcinoma, however, it should be classi-fied as a grade 2 endometrioid carcinoma. To
avoid confusion between these papillary tumors
with different cytologic features and prognosis,
we recommend not using the term “papillary”
as a diagnostic term for any type of endometrial
carcinoma.
Immunohistochemical analysis can assist in
the recognition of serous carcinoma and dis-tinction from endometrioid carcinoma. Serous
carcinomas usually show diffuse, intense reac-tivity for p53 protein which correlates with p53
gene mutations which are found in more than
Malignant Neoplasms—Classification, Grading, and Staging of the Tumor 231
Figure 10.25. Well-differentiated villoglandular
adenocarcinoma. The nuclei lining the papillae have
grade 1 features, in contrast to the high nuclear grade
seen in serous carcinoma. Nuclei are round to oval
and exhibit minimal pleomorphism. Also present is
stromal desmoplasia.
90% of serous carcinomas (Fig. 10.27).85–88
In addition, they have a high proliferation
index reflected in diffuse reactivity for Ki-67
protein in the nuclei.89
Generally, serous carci-noma lacks estrogen and progesterone recep-tors.89–91
Low-grade endometrioid carcinomas,
in contrast, have relatively low prolifera-tion indices, do not express  p53, and show
strong reactivity for estrogen and progesterone
receptors.90–92
Serous carcinoma is a highly malignant form
of endometrial carcinoma that usually occurs in
older women, with a median age in the seventh
decade.15;68;82;84;93;94
Rare cases are seen in
younger women, however.95–97
These tumors
often invade the myometrium deeply and per-meate lymph-vascular spaces. Not uncommonly,
serous carcinoma is minimally invasive and may
even be confined to an endometrial polyp.98;99
Serous carcinoma also can be limited to curet-ting specimens only with no residual serous
carcinoma in the subsequent hysterectomy.100
Serous carcinoma with invasion of less than
1.0 cm has been termed “minimal uterine serous
carcinoma (MUSC.)101
Even with no or minimal
myometrial invasion, serous carcinoma can dis-seminate widely.11;15;68;79;80;84;96;101–106;106a
Patients with serous carcinoma often either
have peritoneal spread at the time of hysterec-tomy or relapse with peritoneal carcinomatosis,
so in this regard they behave like their ovarian
counterparts. Occasional cases also appear to
be multifocal, with associated ovarian serous
carcinoma at the time of diagnosis. Because of
their aggressive behavior, even when superficial
or confined to a polyp, mixed tumors with both
endometrioid and serous patterns in which at
least 25% of the tumor contains a serous
component should be classified as serous
carcinoma.
232 10. Endometrial Carcinoma
Figure 10.26. Moderately differentiated adenocar-cinoma with papillary features. Typical (“endometri-oid”) adenocarcinoma shows a papillary pattern.
Inset: The nuclei show considerable atypia but do
not have features of serous carcinoma, indicating a
FIGO grade 2 carcinoma.
Clear Cell Carcinoma
In clear cell carcinoma, the majority of the cells
have clear, vacuolated cytoplasm because of the
presence of glycogen.3;8;12;21;22;25;37
This tumor can
have a variety of growth patterns, including
tubular, cystic, papillary, and solid (Figs. 10.28
and 10.29). The stroma and papillary cores of
clear cell carcinoma typically have hyalinized
stroma. In some cases the clear cytoplasm is
inconspicuous and the nuclei bulge into the
gland lumens, forming so-called hobnail cells
(Fig. 10.29). Clear cell carcinoma also occurs in
the ovary, cervix, and vagina. In the ovary or
cervix, clear cell tumors in older women have
patterns similar to those found in the
endometrium, but in the vagina and cervix of
women exposed to in utero diethylstilbstrol
(DES), the tubulocystic pattern predominates.
Many examples of clear cell carcinoma
appear to be closely related to serous car-cinoma, and, as mentioned earlier, this tumor
pattern may be admixed with serous carci-noma.68
The similarity with serous carcinoma
is reflected in the high nuclear grade of
many clear cell tumors (Fig. 10.28). Especially
in the papillary and solid patterns, the nuclei
generally are pleomorphic with marked
atypia, although the hyalinized stroma of clear
cell carcinoma is a distinguishing feature.
Macronucleoli and abnormal mitotic figures
usually are present. Occasionally, these tumors,
like serous carcinoma, contain psammoma
bodies. Clear cell carcinoma is another aggres-sive variant of endometrial cancer.8;12;107
Like
serous carcinoma, it tends to occur in older
patients and has a high relapse rate.25;107;108
Although clinically similar to serous carci-noma, high-grade clear cell carcinoma less fre-quently demonstrates the diffuse strong reac-tivity for p53 protein that is seen in serous
carcinoma.89
Malignant Neoplasms—Classification, Grading, and Staging of the Tumor 233
Figure 10.27. Serous carcinoma, p53 immunohistochemistry. Diffuse, strong reactivity for p53 protein is a
characteristic of most serous carcinomas.
Figure 10.28. Clear cell carcinoma. Papillary pat-tern in which the cells have clear, vacuolated cyto-plasm. Inset: The nuclei are high grade (grade 3) with
marked enlargement, irregular outlines, smudged
chromatin, and prominent nucleoli.
Figure 10.29. Clear cell carcinoma. Tubular pattern
in which clear cytoplasm is less conspicuous, and the
nuclei bulge, hobnail fashion, into the glandular
lumens (arrows). The hyalinized stroma is an occa-sional feature of clear cell carcinoma.
Clear cell carcinoma should be differentiated
from secretory carcinoma or endometrioid car-cinoma with clear cell/secretory change. In con-trast to the high nuclear grade of clear cell
carcinoma, other carcinomas with clear or pale
cytoplasm lack high nuclear grade and do not
show the consistent growth patterns of clear
cell carcinoma including papillary growth with
hyalinized stroma cores, hobnail cells, and tubu-locystic arrangements. Some of these carcino-mas with lower grade nuclei tend to be solid,
containing cells with clear cytoplasm that may
represent squamous differentiation. If an
endometrial carcinoma with clear cells does not
show the typical growth patterns with high
nuclear grade, it should be classified as an
endometrioid tumor with secretory change.
A rare case of apparent secretory carcinoma
may show foci with high-grade nuclei, suggest-ing that clear cell carcinoma may be present, so
it is important to thoroughly sample any low-grade tumor with clear cytoplasm to determine
if grade 3 nuclei and a more aggressive neo-plasm may be present.
Rare Histologic Patterns
Primary squamous carcinoma of the endo-metrium does occur, but is rare.109–112
To diag-nose this entity, it is necessary to exclude a
primary cervical carcinoma (see later, Endome-trial Versus Endocervical Carcinoma). Primary
verrucous squamous cell carcinoma also has
been reported to arise in the endometrium.113
Recently, there have been several reports
of transitional cell carcinomas of the endome-trium that are usually admixed with other pat-terns of endometrial carcinoma.114;115
Undifferentiated carcinomas show no evi-dence of glandular or squamous differentia-tion.116
They account for fewer than 2% of
primary endometrial carcinomas. Some undif-ferentiated carcinomas have features resem-bling small cell carcinoma of the lung, whereas
others are composed of large cells that range
from polygonal to spindle in shape. Undifferen-tiated carcinoma often has a solid growth
pattern with extensive necrosis. With the small
cell pattern, the neoplastic cells have scant cyto-plasm and hyperchromatic nuclei with indistinct
nucleoli.117
These neoplasms may show neu-roendocrine differentiation.118–121
Some tumors
with a small cell component are admixed with
typical adenocarcinoma.117
The large-cell
variant is composed of sheets of large epithelial
cells that have a moderate amount of cytoplasm
and large vesicular nuclei with prominent
nucleoli.
Rarely, carcinomas show unusual patterns of
differentiation, such as osteoclastic-type giant
cells or trophoblast.69;122
Choriocarcinoma of the
endometrium may be found in postmenopausal
women, usually resulting from dedifferentiation
of a poorly differentiated carcinoma.122–127
Other
rare histologic types include glassy cell carci-noma,128;129
oxyphilic variant of endometrioid
carcinoma,130
signet-ring cell carcinoma,131
lymphoepithelioma-like carcinoma,132
and giant
cell carcinoma.133
Most of these are high-grade malignancies that occur in older patients.
Other rare tumors include primary yolk sac
tumor,134;135
a -fetoprotein–secreting hepatoid
adenocarcinoma associated with endometrioid
carcinoma,136
and endometrioid carcinoma
associated with Ewing sarcoma/peripheral
primitive neuroectodermal tumor.137–141
Wilms
tumor also may rarely occur as a primary uterine
neoplasm.142;143
On occasion an endometrial neoplasm is
difficult to classify. Small tissue samples or
extensive necrosis may limit the histologic
evaluation. Immunohistochemistry has little
utility for subclassification of primary endome-trial carcinomas, although keratin reactivity can
help determine whether or not a malignant
tumor is a carcinoma. The one exception is the
strong immunoreactivity for p53 that is often
seen in serous carcinoma.
Mixed Mesodermal Tumors
The distinction between a malignant mesoder-mal mixed tumor (MMMT) or carcinosarcoma
and high-grade carcinoma may at times be dif-ficult. In MMMT the epithelial element usually
has features of high-grade endometrial adeno-carcinoma, often serous or clear cell carcinoma,
although endometrioid patterns, including car-cinoma with squamous differentiation, may be
found. Clinicopathologic, immunohistochemi-Malignant Neoplasms—Classification, Grading, and Staging of the Tumor 235
cal, and molecular genetic studies now indicate
that MMMT represents a variant of carcinoma
(sarcomatoid carcinoma), yet it is important to
identify a sarcomatous component because
MMMT is even more aggressive than other
high-grade carcinomas. In such cases, ascertain-ing the presence of a malignant stromal com-ponent, a feature discussed in the next chapter,
becomes important. Often, the sarcomatous
component is readily identified, especially
when heterologous elements, such as cartilage
or rhabdomyoblasts, are present. In many cases,
however, the malignant stroma is composed
only of spindle cells, identified as malignant by
their cellularity, nuclear atypia, and high mitotic
activity. Typically, the sarcomatous component
is intimately associated with, yet distinct from,
the carcinomatous component. In these cases,
keratin immunohistochemical stains can be
helpful for distinguishing the carcinomatous
from the sarcomatous component, as the sarco-matous elements are negative or only focally
positive for keratin.
Staging
In 1988 FIGO revised its staging system from
one that was strictly based on clinical evalua-tion to one based on combined surgical and
histopathologic findings.29
Staging of endome-trial carcinoma (corpus cancer) now employs a
variety of histologic risk factors, including
grade, depth of myometrial invasion, involve-ment of the cervix, and peritoneal cytology
(Table 10.4).27
Endometrial carcinoma confined to the
corpus is stage I, and about three quarters of all
primary endometrial carcinomas are stage I.
Based on the presence and amount of myome-trial invasion, carcinomas are subdivided into
IA when tumor is limited to the endometrium,
IB when tumor invades less than one half the
myometrium, and IC when tumor invades more
than one half the myometrium. With cervical
invasion the tumor is stage II. Stage II is subdi-vided into IIA when there is only endocervical
glandular involvement, and IIB when there is
cervical stromal invasion by endometrial carci-noma. As this assessment is based on the micro-scopic findings in the hysterectomy specimen,
fractional dilation and curettage, with curettage
specimens of the endometrium and endocervix
submitted separately to differentiate stage I
and stage II, is no longer necessary. In fact,
the office-based endometrial biopsy often
establishes the diagnosis and the patient is
spared further endometrial evaluation prior to
hysterectomy.
Although staging is based on surgical–patho-logic analysis of the hysterectomy specimen,
accurate histologic evaluation of biopsies is
important, as grade and histologic findings
influence the planning for surgery and whether
surgical staging and lymph node sampling is
necessary. Furthermore, with a high-grade neo-plasm, a gynecologist should request the assis-236 10. Endometrial Carcinoma
Table 10.4. International Federation of Gynecology and Obstetrics
(FIGO) staging of corpus cancer.
Stage Description
Ia G123 Tumor limited to endometrium
Ib G123 Invasion of less than half of the myometrium
Ic G123 Invasion of more than half of the myometrium
IIa G123 Endocervical glandular involvement only
IIb G123 Cervical stromal invasion
IIIa G123 Tumor invades serosa and/or adenexae and/or positive peritoneal
cytology
IIIb G123 Vaginal metastases
IIIc G123 Metastases to pelvic and/or paraaortic lymph nodes
IVa G123 Tumor invasion of bladder and/or bowel mucosa
IVb Distant metastases including intraabdominal and/or inguinal
lymph node
G, Grade.
tance of a gynecologic oncologist in order that
appropriate staging and therapy are performed.
Endometrial Versus Endocervical
Carcinoma
At times it is difficult to determine whether car-cinoma in an endometrial biopsy involves the
endometrium, the endocervix, or both sites.
Assigning a primary site is important, because
endometrial carcinoma confined to the corpus
is managed differently from endocervical carci-noma. The usual therapy for endometrial
carcinoma is extrafascial total abdominal hys-terectomy and bilateral salpingo-oophorec-tomy. In contrast, surgical management of
invasive endocervical carcinoma is a radical
hysterectomy and pelvic lymph node dissec-tion, an operation with potential for greater
morbidity.
If the tumor in the endometrial biopsy has
the typical endometrioid pattern and the clini-cal information, such as older age and uterine
enlargement, is consistent with an endometrial
cancer, there is little question regarding the
primary site. In contrast, if the tumor has a
pattern that is more commonly found in the
cervix, such as mucinous carcinoma, or carci-noma with extensive squamous differentiation,
then determination of the primary site is more
difficult.
Squamous differentiation can be prominent
in primary endometrial carcinoma and
adenosquamous carcinoma is a well-recognized
variant of endocervical carcinoma. Conse-quently, when curettage specimens show a
mixture of glandular and squamous compo-nents, the differential diagnosis often includes
endometrial and endocervical primary sites.
Adenocarcinoma with a squamous component
at each site has different histologic features. In
endometrial carcinoma, the squamous element
often is intimately associated with glands,
appearing to arise in and differentiate from
the glands. The glandular element typically
predominates in endometrial carcinomas. In
contrast, in endocervical adenosquamous
carcinoma, the squamous element usually pre-dominates and glandular differentiation is
subtle. Furthermore, the cervical neoplasms do
not show the prominent nests of morular
growth with squamous differentiation confined
to gland lumens that are very frequent in
endometrial adenocarcinoma.
Added evidence in favor of an endometrial
primary carcinoma is the presence of associated
hyperplasia. Stromal foam cells also would
suggest a primary tumor of the endometrium.
On the other hand, the presence of cervical
intraepithelial neoplasia or endocervical ade-nocarcinoma in situ, as well as transitions to
normal endocervical epithelium, supports the
diagnosis of cervical carcinoma. In addition, in
a fractional curettage, if more tumor is present
in the endometrial fraction, it is more likely that
the neoplasm is arising in the endometrium.
Conversely, if the endocervical fraction con-tains the bulk of the tumor, the primary tumor
is most likely in the endocervix.
Immunohistochemical and in situ hybridiza-tion studies can help to distinguish endometrial
from endocervical primary tumors. Recent
studies indicate that estrogen and especially
progesterone receptor protein is reactive in
endometrioid type carcinoma, while cervical
adenocarcinoma is not.144–147
In addition, detec-tion of human papilloma virus (HPV) by in situ
hybridization is seen in endocervical but not
endometrial carcinomas.144
The combination of
hormone receptor and HPV in situ analysis
appears to be very useful in this differential
diagnosis. Preliminary data suggest that strong
diffuse expression of p16, which occurs in close
to 100% of cervical squamous and adenocarci-nomas, is either absent or only patchy in
endometrioid carcinoma.148;149
Other immunohistochemical stains have
been applied although the results suggest that
these are less specific for differentiating these
two primary sites. For instance, vimentin fre-quently stains endometrial carcinomas whereas
cervical adenocarcinomas are nega-tive.150;145;146;151
Conversely, carcinoembryonic
antigen (CEA) often is present in endocervical
carcinomas but is less common in endometrial
primary tumors.146;151–154
The majority of endo-cervical carcinomas show abundant, diffuse
intracellular CEA reactivity, whereas only
about half of endometrial carcinomas contain
CEA, and this reactivity is usually focal and at
Malignant Neoplasms—Classification, Grading, and Staging of the Tumor 237
the luminal surface. This immunostaining
pattern for CEA is not useful for endometrial
mucinous carcinoma from endocervical adeno-carcinoma as both of these tumor types tend to
be CEA positive. Differential cytokeratin stain-ing for cytokeratin 7 and cytokeratin 20 is not
useful in distinguishing between these primary
sites.151
Histochemical stains for mucin have
little value in determining the primary site,
however, because endometrial adenocarcinoma
often shows at least focal cytoplasmic mucin
and some endocervical adenocarcinomas show
little cytoplasmic mucin.
If the morphologic features do not clearly
establish the primary site, clinical information
often resolves the problem. The presence or
absence of gross tumor involving the cervix is
important information. If tumor clinically
involves the cervix, then the issue of primary
site is largely irrelevant. The tumor can be
managed as a cervical neoplasm.27
If the tumor
is occult, information on age and menopausal
status can be helpful. Endometrial carcinoma
usually occurs in older patients with an average
age of 55 to 60 years. Endometrial carcinoma
is unusual before the menopause unless the
patient has a predisposing condition, such as
the Stein–Leventhal syndrome. Endocervical
adenocarcinoma, in contrast, has a much wider
age range, with many tumors occurring in
premenopausal women. The average age is
between 45 and 55 years.
Metastatic Carcinoma
The most common extrauterine carcinomas
that metastasize to or extend into the
endometrium arise in the ovary, breast, or gas-trointestinal tract, especially the colon.155–157
Metastases from other primary sites are rare,
but on occasion a tumor from the stomach, pan-creas, or other visceral site metastasizes to the
endometrium. It is very unusual for tumors
from these sites to present with abnormal
vaginal bleeding and to be diagnosed first in an
endometrial biopsy.
Separating metastatic ovarian carcinoma
from an endometrial primary tumor can be
especially difficult, as virtually all patterns of
primary endometrial carcinoma can occur in
primary ovarian epithelial carcinoma.158;159
In
particular, serous and endometrioid carcinomas
are common in the ovary, and these tumors are
histologically identical to their endometrial
counterparts. Despite these difficulties, the
question of ovarian metastasis versus an
endometrial primary site in a biopsy or curet-ting is infrequent. This differential diagnosis
is most likely to occur when the biopsy shows
a serous carcinoma, usually with psammoma
bodies. In such cases it may be impossible to
exclude an ovarian primary site, but metastatic
tumor presenting in biopsies is extremely infre-quent compared to a primary endometrial
serous carcinoma. Furthermore, cases with
involvement of the endometrium and ovary
often appear to represent synchronous primary
tumors rather than metastases. Metastatic
lesions should be considered, however, when
the amount of tumor is relatively scant and is
admixed with more abundant fragments of
benign, nonhyperplastic endometrium. Also,
clinical history of an adnexal mass should alert
the pathologist to the possibility of primary
ovarian neoplasia.
Metastatic carcinoma from other sites,
although rare, can be problematic. Occasion-ally, colon carcinoma may involve the
endometrium and closely simulate a uterine
primary tumor, having an “endometrioid”
pattern (Fig. 10.30). In such cases a history of a
known extrauterine primary tumor or a mass
lesion in the bowel can be essential for estab-lishing the correct diagnosis. Colon adenocarci-noma may have a so-called garland-like
arrangement of glands surrounding areas of
“dirty” necrosis composed of cellular debris
that helps in recognizing the tumor. The glands
of metastatic colon carcinoma have a sharp
luminal border. Endometrial carcinoma, on the
other hand, typically does not show much
necrosis in glandular lumens, and the cells have
an ill-defined, fuzzy luminal border in routine
sections. Mucin stains have little utility in deter-mining the primary site, as endometrial carci-noma can have abundant cytoplasmic mucin.
Immunohistochemical stains for CEA may be
helpful in establishing whether the tumor is
metastatic from the gastrointestinal tract or is
the primary tumor in the endometrium. In
238 10. Endometrial Carcinoma
general, colon primary tumors are diffusely
positive for CEA while endometrial carcino-mas are not. Also, endometrial carcinomas
usually are positive for cytokeratin 7 and nega-tive for cytokeratin 20,151
while metastatic colon
adenocarcinoma usually is strongly positive for
cytokeratin 20 and negative for cytokeratin 7.160
Metastatic breast carcinoma, although rare,
can be especially difficult to diagnose. These
tumors typically infiltrate in solid sheets or
small groups in a linear pattern, often sparing
glands and diffusely invading the stroma (Fig.
10.31). The neoplastic cells may resemble
stromal cells or inflammatory cells, lacking
the organoid arrangements seen in most car-cinomas. In such cases, immunohistochemical
stains for keratin and histochemical stains
for mucin are useful for demonstrating the
epithelial origin of the cells. Also, an immunos-tain for gross cystic disease fluid protein-15
would help identify the primary site of a breast
metastasis. Differential immunostaining for
cytokeratins 7 and 20 has no utility in the
differential diagnosis of endometrial versus
breast carcinoma.161
Clinical Queries and Reporting
It is important to establish the diagnosis of
atypical hyperplasia versus well-differentiated
adenocarcinoma whenever possible. A biopsy
diagnosis of atypical hyperplasia may be
managed medically with progestin therapy with
periodic resampling of the endometrium. This
conservative management is possible because
Clinical Queries and Reporting 239
Figure 10.30. Metastatic colon carcinoma. Irregular
glands with sharply delimited luminal borders are
haphazardly distributed in a reactive, fibroblastic
stroma. This pattern can closely mimic an endome-trial primary, but the haphazard distribution of the
glands in markedly desmoplastic stroma suggests a
metastasis from a gastrointestinal primary. Also, the
small nests of malignant cells (arrow) are not
typical of endometrial carcinoma.
many lesions are reversible. A biopsy diagnosis
of well-differentiated adenocarcinoma, in
contrast, clearly establishes the presence of
malignancy.
Although the criteria for distinguishing atyp-ical hyperplasia from well-differentiated ade-nocarcinoma are well defined, the diagnosis
should also be placed in the appropriate clini-cal context in equivocal cases. For example, in
younger, premenopausal patients, a conserva-tive approach is appropriate when the biopsy
shows an atypical lesion that may represent
well-differentiated adenocarcinoma. Studies
show that these lesions, even if they fulfill cri-teria for adenocarcinoma, are indolent and
reversible in up to 75% of cases with progestin
therapy.162
Thus, in the premenopausal patient
younger than 40 years of age, it is especially
important to be certain that a lesion at least ful-fills the minimal criteria for well-differentiated
adenocarcinoma before establishing a diagno-sis of malignancy. If the glands are atypical but
do not clearly show features of “stromal inva-sion” as defined earlier, then the lesion is best
classified as atypical hyperplasia. These patients
can be managed conservatively and re-biopsied, with one study showing a 90% regres-sion rate.162
In older postmenopausal women,
atypical glands should be viewed even more
suspiciously for the possibility of underlying
carcinoma, and criteria for the histologic diag-nosis of carcinoma can be applied more liber-ally. Most of these patients are best treated by
a total abdominal hysterectomy and bilateral
salpingo-oophorectomy if they are candidates
for surgery.
Once the diagnosis of endometrial adenocar-cinoma is made, the gynecologist often needs
240 10. Endometrial Carcinoma
Figure 10.31. Metastatic breast carcinoma. A solid
sheet of cells representing metastatic breast carci-noma infiltrates the endometrium. The pattern could
be mistaken for an aggregate of stromal or inflam-matory cells, and special stains for mucin and keratin
may be needed to establish the diagnosis. The resid-ual glands show progestin effect with glandular
atrophy secondary to therapy for breast carcinoma.
information on several other aspects of the
tumor. Whenever possible, the FIGO grade of
endometrial carcinoma in biopsies should be
given. For instance, low-grade adenocarcinoma
frequently is confined to the endometrium or is
only superficially invasive of the myometrium;
extrauterine spread is unlikely. Conversely,
high-grade carcinoma, including aggressive his-tologic subtypes of serous and clear cell carci-noma, has greater potential for extrauterine
spread at the time of hysterectomy. If grading
is not possible owing to limited sampling, or to
necrosis or other distortion of the tissue, this
should be noted.
Classifying the tumor according to histologic
subtype also is important. When the tumor has
an endometrioid, villoglandular, secretory, or
mucinous pattern, the type of tumor has little
clinical significance by itself once the tumor has
been assigned the appropriate histologic grade.
Squamous differentiation, too, has little clinical
significance once the tumor is graded but may
be relevant to note for histologic correlation
with any subsequent metastases or recurrences.
On the other hand, serous and high-grade clear
cell carcinomas are aggressive tumors that are
important to identify. The diagnosis of these
tumors indicates that there is an increased risk
for deep myometrial invasion and metastases.
Mixed carcinomas have at least 10% of a
second cell type. An example would be a
mixture of endometrioid and serous patterns. If
25% or more of such tumor is composed of
serous carcinoma, then it should be classified as
such because these mixed neoplasms behave
the same as pure serous carcinoma. Otherwise
it can be termed a “mixed carcinoma,” with a
note describing the different types of carci-noma present.
The amount of tumor in the specimen and
associated lesions such as hyperplasia or polyps
are potentially useful data. For example, a
small amount of tumor may alert the clinician
to difficulties in accurate grading. The pres-ence of associated hyperplasia suggests
estrogenic effects that may influence the
method of therapy. Should the pattern suggest
a metastatic process, the report should clearly
indicate this and suggest the primary sites, if
possible.
References
(1) Smith M, McCartney AJ. Occult, high-risk
endometrial cancer. Gynecol Oncol 1985;
22:154–161.
(2) Bokhman JV. Two pathogenetic types of
endometrial carcinoma. Gynecol Oncol 1983;
15:10.
(3) Ronnett BM, Zaino RJ, Ellenson LH,
Kurman RJ. Endometrial carcinoma. In:
Kurman RJ, ed. Blaustein’s pathology of the
female genital tract. 5th ed. New York:
Springer-Verlag; 2002: 501–559.
(4) Sherman ME. Theories of endometrial car-cinogenesis: A multidisciplinary approach.
Mod Pathol 2000; 13:295–308.
(5) Longacre TA, Hendrickson MR. Diffusely
infiltrative endometrial adenocarcinoma: an
adenoma malignum pattern of myoinvasion.
Am J Surg Pathol 1999; 23:69–78.
(6) Landry D, Mai KT, Senterman MK, Perkins
DG, Yazdi HM, Veinot JP, et al. Endometri-oid adenocarcinoma of the uterus with a
minimal deviation invasive pattern.
Histopathology 2003; 42:77–82.
(7) Alberhasky RC, Connelly PJ, Christopherson
WM. Carcinoma of the endometrium. IV.
Mixed adenosquamous carcinoma. Am J Clin
Pathol 1982; 77:655–664.
(8) Christopherson WM, Alberhasky RC,
Connelly PJ. Carcinoma of the endometrium.
I. A clinicopathologic study of clear cell car-cinoma and secretory carcinoma. Cancer
1982; 49:1511–1523.
(9) Christopherson WM, Connelly PJ,
Alberhasky RC. Carcinoma of the endome-trium. V. An analysis of prognosticators in
patients with favorable subtypes and stage I
disease. Cancer 1983; 51:1705–1709.
(10) Connelly PJ, Alberhasky RC, Christopherson
WM. Carcinoma of the endometrium. III.
Analysis of 865 cases of adenocarcinoma
and adenoacanthoma. Obstet Gynecol 1982;
59:569–575.
(11) Hendrickson M, Ross J, Eifel PJ, Cox RJ,
Martinez A, Kempson R. Adenocarcinoma
of the endometrium: An analysis of 256 cases
with carcinoma limited to the uterine corpus.
I. Pathology review and analysis of prognos-tic variables. Gynecol Oncol 1982; 13:373–
392.
(12) Kurman RJ, Scully RE. Clear cell carcinoma
of the endometrium. An analysis of 21 cases.
Cancer 1976; 37:872–882.
References 241
(13) Melhem MF, Tobon H. Mucinous adenocar-cinoma of the endometrium: A clinico-patho-logical review of 18 cases. Int J Gynecol
Pathol 1987; 6:347–355.
(14) Ross JC, Eifel PJ, Cox RS, Kempson RL,
Hendrickson MR. Primary mucinous adeno-carcinoma of the endometrium. A clinico-pathologic and histochemical study. Am J
Surg Pathol 1983; 7:715–729.
(15) Chen JL, Trost DC, Wilkinson EJ. Endome-trial papillary adenocarcinomas: Two clinico-pathological types. Int J Gynecol Pathol
1985; 4:279–288.
(16) Mittal KR, Schwartz PE, Barwick KW. Archi-tectural (FIGO) grading, nuclear grading,
and other prognostic indicators in stage I
endometrial adenocarcinoma with identifica-tion of high-risk and low-risk groups. Cancer
1988; 61:538–545.
(17) Wilson TO, Podratz KC, Gaffey TA, Malka-sian GD, O’Brien PC, Naessens JM. Evalua-tion of unfavorable histologic subtypes in
endometrial adenocarcinoma. Am J Obstet
Gynecol 1990; 162:418–426.
(18) Zaino RJ, Kurman R, Herbold D, Gliedman
J, Bundy BN, Voet R, et al. The significance
of squamous differentiation in endometrial
carcinoma—Data from a Gynecologic
Oncology Group study. Cancer 1991;
68:2293–2302.
(19) Abeler VM, Kjorstad KE. Endometrial ade-nocarcinoma with squamous cell differentia-tion. Cancer 1992; 69:488–495.
(20) Abeler VM, Kjorstad KE. Serous papillary
carcinoma of the endometrium: A
histopathological study of 22 cases. Gynecol
Oncol 1990; 39:266–271.
(21) Abeler VM, Kjorstad KE. Clear cell carci-noma of the endometrium: A histopatholog-ical and clinical study of 97 cases. Gynecol
Oncol 1991; 40:207–217.
(22) Webb GA, Lagios MD. Clear cell carcinoma
of the endometrium. Am J Obstet Gynecol
1987; 156:1486–1491.
(23) Dunton CJ, Balsara G, McFarland M,
Hernandez E. Uterine papillary serous carci-noma. A review. Obstet Gynecol Survey
1991; 46:97–102.
(24) Fanning J, Evans MC, Peters AJ, Samuel M,
Harmon ER, Bates JS. Endometrial adeno-carcinoma histologic subtypes: Clinical and
pathologic profile. Gynecol Oncol 1989;
32:288–291.
(25) Kanbour–Shakir A, Tobon H. Primary clear
cell carcinoma of the endometrium. A clini-copathologic study of 20 cases. Int J Gynecol
Pathol 1991; 10:67–78.
(26) Kadar N, Malfetano JH, Homesley HD.
Determinants of survival of surgically staged
patients with endometrial carcinoma histo-logically confined to the uterus: Implications
for therapy. Obstet Gynecol 1992; 80:
655–659.
(27) DiSaia PJ, Creasman WT. Clinical gyneco-logic oncology, 6th ed. St. Louis: Mosby-Year
Book, 2002.
(28) Scully RE, Bonfiglio TA, Kurman RE,
Silverberg SG, Wilkinson EJ. Histological
typing of female genital tract tumours. New
York: Springer-Verlag; 1994.
(28a) Silverberg SG, Kurman RJ, Nogales F, Mutter
GL, Kubik-Nuch RA, Tavassoli FA. Tumours
of the uterine corpus. Epithelial tumours and
related lesions. In: World Health Organiza-tion Classification of tumours. Pathology and
genetics of tumours of the breast and female
genital organs. Tavassoli FA, Devilee P, eds.
Lyon: IARC Press; 2003: 221–232.
(29) Benedet JL, Bender H, Jones H, III, Ngan
HY, Pecorelli S. FIGO staging classifications
and clinical practice guidelines in the man-agement of gynecologic cancers. FIGO
Committee on Gynecologic Oncology. Int J
Gynaecol Obstet 2000; 70:209–262.
(30) Hendrickson MR, Ross JC, Kempson RL.
Toward the development of morphologic cri-teria for well-differentiated adenocarcinoma
of the endometrium. Am J Surg Pathol 1983;
7:819–838.
(31) Kurman RJ, Norris HJ. Evaluation of criteria
for distinguishing atypical endometrial hyper-plasia from well-differentiated carcinoma.
Cancer 1982; 49:2547–2557.
(32) Norris HJ, Tavassoli FA, Kurman RJ.
Endometrial hyperplasia and carcinoma.
Diagnostic considerations. Am J Surg Pathol
1983; 7:839–847.
(33) Kraus FT. High-risk and premalignant
lesions of the endometrium. Am J Surg
Pathol 1985; 9(Suppl):31–40.
(34) Ronnett BM, Kurman RJ. Precursor lesions
of endometrial carcinoma. In: Kurman R, ed.
Blaustein’s pathology of the female genital
tract. 5th ed. New York: Springer-Verlag,
2002: 467–500.
(35) Lehman MB, Hart WR. Simple and complex
hyperplastic papillary proliferations of the
endometrium: A clinicopathologic study of
nine cases of apparently localized papillary
lesions with fibrovascular stromal cores and
242 10. Endometrial Carcinoma
epithelial metaplasia. Am J Surg Pathol 2001;
25:1347–1354.
(36) King A, Seraj IM, Wagner RJ. Stromal inva-sion in endometrial carcinoma. Am J Obstet
Gynecol 1984; 149:10–14.
(37) Silverberg SG, Kurman RJ. Tumors of the
uterine corpus and gestational trophoblastic
disease. Atlas of tumor pathology, 3rd series,
Fascicle 3. Washington, DC: Armed Forces
Institute of Pathology, 1992.
(38) Winkler B, Alvarez S, Richart RM, Crum
CP. Pitfalls in the diagnosis of endometrial
neoplasia. Obstet Gynecol 1984; 64:185–
194.
(39) Crum CP, Richart RM, Fenoglio CM. Ade-noacanthosis of the endometrium. A clinico-pathologic study in premenopausal women.
Am J Surg Pathol 1981; 5:15.
(40) Hendrickson MR, Kempson RL. Endome-trial epithelial metaplasias: Proliferations
frequently misdiagnosed as adenocarcinoma.
Report of 89 cases and proposed classifica-tion. Am J Surg Pathol 1980; 4:525–542.
(41) Blaustein A. Morular metaplasia misdiag-nosed as adenoacanthoma in young women
with polycystic ovarian disease. Am J Surg
Pathol 1982; 6:223.
(42) Clement PB, Young RH. Endometrioid car-cinoma of the uterine corpus: A review of its
pathology with emphasis on recent advances
and problematic aspects. Adv Anat Pathol
2002; 9:145–184.
(43) Andersen WA, Taylor PT, Fechner RE,
Pinkerton JAV. Endometrial metaplasia
associated with endometrial carcinoma. Am
J Obstet Gynecol 1987; 157:597–604.
(44) Clement PB. Pathology of the uterine corpus.
Hum Pathol 1991; 22:776–791.
(45) Zaman SS, Mazur MT. Endometrial papillary
syncytial change. A nonspecific alteration
associated with active breakdown. Am J Clin
Pathol 1993; 99:741–745.
(46) Kaku T, Tsukamoto N, Tsuruchi N, Sugihara
K, Kamura T, Nakano H. Endometrial meta-plasia associated with endometrial carci-noma. Obstet Gynecol 1992; 80:812–816.
(47) Rorat E, Wallach RC. Papillary metaplasia of
the endometrium: Clinical and histopatho-logic considerations. Obstet Gynecol 1984;
64:90S–92S.
(48) Sherman ME, Mazur MT, Kurman RJ.
Benign diseases of the endometrium. In:
Kurman RJ, ed. Blaustein’s pathology of
the female genital tract. 5th ed. New York:
Springer-Verlag; 2002: 421–466.
(49) Mazur MT. Atypical polypoid adenomyomas
of the endometrium. Am J Surg Pathol 1981;
5:473–482.
(50) Young RH, Treger T, Scully RE. Atypical
polypoid adenomyoma of the uterus. A
report of 27 cases. Am J Clin Pathol 1986;
86:139–145.
(51) Iezzoni JC, Mills SE. Nonneoplastic endome-trial signet-ring cells. Vacuolated decidual
cells and stromal histiocytes mimicking
adenocarcinoma. Am J Clin Pathol 2001;
115:249–255.
(52) Scully RE, Poulson H, Sobin LH. Interna-tional Histological Classification and Typing
of Female Genital Tract Tumours. Berlin:
Springer-Verlag; 1994.
(53) Demopoulos RI, Dubin N, Noumoff J, et al.
Prognostic significance of squamous differ-entiation in stage I endometrial adenocarci-noma. Obstet Gynecol 1986; 68:245–250.
(54) Zaino RJ, Kurman RJ, Diana KL, Morrow
CP. The utility of the revised International
Federation of Gynecology and Obstetrics
histologic grading of endometrial adenocar-cinoma using a defined nuclear grading
system. A Gynecologic Oncology Group
study. Cancer 1995; 75:81–86.
(55) Nordstrom B, Strang P, Lindgren A,
Bergstrom R, Tribukait B. Endometrial car-cinoma: the prognostic impact of papillary
serous carcinoma (UPSC) in relation to
nuclear grade, DNA ploidy and p53 expres-sion. Anticancer Res 1996; 16:899–904.
(56) Ayhan A, Taskiran C, Yuce K, Kucukali T.
The prognostic value of nuclear grading and
the revised FIGO grading of endometrial
adenocarcinoma. Int J Gynecol Pathol 2003;
22:71–74.
(57) Alkushi A, Lim P, Coldman A, Huntsman D,
Miller D, Gilks B. Assessment of a novel
grading system for endometrial carcinoma.
Mod Pathol 2003; 16:180A.
(58) Christopherson WM, Alberhasky RC,
Connelly PJ. Carcinoma of the endometrium:
II. Papillary adenocarcinoma: A clinical
pathologic study of 46 cases. Am J Clin
Pathol 1982; 77:534–540.
(59) Daniel AG, Peters WA, III. Accuracy of office
and operating room curettage in the grading
of endometrial carcinoma. Obstet Gynecol
1988; 71:612–614.
(60) Cowles TA, Magrina JF, Masterson BJ,
Capen CV. Comparison of clinical and surgi-cal staging in patients with endometrial car-cinoma. Obstet Gynecol 1985; 66:413–416.
References 243
(61) Piver MS, Lele SB, Barlow JJ, Blumenson L.
Paraaortic lymph node evaluation in stage I
endometrial cancer. Obstet Gynecol 1982;
59:97–100.
(62) Larson DM, Johnson KK, Broste SK,
Krawisz BR, Kresl JJ. Comparison of D&C
and office endometrial biopsy in predicting
final histopathologic grade in endometrial
cancer. Obstet Gynecol 1995; 86:38–42.
(63) Obermair A, Geramou M, Gucer F, Denison
U, Graf AH, Kapshammer E, et al. Endome-trial cancer: Accuracy of the finding of a well
differentiated tumor at dilatation and curet-tage compared to the findings at subsequent
hysterectomy. Int J Gynecol Cancer 1999;
9:383–386.
(64) Nielsen AL, Thomsen HK, Nyholm HCJ.
Evaluation of the reproducibility of the
revised 1988 International Federation of
Gynecology and Obstetrics grading system of
endometrial cancers with special emphasis
on nuclear grading. Cancer 1991; 68:2303–
2309.
(65) Zaino RJ, Silverberg SG, Norris HJ, Bundy
BN, Morrow CP, Okagaki T. The prognostic
value of nuclear versus architectural grading
in endometrial adenocarcinoma. A Gyneco-logic Oncology Group study. Int J Gynecol
Pathol 1994; 13:29–36.
(66) Lax SF, Kurman RJ, Pizer ES, Wu L, Ronnett
BM. A binary architectural grading system
for uterine endometrial endometrioid carci-noma has superior reproducibility compared
with FIGO grading and identifies subsets
of advance-stage tumors with favorable and
unfavorable prognosis. Am J Surg Pathol
2000; 24:1201–1208.
(67) Taylor RR, Zeller J, Lieberman RW,
O’Connor DM. An analysis of two versus
three grades for endometrial carcinoma.
Gynecol Oncol 1999; 74:3–6.
(67a) Scholten AN, Smit VTHBM, Beerman H, van
Putten WLJ, Creutzberg CL. Prognostic
significance and interobserver variablility of
histologic grading systems form endometrial
carcinoma. Cancer 2004; 100:764–772.
(68) Hendrickson M, Ross J, Eifel P, Martinez A,
Kempson R. Uterine papillary serous carci-noma. A highly malignant form of endome-trial adenocarcinoma. Am J Surg Pathol
1982; 6:93–108.
(69) Clement PB, Scully RE. Endometrial hyper-plasia and carcinoma. In: Clement PB, Young
RH, ed. Tumors and tumorlike lesions of the
uterine corpus and cervix. New York:
Churchill Livingstone; 1993: 181–264.
(70) Hendrickson MR, Kempson RL. Ciliated
carcinoma–a variant of endometrial adeno-carcinoma. A report of 10 cases. Int J
Gynecol Pathol 1983; 2:1–12.
(71) Gould PR, Li L, Henderson DW, Barter RA,
Papadimitriou JM. Cilia and ciliogenesis in
endometrial adenocarcinomas. Arch Pathol
Lab Med 1986; 110:326–330.
(72) Tobon H, Watkins GJ. Secretory adenocarci-noma of the endometrium. Int J Gynecol
Pathol 1985; 4:328–335.
(73) Dawagne MP, Silverberg SG. Foam cells in
endometrial carcinoma. Gynecol Oncol 1982;
13:67–75.
(74) Silver SA, Sherman ME. Morphologic and
immunophenotypic characterization of foam
cells in endometrial lesions. Int J Gynecol
Pathol 1998; 17:140–145.
(75) Tiltman AJ. Mucinous carcinoma of the
endometrium. Obstet Gynecol 1980; 55:
244–247.
(76) Czernobilsky B, Katz Z, Lancet M, Gaton E.
Endocervical type epithelium in endometrial
carcinoma: A report of 10 cases with empha-sis on histochemical methods for differen-tial diagnosis. Am J Surg Pathol 1980; 4:
481–489.
(77) Young RH, Scully RE. Uterine carcinomas
simulating microglandular hyperplasia. A
report of six cases. Am J Surg Pathol 1992;
16:1092–1097.
(78) Zaloudek C, Hayashi GM, Ryan IP, Powell
CB, Miller TR. Microglandular adenocarci-noma of the endometrium: A form of muci-nous adenocarcinoma that may be confused
with microglandular hyperplasia of the
cervix. Int J Gynecol Pathol 1997; 16:52–59.
(79) Lee KR, Belinson JL. Recurrence in nonin-vasive endometrial carcinoma. Relationship
to uterine papillary serous carcinoma. Am J
Surg Pathol 1991; 15:965–973.
(80) Sherman ME, Bitterman P, Rosenshein NB,
Delgado G, Kurman RJ. Uterine serous
carcinoma—a morphologically diverse neo-plasm with unifying clinicopathologic
features. Am J Surg Pathol 1992; 16:600–610.
(81) Lee KR, Belinson JL. Papillary serous ade-nocarcinoma of the endometrium: A clinico-pathologic study of 19 cases. Gynecol Oncol
1992; 46:51–54.
(82) Carcangiu ML, Chambers JT. Uterine papil-lary serous carcinoma—A study on 108
244 10. Endometrial Carcinoma
cases with emphasis on the prognostic sig-nificance of associated endometrioid carci-noma, absence of invasion, and concomitant
ovarian carcinoma. Gynecol Oncol 1992; 47:
298–305.
(83) Demopoulos RI, Genega E, Vamvakas E,
Carlson E, Mittal K. Papillary carcinoma of
the endometrium: Morphometric predictors
of survival. Int J Gynecol Pathol 1996; 15:
110–118.
(84) Sutton GP, Brill L, Michael H, et al. Malig-nant papillary lesions of the endometrium.
Gynecol Oncol 1987; 27:294–304.
(85) Lax SF, Kendall B, Tashiro H, Slebos RJ,
Hedrick L. The frequency of p53, K-ras muta-tions, and microsatellite instability differs in
uterine endometrioid and serous carcinoma:
evidence of distinct molecular genetic path-ways. Cancer 2000; 88:814–824.
(86) Sherman ME, Bur ME, Kurman RJ. p53 in
endometrial cancer and its putative precur-sors: Evidence for diverse pathways of
tumorigenesis. Hum Pathol 1995; 26:
1268–1274.
(87) Tashiro H, Isacson C, Levine R, Kurman RJ,
Cho KR, Hedrick L. p53 gene mutations are
common in uterine serous carcinoma and
occur early in their pathogenesis. Am J
Pathol 1997; 150:177–185.
(88) Kovalev S, Marchenko ND, Gugliotta BG,
Chalas E, Chumas J, Moll UM. Loss of p53
function in uterine papillary serous carci-noma. Hum Pathol 1998; 29(6):613–619.
(89) Lax SF, Pizer ES, Ronnett BM, Kurman RJ.
Clear cell carcinoma of the endometrium
is characterized by a distinctive profile of
p53, Ki-67, estrogen, and progesterone
receptor expression. Hum Pathol 1998; 29:
551–558.
(90) Carcangiu ML, Chambers JT, Voynick IM,
Pirro M, Schwartz PE. Immunohistochemical
evaluation of estrogen and progesterone
receptor content in 183 patients with
endometrial carcinoma. Part I: Clinical and
histologic correlations. Am J Clin Pathol
1990; 94(3):247–254.
(91) Kounelis S, Kapranos N, Kouri E, Coppola D,
Papadaki H, Jones MW. Immunohistochemi-cal profile of endometrial adenocarcinoma: A
study of 61 cases and review of the literature.
Mod Pathol 2000; 13:379–388.
(92) Lax SF, Pizer ES, Ronnett BM, Kurman RJ.
Comparison of estrogen and progesterone
receptor, Ki-67, and p53 immunoreactivity
in uterine endometrioid carcinoma and
endometrioid carcinoma with squamous,
mucinous, secretory, and ciliated cell differ-entiation. Hum Pathol 1998; 29:924–931.
(93) Chambers JT, Merino M, Kohorn EI, et al.
Uterine papillary serous carcinoma. Obstet
Gynecol 1987; 69:109.
(94) Jeffrey JF, Krepart GV, Lotocki RJ. Papillary
serous adenocarcinoma of the endometrium.
Obstet Gynecol 1986; 67:670.
(95) Woolas RP, Williams K, McCartney AJ.
Uterine papillary serous carcinoma in a 32-year-old with Turner’s syndrome. Gynecol
Oncol 1994; 52:395–401.
(96) Cirisano FD, Jr, Robboy SJ, Dodge RK,
Bentley RC, Krigman HR, Synan IS, et al. Epi-demiologic and surgicopathologic findings of
papillary serous and clear cell endometrial
cancers when compared to endometrioid car-cinoma. Gynecol Oncol 1999; 74:385–394.
(97) Evans-Metcalf ER, Brooks SE, Reale FR,
Baker SP. Profile of women 45 years of age
and younger with endometrial cancer. Obstet
Gynecol 1998; 91:349–354.
(98) Silva EG, Jenkins R. Serous carcinoma in
endometrial polyps. Mod Pathol 1990;
3:120–128.
(99) McCluggage WG, Sumathi VP, McManus DT.
Uterine serous carcinoma and endometrial
intraepithelial carcinoma arising in endome-trial polyps: Report of 5 cases, including 2
associated with tamoxifen therapy. Hum
Pathol 2003; 34:939–943.
(100) Aquino-Parsons C, Lim P, Wong F, Milden-berger M. Papillary serous and clear cell car-cinoma limited to endometrial curettings in
FIGO stage 1a and 1b endometrial adeno-carcinoma: treatment implications. Gynecol
Oncol 1998; 71:83–86.
(101) Wheeler DT, Bell KA, Kurman RJ, Sherman
ME. Minimal uterine serous carcinoma:
Diagnosis and clinicopathologic correlation.
Am J Surg Pathol 2000; 24:797–806.
(102) Carcangiu ML, Tan LK, Chambers JT. Stage
IA uterine serous carcinoma: A study of 13
cases. Am J Surg Pathol 1997; 21:1507–1514.
(103) Baergen RN, Warren CD, Isacson C, Ellenson
LH. Early uterine serous carcinoma: Clonal
origin of extrauterine disease. Int J Gynecol
Pathol 2001; 20(3):214–219.
(104) Gehrig PA, Groben PA, Fowler WC, Jr,
Walton LA, Van Le L. Noninvasive papillary
serous carcinoma of the endometrium.
Obstet Gynecol 2001; 97:153–157.
References 245
(105) Soslow RA, Pirog E, Isacson C. Endometrial
intraepithelial carcinoma with associated
peritoneal carcinomatosis. Am J Surg Pathol
2000; 24:726–732.
(106) Chan JK, Loizzi V, Youssef M, Osann K,
Rutgers J, Vasilev SA, et al. Significance of
comprehensive surgical staging in noninva-sive papillary serous carcinoma of the
endometrium. Gynecol Oncol 2003; 90:
181–185.
(106a) Slomovitz BM, Burke TW, Eifel PJ,
Ramondetta LM, Silva EG, Jhingran A, et al.
Uterine papillary serous carcinoma (UPSC):
A single institution review of 129 cases.
Gynecol Oncol 2003; 91:463–469.
(107) Sakuragi N, Hareyama H, Todo Y, Yamada H,
Yamamoto R, Fujino T, et al. Prognostic sig-nificance of serous and clear cell adenocarci-noma in surgically staged endometrial
carcinoma. Acta Obstet Gynecol Scand 2000;
79:311–316.
(108) Abeler VM, Vergote IB, Kjorstad KE,
Trope CG. Clear cell carcinoma of the
endometrium. Prognosis and metastatic
pattern. Cancer 1996; 78:1740–1747.
(109) Simon A, Kopolovic J, Beyth Y. Primary squa-mous cell carcinoma of the endometrium.
Gynecol Oncol 1988; 31:454–461.
(110) Melin JR, Wanner L, Schulz DM, Cassell EE.
Primary squamous cell carcinoma of the
endometrium. Obstet Gynecol 1979;
53:115–119.
(111) Goodman A, Zukerberg LR, Rice LW, Fuller
AF, Young RH, Scully RE. Squamous cell
carcinoma of the endometrium: A report of
eight cases and a review of the literature.
Gynecol Oncol 1996; 61:54–60.
(112) Houissa-Vuong S, Catanzano-Laroudie M,
Baviera E, Balaton A, Galet B, Gedeon I,
et al. Primary squamous cell carcinoma of the
endometrium: Case history, pathologic find-ings, and discussion. Diagn Cytopathol 2002;
27:291–293.
(113) Ryder DE. Verrucous carcinoma of the
endometrium—a unique neoplasm with long
survival. Obstet Gynecol 1982; 59:78S–80S.
(114) Spiegel GW, Austin RM, Gelven PL. Transi-tional cell carcinoma of the endometrium.
Gynecol Oncol 1996; 60:325–330.
(115) Lininger RA, Ashfaq R, Albores-Saavedra J,
Tavassoli FA. Transitional cell carcinoma of
the endometrium and endometrial carci-noma with transitional cell differentiation.
Cancer 1997; 79:1933–1943.
(116) Abeler VM, Kjorstad KE, Nesland JM.
Undifferentiated carcinoma of the
endometrium. A histopathologic and clinical
study of 31 cases. Cancer 1991; 68:98–105.
(117) Huntsman DG, Clement PB, Gilks CB, Scully
RE. Small-cell carcinoma of the endo-metrium. A clinicopathological study of
sixteen cases. Am J Surg Pathol 1994; 18:364–
375.
(118) Olson N, Twiggs L, Sibley R. Small-cell carci-noma of the endometrium: Light microscopic
and ultrastructural study of a case. Cancer
1982; 50:760–765.
(119) Paz RA, Frigerio B, Sundblad AS, Eusebi V.
Small-cell (oat cell) carcinoma of the
endometrium. Arch Pathol Lab Med 1985;
109:270–272.
(120) Manivel C, Wick MR, Sibley TK. Neuro-endocrine differentiation in mullerian neo-plasms. An immunohistochemical study of a
“pure” endometrial small-cell carcinoma and
a mixed mullerian tumor containing small-cell carcinoma. Am J Clin Pathol 1986; 86:
438–443.
(121) Meydanli MM, Erguvan R, Altinok MT,
Ataoglu O, Kafkasli A. Rare case of neu-roendocrine small cell carcinoma of the
endometrium with paraneoplastic membra-nous glomerulonephritis. Tumori 2003; 89:
213–217.
(122) Pesce C, Merino MJ, Chambers JT, Nogales
F. Endometrial carcinoma with trophoblastic
differentiation. An aggressive form of uterine
cancer. Cancer 1991; 68:1799–1802.
(123) Deligdisch L, Kase NG, Bleiweiss IJ.
Endometrial cancer in elderly women: A his-tologic and steroid receptor study. Gerontol-ogy 2000; 46:17–21.
(124) Kalir T, Seijo L, Deligdisch L, Cohen C.
Endometrial adenocarcinoma with chorio-carcinomatous differentiation in an elderly
virginal woman. Int J Gynecol Pathol 1995;
14:266–269.
(125) Savage J, Subby W, Okagaki T. Adenocarci-noma of the endometrium with trophoblastic
differentiation and metastases as choriocar-cinoma: A case report. Gynecol Oncol 1987;
26:257–262.
(126) Tunc M, Simsek T, Trak B, Uner M.
Endometrium adenocarcinoma with chorio-carcinomatous differentiation: A case report.
Eur J Gynaecol Oncol 1998; 19:489–491.
(127) Black K, Sykes P, Ostor AG. Trophoblastic
differentiation in an endometrial carcinoma.
246 10. Endometrial Carcinoma
Aust N Z J Obstet Gynaecol 1998; 38:
472–473.
(128) Arends J, Willebrand D, Gans HJD, Swaen
GJV, Bosman FT. Adenocarcinoma of the
endometrium with glassy cell features—
Immunohistochemical observations. Histopa-thology 1984; 8:873–879.
(129) Christopherson WM, Alberhasky RC,
Connelly PJ. Glassy cell carcinoma of the
endometrium. Hum Pathol 1982; 13:418–421.
(130) Pitman MB, Young RH, Clement PB,
Dickersin GR, Scully RE. Endometrioid
carcinoma of the ovary and endometrium,
oxyphilic cell type: A report of nine cases. Int
J Gynecol Pathol 1994; 13:290–301.
(131) Mooney EE, Robboy SJ, Hammond CB,
Berchuck A, Bentley RC. Signet-ring cell
carcinoma of the endometrium: A primary
tumor masquerading as a metastasis. Int J
Gynecol Pathol 1997; 16:169–172.
(132) Vargas MP, Merino MJ. Lymphoepithelioma-like carcinoma: An unusual variant of
endometrial cancer. A report of two cases. Int
J Gynecol Pathol 1998; 17:272–276.
(133) Jones MA, Young RH, Scully RE. Endome-trial adenocarcinoma with a component of
giant cell carcinoma. Int J Gynecol Pathol
1991; 10:260–270.
(134) Joseph MG, Fellows FG, Hearn SA. Primary
endodermal sinus tumor of the
endometrium. A clinicopathologic, immuno-cytochemical, and ultrastructural study.
Cancer 1990; 63:297–302.
(135) Ohta M, Sakakibara K, Mizuno K, Kano T,
Matsuzawa K, Tomoda Y, et al. Successful
treatment of primary endodermal sinus
tumor of the endometrium. Gynecol Oncol
1988; 31:357–364.
(136) Hoshida Y, Nagakawa T, Mano S, Taguchi K,
Aozasa K. Hepatoid adenocarcinoma of the
endometrium associated with alpha–fetopro-tein production. Int J Gynecol Pathol 1996;
15:266–269.
(137) Daya D, Lukka H, Clement PB. Primitive
neuroectodermal tumors of the uterus—A
report of four cases. Hum Pathol 1992; 23:
1120–1129.
(138) Hendrickson MR, Scheithauer BW. Primitive
neuroectodermal tumor of the endometrium:
Report of two cases, one with electron micro-scopic observations. Int J Gynecol Pathol
1986; 5:249–259.
(139) Sinkre P, Albores-Saavedra J, Miller DS,
Copeland LJ, Hameed A. Endometrial
endometrioid carcinomas associated with
Ewing sarcoma/peripheral primitive neuro-ectodermal tumor. Int J Gynecol Pathol
2000; 19:127–132.
(140) Sorensen JB, Schultze HR, Madsen EL,
Holund B. Primitive neuroectodermal
tumor (PNET) of the uterine cavity. Eur J
Obstet Gynecol Reprod Biol 1998; 76:181–
184.
(141) Fraggetta F, Magro G, Vasquez E. Primitive
neuroectodermal tumour of the uterus with
focal cartilaginous differentiation. Histo-pathology 1997; 30:483–485.
(142) Bittencourt AL, Britto JF, Fonseca LEJ.
Wilms’ tumor of the uterus: The first report
of the literature. Cancer 1981; 47:2496–2499.
(143) Muc RS, Grayson W, Grobbelaar JJ. Adult
extrarenal Wilms tumor occurring in the
uterus. Arch Pathol Lab Med 2001; 125:
1081–1083.
(144) Staebler A, Sherman ME, Zaino RJ, Ronnett
BM. Hormone receptor immunohistochem-istry and human papillomavirus in situ
hybridization are useful for distinguishing
endocervical and endometrial adenocarcino-mas. Am J Surg Pathol 2002; 26(8):998–1006.
(145) Kamoi S, AlJuboury MI, Akin MR, Silver-berg SG. Immunohistochemical staining in
the distinction between primary endometrial
and endocervical adenocarcinomas: Another
viewpoint. Int J Gynecol Pathol 2002; 21:
217–223.
(146) McCluggage WG, Sumathi VP, McBride HA,
Patterson A. A panel of immunohistochemi-cal stains, including carcinoembryonic
antigen, vimentin, and estrogen receptor, aids
the distinction between primary endometrial
and endocervical adenocarcinomas. Int J
Gynecol Pathol 2002; 21:11–15.
(147) Zhang X, Lin Z, Lee ES, Kim I. Immuno-histochemical profiles of endometrial and
endocervical adenocarcinomas. Mod Pathol
16[217A]. 2003.
(148) Ansari-Lari MA, Staebler A, Ronnett BM.
Distinction of endocervical and endometrial
adenocarcinomas: p16 expression correlated
with human papillomavirus (HPV) DNA
detection by in situ hybridization (ISH). Mod
Pathol 2003; 16:180A.
(149) McCluggage WG, Jenkins D. p16 immunore-activity may assist in the distinction between
endometrial and endocervical adenocarci-noma. Int J Gynecol Pathol 2003; 22:231–
235.
References 247
(150) Dabbs DJ, Geisinger KR, Norris HT. Inter-mediate filaments in endometrial and endo-cervical carcinomas. The diagnostic utility of
vimentin patterns. Am J Surg Pathol 1986;
10:568–576.
(151) Castrillon DH, Lee KR, Nucci MR. Distinc-tion between endometrial and endocervical
adenocarcinoma: An immunohistochemical
study. Int J Gynecol Pathol 2002; 21:4–10.
(152) Cohen C, Shulman G, Budgeon LR. Endo-cervical and endometrial adenocarcinoma:
An immunoperoxidase and histochemical
study. Am J Surg Pathol 1982; 6:151–157.
(153) Tamimi HK, Gown AM, Kimdeobald J, Figge
DC, Greer BE, Cain JM. The utility of
immunocytochemistry in invasive adenocar-cinoma of the cervix. Am J Obstet Gynecol
1992; 166:1655–1662.
(154) Maes G, Fleuren GJ, Bara J, Nap M. The
distribution of mucins, carcinoembryonic
antigen, and mucus-associated antigens in
endocervical and endometrial adenocarcino-mas. Int J Gynecol Pathol 1988; 7:112–122.
(155) Kumar A, Schneider V. Metastases to the
uterus from extrapelvic primary tumors. Int J
Gynecol Pathol 1983; 2:134–140.
(156) Kumar NB, Hart WR. Metastases to the
uterine corpus from extagenital cances. A
clinicopathologic study of 63 cases. Cancer
1982; 50:2163–2169.
(157) Mazur MT, Hsueh S, Gersell DJ. Metastases
to the female genital tract. Analysis of 325
cases. Cancer 1984; 53:1978–1984.
(158) Ulbright TM, Roth LM. Metastatic and inde-pendent cancers of the endometrium and
ovary: A clinicopathologic study of 34 cases.
Hum Pathol 1985; 16:28–34.
(159) Eifel P, Hendrickson M, Ross J, Ballon S,
Martinez A, Kempson RL. Simultaneous pre-sentation of carcinoma involving the ovary
and uterine corpus. Cancer 1982; 50:163–
170.
(160) Gown AM. Immunohistochemical determi-nation of primary sites of carcinomas. J His-totechnol 1999; 22:209–215.
(161) Agoff SN, Grieco VS, Garcia R, Gown AM.
Immunohistochemical distinction of
endometrial stromal sarcoma and cellular
leiomyoma. Appl Immunohistochem Mol
Morphol 2001; 9:164–169.
(162) Randall TC, Kurman RJ. Progestin treatment
of atypical hyperplasia and well–differenti-ated carcinoma of the endometrium in
women under age 40. Obstet Gynecol 1997;
90:434–440.
248 10. Endometrial Carcinoma
MMMT typically arises in the endometrium
and is often first diagnosed by biopsy or curet-tage. The remaining tumors are extremely
infrequent in endometrial curettings. Even
submucosal leiomyomas rarely are sampled by
an endometrial biopsy or curettage. It has been
reported that prior to a hysterectomy the diag-nosis of endometrial stromal sarcoma is missed
in 20% of cases.5
Similarly for leiomyosarcoma,
inaccurate or inconclusive diagnosis on curet-tings ranges from 20% to 80%.5–7
Although
these tumors have morphologic features that
readily assist in their recognition, it is often
impossible to classify them accurately from
biopsy material alone.
Malignant Mixed Mesodermal
Tumor (Carcinosarcoma)
The MMMT is a highly malignant, biphasic
tumor consisting of both epithelial and mes-enchymal components. Also referred to as
“malignant mixed müllerian tumor” or “carci-nosarcoma,” it is predominantly a neoplasm of
the postmenopausal patient. The median age at
diagnosis is about 65 years.1;3
As with endome-trial carcinoma, the patient with MMMT typi-cally presents with postmenopausal bleeding.
The uterus is almost always enlarged, and the
tumor may present as a polypoid mass that
prolapses through the cervical os.
In the past these tumors had been treated as
sarcomas,4
but recent molecular genetic analy-sis indicates that these neoplasms are clonal.8–10
249
11
Other Tumors
Uterine tumors other than benign polyps or
carcinoma are rarely encountered in endome-trial biopsies and curettings. The classification
in Table 11.1 lists most of these other neo-plasms. Among this group of tumors, malignant
mixed mesodermal tumor (MMMT) is the most
common malignancy, yet it accounts for only
1% to 2% of all uterine neoplasms.1– 4
The
Malignant Mixed Mesodermal Tumor
(Carcinosarcoma)  . . . . . . . . . . . . . . . . . . . 249
Pathologic Features  . . . . . . . . . . . . . . . . 250
Differential Diagnosis  . . . . . . . . . . . . . . 253
Clinical Queries and Reporting . . . . . . . 255
Mullerian Adenofibroma and
Adenosarcoma  . . . . . . . . . . . . . . . . . . . . . 256
Pathologic Features  . . . . . . . . . . . . . . . . 256
Differential Diagnosis  . . . . . . . . . . . . . . 259
Clinical Queries and Reporting . . . . . . . 259
Stromal Tumors  . . . . . . . . . . . . . . . . . . . . . 260
Stromal Nodule and Low-Grade
Stromal Sarcoma  . . . . . . . . . . . . . . . . . . 260
High-Grade Stromal Sarcoma  . . . . . . . . 262
Differential Diagnosis  . . . . . . . . . . . . . . 263
Clinical Queries and Reporting . . . . . . . 264
Smooth Muscle Tumors  . . . . . . . . . . . . . . . 264
Leiomyomas  . . . . . . . . . . . . . . . . . . . . . 264
Variants of Leiomyoma  . . . . . . . . . . . . . 264
Tumorlets . . . . . . . . . . . . . . . . . . . . . . . . 265
Leiomyosarcoma  . . . . . . . . . . . . . . . . . . 267
Clinical Queries and Reporting . . . . . . . 267
Rare Neoplasms  . . . . . . . . . . . . . . . . . . . . 267
Lymphoma and Leukemia . . . . . . . . . . . 267
Miscellaneous Tumors  . . . . . . . . . . . . . . 268
Other Lesions and Tumor-Like
Conditions  . . . . . . . . . . . . . . . . . . . . . . . . . 269
They are therefore now thought to represent
metaplastic or sarcomatoid carcinomas.1;3;11–18
Immunohistochemical studies show that both
the carcinomatous and sarcomatous elements
are reactive for keratin and epithelial mem-brane antigen (EMA), although immunos-taining is less diffuse in the sarcomatous
foci.11;13–15;19;20
In addition, the tumors spread
like carcinomas rather than sarcomas, and the
metastases are almost always carcinoma, even
if the bulk of the primary tumor is sarcoma-tous.12;13
Thus the behavior of MMMT is dic-tated by the epithelial component. Nonetheless,
these neoplasms are very aggressive and should
be specifically classified as MMMT or carci-nosarcoma whenever possible.
Pathologic Features
MMMTs are often polypoid and can produce
abundant tissue in a curettage specimen. They
may be small and confined to a polyp.12;21
MMMTs are biphasic tumors characterized
by a combination of carcinoma and
sarcoma.1;3;12;13;21–26
Usually the components are
intimately admixed, with nests and masses of
carcinoma separated by a malignant spindle
cell stroma (Figs. 11.1 and 11.2). The carcinoma
often appears to blend into the sarcoma, with
the cells at the interface lacking clear features
of either carcinoma or sarcoma (Fig. 11.3), but
occasionally the two cell types appear as sepa-rate foci.
The carcinomatous component of MMMT
typically is a high-grade adenocarcinoma that is
usually endometrioid type and may show squa-mous change. Tumors may have an epithelial
component with serous, clear cell, mucinous or
undifferentiated carcinoma (Fig. 11.4), and in
our experience serous carcinoma is one of the
most frequent epithelial elements. The epithe-lial component rarely may have a predominant
pattern of squamous carcinoma.
The sarcomatous component can display a
variety of patterns that have been referred to
as either homologous or heterologous.1;22–24
The sarcoma is considered homologous when
it shows differentiation toward mesenchymal
cells normally present in the uterus. Often the
homologous sarcoma resembles fibrosarcoma
or malignant fibrous histiocytoma composed of
interlacing primitive spindle cells (Fig. 11.4),
but at times it may have features of endome-trial stromal sarcoma or leiomyosarcoma. The
sarcoma is considered heterologous when it
is differentiated into cell types not normally
found in the uterus (Fig. 11.5). The usual
heterologous components are rhabdomy-osarcoma, chondrosarcoma, or osteosarcoma.
Other heterologous components, including
liposarcoma or even glia, may rarely be
found.27;28
Some MMMTs have a sarcomatous compo-nent that contains large cells with abundant,
eosinophilic cytoplasm. These large cells
superficially resemble rhabdomyoblasts, but
they lack sarcomeric filaments and are not
immunoreactive for desmin, myogenin, or
myoglobin. Other MMMTs may have dense
eosinophilic hyaline material between the
spindle cells that is suggestive, but not diag-nostic, of osteosarcoma, lacking unequivocal
osteoid and osteoblastic-like stromal cells.
Neither large, eosinophilic cells nor hyaline
stroma should be construed as evidence of
heterologous differentiation unless the
pattern shows clearly diagnostic features. The
250 11. Other Tumors
Table 11.1. Uterine tumors other than carcinoma
that may be found in biopsies and curettage
specimens.
Mixed epithelial–stromal tumors
Malignant mixed mesodermal tumor (MMMT)
(carcinosarcoma)a
Adenofibroma
Adenosarcoma
Stromal tumors
Stromal nodule
Low-grade stromal sarcoma
High-grade stromal sarcoma
Smooth muscle tumors
Leiomyoma
Variants of leiomyoma
Tumorlets
Leiomyosarcoma
Other neoplasms
Lymphoma/leukemia
Rare tumors
a
These are now considered metaplastic or sarcomatoid
carcinomas.
Figure 11.1. Malignant mixed mesodermal tumor. This typical pattern shows a mixture of adenocarcinoma,
endometrioid-type, and a sarcomatous stroma.
Figure 11.2. Malignant mixed mesodermal tumor. Adenocarcinoma, endometrioid-type, merges into a sar-comatous stroma in the lower portion of this field.
Figure 11.3. Malignant mixed mesodermal tumor. Adenocarcinoma blends into the sarcoma. Several ill-defined clusters of cells (arrows) are difficult to classify as carcinoma or sarcoma.
Figure 11.4. Malignant mixed mesodermal tumor.
The carcinomatous component is serous type with
coarse papillae, and the mesenchyme shows undif-ferentiated sarcoma composed of irregular and hap-hazard spindle-shaped cells. The primitive spindle
cells of the sarcomatous component resemble malig-nant fibrous histiocytoma.
distinction, however, is largely academic, as
discussed later.
Rarely an MMMT may show well-differentiated carcinoma, a relatively low-grade
sarcoma, or a combination of both instead of
the high-grade patterns usually seen in the car-cinoma and sarcoma of MMMTs (Figs. 11.6 and
11.7). These tumors nonetheless show features
of malignancy in both the glands and the
stroma. The glands show, at a minimum, the
features of well-differentiated adenocarcinoma
as outlined earlier (see Chapter 10). Low-grade
sarcoma lacks high cellularity and nuclear pleo-morphism but shows a cellular population of
enlarged mesenchymal cells and a high mitotic
rate of four or more mitoses per 10 high power
fields (HPFs). These latter tumors can be diag-nosed as low-grade MMMT.
Differential Diagnosis
For most cases of MMMT, the differential
diagnosis centers mainly on carcinoma. The
mesenchymal component may be indistinct or
difficult to differentiate from undifferentiated
carcinoma, carcinoma with a spindle cell com-ponent, or high-grade adenocarcinoma with
squamous differentiation (adenosquamous car-cinoma). In the latter case of adenocarcinoma
with squamous change, the squamous com-ponent may have a spindle appearance with
swirling aggregates of elongate, nonkeratiniz-ing squamous cells. The squamous component
is within the center of the glands, in contrast to
the sarcoma of the MMMT, which surrounds
the glands (see Chapter 10, Fig. 10.17). At times
the squamous element may be extensive and
Malignant Mixed Mesodermal Tumor (Carcinosarcoma) 253
Figure 11.5. Heterologous elements in malignant
mixed mesodermal tumors. Left: A nodule of carti-laginous differentiation (arrows) is seen in an area of
otherwise undifferentiated sarcoma. Right: A focus
of rhabdomyosarcoma is present showing multiple
rhabdomyoblasts with abundant eosinophilic cyto-plasm. Although cross striations are not seen in
this field, immunohistochemical stains for desmin
and myoglobin demonstrated cytoplasmic muscle
filaments.
Figure 11.6. Malignant mixed mesodermal tumor. In this case the malignant glands appear well differenti-ated, and the stroma lacks the marked cellularity often seen in these neoplasms.
Figure 11.7. Malignant mixed mesodermal tumor.
Well-formed but malignant glands blend into a sar-comatous stroma. Although both the carcinoma and
the sarcoma lack high-grade features, both compo-nents are malignant, establishing the diagnosis of
mixed mesodermal tumor (same case as Fig. 11.6).
efface glandular structures, thereby simulating
a sarcoma. Immunohistochemical stains for
keratin should help resolve the diagnosis, as
discussed later.
In some cases the sarcoma is very limited
in the curettage specimen and is eclipsed by
a dominant carcinomatous element. Careful
study of all the material may be necessary to
correctly diagnose an MMMT with a subtle sar-comatous component. The sarcomatous areas
may blend imperceptibly into the carcinoma-tous component, making the sarcoma difficult
to distinguish from poorly differentiated carci-noma. In these cases, the malignant spindle cells
have an indistinct interface with the epithelial
component, and there is a zone of cells that
are indeterminate in their differentiation into
epithelium or mesenchyme (Figs. 11.2 and 11.3).
We have found that this feature, the merging
of spindle cells into the clearly carcinomatous
component, is especially useful for recognizing
MMMTs in cases where the sarcoma is largely
overshadowed by the carcinoma.
It is important to scrutinize the stroma
between glands or nests of carcinomatous
epithelium to determine whether the mes-enchyme has malignant features, such as
increased cellularity, pleomorphic nuclei, and a
high mitotic rate. Non-neoplastic mesenchyme
in adenocarcinoma either resembles normal
endometrial stromal or shows desmoplasia,
with cells developing fibroblast-like character-istics. Benign mesenchymal cells have a low
mitotic rate; stromal cellularity may be mildly
increased, but the nuclei remain oval and
uniform with indistinct chromatin. Sarcoma-tous tissue, in contrast, shows increased cellu-larity; the nuclei are irregular and closely
packed; and mitotic figures are numerous,
usually in the range of four or more per 10
HPFs. Atypical mitoses often are present.
In questionable cases where the distinction
between MMMT and poorly differentiated
adenocarcinoma is not clear by routine histol-ogy, immunohistochemical stains for keratin
and EMA, can be useful for demonstrating the
biphasic pattern of the MMMT.11;13–15;19;20;29
The
epithelial component shows intense and gener-alized reactivity for keratin and EMA, as noted
earlier, and although the epithelial markers are
also present in the sarcomatous component,
this staining is less intense. These immunostains
are especially useful, as they highlight the
mixed population of carcinoma and sarcoma in
the MMMT. Mesenchymal markers also can
help to identify an MMMT by highlighting
the sarcomatous areas. Actin, muscle-specific
actin, and desmin immunostains are useful
for demonstrating muscle differentiation in the
spindle cell population. Desmin, myoglobin,
and especially myogenin also assist in demon-strating rhabdomyoblasts.29;30
Clinical Queries and Reporting
MMMT is an important diagnosis, because the
tumor is highly aggressive. Correct diagnosis
prior to hysterectomy alerts the gynecologist to
the high likelihood of extrauterine spread at
the time of laparotomy. In addition, the gyne-cologist may wish to stage the patient with an
MMMT more thoroughly. The question of het-erologous versus homologous sarcoma in an
MMMT is of less importance than the recogni-tion of the MMMT. Heterologous elements
have no influence on the prognosis. When the
epithelial component is serous or clear cell car-cinoma, there is a greater tendency for deep
myometrial invasion and metastatic disease, so
these cell types should be noted.12;31
In some cases biopsy or curettage may show
a tumor with features that suggest an MMMT
but are not conclusive. This problem is most
often seen when the sections show adenocarci-noma with foci of cellular spindle cells around
the glands. The spindle cell component may
show increased mitotic activity but is not con-clusive for a sarcomatous element. In such cases
it is best to clearly describe the changes and to
indicate the possibility that the tumor may be
an MMMT. The knowledge that the tumor may
be an MMMT often is sufficient for clinical
management in specimens with equivocal fea-tures, as a hysterectomy will be performed. The
entire uterus can then be studied in greater
detail to determine whether a sarcomatous
component is present.
An occasional case demonstrates only high-grade sarcoma with no definite epithelial ele-ments, despite immunohistochemical analysis.
In these cases the differential diagnosis
includes a pure sarcoma or an MMMT with
Malignant Mixed Mesodermal Tumor (Carcinosarcoma) 255
predominance of the sarcomatous component.
Practically, this distinction is not important for
directing further therapy. Both types of tumor
are high-grade malignancies with a similar
poor prognosis. Consequently, when the tumor
type is not clear-cut, a diagnosis of “malignant
tumor” with a comment indicating the differ-ential should be made. The diagnosis usually is
clarified by the study of additional tissue. If the
patient is a candidate for surgery, hysterectomy
usually is attempted to stage the tumor, and the
neoplasm can be fully assessed by examining
the entire uterus.
Mullerian Adenofibroma
and Adenosarcoma
Adenofibroma and adenosarcoma are related
tumors.1;3;22;32–37
Both are biphasic tumors with
benign glands regularly distributed in a cellular
mesenchyme. The adenofibroma is benign
whereas the adenosarcoma has low-grade
malignant features. These are rare neoplasms
that usually arise in the corpus, but about 10%
of the tumors originate in the endocervix. They
generally occur in postmenopausal patients,
although approximately 30% of these tumors
are found in patients under the age of 50. Rare
cases have occurred in teenagers. A few cases
of adenosarcoma have been associated with
estrogen use, the Stein–Leventhal syndrome, or
prior pelvic irradiation.32
Adenosarcoma has
been associated with tamoxifen use.31;38
On
clinical examination, adenosarcoma typically is
a large, polypoid tumor that expands the
endometrial cavity and often prolapses through
the cervical os.32;34;39
Adenofibroma may recur within the uterus
but does not metastasize or extend beyond the
uterus. Adenosarcoma, however, recurs in the
vagina or pelvis in as many as 25% of cases and
sometimes disseminates widely as a high-grade
sarcoma. Although the terminology suggests
that there is a clear histologic difference
between the benign adenofibroma and the
malignant adenosarcoma, many of these
tumors have borderline features that make sep-aration of the two tumor types difficult. Of the
two, adenosarcoma is much more common,
accounting for more than 90% of all tumors
in this group of adenofibroma and adenosar-coma.3
Hysterectomy is the treatment of choice
for adenosarcoma, but adenofibroma may
be managed more conservatively with repeat
curettage and hysteroscopy.
Pathologic Features
Adenofibroma is characterized by benign
cytologic features of the epithelium and
stroma.34;36;37
The tumor shows a character-istic pattern of broad-based papillary fronds
covered with bland epithelium and supported
by widely distributed spindle cells (Fig. 11.8).
Low columnar epithelium lines large polypoid
fronds of sparsely cellular mesenchyme and
also forms small glandular infoldings into the
stroma. The epithelium lacks atypia or mitotic
activity. The mesenchyme is composed of
widely spaced spindle cells that also lack
mitotic activity (Fig. 11.9).
Adenosarcoma has a low-power growth
pattern similar to that of adenofibroma charac-terized by leaf-like fronds of mesenchyme
covered by cuboidal to columnar epithelium
(Fig. 11.10). In adenosarcoma, however, the
stroma is more cellular, a feature indicating low-grade malignant growth (Fig. 11.11).1;3;22;32;34;35
This tumor has irregular, often large glands that
form gaping cystic spaces. Besides the surface
papillary growth, the glands often have broad-based papillary infoldings into the lumen. In
adenosarcoma, glands and surface epithelium
lack cytologic features of malignancy, although
they often appear hyperplastic, with mitotic
activity and slight nuclear atypia. Usually the
glands have an endometrioid appearance, but
they may show eosinophilic, mucinous, squa-mous, or clear cell change.
In adenosarcoma, the sarcomatous compo-nent usually resembles endometrial stromal
sarcoma with plump spindle cells in a rich vas-cular background, but the mesenchyme can
have features of fibrosarcoma or leiomyosar-coma and may show heterologous elements.
The cellularity of the stroma often is accentu-ated around the glands, forming hypercellular
cuffs with less cellular mesenchyme further
away from the glands. The mitotic rate in the
256 11. Other Tumors
Figure 11.8. Adenofibroma. The tumor is characterized by branching bulbous papillary fronds with a thin
epithelial lining and sparsely cellular stroma.
Figure 11.9. Adenofibroma. At high magnification
the tumor shows bland fibrous stroma covered by
cuboidal epithelium. Both glandular and stromal ele-ments lack malignant features. It is important that
the entire tumor have these features and that there
be no areas of increased cellularity or stromal atypia
in order to establish the diagnosis of adenofibroma.
Figure 11.10. Adenosarcoma. The tumor is composed of large papillary fronds and glands in a highly cel-lular stroma. In this case the stroma resembles endometrial stroma.
Figure 11.11. Adenosarcoma. The stroma forms hypercellular cuffs around the glands. The glands lack
malignant cytologic features.
stroma is variable, and the reported cases have
ranged between 1 and 40 mitoses per 10 HPFs.
Most cases show four or more mitoses per 10
HPFs, but any convincing mitotic activity of as
little as two or more mitoses per 10 HPFs is suf-ficient to establish a diagnosis of adenosarcoma
in a case that shows an otherwise typical growth
pattern.32
Up to 20% of adenosarcomas have a het-erologous element, such as cartilage, fat,
osteoid, or rhabdomyoblasts, that appears to
have no influence on the overall prognosis of
the tumor.32
In occasional cases adenosarcoma
shows overgrowth of sarcoma, either stromal
sarcoma or, less commonly, leiomyosarcoma
or rhabdomyosarcoma.35;40
Adenosarcoma with
stromal overgrowth is important to recognize,
as this tumor has a higher frequency of myome-trial invasion, recurrence, and metastases than
the typical adenosarcoma.40
Adenosarcoma may also show sex-cord–like
elements characterized by ribbons or nests of
cells, sometimes with tubule formation, resem-bling ovarian sex cord tumors.41
These patterns
also may include cells with abundant clear to
foamy cytoplasm.
Differential Diagnosis
The distinction of adenofibroma from adeno-sarcoma may be difficult, because the mitotic
rate is relatively low in some cases of adenosar-coma. As mentioned earlier, in biopsy material
virtually any convincing mitotic activity in an
otherwise characteristic tumor indicates a diag-nosis of adenosarcoma. In our experience, ade-nofibroma is an extraordinarily rare diagnosis
in biopsy and curettage specimens and should
be made only when no mitoses are found in the
stroma. In occasional cases, however, the dis-tinction between the two lesions may be
extremely difficult, even with repeated evalua-tions, and hysterectomy is necessary to allow
thorough sampling and ensure complete
removal of the lesion.
Either adenofibroma or adenosarcoma may
resemble an endometrial polyp. In contrast to
polyps, these lesions characteristically have a
leaflike configuration, and the biopsy consists of
large fragments of tissue. The increased cellu-larity of the stroma and periglandular cuffing in
adenosarcoma are helpful features in the dif-ferential diagnosis. If a patient has a history
of recurrence of apparent endometrial polyps,
especially large polyps, the differential diagno-sis should include adenosarcoma, as recurrence
of ordinary polyps is unusual.
The atypical polypoid adenomyoma may
resemble adenosarcoma because of its mixture
of atypical glands in a cellular mesenchyme (see
Chapter 8). Unlike adenosarcoma, the mes-enchyme of the atypical polypoid adenomyoma
is composed of short, interlacing fascicles of
smooth muscle cells. The atypical polypoid ade-nomyoma is, in general, a smaller lesion, lacking
the leaflike papillary fronds of adenosarcoma.
In addition, many of the glands in the atypical
polypoid adenomyoma show nests of nonkera-tinizing squamous (morular) change.
Clinical Queries and Reporting
Because adenofibroma and adenosarcoma
are unusual neoplasms, a comment regarding
the nature of the tumor is warranted. The
comment can include a brief description with a
statement regarding the mitotic rate of the
stroma. Other salient features of adenosar-coma, such as heterologous elements or sarco-matous overgrowth, should be noted in a
comment. Sarcomatous overgrowth is an espe-cially important feature, as it is associated with
a higher rate of recurrence. A telephone call can
help to clarify the diagnosis for a gynecologist
who is not familiar with the entity.
In small samples it may be difficult to estab-lish the diagnosis of either adenofibroma or
adenosarcoma, as these neoplasms often
have relatively bland cytologic features. If the
morphologic features suggest either of these
entities but are not clearly diagnostic, the
gynecologist should be informed of the differ-ential in order to allow further evaluation as
clinically indicated. Often with this informa-tion the clinician may elect to proceed with a
hysterectomy in the perimenopausal or post-menopausal patient. Hysteroscopy, magnetic
resonance imaging (MRI), and repeat curettage
can be performed in young women in whom
conservative management is desirable.
Mullerian Adenofibroma and Adenosarcoma 259
Stromal Tumors
Tumors in this category include the benign
stromal nodule, low-grade stromal sarcoma,
and high-grade stromal sarcoma.1;22;42–49
These neoplasms frequently involve the
endometrium, so endometrial biopsy or curet-tage may sample them. Low-grade stromal
sarcoma recurs in up to one half of cases, often
in the pelvis or abdomen following hysterec-tomy.43–45;48;50–52
Relapses may occur many years
after hysterectomy. This lesion also has been
termed “endolymphatic stromal myosis” or
“stromatosis.”44;50;53;54
High-grade stromal
sarcomas are aggressive neoplasms that spread
widely.1;43;45;46;49
Most patients with high-grade
stromal sarcoma succumb within 3 years.
Stromal tumors occur over a wide age range,
with a mean age in the fifth decade in most
studies. The majority of low-grade stromal sar-comas occur in women under the age of 50
years,50
whereas high-grade stromal sarcoma
generally occurs in older women.43;49
These
tumors frequently present with abnormal
bleeding, although they may present with pelvic
pain, with other, nonspecific complaints, or even
with distant metastases to the lung or other
sites.44–46;48
Some cases are not detected by
curettage and are diagnosed only in hysterec-tomy specimens.43
Stromal Nodule and Low-Grade
Stromal Sarcoma
The stromal nodule and low-grade stromal
sarcoma are composed of cells that resemble
endometrial stroma of the normal proliferative
phase.1;42;43;45;47
Cytologically these lesions are
identical and are distinguished by their growth
patterns (see later). The cells are oval and
uniform and are supported by a prominent
network of small vascular spaces (Figs. 11.12
260 11. Other Tumors
Figure 11.12. Low-grade stromal sarcoma. The
tumor is composed of a uniform population of oval
mesenchymal cells that resemble the stromal cells of
proliferative phase endometrium in a background of
multiple small vessels. The circumferential orienta-tion of the stromal cells around the blood vessels is
a characteristic feature.
and 11.13). The mitotic rate is variable, but most
tumors show no more than three mitotic figures
per 10 HPFs. Some tumors can show high
mitotic counts, however, and mitotic counts are
not a criterion to distinguish stromal nodules
or low-grade or high-grade stromal sarcoma.42
Stromal nodules and low-grade stromal
sarcoma do not show necrosis or cellular pleo-morphism (Fig. 11.12). Occasional cases show
small aggregates of foam cells admixed.55
Small
amounts of smooth muscle may be interspersed
between areas with typical stromal differentia-tion, but this finding does not alter the diagno-sis.56;57
In fact, immunohistochemical stains
show reactivity of many of the cells for actin
although stromal tumors demonstrate either
weak or absent reactivity to desmin.55;57–60
The
smooth muscle differentiation may result in a
“starburst” pattern of collagen deposition.56
Another immunohistochemical feature of
stromal tumors is reactivity for CD10, which is
usually diffuse and moderate to strong in inten-sity within these tumors.57;58;61;62
These tumors
can show focal reactivity for keratin63
although
they generally are keratin negative.57
The stromal nodule is distinguished from
the low-grade stromal sarcoma by the growth
pattern, a feature that usually cannot be
appreciated in curettage specimens.1;42;45;47;64
The stromal nodule is a circumscribed tumor
with a rounded, pushing interface with the
myometrium or endometrium. Low-grade
stromal sarcoma, in contrast, has an infiltrative
pattern of growth, showing irregular invasion of
the myometrium, often accompanied by
endovascular growth.
Low-grade stromal tumors may show sex-cord–like patterns with nests and trabeculae of
Stromal Tumors 261
Figure 11.13. Low-grade stromal sarcoma. Higher
magnification of the tumor shows a monotonous
population of cells with small nuclei and scant cyto-plasm. In this case mitotic figures are infrequent, but
the cytologic features rather than the mitotic rate
establish the diagnosis of a low-grade stromal tumor.
The pattern in this field could represent either a
stromal nodule or a low-grade stromal sarcoma. To
firmly establish the diagnosis of low-grade stromal
sarcoma, it is necessary to identify infiltrating
margins or endovascular growth, features that
usually cannot be assessed in curettage specimens.
condensed cells forming distinctive epithelial-like patterns that resemble ovarian sex cord
tumors (Fig. 11.14).45;47;56;60;65;66
These epithelioid
nests often form anastomosing trabeculae that
yield a plexiform pattern. The nests and cords
may be solid or they may form tubules with a
central lumen. Immunohistochemical stains
for actin and desmin show reactivity in the sex-cord–like areas that resembles that found in
the spindle cell component.59;60;66
Sex-cord–like
areas also may be immunoreactive for keratin.60
Stromal sarcoma also can show a component of
epithelioid or rhabdoid appearing cells with
abundant eosinophilic cytoplasm.67;68
Stromal tumors can also show scattered foci
of benign endometrial glands and may even
show extensive endometrial glandular differen-tiation, but the presence of glands does not
alter the diagnosis.1;69
Tumors with this finding
should not be classified as carcinosarcomas, as
the glands are benign. Stromal tumors with
glandular differentiation are distinguished
from adenosarcomas by their growth pattern,
their lack of leaflike polypoid fronds, and the
presence of a prominent vascular framework in
the tumor. They resemble low-grade endome-trial stromal sarcoma except for the glandular
differentiation.
High-Grade Stromal Sarcoma
High-grade stromal sarcoma is a tumor with
obvious malignant characteristics, including
increased cellularity, pleomorphism, nuclear
atypia, and a high mitotic rate.1;43;46;49
Atypical
mitotic figures and foci of necrosis are usually
262 11. Other Tumors
Figure 11.14. Low-grade stromal sarcoma. Left:
Portion of low-grade stromal sarcoma has an epithe-lioid pattern in which the cells show a vague plexi-form arrangement with interlacing linear arrays. A
small amount of endometrial surface epithelium is
present to the left of the field. Right: At high magni-fication the cells have epithelioid qualities with dis-cernible cell borders. The cells remain monotonous,
however, and are distributed around a network of
many small vascular spaces.
present. An occasional tumor shows some
resemblance to low-grade stromal sarcoma,
with a population of more uniform ovoid cells
but with a high mitotic rate and necrosis. Many
of these tumors, however, demonstrate nuclear
pleomorphism and marked atypia, features that
result in little resemblance to normal endome-trial stroma. The term “high-grade undifferen-tiated sarcoma” has been proposed for these
tumors, since their histogenetic relationship to
endometrial stroma is less certain.49
Regardless
of terminology, this is an aggressive, high-grade
sarcoma with a grave prognosis and is therefore
important to recognize.
Differential Diagnosis
Stromal nodules and low-grade stromal sarco-mas have very characteristic histologic features
that limit the differential diagnosis to only a few
other lesions. It is most important, however, to
attempt to distinguish these tumors from each
other. The growth pattern determines whether
a tumor is a stromal nodule or a low-grade
stromal sarcoma. Usually this feature is not dis-cernible in biopsy material, as the interface with
normal tissues is, at best, fragmented. Endovas-cular growth and infiltrative growth cannot be
evaluated in biopsies. For these reasons it may
not be possible to distinguish between these
lesions in biopsy material (see later, Clinical
Queries and Reporting).
Most high-grade stromal sarcomas are less
differentiated and lack a monotonous cell pop-ulation. In these cases it may be difficult to
determine whether the tumor is a high-grade
stromal sarcoma or another type of sarcoma,
such as leiomyosarcoma. Practically, this dis-tinction usually makes little difference, as the
clinical management rests on the recognition of
high-grade sarcoma rather than the histologic
subtype. If the differential includes poorly dif-ferentiated or undifferentiated carcinoma with
a sheetlike proliferation of small, monotonous
cells, immunostains for keratin, EMA, CD10,
actin, and desmin should resolve the question.
Stromal tumors show reactivity for CD10, actin
and, focally, desmin whereas carcinoma does
not. Keratin may be focally positive in stromal
tumors,63
so focal reactivity for this antigen
does not necessarily indicate carcinoma. EMA,
however, in our experience, is more specific as
a marker for carcinoma and is not found in
smooth muscle or stromal cells.
At times stromal tumors show foci of smooth
muscle, and sometimes there is a prominent
mixture of stromal and smooth muscle spindle
cells. For practical purposes these tumors
should be classified as stromal tumors, because
their behavior will be similar to that of a pure
stromal neoplasm. If immunohistochemical
stains are employed to help make the distinc-tion, then the diagnosis of a smooth muscle
tumor should be reserved for those cases that
show extensive cytoplasmic staining for desmin.
Tumors that show extensive smooth muscle dif-ferentiation (>30%) can be classified as mixed
endometrial stromal–smooth muscle tumors.70
Cellular leiomyomas (see later) can be diffi-cult to distinguish from low-grade stromal
tumors, as both lesions show closely spaced,
small, spindle-shaped cells. In curettage
specimens, cellular leiomyomas are very
unusual, however, and stromal tumors are more
likely to be biopsied. Low-grade stromal
tumors also show the characteristic pattern of
uniform cells in a network of small vascular
channels, which is distinctive for stromal tumors
but not for smooth muscle neoplasms. Stromal
tumors also may have aggregates of foam
cells.55
Conversely, smooth muscle tumors have
a fascicular pattern with interlacing bundles of
elongate cells. Cellular leiomyomas also have
thicker vessels with muscular walls and
artifactual, cleft-like spaces separating some of
the fascicles.55
Immunohistochemistry also is
helpful in distinguishing a stromal tumor from
a cellular leiomyoma. Stromal tumors are
generally negative for desmin and h-caldesmon
whereas cellular leiomyomas are positive.57;71
Although stromal tumors tend to be diffusely
reactive for CD10, reactivity has also been seen
in some cellular leiomyomas, so this antigen
should be interpreted in the context of a panel
of antibodies.57
Occasionally, aggregates of histiocytes from
the uterine cavity may present as a sheet of
monotonous cells in an endometrial biopsy
that can superficially mimic a proliferation of
stromal cells. Despite the resemblance at low
Stromal Tumors 263
magnification, masses of histiocytes lack the
vascular framework that uniformly accompa-nies stromal cell proliferations. Immunohisto-chemical stains for histiocytes, such a lysozyme,
KP-1, or HAM 56 are useful for identifying
these cells.
Clinical Queries and Reporting
Low-grade stromal cell proliferations are diffi-cult to classify with precision, as the interface
with myometrium is not represented in biopsy
material. Consequently, when the sections
demonstrate an apparent low-grade stromal
tumor, it is necessary to clearly indicate the
differential diagnosis. In perimenopausal or
postmenopausal women, a total abdominal
hysterectomy should be performed in order to
establish the diagnosis. In young women a con-servative approach can be undertaken. If the
uterus is not enlarged, repeat curettage and
simultaneous hysteroscopy can be performed.
In addition, an MRI scan should be performed
to rule out the presence of intramural involve-ment. If all of these tests are negative, careful
follow-up and a repeat curettage in 3 to 6
months is appropriate. Continued long-term
follow-up for many years is necessary, as
low-grade stromal sarcomas may recur after
symptom-free intervals of up to 20 years.
In cases of high-grade stromal sarcoma
where the tumor cells are poorly differentiated,
the separation of stromal sarcoma from other
forms of sarcoma may be extremely difficult
in small samples of tissue. Establishing the
presence of high-grade sarcoma is sufficient
to guide further therapy, however, and the sub-classification of the neoplasm has little clinical
relevance.
Smooth Muscle Tumors
Leiomyomas
Smooth muscle tumors are rarely encountered
in endometrial biopsy and curettage specimens.
Those sampled usually are benign leiomyomas.
Tissue from a submucosal leiomyoma (“sub-mucous fibroid”) or a pedunculated intracavi-tary leiomyoma occasionally is obtained. These
specimens show fragments of smooth muscle,
although the tissue usually is too fragmented to
allow definitive diagnosis of a leiomyoma. In
premenopausal women, leiomyomas can affect
endometrial development. Endometrium over-lying a leiomyoma may show focal hypoplasia
or atrophy, being thin with sparse glands that
are underdeveloped and small compared to
surrounding tissue (Fig. 11.15).72
Endometrium
adjacent to leiomyomas also can become dis-torted.72
The glands may lose their perpen-dicular orientation relative to the surface
epithelium. These focal distortions of endome-trial growth can yield a pattern of irregular
maturation, with fragments of normally devel-oped proliferative or secretory endometrium
adjacent to fragments with irregular or poorly
developed glands. Such morphologic abnormal-ities of the glands generally are nonspecific,
however, and are not sufficient to establish the
diagnosis of a submucous leiomyoma. Often
the endometrial tissue from biopsy material
shows either no demonstrable effects from the
leiomyomas or nonspecific glandular and
stromal breakdown, and in these cases it is not
possible to diagnose leiomyomas from a biopsy
or curettage specimen.
Variants of Leiomyoma
There are a number of variants of benign
leiomyoma characterized either by increased
mitotic activity or by deviation from the typical
pattern of interlacing spindle cells, includ-ing cellular leiomyoma, leiomyoma with
bizarre nuclei (atypical leiomyoma), epitheli-oid leiomyoma, and mitotically active leiomy-oma.1;22;47;73–80
As previously indicated, it is very
unusual to encounter these tumors on biopsy or
curettage.
Bizarre, cellular, and epithelioid leiomyomas
have mitotic rates of four or fewer per 10
HPFs.47;73;78;79
Generally the mitotic rate is very
low, averaging less than one mitosis per 10
HPFs. Mitotically active leiomyomas have high
mitotic rates, up to 15 mitoses per 10 HPFs, but
lack cytologic atypia.76;77;81
These tumors often
are submucosal.76
In the rare event that any of
these lesions are sampled by curette, they are
fragmented and only partially removed (Fig.
11.16). Sampling is therefore inadequate and
precludes an accurate diagnosis. It is impossible
264 11. Other Tumors
to be certain that the tumor would not have a
higher mitotic rate in the portion that is not
removed or that more marked nuclear atypia is
not present elsewhere. Also, other features,
such as size of the lesion and interface of the
tumor with normal myometrium, cannot be
assessed. The differential diagnosis of stromal
tumor and cellular leiomyoma is discussed in
the section on stromal tumors.
Tumorlets
Small, circumscribed mesenchymal prolifera-tions called tumorlets occasionally may be
found in endometrial curettage specimens. One
such lesion is the plexiform tumorlet, a rare
small epithelioid leiomyoma that usually occurs
in the myometrium but that can involve the
endometrium.34;82
Often these are multifocal. In
the endometrium they typically occur near the
endometrial–myometrial interface and show
transitions to normal endometrial stroma.
These lesions form circumscribed microscopic
nodules and are composed of small, oval to
polygonal cells in an organoid arrangement,
typically forming small anastomosing trabecu-lae or cords with an intervening hyaline matrix.
In another pattern of tumorlet or small
leiomyoma that may be seen in curettage spec-imens, the cells have a more spindled appear-ance (Fig. 11.17). In these small proliferations,
the cells interlace in a less well defined plexi-form pattern, and although they are circum-scribed, the cells at the periphery blend into
surrounding endometrial stroma. Their inti-mate relationship to endometrial stroma is
another example of how stromal and smooth
muscle cells can coexist with transitional cells
that bridge the morphology of these two cell
types.
Tumorlets, unlike other leiomyoma vari-ants, are sufficiently small that they can be
Smooth Muscle Tumors 265
Figure 11.15. Endometrium overlying a leiomyoma.
A fragment of endometrium is markedly thinned
and atrophic, with a thin underlying layer of smooth
muscle. Other fragments showed a secretory phase
pattern. These findings are suggestive of a leiomy-oma but require clinical correlation. In this case,
subsequent hysterectomy showed submucosal
leiomyomas.
Figure 11.16. Portion of bizarre leiomyoma in curet-tings. A fragment of smooth muscle with moderate
nuclear pleomorphism underlies a fragment of
endometrium. Despite the nuclear atypia, the smooth
muscle lacked mitotic activity. Nonetheless, hysterec-tomy is necessary to exclude a leiomyosarcoma.
Figure 11.17. Tumorlet. Left: The lesion forms a
small, circumscribed nodule within the endometrial
stromal that is composed of an organoid, plexi-form–like arrangement of ovoid cells. Right: At
higher magnification the tumorlet shows strands of
small benign cells with intervening hyaline zones.
The lesion blends imperceptibly into surrounding
endometrial stroma. Lesions such as this are invari-ably benign and of no clinical significance.
confidently recognized in curettage specimens.
When they occur in the endometrium, they
often are completely removed by curettage.
They have no known clinical significance and
are noteworthy only for their unusual pattern.
They should be recognized to prevent misclas-sification of a more serious lesion.
Leiomyosarcoma
Leiomyosarcoma typically occurs in patients
older than 40 years of age. Usually the neo-plasm is confined to the myometrium, but it
may involve the endometrial cavity and be
sampled in curettings. Leiomyosarcomas show
features of high-grade sarcoma, with closely
spaced hyperchromatic nuclei showing pleo-morphism, a high mitotic rate, and abnormal
mitotic figures.1;22;47;83
Tumor necrosis usually is
present, too. The mitotic rate should be greater
than five per 10 HPFs and often is well over 10
per 10 HPFs; the cellularity and nuclear atypia
combined with the mitotic index are used in
making the diagnosis. The cells grow in inter-lacing fascicles with abundant eosinophilic fib-rillary cytoplasm and elongate nuclei with blunt
ends. Epithelioid variants occur in which the
cells are polygonal rather than spindle shaped.47
Immunohistochemical evaluation of
leiomyosarcoma shows diffuse cytoplasmic
reactivity for actin, muscle-specific actin, and
desmin, but because of the overlap of the
immunophenotype with stromal sarcoma,
morphologic analysis is generally the best
method of classifying this neoplasm.
Clinical Queries and Reporting
Smooth muscle tumors other than ordinary
leiomyomas are so unusual in endometrial
samples that they require a comment regarding
their appearance and possible malignant poten-tial. Unless the tumor is obviously malignant,
it probably will be necessary to describe the
lesion and wait for the hysterectomy to deter-mine the biologic potential of the tumor. Even
in hysterectomy specimens, where thorough
sampling is possible, an occasional tumor
remains borderline, and some have proposed
the term “smooth muscle tumors of uncertain
malignant potential”47;77;84
or “smooth muscle
tumor of low malignant potential”78
for equiv-ocal lesions.
High-grade leiomyosarcoma usually is
readily recognized. Often precise classification
of the sarcoma may be difficult from a biopsy
specimen, but in general, recognition of high-grade sarcoma is sufficient for clinical manage-ment. A brief microscopic description with a
mitotic count helps to clarify the diagnosis.
Rare Neoplasms
Lymphoma and Leukemia
Lymphoma or leukemic infiltration of the
endometrium is a highly unusual finding in
biopsy specimens.85;86
When this does occur, it
commonly is found in a patient with known dis-seminated disease. Primary involvement is rare,
and lymphoma is more common than leukemia.
Most of the lymphomas that involve the corpus
are diffuse large cell type or follicular of the
small-cleaved cell, large cell, or mixed type.86;87
Other types of lymphoma are rare. Diffuse
small cell and mixed small and large cell pat-terns, follicular, undifferentiated lymphoma
of Burkitt and non-Burkitt type, primary T-cell
lymphoma, and Hodgkin disease all have been
reported, however.86–90
Leukemic infiltrates
most commonly are attributable to myeloge-nous leukemia and may present as granulocytic
sarcoma.86
Involvement of the endometrium by lym-phoma or leukemia results in extensive infiltra-tion of the stroma with atypical lymphoid cells
that may envelop a few residual glands (Fig.
11.18). The process should be distinguished
from severe endometritis which may show
scattered lymphoid follicles, transformed
lymphocytes, and immunoblasts.86;91
The mixed
inflammatory infiltrate of endometritis and the
frequent presence of a neutrophilic infiltrate in
the glands and surface epithelium can be
helpful features for separating inflammation
from a lymphoid neoplasm.
Sheets of malignant lymphoid cells in the
endometrium also may mimic a high-grade car-cinoma or a stromal tumor. Immunohistochem-Rare Neoplasms 267
ical stains for leukocyte common antigen and
other lymphoid markers as well as keratin and
vimentin can help in establishing the correct
diagnosis.
Miscellaneous Tumors
It is extremely unusual to encounter pure
mesenchymal tumors other than stromal and
smooth muscle neoplasms in endometrial
biopsy and curettage specimens. An occasional
case of a heterologous sarcoma, such as chon-drosarcoma, osteosarcoma, angiosarcoma, or
pleomorphic rhabdomyosarcoma, has been
reported in the uterine cavity.92–99
Some of these
neoplasms may represent MMMTs in which the
carcinomatous component is indistinct or the
stromal component has overgrown the carcino-matous elements. All of these neoplasms repre-sent high-grade sarcomas that show aggressive
behavior similar to MMMTs. High-grade
sarcoma that resembles malignant fibrous histi-ocytoma or fibrosarcoma is best classified as a
variant of high-grade endometrial stromal
sarcoma. Other sarcomas, such as alveolar soft
part sarcoma and malignant rhabdoid tumor,
also have been reported, with most cases occur-ring in the fourth or fifth decades.100–105
Heman-giopericytoma is a diagnosis that should be
used with great caution; most of the reported
cases of so-called hemangiopericytoma of the
uterus actually represent endometrial stromal
sarcoma.56
Uterine hemangiopericytoma is van-ishingly rare, if it exists at all.
Benign mesenchymal tumors other than
leiomyomas are extremely rare. Occasional
lipomas and hemangiomas have been encoun-tered, but they are so unusual that other, more
268 11. Other Tumors
Figure 11.18. Lymphoma. Diffuse large-cell lym-phoma infiltrates the endometrium and encompasses
a residual benign gland. The patient presented with
abnormal bleeding and had no known disease prior
to endometrial biopsy. Shortly after endometrial
sampling, however, she presented with symptoms of
extrauterine disease. The diffuse infiltrate of large,
malignant lymphoid cells contrasts with the mixed
inflammatory infiltrate, including plasma cells, seen
in chronic endometritis.
common primary processes should be excluded
before diagnosing these rare lesions in a biopsy
specimen. For example, adipose tissue in an
endometrial biopsy usually represents omental
fat indicating a uterine perforation rather than
a lipomatous tumor or focal fatty change of the
endometrium. In these instances the clinician
should be contacted immediately. Likewise,
lesions with prominent vascularity usually rep-resent a low-grade stromal tumor or vascular
ectasia at the surface of a polyp rather than
hemangioma.
Other primary tumors of the corpus that
may be encountered in endometrial biopsies or
curettings are extremely unusual.These oddities
include yolk sac tumor,106;107
teratoma,88;108
prim-itive neuroectodermal tumor,109;110
paragan-glioma,111–113
glioma,27
and Wilms tumor.114;115
Adenomatoid tumors of the myometrium only
rarely involve the endometrium but may be
diagnosed in endometrial samples.116
Other Lesions and
Tumor-Like Conditions
On occasion, an endometrial sample may
contain fragments of bone or cartilage that
represent residual fetal parts from an earlier
abortion.117
Usually the specimen consists of
scattered fragments of ossified material without
specific relation to the other fragments of
endometrial tissue present. These foci may be
accompanied by an inflammatory response.
Calcification also can occur in a variety of other
circumstances.118–124
Sometimes the calcification
is extensive, leading to ossification.121–123
A few
cases of psammoma body formation in associa-tion with exogenous hormone use have been
reported.119;124
Psammoma bodies without asso-ciated epithelium may be a sign of extrauterine
serous carcinoma, however. Rarely, gland
contents or stroma may show microscopic areas
Other Lesions and Tumor-Like Conditions 269
Figure 11.19. Tissue from uterine septum. Biopsy
of endometrium from a septum confirmed by hys-teroscopy. The tissue shows poorly developed secre-tory endometrium overlying smooth muscle. The his-tologic findings are not diagnostic by themselves and
require clinical correlation.
of dystrophic calcification that have no appar-ent significance, possibly representing dys-trophic calcification of inspissated mucus or
cellular debris.
Sampling of a uterine septum in a pre-menopausal patient may yield fragments of
poorly developed endometrium with underly-ing smooth muscle (Fig. 11.19). These changes
resemble the changes encountered adjacent to
submucosal leiomyomas, and the differential
diagnosis requires clinical correlation.
References
(1) Zaloudek CJ, Hendrickson MR. Mesenchymal
tumors of the uterus. In: Kurman RJ, ed.
Blaustein’s pathology of the female genital
tract. New York: Springer-Verlag; 2002: 5th ed.
561–615.
(2) Clement PB. Tumorlike lesions of the uterine
corpus. In: Clement PB, Young RH, eds.
Tumors and tumorlike lesions of the uterine
corpus and cervix. New York: Churchill Liv-ingstone; 1993: 137–179.
(3) Clement PB, Scully RE. Tumors with mixed
epithelial and mesenchymal elements. In:
Clement PB, Young RH, eds. Tumors and
tumorlike lesions of the uterine corpus and
cervix. New York: Churchill Livingstone; 1993:
329–370.
(4) DiSaia PJ, Creasman WT. Clinical gynecologic
oncology, 6th ed. St. Louis: Mosby-Year Book,
2002.
(5) Goff BA, Rice LW, Fleischhacker D, Muntz
HG, Falkenberry SS, Nikrui N, et al. Uterine
leiomyosarcoma and endometrial stromal
sarcoma: Lymph node metastases and sites of
recurrence. Gynecol Oncol 1993; 50:105–109.
(6) Leibsohn S, d’Ablaing G, Mishell DRJ,
Schlaerth JB. Leiomyosarcoma in a series
of hysterectomies performed for presumed
uterine leiomyomas. Am J Obstet Gynecol
1990; 162:968–976.
(7) Kahanpaa KV, Wahlstrom T, Grohn P,
Heinonen E, Nieminen U, Widholm O. Sarco-mas of the uterus: A clinicopathologic study of
119 patients. Obstet Gynecol 1986; 67:417–424.
(8) Kounelis S, Jones MW, Papadaki H, Bakker A,
Swalsky P, Finkelstein SD. Carcinosarcomas
(malignant mixed mullerian tumors) of the
female genital tract: Comparative molecular
analysis of epithelial and mesenchymal com-ponents. Hum Pathol 1998; 29:82–87.
(9) Thompson L, Chang B, Barsky SH. Mono-clonal origins of malignant mixed tumors
(carcinosarcomas). Evidence for a divergent
histogenesis. Am J Surg Pathol 1996; 20:
277–285.
(10) Wada H, Enomoto T, Fujita M, Yoshino K,
Nakashima R, Kurachi H, et al. Molecular evi-dence that most but not all carcinosarcomas
of the uterus are combination tumors. Cancer
Res 1997; 57:5379–5385.
(11) Debrito PA, Silverberg SG, Orenstein JM.
Carcinosarcoma (malignant mixed mullerian
(mesodermal) tumor) of the female genital
tract—immunohistochemical and ultrastruc-tural analysis of 28 cases. Hum Pathol 1993;
24:132–142.
(12) Silverberg SG, Major FJ, Blessing JA, Fetter B,
Askin FB, Liao S-Y, et al. Carcinosarcoma
(malignant mixed mesodermal tumor) of
the uterus. A Gynecologic Oncology Group
pathologic study of 203 cases. Int J Gynecol
Pathol 1990; 9:1–19.
(13) Bitterman P, Chun B, Kurman RJ. The signifi-cance of epithelial differentiation in mixed
mesodermal tumors of the uterus. A clinico-pathologic and immunohistochemical study.
Am J Surg Pathol 1990; 14:317–328.
(14) George E, Manivel JC, Dehner LP, Wick
MR. Malignant mixed mullerian tumors: An
immunohistochemical study of 47 cases, with
histogenetic considerations and clinical corre-lation. Hum Pathol 1991; 22:215–223.
(15) Meis JM, Lawrence WD. The immunohisto-chemical profile of malignant mixed mullerian
tumor. Overlap with endometrial adenocarci-noma. Am J Clin Pathol 1990; 94:1–7.
(16) Colombi RP. Sarcomatoid carcinomas of
the female genital tract (malignant mixed
mullerian tumors). Semin Diagn Pathol 1993;
10:169–175.
(17) Zelmanowicz A, Hildesheim A, Sherman
ME, Sturgeon SR, Kurman RJ, Barrett RJ,
et al. Evidence for a common etiology for
endometrial carcinomas and malignant mixed
mullerian tumors. Gynecol Oncol 1998; 69:
253–257.
(18) Sreenan JJ, Hart WR. Carcinosarcomas of the
female genital tract. A pathologic study of 29
metastatic tumors: Further evidence for the
dominant role of the epithelial component
and the conversion theory of histogenesis. Am
J Surg Pathol 1995; 19(6):666–674.
(19) Geisinger KR, Dabbs DJ, Marshall RB. Malig-nant mixed mullerian tumors. An ultrastruc-270 11. Other Tumors
tural and immunohistochemical analysis with
histogenetic considerations. Cancer 1987; 59:
1781–1790.
(20) Auerbach HE, Livolsi VA, Merino MJ. Malig-nant mixed mullerian tumors of the uterus.
An immunohistochemical study. Int J Gynecol
Pathol 1988; 7:123–130.
(21) Dinh TV, Slavin RE, Bhagavan BS, Hannigan
EV, Tiamson EM, Yandell RB. Mixed muller-ian tumors of the uterus. A clinicopathologic
study. Obstet Gynecol 1989; 74:388–392.
(22) Silverberg SG, Kurman RJ. Tumors of the
uterine corpus and gestational trophoblastic
disease. Atlas of tumor pathology, 3rd series,
Fascicle 3. Washington, DC: Armed Forces
Institute of Pathology, 1992.
(23) Norris HJ, Roth E, Taylor HB. Mesenchymal
tumors of the uterus. II. A clinical and patho-logic study of 31 mixed mesodermal tumors.
Obstet Gynecol 1966; 28:57–63.
(24) Norris HJ, Taylor HB. Mesenchymal tumors
of the uterus. III. A clinical and pathologic
study of 31 carcinosarcomas. Cancer 1966;
19:1459–1465.
(25) Gagne H, Tetu B, Blondeau L, Raymond PE,
Blais R. Morphologic prognostic factors
of malignant mixed mullerian tumor of the
uterus: A clinicopathologic study of 58 cases.
Mod Pathol 1989; 2:433–438.
(26) Schweizer W, Demopoulos R, Beller U, Dubin
H. Prognostic factors for malignant mixed
mullerian tumors of the uterus. Int J Gynecol
Pathol 1990; 9:129–136.
(27) Young RH, Kleinman GM, Scully RE. Glioma
of the uterus. Report of a case with comments
on histogenesis. Am J Surg Pathol 1981;
5:695–699.
(28) Gersell DJ, Duncan DA, Fulling KH. Malig-nant mixed mullerian tumor of the uterus
with neuroectodermal differentiation. Int J
Gynecol Pathol 1989; 8:169–178.
(29) Azumi N, Czernobilsky B. Immunohistochem-istry. In: Kurman RJ, ed. Blaustein’s pathology
of the female genital tract. New York:
Springer-Verlag’ 5th ed. 2002: 1251–1276.
(30) Cerilli LA, Wick MR. Immunohistology of
soft tissue and osseous neoplasms. In: Dabbs
DJ, ed. Diagnostic immunohistochemistry.
Philadelphia: Churchill Livingstone; 2002:
59–112.
(31) Clement PB, Oliva E, Young RH. Mullerian
adenosarcoma of the uterine corpus associ-ated with tamoxifen therapy: A report of six
cases and a review of tamoxifen-associated
endometrial lesions. Int J Gynecol Pathol
1996; 15:222–229.
(32) Clement PB, Scully RE. Mullerian adenosar-coma of the uterus: A clinicopatholgical analy-sis of 100 cases with a review of the literature.
Hum Pathol 1990; 21:363–381.
(33) Clement PB, Scully RE. Mullerian adenosar-coma of the uterus: A clinicopathologic
analysis of ten cases of a distinctive type of
Mullerian mixed tumor. Cancer 1974; 34:
1138–1149.
(34) Zaloudek CJ, Norris HJ. Adenofibroma and
adenosarcoma of the uterus. A clinicopatho-logic study of 35 cases. Cancer 1981; 48:
354–366.
(35) Kaku T, Silverberg SG, Major FJ, Miller A,
Fetter B, Brady MF. Adenosarcoma of the
uterus: A Gynecologic Oncology Group clini-copathologic study of 31 cases. Int J Gynecol
Pathol 1992; 11:75–88.
(36) Abell MR. Papillary adenofibroma of the
uterine cervix. Am J Obstet Gynecol 1971;
110:990–993.
(37) Vellios F, Ng AB, Reagan JW. Papillary ade-nofibroma of the uterus. A benign mesodermal
mixed tumor of mullerian origin. Am J Clin
Pathol 1973; 60:543–551.
(38) Bocklage T, Lee KR, Belinson JL. Uterine
mullerian adenosarcoma following adenomy-oma in a woman on tamoxifen therapy.
Gynecol Oncol 1992; 44:104–109.
(39) Kerner H, Lichtig C. Mullerian adenosarcoma
presenting as cervical polyps—a report of
seven cases and review of the literature.
Obstet Gynecol 1993; 81:655–659.
(40) Clement PB. Mullerian adenosarcomas of the
uterus with sarcomatous overgrowth. A clini-copathological analysis of 10 cases. Am J Surg
Pathol 1989; 13:28–38.
(41) Clement PB, Scully RE. Mullerian adenosar-comas of the uterus with sex cord–like
elements. Am J Clin Pathol 1989; 91:664–672.
(42) Tavassoli FA, Norris HJ. Mesenchymal tumors
of the uterus. VII. A clinicopathologic study of
60 endometrial stromal nodules. Histopathol-ogy 1981; 5:1–10.
(43) Chang KL, Crabtree GS, Lim-Tan SK,
Kempson RL, Hendrickson MR. Primary
uterine endometrial stromal neoplasms. A
clinicopathologic study of 117 cases. Am J
Surg Pathol 1990; 14:415–438.
(44) Hart WR, Yoonessi M. Endometrial stromato-sis of the uterus. Obstet Gynecol 1977;
49:393–403.
References 271
(45) Norris HJ, Taylor HB. Mesenchymal tumors of
the uterus. I. A clinical and pathological study
of 53 endometrial stromal tumors. Cancer
1966; 19:755–766.
(46) Yoonessi M, Hart WR. Endometrial stromal
sarcomas. Cancer 1977; 40:898–906.
(47) Clement PB. Pure mesenchymal tumors. In:
Clement PB, Young RH, eds. Tumors and
tumorlike lesions of the uterine corpus and
cervix. New York: Churchill Livingstone; 1993:
265–328.
(48) Fekete PS, Vellios F. The clinical and histologic
spectrum of endometrial stromal neoplasms:
A report of 41 cases. Int J Gynecol Pathol
1984; 3:198–212.
(49) Evans HL. Endometrial stromal sarcoma and
poorly differentiated endometrial sarcoma.
Cancer 1982; 50:2170–2182.
(50) Piver MS, Rutledge FN, Copeland L, Webster
K, Blumenson L, Suh O. Uterine endolym-phatic stromal myosis: A collaborative study.
Obstet Gynecol 1984; 64:173–178.
(51) Mansi JL, Ramachandra S, Wiltshaw E, Fisher
C. Endometrial stromal sarcomas. Gynecol
Oncol 1990; 36:113–118.
(52) Larson B, Silfversward C, Nilsson B,
Pettersson F. Endometrial stromal sarcoma of
the uterus. A clinical and histopathologic
study. The Radiumhemmet series 1936–1981.
Eur J Obstet Gynecol Reprod Biol 1990;
35:239–249.
(53) Gloor E, Schnyder P, Cikes M, et al. Endolym-phatic stromal myosis: surgical and hormonal
treatment of extensive abdominal recurrence
20 years after hysterectomy. Cancer 1982;
50:1888–1893.
(54) Thatcher SS, Woodruff JD. Uterine stromato-sis: A report of 33 cases. Obstet Gynecol 1982;
59:428–434.
(55) Oliva E, Young RH, Clement PB, Bhan AK,
Scully RE. Cellular benign mesenchymal
tumors of the uterus. A comparative morpho-logic and immunohistochemical analysis of 33
highly cellular leiomyomas and six endome-trial stromal nodules, two frequently confused
tumors. Am J Surg Pathol 1995; 19:757–768.
(56) Oliva E, Clement PB, Young RH. Endometrial
stromal tumors: An update on a group of
tumors with a protean phenotype. Adv Anat
Pathol 2000; 7:257–281.
(57) Oliva E, Young RH, Amin MB, Clement PB.
An immunohistochemical analysis of endome-trial stromal and smooth muscle tumors of
the uterus: A study of 54 cases emphasizing
the importance of using a panel because of
overlap in immunoreactivity for individual
antibodies. Am J Surg Pathol 2002; 26:
403–412.
(58) Agoff SN, Grieco VS, Garcia R, Gown AM.
Immunohistochemical distinction of endome-trial stromal sarcoma and cellular leiomyoma.
Appl Immunohistochem Mol Morphol 2001;
9:164–169.
(59) Farhood AI, Abrams J. Immunohistochemistry
of endometrial stromal sarcoma. Hum Pathol
1991; 22:224–230.
(60) Lillemoe TJ, Perrone T, Norris HJ, Dehner LP.
Myogenous phenotype of epithelial–like areas
in endometrial stromal sarcomas. Arch Pathol
Lab Med 1991; 115:215–219.
(61) McCluggage WG, Sumathi VP, Maxwell P.
CD10 is a sensitive and diagnostically useful
immunohistochemical marker of normal
endometrial stroma and of endometrial
stromal neoplasms. Histopathology 2001;
39:273–278.
(62) Chu PG, Arber DA, Weiss LM, Chang KL.
Utility of CD10 in distinguishing between
endometrial stromal sarcoma and uterine
smooth muscle tumors: An immunohisto-chemical comparison of 34 cases. Mod Pathol
2001; 14:465–471.
(63) Binder SW, Nieberg RK, Cheng L, al Jitawi S.
Histologic and immunohistochemical analysis
of nine endometrial stromal tumors: An unex-pected high frequency of keratin protein pos-itivity. Int J Gynecol Pathol 1991; 10:191–197.
(64) Dionigi A, Oliva E, Clement PB, Young RH.
Endometrial stromal nodules and endometrial
stromal tumors with limited infiltration: A
clinicopathologic study of 50 cases. Am J Surg
Pathol 2002; 26:567–581.
(65) Clement PB, Scully RE. Uterine tumors
resembling ovarian sex-cord tumors. A clini-copathologic analysis of fourteen cases. Am J
Clin Pathol 1976; 66:512–525.
(66) McCluggage WG, Shah V, Walsh MY, Toner
PG. Uterine tumour resembling ovarian sex
cord tumour—evidence for smooth muscle
differentiation. Histopathology 1993;
23:83–85.
(67) McCluggage WG, Date A, Bharucha H, Toner
PG. Endometrial stromal sarcoma with sex
cord–like areas and focal rhabdoid differenti-ation. Histopathology 1996; 29:369–374.
(68) Oliva E, Clement PB, Young RH. Epithelioid
endometrial and endometrioid stromal
tumors: A report of four cases emphasizing
their distinction from epithelioid smooth
muscle tumors and other oxyphilic uterine and
272 11. Other Tumors
extrauterine tumors. Int J Gynecol Pathol
2002; 21:48–55.
(69) Clement PB, Scully RE. Endometrial stromal
sarcomas of the uterus with extensive
endometrioid glandular differentiation: A
report of three cases that caused problems in
differential diagnosis. Int J Gynecol Pathol
1992; 11:163–173.
(70) Oliva E, Clement PB, Young RH, Scully RE.
Mixed endometrial stromal and smooth
muscle tumors of the uterus: A clinicopatho-logic study of 15 cases. Am J Surg Pathol 1998;
22:997–1005.
(71) Nucci MR, O’Connell JT, Huettner PC, Cviko
A, Sun D, Quade BJ. h-Caldesmon expression
effectively distinguishes endometrial stromal
tumors from uterine smooth muscle tumors.
Am J Surg Pathol 2001; 25:455–463.
(72) Deligdisch L, Loewenthal M. Endometrial
changes associated with myomata of the
uterus. J Clin Pathol 1970; 23:676–680.
(73) Kempson RL, Hendrickson MR. Pure mes-enchymal neoplasms of the uterine corpus:
selected problems. Semin Diagn Pathol 1988;
5:172–198.
(74) Clement PB. Pathology of the uterine corpus.
Hum Pathol 1991; 22:776–791.
(75) Evans HL, Chawla SP, Simpson C, et al.
Smooth muscle tumors of the uterus other
than ordinary leiomyoma. A study of 46 cases,
with emphasis on diagnostic criteria and
prognostic factors. Cancer 1988; 62:2239–
2247.
(76) Prayson RA, Hart WR. Mitotically active
leiomyomas of the uterus. Am J Clin Pathol
1992; 97:14–20.
(77) O’Connor DM, Norris HJ. Mitotically active
leiomyomas of the uterus. Hum Pathol 1990;
21:223–227.
(78) Hendrickson MR, Longacre TA, Kempson
RL. The uterine corpus. In: Sternberg SS,
ed. Diagnostic surgical pathology. 3rd ed.
Philadelphia: Lippincott Williams & Wilkins’
1999: 2203–2305.
(79) Kurman RJ, Norris HJ. Mesenchymal tumors
of the uterus. VI. Epithelioid smooth muscle
tumors including leiomyoblastoma and clear
cell leiomyoma. Cancer 1976; 37:1853–1865.
(80) Mazur MT, Kraus FT. Histogenesis of mor-phologic variations in tumors of the uterine
wall. Am J Surg Pathol 1980; 4:59–74.
(81) Perrone T, Dehner LP. Prognostically
favorable “mitotically active” smooth-muscle
tumors of the uterus. A clinicopathologic study
of 10 cases. Am J Surg Pathol 1988; 12:1–8.
(82) Kaminski PF, Tavassoli FA. Plexiform tumor-let: A clinical and pathologic study of 15 cases
with ultrastructural observations. Int J
Gynecol Pathol 1984; 3:124–134.
(83) Taylor HB, Norris HJ. Mesenchymal tumors of
the uterus. IV. Diagnosis and prognosis of
leiomyosarcomas. Arch Pathol 1966; 82:40–44.
(84) Kempson RL, Hendrickson MR. Pure mes-enchymal neoplasms of the uterine corpus:
selected problems. Semin Diagn Pathol 1988;
5:172–198.
(85) Harris N, Scully R. Malignant lymphoma
and granulocytic sarcoma of the uterus and
vagina. A clinicopathologic analysis of 27
cases. Cancer 1984; 53:2530–2545.
(86) Ferry JA, Young RH. Malignant lymphoma,
pseudolymphoma, and hematopoietic disor-ders of the female genital tract. Pathol Annu
1991; 26:227–263.
(87) Vang R, Medeiros LJ, Ha CS, Deavers M. Non-Hodgkin’s lymphomas involving the uterus:
A clinicopathologic analysis of 26 cases. Mod
Pathol 2000; 13:19–28.
(88) Clement PB. Miscellaneous primary neo-plasms and metastatic neoplasms. In: Clement
PB, Young RH, eds. Tumors and tumorlike
lesions of the uterine corpus and cervix. New
York: Churchill Livingstone; 1993: 371–418.
(89) Masunaga A, Abe M, Tsujii E, Suzuki Y,
Ohgida T, Toyama M, et al. Primary uterine
T-cell lymphoma. Int J Gynecol Pathol 1998;
17:376–379.
(90) van de RM, Kamel OW, Chang PP, Lee A,
Warnke RA, Salhany KE. Primary low-grade
endometrial B-cell lymphoma. Am J Surg
Pathol 1997; 21:187–194.
(91) Young RH, Harris NL, Scully RE. Lymphoma-like lesions of the lower female genital tract.
A report of 16 cases. Int J Gynecol Pathol
1985; 4:289–299.
(92) Vakiani M, Mawad J, Talerman A. Heterolo-gous sarcomas of the uterus. Int J Gynecol
Pathol 1982; 1:211–219.
(93) De Young B, Bitterman P, Lack EE. Primary
osteosarcoma of the uterus: Report of a case
with immunohistochemical study. Mod Pathol
1992; 5:212–215.
(94) Clement PB. Chondrosarcoma of the uterus:
Report of a case and review of the literature.
Hum Pathol 1978; 6:726–727.
(95) Chiarle R, Godio L, Fusi D, Soldati T, Palestro
G. Pure alveolar rhabdomyosarcoma of the
corpus uteri: Description of a case with
increased serum level of CA-125. Gynecol
Oncol 1997; 66:320–323.
References 273
(96) Holcomb K, Francis M, Ruiz J, Abulafia O,
Matthews RP, Lee YC. Pleomorphic rhab-domyosarcoma of the uterus in a post-menopausal woman with elevated serum
CA125. Gynecol Oncol 1999; 74:499–501.
(97) McCluggage WG, Lioe TF, McClelland HR,
Lamki H. Rhabdomyosarcoma of the uterus:
Report of two cases, including one of the
spindle cell variant. Int J Gynecol Cancer
2002; 12:128–132.
(98) Ordi J, Stamatakos MD, Tavassoli FA. Pure
pleomorphic rhabdomyosarcomas of the
uterus. Int J Gynecol Pathol 1997; 16:369–377.
(99) Schammel DP, Tavassoli FA. Uterine angiosar-comas: A morphologic and immunohisto-chemical study of four cases. Am J Surg Pathol
1998; 22:246–250.
(100) Guillou L, Lamoureux E, Masse S, Costa J.
Alveolar soft–part sarcoma of the uterine
corpus: Histological, immunocytochemical
and ultrastructural study of a case. Virchows
Archiv [A] Pathol Anat 1991; 418:467–471.
(101) Cho KR, Rosenchein NB, Epstein JI. Malig-nant rhabdoid tumor of the uterus. Int J
Gynecol Pathol 1989; 8:381–387.
(102) Gray GJ, Glick A, Kurtin P, Jones H. Alveolar
soft part sarcoma of the uterus. Hum Pathol
1986; 17:297–300.
(103) Nolan N, Gaffney E. Alveolar soft part
sarcoma of the uterus. Histopathology 1990;
16:97–99.
(104) Radig K, Buhtz P, Roessner A. Alveolar soft
part sarcoma of the uterine corpus. Report of
two cases and review of the literature. Pathol
Res Pract 1998; 194:59–63.
(105) Gaertner EM, Farley JH, Taylor RR, Silver
SA. Collision of uterine rhabdoid tumor and
endometrioid adenocarcinoma: A case report
and review of the literature. Int J Gynecol
Pathol 1999; 18:396–401.
(106) Joseph MG, Fellows FG, Hearn SA. Primary
endodermal sinus tumor of the endometrium.
A clinicopathologic, immunocytochemical,
and ultrastructural study. Cancer 1990; 63:
297–302.
(107) Ohta M, Sakakibara K, Mizuno K, Kano T,
Matsuzawa K, Tomoda Y, et al. Successful
treatment of primary endodermal sinus tumor
of the endometrium. Gynecol Oncol 1988; 31:
357–364.
(108) Martin E, Scholes J, Richart R, Fenoglio C.
Benign cystic teratoma of the uterus. Am J
Obstet Gynecol 1979; 135:429–431.
(109) Daya D, Lukka H, Clement PB. Primitive neu-roectodermal tumors of the uterus—A report
of four cases. Hum Pathol 1992; 23:1120–1129.
(110) Hendrickson MR, Scheithauer BW. Primitive
neuroectodermal tumor of the endometrium:
Report of two cases, one with electron micro-scopic observations. Int J Gynecol Pathol
1986; 5:249–259.
(111) Beham A, Schmid C, Fletcher CDM, Aubock
L, Pickel H. Malignant paraganglioma of the
uterus. Virchows Arch [A] 1992; 420:453–457.
(112) Tavassoli F. Melanotic paraganglioma of the
uterus. Cancer 1986; 58:942–948.
(113) Young TW, Thrasher TV. Nonchromaffin para-ganglioma of the uterus. Arch Pathol Lab Med
1982; 106:608–609.
(114) Bittencourt AL, Britto JF, Fonseca LEJ.
Wilms’ tumor of the uterus: The first report of
the literature. Cancer 1981; 47:2496–2499.
(115) Muc RS, Grayson W, Grobbelaar JJ. Adult
extrarenal Wilms tumor occurring in the
uterus. Arch Pathol Lab Med 2001; 125:
1081–1083.
(116) Carlier MT, Dardick I, Lagace AF, Sreeram V.
Adenomatoid tumor of the uterus: presenta-tion in endometrial curettings. Int J Gynecol
Pathol 1986; 5:69–74.
(117) Sherman ME, Mazur MT, Kurman RJ. Benign
diseases of the endometrium. In: Kurman RJ,
ed. Blaustein’s pathology of the female genital
tract. 5th ed. New York: Springer-Verlag;
2002:421–466.
(118) Alpert LC, Haufrect EJ, Schwartz MR.
Uterine lithiasis. Am J Surg Pathol 1990; 14:
1071–1075.
(119) Herbold DR, Magrane DM. Calcifications of
the benign endometrium. Arch Pathol Lab
Med 1986; 110:666–669.
(120) Untawale VG, Gabriel JB, Chauhan PM. Cal-cific endometritis. Am J Obstet Gynecol 1982;
144:482–483.
(121) Ceccacci L, Clancy G. Endometrial ossifica-tion: Report of an additional case. Am J
Obstet Gynecol 1981; 141:103–104.
(122) Bhatia NN, Hoshiko MG. Uterine osseous
metaplasia. Obstet Gynecol 1982; 60:256–259.
(123) Waxman M, Moussouris HF. Endometrial
ossification following an abortion. Am J
Obstet Gynecol 1978; 130:587–588.
(124) Valicenti JF, Priester SK. Psammoma bodies of
benign endometrial origin in cervicovaginal
cytology. Acta Cytol 1977; 21:550–552.
274 11. Other Tumors
patients with severe abnormal endometrial
bleeding.1–5;7–11
Complications of D&C can include hemor-rhage, infection, or perforation, although each
of these appears to occur at a rate of between
4 and 6 per 1000 procedures.1;12
Because many
patients who undergo D&C do not come to
hysterectomy, the overall sensitivity and speci-ficity of this technique have been difficult to
determine. Several studies found that D&C
missed a significant number of polyps, hyper-plasias, and carcinomas.13–15
In fact, one study
found that D&C performed immediately
before hysterectomy often sampled less than
half of the cavity, suggesting that this procedure
may fail to fully document significant endome-trial pathology.13
Curettage or biopsy before
hysterectomy for leiomyomas also has little
value, rarely identifying a significant lesion,16
although patients with leiomyoma and abnor-mal bleeding may rarely have a malignancy.17
Several investigators also have found discrep-ancies between the grade of endometrial carci-noma in curettage specimens as compared to
hysterectomy specimens.9;11;18–21
Strictly applied, the term “endometrial
biopsy” refers to a limited sampling procedure
that does not require endocervical dilation
prior to sampling. Endometrial biopsy is
relatively painless and does not require the
anesthetic used for D&C. It is usually an office-based procedure.3–5;22
These samples are taken
either with a small sharp curette, such as the
Novak or Randall curette, or with a flexible
plastic cannula that uses suction to aspirate
275
12
Methods of Endometrial Evaluation
Endometrial Sampling
Techniques
There are several methods of sampling the
endometrium. The “gold standard” is dilation
and curettage (D&C), which requires dilation
of the cervix to allow insertion of a curette
into the endometrial cavity.1–5
This technique
allows for the most thorough sampling of the
endometrium but requires anesthesia for cervi-cal dilation. The curette is drawn across the
anterior and posterior endometrial surfaces,
scraping the tissue free. D&C also readily
allows for a fractional curettage with sampling
of both the endometrial and the endocervical
mucosa. Fractional sampling is especially useful
for evaluating possible endocervical pathology,
such as extension of endometrial adenocarci-noma to the endocervix.6
D&C is most com-monly used in situations in which more
extensive sampling of the endometrium is
needed to exclude significant pathology or to
remove as much endometrium as possible in
Endometrial Sampling Techniques  . . . . . . 275
Noninvasive Methods of Endometrial
Evaluation  . . . . . . . . . . . . . . . . . . . . . . . . . 276
Hysteroscopy  . . . . . . . . . . . . . . . . . . . . . 276
Ultrasound . . . . . . . . . . . . . . . . . . . . . . . 277
Magnetic Resonance Imaging  . . . . . . . . 277
Histologic Techniques  . . . . . . . . . . . . . . . . 277
Frozen Section . . . . . . . . . . . . . . . . . . . . . . 278
Immunohistochemistry  . . . . . . . . . . . . . . . 279
the tissue. The Pipelle endometrial aspirator
(Cooper Surgical, Shelton, CT) is the most
widely used of these devices. Limited sampling
techniques are especially useful for obtaining
smaller specimens for endometrial dating in
infertility patients or for evaluating the
response of endometrial tissue to steroid
hormone therapy of various types. Hyperplasia
and neoplasia can be accurately diagnosed by
the endometrial biopsy, however,23–28
and it is
possible to perform limited fractional sampling
of endocervical as well as endometrial tissue
using some biopsy devices. Because these
various biopsy procedures can be done in the
office rather than the operating room and
because they yield sufficient specimens for
diagnosis in most cases, they are cost-effective
methods for endometrial evaluation.1;2;22
The Pipelle and related devices have
received widespread clinical usage because
they are simple to use, cost effective, and reli-able for giving adequate tissue samples in most
cases.12
The Pipelle-type device uses a hand-held piston to generate negative pressure and
aspirate tissue through a narrow cannula
inserted into the endometrial cavity. The Pipelle
does change the pattern of tissue fragmenta-tion, yielding cylinders of tissue with small
portions of endometrium mixed with fresh
blood clot. Comparisons of the Pipelle sampling
device with other, more traditional, sampling
mechanisms show no significant difference in
the overall quality of tissue taken for evalua-tion,28–39
although some studies find that the
Pipelle technique samples much less of the
endometrial surface than other biopsy de-vices.37
Limited sampling with these devices
may lead to some under diagnosis of significant
abnormalities, however.25;27;40–42
The Pipelle also
has limited sensitivity for detecting intrauterine
gestation and excluding an ectopic pregnancy.40
Other aspiration devices, such as the Vabra
aspirator (Berkeley Medevices, Berkeley, CA)
or the Tis-U-Trap (Milex Products, Chicago,
IL), use a mechanical vacuum to extract tissue
into a tissue collection apparatus.1;22;29;30;37;43
The
cannula for these devices is thin, ranging from
3 to 4 mm, so general anesthesia is not required.
This technique tends to result in extensive frag-mentation of the tissue, but the overall quality
is comparable to that of a D&C specimen for
diagnosis. Another advantage of this method
is that it samples much of the endometrium.
Endometrial brush biopsy using the Tao Brush
(Cook OB/GYN, Bloomington, IN) also has
been effective in detecting endometrial abnor-malities.44;45
This technique uses a brush to
remove tissue for both histology and cytology,
and requires a special fixative and centrifuga-tion to prepare the material.
An aspiration technique called suction curet-tage is used in evacuating early (first trimester)
abortion specimens. The procedure requires
cervical dilation and is often done under local
anesthesia (paracervical block), as general
anesthesia increases the risk of perforation, vis-ceral injury, and hemorrhage during extraction
of the aborted gestation.46;47
In very early preg-nancy, however, endometrial aspiration, which
is often termed “menstrual extraction,” can be
performed using a small plastic cannula without
anesthesia or dilation. After the first trimester,
but generally before the 20th week, abortion
can be performed by dilation and evacuation
(D&E), a technique that employs gradual
cervical dilation using an osmotic dilator
(Laminaria japonica).46;47
Noninvasive Methods of
Endometrial Evaluation
Hysteroscopy
Hysteroscopy with fiberoptic illumination is
widely used for visualizing the endometrium
and allowing directed biopsy or excision of
lesions.3;48–50
Hysteroscopy, especially with a
large-diameter hysteroscope, may require local
or general anesthesia, and in some patients cer-vical dilation is necessary. With small-diameter
scopes, this can be an office-based procedure,
however. The technique is usually performed
by distending the endometrial cavity to allow
visualization, a procedure termed “panoramic
hysteroscopy.”3;51
For the larger scopes the
distending medium often is dextran, although
other substances, such as 5% dextrose and
water and carbon dioxide gas, may be used. A
nondistending technique known as “contact
hysteroscopy” does not require a distending
medium. In this technique the surface to be
276 12. Methods of Endometrial Evaluation
viewed is touched, and lesions are identified by
their contour, color, vascular pattern, and
spatial relationships.3;52
Hysteroscopy has the advantage of giving
directed biopsy specimens, in contrast to the
blind biopsy offered by other procedures. It is
useful for evaluation of women with abnormal
uterine bleeding. It can reveal polyps or small
submucosal leiomyomas and enhances clini-copathologic correlations. The technique is
useful before and after D&C to make certain
that lesions such as polyps or adhesions are
removed by the curettage. In fact, hysteroscopy
with endometrial resection may provide supe-rior detection of focal endometrial lesions
compared to D&C alone.14
In addition, hys-teroscopy can help in evaluation of women with
repetitive abortions who may have a congenital
abnormality, such as a septum. This procedure
also can be used to determine the extent and
possible cervical extension of endometrial
carcinoma. The technique of hystero-scopically directed transcervical resection of
the endometrium can be used as a therapy for
dysfunctional uterine bleeding, obliterating the
endometrium.49;53
Ultrasound
Transvaginal ultrasound is another adjunctive
technique for examining the endometrium.54–70
Sonography with a transvaginal probe evalu-ates the thickness and morphology of the
endometrium. The technique permits measure-ment of the thickness of the combined anterior
and posterior endometrium, which is referred
to as the endometrial “stripe.” This parameter
can assist in determining pathologic and physi-ologic changes in the endometrium.55;61
In post-menopausal patients a thin endometrial stripe
of less than 4 or 5 mm indicates that a signifi-cant pathologic lesion of the endometrium is
unlikely,58;71
whereas a stripe thicker than 5 mm
suggests the presence of polyps, hyperplasia,
or carcinoma. In addition, this procedure can
help to determine the presence or absence of
myometrial invasion by endometrial carci-noma. Ultrasonography also is useful for deter-mining the degree of development of the
endometrium in the secretory phase by deter-mining its thickness and texture.54–57
This tech-nique cannot replace biopsy for accurate eval-uation of endometrial morphology, however.60
Both transvaginal and transabdominal ultra-sound are useful for assessing the possible pres-ence of an ectopic pregnancy. When ectopic
pregnancy is suspected, sonography can deter-mine whether a gestation is in utero or tubal.72
Both methods of ultrasound also are used in the
diagnosis of gestational trophoblastic disease,
especially hydatidiform mole.73
Magnetic Resonance Imaging
Magnetic resonance imaging (MRI) provides a
clear view of the uterine anatomy that is espe-cially useful in the evaluation of tumors.3;74–79
MRI demonstrates the endometrial–myome-trial interface or “junctional zone,” so it can be
used to assess myometrial invasion by endome-trial carcinoma.80
It also can demonstrate
myometrial invasion in gestational trophoblas-tic disease.81
On occasion this technique is
useful for careful follow-up or assessment of
other forms of uterine neoplasia, such as
stromal tumors or leiomyomas. MRI is time
consuming and expensive, however, and it is not
practical for routine evaluation of non-neoplastic conditions.
Histologic Techniques
Gross examination of endometrial tissue is
generally not reliable for selecting material for
microscopy. Consequently, in most cases, the
whole tissue specimen should be submitted. For
abortion specimens containing abundant tissue,
three cassettes are sufficient to verify the pres-ence of placental tissue (chorionic villi or tro-phoblast). If gross examination shows placental
tissue, however, one cassette will be sufficient
if the study is intended only to document
the presence of an intrauterine gestation. An
exception is examination of hydatidiform mole.
A minimum of four tissue blocks should be sub-mitted to ensure adequate assessment of the
chorionic villi, including the degree of tro-phoblastic hyperplasia and atypia.
Proper technique is requisite to ensure ade-quate histologic evaluation. Biopsy tissue that
suffers from suboptimal fixation, processing, or
Histologic Techniques 277
sectioning will have artifacts that hinder micro-scopic evaluation. Fixation of endometrial
tissue often is difficult because of the large
amount of blood that is admixed with the tissue
fragments. Some pathologists advocate special
fixatives such as Bouin’s for endometrial biop-sies, since they offer excellent cytologic detail,82
but formalin is the most widely available and
accepted fixative and, in our opinion, is the fix-ative of choice. Acid-containing fixatives such
as Bouin’s degrade DNA, limiting any type of
molecular analysis of the tissues.
Before processing, tissue fragments should
be separated from as much blood as possible.
Well-fixed tissue can be placed in a tea strainer
and briefly rinsed with water to remove some
blood before placing into a cassette. Wrapping
the tissue in thin, porous paper (tissue wrap or
lens paper) or placing tissue in a porous “biopsy
bag” or tea bag in the cassette prevents loss of
small tissue fragments. Experience has shown
that sponges used to hold small specimens
in cassettes cause artifacts and distort the
three-dimensional configuration of the tissue.83
During processing, immersion in alcohol–for-malin removes some of the blood, which aids in
subsequent sectioning. Modern tissue proces-sors using vacuum provide optimal dehydration
and penetration of paraffin into tissue. In our
experience, ethanol is a better dehydrating
agent than denatured alcohol. To achieve
optimum processing, it is necessary to change
reagents in the processor daily.
Specimens from endometrial biopsies and
curettings are among the more difficult tissues
to section, because they are highly fragmented
and bloody. The paraffin-embedded tissue
tends to be dry, resulting in shatter and a
“venetian blind” effect. Warming the block in
warm water and then applying ice to the surface
of the block facilitates even sectioning, with
decreased fragmentation and shatter. Speci-mens should be cut at 4 to 6 mm.
The paraffin blocks should be cut at multiple
levels (two or three) in most cases. Multiple
levels, or step sections, are especially important
for smaller samples embedded in one or two
cassettes. Step sections provide the most com-prehensive study of the tissue, allowing the
pathologist to assess the three-dimensional
aspects of the tissue, and are especially useful
for endometrial samples, because the tissue
tends to be highly fragmented and haphazardly
oriented. Furthermore, levels on the block can
uncover occasional subtle abnormalities that
would not be noticed if only a single section
was reviewed. For example, levels may clarify
the presence of a polyp or they may reveal
that an apparent polypoid structure simply
represents tangential sectioning of normal
endometrium. Levels also help to determine
whether apparently disordered glands repre-sent a true abnormality or are simply an
artifact of the procedure. Even endometrial
biopsies for histologic dating in infertility
patients benefit from multiple levels; fre-quently the histologic date is correctly adjusted
by identifying more advanced secretory
changes in step sections.
Routine hematoxylin and eosin (H&E)
stains generally suffice for the diagnosis of most
specimens. Other histochemical stains are
rarely necessary. The use of the periodic
acid–Schiff (PAS) stain to demonstrate glyco-gen in the early secretory phase has no advan-tage over careful examination of routine H&E
sections for subnuclear vacuoles. Biopsies
showing granulomatous inflammation should
be stained for acid-fast and fungal organisms.
Tissue Gram stains are not useful for evalua-tion of most cases of endometritis. Stains for
epithelial mucin, such as mucicarmine and
alcian blue, are useful for establishing the
diagnosis of adenocarcinoma in a poorly differ-entiated malignant tumor. Mucin stains have
little utility for determining endometrial versus
endocervical primary sites, however, because
tumors at either site show variable amounts of
cytoplasmic and luminal mucin (see Chapter
10).
Frozen Section
Frozen sections can be useful in the evaluation
of occasional cases. Usually, however, frozen
sections cause significant artifacts in endome-trial tissue, as the tissue often is edematous and
contains considerable blood. These tissues have
very different consistency and water content
278 12. Methods of Endometrial Evaluation
compared to other specimens, such as lymph
nodes or breast tissue. Consequently, laborato-ries that routinely use frozen section for the
latter tend to have greater difficulty obtaining
sections from endometrial samples.
On occasion a frozen section is requested just
prior to hysterectomy in a perimenopausal or
postmenopausal woman with abnormal uterine
bleeding to determine whether carcinoma is
present. This technique is helpful if the tissue is
clearly benign or clearly malignant. The
subtleties of glandular patterns, which are
crucial in distinguishing atypical hyperplasia
from well-differentiated adenocarcinoma, can
be substantially obscured by artifacts caused by
the frozen section technique, however. A better
method of assessing the endometrium preoper-atively is to obtain an office-based biopsy.
Formalin-fixed specimens can be rapidly
processed and reported, offering greater diag-nostic accuracy.
One other occasional application of frozen
section is in the evaluation of the patient with
a possible ectopic pregnancy. Frozen section
can help establish the presence or absence of
intrauterine trophoblastic tissue in selected
cases. Usually, however, measurement of serum
progesterone, serum  b -human chorionic
gonadotropin (b-hCG) measurements, and
transvaginal ultrasound can be used to assess
the possible presence of an ectopic gestation
before resorting to curettage.84
When curettings
are obtained, an attempt should be made to
visualize villi by floatation of the specimen in
saline before resorting to frozen section.
Immunohistochemistry
Accurate interpretation of most endometrial
biopsies depends primarily on evaluation of
well-fixed and carefully prepared H&E sec-tions. In an occasional case, however, optimal
assessment of an abnormality is aided by
immunohistochemical stains.85–87
Immunohisto-chemistry generally is most helpful either to
assess trophoblastic tissue or to evaluate a neo-plasm. Despite the large number of antibodies
available, only a few are useful adjuncts for
the diagnosis of most endometrial lesions. The
applications of immunohistochemistry for spe-cific diagnoses also are discussed in greater
detail in the relevant chapters. The following is
a brief summary of instances in which immuno-histochemistry can assist in the diagnosis.
For trophoblastic tissue, one of the most
useful immunostains is keratin. Because tro-phoblastic cells are epithelial, any type of
trophoblast (cytotrophoblast, intermediate
trophoblast, or syncytiotrophoblast) is immu-noreactive to keratin unless fixation and preser-vation have masked the presence of the
filaments. Consequently, a keratin stain can be
very useful for demonstrating trophoblastic
cells, especially intermediate trophoblast, in
specimens in which chorionic villi and tro-phoblastic cells are not clearly evident. An
example would be identification of trophoblast
in assessing the possible presence of an ectopic
pregnancy.88–92
In these cases the infiltrate of
intermediate trophoblast at the placental
implantation site can be very difficult to distin-guish from decidua (see Chapter 3). In addition
to keratin, the placental hormones human
chorionic gonadotropin-b (hCG-b ), human pla-cental lactogen (hPL), and Mel-CAM (CD 146)
are produced by syncytiotrophoblast and inter-mediate trophoblast. Immunostains for these
proteins, especially hPL and Mel-CAM, which
are present in intermediate trophoblast at the
placental implantation site, can be helpful in
ruling out an ectopic pregnancy.88;89;91–96
Demonstration of hCG, hPL, Mel-CAM, and
inhibin-a also is useful in establishing the diag-nosis of choriocarcinoma, and placental site
trophoblastic tumor (PSTT), and epithelioid
trophoblastic tumor when routine H&E sec-tions fail to clearly demonstrate the diagnostic
histologic features of these neoplasms (see
Chapter 4).93;95;97–99
Inhibin-a also is a helpful
immunohistochemical marker for the epi-thelioid trophoblastic tumor.99;100
Differential
staining of trophoblastic and proliferation
markers also helps distinguish between the dif-ferent types of trophoblastic tumors. Chorio-carcinoma is more strongly reactive with hCG
while the PSTT generally shows more staining
for hPL. The Ki-67 proliferation index is also
much higher in choriocarcinoma (> 50%) as
compared to PSTT (15% to 20%).101
The
Immunohistochemistry 279
epithelioid trophoblastic tumor stains best with
inhibin-a and p63 but shows only limited stain-ing for hCG, hPL, and Mel-CAM.95;100;100a
The
Ki-67 proliferation index of this tumor is
approximately 20%.95
The proliferation marker
Ki-67 also can be useful in distinguishing an
exaggerated placental site from the placental
site trophoblastic tumor (PSTT). The exagger-ated placental site has no mitotic activity and a
Ki-67 index near zero while in PSTT the pro-liferation index is 14% ± 6.9%.101
Immunohistochemistry can also be helpful in
diagnosis of the placental site nodule. The inter-mediate cells of this lesion are reactive for
keratin and EMA as well as placental alkaline
phosphatase (PLAP), inhibin-a, and p63. The
Ki-67 proliferation index is less than 10%. The
placental site nodule is only focally reactive for
hPL and Mel-CAM while the PSTT stains more
diffusely for these antigens.99
Also, the PSTT is
not reactive for PLAP.
In early pregnancy the endometrial glands
are immunoreactive for S-100 protein, and
this staining disappears after the 12th week of
gestation.102;103
Normal proliferative and
secretory endometrium and hyperplastic and
neoplastic glands do not stain for S-100 pro-tein. No antibodies assist in distinguishing
atypical hyperplasia from well-differentiated
adenocarcinoma.
Endometrial intraepithelial carcinoma (EIC)
(see Chapter 9) and serous and clear cell carci-nomas of the endometrium (see Chapter 10)
show diffuse and strong reactivity for Ki-67,
which demonstrates the high proliferative
index of these lesions.86;104–106
EIC and serous
and clear cell carcinoma also are often strongly
immunoreactive for p53, while endometrioid-type endometrial carcinomas are generally
not.86;105
Estrogen receptors also are usually
absent in serous and clear cell carcinoma,
in contrast to their presence in low-grade
endometrioid carcinomas.86;105;107
In the evaluation of neoplasia, immunohisto-chemical stains may assist in the differential
diagnosis of endometrial and endocervical
primary adenocarcinoma. Endometrial carci-noma generally is immunoreactive for estrogen
and progesterone receptor protein whereas
endocervical carcinoma is not.108–111
In addition,
detection of human papilloma virus (HPV) by
in situ hybridization is seen in endocervical but
not endometrial carcinomas.108
Also, virtually
all of the usual types of endocervical carcino-mas react with p16 whereas most of the typical
endometrioid carcinomas arising in the uterine
corpus do not.112;113
Other immunostains also
may be helpful but less specific in this distinc-tion of primary site. For instance, vimentin
frequently stains endometrial carcinomas
while cervical adenocarcinomas are nega-tive.114;85;115;109;110;116
Conversely, carcinoembry-onic antigen (CEA) often is present in
endocervical carcinomas but is less common
in endometrial primary tumors110;115–118
(see
Chapter 10). However, neither vimentin nor
CEA is a completely specific marker for
primary site.
Keratin immunostains also can help to deter-mine whether a solid proliferation of cells rep-resents an epithelial tumor or a lesion of
mesenchymal or lymphoid cells. If the lesion
represents a malignant mixed mesodermal
tumor (MMMT) (carcinosarcoma), keratin
staining also is useful for highlighting the
biphasic nature of the tumor, with the keratin-positive epithelial component standing out
against the background of mostly nonreac-tive sarcomatous cells (see Chapter 11).119–124
Epithelial membrane antigen (EMA) also
stains epithelial components. The sarcomatous
spindle cell component may stain focally with
keratin and EMA, but this staining is limited
and less intense than the reactivity of the
clearly carcinomatous component.119–125
In assessing a possible MMMT, other
markers can be useful for establishing the pres-ence of sarcomatous elements, although the
subtype of sarcoma has no influence on the
prognosis of the lesion. The sarcomatous
component typically is reactive for vimentin
and actin, and if the sarcomatous component
includes leiomyosarcoma or rhabdomyosar-coma, muscle-specific actin and desmin reactiv-ity also is found.119;120;123–125
Myoglobin and
myogenin are more specific stains for rhab-domyoblasts.122;123
Occasionally other stains are
useful. A tumor with glial differentiation will
stain for S-100 protein or glial fibrillary acidic
protein. Cartilaginous tissue is immunoreactive
280 12. Methods of Endometrial Evaluation
for S-100 protein. Although immunohisto-chemical stains are useful adjuncts for tumor
diagnosis, correlation of histologic features with
immunoreactivity is essential for proper classi-fication of cell types.
Endometrial stromal tumors are immunore-active for CD10, actin, and, rarely, desmin.126–128
Smooth muscle tumors are more diffusely reac-tive for desmin and also are positive for h-caldesmon.129;130
Both stromal and smooth
muscle tumors also may show immunoreactiv-ity for keratin, although the staining usually
is focal, so positive staining for keratin needs
to be carefully assessed, especially for the
number and intensity of positive cells. In
contrast to keratin, EMA is much more specific
as an epithelial marker in our experience. It is
not present in stromal or smooth muscle
tumors.
Other applications of immunohistochemistry
are relatively infrequent. On occasion the
endometrium will contain metastatic tumor
from an unknown primary site, and immuno-histochemical evaluation can help determine
the type of tumor present.131
An immunostain
for S-100 protein can help identify metastatic
melanoma. Metastatic carcinoma from the gas-trointestinal tract typically is immunoreactive
for cytokeratin 20 and CEA, whereas primary
endometrial cancer usually is not. Metastatic
breast carcinoma often shows immunostaining
for gross cystic disease fluid protein-15.131
Lym-phoma and leukemia can be characterized
using a number of lymphoid markers.132
Anti-bodies for herpesvirus and cytomegalovirus
can help establish the presence of these viral
infections.
References
(1) Grimes DA. Diagnostic dilation and curet-tage. A reappraisal. Am J Obstet Gynecol
1982; 142:1–6.
(2) Smith J, Schulman H. Current dilation and
curettage. A need for revision. Am J Obstet
Gynecol 1985; 65:516–518.
(3) Droegemueller W, Katz V. Diagnostic proce-dures. In: Stenchever MA, Droegemueller W,
Herbst AL, Mishell DR, Jr, eds. Comprehen-sive gynecology. 4th ed. St. Louis: Mosby;
2001: 219–249.
(4) Butler WJ. Normal and abnormal uterine
bleeding. In: Rock JA, Jones HW, III, eds.
Te Linde’s operative gynecology. 9th ed.
Philadelphia: Lippincott Williams & Wilkins;
2003: 457–481.
(5) Friedman F, Brodman ML. Endometrial sam-pling techniques. In: Altchek A, Deligdish L,
eds. The uterus: Pathology, diagnosis and
management. New York: Springer-Verlag;
1991: 155–162.
(6) Caron C, Tetu B, Laberge P, Bellemare G,
Raymond P–E. Endocervical involvement by
endometrial carcinoma on fractional curet-tage: A clinicopathological study of 37 cases.
Mod Pathol 1991; 4:644–647.
(7) Nickelsen C. Diagnostic and curative value
of uterine curettage. Acta Obstet Gynecol
Scand 1980; 65:693–697.
(8) Schlaerth JB, Morrow CP, Rodriguez M.
Diagnostic and therapeutic curettage in ges-tational trophoblastic disease. Am J Obstet
Gynecol 1990; 162:1465–1471.
(9) Daniel AG, Peters WA, III. Accuracy of office
and operating room curettage in the grading
of endometrial carcinoma. Obstet Gynecol
1988; 71:612–614.
(10) Berkowitz RS, Desai U, Goldstein DP,
Driscoll SG, Marean AR, Berstein MR.
Pretreatment curettage—a predictor of
chemotherapy response in gestational tro-phoblastic disease. Gynecol Oncol 1980;10:
39–43.
(11) Cowles TA, Magrina JF, Masterson BJ,
Capen CV. Comparison of clinical and surgi-cal staging in patients with endometrial car-cinoma. Obstet Gynecol 1985; 66:413–416.
(12) Spencer CP, Whitehead MI. Endometrial
assessment re-visited. Br J Obstet Gynaecol
1999; 106:623–632.
(13) Stock RJ, Kanbour L. A pre-hysterectomy
curettage. Obstet Gynecol 1975; 45:537–560.
(14) Epstein E, Ramirez A, Skoog L, Valentin L.
Dilatation and curettage fails to detect most
focal lesions in the uterine cavity in women
with postmenopausal bleeding. Acta Obstet
Gynecol Scand 2001; 80:1131–1136.
(15) Bettocchi S, Ceci O, Vicino M, Marello F,
Impedovo L, Selvaggi L. Diagnostic inade-quacy of dilatation and curettage. Fertil Steril
2001; 75:803–805.
(16) Moller LMA, Berget A. Prehysterectomy
curettage in women with uterine fibromy-omata is not worthwhile. Acta Obstet
Gynecol Scand 1993; 72:374–376.
References 281
(17) Duplantier N, Finan MA, Barbe T. Necessity
of endometrial biopsy in women with
enlarged uteri and a preoperative diagnosis
of uterine leiomyomata. J Reprod Med 2003;
48:23–27.
(18) Soothill PW, Alcock CJ, MacKenzie IZ. Dis-crepancy between curettage and hysterec-tomy histology in patients with stage I
uterine malignancy. Br J Obstet Gynaecol
1989; 96:478–481.
(19) Piver MS, Lele SB, Barlow JJ, Blumenson L.
Paraaortic lymph node evaluation in stage I
endometrial cancer. Obstet Gynecol 1982;
59:97–100.
(20) Sant Cassia LJ, Weppelmann B, Shingleton
H, Soong SJ, Hatch K, Salter MM. Man-agement of early endometrial carcinoma.
Gynecol Oncol 1989; 35:362–366.
(21) Obermair A, Geramou M, Gucer F, Denison
U, Graf AH, Kapshammer E, et al. Endome-trial cancer: accuracy of the finding of a well
differentiated tumor at dilatation and curet-tage compared to the findings at subsequent
hysterectomy. Int J Gynecol Cancer 1999;
9:383–386.
(22) Feldman S, Berkowitz RS, Tosteson ANA.
Cost-effectiveness of strategies to evaluate
postmenopausal bleeding. Obstet Gynecol
1993; 81:968–975.
(23) Larson DM, Johnson KK, Broste SK,
Krawisz BR, Kresl JJ. Comparison of D&C
and office endometrial biopsy in predicting
final histopathologic grade in endometrial
cancer. Obstet Gynecol 1995; 86:38–42.
(24) Agostini A, Shojai R, Cravello L, Rojat-Habib MC, Roger V, Bretelle F, et al.
Endometrial biopsy during outpatient hys-teroscopy: Evaluation and comparison of two
devices. Eur J Obstet Gynecol Reprod Biol
2001; 97:220–222.
(25) Dijkhuizen FP, Mol BW, Brolmann HA,
Heintz AP. The accuracy of endometrial
sampling in the diagnosis of patients with
endometrial carcinoma and hyperplasia: a
meta-analysis. Cancer 2000; 89:1765–1772.
(26) Ferry J, Farnsworth A, Webster M, Wren B.
The efficacy of the pipelle endometrial
biopsy in detecting endometrial carcinoma.
Aust N Z J Obstet Gynaecol 1993; 33:76–78.
(27) Guido RS, Kanbour-Shakir A, Rulin MC,
Christopherson WA. Pipelle endometrial
sampling. Sensitivity in the detection of
endometrial cancer. J Reprod Med 1995;
40:553–555.
(28) Stovall TG, Photopulos GJ, Poston WM,
Ling FW, Sandles LG. Pipelle endometrial
sampling in patients with known endometrial
carcinoma. Obstet Gynecol 1991; 77:954–956.
(29) Koonings PP, Moyer DL, Grimes DA. A ran-domized clinical trial comparing Pipelle and
Tis-u-trap for endometrial biopsy. Obstet
Gynecol 1990; 75:293–295.
(30) Kaunitz AM, Masciello A, Ostrowski M,
Rovira EZ. Comparison of endometrial
biopsy with the endometrial Pipelle and
Vabra aspirator. J Reprod Med 1988; 33:
427–431.
(31) Hill GA, Herbert CMI, Parker RA, Wentz
AC. Comparison of late luteal phase end-ometrial biopsies using the Novak curette or
Pipelle endometrial suction curette. Obstet
Gynecol 1989; 73:443–446.
(32) Eddowes HA, Read MD, Codling BW.
Pipelle—a more acceptable technique for
outpatient endometrial biopsy. Br J Obstet
Gynaecol 1990; 97:961–962.
(33) Silver MM, Miles P, Rosa C. Comparison of
Novak and Pipelle endometrial biopsy instru-ments. Obstet Gynecol 1991; 78:828–830.
(34) Stovall TG, Ling FW, Morgan PL. A prospec-tive, randomized comparison of the Pipelle
endometrial sampling device with the Novak
curette. Am J Obstet Gynecol 1991;
165:1287–1290.
(35) Fothergill DJ, Brown VA, Hill AS. Histologi-cal sampling of the endometrium—a com-parison between formal curettage and the
Pipelle sampler. Br J Obstet Gynaecol 1992;
99:779–780.
(36) Ferry J, Farnsworth A, Webster M, Wren B.
The efficacy of the Pipelle endometrial
biopsy in detecting endometrial carcinoma.
Aust N Z J Obstet Gynaecol 1993; 33:76–78.
(37) Rodriguez GC, Yaqub N, King ME. A
comparison of the Pipelle device and the
Vabra aspirator as measured by endometrial
denudation in hysterectormy specimens:
The Pipelle device samples significantly less
of the endometrial surface than the Vabra
aspirator. Am J Obstet Gynecol 1993; 168:
55–59.
(38) Check JH, Chase TS, Nowroozi K, Wu CH,
Chern R. Clinical evaluation of the Pipelle
endometrial suction curette for timed
endometrial biopsy. J Reprod Med 1989; 34:
218–220.
(39) Henig I, Chan P, Treadway PR, Maw BM,
Gullett AJ, Chatwood M. Evaluation of the
282 12. Methods of Endometrial Evaluation
Pipelle curette for endometrial biopsy.
J Reprod Med 1989; 34:786–789.
(40) Barnhart KT, Gracia CR, Reindl B, Wheeler
JE. Usefulness of pipelle endometrial biopsy
in the diagnosis of women at risk for ectopic
pregnancy. Am J Obstet Gynecol 2003;
188:906–909.
(41) Ong S, Duffy T, Lenehan P, Murphy J.
Endometrial pipelle biopsy compared to
conventional dilatation and curettage. Ir J
Med Sci 1997; 166:47–49.
(42) Ries A, Singson P, Bidus M, Barnes JG. Use
of the endometrial pipelle in the diagnosis of
early abnormal gestations. Fertil Steril 2000;
74:593–595.
(43) Einerth Y. Vacuum curettage by the Vabra
method. A simple procedure for endometrial
diagnosis. Acta Obstet Gynecol Scand 1982;
61:373–376.
(44) Wu HH, Casto BD, Elsheikh TM. Endome-trial brush biopsy. An accurate outpatient
method of detecting endometrial malignancy.
J Reprod Med 2003; 48:41–45.
(45) Del Priore G, Williams R, Harbatkin CB,
Wan LS, Mittal K, Yang GC. Endometrial
brush biopsy for the diagnosis of endometrial
cancer. J Reprod Med 2001; 46:439–443.
(46) Stubblefield PM. Conception control: contra-ception, sterilization, and pregnancy termi-nation. In: Kistner RW, ed. Gynecology:
Principles and practice. 4th ed. Chicago: Year
Book Medical Publishers; 1986: 583–621.
(47) Grimes DA. Management of abortion. In:
Rock JA, Jones HW, III, eds. Te Linde’s oper-ative gynecology. 9th ed. Philadelphia: Lip-pincott Williams & Wilkins; 2003: 483–505.
(48) Friedler S, Margalioth EJ, Kafka I, Yaffe H.
Incidence of post-abortion intra-uterine
adhesions evaluated by hysteroscopy. A
prospective study. Hum Reprod 1993;
8:442–444.
(49) Hellen EA, Coghill SB, Shaxted EJ. The
histopathology of transcervical resection of
the endometrium—an analysis of 200 cases.
Histopathology 1993; 22:361–365.
(50) Baggish MS. Operative hysteroscopy. In:
Rock JA, Jones HW, III, eds. Te Linde’s oper-ative gynecology. 9th ed. Philadelphia: Lip-pincott Williams & Wilkins; 2003: 379–411.
(51) Valle RF. Office hysteroscopy (panoramic
hysteroscopy). In: Baggish MS, Barbot J,
Valle RF, eds. Diagnostic and operative hys-teroscopy: A text and atlas. St. Louis: Mosby;
1999: 171–183.
(52) Vallee RF, Baggish MS. Instrumentation
for hysteroscopy. In: Baggish MS, Barbot J,
Valle RF, eds. Diagnostic and operative hys-teroscopy: A text and atlas. 2nd ed. St. Louis:
Mosby; 1989: 97–126.
(53) Daniell JF, Kurtz BR, Ke RW. Hysteroscopic
endometrial ablation using the rollerball
electrode. Obstet Gynecol 1992; 80:329–332.
(54) Rogers PAW, Polson D, Murphy CR, Hosie
M, Susil B, Leoni M. Correlation of endome-trial histology, morphometry, and ultrasound
appearance after different stimulation proto-cols for in vitro fertilization. Fertil Steril
1991; 55:583–587.
(55) Fleischer AC, Gordon AN, Entman SS,
Kepple DM. Transvaginal scanning of the
endometrium. J Clin Ultrasound 1990; 18:
337–349.
(56) Khalifa E, Brzyski RG, Oehninger S, Acosta
AA, Muasher SJ. Sonographic appearance
of the endometrium—The predictive value
for the outcome of in vitro fertilization in
stimulated cycles. Hum Reprod 1992; 7:677–
680.
(57) Dickey RP, Olar TT, Curole DN, Taylor SN,
Rye PH. Endometrial pattern and thickness
associated with pregnancy outcome after
assisted reproduction technologies. Hum
Reprod 1992; 7:418–421.
(58) Sheth S, Hamper UM, Kurman RJ. Thickened
endometrium in the postmenopausal
woman—sonographic—pathologic correla-tion. Radiology 1993; 187:135–139.
(59) Dorum A, Kristensen GB, Langebrekke A,
Sornes T, Skaar O. Evaluation of endometrial
thickness measured by endovaginal ultra-sound in women with postmenopausal bleed-ing. Acta Obstet Gynecol Scand 1993; 72:
116–119.
(60) Doherty CM, Silver B, Binor Z, Molo MW,
Radwanska E. Transvaginal ultrasound and
the assessment of luteal phase endometrium.
Am J Obstet Gynecol 1993; 168:1702–1709.
(61) Fleischer AC. Transvaginal sonography of
endometrial disorders. In: Fleischer AC,
Manning FA, Jeanty P, Romero R, eds.
Sonography in obstetrics and gynecology.
Principles and practice. 6th ed. New York:
McGraw-Hill; 2001: 979–1000.
(62) Iha T, Shen H, Kanazawa K. Hysteroscopy
to detect stage IA well-differentiated
endometrioid adenocarcinoma of the endo-metrium. Acta Obstet Gynecol Scand 2003;
82:378–384.
References 283
(63) Giorda G, Crivellari D, Veronesi A, Perin T,
Campagnutta E, Carbone A, et al. Compari-son of ultrasonography, hysteroscopy, and
biopsy in the diagnosis of endometrial lesions
in postmenopausal tamoxifen–treated
patients. Acta Obstet Gynecol Scand 2002;81:
975–980.
(64) Panda JK. One-stop clinic for post-menopausal bleeding. J Reprod Med 2002;
47:761–766.
(65) Gull B, Karlsson B, Milsom I, Granberg S.
Can ultrasound replace dilation and curet-tage? A longitudinal evaluation of post-menopausal bleeding and transvaginal
sonographic measurement of the endo-metrium as predictors of endometrial
cancer. Am J Obstet Gynecol 2003; 188:401–
408.
(66) Davidson KG, Dubinsky TJ. Ultrasono-graphic evaluation of the endometrium in
postmenopausal vaginal bleeding. Radiol
Clin North Am 2003; 41:769–780.
(67) Giusa-Chiferi MG, Goncalves WJ, Baracat
EC, Albuquerque Neto LC, Bortoletto CC,
de Lima GR. Transvaginal ultrasound,
uterine biopsy and hysteroscopy for post-menopausal bleeding. Int J Gynaecol Obstet
1996; 55:39–44.
(68) Langer RD, Pierce JJ, O’Hanlan KA,
Johnson SR, Espeland MA, Trabal JF, et al.
Transvaginal ultrasonography compared
with endometrial biopsy for the detection of
endometrial disease. Postmenopausal Estro-gen/Progestin Interventions Trial. N Engl J
Med 1997; 337:1792–1798.
(69) Smith–Bindman R, Kerlikowske K, Feldstein
VA, Subak L, Scheidler J, Segal M, et al.
Endovaginal ultrasound to exclude endome-trial cancer and other endometrial abnor-malities. JAMA 1998; 280:1510–1517.
(70) Weber AM, Belinson JL, Bradley LD, Pied-monte MR. Vaginal ultrasonography versus
endometrial biopsy in women with post-menopausal bleeding. Am J Obstet Gynecol
1997; 177:924–929.
(71) Goldchmit R, Katz Z, Blickstein I, Caspi B,
Dgani R. The accuracy of endometrial
Pipelle sampling with and without sono-graphic measurement of endometrial thick-ness. Obstet Gynecol 1993; 82:727–730.
(72) Fleischer AC, Diamond MP, Cartwright PS.
Transvaginal sonography of ectopic preg-nancy. In: Fleischer AC, Manning FA, Jeanty
P, Romero R, eds. Sonography in obstetrics
and gynecology. Principles and practice. 6th
ed. New York: McGraw-Hill; 2001: 113–138.
(73) Fleischer AC, Jones HWI. Sonography of
trophoblastic diseases. In: Fleischer AC,
Manning FA, Jeanty P, Romero R, eds.
Sonography in obstetrics and gynecology,
principles and practice. 6th ed. New York:
McGraw-Hill; 2001: 843–852.
(74) Lange RC, Buberg AC, McCarthy SM. An
evaluation of MRI contrast in the uterus
using synthetic imaging. Magnet Reson Med
1991; 17:279–284.
(75) Brown HK, Stoll BS, Nicosia SV, Florica JV,
Hambley PS, Clarke LP, et al. Uterine junc-tional zone: Correlation between histologic
findings and MR imaging. Radiology 1991;
179:409–413.
(76) Scoutt LM, Flynn SD, Luthringer DJ,
McCauley TR, McCarthy SM. Junctional
zone of the uterus: correlation of MR
imaging and histologic examination of
hysterectomy specimens. Radiology 1991;
179:403–407.
(77) Schneider G. Pelvis. In: Runge VM, ed. Clin-ical MRI. Philadelphia: WB Saunders. 2002:
353–382.
(78) Ascher SM, Reinhold C. Imaging of cancer of
the endometrium. Radiol Clin North Am
2002; 40:563–576.
(79) Dueholm M, Lundorf E, Olesen F. Imaging
techniques for evaluation of the uterine
cavity and endometrium in premenopausal
patients before minimally invasive surgery.
Obstet Gynecol Surv 2002; 57:388–403.
(80) Frei KA, Kinkel K. Staging endometrial
cancer: Role of magnetic resonance imaging.
J Magnet Reson Imaging 2001; 13:850–855.
(81) Barton JW, McCarthy SM, Kohorn EI, Scoutt
LM, Lange RC. Pelvic MR imaging findings
in gestational trophoblastic disease, incom-plete abortion, and ectopic pregnancy: are
they specific? Radiology 1993; 186:163–168.
(82) Buckley CH, Fox H. Biopsy pathology of
the endometrium. 2nd ed. London: Arnold;
2002.
(83) Kepes JJ, Oswald O. Tissue artefacts caused
by sponge in embedding cassettes. Am J Surg
Pathol 1991; 15:810–812.
(84) Carson SA, Buster JE. Ectopic pregnancy.
N Engl J Med 1993; 329:1174–1181.
(85) Azumi N, Czernobilsky B. Immunohisto-chemistry. In: Kurman RJ, ed. Blaustein’s
pathology of the female genital tract. 5th ed.
New York: Springer-Verlag; 2002: 1251–1276.
284 12. Methods of Endometrial Evaluation
(86) McCluggage WG. Recent advances in im-munohistochemistry in gynaecological patho-logy. Histopathology 2002; 40:309–326.
(87) Soslow RA, Isacson C. Diagnostic immuno-histochemistry of the female genital tract. In:
Dabbs DJ, ed. Diagnostic immunohistochem-istry. New York: Churchill Livingstone; 2002:
486–516.
(88) O’Connor DM, Kurman RJ. Intermediate
trophoblast in uterine curettings in the diag-nosis of ectopic pregnancy. Obstet Gynecol
1988; 72:665–670.
(89) Kaspar HG, To T, Dinh TV. Clinical use of
immunoperoxidase markers in excluding
ectopic gestation. Obstet Gynecol 1991;
78:433–437.
(90) Daya D, Sabet L. The use of cytokeratin as a
sensitive and reliable marker for trophoblas-tic tissue. Am J Clin Pathol 1991; 95:137–141.
(91) Yeh IT, O’Connor DM, Kurman RJ. Inter-mediate trophoblast: further immunocyto-chemical characterization. Mod Pathol 1990;
3:282–287.
(92) Sorensen FB, Marcussen N, Daugaard HO,
Kristiansen JD, Moller J, Ingerslev HJ.
Immunohistological demonstration of inter-mediate trophoblast in the diagnosis of
uterine versus ectopic pregnancy—a retro-spective survey and results of a prospective
trial. Br J Obstet Gynaecol 1991; 98:463–469.
(93) Kurman RJ, Main CS, Chen HC. Inter-mediate trophoblast: A distinctive form of
trophoblast with specific morphological,
biochemical and functional features. Placenta
1984; 5:349–370.
(94) Angel E, Davis JR, Nagle RB. Immunohisto-chemical demonstration of placental hor-mones in the diagnosis of uterine versus
ectopic pregnancy. Am J Clin Pathol 1985;
84:705–709.
(95) Shih I-M, Mazur MT, Kurman RJ. Gesta-tional trophoblastic disease and related
lesions. In: Kurman RJ, ed. Blaustein’s
pathology of the female genital tract. 5th ed.
New York: Springer-Verlag; 2002: 1193–1247.
(96) Shih I, Wang T, Wu T, Kurman RJ, Gearhart
JD. Expression of Mel-CAM in implantation
site intermediate trophoblastic cell line, IST-1, limits its migration on uterine smooth
muscle cells. J Cell Sci 1998; 111 (Pt
17):2655–2664.
(97) Kurman RJ, Young RH, Norris HJ, Main CS,
Lawrence WD, Scully RE. Immunocyto-chemical localization of placental lactogen
and chorionic gonadotropin in the normal
placenta and trophoblastic tumors, with
emphasis on intermediate trophoblast and
the placental site trophoblastic tumor. Int J
Gynecol Pathol 1984; 3:101–121.
(98) Young RE, Kurman RJ, Scully RE. Proli-ferations and tumors of intermediate tro-phoblast of the placental site. Semin Diagn
Pathol 1988; 5:223–237.
(99) Shih IM, Kurman RJ. The pathology of inter-mediate trophoblastic tumors and tumor-like
lesions. Int J Gynecol Pathol 2001; 20:31–47.
(100) Shih IM, Kurman RJ. Epithelioid tro-phoblastic tumor: A neoplasm distinct from
choriocarcinoma and placental site tro-phoblastic tumor simulating carcinoma. Am
J Surg Pathol 1998; 22:1393–1403.
(100a) Shih IM, Kurman RJ. p63 expression is useful
in the distinction of epithelioid trophoblastic
and placental site trophoblastic tumors by
profiling trophobastic subpopulations. Am J
Surg Pathol (In press).
(101) Shih IM, Kurman RJ. Ki-67 labeling index in
the differential diagnosis of exaggerated pla-cental site, placental site trophoblastic tumor,
and choriocarcinoma: A double immunohis-tochemical staining technique using Ki-67
and Mel-CAM antibodies. Hum Pathol 1998;
29:27–33.
(102) Nakamura Y, Moritsuka Y, Ohta Y, Itoh S,
Haratake A, Kage M, et al. S-100 protein in
glands within decidua and cervical glands
during early pregnancy. Hum Pathol 1989;
20:1204–1209.
(103) Agarwal S, Singh UR. Immunoreactivity with
S100 protein as an indicator of pregnancy.
Indian J Med Res 1992; 96:24–26.
(104) Ronnett BM, Kurman RJ. Precursor lesions
of endometrial carcinoma. In: Kurman R, ed.
Blaustein’s pathology of the female genital
tract. 5th ed. New York: Springer-Verlag;
2002: 467–500.
(105) Lax SF, Pizer ES, Ronnett BM, Kurman RJ.
Clear cell carcinoma of the endometrium is
characterized by a distinctive profile of p53,
Ki-67, estrogen, and progesterone receptor
expression. Hum Pathol 1998; 29:551–558.
(106) Wheeler DT, Bell KA, Kurman RJ, Sherman
ME. Minimal uterine serous carcinoma:
Diagnosis and clinicopathologic correlation.
Am J Surg Pathol 2000; 24:797–806.
(107) Lax SF, Pizer ES, Ronnett BM, Kurman RJ.
Comparison of estrogen and progesterone
receptor, Ki-67, and p53 immunoreactivity in
References 285
uterine endometrioid carcinoma and
endometrioid carcinoma with squamous,
mucinous, secretory, and ciliated cell differ-entiation. Hum Pathol 1998; 29:924–931.
(108) Staebler A, Sherman ME, Zaino RJ, Ronnett
BM. Hormone receptor immunohistochem-istry and human papillomavirus in situ
hybridization are useful for distinguishing
endocervical and endometrial adenocarcino-mas. Am J Surg Pathol 2002; 26:998–1006.
(109) Kamoi S, AlJuboury MI, Akin MR, Silver-berg SG. Immunohistochemical staining in
the distinction between primary endometrial
and endocervical adenocarcinomas: Another
viewpoint. Int J Gynecol Pathol 2002; 21:
217–223.
(110) McCluggage WG, Sumathi VP, McBride HA,
Patterson A. A panel of immunohistochemi-cal stains, including carcinoembryonic
antigen, vimentin, and estrogen receptor, aids
the distinction between primary endometrial
and endocervical adenocarcinomas. Int J
Gynecol Pathol 2002; 21:11–15.
(111) Zhang X, Lin Z, Lee ES, Kim I. Immunohis-tochemical profiles of endometrial and endo-cervical adenocarcinomas. Mod. Pathol.
16[217A]. 2003 (Abstr).
(112) Ansari–Lari MA, Staebler A, Ronnett BM.
Distinction of endocervical and endometrial
adenocarcinomas: p16 expression correlated
with human papillomavirus (HPV) DNA
detection by in situ hybridization (ISH).
Mod. Pathol. 16, 180A. 2003 (Abstr).
(113) McCluggage WG, Jenkins D. p16 immunore-activity may assist in the distinction between
endometrial and endocervical adenocarci-noma. Int J Gynecol Pathol 2003; 22:231–235.
(114) Dabbs DJ, Geisinger KR, Norris HT. Inter-mediate filaments in endometrial and endo-cervical carcinomas. The diagnostic utility of
vimentin patterns. Am J Surg Pathol 1986;
10:568–576.
(115) Tamimi HK, Gown AM, Kimdeobald J, Figge
DC, Greer BE, Cain JM. The utility of
immunocytochemistry in invasive adenocar-cinoma of the cervix. Am J Obstet Gynecol
1992; 166:1655–1662.
(116) Castrillon DH, Lee KR, Nucci MR. Distinc-tion between endometrial and endocervical
adenocarcinoma: An immunohistochemical
study. Int J Gynecol Pathol 2002; 21:4–10.
(117) Cohen C, Shulman G, Budgeon LR. Endo-cervical and endometrial adenocarcinoma:
An immunoperoxidase and histochemical
study. Am J Surg Pathol 1982; 6:151–157.
(118) Maes G, Fleuren GJ, Bara J, Nap M. The
distribution of mucins, carcinoembryonic
antigen, and mucus-associated antigens in
endocervical and endometrial adenocar-cinomas. Int J Gynecol Pathol 1988; 7:112–
122.
(119) Debrito PA, Silverberg SG, Orenstein JM.
Carcinosarcoma (malignant mixed mullerian
(mesodermal) tumor) of the female genital
tract—immunohistochemical and ultrastruc-tural analysis of 28 cases. Hum Pathol 1993;
24:132–142.
(120) George E, Manivel JC, Dehner LP, Wick MR.
Malignant mixed mullerian tumors: An
immunohistochemical study of 47 cases, with
histogenetic considerations and clinical cor-relation. Hum Pathol 1991; 22:215–223.
(121) Geisinger KR, Dabbs DJ, Marshall RB.
Malignant mixed mullerian tumors. An ultra-structural and immunohistochemical analysis
with histogenetic considerations. Cancer
1987; 59:1781–1790.
(122) Auerbach HE, Livolsi VA, Merino MJ. Malig-nant mixed mullerian tumors of the uterus.
An immunohistochemical study. Int J
Gynecol Pathol 1988; 7:123–130.
(123) Meis JM, Lawrence WD. The immunohisto-chemical profile of malignant mixed muller-ian tumor. Overlap with endometrial
adenocarcinoma. Am J Clin Pathol 1990;
94:1–7.
(124) Costa MJ, Khan R, Judd R. Carcinosarcoma
(malignant mixed mullerian [mesodermal]
tumor) of the uterus and ovary. Correlation
of clinical, pathologic, and immunohisto-chemical features in 29 cases. Arch Pathol
Lab Med 1991; 115:583–590.
(125) Bitterman P, Chun B, Kurman RJ. The signif-icance of epithelial differentiation in mixed
mesodermal tumors of the uterus. A clinico-pathologic and immunohistochemical study.
Am J Surg Pathol 1990; 14:317–328.
(126) Farhood AI, Abrams J. Immunohistochem-istry of endometrial stromal sarcoma. Hum
Pathol 1991; 22:224–230.
(127) Franquemont DW, Frierson HF, Mills SE.
An immunohistochemical study of normal
endometrial stroma and endometrial stromal
neoplasms—evidence for smooth muscle
differentiation. Am J Surg Pathol 1991;
15:861–870.
(128) Lillemoe TJ, Perrone T, Norris HJ, Dehner
LP. Myogenous phenotype of epithelial–like
areas in endometrial stromal sarcomas. Arch
Pathol Lab Med 1991; 115:215–219.
286 12. Methods of Endometrial Evaluation
(129) Nucci MR, O’Connell JT, Huettner PC,
Cviko A, Sun D, Quade BJ. h-Caldesmon
expression effectively distinguishes endome-trial stromal tumors from uterine smooth
muscle tumors. Am J Surg Pathol 2001;
25:455–463.
(130) Oliva E, Young RH, Amin MB, Clement PB.
An immunohistochemical analysis of endo-metrial stromal and smooth muscle tumors of
the uterus: A study of 54 cases emphasizing
the importance of using a panel because of
overlap in immunoreactivity for individual
antibodies. Am J Surg Pathol 2002; 26:
403–412.
(131) Dabbs DJ. Carinomatous differentiation and
metastatic carcinoma of unknown primary.
In: Dabbs DJ, ed. Diagnostic immunohisto-chemistry. New York: Churchill Livingstone;
2002: 163–196.
(132) Gocke CD. Non-Hodgkin’s lymphoma. In:
Dabbs DJ, ed. Diagnostic immunohistochem-istry. New York: Churchill Livingstone; 2002:
113–134.
References 287
A
Ablation therapy, foreign body
granulomatous response
following, 155
Abnormal uterine bleeding,
100–120, 139–140, 203
in adolescence, 3
causes of, 118
clinical queries, 119
clinical terms for, 2
endometrial breakdown,
bleeding, features of, 101
estrogen-related, 109–115
atrophy, 113–115
disordered proliferative
phase, 112–113
with glandular, stromal
breakdown, 109–112
persistent proliferative
phase, 112–113
features of, 101
glandular, stromal breakdown
with artifactual crowding,
107
fibrin thrombus with, 104
hyalinized stroma secondary
to, 107
morphologic features,
100–108
glandular lumens, debris in,
108
nonmenstrual secretroy phase,
118
in perimenopausal years, 3
in postmenopausal years, 3
progesterone-related, 115–119
abnormal secretory
endometrium, 117–118
mimicking carcinoma, 214–215
Artifacts, 23–26
focal artifactual gland
crowding, 25
framentation, 24
hobnail-like, 191
late secretory endometrium,
22, 23
mimicking carcinoma, 215–216
telescoping, 24, 25
Atrophic polyp, 168, 169
Atrophy with estrogen-related
bleeding, 113–115
Atypical polypoid adenomyoma,
170–173
B
Benign changes mimicking
carcinoma, 214
Arias-Stella reaction, 214–215
atypical polypoid
adenomyoma, 214
contaminants, 215–216
epithelial cytoplasmic change,
214
necrosis, 215–216
pregnancy, 214–215
tissue artifacts, 215–216
Benign polyp, 165, 166, 167
with atypical stromal cells, 166
Biopsy, endometrial, 1–6
abortion, 34–66
bleeding, uterine,
dysfunctional, 100–120
carcinoma, 208–248
cytoplasmic change, 178–207
ectopic pregnancy, 34–66
endometritis, 147–162
irregular shedding, 117
luteal phase defects, 117
reporting of, 119
in reproductive years, 3
Abortion
ectopic pregnancy, 34–66
pathologic changes, 58–59
Actinomycosis, 155
Adenocarcinoma
clear cell, 233–235
endometrial, World Health
Organization
classification, 216
FIGO grade 1, 212, 213
grading, 216–221
mucinous, 225–227
with papillary features, 232
serous, 227–232
staging, corpus cancer, 236
with squamous differentiation,
223–225
Adenofibroma, 256–260
clinical queries, 259
differential diagnosis, 259
pathologic features, 256–259
reporting of, 259
Adenomyoma, atypical polypoid,
170–173
Adenosarcoma, 256–260
differential diagnosis, 259
pathologic features, 256–259
reporting of, 259
Adhesions, 174–175
Adolescence, abnormal uterine
bleeding in, 3
Arias-Stella reaction, 39–41
gestational endometrium with,
41
289
Index
Biopsy, endometrial (cont.):
evaluation methods, 275–288
gestational trophoblastic
disease, 67–99
hormones, effects of, 121–146
hyperplasia, 178–207
infertility evaluation, 7–33
interpretation of, 2–5
intraephithelial carcinoma,
178–207
malignant mixed mesodermal
tumor, 249–256
normal endometrium, 7–33
polyps, 163–177
pregnancy, 34–66
rare neoplasms, 267–269
smooth muscle tumors,
264–367
stromal tumors, 260–264
tumor-like conditions, 269–270
Bleeding, uterine, dysfunctional,
100–120, 139–140, 203
causes of, 118
clinical queries, 119
clinical terms for, 2
estrogen-related, 109–115
disordered proliferative
phase, 112–113
with glandular, stromal
breakdown, 109–112
persistent proliferative
phase, 112–113
features of, 101
glandular, stromal breakdown
with artifactual crowding,
107
fibrin thrombus with, 104
hyalinized stroma secondary
to, 107
morphologic features,
100–108
glandular lumens, debris in,
108
nonmenstrual secretroy phase,
118
progesterone-related,
115–119
abnormal secretory
endometrium, 117–118
irregular shedding, 117
luteal phase defects, 117
reporting, 119
Breast carcinoma, metastatic,
240
differential diagnosis,
193–195
immunostaining, 195
Cervical contaminants, 26, 27
Choriocarcinoma, 81–85
differential diagnosis, 84–85
general features, 81
hydatidiform mole,
distinguished, 78–79
pathologic features, 81–84
Chorionic villi
after first trimester, 59–61
first trimester, 57–58
Ciliated cell change, epithelial,
196–198
Clear cell carcinoma, 233–235
Clear cell change, epithelial,
201–202
Clinical history, 2–5
Clomiphene citrate, 136–137
Colon carcinoma, metastatic, 239
Combined estrogen/progestin
replacement therapy,
menopausal women,
131–132, 139
Confluent gland pattern, 210–211
Contaminants, 23–26
cervical, 26, 27
mimicking carcinoma, 215–216
Contraceptives, progestins,
122–130
morphologic features of, 123
response patterns, 123–130
stromal changes, 128–130
Cytogenetics, hydatidiform mole,
69
Cytomegalovirus, 155–156
D
Danazol, 137, 138
Dating, endometrial
early secretory endometrium,
15
interval endometrium, 14
late secretory endometrium, 17
days 23–24, 16
days 26–27, 19
predecidua, 18
mid-secretory endometrium,
days 20–21, 16
normal cycling, 10–21
pitfalls in, 21–23
secretory phase, 13
Decidua, 49, 50
290 Index
C
Carcinoma. See also under
specific type of
carcinoma
breast, metastatic, 240
colon, metastatic, 239
endometrial, 208–248
adenocarcinoma, with
squamous differentiation,
223–225
Arias-Stella reaction,
214–215
atypical polypoid
adenomyoma, 214
benign changes mimicking,
214
classification of, 216
clear cell carcinoma,
233–235
clinical queries, 239–241
confluent gland pattern,
210–211
contaminants, 215–216
desmoplastic stroma,
altered, 212
endocervical-like
endometrial carcinoma,
227
endometrial vs. endocervical
carcinoma, 237–238
epithelial cytoplasmic
change, 214
fibrous stroma, altered, 212
grading, 216–221
malignant neoplasms,
216–239
metastatic carcinoma,
238–239
mixed mesodermal tumors,
235–236
mucinous carcinoma,
225–227
necrosis, 215–216
papillary pattern, 212–214
pregnancy, 214–215
rare histologic patterns, 235
reporting of, 239–241
serous carcinoma, 227–232
staging, 236–237
well-differentiated, diagnosis
criteria, 209–214
intraepithelial carcinoma,
193–195
behavior, 195
Decidualized stroma, 38
Desmoplastic stroma, altered,
212
Dysfunctional uterine bleeding,
100–120, 139–140, 203
causes of, 118
clinical queries, 119
clinical terms for, 2
endometrial breakdown,
bleeding, features of, 101
estrogen-related, 109–115
atrophy, 113–115
disordered proliferative
phase, 112–113
with glandular, stromal
breakdown, 109–112
persistent proliferative
phase, 112–113
glandular, stromal breakdown
with artifactual crowding,
107
fibrin thrombus with, 104
morphologic features,
100–108
glandular lumens, debris in,
108
nonmenstrual secretroy phase,
118
progesterone-related, 115–119
abnormal secretory
endometrium, 117–118
irregular shedding, 117
luteal phase defects, 117
reporting of, 119
E
Early gestational endometrium,
35–37
cycle of conception, 36
decidualized stroma, 37
Early secretory endometrium, 15
Ectopic pregnancy, 34–66
endometrium associated with,
62–63
Endocervical-endometrial polyp,
mixed, 170
Endocervical-like endometrial
carcinoma, 227
Endometrial adenocarcinoma,
World Health
Organization
classification, 216
Endometrial biopsy, 1–6
abortion, 34–66
reporting of, 239–241
serous carcinoma, 227–232
staging, 236–237
well-differentiated, diagnosis
criteria, 209–214
Endometrial dating
early secretory endometrium,
15
interval endometrium, 14
late secretory endometrium,
16–19
granular lymphocytes, 18
predecidua, 18
mid-secretory endometrium,
days 20–21, 16
morphologic features, 11
normal cycling, 10–21
pitfalls in, 21–23
Endometrial glands, 35–43
Endometrial hyperplasia,
178–207
atypical, 182–188
morphologic features of,
186
behavior, 192–193
classification of, 179
differential diagnosis,
188–192
disordered proliferative
endometrium, differential
diagnosis, 189
endometritis, differential
diagnosis, 189–190
hobnail-like artifact, 191
polypoid adenomyoma,
atypical, differential
diagnosis, 190–192
polyps, differential diagnosis,
189
terminology, 179
without atypia, 180–182
complex hyperplasia,
181–182
simple hyperplasia,
180–181
World Health Organization
classification, 179
Endometrial intraepithelial
carcinoma, 193–195
behavior, 195
differential diagnosis, 193–195
immunostaining, 195
Endometrial vs. endocervical
carcinoma, 237–238
Index 291
adenosarcoma, 256–260
bleeding, uterine,
dysfunctional, 100–120
carcinoma, 208–248
cytoplasmic change, 178–207
ectopic pregnancy, 34–66
endometritis, 147–162
evaluation methods, 275–288
gestational trophoblastic
disease, 67–99
hormones, effects of, 121–146
hyperplasia, 178–207
indications for, 1–2
infertility, 3–4, 7–33
malignant mixed meodermal
tumor, 249–256
Müllerian adenofibroma,
256–260
normal endometrium, 7–33
polyps, 163–177
pregnancy, 34–66
rare neoplasms, 267–269
smooth muscle tumors,
264–367
stromal tumors, 260–264
tumor-like conditions, 269–270
Endometrial breakdown,
bleeding, features of, 101
Endometrial carcinoma, 208–248
adenocarcinoma, with
squamous differentiation,
223–225
benign changes mimicking,
214–216
classification of, 216
clear cell carcinoma, 233–235
clinical queries, 239–241
confluent gland pattern,
210–211
desmoplastic stroma, altered,
212
endocervical-like endometrial
carcinoma, 227
endometrial vs. endocervical
carcinoma, 237–238
fibrous stroma, altered, 212
grading, 216–221
malignant neoplasms, 216–239
metastatic carcinoma, 238–239
mixed mesodermal tumors,
235–236
mucinous carcinoma, 225–227
papillary pattern, 212–214
rare histologic patterns, 235
Endometritis, 147–162
ablation therapy, foreign body
granulomatous response,
155
abnormal glandular
development, 152–153
clinical queries, 160–161
differential diagnosis, 158–160
endometrial hyperplasia,
differential diagnosis,
189–190
glandular, stromal breakdown,
154
infections, 154–158
actinomycosis, 155
cytomegalovirus, 155–156
granulomatous
inflammation, 154–155
herpesvirus endometritis,
156, 157
mycoplasma, 156–158
inflammatory cells, 148–151
morphologic features, 148
nonspecific, 147–154
pseudoactinomycotic radiate
granules, 160
reporting of, 160–161
stromal changes, 151–152
Endometrium, normal, 7–33
artifacts, 23–26
framentation, 24
late secretory endometrium,
22, 23
telescoping, 24, 25
clinical queries, 30–31
contaminants, 23–26
endometrial dating
normal cycling, 10–21
pitfalls in, 21–23
secretory phase, 13
histiocytes, 28
lower uterine segment, 10
luteal phase defect, 26–30
menstrual endometrium, 19–21
in pregnancy, 37–39
proliferative phase
endometrium, 11–13
reporting of, 30–31
secretory phase endometrium,
12–19
undatable endometrium,
causes of, 29
Eosinophilic cell change,
epithelial, 198–199
G
Gestational endometrium, 38, 39
with clear cell change, 42
with optically clear nuclei, 43
Gestational trophoblastic
disease, 67–99
hydatidiform mole, 67–81
choriocarcinoma,
distinguished, 78–79
clinical queries, 80–81
complete, 70–74
complete/partial,
distinguished, 76–77
cytogenetics, 69
differential diagnosis, 76–79
exaggerated placental
implantation site, 74
gestational trophoblastic
disease, postmolar, 79–80
hydropic abortus compared,
68, 70, 77–78
invasive, 79–80
partial, 74–76
pathologic features, 69–76
persistent complete, 79
postmolar, 79–80
reporting of, 80–81
trophoblastic neoplasms, 81–94
choriocarcinoma, 81–85
clinical queries, 94
epithelioid trophoblastic
tumor, 91–94
placental site trophoblastic
tumor, 85–91
reporting of, 94
World Health Organization
classification, modified, 68
Glandular, stromal breakdown,
100, 154
morphologic features, 100–108
Glandular changes in pregnancy,
41–43
Glandular development,
abnormal, 152–153
Glandular lumens, debris in, 108
Gonadotropin-releasing
hormone agonists, 138
Grading of endometrial
carcinoma, 216–221
H
Herpesvirus endometritis, 156,
157
Histiocytes, endometrial, 28
292 Index
Epithelial cytoplasmic change,
195–204
ciliated cell change, 196–198
clear cell change, 201–202
differential diagnosis, 202–204
eosinophilic cell change,
198–199
mimicking carcinoma, 214
mucinous change, 199–201
secretory change, 201–202, 203
hobnail, 202
squamous differentiation, 196
Epithelioid trophoblastic tumor,
91–94
differential diagnosis, 93–94
pathologic features, 91–93
trophoblastic neoplasms,
differential diagnosis,
93–94
Estrogen/progestin replacement
therapy, 131–132, 139
Estrogen-related bleeding,
109–115
atrophy, 113–115
disordered proliferative phase,
112–113
with glandular, stromal
breakdown, 109–112
persistent proliferative phase,
112–113
Estrogenic hormones, 122. See
also Estrogen
Evaluation methods,
endometrium, 275–288
frozen section, 278–279
histologic techniques, 277–278
hysteroscopy, 276–277
immunohistochemistry,
279–281
magnetic resonance imaging,
277
noninvasive methods, 276–277
sampling techniques, 275–276
ultrasound, 277
F
Fibrin thrombus with glandular,
stromal breakdown, 104
Fibrinoid, 53
Foreign body granulomatous
response, with ablation
therapy, 155
Framentation artifacts, 24
Frozen section, 278–279
Histologic evaluation,
endometrium, 8–10
techniques for, 277–278
Hobnail-like artifact,
proliferative
endometrium, 191
Hobnail secretory change,
epithelial cytoplasmic, 202
Hormones, effects of, 4, 121–146.
See also under specific
hormones
clinical queries, 139–140
reporting of, 139–140
Human chorionic gonadotropin,
137–138
Human menopausal
gonadotropins, 137–138
Hyalinized stroma secondary to
glandular, stromal
breakdown, 107
Hydatidiform mole, 67–81
choriocarcinoma,
distinguished, 78–79
clinical queries, 80–81
complete, 70–74
partial, distinguished, 76–77
cytogenetics, 69
differential diagnosis, 76–79
early complete, 73
gestational trophoblastic
disease, postmolar, 79–80
invasive, 79–80
partial, 74–76
pathologic features, 69–76
hydropic abortus, compared,
70
persistent complete, 79
persistent trophoblast
following, 80
reporting of, 80–81
Hydropic abortus, hydatidiform
mole, comparison, 70,
77–78
cytogenetic, 68
pathologic features, 70
Hydropic change in abortion,
58–59
Hyperplasia, endometrial,
178–207
atypical, 182–188
morphologic features of, 186
behavior, 192–193
classification of, 179
differential diagnosis, 188–192
secretory phase endometrium,
12–19
undatable endometrium,
causes of, 29
Infertility therapy, hormones
treatment, 140. See also
Hormones
Intermediate trophoblast, keratin
immunoreactivity of, 54
Interval endometrium, 14
Intraephithelial carcinoma,
endometrial, 193–195
behavior, 195
differential diagnosis, 193–195
immunostaining, 195
Irregular shedding, 117
L
Late secretory endometrium, 17
days 23–24, 16
days 26–27, 19
predecidua, 18
Leiomyoma, 264
endometrium overlying, 265
variants of, 264–265
Leiomyosarcoma, 267
Leukemia, 267–268
Luteal phase defects, 26–30
progesterone-related bleeding,
117
Lymphoma, 267–268
M
Magnetic resonance imaging, as
evaluation method, 277
Malignant mixed mesodermal
tumor, 249–256
clinical queries, 255–256
differential diagnosis, 253–255
pathologic features, 250–253
reporting of, 255–256
Malignant neoplasms, 216–239.
See also specific neoplasm
Menopausal women, combined
estrogen/progestin
replacement therapy,
131–132, 139. See also
Estrogen; Progestin
Menstrual endometrium, 19–21
Metaplasia. See Epithelial
cytoplasmic change
Metastatic carcinoma, 238–240
breast, 240
colon, 239
Index 293
disordered proliferative
endometrium, differential
diagnosis, 189
endometritis, differential
diagnosis, 189–190
hobnail-like artifact,
proliferative
endometrium, 191
polypoid adenomyoma,
atypical, differential
diagnosis, 190–192
polyps, differential diagnosis,
189
terminology, 179
without atypia, 180–182
complex hyperplasia,
181–182
simple hyperplasia, 180–181
World Health Organization
classification, 179
Hysteroscopy, 276–277
I
Immature chorionic villi, 57
with hydropic change, 59, 60
with villous fibrosis, 60
Immunohistochemistry, 279–281
Indications for endometrial
biopsy, 1–2
Infertility evaluation, 7–33
abnormal secretory phase
patterns, 26–30
artifacts, 23–26
cervical contaminants, 26, 27
clinical queries, 30–31
endometrial biopsy, 3–4
endometrial dating
early secretory
endometrium, 15
interval endometrium, 14
late secretory endometrium,
16–19
mid-secretory endometrium,
days 20–21, 16
morphologic features, 11
normal cycling, 10–21
pitfalls in, 21–23
secretory phase, 13
histologic evaluation, 8–10
luteal phase defect, 26–30
menstrual endometrium, 19–21
proliferative phase
endometrium, 11–13
reporting of, 30–31
Mid-secretory endometrium,
days 20–21, 16
Mixed endometrial-endocervical
polyp, 170, 171
Mixed mesodermal tumors,
235–236
Mole, hydatidiform. See
Hydatidiform mole
Mucinous carcinoma, 225–227
Müllerian adenofibroma,
adenosarcoma, 256–260
Mycoplasma, 156–158
N
Noninvasive evaluation methods,
276–277
Normal endometrium, 7–33
abnormal secretory phase
patterns, 26–30
artifacts, 23–26
fragmentation, 24
telescoping, 24, 25
clinical queries, 30–31
contaminants, 23–27
endometrial dating
early secretory
endometrium, 15
interval endometrium, 14
late secretory endometrium,
16–19
mid-secretory endometrium,
days 20–21, 16
normal cycling, 10–21
pitfalls in, 21–23
histiocytes, 28
histologic evaluation, 8–10
lower uterine segment, 10
luteal phase defect, 26–30
menstrual endometrium, 19–21
proliferative phase
endometrium, 11–13
reporting of, 30–31
undatable endometrium,
causes of, 29
Nuclear grading,
adenocarcinoma, 220
nuclear grade 1, 218
nuclear grade 2, 219
nuclear grade 3, 220
O
Oral contraceptives
progestins, 122–130
morphologic features of, 123
Postmenopausal years, causes of
abnormal uterine bleeding
in, 3
Predecidua, endometrial dating,
18
Pregnancy, 34–66
Arias-Stella reaction, 39–41
chorionic villi
after first trimester, 59–61
first trimester, 57–58
clinical queries, 63–64
complications of, 35
decidua, 49, 50
decidualized stroma, 38
early gestational endometrium,
35–37
cycle of conception, 36
decidualized stroma, 37
ectopic, 34–66
endometrium associated
with, 62–63
fibrinoid, 53
first trimester placential
development, events in, 58
gestational endometrium, 38,
39
with clear cell change, 42
with optically clear nuclei,
43
glandular changes, 35–43
histologic changes of
endometrium in, 36
hydropic change, in abortion,
58–59
immature chorionic villi, 57
with hydropic change, 59, 60
with trophoblast, 45
with villous fibrosis, 60
intermediate trophoblast,
keratin immunoreactivity
of, 54
later pregnancy, endometrium
in, 37–39
mimicking carcinoma, 214–215
myometrium, 51
pathologic changes, in
abortion, 58–59
placenta accreta, 62
placental implantation site,
48–53
exaggerated, 53–54
placental polyps, 61
placental site nodule, 54–57
reporting of, 63–64
294 Index
response patterns, 123–130
stromal changes, 128–130
P
Papillary pattern, 212–214
Perimenopausal years, causes of
abnormal uterine bleeding
in, 3
Pink cell change. See
Eosinophilic cell change
Placenta accreta, 62
Placental implantation site, 48–53
exaggerated, 53–54
hydatidiform mole, 74
Placental polyps, 61
Placental site nodule, 54–57
Placental site trophoblastic
tumor, 85–91
differential diagnosis, 90–91
general features, 85
pathologic features, 85–90
Polyps, 163–177
adenomyoma, atypical
polypoid, 170–173
endometrial hyperplasia,
atypical, differential
diagnosis, 190–192
mimicking carcinoma, 214
polypoid, atypical, 170–173
adenomyamaous polyp,
170–173
adhesions, 174–175
atrophic polyp, 168, 169
benign, 165, 166, 167
with atypical stromal cells,
166
classification of, 164–173
clinical queries, 175–176
common polyps, 167–170
differential diagnosis, 173–175
endometrial hyperplasia,
differential diagnosis,
189
functional pattern, 168–170
mixed endometrial-endocervical pattern, 170,
171
proliferative/hyperplastic
pattern, 167–168
reporting of, 175–176
squamous change in, 168
Postmenopausal hormone
replacement, 131–132,
139
stroma, 35–43
trophoblastic cells, 43–62
of early pregnancy, 46
in first trimester, 44
immunohistochemistry of,
46–48
villi, 43–62
villous trophoblast, after first
trimester, 59–61
Products of conception, 4
Progesterone-related bleeding,
115–119
abnormal secretory
endometrium, 117–118
irregular shedding, 117
luteal phase defects, 117
Progestin effects, morphologic
features of, 123
Progestin-like effects with no
hormone use, 132–133
Progestins/oral contraceptives,
122–130
morphologic features of, 123
response patterns, 123–130
stromal changes, 128–130
Proliferative endometrium,
endometrial hyperplasia,
differential diagnosis,
189
Proliferative phase
endometrium, 11–13
Pseudoactinomycotic radiate
granules, 160
R
Reloxifene, 136
Rohr’s stria, 53
S
Sampling techniques, 275–276
Secretory phase endometrium,
12–19
normal, 9
Serous carcinoma, 227–232
histologic features of, 229
immunohistochemistry, 233
Shedding, irregular,
progesterone-related
bleeding, 117
Smooth muscle tumors, 264–367
clinical queries, 267
leiomyomas, 264
variants of, 264–265
leiomyosarcoma, 267
clinical queries, 94
epithelioid trophoblastic
tumor, 91–94
placental site trophoblastic
tumor, 85–91
reporting of, 94
World Health Organization
classification, modified, 68
Trophoblastic neoplasms, 81–94
choriocarcinoma, 81–85
differential diagnosis, 84–85
general features, 81
pathologic features, 81–84
clinical queries, 94
epithelioid trophoblastic
tumor, 91–94
differential diagnosis, 93–94
pathologic features, 91–93
placental site trophoblastic
tumor, 85–91
differential diagnosis, 90–91
pathologic features, 85–90
reporting of, 94
Tubal metaplasia. See Ciliated
cell change
Tumorlets, 265–267
U
Ultrasound, for evaluation, 277
Undatable endometrium, causes
of, 29
Uterine bleeding, dysfunctional,
100–120, 139–140, 203
in adolescence, 3
causes of, 118
clinical queries, 119
clinical terms for, 2
endometrial breakdown,
bleeding, features of, 101
estrogen-related, 109–115
atrophy, 113–115
disordered proliferative
phase, 112–113
with glandular, stromal
breakdown, 109–112
persistent proliferative
phase, 112–113
features of, 101
glandular, stromal breakdown
with artifactual crowding,
107
fibrin thrombus with, 104
glandular lumens, debris in,
108
Index 295
reporting of, 267
tumorlets, 265–267
Staging of endometrial
carcinoma, 236–237
Stromal tumors, 260–264
clinical queries, 264
differential diagnosis, 263–264
high-grade stromal sarcoma,
262–263
low-grade stromal sarcoma,
260–262
reporting of, 264
T
Tamoxifen, 133–136
polyp, 135
with secretory change, 136
Telescoping artifacts, 24, 25
Tissue artifacts, mimicking
carcinoma, 215–216
Trophoblastic cells, 43–62. See
also Trophoblastic disease,
gestational
of early pregnancy, 46
first trimester
morphologic features, 44
immunohistochemistry of,
46–48
Trophoblastic disease,
gestational, 67–99
hydatidiform mole, 67–81
choriocarcinoma,
distinguished, 78–79
clinical queries, 80–81
complete, 70–74
complete/partial,
distinguished, 76–77
cytogenetics, 69
differential diagnosis, 76–79
exaggerated placental
implantation site, 74
gestational trophoblastic
disease, postmolar, 79–80
hydropic abortus compared,
68, 70, 77–78
invasive, 79–80
partial, 74–76
pathologic features, 69–76
persistent complete, 79
persistent trophoblast
following, 80
reporting of, 80–81
trophoblastic neoplasms, 81–94
choriocarcinoma, 81–85
Uterine bleeding, dysfunctional,
(cont.):
nonmenstrual secretroy phase,
118
in perimenopausal years, 3
in postmenopausal years, 3
progesterone-related,
115–119
abnormal secretory
endometrium, 117–118
irregular shedding, 117
Villous trophoblast, after first
trimester, 59–61
W
World Health Organization
classification
endometrial adenocarcinoma,
216
endometrial hyperplasia, 179
gestational trophoblastic
disease, 68
296 Index
luteal phase defects, 117
reporting of, 119
in reproductive years, 3
Uterine septum, tissue from,
269
V
Villi, in pregnancy, 43–62
Villoglandular adenocarcinoma,
well-differentiated,
231

One comment on “Diagnosis of Endometrial Biopsies and Curettings

  1. […] Anastrozole With or Without Lonafarnib in Treating Postmenopausal Women With Stage IIIB, Stage IIIC, or Stage IV Breast CancerNeoadjuvant Epirubicin, Cyclophosphamide, and Paclitaxel With or Without Gemcitabine in Treating Women Who Are Undergoing Surgery for Early Breast CancerIs This What You’re Searching For Concerning Breast Cancer Staging ChartIf My 13 Year Old Know This Much About Breast Cancer Staging Chart, You Should TooStage 3 breast cancerBreast Cancer Awareness Essay? Cu? They are some s? Symptoms of the cancer?Find a doctor in your area of ​​mesotheliomaDiagnosis of Endometrial Biopsies and Curettings […]

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Follow

Get every new post delivered to your Inbox.

Join 408 other followers

%d bloggers like this: